Since the following sections focus primarily on direct gas-sensing mechanisms, I define gas sensor proteins as proteins that require gases (or their solutes) as substrates for enzyme activity. Gasoreceptors are proteins in which either the binding or lack of binding of a gaseous molecule (or its solute) as a ligand can trigger a cellular signal or response (Anbalagan, 2024a). Gasoreceptors can bind a gaseous molecule either via a metal cofactor or in a metal cofactor-independent manner. Gasoreceptors can be any gas binding-based allosterically-regulated protein with various functions, including enzyme, transcription factor, and ion channel activities (Aono, 2017; Anbalagan, 2024b). For instance, mammalian soluble guanylate cyclase (sGC) and the Escherichia coli Direct Oxygen Sensor (DosP) phosphodiesterase are NO and O₂ gasoreceptors, respectively. While sGC is well known for its role in vasodilation in mammals, DosP regulates E. coli biofilm formation and motility (Ignarro and Freeman, 2017; Delgado-Nixon et al., 2000; Shimizu, 2013). Based on the classification of signal transduction systems (STSs) in prokaryotes, the sGC and DosP can potentially be described as a gasoreceptor that function in one-component STSs (Ulrich et al., 2005; Wuichet et al., 2010). However, the classification of sGC is controversial because the cGMP-generating catalytic site forms at the interface of heterodimer. Therefore, a more appropriate classification for sGC could may be as a co-component STS. Finally, diverse cellular signaling events mediated by gases acting as ligands for gasoreceptors are unified under the umbrella term “gasocrine signaling” (Anbalagan, 2024a).

To the best of my knowledge, the majority of the O₂-sensing mechanisms in both plants and erythrocyte-containing vertebrates are largely based on O₂ sensors (Hammarlund et al., 2020). These sensors can belong to major enzyme classes, such as dioxygenases, monooxygenases, and oxidases, all of which require O₂ as a substrate. However, numerous O₂-dependent enzymes (more than 200 enzymes in humans alone) can potentially be considered as O₂ sensors. Depending on their cellular localization, these proteins could potentially perform O₂-sensing roles in their respective cellular regions or organelles (Li et al., 2023). Amongst these enzymes, Prolyl hydroxylase domain proteins (PHDs) are well-known O₂ sensor proteins in humans and mice. These enzymes require O₂ for hydroxylase activity and regulate the Hypoxia-inducible factor 1-alpha/Prolyl Hydroxylase Domain Protein 2/von Hippel-Lindau (HIF1α/PHD2/VHL) pathway (Hammarlund et al., 2020; Li et al., 2023; Ratcliffe and Keeley, 2025). In plants, cysteine oxidases along with the N-degron pathway, and reactive oxygen species-mediated signaling are considered major mechanisms by which plants sense O₂ (Hammarlund et al., 2020; Holdsworth and Gibbs, 2020).

However, the precise identity of vertebrate and plant O₂ gasoreceptors similar to the mammalian NO gasoreceptor sGC or the E. coli O₂ gasoreceptor phosphodiesterase (DosP) remains unclear. Even in chemoreceptors cells such as the glomus cells of the carotid body and pulmonary artery smooth muscle cells, the identity of O₂ gasoreceptors are unknown (Weissmann et al., 2006; Moreno-Domínguez et al., 2020; Gao et al., 2022). In the carotid body, cytochrome c oxidase (complex IV) of the mitochondrial electron transport chain has been reported as an O₂ sensor (Moreno-Domínguez et al., 2020; Peng et al., 2010; Kumar and Prabhakar, 2012; Peng et al., 2020. Nevertheless, it is unclear whether these sensors bind O₂ and function as O₂ gasoreceptors or respond to other accumulating molecules, such as reactive oxygen species, H₂S, NO, or phosphatidic acid (Gao et al., 2022; Peng et al., 2023; Aragonés et al., 2001). In considering O₂ per se as a gaseous signaling molecule, I proposed it as an essential gasotransmitter, similar to “essential” amino acids (Anbalagan, 2024c).

