In the late 1990s, research and development of hemoglobin (Hb)-based oxygen carriers (HBOCs), or “blood substitutes,” reached the highest levels of activity with several products reaching phase 3 pivotal clinical trials. Unfortunately, clinical evaluation of HBOCs in humans has been disappointing, with toxicity leading to the discontinuation of their development. The most reported adverse events in human trials included hypertension, cerebrovascular events, renal dysfunction, pancreatic and liver enzyme elevation, and myocardial events (Natanson et al., 2008). The reaction of cell-free Hb with nitric oxide generated by the vascular endothelial lining of blood vessels leading to vasoconstriction and elevation in blood pressure is the most likely mechanism underlying these effects (Gould et al., 1998; Olson et al., 2004). The second proposed mechanism is the rapid unload of oxygen by cell-free Hb, leading to vasoconstriction as an autoregulatory response (Winslow, 2013). In recent years experimental evidence emerged to support a third mechanism which focuses on heme-mediated side reactions that ultimately lead to tissue toxicity (Alayash, 2004; Alayash, 1999).

Heme-mediated redox reactions of Hb occur in red blood cells (RBCs), despite the presence of several reducing and antioxidant mechanisms within these cells. These reactions are accelerated in disease states and during room storage temperature of blood in blood banks. In both sickle cell and aged RBCs these reactions yield highly reactive species of Hb which can target band 3 proteins and lead to structural changes in RBCs (Strader et al., 2020; Kassa et al., 2024).

Outside the RBCs, free Hb undergoes rapid conversion to highly oxidized species at an accelerated pace, leading to tissue damage if uncontrolled. The heme prosthetic group in the ferrous (HbFe²⁺) state reversibly binds oxygen, leading to oxidation of the heme iron to the ferric form (HbFe³⁺) in a process known as autoxidation (Alayash and Wilson, 2022). The slow autoxidation of the oxygen-carrying HbFe²⁺→HbFe³⁺, a non-oxygen carrying ferric (metHb), is a single electron process which leads to the production of superoxide ions (O2^•−) that rapidly dismutate to form hydrogen peroxide (H₂O₂). The resulting H₂O₂ triggers a cascade of oxidative reactions with either ferrous or ferric Hb. These reactions lead to the generation of a potent oxidizing ferryl Hb (HbFe⁴⁺) species, as well as a secondary radical from the reaction between H₂O₂ and metHb (Kassa et al., 2024). However, unlike classical peroxidases, Hb is unable to harness this radical (^•HbFe⁴⁺), which then migrates to a group of amino acids in an area on the protein known as the oxidation “hotspot”, primarily targeting βCys93 site (Alayash and Wilson, 2022).

Chemical and genetic alterations of Hb to generate cell-free HBOCs have been found in some cases to lead to accelerated oxidative reactions and formation of a persistent ferryl species (Silaghi-Dumitrescu et al., 2007). Ferryl Hb once formed is known to be a highly reactive molecule that impacts cellular and subcellular components including the mitochondria. Mitochondria are central to cellular bioenergetics generating most ATPs as the “powerhouse,” through a tightly controlled oxidative phosphorylation (OXPHOS) and play a critical balancing act in maintaining oxygen homeostasis (Jana and Alayash, 2025). Mild to moderate alteration in cellular bioenergetics is seen in many chronic diseases and therefore represents a potential therapeutic strategy. Mitochondrial bioenergetic analyses have not been explored extensively in the context of HBOC cytotoxicity. Recently however, our group and others have shown that heme and cell-free Hb-mediated disruption of mitochondrial bioenergetic function in different cell types (Cardenes et al., 2014; Higdon et al., 2012; Chintagari et al., 2016; Kassa et al., 2015). We found that exposure of E10 cells to ferryl Hb resulted in a drop in hyperpolarized cell percentage over untreated control, thus indicating a significant mitochondrial depolarization or compromised mitochondrial respiration. We also found heme-mediated loss of mitochondrial trans-membrane potential (Δψm) following exposure of E10 cells to hemin (10 µM) for 24 h (Chintagari et al., 2016).

