Since many Bacillaceae strains exhibit strong protease activity, it is difficult to purify intact cytochrome c. Although Bacillaceae strains should not have soluble cytochromes c due to the lack of the periplasmic space in the cell, several soluble cytochromes c have been purified. It has been reported that B. subtilis 168 possesses two types of membrane-binding cytochromes c-550 and c-551. Cytochrome c-551 has a molecular mass of 10 kDa and is composed of 92 amino acid residues, including 14 basic amino acid residues in its processed form, and binds to the membrane via a diacyl-glyceryl-cysteine moiety by modification of its N-terminal sequence (Figure 2 and Table 1; Bengtsson et al., 1999). Cytochrome c-551 is the counterpart of cytochrome c to cytochrome c-550 in obligate alkaliphilic E. clarkii. The midpoint redox potential was considered to be > + 100 mV and its pI was 3.8. However, the function of cytochrome c-551 has not yet been elucidated. It is speculated that cytochrome c-551 is involved in the electron transfer between cytochrome bc₁ and cytochrome c oxidase. The phylogenetic position of cytochrome c-551 was similar to that of Bacillus licheniformis SCDB cytochrome c (Figure 3). Cytochrome c-550 has a molecular mass of 13 kDa and is composed of 120 amino acid residues with a membrane anchor domain consisting of a single α-helical transmembrane segment of a hydrophobic polypeptide comprising 30 amino acid residues and a heme- containing main body protein comprising approximately 74 residues with a calculated pI of 5.4 (von Wachenfeldt and Hederstedt, 1990). The midpoint redox potential of cytochrome c-550 is + 178 mV (von Wachenfeldt and Hederstedt, 1993). The function of cytochrome c-550 is electron transfer from cytochrome b₆c to the cytochrome caa₃ terminal oxidase, and the main supercomplex formed by caa₃ oxidase, the b₆c complex, and cytochrome c-550 with ATP synthase has been isolated (García Montes de Oca et al., 2012).

The abundance of membrane-bound cytochrome c was higher at pH 10 than that at pH 7 in facultative alkaliphilic S. cohnii YN-2000 (Yumoto et al., 1991). The elution profile of anion exchange chromatography loading of the solubilized membrane fraction by the detergent, Triton X-100 exhibited a larger amount of cytochrome c-553 in cells grown at pH 10 than in cells grown at pH 7 (Yumoto et al., 1991). Cytochrome c-553 has a molecular mass of 10.5 kDa by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with a pI of 3.9 (Table 1). If cytochrome c-553 of strain YN-2000 is the same as that in S. cohnii DSM 6307^T, it consists of 93 amino acid residues including seven basic amino acid residues in the processed protein. The midpoint redox potential was + 87 mV in the pH range of 6–8. The native molecular mass, as determined by gel filtration, was 37 kDa. Therefore, it was suggested that cytochrome c-553 forms a tetramer in its native form in solution or in its original membrane-binding form.

The phylogenetic position of S. cohnii DSM 6307^T cytochrome c clustered with that of Sporosarcina pasteurii NCIM 2477 is shown in Figure 3. Cytochrome c oxidase in S. cohnii YN-2000, cytochrome aco₃, which is a different type from other Bacillaceae cytochrome c oxidases was purified and characterized (Qureshi et al., 1990). Cytochrome aco₃ reacts with cytochrome c-553, and the terminal enzymatic activity is greatly enhanced in the presence of poly-L-lysine, which accelerates interaction of two negatively charged molecules: cytochrome c-553 and cytochrome aco₃ (Yumoto et al., 1993).

The three-dimensional (3D) structure of cytochrome c in alkaliphilic Bacillaceae was first studied in S. pasteurii cytochrome c-553 (Benini et al., 2000). Cytochrome c-553 has a molecular mass of 9.6 kDa and consists of 92 amino acids having a midpoint redox potential of + 47 mV and a predicted pI of 3.96 (Benini et al., 1998). Most of the charges were localized on the opposite side of that exposed to the heme edge. This localization may be related to H⁺ transfer on the outer surface of the membrane. The correlation between heme solvent accessibility and entropy suggests a direct link between the major determinant of the electrical potential (entropy) and a structural parameter (heme solvent exposure). The low midpoint redox potential of cytochrome c could be attributed to the decrease in reduction entropy caused by the extrusion of water molecules from the reduction of the hydration shell in this protein. This event affects solvent accessibility to the heme during its reduction. Thus, analysis of the 3D-structure revealed important structural characteristics that explain the low redox potential and retention of H⁺ at the outer surface of the membrane of the cytochrome c in the alkaliphile.

