Acidithiobacillus ferrooxidans ATCC 23270^T, L. ferrooxidans DSM 2705^T, and L. ferriphilum DSM 14647^T were cultured autotrophically on soluble ferrous ions at 30°C in the medium described elsewhere (Tuovinen and Kelly, 1973), adjusted to 158 mM FeSO₄⋅7H₂O and pH 1.5. Acidihalobacter ferrooxidans 14175, Sulfobacillus thermosulfidooxidans 9293^T, Alicyclobacillus ferrooxydans 22381^T, Acidimicrobium ferrooxidans 10331^T, Ferrimicrobium acidiphilum, 19497^T, Acidiplasma aeolicum 18409^T, Ferroplasma acidiphilum 12658^T, Metallosphaera sedula 5348^T, and Sulfolobus metallicus 6482^T were all obtained from the DSM. Each organism was cultured on 20 mM ferrous sulfate at pH 1.5 using the relevant mixotrophic medium and growth temperature recommended for each microorganism in the DSM media guide. Cells grown to late stationary phase were harvested by centrifugation, washed twice with 0.02 M H₂SO₄, and resuspended in sufficient 0.02 M H₂SO₄ to achieve a stock suspension of approximately 1 × 10¹⁰ cells/ml. Each stock suspension was stored at 4°C for no longer than a week while spectroscopic experiments were conducted on aliquots of the cells.

Absolute numbers of microorganisms were determined by electrical impedance measurements in a Multisizer 4 particle counter (Beckman Coulter, Inc., Brea, CA, USA) fitted with a 30-μm aperture (Blake and Griff, 2012; Li et al., 2015). The instrument was programmed to siphon 50 μl of sample that contained Isoton II as the electrolyte. The current applied across the aperture was 600 μA. Voltage spikes attendant with impedance changes as microorganisms passed through the aperture were monitored with an instrument gain of four.

Absorbance measurements on intact cells in suspension were conducted in an OLIS CLARiTY 1000A spectrophotometer (On Line Instrument Systems, Inc., Bogart, GA, USA) as described previously (Blake and Griff, 2012; Li et al., 2015). In a typical experiment, identical 8-ml solutions of 0.02 M sulfuric acid, pH 1.5, were added to both the sample and reference observation cavities of the spectrophotometer. A volume was withdrawn from the sample cavity and replaced with an equal volume of suspended cells. The contents of both observation cavities were maintained at the respective growth temperature of each organism using a model TC-1 Peltier temperature control element from Quantum Northwest (Liberty Lake, WA, USA). After recording a stable baseline, 40 μl of a 200 mM solution of ferrous sulfate, pH 1.5, were added to the sample observation chamber to create a 1.0 mM solution in reduced iron. This concentration of the ferrous ion electron donor was always in excess to the electron accepting capacity of the molecular oxygen electron acceptor in the reaction mixtures, which ranged from 236 to 150 μM in the mixtures at temperatures from 30 to 65°C, respectively. Raw absorbance spectra were subsequently collected at a rate of 6.2/s for several minutes. Iron-dependent absorbance changes in the suspensions of cells were complete in the time that it took to add the electron donor, close the chamber, and initiate the data collection. The resulting absorbance changes were stable for the entire time that raw data were collected. These raw absorbance values were subsequently converted to equivalent absorbance values per cm using Fry’s method (Fry et al., 2010) with analysis software provided by OLIS, Inc.

The principal features of the ICAM employed herein to conduct absorbance measurements on intact cells in turbid suspensions were described earlier (Blake and Griff, 2012; Li et al., 2015). This novel design for a spectrophotometer imposes two consequences. First, the multiple transversals around the interior of the observation cavity means that the incident exciting light experiences a much longer effective path length than it would in an equivalent linear spectrophotometer where the transmitted light experiences only one pass through a filled cuvette. Consequently, absorbance measurements in this ICAM result in a much greater sensitivity than would equivalent measurements in a standard linear spectrophotometer. Second, to the extent that the exciting incident light can be made to be totally diffuse, there are no longer deleterious consequences to absorbance measurements from turbid suspensions that scatter light. If the exciting light is already randomly scattered to a maximum extent, then additional light scattering by the turbid sample will have no immediate consequences on the integrity of the absorbance measurement. Thus it is possible to conduct absorbance measurements in situ in whole cells under physiological solution conditions.

The Proteobacteria and the Nitrospirae are the two phyla of Gram-negative eubacteria that contain obligately acidophilic members that respire aerobically on soluble iron (Bonnefoy and Holmes, 2012). These bacteria are distributed among at least four genera in the Proteobacteria: Acidithiobacillus, Acidiferrobacter, Acidihalobacter and Ferrovum. There are four species recognized within the genus Acidithiobacillus that oxidize iron, while there are at least two species within the genus Acidihalobacter that also do so. We chose Ah. ferrooxidans and At. ferrooxidans to represent the Proteobacteria phylum in the in situ spectroscopic studies reported herein.

