1.18 Supernumerary subunits
1.18.1 Bovine supernumerary subunit IV and yeast subunit
Subunit IV of bovine CcO is the largest supernumerary subunit (~17 kDa). The structure is superimposed on the homologous subunit 5 of yeast CcO shown in Figure 1-20. The hydrophilic C terminal domain projects into the IMS close to the cyt c binding site on subunit II, and a hydrophilic N terminal domain projects into the matrix [34]. The single membrane spanning α-helix is closely associated with core subunit I.
Figure 1-20. Subunit 5 of yeast CcO (red) is in the same structural location as subunit IV (blue) of bovine CcO.
Although the yeast 5A isoform [41] is overlaid with the bovine IV-2 isoform (PDB:1V54), the pairing of specific isoforms (5A with IV-1/IV-2 and 5B with IV-2/IV-1) is not possible based on their modest ~20 % sequence identities [26,41].
Subunit IV of bovine CcO has gained interest since it is thought to be a key regulatory subunit that binds ATP (allosteric inhibitor) or ADP (allosteric activator) [122,123], and is part of a negative feedback loop mechanism of respiratory control
53 [136-138]. Moreover, subunit IV has phosphorylation sites and S58 phosphorylation (only conserved in mammalian CcOs), is reported to modulate the respiratory activity by controlling ATP allosteric inhibition [139]. Subunit IV-2 expression is limited to lung, neurons and fetal muscle whilst IV-1 expression occurs across all tissue types [140,141]. Moreover, IV-2 expression is induced by hypoxic conditions under the control of RBPJ, CXXC5 and CHCHD2 transcription factors [142], or by toxins [143].
Yeast CcO is a useful model system to study this subunit, since its subunit 5 is homologous to mammalian subunit IV [26], and is the only yeast CcO supernumerary subunit to have isoforms (5A and 5B) [144,145]. These are encoded by single copy genes; COX5A on chromosome 14 and COX5B on chromosome 9. They share 67 % nucleotide (66 % amino acid) identity and are thought to have arisen from gene duplication followed by sequence divergence [145]. One or other polypeptide is essential for CcO assembly. As with the mammalian subunit IV isoforms, the expression pattern of subunits 5A and 5B is controlled by oxygen (and haem) concentration [146,147]. COX5A is expressed above 1 µM O2 and COX5B under low
(<1 µM O2) oxygen concentrations [148]. The O2 regulatory pathway of subunit 5
isoform expression is summarised in Figure 1-21. O2 stimulates haem synthesis that
then binds to and activates the transcription inducer Hap2/3/4/5 complex of COX5A; haem also activates Hap1 that induces the expression of ROX1, a repressor of COX5B expression. Under low oxygen levels, haem levels fall and Hap1 and Hap2/3/4/5 are no longer activated. This prevents COX5A transcription and loss of ROX1 transcription allows derepression of a set of hypoxic genes including COX5B [149,150].
Figure 1-21. The oxygen sensing and regulatory pathway of subunit 5A and 5B of yeast CcO.
COX5A is expressed at high oxygen concentrations (>1 µM) and COX5B is expressed under low oxygen concentrations <1 µM.
To enable COX5B to be expressed and subunit 5B to be assembled into CcO, under normal growth conditions, the ROX1 gene was mutated in a ΔCOX5A-deleted strain [147]. However, the CcO level was lowered by a factor of 2-3 compared to the control COX5A-expressing strain. Intriguingly, it was observed that the 5B isozyme had a turnover number that was 2-3 times faster than that of the 5A isozyme [28,151]. Thus subunit 5 isoforms were proposed to have an activity regulating role by allosterically altering the protein environment around haems a and a3 [28]. This effect on core
catalytic activity has been explored further in Results Chapter 7.
1.19 Aims
The aims of each of the five results chapters are listed here and are described in detail within each results chapter.
Results Chapter 3.
To associate redox-induced IR changes in the 1700-1000 cm-1 IR range in bovine CcO with specific metal centre transitions by exploiting the mixed valence (MV) states of CcO ligated with either cyanide (CN) or carbon monoxide (CO) [159,160][161].
Results Chapter 4.
To associate the redox-induced IR changes of functionally significant protonated carboxyl groups in the 1800-1700 cm-1 IR range in bovine CcO with specific redox groups by exploiting the MV states of CcO ligated with CN or CO.
COX5A
COX5B COX5B
High oxygen and haem levels (>1µM O
2)
ROX1
Low oxygen and haem levels (<1 µM O 2) ROX1 All subunits including COX5A Overview
55 Results Chapter 5.
To electrochemically induce a phenoxyl radical on Tyr-His model compounds designed to resemble the conserved feature in CcO and combine the optimal electrochemical conditions with ATR-FTIR spectroscopy to record reduced minus oxidised (radical) IR difference spectra. These spectra were used together with published spectra to tentatively assign tyrosine radical and Tyr-His associated IR bands in the PM (oxyferryl associated radical) minus oxidised IR difference spectrum of
bovine CcO.
Results Chapter 6.
To develop a protocol to reconstitute yeast WT CcO into liposomes using bovine CcO as a control. This would allow the proton translocation activity of yeast WT CcO to be quantitatively assayed. To then compare the proton/electron stoichiometry to that of the extreme 4H mutant (Q411L/Q413L/S458A/S455A). This would provide a means to assess the possible role of the H channel as a route for translocated protons in yeast mitochondrial CcO.
Results Chapter 7.
To investigate the origin of the increased core catalytic activity of 5B isozyme compared to 5A isozyme [147]. In order to do so, mutant strains were constructed by Brigitte Meunier following a BBSRC funded 3 week visit to her laboratory. The mutant strains expressed wholly subunit 5B, by replacing the COX5A open reading frame with COX5B, so that it was under the control of the COX5A promoter. This gene replacement was achieved at the COX5A nuclear locus or on a centromeric plasmid (1 copy per cell). Such mutants expressed the 5B isozyme to wild type (5A isozyme) levels under aerobic growth conditions, and without the complication of secondary effects caused by mutation of ROX1-encoded transcription factor. This allowed comparison of the catalytic properties of the 5A and 5B isozymes that have been expressed under identical conditions.