Heme is a pro-oxidant molecule that participates in the formation of oxidative free radicals leading to oxidative damage (51). In a model of heme overload using nude mice, excessive heme in the plasma promotes the production of reactive oxygen species (ROS) and reduces the effectiveness of nitric oxide (NO), thereby affecting the expansion of blood vessels (52). HO-1 degrades heme, and therefore has anti-oxidant activity. Balla et al (51) showed that the combination of heme and ferritin, together with HO-1, prevents oxidative stress. In another study performed in cultured alveolar epithelial cells, HO-1 overexpression was found to lead to a nearly 3-fold increase in ferritin, which was accompanied by a compensatory increase in transferrin receptors, subsequently altering the distribution of iron in the cells and protecting cells from iron toxicity induced by heme degradation (53).

The anti-inflammatory activity of CO and CO-releasing molecules (CORM) has been well documented in various animal models. Increased HO-1 expression or CO prevents impairment of the blood-brain barrier (BBB), cerebral microvascular congestion and neuroinflammation (54). Mechanistically, this protection is achieved by the binding of CO to hemoglobin to prevent its oxidation and the production of free hemoglobin (54). In addition, a carrageenan-induced inflammation model showed that CO inhibits neutrophil migration and white blood cell adhesion (55). On the other hand, CORM-2 reduces the accumulation of inflammatory cells, the expression of intercellular adhesion molecule-1 (ICAM-1) and the activation of nuclear factor κB (NF-κB) in septic mice (56). In vitro experimental study confirmed that the mechanisms behind the anti-inflammatory effect of CORM-2 are related to the decreased expression of NF-κB-dependent vascular endothelial cell adhesion molecules and secretion of inducible nitric oxide synthase (iNOS) (56). Using an animal model related to human sepsis (cecal ligation and puncture in rodents), CORM-2 or CORM-3 was found to reduce the migration of polymorphonuclear (PMN) leukocytes to purulent tissues, thereby attenuating inflammation and saving animals from succumbing to sepsis (57–59). In HO-1-deficient mice, treatment with CORM-2 was found to ameliorate the severity of sepsis associated with Enterococcus faecalis infection (60). CORM-3 was found to have a bactericidal effect on Pseudomonas aeruginosa, thereby reducing the mortality of mice with bacteremia (61). Additionally, CO exerts prominent anti-inflammatory actions in part by reducing the release of tumor necrosis factor α (TNF-α), interleukin (IL)-1, macrophage inflammatory protein (MIP)-1 and IL-6 from activated macrophages (62,63).

The beneficial effects of CO on the cardiovascular system have been gradually discovered. Due to severe arterial thrombosis, HO-1-deficient mice died within four days following an aortic transplantation; however, treatment with CORM-2 was found to improve the survival rate of HO-1-deficient mice (64). The underlying protective mechanisms are related to the inhibition of platelet aggregation through the activation of guanylate cyclase (65). Endogenous CO binds to and activates sGC, subsequently increasing the concentration of cGMP and inducing a vasodilatory response (50,66). In a mouse model of myocardial infarction caused by coronary artery ligation, intravenous infusion of CORM-3 prior to reperfusion was found to reduce infarct size, fibrillation and tachycardia (67,68). The mechanisms underlying CORM-3-mediated cardiovascular protection involve potassium channels, since a small amount of CORM-3 is lost when mitochondrial ATP-dependent potassium channel inhibitors are used (67,68). Another study also demonstrated that HO-1 promotes angiogenesis through CO (69), and improves the survival of tissues surrounding the blood supply by inhibiting apoptosis of endothelial cells (70).

Many scholars have reported that CO plays an important role in organ transplantation and preservation. CO in a gaseous form or as a CORM can be used as a preservation adjuvant in organ preservation solution. Using a model of allogeneic heart transplantation in rodents, continuous inhalation of CO or CORM can prolong the survival time of the transplanted heart (71,72). The cardioprotective effect of CO gas is achieved by inhibiting platelet aggregation and endothelial cell apoptosis, as well as producing endogenous vasoconstrictors, thereby improving the microcirculation of the graft (71,72). Compared with the untreated control group, rabbit kidneys washed with CORM-3 or CORM-1 and then stored in a cold solution for 24 h showed higher rates of perfusion flow, glomerular filtration, and sodium and glucose reabsorption (73). The above-mentioned mechanisms of renal protection are thought to be related to CO-medicated expansion of blood vessels, inhibition of cell apoptosis and promotion of angiogenesis (73).

