With the exception of coenzyme Q10 deficiency , there is currently no established treatment for patients with RC defects. Pharmacological stimulation of the biogenesis of the RC is a promising new approach for the treatment of these disorders. Two compounds have been shown to be effective to this purpose. BZF, which has been successfully employed in patients with defects of mitochondrial beta-oxidation, was shown to be effective also in a mouse model of COX deficiency due to COX10 mutations , and in cells of patients with COX deficiency due to COX10 mutation and with unknown defects. However a recent study could not confirm the efficacy of BZF in mice with defects in SCO2
COX15, or SURF1, and it was found to cause severe hepatomegaly in both control and COX-deficient animals, whereas AICAR, another activator of the PPARgamma/PGC-1alpha pathway, could rescue the phenotype in these animals .
Yet, these negative results concerning the efficacy of BZF should be taken with caution for two reasons. Mice (and rodents in general) are not a good model to study BZF. In fact it has been shown that there are marked species-specific differences in the effects of BZF, which is known to cause hepatomegaly in mice and rats, but not in other mammals such as dogs and monkeys  (or humans). The biological bases of this phenomenon are not clear. In fact, BZF is commonly used in patients and elevation of liver enzymes is an infrequent side effect in clinical use . Moreover, our results show that the therapeutic window for BZF is very narrow (see below), therefore dosage is a critical issue when administering BZF to animals (or patients), and excessive doses may result in lack of response. A possible explanation for the observed lack of effectiveness of BZF in the COX deficient mice models, could be that the plasma levels of the drug fell outside the narrow therapeutic range (animals were administered BZF in drinking water and serum levels were not monitored). As pointed out by Djouadi and Bastin mice received a dose of BZF 100-fold the therapeutic dose used in patients .
Although the effects of AICAR in COX deficient mice were striking, this compound is still not approved for medical use, while BZF and copper are routinely used in patients. We have therefore examined the effectiveness of BZF (with or without copper) in human cells with SCO2 deficiency, to assess the potential use of these drugs in patients with SCO2 mutations. Our data support the efficacy of the combined treatment with these two compounds. Although in principle we cannot rule out the hypothesis that the effect we observed occurs only in cultured cells, and not in whole tissues, nevertheless patients with beta oxidation defects receiving BZF therapy displayed an increase in COX activity in their muscles . Our results highlight some important points that should be taken into account when designing experiments in non rodent animal models, or therapeutic trials in patients.
First, the response of COX activity to BZF treatment was similar (an increase of about 40% compared to baseline levels) in all cell lines we studied (primary fibroblasts, HeLa, HEK293, and SEM cells), and the same relative increase was also detected in COX-deficient cells (both in SCO2 and SURF1 mutants), indicating that BZF does not increase efficiency of the assembly process, but simply stimulates COX biogenesis as a whole. The increase of COX activity after BZF treatment was accompanied by increase of cellular ATP production in SCO2 cells.
However, the precise mechanism of action of BZF is unclear . In fact we detected an increase in enzymatic activities of complex I, III, and IV, whereas other mitochondrial proteins, both matrix enzymes such as CS and OAT, or localized to the mitochondrial inner membrane, such as complex II and SCO2 protein itself, were unaffected by BZF treatment, ruling out an effect of BZF on mitochondrial biogenesis as a whole. Coenzyme Q10 levels in deficient cells were similarly unaffected. BZF appears to be stimulating only RC complexes which contain mitochondrial DNA encoded subunits, and not mitochondrial biogenesis as a whole. These findings argue against an involvement of the PGC1alpha pathway, but the exact mechanism of action remains to be elucidated. The apparently different mechanism of action of AICAR and BZF suggest that these two compound could also be used in synergy. Future work will be aimed to test this hypothesis.
Second, we noted that the effect of BZF on COX activity peaks at a BZF concentration of 400 μM in the culture medium. However, higher BZF concentrations (600 μM) do not result in a plateau of the effect, but we observed instead a relative reduction of COX activity. A slight decrease of COX activity with 500 μM BZF was in fact noted also by Bastin and coworkers  and they did not test higher BZF doses. These results are critical for therapeutic trials, because they imply that plasma BZF levels must be closely monitored in patients, since the therapeutic window for this compound appears to be very narrow, and doses too high or too low will be ineffective. The reason for the observed decrease in efficacy at higher BZF concentrations is not clear. We observed a minor increase in susceptibility to apoptosis in cells treated with 600 μM BZF, but only after incubation with H2O2. Some sort of negative feedback mechanism could be acting, but further work is needed to address this issue.
Third, and most important, we demonstrated a synergistic effect between copper and BZF. The effect of BZF alone is relatively modest, but we had shown previously that CuCl2 can rescue COX activity in SCO2 mutant cells in a dose-dependent manner with complete recovery of COX activity at 200 μM CuCl2 . Preliminary trials in patients have also yielded promising results, although one major drawback is the toxicity of copper . Our present data demonstrate that CuCl2 and BZF have an additive effect, which permits to employ lower doses of each compound and still achieve complete normalization of COX activity in patient’s cells. It will be possible to avoid at least part of the toxicity related to high serum copper levels, while employing BZF at doses lower than the peak effective levels of 400 μM.