Response on "commentary on "using resonance synchronous spectroscopy to characterize the reactivity and electrophilicity of biologically relevant sulfane sulfur". Evidence that the methodology is inadequate because it only measures unspecific light scattering". The evidence is incorrect.

Sulfane sulfur, including HSnH and RSnH, n ≥ 2; RSnR, n ≥ 3, contains zero-valent sulfur (S0). It is a newly discovered cellular component with important physiological functions, including redox homeostasis maintenance and signaling [1]. Due to the diversity of sulfane sulfur species, their chemical properties are still largely unknown. We recently discovered that biologically relevant sulfane sulfur species display strong optical signals when analyzed by resonance synchronous spectroscopy (RS2), in which excitation and emission wavelengths are essentially identical [2]. We reported that several sulfane sulfur species, including inorganic polysulfide (H2Sn, HSn−, and Sn2−), glutathioine persulfide (GSSH), protein persulfide, and organic polysulfide (RSnH, n ≥ 2 and RSnR, n ≥ 3) have RS2 signals, which are affected by pH if the sulfane sulfur species undergo protonation and deprotonation [2].

After our publication, Cuevasanta et al. published a commentary, claiming that RS2 does not measure soluble sulfane sulfur but elemental sulfur particles derived from soluble sulfane sulfur [3]. However, they only did two inappropriate experiments without any quantification, leading to a wrong conclusion that our method “only measures unspecific light scattering”.

For the first experiment, they showed colloidal sulfur, prepared by vortexing sulfur powder into water, also displayed RS2 signals, suggesting that our reported RS2 of sulfane sulfur is due to light scattering of sulfur particles [3]. We performed a similar experiment, diluting inorganic polysulfide (26 mM stock in an alkaline solution [4]) to 1.5 μM in 50 mM Tris buffer (pH 7.4) for RS2 analysis. We then prepared colloidal sulfur by vortexing sulfur powder in the same buffer [3]. The suspension was allowed to settle for 1 hour, and the supernatant was diluted with equal volume of the same buffer before RS2 analysis. The data are presented as R2S2 in which the buffer's RS2 is corrected [5]. This correction is necessary, as Tris buffer has background RS2 signals (Fig. 1A in [2]). Cuevasanta et al. used water instead of a buffer and did not use R2S2 [3]. The R2S2 spectra are similar but different when compared via overlaying, as the colloidal sulfur spectrum is red-shifted (Fig. 1, black vs. blue). The RS2 signal of the polysulfide solution was unstable and mostly disappeared after 5 min (Fig. 1, black, red, green, and grey), while that of colloidal sulfur was stable, showing no reduction within 30 min (Fig. 2). Thus, the RS2 signals of polysulfide and colloidal sulfur are different.

Fig. 1

RSspectra of polysulfide and colloidal sulfur. Black and red curves: Polysulfide stock with sulfide in alkaline solution under anaerobic conditions was diluted to 1.5 μM in 50 mM Tris buffer (pH 7.4). Black, immediately; red, after 1 min; green, after 3 min, grey, after 5 min. Blue curve: Colloidal sulfur was prepared in the Tris buffer by vortexing. The colloidal sample was diluted with equal volume of the same buffer before RS2 analysis. Pink curve: Elemental sulfur was dissolved in acetone and diluted to 1.5 μM in the Tris buffer. R2S2 was obtained by correcting the RS2 signal of the buffer [5]. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

The RS2 of the colloidal sulfur solution (Fig. 1 legend) was stable within 30 min. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

We have showed the presence of elemental sulfur S8 in inorganic polysulfide at neutral pH [2]. Polysulfide stock is prepared in alkaline solutions with sulfide in access under anaerobic conditions [4], and it is mainly present as long chain polysulfide species [2]. When the stock was diluted to 1.5 μM in 50 mM Tris buffer (pH 7.4), S8 could form at the relative neutral pH [2]. To test whether S8 was mainly responsible for RS2, we dissolved elemental sulfur in acetone (15 mM) and diluted it to 1.5 μM in 50 mM Tris buffer (pH 7.4). The R2S2 signal of S8 was much weaker, about 4.6-fold lower than that of 1.5 μM inorganic polysulfide (Fig. 1, pink). Thus, when inorganic polysulfide is diluted in 50 mM Tris buffer (pH 7.4), the initial RS2 signal is mainly from the polysulfide. The rapid loss of the signal is likely due to the conversion to S8 or the oxidation by O2. The produced S8 should aggregate into fine particles similar to that of S8, obtained via diluting sulfur stock in acetone into the same buffer; both should display reduced RS2 signals likely because of scattering and the sulfane sulfur property of S8 (Fig. 1), as RS2 is often used to analyzed aggregates of dye molecules [5,6].

For the second experiment, Cuevasanta et al. used 1 mM H2O2 to oxidize 1 mM H2S and claimed that the reaction also produced the reported signal [3]. They showed that the obtained signal was from small particles via light scattering. However, they did not show how long it took to generate the signal and how much sulfur particles were generated from 1 mM H2S. We repeated their experiment and could not detect the signal within 30 min. This is likely due to the high concentrations of H2O2 used in their experiment. Since H2O2 reacts with H2S at a much slower rate (0.46 M−1s−1, the 2nd rate constant [2]) than with sulfane sulfur (23.76 M−1s−1 for GSSH reacting with H2O2 [2]), the produced polysulfide is not accumulated but rapidly oxidized by H2O2. The method cannot be used to prepare inorganic polysulfide, which is unstable even without H2O2 (Fig. 1). They misused our kinetic assay [3]. We used excess sulfide to react with 50 μM H2O2 and monitored polysulfide production by using RS2; we only used the data from the first 3 min to obtain the initial rate, from which the rate constant was calculated [2].

The R2S2 spectrum of GSSH at pH 6 (Fig. 2A of original paper [2]) has a maximum around 650 nm. In comparison, the R2S2 spectrum of colloidal sulfur in 50 mM phosphate buffer (pH 6) is essentially the same as that in 50 mM Tris buffer, pH 7.4, significantly different from that of GSSH [3]. As presented in Fig. 1C&D of our original paper [2], the R2S2 spectra of the commercially available Bis(methyl) trisulfide (CH3-SSS-CH3) and Bis[3-(triethoxysilyl)propyl] tetrasulfide are also different from that of colloidal sulfur. Thus, there is no evidence to suggest that these compounds decay to elemental sulfur during our assay. Cuevasanta et al. suggested that the reaction of these compounds with H2O2 or SSP4 (sulfane sulfur probe 4) as we tested is through colloidal sulfur without any supporting evidence [3]. We have not found any other reports suggesting that colloidal sulfur is an intermediate in these reactions.

RS2 is a data acquisition method by using a fluorometer. RS2 signals can be contributed by scattering, on-fluorescence, and possible Stokes’ shifted fluorescence [6]. Further, resonant Rayleigh scattering, caused by molecular polarity, may also contribute to RS2 [7,8]. The electrophilic property of sulfane sulfur could be polar when containing a thiosulfoxide bond [2]. We did not observe any fluorescence besides RS2 signals for all tested sulfane sulfur. However, a compound does not have to be fluorescent to give RS2 signals, as evidence by the RS2 spectrum of 50 mM Tris buffer (Fig. 1A in [2]). In short, the RS2 property of sulfane sulfur is unexpected, but is real.

Funding

The work was financially supported by a grant from the National Key Research and Development Program of China (2016YFA0601103).