Introduction

As a rule, the loss of iron-sulfur (FeS) clusters is very damaging to the function of FeS proteins. Nitric oxide (NO) radicals provide a prominent pathway for disruption of iron-sulfur clusters, and lead to the formation of dinitrosyl-iron complexes (DNICs). The spectroscopic properties of such DNIC motifs are discussed in Chapter 2, and the formation of protein-bound DNIC is the subject of Chapters 4-6. Usually, the transformation of FeS clusters to DNICs is highly disruptive. However, in the case of the SoxR protein, this transformation converts the protein into a potent activator of gene transcription. SoxR is a homodimer protein containing a pair of [2Fe—2S] clusters, and it becomes activated when cells are exposed to superoxide or NO. Activation of this protein stimulates the transcription of the soxS gene as a first step in a cascade to activate regulon promoters. The activity of SoxR is regulated through the state of its [2Fe—2S] clusters. Two distinct regulatory mechanisms have been recognized: First, regulation via the redox state since the protein is transcriptionally active only in oxidized state. Second, via the reversible assembly-disassembly of the [2Fe—2S] clusters. Interestingly, the reconstruction of the clusters seems to be a tightly controlled process in living cells: Although the DNICs formed in purified SoxR by treatment with pure NO gas are quite stable in vitro, the SoxR DNICs formed in intact cells are very rapidly replaced by normal reduced [2Fe—2S] centers, with a consequent deactivation

* Author for Correspondence. E-mail: [email protected]

of SoxR. This rapid turnover is not exclusive to SoxR. DNICs formed in ferredoxin are also rapidly turned over and replaced by unmodified metal centers. In vitro, this turnover can be mimicked by treatment with L-cysteine (which non-enzymatically removes the DNICs) and the IscS cysteine desulfurase, which supplies inorganic sulfide for reconstructing the [2Fe-2S] centers. The active repair of DNICs appears to be a general process that helps counteract the toxicity of NO exposure. The reversibility of the transformation FeS — DNIC also makes it well suited to a role in signal transduction.

Reactive molecules are generated routinely in biological systems, both actively as cellular products with defined roles, and inadvertently as the hazardous by-products of normal metabolism. Oxygen radicals are endogenously formed in both ways [1]: they are the unavoidable by-products of aerobic metabolism, through autooxidation reactions of electron transport components and other cellular molecules; oxygen radicals are also generated actively in large amounts during the inflammatory response of immune cells such as macrophages and neutrophils. Oxygen radicals actively produced at lower levels are involved in signaling pathways, for example in response to some growth factors [2]. Another physiological example is NO, which at high levels contributes to the cytotoxic action of activated macrophages [3], but at much lower levels is employed for intercellular signaling and perhaps other regulatory purposes [4,5]. NO is also a metabolic intermediate for denitrifying bacteria; disruption of regulation in this process can cause large amounts of the NO intermediate to accumulate, in some circumstances enough to inhibit the growth of the NO-generating bacteria or their neighbors [6].

Radical species cause broad cellular damage. Superoxide (O2-), the primary product of autooxidation reactions and of the NADPH-consuming oxidases, reacts directly with only a few types of molecules, notably protein FeS centers [7]. Such reactions often inactivate enzymes. Aconitases [8] are a well-studied example, and have been discussed in Chapter 6. Superoxide is rapidly converted to H2O2 by superoxide dismutase (SOD); H2O2 can then react with reduced transition metals such as iron or copper to generate hydroxyl radical, which can damage all classes of macromolecules [7]. NO also reacts with FeS centers [9,10], as well as with heme [11], in both cases often leading to protein inactivation. Additional NO-dependent damage in cells derives from downstream reaction products of NO with O2, O2- or thiols [12]. The resulting reactive species can damage DNA, proteins, or membranes. Thus, all the major cellular components are at risk for damage from free radicals.

In the face of this threat, complex cellular systems have evolved to offset the toxic threat of oxidative or NO damage [13,14]. SOD and catalase remove O2- and H2O2, respectively, while small compounds such as glutathione and a-tocopherol neutralize radicals. Protein FeS centers are repaired or replaced [15], and DNA lesions are corrected by complex enzymatic systems [16]. Collectively, these defenses offset the threat of mutations or cell death that could result from the accumulation of free radical damage.