In the main olfactory epithelium sensory neurons (type B cells) of mice, it has been reported that Soluble guanylate cyclase 1 subunit beta 2 (GUCY1B2) and Transient receptor potential cation channel subfamily C member 2 (TRPC2) mediate calcium influx responses under low O₂ conditions. However, it is unclear whether these proteins act as O₂ gasoreceptors (Bleymehl et al., 2016). Similar calcium influx-related responses to acute hypoxia have also been reported in astrocytes, which appear to function independently of peripheral chemoreceptor O₂-sensing mechanisms. Specifically, in astrocytes isolated from the mice parafacial respiratory group and the retrotrapezoid nucleus, calcium influx is regulated by the differential accumulation of Transient receptor potential cation channel, subfamily A, member 1 (TRPA1) in the plasma membrane under hypoxic conditions (Uchiyama et al., 2020). PHD2 and Neural precursor cell-expressed developmentally downregulated gene 4-1 (NEDD4-1) E3 ubiquitin ligase-dependent trafficking of TRPA1 channels are implicated in this process; however, it is unclear whether these proteins perform the role as O₂ gasoreceptors. Finally, the “hypoxia sensor” in astrocytes is considered to reside inside the mitochondria, thus suggesting a similar mechanism to that of glomus cells, in which mitochondrial cytochrome c oxidase has been reported as an O₂ sensor protein, but not an O₂ gasoreceptor (Moreno-Domínguez et al., 2020; Gao et al., 2022; Angelova et al., 2015). Overall, the identity of O₂ gasoreceptors in vertebrates and plants remains unknown.

Androglobin (ADGB) which is known to expressed in mammalian testis, female reproductive tract, lungs and brain is a promising candidate for an O₂ gasoreceptor in one-component STS (Koay et al., 2021; Hoogewijs et al., 2012). The circular permuted globin domain of ADGB can bind O₂ and NO and based on its binding kinetics, ADGB has been proposed to have an O₂ sensing rather than transport role (Hoogewijs et al., 2012; Reeder et al., 2024). ADGB is necessary for sperm development in mice and ciliogenesis in cell cultures (Koay et al., 2021; Keppner et al., 2022). Mice ADGB^-/- mutants exhibit defective maturation of elongating spermatids, abnormal sperm shape, defects in sperm microtubule and mitochondrial organization (Keppner et al., 2022). However, it is unclear whether ADGB’s protease activity (either autolytic or on its targets) is regulated by its O₂-binding status. At least for one of its binding partner and protease activity target, Septin 10 (SEPT10), ADGB’s protease activity is unaffected by chronic hypoxia (24 h at 0.2% O₂) in cell culture-based experiments. Finally, ADGB is not the only O₂ binding protein that can be considered as a candidate O₂ gasoreceptor.

The role of archaeal, bacterial, and yeast protoglobins and flavohemoglobins was proposed to be that of enzymes such as nitrite reductase, NO dioxygenase, denitrosylase, and alkylhydroperoxide reductase, rather than that of gas transport (Zhu and Riggs, 1992; Durner et al., 1999; Bonamore et al., 2003; Corker and Poole, 2003; Freitas et al., 2004; Ascenzi et al., 2014). Similar enzymatic activity has also been reported in plant phytoglobins (Mukhi et al., 2017; Villar et al., 2020). Oxygenated hemoglobin is a NO dioxygenase, while deoxygenated hemoglobin (with 40%–60% oxyHb saturation) and deoxygenated form of its paralogs and orthologs also exhibit nitrite reductase activity. They use co-substrates, such as nitrite and protons, to generate NO, a gaseous signaling molecule and gasotransmitter with a well-known role in vasodilation, among other functions (Huang et al., 2005; Angelo et al., 2008). Given its use of nitrite and protons as a substrate, hemoglobin can be considered one of the nitrite and proton sensor proteins. However, O₂ and CO are competitive inhibitors of deoxygenated hemoglobin’s nitrite reductase activity and resulting signaling response, thus warranting reconsideration of hemoglobin’s role as a gas-transporting protein.