In an isolated rat heart Langendorff perfusion system we found that perfusion with a ferric form of crosslinked Hb caused significant impairment of mitochondrial cytochrome c oxidase. We also observed many fused mitochondria that maintained their round shape following hypoxic perfusion, but with partial loss of cristae membrane was observed following perfusion with ferric Hb (Edmondson et al., 2019).

Here, we incubated 4 different chemically modified HBOCs (human and bovine) in different redox (ferrous, ferric and ferryl) states with human-derived cardiomyocytes (AC16), and we monitored mitochondrial respiration of cells and the enzymatic activities of their respective complexes. We noted very little differences in the ETC activities among the ferrous forms of HBOCs vs. their native unmodified Hbs. However, ferric, and most notably ferryl forms of HBOCs were more damaging and more greatly initiated cellular and subcellular changes.

Because of the confidentiality in disclosing the nature of HBOC toxicity by manufacturers and regulatory authorities, very limited information on their side effects during human clinical trials were made in the open literature. Recently we were able to compare side by side biochemical and biophysical characteristics of all HBOCs that undergone clinical evaluation in variety of clinical settings after these HBOCs were made available to our group (Meng et al., 2018). In a follow up publication by Jahr et al. (2025), a correlation between these unique features reported by Meng et al., and their clinical outcome were carried out recently to specifically draw some general lessons that will hopefully enable the design of a safe and effective product. Based on this publication, clinical trials, and clinical outcomes were added to the biophysical and biochemical characteristics of the various products that enabled the formation of more definitive conclusions (Jahr et al., 2025).

A recent analysis of the incidence of myocardial infarctions (MI) in patients infused with 5 clinically tested HBOCs (4 out the 5 HBOCs were used in the current study) that focused on pharmacokinetic and biochemical parameters to identify correlations that are suggestive of cause-and-effect. There were positive correlations between MI incidence and HBOC dose, size, intravascular half-life, and area under the plasma concentration (total exposure of blood and endothelium to HBOCs versus time curve (AUC). Furthermore, MI incidence was positively correlated with initial rates of HBOC autoxidation, oxidation by nitric oxide, and AUCs estimated for these HBOC oxidation products. These observations imply that increased MI risk after HBOC infusion is due to intravascular reactions which exacerbate oxidative stress (Estep, 2025). The correlation between HBOC infusion, oxidative events and enhanced MI incidence may be the most rigorous proof to date of the hypothesis that oxidative stress increases MI risk (Estep, 2025).

In this study we examined the impact of different HBOC oxidative states (oxy/ferrous, ferric and ferryl) on cultured human cardiac cells. Specifically, we focused on the role of these species in cellular injury and mitochondrial respiration, specifically the correlation between HBOCs own oxidative reactions and consequent impact on cellular and subcellular components of cardiac myocytes. To understand potential mechanisms of toxicity of the HBOCs on human cardiac cells, we used AC16 cells which are possibly the only commercially available cardiac cell line of human origin derived from primary human ventricular tissue and can be induced to differentiate into cardiomyocytes. As shown in Table 1, autoxidation rates, ferryl levels as well as the rates of heme release from the ferric forms of these HBOCs, it is quite possible that increases in the ferrylHb activities play a key role in changes seen at the cellular and subcellular (ETC.) levels as described here. Generally, higher rates of autoxidation in human and bovine Hbs were matched by higher rates of oxidative modifications and ferrylHb formation. As reported in this table, chemically modified human and bovine Hbs autoxidize at rates 1.7–3.0-fold higher than their native unmodified Hbs, likely due to the sit-specific modification(s) on these proteins. Ferryl heme activities were also higher in the modified human and bovine Hbs, some 1.4 to 1.87-fold higher than their respective unmodified Hbs. It is interesting to note that the reduction in heme released from the oxidized forms of these HBOCs (due likely) to their specific chemistry of each HBOC was reduced by almost 1.4-fold.