The first purified and characterized cytochrome c from an alkaliphilic Bacillaceae strain was cytochrome c-552 isolated from A. pseudofirmus RAB (Davidson et al., 1988). Cytochrome c-552 has a molecular mass of 16.5 kDa with an acidic pI of 3.4. The midpoint redox potential of this cytochrome is + 66 mV at pH 7, which decreases at pH > 8.3, depending on the pH increase. According to resonance Raman spectroscopy, this pH-dependent decrease in redox potential may be attributed to a switch in the sixth ligand of heme c from methionine to histidine when the oxidized cytochrome is reduced (Larsen et al., 1990). Cytochrome c-552 is autooxidizable and is purified in a soluble form. These characteristics may arise due to the partial digestion of cytochrome c by intrinsic proteinases during purification. Although most strains belong to the same species as A. pseudofirmus RAB are obligate alkaliphiles (Nielsen et al., 1995), whereas A. pseudofirmus OF4 is a facultative alkaliphile. If cytochrome c-552 in A. pseudofirmus RAB is the same as that in the corresponding cytochrome c in A. pseudofirmus OF4, cytochrome c-552 consists of 105 amino acid residues and exhibits the same branching with cytochrome c in obligate alkaliphilic A. alcalophilus DSM 485^T (Figure 3).

The cell extract of obligate alkaliphilic E. clarkii K24-1U exhibits higher cytochrome c content than that in neutralophilic B. subtilis under high-aeration conditions (4.5-fold). The cytochrome c abundance in E. clarkii K24-1U further increases under low-aeration conditions (6.3-fold) (Hijikata, 2004; Matsuno et al., 2018). Cytochrome c-550 purified from E. clarkii K24-1U, exhibited very weak protease activity (Ogami et al., 2009). Characterization of cytochrome c-550, determination of its gene sequence, and constricted recombinant expressing C18M anchorless mutant protein enables us to understand the molecular features in native and anchorless cytochrome c molecules. Native cytochrome c-550 and C18M cytochrome c-550 had molecular weights of 11,083–11,115 in the analysis of the separated fractions in HPCL and 10,543 by Matrix-assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry (MALDI-TOF-MS) with an acidic pI of 4.1. The midpoint redox potential was + 83 mV in the redox titration (Table 1). The purified cytochrome c-550 exhibited two bands of 17 kDa and 20 kDa in SDS-PAGE. The native molecular mass determined by gel filtration was 40 kDa in both native and C18M cytochromes c. Therefore, it is suggested that cytochrome c-550 forms a tetramer in its native form in solution or in its original membrane-binding form. Cytochrome c-550 exhibited only one band of 23 kDa in blue native-PAGE for both native and C18M cytochromes c-550. These numbers of SDS-PAGE and blue native-PAGE of molecular mass correspond to the dimer based on the molecular mass of the minimum unit of molecular mass determined by MALDI-TOF-MS using fractions separated by HPLC and its gene sequence. These two bands in SDS-PAGE were shifted to 10 and 15 kDa in C18M cytochrome c-550. Therefore, the appearance of the two bands in SDS-PAGE was not attributed to the N-terminal modifications but to the internal specific sequence located near the N-terminal sequence of Asn²¹-Asn⁴³ because the specific sequence lacking S. cohnii YN-2000 tetrameric cytochrome c exhibited only one monomer band in SDS-PAGE (Yumoto et al., 1991). The above results suggested that the binding molecular mechanisms of dimeric and tetrameric (dimeric plus dimeric) forms are different.

The determined gene sequence of cytochrome c-550 revealed the hydrophobic N-terminal sequence Met¹-Ala¹⁷ as the signal peptide sequence (Ogami et al., 2009; Figure 2). In addition, MALDI-TOF-MS analysis using fractionated by HPLC of the native protein and fatty acid analysis of cytochrome c revealed the structure of the modified moiety of the protein. During the processing of mature cytochrome c production, expressed cytochrome c-550 is translocated to the extracellular side of the membrane by the signal peptide. After the signal peptide was dissociated from the main body protein, terminal Cys¹⁸ was modified by attaching diacylglycerol and acetyl moieties. Cytochrome c-550 binds to the fatty acid length of C₁₅ in the internal moiety and different chain lengths of fatty acids, C₁₅, C₁₆, and C₁₇ in the external moiety via glycerol-Cys¹⁸. Molecular binding species are the major components of membrane fatty acids. Therefore, if the amount of cytochrome c-550 expressed fluctuates depending on the culture conditions, there will be a little influence on the fatty acid composition of the membrane.