Figure 1 shows the reduced minus oxidized difference spectrum that was observed when intact cells of air-oxidized Ah. ferrooxidans were exposed to excess ferrous ions at 35°C and pH 1.5. This difference spectrum was different from that obtained previously with At. ferrooxidans (Li et al., 2015) in a number of ways. First, the difference spectrum in Figure 1 exhibited no evidence of typical reduced cytochrome c-dependent absorbance changes around 417 or 550 nm. In contrast, the participation of c-type cytochromes in the comparable difference spectrum obtained previously with At. ferrooxidans was indicated by prominent peaks at 417, 520, and 551 nm. Second, the reduced peak at 565 nm in Figure 1 was clearly in the spectral region normally attributed to the reduced peaks of b-type cytochromes, a feature that was absent in the spectrum obtained using intact At. ferrooxidans. Third, the first α peak encountered in the difference spectrum of Ah. ferrooxidans had a reduced peak at 608 nm, some 10 nm red-shifted from the equivalent peak seen in At. ferrooxidans at 598 nm. Finally, there was no evidence of a broad trough of negative absorbance in the spectrum in Figure 1 that one could attribute to the iron-dependent reduction of a rusticyanin-like molecule. In contrast, iron-reduced intact At. ferrooxidans exhibited a broad trough of negative absorbance in the difference spectrum from 500 to 650 nm which was consistent with the hypothesis that concentrated amounts of the blue copper protein rusticyanin were reduced by soluble iron.

Spectral properties similar to those shown in Figure 1 have been reported for terminal oxidases expressed by two other Gram-negative eubacteria: Thermus thermophilus, a thermophile that grows at 70°C (Zimmerman et al., 1988); and Rhodothermus marinus, a thermohalophilic bacterium (Verissimo et al., 2007). The positions of the reduced α and β peaks varied between these two bacteria from 600 to 613 nm and from 577 to 562 nm, respectively. These absorbance properties were attributed to a cytochrome ba₃ terminal oxidase expressed by these two extremophilic eubacteria. A working hypothesis is that the absorbance properties shown in Figure 1 represent a cytochrome ba₃-type terminal oxidase that participates in the aerobic iron respiratory chain of the obligately halophilic and acidophilic Ah. ferrooxidans. In any case, it is evident that At. and Ah. ferrooxidans utilize different prominent redox-active prosthetic groups and biomolecules for respiratory electron transfer when they respire aerobically on soluble iron.

Leptospirillum is the only genus within the Nitrospirae that is known to contain acidophilic members that respire aerobically on iron. The reduced minus oxidized difference spectra contrasted above were quite different from that observed when intact cells of L. ferrooxidans were mixed with excess ferrous ions at 30°C and pH 1.5 (Blake and Griff, 2012).

The difference spectrum obtained with L. ferrooxidans had a reduced α peak at the unusual position of 579 nm and represents a unique iron-responsive cytochrome that has not been reported for members of the Proteobacteria (Singer et al., 2008). Spectral changes attributable to blue copper proteins or typical cytochromes a, b, or c have not been reported in the genus Leptospirillum and were not visible in the ICAM spectra reported earlier (Blake and Griff, 2012). A working hypothesis is that members of the Nitrospirae and the Proteobacteria express and utilize different redox-active biomolecules to conduct aerobic respiration on iron.

The Firmicutes and the Actinobacteria are the two phyla of Gram-positive eubacteria that contain obligatory acidophilic members that respire aerobically on soluble iron (Bonnefoy and Holmes, 2012). These acidophilic bacteria are distributed among at least three genera in the Firmicutes: Sulfobacillus, Alicyclobacillus, and Acidibacillus. Sulfobacillus and Alicyclobacillus contain at least five and four separate species, respectively, that respire on iron. We chose S. thermosulfidooxidans and Alb. ferrooxydans to represent the Firmicutes phylum in the in situ spectroscopic studies reported herein.

Figure 2 shows the reduced minus oxidized difference spectra that were observed when intact cells of S. thermosulfidooxidans and Alb. ferrooxydans were mixed with excess ferrous ions at pH 1.5 and 50 and 30°C, respectively. Although the two difference spectra differed by several nanometers in both their reduced Soret and α peaks, we judged the two spectra to be sufficiently similar so as to represent the same type of heme prosthetic group embedded in slightly different protein environments. Because the positions of both reduced α peaks were greater than 600 nm, our hypothesis is that the spectra in Figure 2 represent the respective terminal oxidases in the aerobic iron respiratory chains of these two Firmicutes.

Acidophilic bacteria that respire aerobically on soluble iron are distributed among as least five genera in the phylum Actinobacteria: Acidimicrobium, Ferrimicrobium, Ferrithrix, Acidithrix, and perhaps Acidithiomicrobium. We chose Am. ferrooxidans and Fm. acidiphilum to represent the Actinobacteria phylum in the in situ spectroscopic studies reported herein. Figure 3 shows the reduced minus oxidized difference spectra that were observed when intact cells of Am. ferrooxidans and Fm. acidiphilum were mixed with excess ferrous ions at pH 1.5 and 45° and 32°C, respectively. As was the case with the two spectra shown in Figure 2, the two difference spectra shown in Figure 3 differed by only a few nanometers in both their reduced Soret and α peaks. Once again, we judged the two spectra in Figure 3 to be sufficiently similar so as to represent the same type of heme prosthetic group embedded in slightly different protein environments.

