Although HOBS are omnipresent, they have received relatively little attention, especially as a protein source for human consumption. To be sold for human consumption within the European Union, the bacterial protein has to be accepted by the European Food Safety Authority. One of the rules is that the bacterium has to grow in pure culture. Because of this, we review a single species rather than mixotrophic cultures. A HOB with a high potential for this production process is Cupriavidus necator. It has a protein content of up to 70% of the dry weight and a high-quality amino acid profile [14, 20]. This is also the best-described species of HOB in literature, taking into account the different names the species has had (Alcaligenes eutropha, Hydrogenomonas eutrophus, and Wautersia eutropha). The full genomes have been sequenced for the H16 strain [37, 38] and for the JMP134 strain [39]. All together, this seems to be the most suited strain for the near future.

C. necator is mainly known for its ability to produce polyhydroxyalkanoates, and it is the model organism for this type of metabolism [40]. Polyhydroxybutyric acid is formed under autotrophic growth and nutrient limiting conditions [41].

The ratio of the gas mixture greatly influences the growth rate of the bacteria. Since a mixture of hydrogen and oxygen gas can easily explode, the preferences of the bacteria and safety must be taken into account. For oxygen, the lower explosion limit is 6 vol% in excess of hydrogen gas [42]. Fortunately, 4 vol% O₂ in the gas phase was found to be optimal for the growth of C. necator [43].

For the growth rate, a maximal growth rate of 0.12 h.^-1 was found by Yu and Lu [44], in an experimental study. This was at 30°C and with H₂:O₂:CO₂ equaling 7:2:1 in the gas phase. A reaction heat of 1764.9 kJ C.mol⁻¹ was observed until a density of 7.7g/L was reached for the dry weight. In the exponential phase, the following equation is applied for the growth of the bacteria:

CO₂ + 7.77H₂ + 2.87O₂ + 0.24NH₃ → CH_1.68O_0.46N_0.24 + 7.28H₂O

The energy efficiency of electrical energy from biomass was found to be 6,4% under autotrophic conditions [45]. In comparison, for microalgae, maximal energy efficiency of 4-6% was found to go from electrical energy to biomass [46].

As shown above, the maximum growth rate of 0.12 h.^-1 can be obtained under ideal conditions. However, if nutrients are not supplied in the right ratio, the microbe will not develop protein or will grow slower. Carbon dioxide concentration is an important factor in the growth rate of bacteria [47]. If CO₂ limiting conditions occur, the hydrogen efficiency will decrease drastically because the bacteria will not have enough carbon available to convert to cellular material [45]. Another factor for growth is nitrogen concentration, which has a big impact on the specific growth rate. For the bacterium C. eutrophus, a species related to C. necator, the specific growth rate drops more than 30% when 50% of the necessary nitrogen is available [48]. If nitrogen-limiting conditions occur, the carbon will be used for the formation of PHB, which is not a protein [47].

Hydrogenases are enzymes which catalyse the cleavage of H₂ into hydrogen ions (H⁺) and electrons. The electrons from H₂ oxidation are used to generate ATP or produce reducing factors like NADH in C. necator [50-52]. C. necator carries four types of hydrogenases, but one of the hydrogenases is of no interest to this chapter due to its inactivity in the lithoautotrophic pathway [53]. The three that are relevant are regulatory hydrogenase (RH), membrane-bound hydrogenase (MBH), and soluble hydrogenase (SH) [52, 54, 55]. The active site of all three enzymes contains a nickel-iron ([Ni-Fe]) metal complex that acts as the catalyst. Via redox reactions, this metal group can convert H₂ to separate hydrogen ions and electrons [56]. Oxygen can also bind to the [Ni-Fe] catalyst, which deactivates the enzyme. To survive in an oxygen-rich environment, all these three hydrogenases need to be modified to withstand oxidation by O₂. This is done in a different way for each hydrogenase, which will be elaborated on accordingly.