The sulfate-reducing bacterium Desulfovibrio gigas possesses cytochrome c₃, which has four hemes per protein molecule (Coutinho and Xavier, 1994). It has been reported that cytochrome c₃ transfers H⁺ via a cooperative H⁺/e^– linkage (redox-Bohr effect) (Louro et al., 1997; Messias et al., 2006). These hemes exhibit different redox potentials by direct measurement on the electrode attached as self-assembled monolayers (SAMs) and by redox titration using mediators in a solution. This difference in redox potential in cytochrome c suggests changes in the redox potential depending on the distance between each heme (heme I–IV) and the electrode. Shorter distances between heme c and the electrode resulted in larger differences in the redox potential of the cytochrome c molecule between in solution status and the electrode-attached state of the cytochrome c molecule. These differences are attributed to the strength of the electric field (Coulomb force), which depends on the distance of each heme from the electrode (Rivas et al., 2005). The redox potential of cytochrome c-550 of E. clarkii measured by redox titration was + 83 mV (in a buffer solution) whereas that of cytochrome c-550 immobilized on the gold electrode using a SAMs of 2-amino-6-purinethiol was + 7 mV. These facts suggest that if the distances between the membrane surface and the main body cytochrome c differ depending on the chain length of fatty acids (C₁₅, C₁₆, and C₁₇), the redox potential of heme c may change depending on their distance from the membrane surface. These electric field strength-dependent differences in redox potential may have an important role for H⁺ transfer on the outer surface of the membrane in electron transfer-coupled H⁺ transfer mechanisms.

The amino acid sequence of E. clarkii K24-1U cytochrome c-550 deduced from the determined gene sequence was very similar to that of cytochrome c in E. clarkii DSM 8720^T (only one residue difference among 118 residues) (Figure 2). We collected data on bioenergetic parameters, such as O₂ consumption and ATP production rates, in E. clarkii DSM 8720^T. Therefore, we can consider these bioenergetic data combined with the data for E. clarkii K24-1U cytochrome c-550. Multiple alignments of two strains of E. clarkii cytochrome c with other cytochromes c from obligate and the facultative alkaliphilic and the neutralophilic Bacillaceae strains were constructed (Figure 2). Cytochrome c-550 contains much fewer basic residues (only two, including His for axial ligand of the heme c) compared with cytochrome c from facultative alkaliphiles and neutralophiles. This scarcity of basic amino acids in the obligate alkaliphilic Bacillaceae may be related to the acidic nature of the molecule, which expels OH^– and attracts H⁺ to protect the environment in the outer surface of the membrane. In the entire molecular structure, the particular part of the amino acid sequence of Asn₂₁-Asn₄₃ in the cytochrome c of the two strains of E. clarkii, which consisted of eight Asn residues and six acidic amino acid residues in 23 residues (Figure 2). The corresponding amino acid sequence of Asn₂₄-Asn₃₇ (Asn₇-Asn₂₀ in the processed protein base) was observed neither in the facultative alkaliphilic nor the neutralophilic Bacillaceae strains. This means that obligate alkaliphiles express cytochromes c, which express an even more effective protein for the alkaline environment adaptation than the neutralophilic or the facultatively alkaliphilic Bacillaceae.

Theoretically, the Asn (N) residue could influence H⁺-transfer. However, there have been few examples of the interpretation that accounts for the contribution of Asn in H⁺ transfer due to the weak hydrogen binding of this individual residue. In the catalytic reaction of methyltetrahydrofolate (MTHF) and corrinoid-iron sulfur protein (CFeSP) methyltransferase (MeTr), the methyl group of CH₃-H₄folate in MTHF is transferred to cob (I) amide. This reaction requires electrophilic activation of the methyl group of MTHF, which involves H⁺ transfer to the N5 group of the pterin ring of MTHF. This H⁺ transfer reaction is possible by the extended H-bond network including Asn¹⁹⁹, Asp¹⁶⁰, and a water molecule (Doukov et al., 2007). Thus, although Asn exhibits weak hydrogen bonds, the overall effect of this cumulative H-bond network is significant in this series reactions. Considering the cooperation of Asn and other H⁺ transferable acidic amino acid residues such as Asp or Glu, the amino acid sequence of Asn₂₁-Asn₄₃ in the cytochrome c of E. clarkii strains produces a cumulative H-bond network in the corresponding region.