We have previously shown that ferric Hb loses heme at rates substantially higher than ferryl Hb. This was also supported by a higher expression of heme oxygenase-1 (HO-1) when ferric Hb was added to cultured lung alveolar squamous epithelial cells (E10) (Kassa et al., 2016). However, ferryl Hb once formed can be more damaging to mitochondrial respiration in cardiac muscle cells than ferrous or ferric as evidenced by the data presented in this study. Although, relative contributions of different oxidative mechanisms in a specific tissue can be greatly influenced by the physiological status and function of the tissue (e.g., brain, heart, lungs etc.) (Granger and Kvietys, 2015), however, in a previous study using lung epithelial cells we observed a similar phenomenon where ferryl Hb exposure caused a robust translocation of HO-1 to mitochondria with a stronger impairment of mitochondrial respiration and loss of mitochondrial transmembrane potential compared to ferric Hb (Chintagari et al., 2016).

Lactate dehydrogenase (LDH), is an enzyme found in almost all tissues involves in cell energy production which can also be used to identify location and severity of tissue damage was used in our experiments as a reliable marker for HBOC’s mediated cytotoxicity. As shown in Figure 1 when myocytes are incubated with the ferrous form of HBOCs, there was little or no change in LDH levels with human derived HBOCs as opposed to slight or moderate increase in LDH levels released by myocytes in the presence of bovine HBOCs. The incubation of myocytes with the ferric forms of human and bovine derived HBOCs led to almost double the LDH release, particularly more so with bovine HBOCs. Using the ferryl species of these HBOCs we noticed substantial increase in cell injury and LDH release to as high as almost 600% (6 times) over untreated cells in the case of bovine derived HBOCs. Interesting to note that both Oxyglobin and Hemopure autoxidize at higher rates than their native Hb and human derived HBOCs and these higher rates and ferryl formation correlated with higher levels of LDH injury in myocytes.

The enhanced oxidation of HBOCs in plasma to methemoglobin has been previously noted in clinical trials (Sprung et al., 2002; Olofsson et al., 2006) despite human capacity to reduce plasma methemoglobin (Vandegriff et al., 1997). However, this reductive capacity may, however, be compromised or overwhelmed in patients with pre-existing endothelial dysfunction at higher HBOC doses (Biro, 2012). Spontaneous oxidation of heme iron within HBOCs in the presence of air (autoxidation) or in the presence of oxidants (oxidation) compromise the ability of HBOCs to deliver oxygen to tissues. In animal models (Lee et al., 1995) and in some clinical trials (Fitzgerald et al., 2011) metHb accumulation up to 30%–40% of total Hb were reported to comprise tissue oxygenation.

It has been known for some time that oxidative stress is correlated with increased MI risk (Wang and Kang, 2020). This is not surprising since the incidence of MI is highly correlated with the initial rate of HBOC autoxidation. The correlation between HBOC infusion, oxidative events and enhanced MI incidence is consistent with the notion that oxidative stress increases MI risk and HBOCs play a critical role in these events (Estep, 2025).

Mitochondria are known to play a critical role in cardiac energy output and are therefore critically important in energy-demanding cardiac muscle (Dai and Rabinovitch, 2009). The oxidation of several biological fuels e.g., NADH, pyruvate or succinate, is accomplished via the mitochondrial ETC that provides the energy required (Lesnefsky et al., 2001). Here, we investigated the biological significance of altered bioenergetics as a possible mechanism of cardiotoxicity in HBOCs using human-derived cardiomyocytes. None of the ferrous Hbs and HBOCs had noticeable effect on any of the respiratory chain complexes of the mitochondrial ETC. However, we found that the ferric forms of the Hbs and HBOCs caused mild inhibition on each ETC complex.

In a previous study, we evaluated myocardial performance and biochemical alterations in rat left ventricle (LV) in response to perfusion with a specific HBOC (DCLHb) following a brief hypoxic episode (Edmondson et al., 2019). LV performance recovery was enhanced following ferrous DCLHb infusion, indicating a cardioprotective effects possibly due to its ability to adequately oxygenate tissues during perfusion. In contrast, ferric DCLHb infusion caused heart lesion formation, which was partially reversed by ascorbate (a mild reducing agent) co-infusion, emphasizing the oxidative toxicity associated with ferric ions. This study showed that heme rapidly released by ferric Hb can also lead to oxidative radical species production and iron accumulation which was directly associated with diminished LV performance with an extensive damage in myocardial mitochondrial ultrastructure (Edmondson et al., 2019).