The RH is located in the cytoplasm of C. necator and functions as a hydrogen sensor [54, 57]. When there is H₂ present, the RH activates regulators, which promote gene transcription of the MBH and SH operons [54]. The oxygen resistance of the RH is based on the access of O₂ to the active site. Special gas channels in the amino acid sequence prevent oxygen from reaching the active site by using bulky amino acid residues that make the channels too narrow for oxygen. The smaller H₂ molecule is guided through these channels to the active site [58].

The MBH is located at the outside of the cell membrane. It contains a cytochrome b subunit, which transfers electrons from the hydrogenase to the ubiquinone pool of C. necator [50]. The oxygen resistance of MBH can be explained by the presence of a [4Fe-3S] cluster next to the [Ni-Fe] site [59]. This [4Fe-3S] structure is responsible for the reduction of oxygen to water with hydrogen and electrons, as presented in reaction 2. First, four electrons from C. necator are transported to the catalytic centre via a reverse electron flow, where they reduce the bound O₂. Subsequent reaction with four protons (H⁺) forms two H₂O molecules that can leave the catalytic centre via special water channels in the enzyme. By reducing the bound oxygen to water at the active site, the enzyme is free again to undergo any subsequent hydrogen cleavage reactions [60]. Binding of oxygen to the metal catalyst temporarily inhibits the activity of the enzyme. Compared to anaerobic conditions, the activity of the isolated MBH hydrogenase is decreased by approximately 40% at an oxygen concentration of 5 vol%. The activity is retained for 40% under a dissolved oxygen concentration of 20 vol% [61].

4H⁺ + 4e- + O₂ ↔ 2 H₂O (reaction 2)

In the cytoplasm of C. necator, the SH is located [62]. The catalytic [Ni-Fe] site is connected to a NAD⁺ reductase subunit with a relay of [4Fe-4S] and [2Fe-S2] complexes, which facilitate electron transport between the two sites [63]. The electrons produced by the hydrogenase are used by the reductase to reduce NAD⁺ to NADH, which is an important cofactor in the metabolism of living organisms [52]. The oxygen resistance of SH is regulated by multiple different dynamics. The first is the inhibition of O₂ diffusion through gas channels by specific amino acid residues, decreasing the amount of oxygen that reaches the active site of the enzyme [64]. Secondly, by reduction of O₂ bound to the active site, the SH can free itself, similar to the oxygen-resistant mechanism of MBH. The key difference between them is that SH has two binding sites for oxygen instead of one, the catalytic [Ni-Fe] site and the NAD⁺ binding site. Additionally, during the reduction of oxygen in the SH, there is not only the production of water but also hydrogen peroxide (H₂O₂) and low amounts of superoxide (O₂-) [65, 66]. These pathways thus ensure the reduction of oxygen bound to the catalytic SH site of C. necator, which allows growth in oxygen-containing environments. In the SH, oxygen present in the environment has only a small effect on enzyme activity. Compared to anaerobic conditions, no significant decrease in activity is visible at an oxygen concentration of 20%. Additionally, the activity of the isolated SH hydrogenase decreases with 20% when 80 vol% oxygen is present. At 40% oxygen, the maximal oxygen concentration for the growth of C. necator, 3% of the electrons produced by H₂ oxidation are reversibly used to reduce the oxygen bound to the active site [66].

C. necator has the ability to fixate CO₂ from its environment. The Calvin-Benson-Bassham (CBB) cycle is utilized for the fixation of CO₂ [16]. C. necator’s H16 strain uses, amongst others, intermediates of the reverse tricarboxylic cycle (rTCA) to support heterotrophic growth [67]. Under autotrophic conditions, C. necator fixes carbon dioxide via the Calvin-Benson-Bassham (CBB) cycle, which is more similar to plant protein, which is derived from plants using the CBB cycle with rubisco as the protein [16]. The bacterium further also has an oxidative-TCA cycle which is more active under mixotrophic conditions, such as when oxidizing hydrogen, than under heterotrophic conditions [68]. In the reductive tricarboxylic acid (rTCA) cycle two molecules of CO₂ are used for the biosynthesis of acetyl-CoA. This makes rTCA a reverse process of the oxidative TCA cycle, also known as Krebs cycle. C. necator makes use of the (r)TCA cycle during cultivation in bioreactors [68-71].