Based on the reported structure of membrane-bound cytochrome c in Bacillus subtilis (David et al., 2000), the Asn-rich segment, Asn₂₁-Asn₄₃ (Asn₄-Asn₂₆ in the processed protein base) in E. clarkii cytochrome c is presumed to be located in the α-helical domain between the membrane anchor and the main body cytochrome c molecule. Therefore, it is considered that this region is located between the outer surface of the membrane and the main body of cytochrome c. Based on the tetrameric structure of this cytochrome c molecule, the Asn-rich segment, Asn₂₁-Asn₄₃, further accumulated in the corresponding location. It can be presumed that this accumulated Asn-rich segment, contributing to the intense H-bond network, influences H⁺ transfer on the outer surface of the membrane associated with the redox reaction of the main body of cytochrome c. The corresponding segment Asn₂₃-Asn₃₇ (Asn₆-Asn₂₀ in the processed protein base) was observed only in the obligate alkaliphilic Bacillaceae (Figure 2). If the corresponding region influences the H-bond network in the space between the main body cytochrome c molecule and the outer surface of the membrane, the segment of Glu₂₅-Lys₄₆ (Glu₆-Lys₂₇ in the processed protein base) in S. cohnii DSM 6307^T may also influence the H-bond network (Figure 2). The abundance ratio of acidic to basic amino acids was higher in the facultative alkaliphilic Bacillaceae than in the neutralophilic Bacillaceae (Table 1 and Figure 2). A similar region found in membrane-bound cytochromes c in alkaliphilic Bacillaceae also contain other membrane-bound protein molecules, such as subunit II of aco₃-type or caa₃-type cytochrome c oxidase, which have abundant acidic residues and a few basic residues (Denda et al., 2001; Noor et al., 2014). This region is located in the external hydrophilic cytochrome c-binding domain of subunit II of the aco₃-type or caa₃-type cytochrome c oxidase. This shows that the H-bond network construction strategy, which is found in membrane-bound cytochrome c, is also found in the subunit structure of other protein molecules that have subunits containing cytochrome c segments. This configuration is probably connected to cytochrome c, which exhibits a low redox potential (< + 100 mV) (Orii et al., 1991). It can be predicted that the abundance of acidic amino acids, tetrameric structures, and low redox potentials are important for redox-coupled H⁺ transfer on the outer membrane surface through the H-bond network.

One of the alkaline adaptation mechanisms of alkaliphilic Bacillaceae is based on the enhancement of H⁺ usage efficiency. This enhanced the efficacy of H⁺ utilization in ATP synthesis and retention of H⁺ in the vicinity of the outer surface of the membrane, compared to in neutralophilic Bacillaceae. In fact, it has been suggested that the number of H⁺ translocated across the biological membrane from the intracellular side to the extracellular side by the respiratory chain in alkaliphilic Bacillaceae was significantly lower than that in neutralophilic Bacillaceae (Hirabayashi et al., 2012; Goto et al., 2016). The oxygen consumption rate based on the endogenous substrates of B. subtilis IAM 1026 (growth in high aeration, pH 7; O₂ consumption was measured at pH 7) was 0.50 ± 0.06 μmol O₂⋅min^–1⋅mg cell protein^–1, while that of the obligate alkaliphilic E. clarkii DSM 8720^T [in high aeration, growth at pH 10; O₂ consumption was measured at pH 10) was 0.19 ± 0.04 μmol O₂⋅min^–1⋅mg cell protein^–1 (Table 2). The low oxygen consumption value is comparable to the data obtained using A. pseudofirmus OF4 (0.21 μmol O₂⋅min^–1⋅mg cell protein^–1 (based on indigenous substrates in the presence of Na⁺)] (Guffanti et al., 1986). A. alcalophilus and A. pseudofirmus RAB exhibited the same O₂ consumption rate (0.49 μmol O₂⋅min^–1⋅mg cell protein^–1) when L-malate is used as a substrate (Lewis et al., 1980). Therefore, the low oxygen consumption rate of E. clarkii DSM 8720^T may be attributed to the absence of the extracellular substrate. A low oxygen consumption rate compared to that of B. subtilis IAM 1026 under high aeration was also observed in facultative alkaliphilic S. cohnii YN-2000 (0.20 ± 0.07 μmol O₂⋅min^–1⋅mg cell protein^–1; O₂ consumption was measured at pH 10) under high-aeration condition. These relatively low oxygen consumption rates, compared with B. subtilis under high-aeration conditions in obligate and facultative alkaliphiles, may be attributed to the hindered translocation of the positively charged H⁺ by the respiratory chain via the attraction force of the intracellular negative charge due to the large ΔΨ. Thus, H⁺ transfer coupled with electron transfer in the respiratory chain may be deterred. The reduction rate of heme a in cytochrome c oxidase in vesicles reportedly increases when ΔΨ is abolished during steady-state respiration using cytochrome c (Gregory and Ferguson-Miller, 1989). This is probably due to the ΔΨ influencing the electron transfer from cytochrome c to heme a. Although it is expected that ΔΨ alters the redox potential of heme a, ΔΨ potentially inhibits the electron transfer from the outer side of the membrane (cytochrome c) to the inner membrane heme a due to repulsion between negative charges (electron negative charge vs. inner membrane negative charge by ΔΨ). Therefore, the difficulties in H⁺ transfer and electron transfer across the membrane under high ΔΨ may produce a low oxygen consumption rate in high aeration in E. clarkii DSM 8720^T.