As previously mentioned, HO-1 generates biliverdin, which is then converted to bilirubin. Both bilirubin and biliverdin are effective oxidative inhibitors (74). Human biliverdin reductase (hBVR), a Ser/Thr/Tyr kinase, is activated by insulin and free radicals (75). hBVR is central to the activation as well as the nuclear import and export of extracellular signal-regulated kinase (ERK)1/2 (76). Insulin increases BVR tyrosine phosphorylation and increases glucose uptake when BVR is knocked down by small interfering RNA, suggesting an important function of BVR in insulin signaling (77). HO-1-derived bilirubin is an efficient scavenger of reactive oxygen and nitrogen species (RONS) (78). In macrophages, biliverdin induces the phosphorylation of endothelial NO synthase (eNOS) at Ser1177, which generates eNOS. It is well recognized that NO plays a crucial role in regulating a wide spectrum of vascular functions, including vasorelaxation, inhibition of leukocyte-endothelial adhesion, vascular smooth muscle cell migration and proliferation, and platelet aggregation (79).

In summary, each product generated by HO-1 plays an important role in protecting cells and tissues through a variety of mechanisms.

The survival rate of skin flaps is related to ischemia-reperfusion injury and surgical trauma. Ischemia-reperfusion injury, if not treated in a timely and appropriate manner, may cause unfavorable outcomes for patients. Both surgical trauma and ischemia-reperfusion may initiate inflammatory responses, causing white blood cells to roll and adhere to the capillary vein (80–82). In the early reperfusion phase, the accumulation of activated leukocytes and reactive oxygen metabolites aggravates the inflammatory response and ultimately impairs endothelial integrity (83). The initiation of the stress response leads to increased HO-1, which significantly reduces the adhesion of white blood cells to the endothelial surface and lessens the impairment of the integrity of venous endothelium (84). Rücker et al (84) studied whether stress conditioning-induced HO-1 could prevent an inflammatory response in transferred osteomyocutaneous flaps. In all tissues analyzed, control flaps presented with significant leukocyte adherence in postcapillary venules, increased expression of intercellular adhesion molecule-1 (ICAM-1), and disruption of endothelial integrity. In contrast, stress conditioning induced considerable HO-1 expression, which coincided with a significant reduction of leukocyte adherence, ICAM-1 expression, and endothelial hyperpermeability. Correspondingly, the inhibition of HO-1 by tin protoporphyrin IX completely abolished the stress conditioning-induced amelioration of the inflammatory response (84). Hence, the protective effect elicited by stress conditioning is mainly mediated by the induction of HO-1, which reduces oxygen free radicals, ICAM-1, adherence of white blood cells to the endothelial surface, endothelial permeability and macromolecular extravasation, ultimately reducing the overall inflammatory response.

Rats undergoing flap reconstruction surgery experience lipid peroxidation and vascular damage, which is related to ischemia-reperfusion injury (85). Ischemia-reperfusion injury involves a complex oxidation process, which is closely related to flap survival (86). Oxidative stress induces an excessive activation of inflammatory processes, which increases ROS. ROS are molecules known to cause tissue damage through multiple mechanisms, including altering the structure and chemistry of proteins, lipids, and nucleic acids. Hence, reducing the release of ROS or removing excessive ROS protects tissues from ischemia-reperfusion injury (87). Lin et al (88) reported that CORM-2 induces HO-1 expression, thereby reducing protein kinase C (PKC)/amino acid-rich tyrosine kinase 2 (Pyk2)-dependent production of ROS. Lin et al (28) showed that ginkgolide B reduced skin flap necrosis and improved the survival of island perforator flaps by inhibiting endoplasmic reticulum stress and oxidative stress. In this study, they further showed that ginkgolide B activated nuclear factor erythroid 2-related factor 2 (Nrf2) signaling. Following its activation, Nrf2 is transported to the nucleus, where it binds to antioxidant response elements located in the cis-regulatory sequences of antioxidant-related enzymes and proteins and increases their expression (89).