Another level of protection comes from the regulated expression of cellular defense systems [17]. The response to oxidative stress coordinates the regulation of numerous genes and proteins to reduce the formation of radicals, neutralize them more effectively, protect macromolecules from damage, and repair lesions when they do occur [13]. For the systems in which the biochemistry of signal transduction has been clarified, it is often the case that the triggering reaction (e.g. oxidation or nitrosylation of FeS centers) is one that usually inactivates protein function. In the oxidative stress sensors, these damaging reactions have been exploited for purposes of signaling [18].

In this chapter, we will concentrate on a particular type of radical signaling by NO, namely the transformation of the FeS centers in SoxR to protein-bound DNICs. Such transformations have been detected in many different iron-sulfur proteins and usually inhibit the enzymatic activity of the affected protein. SoxR is unusual in this respect as the transformation activates the enzyme. This property is crucial for the role of SoxR as an oxidative stress regulator [19]. In this protein, the reversible formation of DNIC provides an important mechanism for intracellular signaling.

DNICs AND NITRIC OXIDE TOXICITY

The first indication for the NO-mediated modification of FeS proteins in vivo came from electron paramagnetic resonance (EPR) spectroscopy studies of Clostridium botulinum [20]. Cultured C. botulinum cells show a prominent EPR absorption near gav = 1.94 from intact FeS proteins. Upon exposure to the combination of sodium nitrite and a reductant such as ascorbate, the EPR signal at gav = 1.94 was lost and replaced with a new signal at gav = 2.04. The g-factor and lineshape indicated that the FeS clusters were converted to protein-bound DNIC. The conversion was caused by the NO radicals released by the reduction of nitrite. The same EPR signal at gav = 2.04 was later observed in lipopolysaccharide-activated macrophages [21-24] and in cultured hepatocytes and macrophages treated with exogenous NO [25]. More recently, Pieper et al. [26] reported that formation of protein-bound DNIC in post-operative day-4 allografts. The formation of DNIC coincided with a decrease of intact FeS clusters, and indicated that FeS proteins are the target of NO cytotoxicity in acute cardiac allograft rejection. More details of this phenomenon can be found in Chapter 19 of this book. Collectively, these studies confirm that pathophysiological levels of NO can transform FeS clusters to protein-bound DNIC in cells and tissues.

In E. coli cells alone, thus far over 180 different FeS proteins have been identified [27]. To explore whether all these FeS proteins may be modified by NO in cells, we (H. Ding laboratory) treated E. coli cells with NO using the Silastic tubing NO delivery system at a rate of ~50nM NO per second, as described by Tannenbaum's group [28]. This NO release rate is comparable to that under pathophysiological conditions [28,29]. Cell extracts prepared from the NO-treated E. coli cells were then fractionated using gel filtration chromatography (Superdex 200). Each eluted fraction was then analyzed by EPR (Fig. 1A). The amplitude of the EPR signal at gav = 2.04 of the protein-bound DNIC and the protein concentration in each fraction were plotted together as a function of the fraction numbers (Fig. 1B). This analysis showed that protein-bound DNICs were present in almost all eluted fractions, indicating that the DNICs were distributed across a broad range of protein sizes in NO-treated E. coli cells.

SoxR AS A SENSOR OF OXIDATIVE STRESS OR NITRIC OXIDE

The SoxR protein is the master regulator of a response to oxidative stress triggered by redox-cycling agents such as paraquat, which generate large amounts of O2- in cells. SoxR itself

Fig. 1. Diversity of DNIC proteins recovered from NO-treated E. coli. A cell extract from bacteria treated with NO was fractionated by gel filtration chromatography. Individual fractions were analyzed by EPR spectroscopy for the gav = 2.04 signal characteristic of DNIC [numbered traces shown in (Panel A)]. Panel B shows the 280-nm absorbance and EPR signal amplitude in the gel filtration column fractions.