In the measurement of the H⁺/O ratio in alkaliphiles and neutralophiles, each measured value was obtained by culturing and measuring at pH 10 and pH 7, respectively. Since Bacillaceae strains do not contain complex I, which translocates 4 H⁺/O, the theoretical H⁺/O ratio in the case of Bacillaceae strains is thought to be 4 H⁺ from complex III plus 2 H⁺ from complex IV, equal to 6 H⁺. The H⁺/O ratio of thermophilic Bacillaceae (Geobacillus) is 6–7 (Chichen et al., 1981; Sone and Fujiwara, 1991), while those in B. subtilis, Bacillus megaterium, and Bacillus licheniformis have been reported to be approximately 4 (Jones et al., 1975). In contrast, the H⁺/O ratio in another Bacillaceae, Brevibacillus brevis was reported to be 5.01 ± 0.26 (Yaginuma et al., 1997). The variation in the H⁺/O ratio depending on the Bacillaceae strain may be attributed to differences in the electron flow paths in the respiratory chain.

The H⁺/O ratios under high- and low-aeration conditions in obligate alkaliphilic E. clarkii DSM 8720^T were 2.2 ± 0.2 and 0.6 ± 0.1, respectively (Table 2). These values were much lower than the theoretical values. The H⁺/O ratios under high-aeration condition in facultative alkaliphilic S. cohnii YN-2000 and neutralophilic B. subtilis were 2.8 ± 0.8 and 4.9 ± 0.1, respectively (Table 2). The H⁺/O ratio in the facultative alkaliphilic strain was higher than that in the obligate alkaliphilic strain, whereas the value was lower than that in the neutralophilic strain and the theoretical values. To date, it is unclear why the H⁺/O ratios in alkaliphilic bacteria are much lower than the theoretical value. The possible reasons for these values might be due to the inhibition of translocation H⁺ by the larger ΔΨ compared with that in the neutralophiles or that the H⁺ translocated by the respiratory chain may remain on the outer surface of the membrane. However, for measuring the H⁺/O ratios, 0.15 μg/mL valinomycin and 0.85 mM KCl, which disrupt ΔΨ across the membrane, were used. Therefore, it is difficult to determine whether the reason for the low H⁺/O ratios in alkaliphiles can be attributed to ΔΨ alone.

There is a possibility that the extremely low H⁺/O ratio in obligate alkaliphilic E. clarkii DSM 8720^T is related to the Asn-rich distinct amino acid sequence in the membrane-bound cytochrome c, which may influence to the H-bond network probably related to the transfer of H⁺ at the outer surface of the membrane. It is possible that this specific structure in E. clarkii cytochrome c spools up translocated H⁺ via the respiratory chain on the outer surface of the membrane.

