Sulfiredoxin Specifically Catalyzes the Deglutathionylation of 2-Cys Peroxiredoxin

Abstract

Glutathionylation은 단백질의 cysteine 잔기와 glutathione 사이에 일어나는 disulfide 결합으로 세포내 reactive oxygen specie의 증가나 그로 인한redox 상태의 변화에 의해 일어날수 있다. Glutathionylation되어진 단백질들은 세포 증식, 분화, 세포주기 진행, 전사 활성도, 신진대사 등에 영향을 줄 수 있기 때문에 세포내 redox 조절의 일반적인 mechanism으로 여겨지고 있다. Glutathionylation은 또한 높은 수준의 oxidative stress 상황에서 단백질의 thiol 잔기가 비가역적 변화를 겪어 단백질이 처하는 위험으로부터 보호할 수 있다. 단백질의 cycteine 잔기가 일단 glutathionylation이 되면 순간적으로 활성을 잃게되면서 sulfinic이나 sulfonic acid와 같은 비가역적 형태로 전환되는 것을 막아주기 때문이다. Peroxiredoxin(Prx)은 thiol-containing protein들로부터 전자를 받아 hydrogen peroxide와 alkyl hydroperoxide를 물과 alcohol로 환원시키는 peroxidase이다. 그 중 Prx I은 세포내 양이 많고 여러 종에 보편적으로 존재하는 isoform으로써 4개의 cysteine을 가지고 있다. N-terminal에 존재하는 conserved cysteine은 peroxidase 반응에서 peroxide를 공격하면서 sulfenic acid가 되고 이것은 곧 또 다른 subunit의 C-terminal cysteine과 반응하여 subunit간 disulfide 결합을 이루게 된다. 그러나 active site cysteine은 촉매과정중 sulfinic acid 형태로 hyperoxidation이 될수 있고 이때에는 peroxidase 활성을 잃게 된다. Sulfiredoxin (Srx)은 이러한 sulfinic acid 형태의 Prx를 환원시키는 효소로 알려졌고 최근에는 glutathionylation을 다시 환원시키는 기능도 있다고 보고되었지만 그 mechanism은 아직 밝혀지지 않았다. 첫번째 부분에서는 Prx I 단백질의 glutathionylation을 단계적인 double alkylation과 질량 분석으로 연구하였는데, Prx I의 세 개 cysteine잔기에 glutathionylation이 될수 있음을 보였으며 그 cysteine은 52, 83, 그리고173 잔기임을 밝혔다. 두번째 부분에서는 Grx I 또는 Srx에 의해 촉진되는 Prx I의 deglutathionylation에 관한 연구가 수행되었다. Glutathionylation의 중요한 특성은 세포내 redox환경이 복원되었을때 그것이 가역적으로 효소에 의해 환원되어질 수 있다는 것이다. Grx I 또는 Srx에 의해 매개되는 Prx I의 각 cysteine의 deglutathionylation을 측정하기 위하여 한 곳의 cysteine에만 glutathionylation이 될수 있는 Prx I의 double cysteine mutant를 만들었다. 결과적으로, Cys 52의 deglutathionylation은 Srx보다 Grx I에 더 의존적이었지만, Cys 83과 Cys 173의 deglutathionylation은 Grx I이 전혀 관여하지 않았고 오직 Srx에 의존성을 보였다. 추가적으로, 세번째 부분에서는 LC-MS 분석을 통하여 Srx가 Prx I의 deglutathionylation 반응중에 자신의 cysteine 잔기를 통한glutathionylated intermediate를 형성함을 보였다. 또한 서로 결합하지 않는 mutant를 이용한 deglutathionylation 실험을 통해 Srx가 Prx I과의 직접 결합이 있을때만 deglutathionylation을 촉진할수 있음을 보였다. 세포내에서 일어나는 Prx I의 glutathionylation과 그에 미치는 Srx의 영향을 보기 위하여 네번째 부분에서는 BioGEE (Biotinylated glutathione ethyl ester)를 도입하였다. BioGEE는 세포안으로 들어가 세포내 esterase에 의해 ethyl ester가 잘려지고서 단백질의 cysteine에 glutathionylation을 시킬수 있다. 세포에 BioGEE를 처리한후 glutathionylation되어진 단백질을 streptavidin으로 침전시켜 확인할수 있다. BioGEE 방법을 이용하여A549 세포에서 H2O2를 처리한 후 Prx I의 glutathionylation이 증가한 것을 확인하였다. 또한, RNA interference로 세포내 Srx를 고갈시킨 경우에는 Prx I의 glutathionylation이 현저히 증가함을 확인하였고 Srx를 overexpression 시킨 경우에는 Prx I의 glutathionylation이 줄어든 것을 관찰하였다. 그러나, 대조적으로Srx와 결합하지 않는 Prx V의 경우에는 Srx의 발현 양이 변하더라도 glutathionylation에 차이를 보이지 않았다. 마지막 부분에서는 Prx I glutathionylation이 oligomerization과 chaperone activity에 미치는 영향을 보기 위하여, 환원 상태와 glutathionylation 상태의 Prx I을 가지고 oligomerization과 chaperone activity를 측정하였다. 2-Cys Prx의 경우 oxidative stress 조건에서 molecular chaperone기능으로의 전환을 겪는다는 것이 보고 되었고, low molecular weight 종에서 high molecular weight complex로의 구조적 전환이 그것의 기능적 전환과 연관이 있음이 알려졌다. Glutathionylated Prx I으로 수행된 analytical ultracentrifugation 실험 결과, glutathionylation이 Prx I의 decameric 구조를 깨뜨리고 주로 dimer와 monomer 형태의 구조로 변화시켰음을 보였다. 또한 glutathionylated Prx I은 reduced Prx I과 비교해서 단백질의 thermal aggregation을 막아주는 능력으로 표시하는 chaperone activity가 저하되었음을 보였다. 