An airlift reactor has been proposed for the growth of HOBS, taking the latter limitations into consideration. Airlift reactors consist of an aerated riser and a downcomer, indicated in Fig. (2), which induce hydrostatic pressure differences over the tank [76]. A liquid flow over the longitudinal direction of the tank is induced, leading to a decrease in gas hold-ups and improvement in oxygen transfer. A small headspace would be advisable in this case, which would keep the hydrogen and oxygen concentration within the explosion limits (hydrogen: 5-75%) [77].

A good alternative would be a packed bed bioreactor, which consists of a column stacked with solid materials. The liquid phase spreads as a layer over the column, whereas the gas stream moves in the counter direction [78]. Hereby, a larger specific surface is created, allowing for high kOLa values at low gassing rates (< 0.02 vvm) and ideal stream velocities [45]. In addition, the pressure build-up will be considerably lower, meaning less electrical power will be needed to acquire a certain gas velocity.

Most research has been done so far with HOBS in conventional airlift bioreactors and several ways HOBS growth can be optimized have already been discussed. Other, more rigorous, ways of improving mass transfer area, however available, such as the use of Microbial fuel cells (MFCs) and Microbial Electrolysis cells (MECs) [79, 80]. In these cases, hydrogen is produced by microorganisms breaking down e.g., organic matter, resulting in hydrogen yields up to 100% that are directly introduced into the liquid [81]. However, these new methods need organic sources, whereas the principal advantage of hydrogen-oxidizing bacteria is based on their growth in renewable sources and waste streams.

On the other hand, there are emerging technologies researching the use of combined reactors for HOB growth and electrolysis, without the use of extra micro-organisms. These reactors aim to optimize the hydrogen gas production and delivery into the liquid stream and show much potential for the growth of HOBS in waste streams. In this case, HOBS would be grown in a two-chamber cell, which is available in various shapes and sizes. As these techniques are incredibly novel and still under research as well, HOB growth in them is not realistic in the near future [82, 83].

When looking at the potential of HOB, the quality and availability of its protein determine the potential use and value. The amino acid (AA) composition of C. necator happens to be slightly superior compared to other bacterial proteins and is nearly equal to the milk protein casein, which is another frequently used protein [29].

All essential amino acids (histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan and valine) are present in C. necator (Fig. 3), which means it suffices as a singular protein source in one's diet. To cover the daily essential amino acid requirements of an average adult (62 kg) 72-96 g is required, depending on the growing medium of C. necator [98]. 121 g of soybean is needed to accomplish the same amino acid concentrations [99]. Overall, C. necator is comparable with soy and whey protein concentrates and could therefore be a good protein alternative based on its AA composition. The study of Ritala et al. [22] comparing SCPs from algae, fungi and bacteria agrees that C. necator has a similar or even better AA composition than soy and algae. Other HOB were investigated by Volova & Barashkov [34], who compared the AA profile of Alcaligenes eutrophus Z1 and Ralstonia eutropha B5786 to that of yeast, microalgae and casein, and concluded their protein profiles were similar. Thereby the protein content of these bacteria was 70% d/w, which was much higher than the 50 and 15% in yeast and wheat grain.

To describe the small differences in the AA-composition, HOB have a relatively high level of alanine and low amount of cysteine. The levels of alanine, glycine, methionine, tyrosine and valine of C. necator exceed the levels of both soy protein isolate and whey protein isolate. For aspartic acid, glumatic acid, histidine, isoleucine, lysine, proline and serine levels C. necator scores under the levels of its competitors. Even though the lysine level of C. necator is the lowest in relation to the others, Calloway & Kumar [14] suggested C. necator as an excellent supplement to cereal grains.

Fig. (3). Overview of amino acid composition in C. necator {Foster, 1964 #224}, soy and whey protein concentrate {Kalman, 2014 #228}.