The H⁺/O ratio was found to be 2.8 even under low-aeration conditions of B. subtilis (Table 2). Since the pumped H⁺ was measured in bulk in the estimation of H⁺/O, the measured values may be influenced by the activity of the respiratory chain. Thus, the H⁺/O ratio may change depending on the cell activity. In addition, there is a possibility that the expressed respiratory components may change depending on the culture conditions. Therefore, it is difficult to compare the values directly among the reported studies. Cyanide-insensitive Non-Proteinaceous Substance (NPS) with a molecular weight of 622 was found in the respiratory system of the obligate alkaliphilic E. polygoni YN-1^T (Higashibata et al., 1998). Thus, it can be predicted that NPS does not influence to H⁺ translocation in the respiratory system. The terminal enzyme was a caa₃-type cytochrome c oxidase, constituting up to only 10% of the total oxygen reducing activity, while 90% of the respiratory activity was attributed to cyanide-insensitive NPS in strain YN-1^T. If the NPS is widely distributed in alkaliphilic Bacillaceae, it may lead to the reduction of the H⁺/O ratio. NPS may prevent the production of excess energy during the later growth phase in the effective energy production systems of alkaliphiles.

Since the O₂ consumption rate is considered to be the turnover rate of the respiratory chain, by using the measured and theoretical H⁺/O values, the amount of H⁺ translocated by the respiratory chain when two electrons (e^–) reduce oxygen (1/2O₂) was estimated (2e^– + 2H⁺ + 1/2O₂ → H₂O) (Table 2). Based on the measured and theoretical H⁺/O values, translocated H⁺ by the respiratory chain is predicted to higher under low- aeration conditions than under high-aeration conditions in E. clarkii DSM 8720^T. This prediction could be attributed to the larger ΔΨ in the high-aeration condition than that in the low-aeration condition, as described in the oxygen consumption rate section. Although the amount of translocated H⁺, based on the measured H⁺/O values, was lower in E. clarkii DSM 8720^T [0.84 (min^–1⋅mg protein^–1)] than in S. cohnii YN-2000 (1.04) the translocated amounts of H⁺ using theoretical H⁺/O values were almost the same (E. clarkii DSM 8720^T, 2.28; S. cohnii YN-2000, 2.4). The number of translocated H⁺ per 1/2O₂ consumption in E. clarkii DSM 8720^T is lower than that in B. subtilis IAM 1026 (6.0, high aeration) on the basis of the theoretical value. Although the above comparison used the theoretical values, it is necessary to reconsider the reality of the measured H⁺/O ratio in consideration of the actual contribution of the NPS, as described above.

If the measured values of H⁺/O reflect the actual value of the number of H⁺, it could mean that a very small number of H⁺ were translocated by the respiratory chain compared to the theoretical values. In addition, the number of translocated H⁺ in alkaliphiles was much lower than that in neutralophilic B. subtilis. The results described above are considered to indicate the physiological strategy of maintaining H⁺ on the outer surface of the membrane as much as possible and operating the cell system in an environment lacking sufficient H⁺ for ATP production and high ΔΨ as a result of the physiological environmental adaptation at high pH. Therefore, although these estimated H⁺/O values may be unrealistic, it is expected that the respiratory H⁺ translocation rate of alkaliphilic bacteria is probably lower than that of neutral bacteria.