Brouard et al (90) studied the anti-apoptotic potential of HO-1 in cultured endothelial cells and found that CO, the product of heme digested by HO-1, activates the p38 mitogen-activated protein kinase (MAPK) pathway to inhibit endothelial cell apoptosis. But the HO-1 agonist heme arginate (HA) failed to improve the survival rate of ischemic flaps (91). In vitro studies have shown that HA is cytotoxic to keratinocytes (91). Brouard et al (90) also found that CO causes vasodilation, which is part of the mechanism by which HO-1 protects skin flaps from chronic ischemia-induced injury (90). Chronic ischemia was found to increase the expression of HO-1, especially on the first and third days post-operation. Increased HO-1 expression in turn induces arteriole expansion and promotes perfusion, which can maintain sufficient capillary perfusion density in the flap and reduce flap necrosis (92). The inhibition of endogenous HO-1 by tin protoporphyrin IX was found to completely eliminate arteriole expansion and hyperperfusion, resulting in a significant decrease in functional capillary density, a significant increase in cell apoptosis, and skin flap necrosis, but aging can reverse this protective effect (92–94). Consistent with the above findings, Kubulus et al (95) reported that delay-associated tissue protection can be obtained by HO-1-mediated attenuation of microcirculatory dysfunction. Increased HO-1 activity protects tissues by biliverdin-associated anti-oxidative actions and/or CO-mediated vasodilation. This is supported by findings that show that trolox treatment after selective blockade of HO-1 reduced necrosis only until day three after flap creation and did not affect the manifestation of microcirculatory disorders. These observations suggested that: i) oxidative stress contributes to the initial development of flap necrosis; ii) anti-oxidative treatment counteracts oxidative stress-induced cell damage; iii) the initial oxidative stress does not define the degree or severity of final flap necrosis; and iv) HO-1 prevents flap necrosis induced by chronic ischemic conditioning (delay) primarily through its vasodilatory action and by improving microvascular perfusion (95). Kubulus et al (23) further showed that cooling induced a marked expression of HO-1 without induction of the Hsp70 protein, which was accompanied by significant improvement in microvascular perfusion. Thus, HO-1 also plays a role in the cooling-mediated amelioration of microcirculation, which results in a significant reduction in final flap necrosis. In another independent study, Sun et al (96) showed that preconditioning with isoflurane promotes hypoxic stabilization and induces the accumulation of hypoxia-inducible factor 1α (HIF-1 α), while HIF-1 αmediates transcriptional activation of the HO-1 gene in response to hypoxia to induce angiogenesis.

However, research has also shown that HO-1 may have dual functions. For example, HO-1 has a dual role in cancer cells. The levels of cellular iron and ROS are the determining factors for the role of HO-1, in which excessive cellular iron and ROS tend to push HO-1 from a protective role to that of a perpetrator. In general, a moderate level of HO-1 activation exerts a cytoprotective effect, while the over-activation of HO-1 becomes cytotoxic due to the excessive increase of labile Fe²⁺ behind the buffering capacity of ferritin (97). Similarly, HO-1 also has a dual role in nerve cells. Nrf2-dependent activation of HO-1 is generally linked to protective effects in neurons and glial cells, while Nrf2-independent activation of HO-1, which often involves AP-1 or NF-κB, seems to exert neurotoxic effects. Indeed, HO-1 expression is associated with neuronal damage and neurodegeneration, especially in Alzheimer's and Parkinson's disease (98).

The application of CO, the catalytic product of HO-1, within a clinical setting also has the following issues. For example, the safety constraints of CO require clinical dosing to maintain levels of carboxyhemoglobin (CO-Hb) that are under 14%, which is in stark contrast to the preclinical frontrunner studies with protocols that had sustained CO-Hb levels above 20% (99). Also, the pharmacokinetics of CO inhalation protocols varied significantly by species. Rodent models achieve therapeutic levels quickly, whereas the time to steady state in humans is roughly three times longer based on respiratory rate and cardiac output (100).

Similarly, the application of HO-1 or its metabolites to the field of flap surgery will also face the above-mentioned issues. Instead, perhaps it would be possible to: i) use HO-1 inhibitors (such as tin protoporphyrin) to repress its overexpression; ii) use an occasional monitoring system to achieve feedback to control the level of CO in the body; or iii) search for molecules that promote the release of CO or solutions that can dissolve CO, and apply them locally to reduce systemic adverse reactions.