The Incomplete Glutathione Puzzle: Just Guessing at Numbers and Figures?

Abstract

Significance: Glutathione metabolism is comparable to a jigsaw puzzle with too many pieces. It is supposed to comprise (i) the reduction of disulfides, hydroperoxides, sulfenic acids, and nitrosothiols, (ii) the detoxification of aldehydes, xenobiotics, and heavy metals, and (iii) the synthesis of eicosanoids, steroids, and iron-sulfur clusters. In addition, glutathione affects oxidative protein folding and redox signaling. Here, I try to provide an overview on the relevance of glutathione-dependent pathways with an emphasis on quantitative data. Recent Advances: Intracellular redox measurements reveal that the cytosol, the nucleus, and mitochondria contain very little glutathione disulfide and that oxidative challenges are rapidly counterbalanced. Genetic approaches suggest that iron metabolism is the centerpiece of the glutathione puzzle in yeast. Furthermore, recent biochemical studies provide novel insights on glutathione transport processes and uncoupling mechanisms.

Critical issues: Which parts of the glutathione puzzle are most relevant? Does this explain the high intracellular concentrations of reduced glutathione? How can iron-sulfur cluster biogenesis, oxidative protein folding, or redox signaling occur at high glutathione concentrations? Answers to these questions not only seem to depend on the organism, cell type, and subcellular compartment but also on different ideologies among researchers.

Future directions: A rational approach to compare the relevance of glutathione-dependent pathways is to combine genetic and quantitative kinetic data. However, there are still many missing pieces and too little is known about the compartment-specific repertoire and concentration of numerous metabolites, substrates, enzymes, and transporters as well as rate constants and enzyme kinetic patterns. Gathering this information might require the development of novel tools but is crucial to address potential kinetic competitions and to decipher uncoupling mechanisms to solve the glutathione puzzle. Antioxid. Redox Signal. 27, 1130-1161.

Keywords: compartmentalization; concentration; function; glutathione; kinetics; rate constant.

FIG. 1.

Electrophilic metabolites and substrates that are considered to play a role in glutathione metabolism. Interaction sites between the electrophile and the thiolate group of glutathione (or the glutathione-dependent enzyme) are indicated by arrowheads. Enzymes are listed in Table 1. See Ref. (60) for details on the enzymatic mechanisms.

FIG. 2.

Subcellular distribution and transport of glutathione. Pores, transporters, and established as well as hypothetical glutathione contents are highlighted. Enzymes and metabolic pathways are omitted for clarity. Unknown transporters are shaded in gray and suggested transport processes are highlighted by dotted arrows. Please note that the transporters might differ among eukaryotes and that the pathways are partially based on in vitro data. For example, the carrier proteins for the import of GSH into the mitochondrial matrix have been characterized in mammals but not in yeast and the import of S-d-lactoylglutathione has been demonstrated for isolated rat mitochondria. GSX, sum of glutathione conjugates, GSOH, GSNO, hemithioacetals, and so on; IMS, mitochondrial intermembrane space; 2-OG, 2-oxoglutarate; P_i, inorganic phosphate; SLG, S-d-lactoylglutathione. For transporters and details, see the Subcellular Distribution and Transport of Glutathione section.

FIG. 3.

Effect of variable substrate concentrations on GSH-dependent enzyme kinetics. In this example, the enzyme “E” uses a ping-pong mechanism to convert the electrophilic substrate “S.” The shown patterns are typical for Grx-catalyzed reductions of glutathionylated disulfide substrates, yielding a thiol and GSSG (GSH + RSSG→RSH + GSSG). Similar patterns are found for selected GPx or Prx that catalyze the GSH-dependent reduction of hydroperoxides. (A) The normalized reaction rates for variable concentrations of “S” at four different constant concentrations of GSH are plotted directly according to the Michaelis–Menten theory (left panel) and in double reciprocal form according to the Lineweaver–Burk theory (right panel). The data can be usually fitted using the Michaelis–Menten equation (with k_cat^app and K_m^app values replacing k_cat and K_m as shown in Eq. 2), yielding hyperbolic and linear curves in the direct and double reciprocal plot, respectively. The parallel lines in the Lineweaver–Burk plot are indicative for the ping-pong mechanism with the constant slope K_m^app/k_cat^app. (B) The apparent kinetic parameters at different GSH concentrations from panel A are analyzed in a (double reciprocal) secondary plot. Linear fits for enzymes with infinite k_cat and K_m value pass through the origin (enzyme A). The true K_m and k_cat value of other enzymes can be calculated from the x- and y-axis intercepts, respectively (enzyme B). The reciprocal slope of the linear fit corresponds to a second-order rate constant k₂, in this case for the reaction between the glutathionylated enzyme “F” and GSH.