Despite its nitrite reductase-dependent signaling role in vivo, hemoglobin and its paralogs are only referred to as a transporter or metabolic sensor, not an O₂ gasoreceptor (Adepu et al., 2024). The well-known role of hemoglobin in transporting O₂ and its low dissociation constant (K_D) for O₂ preclude its classification as an O₂ gasoreceptor (Hammarlund et al., 2020). However, dissociation constants must be considered not from the perspective of in vitro studies but rather from the point of physiological O₂ concentration in vivo, where hemoglobin can be present even in a transient manner. Moreover, reversible binding is not unique to gas-transporting proteins; it is also a feature of previously well-accepted gasoreceptors.

Both hemoglobin and the NO gasoreceptor sGC exhibit reversible binding of O₂ and NO, respectively (Kharitonov et al., 1997; Winger et al., 2007). sGC binds NO with a relatively higher affinity (dissociation constant K_D of 4.2 pM - 54 nM), but dissociates more freely due to its rapid dissociation kinetics. In contrast, O₂ binds to hemoglobin with a relatively lower affinity (K_D of 1–10 μM) and dissociates less freely (Winger et al., 2007; Unzai et al., 1998; Tsai et al., 2012). However, since the concentration of O₂ in typical vertebrate tissues (5–50 μM) is higher than the concentration of NO (100pM-5 nM), hemoglobin’s relatively low K_D for O₂ should not exclude its consideration as an O₂ gasoreceptor in a microenvironment-dependent manner (Hall and Garthwaite, 2009). Similar arguments can be made for CO (K_D of 1–10 nM) and for other gases that can competitively inhibit hemoglobin (e.g., NO, H₂S, HCN) (Yuan et al., 2022).

Finally, in vertebrates with erythrocytes, the role of hemoglobin in gas transport largely depends on erythrocytes being passively carried in the bloodstream and O₂ being delivered to cells due to the Bohr effect (Malte et al., 2021). In human and mice, the NO gasoreceptor sGC is expressed not only in “immobile” vascular smooth muscle cells. Relatively mobile cells, such as platelets, neutrophils, and macrophages, also express sGC (Ciuman et al., 2006; Makhoul et al., 2018). Similarly, hemoglobin is expressed not only in mobile cells, such as mature erythrocytes. Hemoglobin subunit expression has been reported in alveolar epithelial cells, endothelial cells, cortical neurons, A9 dopaminergic neurons in the substantia nigra, cortical and hippocampal astrocytes, oligodendrocytes, epidermal keratinocytes, and chondrocytes and even in organelles such as mitochondria and membraneless condensates (Hedy’s) (Bhaskaran et al., 2005; Biagioli et al., 2009; Richter et al., 2009; Shephard et al., 2014; Brunyanszki et al., 2015; Tahara et al., 2022; Zhang F. et al., 2023; Reed et al., 2025). These findings suggest that, similar to sGC, hemoglobin is expressed in various mobile and immobile cells. However, the precise reason hemoglobin is still accepted as an O₂ transport protein in relatively immobile cells is unclear. Is it due to hemoglobin’s diffusion or intrinsic mobility within a cell (Kutchai and Staub, 1969; Richardson et al., 2020)? Or is it due to the interaction with other mobile proteins, entropy or the physical activity of immobile cells or animals? Therefore, hemoglobin, a multifunctional protein that transport gases, can also potentially act as a gasoreceptor.

The receptors in one-component STSs are evolutionarily advanced. They have at least two protein domains that perform different functions: one binds the ligand, and the other signals (Ulrich et al., 2005). Since simple gases are evolutionarily ancient than peptides, proto-receptors for gases probably consisted of proteins or nucleic acids with fewer domains (Harold et al., 2023; Anbalagan, 2024b).