To understand the effect of H⁺ translocated by the respiratory chain under high aeration conditions, the maximum ATP synthesis rates of obligate (i.e., E. clarkii DSM 8720^T) and facultative (i.e., S. cohnii YN-2000) alkaliphiles and neutralophile (i.e., B. subtilis IAM1026) were compared on the basis of the data described in Table 2. The ATP synthesis rates of E. clarkii DSM 8720^T were higher than those of S. cohnii YN-2000 and B. subtilis IAM 1026. This suggests that the intensity of H⁺, which is translocated by the respiratory chain to drive ATP synthesis through the ATPase, is higher than that of other bacteria, while the number of H⁺ retaining on the outer surface of the membrane is larger than in other bacteria under low aeration. This H⁺ accumulation is attributed to the H⁺-bond network producing cytochrome c. In addition, since E. clarkii DSM 8720^T growth rate in low aeration conditions exhibits equivalent to S. cohnii in high aeration conditions and much higher than B. subtilis in low aeration conditions (Table 2), the H⁺ potential in the vicinity of the outer surface of the membrane is expected to be much higher than that measured in the bulk base. This efficient H⁺ usage may also be attributed to the H⁺ capacitor strategy on the outer surface of the membrane (Matsuno et al., 2018). The increase of cytochrome c-550, which possesses a specific amino acid sequence between the main body protein and membrane-anchoring part facilitates the formation of an H⁺ capacitor via the H-bond network. The H⁺ capacitor strategy may also influence to the maximum ATP production rate per H⁺ translocation by the respiratory chain. The predicted H⁺ utilization strategy for ATP production in high and low aeration conditions in E. clarkii DSM 8720^T is illustrated in Figure 4. The maximum ATP production rate per H⁺ translocated by the respiratory chain of E. clarkii DSM 8720^T under high-aeration conditions was 11.5 nmol ATP ⋅ translocated H^+–1 (theoretical base), which is much higher than the 4.0 nmol ATP observed in S. cohnii YN-2000 and the neutralophilic B. subtilis IAM 1026 (0.33 nmol ATP ⋅ translocated H^+–1). Thus, H⁺ efficiency under high-aeration conditions in E. clarkii is much higher than that in the S. cohnii and B. subtilis. This is attributed to the much higher ΔΨ in E. clarkii than that in the other bacteria.

H⁺ extrusion in the bulk phase by the respiratory chain in E. clarkii K24-1U was observed to be very slow (Yoshimune et al., 2010). It appears that the H⁺ extrusion into the bulk phase occurs after H⁺ retention at sites in the vicinity of the outer surface of the membrane. Therefore, H⁺, which is extruded by the respiratory chain, does not appear to be directly extruded into the bulk phase. However, the H⁺, extruded by the respiratory chain the retained on the outer surface of the membrane is released into the bulk phase by the addition of valinomycin or ETH-157, which disrupts ΔΨ (Yoshimune et al., 2010). Therefore, ΔΨ prevents the release of H⁺, which is translocated by the respiratory chain into the bulk phase. The rate of H⁺ translocation in the presence of valinomycin in E. clarkii K24-1U was approximately 70% that in B. subtilis under the same experimental conditions. This suggests that the intrinsic H⁺ translocation frequency was reduced by 70% in the alkaliphile as a strategy of reducing H⁺ utilization frequency. Accumulation of H⁺ at the outer surface of the membrane can also be demonstrated by addition of monensin, which is a Na⁺/H⁺ exchange reagent (Yoshimune et al., 2010). Monensin translocates H⁺, previously located to the extracellular surface of the membrane, into the intracellular space, and Na⁺ is located to the intracellular surface of the membrane. The transfer of H⁺, which is extruded by the respiratory chain into the bulk phase is further retarded by the introduction of monensin, which translocates H⁺ located extracellular surface, into intracellular space. These results indicate that ΔΨ influences to the retardation of H⁺ at the outer membrane surface in alkaliphilic Bacillaceae.

The much higher per H⁺ efficiency for ATP production in E. clarkii DSM8720^T than in B. subtilis IAM 1026 may be attributed to the combination of the H⁺ capacitor effect by cytochrome c and high ΔΨ and a high ΔΨ in the vicinity of the outer surface of the membrane. Although the contribution of the H⁺ capacitor effect is unknown, it is difficult to account for the difference in H⁺ efficiency for ATP production between E. clarkii DSM8720^T and B. subtilis IAM 1026 based on the difference in the estimated ΔΨ values (Table 2). This may be attributed to either or both the reasons described below. Although ΔΨ was estimated by substitutions of fluorescent changes in the reaction of respiration, it was calibrated by the diffusion potential produced by the different concentrations of K⁺ between intra- and extra-cellular vesicles plus valinomycin. One possible reason for this is that the diffusion potential based on valinomycin plus K⁺ was not reflected by the real ΔΨ for H⁺ across the membrane. It has been reported that the artificially imposed diffusion potential based on K⁺ has an inferior effect compared to the equal intensity of ΔΨ produced by the respiratory chain in A. pseudofirmus OF4 (Guffanti and Krulwich, 1994). Another reason is that the estimated ΔΨ is different from the real values in the vicinity of the outer surface of the membrane. It seems reasonable to assume that the force attributed to ΔΨ for H⁺ would be larger for H⁺ located closer to the outer surface of the membrane than H⁺ located in the extracellular bulk phase.