The bioavailability of HOB, soy and whey proteins are displayed in Table 1. Nitrogen digestibility is the percentage of the absorbed nitrogen of the total nitrogen intake [14]. The biological value is based on the amount of nitrogen used for tissue formation divided by the nitrogen absorbed from food and multiplied by 100, expressed as percentage of nitrogen utilized [17]. The biological value indicates how efficiently the nitrogen is used by the body. The net nitrogen utilization is calculated by multiplying nitrogen digestibility by the biological value and then dividing by 100 [14]. These values were given for soy protein and whey protein [17]. The nitrogen digestibility values of soy and whey proteins, however, were calculated with the above-mentioned formula. As there is a difference in metabolism and natural diet between rats and humans, it should be taken into account that the study of Calloway & Kumar [14] used rats to test the bioavailability, whereas soy’s and whey’s values were obtained by humans.

The nitrogen digestibility of C. necator is significantly higher than that of soy and whey protein. Looking at the biological values and net nitrogen utilization, C. necator scores higher than soy, but lower than whey. These values suggest that C. necator is a better protein alternative than soy.

From an economical perspective, the production costs of hydrogen oxidizing bacteria should not be higher than already existing protein sources. This is already true in the case of HOB competing with space food [97]. However, when using HOB-protein as a supplement, soy and whey concentrate are considered competitors. 71-75% of C. necator’s dry weight consists of protein, which is respectively around 63% and 90% for soy and whey concentrate [14, 20]. Prices of soy protein are around 1 Euro/kg as competing for plant protein and 2 Euro/kg for the high-quality fish protein [72]. In a study by Zhang et al. [100], it was found that, with electricity costs at V0.05 kW h-1, growing a HOB culture was feasible. However, this study used a mixed culture of HOB, producing several valuable bioproducts, among which proteins. Also, no exact costs or market prices were mentioned.

The growth of HOB uses minor land occupation and water compared to the conventional ways of growing proteins such as soybeans, especially when the wind is used as an energy source [101]. It proves to be no threat to either biodiversity or the surrounding ecosystem since there is no chance of eutrophication, soil water pollution and extensive water usage. Furthermore, the production of the hydrogen oxidizing bacteria does not cause an increase in the emission of CO₂, methane, or nitrous oxide, as there is no net emission of these gasses [20]. This means that for the growth in the bioreactor the carbon cycle is conserved, and there is zero additional carbon dioxide emission in the production of the product [102]. In addition, there is no fluctuation in price or yield due to seasonality, ensuring food security and stable pricing. Therefore, HOB manufacturing can provide reliable production close to the point of consumption.

Given the previously discussed topics, C. necator protein can compete qualitatively and quantitatively with soy protein and whey protein amongst some other SCP like fungi and algae. The technology needed to overcome the gas issues to cultivate HOB, turning it into a protein-rich ingredient, is available. There is more investigation needed on the upscaling process of the production, but this is not the limiting factor as technologies are making progress [26, 103].

Regarding food safety, especially RNA content, endotoxins, and potential allergy symptoms should be taken into account [22]. When the bacteria are grown too fast, or their RNA is not processed, the high nucleic acid content resulting in uric acid will be toxic in large quantities (>2 gram per day) [12, 23, 104]. There are several methods developed to reduce the RNA content for human consumption [12]. For example, RNAase could be used at a temperature to 60-70 °C for 20 min, alkaline hydrolysis, autolytic methods and chemical extraction. The production process of the final product of HOB-based SCP should be done carefully. Gram-negative bacteria like C. necator appear to contain endotoxins integrated in the cell walls. Endotoxins could cause inflammatory reactions [18, 105]. To inactivate endotoxins, hydrogen peroxide is used for biosynthesis of poly(3-hydroxybutyrate-co-4-hydroxybutyrate) produced by C. necator [106]. However, it is unclear how these treatments affect the quality of the proteins in HOB. An alternative method to inhibit endotoxin formation would be suppressing or the genes responsible for the toxins [18]. Allergic reactions could be minimized by destroying the cell wall using either mechanical techniques (soniciation, blending, agitation with beads, etc.) or enzymatic and chemical treatments [15, 107]. The specific details of the treatments are dependent on the species or strain and the quantity of required protein.