All four of the difference spectra shown in Figures 2 and 3 are highly similar. The wavelengths of maximum absorbance of the reduced Sorets and α peaks in all four difference spectra support the hypothesis that each spectrum represents an a-type heme that is part of the terminal oxidase in its respective microorganism. The slight differences among the four spectra are presumed to be due to subtle structural differences among the respective globins that bind and function using the same a-type porphyrins. Thus Gram-positive eubacteria that respire aerobically on iron do so using basically the same principal redox-active prosthetic group in their respective terminal oxidases, regardless of the phylum or genus into which the bacterium is assigned on the basis of the sequence of its 16S ribosomal RNA. We posit that the Gram-positive eubacteria express a unique type of electron transfer pathway and strategy to accomplish aerobic respiration on soluble iron.

The Crenarchaeota and Euryarchaeota are the two phyla of Archaea that contain obligately acidiphilic members that respire aerobically on soluble iron (Bonnefoy and Holmes, 2012). These Archaea are distributed among two genera in the acidophilic Euryarchaeota: Acidiplasma and Ferroplasma. Acidiplasma and Ferroplasma contain at least two and three separate species, respectively, that respire aerobically on soluble iron. We chose Ap. aeolicum and Fp. acidiphilum to represent the Crenarchaeota phylum in the in situ spectroscopic studies reported herein.

Figure 4 shows the reduced minus oxidized difference spectra that were observed when intact cells of Ap. aeolicum and Fp. acidiphilum were mixed with excess ferrous ions at pH 1.5 and 40° and 35°C, respectively. The absorbance of the reduced α peak observed with Ap. aeolicum had a maximum value at 583 nm, but there was also a discernible shoulder at longer wavelengths. The corresponding α peak observed with iron-reduced Fp. acidiphilum exhibited two distinct peaks of absorbance, a more intense peak at 583 nm and a less intense peak at 594 nm. Others have reported the existence of an a₅₈₃aa₃-type of terminal oxidase in S. tokodaii, a member of the phylum Euryarchaeota (Iwasaki et al., 1995; Schafer, 1996; Schafer et al., 1996). The a₅₈₃ component in the latter organism is also represented as heme A_S, an a-type heme with a formyl group on ring 1 and a hydroxyethylgeranylgeranyl side chain on ring 2 of the a-heme frame (Lubben and Morand, 1994). The accompanying aa₃ component of the terminal oxidase in S. tokodaii had a reduced α peak at 603 nm. Because the existence of a reduced α peak at 583 nm is rare, we hypothesize that the terminal oxidases expressed by these two members of the Crenarchaeota represent a novel a₅₈₃aa₃-type of terminal oxidase where the aa₃ component differs slightly from that expressed in S. tokodaii.

Acidophilic bacteria that respire aerobically on soluble iron are distributed among four genera in the phylum Euryarchaeota: Metallosphaera, Sulfolobus, Acidianus, and Sulfurococcus. We chose M. sedula and S. metallicus to represent the Euryarchaeota phylum in the in situ spectroscopic studies reported herein. Figure 5 shows the reduced minus oxidized difference spectra that were observed when intact cells of each Archaea were mixed with excess ferrous ions at pH 1.5 and 60°C. Although the two difference spectra differed by several nanometers in all three peaks, we judged the two spectra to be sufficiently similar so as to represent the same prosthetic groups embedded in slightly different protein environments.

The first observation was that the reduced Soret and α peaks at around 422/423 and 578/579 nm, respectively, were remarkably similar to those reported for L. ferrooxidans (Singer et al., 2008; Blake and Griff, 2012). Thus one can hypothesize that the same heme prosthetic group is expressed and utilized for aerobic respiration on soluble iron by members of both the eubacterial Nitrospirae and the archaeal Euryarchaeota phyla. Interestingly, there is no evidence for genes in M. sedula 5348^T, M. cuprina Ar-4, or S. metallicus 6482^T that encode a protein similar to the cytochrome 579 that is expressed in L. ferrooxidans 2705^T.

The second observation was that no spectral evidence was obtained for the participation of an a₅₈₃aa₃-type of terminal oxidase in either of these two representatives of Euryarchaeota, despite its existence in S. tokodaii (another member of the Euryarchaeota) and the participation of this latter terminal oxidase during respiration in the Crenarchaeota. The third observation was that a novel spectral species with a reduced peak at around 485 nm was observed when both Euryarchaeota were mixed with an excess of soluble ferrous ions. No evidence for a similar spectral peak was evident from the data reported for Leptospirillum. We know of no redox-active prosthetic group that exhibits a reduced peak in the vicinity of 485 nm. We hypothesize that this unexpected spectral intermediate represents a heretofore unknown and uncharacterized prosthetic group that is expressed and exploited by these Euryarchaeota as they respire aerobically on soluble iron.