Negative ion capacity was estimated in the cell extracts containing inside-out membrane vesicles of obligate alkaliphilic E. clarkii DSM 8720^T and facultative alkaliphilic S. cohnii YN-2000 grown at pH 10 and neutralophilic B. subtilis IAM 1026 grown at pH 7 under high-aeration conditions (Goto et al., 2016). The negative ion capacity of intracellular contents in the two strains of alkaliphiles increased with measured pH between pH 6 and pH 8, whereas that in B. subtilis IAM 1026 barely changes between pH 6 and pH 10. The negative ion capacity of E. clarkii DSM 8720^T and S. cohnii YN-2000 was 4.7-times and 4.1-times higher than that of B. subtilis IAM 1026 based on each corresponding intracellular pH. The intracellular materials obtained by two alkaliphiles grown under low-aeration conditions exhibited approximately 72% intensity of negative ion capacity than those under high-aeration conditions at the corresponding intracellular pH (Goto et al., 2016). The intensities of the intracellular negative-ion capacity and the corresponding measured ΔΨ were correlated. Therefore, it is considered that the intracellular negative-ion capacity influences to ΔΨ, which is generated across the membrane. However, it is unclear whether the estimated negative ion capacity includes membrane-permeable ions. The high negative ion capacity in the alkaliphilic strains is probably attributed to the intracellular acidic substances, including proteins, at higher intracellular pH (ca. pH 8–8.2) than in the ordinary neutrophilic strains (pH 6–7). Protein pI distribution estimated using the genomes of alkaliphilic, neutralophilic, and acidophilic strains indicates a preference for acidic proteins (pI 4.01–5) in alkaliphiles (Lebre and Cowan, 2020).

Electron transfer-coupled H⁺ transfer in horse heart cytochrome c was examined by SAMs cytochrome c attached to an Ag electrode produced at different distances connected to the different chain lengths (C₆–C₁₆) of ω-carboxyl alkanethiols (Murgida and Hildebrandt, 2001). Only in the case where the distance between cytochrome c and the electrode is short (C₂), a difference in the electron transfer rate of the H⁺/D⁺ effect is observed: 132 s^–1 in H₂O₂ and 33 s^–1 in D₂O. H⁺/D⁺ effect does not occur in cytochrome c in the solution. Therefore, these results suggest that the prerequisites for electron transfer coupled with H⁺ transfer are (i) cytochrome c forms accumulated, such as a monolayer on the electrode, (ii) accumulated cytochrome probably should be lined up in the correct orientation, and (iii) cytochrome c is affected by Coulomb force from the electrode, which may correspond to ΔΨ on the membrane surface. Considering the high ATP productivity in alkaliphilic Bacillaceae, it can be assumed that the actual ΔΨ in the vicinity of the outer surface of the membrane is much higher than the estimated values. In addition, as described above, the redox potential of cytochrome c-550 immobilized on the gold electrode using SAMs was much lower than that in the redox titration. This suggests that Coulomb force from the electrode induces conformational changes around heme c, altering the redox potential (Murgida and Hildebrandt, 2004). Owing to this change in the redox potential of cytochrome c, the difference in the redox potential between cytochrome c-550 and the direct electron-accepting prosthetic group in cytochrome c oxidase (i.e., heme a) becomes larger. This means that a very large redox potential gap [+ 7 mV → ca + 250 mV (e.g., heme a)] may be necessary to transfer electrons from cytochrome c, which is located on the outer surface of the membrane, to the inner prosthetic component (e.g., heme a), which is located on the inner membrane under high ΔΨ and consists of strong intracellular negative charges. It is possible that this ΔΨ-dependent redox potential change mechanism regulates electron transfer in the respiratory chain depending on the intensity of the electric field on the outer surface of the membrane of E. clarkii.

The above-described are alkaliphilic Bacillaceae strains that utilize outer-surface-membrane-based ΔΨ consisting of the Donnan potential, which is attributed to the high intracellular pH and the intracellular acidic substances, including proteins. In accordance with the large ΔΨ, the respiratory chain is modified to match the harsh conditions for electron flow from the extracellular side to the intracellular side and translocation of H⁺ from the intracellular side to the extracellular side. The combination of enhanced ΔΨ and specifically modified respiratory components in obligately alkaliphilic E. clarkii, and influenced the enhancement of respiratory extruded H⁺ protection for ATP production. This system will make a limited amount of H⁺ present in the harsh environment even more effective for utilization of H⁺.