Fig. 1. Diversity of DNIC proteins recovered from NO-treated E. coli. A cell extract from bacteria treated with NO was fractionated by gel filtration chromatography. Individual fractions were analyzed by EPR spectroscopy for the gav = 2.04 signal characteristic of DNIC [numbered traces shown in (Panel A)]. Panel B shows the 280-nm absorbance and EPR signal amplitude in the gel filtration column fractions.

appears to control only one promoter in E. coli, that of the soxS gene, which is positioned head-to-head with the soxR gene in E. coli [30,31] and Salmonella [32]. The elevated levels of SoxS protein resulting from SoxR activation stimulate expression of the many genes of the E. coli soxRS defense system through specific binding of the target promoters and recruitment of RNA polymerase [33]. The activation of the soxRS regulon genes enhances bacterial resistance against redox-cycling agents such as paraquat (but not to simple oxidants such as H2O2), as well as to a broad range of antibiotics and to organic solvents [17]. The oxidant resistance phenotype involves many functions, including some with obvious antioxidant roles (e.g. SOD) or functions in cellular repair (e.g. the DNA repair enzyme endonuclease IV). In contrast, the increased antibiotic resistance phenotype was quite unexpected [34]. This SoxR-mediated resistance against antibiotics and organic solvents involved the upregulation of the efflux pumps in the inner membrane [35] and modified the expression of porins in the outer membrane [36]. These effects show that the range of metabolic and biochemical functions affected by the soxRS system is very broad, and suggests possible evolutionary origins of the system in response to various environmental exposures [37].

The soxRS system was discovered and characterized initially by studies of the response to redox-cycling agents, which generate increased levels of superoxide in the cell [38]. An initial survey of various oxidative and other agents was consistent with a specific response to O2- or superoxide-generating agents [39], and even revealed a previously unsuspected potent redox-cycling activity of a DNA-damaging carcinogen [40]. For various reasons, exposure to NO (pure gas, delivered through gas-permeable tubing) caused significant activation of soxRS [41]. Notably, this activation in response to NO was independent of oxygen, and was even somewhat more efficient under anaerobic than aerobic conditions [41]. NO donors such as "NONOate" were also able to activate SoxR in vivo [42]. Activation of the soxRS system by NO improved the chance of survival of phagocytosed E. coli in activated murine macrophages [41]. A direct measurement of soxS promoter activity in phagocytosed bacteria indicated about 30-fold induction 8h after their initial uptake by the macrophages. This effect was dependent on a functional soxR gene and abrogated by incubation of the bacteria-containing macrophages with the an NO synthase inhibitor [43]. In contrast, soxRS did not contribute significantly to the survival of Salmonella in activated macrophages [44], although it should be noted in this context that S. enterica has evolved to reside in the macrophage phagosome, unlike E. coli.

The activation of SoxR by NO was somewhat surprising. Previous studies had established that SoxR could be activated by one-electron oxidation of the protein's [2Fe—2S] centers. This redox activation had been monitored by EPR in intact bacteria, as well as with purified SoxR [45-49]. In vitro and in vivo studies indicated that the transcriptional activity was dependent on intact [2Fe—2S] centers in the protein [50,51]. Against this background, it seemed unlikely that NO could activate SoxR by formation of the FeS centers into DNICs. However, a detailed study [52] showed that this was indeed the case: an EPR signal characteristic of SoxR-bound DNIC was formed in intact E. coli treated with NO. The intensity of this DNIC signal correlated perfectly with the transcriptional activation of SoxR. Furthermore, treatment of purified SoxR yielded a DNIC-containing protein that could be repurified and which exhibited potent transcriptional activity [52]. An interesting observation during these studies was the contrast between the stability of the SoxR DNIC in vitro, and its rapid disappearance in vivo after termination of the NO exposure [52]. The experiments show that cells actively promote this turnover.

REPAIR OF PROTEIN DNICs

To search for the cellular factors responsible for removing and replacing the protein-bound DNIC, ferredoxin DNICs were prepared from either NO-treated E. coli or by treating ferre-doxin [2Fe—2S] directly with NO in vitro [53]. EPR spectroscopy was used to measure the ferredoxin DNIC [54]. When the ferredoxin DNIC (Fig. 2, trace a) was incubated g = 2.04

Was this article helpful?

0 0

Post a comment