In the current study, we also observed that mitochondrial respiration was impacted by a mechanism involving impairment of mitochondrial ETC function especially the cytochrome c oxidase (complex IV) activity in agreement with our in vivo studies (Edmondson et al., 2019). It is generally accepted that cardiac contractility is strongly dependent on mitochondria, and that depletion of ATP is the main factor involved in myocardial dysfunction (Wal et al., 2023; Neubauer, 2007; Gupta et al., 2012). We have seen earlier that ferric Hbs can in fact facilitate uncoupling leading to upregulation of oxygen consumption in various cultured mammalian cells (Chintagari et al., 2016; Jana et al., 2017). A recent investigation showed that ferric Hb is the primary driver of cell-free Hb-induced lung microvascular mitochondrial dysfunction. Activation of the mitochondrial permeability transition (mPTP) and the release of mtDNA are a feature of ferric cell-free Hb mediated injury (Riedmann et al., 2025). Heme released from oxidatively unstable Hb can also cause activation of TLR4. Activated TLR4 and carbon monoxide (CO) are both known to cause the uncoupling of mitochondrial respiration (Kaczara et al., 2015; Kaczara et al., 2016). In contrast, ferryl forms of both human and bovine Hbs and their HBOC analogs were significantly damaging to all mitochondrial ETC complexes, further indicating the high reactivity and toxicity of the ferryl ions. Apart from the direct oxidative toxicity of the ferryl ions, many other secondary processes like heme and iron accumulation, oxidation of lipids and proteins can all contribute to the inhibition of bioenergetic functions and a general cellular toxicity (Edmondson et al., 2019).

The molecular basis of toxicity mediated by the ferric/ferryl hemes towards cellular systems and especially their effect on the bioenergetics of the mitochondria is not fully understood. We have previously shown the impact of ferric/ferryl forms on heart tissues (Edmondson et al., 2019) and in cultured endothelial/epithelial cells through possibly an electron transfer pathway from these redox active molecules in target cells (Kassa et al., 2015). We have also shown in our earlier study that ferric Hb shows high degree of heme release than ferryl Hb with concomitant HO-1 expression which can possibly be a mechanism behind ferric Hb/HBOCs mediated complex IV inactivation seen in this study (Kassa et al., 2016). We and other studies have shown HO-1 translocation to mitochondria under different stress conditions, indicating an interesting link between HO-1 overexpression and mitochondrial dysfunction (Bindu et al., 2015; Slebos et al., 2007; Kassa et al., 2015). Mitochondrial accumulation of HO-1 can lead to localized generation of toxic levels of CO upon degradation of the heme, thus leading to complex IV inhibition (Song et al., 2006). It is also possible that HO-1 translocation to mitochondria can be a compensatory protective effect of the HO-1 protein by scavenging extra heme load (Bindu et al., 2015; Slebos et al., 2007).

Several antioxidative strategies have been designed recently to target Hb-mediated oxidative pathways to control these reactions, specifically by the ferryl heme. These included the use of ferryl specific antioxidants, such as caffeic acid (Kassa et al., 2021), curcumin (Hicks et al., 2024), (−)-epigallocatechin gallate (EGCG) (Jia and Alayash, 2008), in addition to re-engineering of more oxidatively stable Hbs.

Apart from the direct oxidative toxicity of the ferryl ions, many other secondary processes like heme and iron accumulation, oxidation of lipids and proteins can all contribute to the inhibition of bioenergetic functions and a general cellular toxicity (Edmondson et al., 2019).

In summary, we showed in a simple cell culture system that a correlation exists between HBOC’s redox states and in particular higher oxidation ferryl heme and cellular and subcellular changes including changes in the mitochondria which may explain the reported cellular changes leading to cardiac lesion. It remains unclear however, the full extent of the molecular basis of HBOC-mediated mitochondrial dysfunction in the hearts. This remains as one limitation of this study. It follows therefore that any protective strategie(s) should include some design elements that minimize Hb’s oxidative reactions to control/reduce the toxicity of HBOC formulations (Bian and Chang, 2015). Finally, identifying how HBOCs may impact patient redox status and endothelial function would improve patient selection (Estep, 2025). In sum, understanding the mechanisms of HBOC toxicity will allow for the design of safer blood substitutes and help identify patient populations that are most likely to benefit from HBOC administration.