The concept of proto-receptors for peptides has been proposed before, but a clear definition has yet to be established (Root-Bernstein, 2005). I define a proto-receptor as an evolutionarily ancient minimalistic receptor type, in which the signaling and sensing domains completely overlap, and which lacks any additional sites for allosteric regulation. Signaling due to proto-receptors can be considered to function in a proto-component STS, that predates one- and split-component STSs. In simple terms, ancestral proteins lacking allosteric regulation can be considered as proto-receptors for the competitive inhibitors that bind and inhibit their activity. For both proto-cells and also modern cells, the absence of a signal can itself serve as a signal. As an analogy, consider a traffic signal that stopped working at a busy intersection. Drivers approaching the intersection would still interpret it as a message to proceed with caution at their own risk. Extending the concept of proto-receptors to gaseous ligands, I define a proto-gasoreceptor as a protein that becomes competitively inhibited by the binding of a gaseous molecule. This results in an absence of cellular signals or responses. O₂ and hydrogen cyanide (HCN)-binding protein Campylobacter jejuni truncated hemoglobin P (Cj-trHbP) has been demonstrated to exhibit peroxidase-like activity in vitro (Lu et al., 2007; Bolli et al., 2008; Ascenzi and Pesce, 2017). Thus, Cj-trHbP can be considered as a candidate O₂ and HCN proto-gasoreceptor in a proto-component STS.

However, the allosteric regulation of protons, CO₂ and 2,3-Bisphosphoglycerate on mammalian hemoglobin evolved after the heme-containing globin domain had already emerged (Faggiano et al., 2022; Storz, 2025). Therefore, mammalian hemoglobin can be considered as a proto-gasoreceptor derivative (Vinogradov et al., 2006).

In mature erythrocytes in humans, hemoglobin is abundant protein (at 5 mM), making up about 90%–95% of their cellular protein content (Pisciotta and Sullivan, 2008). Along with the erythrocytes, hemoglobin is cyclically carried to hypoxic capillaries, with erythrocyte residence time ranging from sub-second to a few seconds depending on tissue vascularization and blood flow rates (Pittman, 2013). Erythrocyte’s transit time will be further affected by the transit time of other cell types in the capillaries or the pathological state of the vasculature, such as lumen narrowing observed in Atherosclerosis (Wang and Popel, 1993; Hogg et al., 1994). These observation suggest that it could be relatively easier for hemoglobin to be deoxygenated not only in healthy capillaries in physiological conditions but also of vasculature in pathological conditions. In a low-O₂ environment in vitro and under hypoxic conditions in vivo, deoxygenated hemoglobin in erythrocytes exhibits nitrite reductase activity, generating NO in a NO synthase-independent manner. This process has also been associated with vasodilation in human subjects, a canine model of hypotonic intravascular hemolysis, and the inhibition of mitochondrial respiration in cell cultures (Brooks, 1937; Nagababu et al., 2003; Basu et al., 2007; Minneci et al., 2008; Gladwin and Kim-Shapiro, 2008; Dent et al., 2021; Helms and Kim-Shapiro, 2013).

Apart from mature erythrocytes, the nitrite reductase activity of the hemoglobin α (HBA) subunit has also been demonstrated in the endothelium of mouse resistance arteries (Keller et al., 2022). Furthermore, endothelial cell-specific conditional Hba ^−/− mutant mice exhibited reductions in multiple parameters, including nitrite consumption, hypoxia-induced vasodilation, and exercise capacity. These results suggest a physiological role for HBA-derived NO under hypoxic conditions. Similarly, the nitrite reductase activity of myoglobin reduces myocardial infarction in mice via NO upon nitrite treatment; this response is lacking in myoglobin−/− mutants (Hendgen-Cotta et al., 2008).

However, the relatively low rate constant of human hemoglobin (1–10 M^-1s^-1), the local concentration of the co-substrate nitrite (200–300 nM in erythrocytes) and endogenous NO scavenging mechanisms, make it difficult to determine whether NO generated from deoxygenated hemoglobin acts as a signaling molecule in physiological or pathological conditions without an exogenous nitrite supply (Dent et al., 2021; Dejam et al., 2005; Wang and Kluger, 2016; Chamchoi et al., 2018).

