Reply to Giamogante et al.: The effect of low cyanide on O₂ consumption is best observed in physiological, rather than reductionist, systems

The core message of our recent report (1) is that nanomolar concentrations of cyanide increase O₂ consumption rate (OCR) and Complex IV (CcOX) activity and elevate adenosine triphosphate (ATP) levels in intact human cells (1). This effect—directionally opposite of the well-known inhibitory effect that cyanide exerts at higher concentrations—was demonstrated using several independent methods, including extracellular flux analysis and a fluorescent ATP:adenosine diphosphate biosensor (1). While preparing a review on cyanide’s cellular effects (2) we became aware of an earlier study, which, independently from us, also observed that nanomolar concentrations of cyanide increase OCR in rat brain endothelial cells (3).

In their letter, Giamogante et al. (4) present OCR measurements in isolated mouse liver mitochondria and detect no significant activating effect of 1 nM cyanide (although a hint for a ∼5% activating effect may be present, this is not statistically significant with the large standard generated from n = 2 to 4 replicates used). In our own studies we avoided using isolated mitochondria because such preparations often respond differently than intact cells (5, 6). When in our experiments we tested the effect of 0.1 nM cyanide in isolated bovine CcOX we also only observed a slight (∼10%) activation (1). Thus, the stimulating effect of low cyanide on OCR is best observable in intact cell-based systems, as opposed to reductionist models.

Giamogante et al. express concerns (4) about our findings (1) showing (using Western blotting and other methods) that cyanide catalyzes CcOX deglutathionylation. Previous work shows that CcOX undergoes glutathionylation (7) and glutathione inhibits CcOX activity (8). Since cyanide promotes disulfur cleavage (8), we hypothesized that a cyanide-induced reversion of S-glutathionylation may occur. Prior to conducting cell-based studies, we assessed the effect of cyanide in bovine serum albumin functionalized with 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) (BSA-TNB). The characteristic band of 330 nm (9) disappeared in response to addition of cyanide (Fig. 1A).

Fig. 1.

(A) Effect of 100 µM cyanide on BSA-TNB. A stock solution of BSA (27 µM) in sodium phosphate buffer (100 mM, pH 7.0) was mixed with 1 mM DTNB and incubated in the absence of light for 30 min. The solution was then washed in a PD-10 MiniTrap G-25 desalting column, in order to wash out TNB and unreacted DTNB, thus isolating BSA-TNB, characterized by a spectral “shoulder” at 330 nm. After incubation of the protein with 100 µM cyanide (or vehicle) for 1 h, the protein sample was loaded in a desalting column in order to wash out the TNB released after cyanide incubation and the spectral profile of the samples was acquired using a Tecan Infinite M200 Pro spectrophotometer. The 330-nm band disappears after cyanide treatment, suggesting that protein–TNB disulfur is cleaved in the presence of cyanide. These findings support the notion that cyanide may interfere with protein posttranslational modifications involving disulfur bonds, such as protein S-glutathionylation. (B) Effect of oxidized glutathione on the number of total free cysteines in bovine CcOX. CcOX from bovine heart (C5499, Sigma; 5 mg/mL, in 25 mM Tris⋅HCl buffer solution, pH 7.8, with 5 mM EDTA and 39 mM n-dodecyl β-d-maltoside) was incubated with 5 mM oxidized glutathione (glutathione disulfide, GSSG) (or vehicle) at room temperature for 2 h. Next, the samples were loaded on a desalting column to wash out the excess GSSG and further incubated with 0.1 nM KCN (or vehicle) for 1 h, followed by another washing step, to wash out KCN and the released reduced glutathione (GSH). Protein samples were then incubated with 1 mM DTNB for 30 min in the absence of light to quantify free cysteines as described (10). The results show that GSSG decreases the number of protein-free cysteines, and this is partly restored by cyanide. Thus, cyanide can promote the release of cysteine-bound glutathione. Data shown as mean ± SEM of n ≥ 3. **P < 0.01 indicates an increased number of free cysteines in response to glutathione (+GSSG vs. control, CTR); ^#P < 0.05 indicates restoration of free cysteines by cyanide (GSSG + KCN vs. GSSG).

In another preliminary experiment, we found that oxidized glutathione decreased the total free cysteines of CcOX; this was partially reversed by nanomolar cyanide (Fig. 1B). The mechanism of cyanide’s action on glutathionylated proteins likely involves nucleophilic displacement: The disulfide cysteine residue covalently bound to glutathione reacts with cyanide, yielding a free-cysteine residue and releasing CN-glutathione. This latter species may then be reconverted to cyanide in the cellular environment. Thus, the effect of cyanide may be catalytic. The reaction of cyanide with DTNB presented by Giamogante et al. (4) is not suitable to model such a reaction. Likewise, the rate constants between cyanide and the a3-CuB center of CcOX (4) are not relevant in the context of the proposed interaction.

Nevertheless, it is clear that the mechanism of the deglutathionylating effect of cyanide requires further characterization; it is also likely that the effects of cyanide on CcOX activity and on cellular O₂ consumption involve multiple biochemical mechanisms.

In conclusion, measurements of O₂ consumption in isolated mitochondria, or in vitro calculations of reaction rate constants in reductionist systems, do not invalidate the conclusion—made by at least two independent laboratories (1, 3)—that in intact cells nanomolar concentrations of cyanide increase OCR. This mediates physiological effects: stimulation of cell proliferation (1) and cytoprotection (3).