FIG. 4.

Comparison of normalized hydroperoxidase activities at variable H₂O₂ concentrations. Normalized reaction rates at (A) high, (B) intermediate, and (C) low H₂O₂ concentrations are shown from left to right. Kinetic parameters for equine catalase (k = 3.5 × 10⁷ M⁻¹s⁻¹), bovine GPx1 at 4 mM GSH (k_cat^app = 3 × 10³ s⁻¹, K_m^app = 3 × 10⁻⁵ M), and recombinant peroxiredoxin AhpC from Salmonella typhimurium (k_cat^app = 52 s⁻¹, K_m^app = 1.4 × 10⁻⁶ M) are based on reported measurements by Ogura (206), Flohe et al. (89), and Parsonage et al. (220), respectively. The parameters for the modeled sensitive Prx isoform are k_cat^app = 100 s⁻¹ and K_m^app = 1 × 10⁻⁵ M.

FIG. 5.

Variable enzyme concentrations have drastic effects on reaction rates. Reaction rates for hypothetical hydroperoxidases A–D are shown from left to right at three different enzyme concentrations and variable H₂O₂ concentrations. The kinetic parameters of enzymes A–D are identical to the enzymes in Figure 4. The relevance of each enzyme is depicted below each plot. (A) Model with an identical concentration of 1 μM for all four proteins. (B) Model with enzyme concentrations differing by two orders of magnitude. (C) Model with enzyme concentrations differing by four orders of magnitude. Similar enzyme concentrations are found in Table 2.

FIG. 6.

Kinetic patterns depend on the enzyme and substrate. Lineweaver–Burk plots are shown for three different enzyme–substrate combinations. (A) Ping-pong patterns of ScGrx7 with the model substrate l-cysteine-glutathione disulfide (GSSCys) at three different GSH concentrations based on data from Ref. (22). The reciprocal slope (the k_cat^app/K_m^app value) remains constant because an increase of the k_cat^app value is coupled to an increase of the K_m^app value. (B) Sequential kinetic patterns of ScGrx7 with the model substrate HEDS at three different GSH concentrations based on data from Ref. (22). The reciprocal slope (the k_cat^app/K_m^app value) and the k_cat^app values vary, whereas the K_m^app values remain constant. (C) pH-dependent sequential kinetic patterns of PfGlo2 at four different concentrations of SLG based on data from Ref. (280). The reciprocal slope (the k_cat^app/K_m^app value) and the K_m^app values vary, whereas the k_cat^app values remain constant. HEDS, bis(2-hydroxyethyl)disulfide; SLG, S-d-lactoylglutathione.

FIG. 7.

The modular architecture of enzymes with ping-pong mechanisms facilitates the gain of novel functions. Two evolutionary scenarios are shown to illustrate how point mutations affect the geometric and electrostatic complementarity between the enzyme and substrate. For enzymes with ping-pong mechanisms, mutations can selectively alter the substrate preferences (scenario a) and/or kinetics (scenario b). This can result in trapped enzyme species and novel functions, for example, in redox signaling or iron–sulfur cluster biogenesis. The model may apply to all kinds of enzymes with ping-pong mechanisms such as Grx, Prx, GPx, GR, and TrxR [see Ref. (60) for details on the respective enzyme mechanisms]. Please note that the relevant enzyme–substrate interactions might be very temporary. Furthermore, they are not restricted to the reaction center and also include the protein surface that can affect the substrate recruitment.

FIG. 8.

The modular architecture of enzymes with ping-pong mechanisms allows a kinetic uncoupling of the half-reactions. Grx catalysis with glutathionylated substrates (GSSR) requires two distinct glutathione interaction sites because of the transition-state geometry of thiol/disulfide exchange reactions (22, 60, 183). The upper and lower parts illustrate differences and similarities between active and inactive Grx isoforms. Both groups have partially conserved residues for the interaction with the GSSR substrate at the so-called scaffold site. This interaction results in the glutathionylation of the active site cysteine residue. The scaffold site can also interact with glutathione that is bound to other groups (X) such as glutathione conjugates or iron–sulfur clusters (161, 164, 183, 302). Inactive Grx lack the hydroxyl group of a conserved tyrosine residue in a CxY(C/S)-motif and have a bulky WP-motif that probably interferes with the GSH interaction at a predominantly uncharacterized activator site (22, 60, 63, 183). Thus, the reductive half-reaction of inactive Grx is either very slow or cannot take place anymore because of an incompatible geometric and electrostatic complementarity between the enzyme and GSH (22, 60, 65).