Overall, at least in laboratory settings, the nitrite reductase activity of human and mouse hemoglobins is insufficient to label hemoglobin as a proto-gasoreceptor-like protein (Figure 1B). Nevertheless, in the case of cyanobacteria Synechocystis hemoglobin (SynHb), rice nonsymbiotic hemoglobin (nsHB1), and Arabidopsis thaliana phytoglobins (AHb1 and AHb2), which encounter relatively high nitrite concentration in vivo and exhibit nitrite reductase activity with relatively high rate constants, such hemoglobin orthologs are more promising candidates for O₂ proto-gasoreceptor or as O₂ proto-gasoreceptor derivates (Sturms et al., 2011; Tiso et al., 2012; Kumar et al., 2016). Vertebrates hemoglobin could perhaps act as an endogenous nitrite reductase with receptor-like signaling activity under conditions when nitrite concentration could be higher due to excess of environment-derived nitrite, dietary activities, nitrite enrichment responses, or the lack of homeostasis in the nitrite synthesis pathways (Washio and Takahashi, 2025; Sokal-Dembowska et al., 2025).

A split-component STS is a type of cellular signaling mechanism. It involves two proteins: one that constitutes a receptor or input domain, which binds or senses an environmental or intracellular stimulus, and another that constitutes a signal transducer with an output domain, which triggers a response. The ligand-binding status can influence the complex formation and activity of the signal transducer protein. In human and mice, the glucokinase regulator (GCKR, also known as GKRP or glucokinase regulatory protein) and glucokinase are examples of split-component STSs. The glucokinase regulator acts as a receptor for fructose 1-phosphate and fructose 6-phosphate, with glucokinase serving as its signal transducer (Beck and Miller, 2013; Anbalagan, 2025).

The presence of receptor in split-component STS for phosphorylated fructose in mammals raises the question of the identity of O₂ gasoreceptors acting in split-component STSs (Figure 1C). In theory, any O₂-binding protein that can physically interact with an enzyme and affect its activity could be considered an O₂ gasoreceptor in split-component STSs. In this context, the major O₂-binding proteins hemoglobin, cytoglobin, neuroglobin, and myoglobin could be considered potential O₂ gasoreceptors and their interacting enzymes could be considered potential signal transducers, provided that the O₂ binding state-dependent interaction with the enzyme promotes or inhibits the enzyme activity. As detailed in the following sections, hemoglobin interacts with various proteins in the erythrocytes, endothelial cells, and mitochondria. The precise reasons for the hemoglobin subunits’ interaction with proteins that exhibit diverse functions are unclear. Either hemoglobin “hitchhikes” across various proteins while transporting O₂, or its transport role is regulated. Alternatively, hemoglobin may be an O₂ gasoreceptor or signal transducer in split-component STS, with some of the interacting proteins acting as signal transducers and receptors, respectively.

The term “receptor-based signaling” is generally well accepted when a nuclear response is triggered as part of the ligand-receptor activation and signal transduction (Roberts and Kruchten, 2016). However, extracellular and intracellular receptors have been accepted in nuclei-less cells, such as mature erythrocytes and platelets (Tanneur et al., 2006; Vona et al., 2019; Jacobson et al., 2011; Balabin et al., 2018). This suggests that a nuclear response is unnecessary when considering a protein that is strongly allosteric regulated by ligand binding and can trigger a physiological cellular response.

However in mature erythrocytes, if we consider hemoglobin to be an O₂ gasoreceptor in split-component STS, what is the identity of the signal transducers (Ellsworth et al., 2009)? Another question to consider is what can serve as a signal transducer in split-component STS? Can it only be enzymes, or could it also be other classes of proteins, such as membranal anion exchangers? Since, receptors in one-component STSs can be both ion channels (e.g., mammalian temperature-sensing TRPV ion channels) and enzymes (e.g., E. coli O₂ gasoreceptor DosP phosphodiesterase), why not consider ion channels or anion exchangers as signal transducers (Figure 1C) (Dhaka et al., 2006).