As HOB-protein is not yet on the EU market, it is likely to be considered as a novel food. However, before being considered novel foods, the HOB-protein product must go through an authorization and consultation process concerning implementing acts according to the European Food Safety Authority's (EFSA) regulations [108]. When confirmed as novel food by the recipient EU country, the Commission will publish information on the Commission’s website meaning that the novel food can be lawfully placed on the EU market [109]. The conditions for microorganisms as novel food are the identification of species, sufficient information for characterisation (genetic typing) at strain level by internationally accepted molecular methods, and naming of strains according to the International Code of Nomenclature [108]. In other words, HOB-protein must be produced from a pure culture. Unlike a mixed culture, which is more efficient and cheaper to produce, a pure culture is commercially unattractive as it is not cheap enough to compete with soy [110]. The need for sterilisation, and lack of symbiosis between strains make pure culturing a costly process. As long as the European Union does not adjust its regulations, this will be the bottleneck for successful implementation of HOB-protein in Europe. Consequently, countries where the regulations are less strict (e.g. US, UK, China), might succeed sooner.

HOB is produced at a large scale for bioplastics, feed (for fish and livestock) and bacterial oils [25]. To give an image of the efficiency of production of HOB for bioplastics (polyhydroxylalkanoates). A cell concentration of 150-160 kg/m^3 over a course of 60 hours in a pilot plant of 150L was obtained [111]. Some companies producing HOB for bioplastic or feed are Kiverdi (plastics) company from 2011 and Novonutrients in cooperation with Deep Branch Biotechnology (feed) [25]. Since 2019, Kivirdi also focused on producing HOB to be the first in producing meat based on air. Their product is called air protein, which is a protein powder ingredient [112]. Novonutrients started around that same year with HOB production for human consumption. Other promising companies producing HOB-based protein as food are the Finnish Solar Foods company and the Belgian Avecom company. The Belgian company Avecom collaborated with the Dutch KWR for the Power to Protein project. In the project they investigated the use HOB for food [113]. They have run a pilot bioreactor (580L) in which they produced biomass with 49% SCP and 85% digestibility (in vitro) [95]. Lab scale values were not reached, but it is a promising start. Solar foods were founded in 2017 and produce Solein as protein ingredient for human consumption [96]. In 2020 it raised funding of almost 25 million euros for ‘producing a unique single-cell protein out of thin air’ project [114]. This is a promising start for implementation of HOB-protein as ingredient on a large scale.

Heuristics are important in people's willingness to accept products. The association and emotions linked to the type of product can play a crucial role. Naturalness and associations with healthiness of a product can increase positive association. Trust is also an essential heuristic. For acceptance of the idea of shared values is necessary, the consumer needs to believe the producers have the same ideas regarding the goal of the products as they have themselves [116].

What can help build a more positive association lies within the framing of the product. This can be seen in the framing of cultured meat. The way in which it is described affects the willingness to accept [117]. An example of positive framing is by using labels. Labels for organic products from an independent accredited institute can be used for differentiation by brands. Positive perception of brands due to organic labelling increases on global, local and private scale and consumers are more willing to pay a premium price for the brand [118].

The psychological factors that could play are role are food technological neophobia. Food technology neophobia refers to the willingness of people to eat novel technological products. There is a scale which can measure this [119] and it can be used in research on the acceptance of products. For example, with the production of 3D printed food, food technology neophobia is high and hard to overcome [120]. For hydro-oxidizing bacteria, further research is needed.

However, coming from a structuralist perspective, House [115] argues that acceptance should not be focused on getting people to try the product but on incorporating the product into their daily lives. Though characteristics such as sustainability could persuade people to buy a product once, other structural factors such as price, taste and availability are much more important in getting people to buy it more often. Furthermore, House [115] states that in the initial stage of production, the focus should be on “early adopters” who are eager to try new products rather than the average consumers who will likely follow later.

The above-mentioned theories on acceptance of similar novel food products can be used when thinking about the acceptance of HOB. However, it must be noted that every situation is different, and what can be used for one product may not work for another. Looking at the context is always important, and these theoretical frameworks can be used as tools to apply to the specific case of hydrogen-oxidizing bacteria.