In erythrocytes, hemoglobin binds to solute carrier family 4 member 1 (SLC4A1), also known as band 3 anion exchanger (AE1), in a manner dependent on its O₂ binding state. This affects the function of SLC4A1 in exchanging bicarbonate (HCO₃ ⁻) for chloride (Cl⁻) ions in erythrocytes. This interaction can influence erythrocyte cell physiology including their acid-base balance, metabolic regulation, membrane stability, survival, and lifespan (Chu et al., 2016; Cendali et al., 2024; Jennings, 2021). Therefore, in mature erythrocytes, I propose hemoglobin as an O₂ gasoreceptor and SLC4A1 as one of its signal transducers in split-component STS.

In mature erythrocytes, another potential signal transducer for hemoglobin in split-component STS is the deoxygenated hemoglobin itself. Heterogenous saturation within hemoglobin tetramers has been reported, and hemoglobin primarily exhibits nitrite reductase activity in its tetrameric form (Huang et al., 2005). The fact that peak nitrite reductase activity occurs at 40%–60% O₂ saturation of hemoglobin suggests that the oxygenated hemoglobin may act as O₂ gasoreceptors while the other bound deoxygenated hemoglobin that exhibit nitrite reductase activity could be signal transducer (Huang et al., 2005; Angelo et al., 2008). However, this proposal have the same bottleneck that I proposed earlier in the proto-gasoreceptor derivative section. Another alternative consideration is the role of oxygenated hemoglobin’s NO dioxygenase activity as a signal transducer, and deoxygenated hemoglobin as an O₂ gasoreceptor in split-component STS. However, the nitrate-specific signaling mechanisms derived from hemoglobin are unclear (Gardner, 2012).

In endothelial cells, a potential signal transducer for hemoglobin in split-component STS is the endothelial NO synthase (eNOS or NOS3), which synthesizes NO. Hemoglobin α and β subunits can interact with endothelial eNOS in mice arterial endothelial cells (Straub et al., 2012; Marozkina et al., 2021). While these interactions seem to play an important role in NO signaling, it is still unknown whether the physical interaction between hemoglobin subunits and eNOS and the activity of eNOS depend on the O₂ binding state (Straub et al., 2014).

In the mitochondria, a potential signal transducer for hemoglobin in split-component STS is the adenosine triphosphate (ATP) synthase (Shephard et al., 2014; Brunyanszki et al., 2015). The physical interaction between hemoglobin β and ATP synthase (ATP5A) has been observed in lysates from rat liver and Drosophila mitochondria (Ebanks et al., 2023). However, it is unclear whether the mitochondrial complex formation or dissociation in vivo can regulate ATP synthase activity depending on hemoglobin’s binding status to O₂. Although the role of ATP synthase is well known for its synthesis in ATP, ATP synthase is also a negative regulator of the mitochondrial permeability transition pore, suggesting potential response in the mitochondria (Picard and Shirihai, 2022; Pekson et al., 2023).

Nevertheless, for mitochondrial ATP synthase to be considered a signal transducer, precise knowledge of the intracellular signaling pathways downstream of its products ATP, H₂O and proton is required (if ATP synthase activity is the only enzyme activity exhibited by ATP synthase). ATP is well known as a substrate for numerous enzymes. Based on the rationale used for O₂ sensors, many ATP-dependent enzymes can also be considered as ATP sensors (Li et al., 2023). However, most of this signaling will not be considered signaling due to ATP receptors unless the ATP sensor is also strongly allosterically regulated by ATP binding at a site distant from the catalytic site. Currently, only a few extracellular ATP-binding proteins, such as P2X ion channels and P2Y G-protein-coupled receptors, are accepted as ATP or purinergic receptors (Ledderose and Junger, 2020). It is unclear whether similar ATP receptors exist intracellularly in the cytoplasm or various organelles (Fountain et al., 2007). Likewise aquareceptors and proton receptor-mediated signaling pathways too must be considered (Anbalagan, 2024d). Once all these signaling pathways acting downstream of mitochondrial ATP synthase are identified, ATP synthase can be considered a signal transducer in split component-STS.