본 연구에서는, Prx I 단백질의 glutathionylation과 Srx에 의해 조절되는 deglutathionylation의 생화학적 분석이 수행되었다. Prx I이 Cys 52, 83, 그리고 173 residue에 glutathionylation 될 수 있음을 보였고 각 cysteine의 deglutathionylation 분석 결과, Cys 52의 deglutathionylation만이 Srx보다Grx I에 의해 영향을 받았고 나머지 Cys 83과 Cys 173의 deglutathionylation에는 Srx 만이 작용하였다. 또한, Srx depletion은 H₂O₂를 처리한 A549 세포에서 Prx I의 glutathionylation을 증가시켰지만, Prx V에는 영향이 없었다. Glutathionylation은 또한 Prx I의 native decameric 구조를 dimer-monomer위주의 구조로 변환시켰으며 chaperone activity를 저해하였다.;Protein glutathionylation involves the formation of a mixed disulfide bond between glutathione (GSH) and protein cysteine residue. Reversible protein glutathionylation constitutes a redox-mediated regulatory mechanism. It plays a key role in cellular regulation and cell signaling, because glutathionylated proteins are involved in various physiological processes such as growth, differentiation, cell cycle progression and metabolism. Glutathionylation can also protect protein thiol from hyperoxidation to form its irreversible oxidatively modified derivatives, since glutathionylated thiol cannot undergo further oxidation while glutathionylated protein can be recovered by deglutathionylation reaction. Peroxiredoxins (Prxs) are a family of peroxidases that reduce hydrogen peroxide and alkyl hydroperoxides to water and alcohol, respectively, with the use of reducing equivalents provided by physiological thiols such as thioredoxin. Peroxiredoxin I (Prx I), the abundant and ubiquitously expressed member of the Prxs family, contains four cysteine residues. A conserved cysteine residue at the NH2-terminal region is critical for its peroxidase activity. During the peroxidase reaction, peroxidatic cysteine attacks the peroxide substrate and is oxidized to a cysteine sulfenic acid (Cys-SOH), which then reacts with the COOH-terminal conserved Cys-SH of the second subunit of the dimer to form an intersubunit disulfide bond. The active site cysteine of Prx, however, can be hyperoxidized to form sulfinic acid (Cys-SO₂H) during catalysis, which leads to the loss of its peroxidase activity. It has been previously demonstrated that sulfiredoxin (Srx) is responsible for the reduction of the sulfinic form of 2-Cys Prxs in the presence of ATP and Mg^(2+). In addition, recently Srx also has been shown to function as a deglutathionylating enzyme for a number of glutathionylated proteins such as actin and protein tyrosine phosphatase 1B. However, the mechanism of deglutathionylation catalyzed by Srx had not been elucidated. In the first part of this study, purified human Prx I was glutathionylated by oxidized glutathione. The glutathionylation sites were identified using a stepwise double alkylation method coupled with mass spectrometric analysis. The result revealed that three of the four cysteine residues in Prx I can be glutathionylated in vitro. These three cysteine residues were identified as Cys^(52), Cys^(83), and Cys^(173). In the second part, deglutathionylation of Prx I catalyzed by Srx or Grx I was examined. To avoid complexity, double cystein mutants were used such that the deglutathionylation of each cysteine by Srx or Grx I could be individually investigated. The results revealed that Srx deglutathionylated all three glutathionylated cysteine residues with equal catalytic efficiency, and comparatively Grx I was a better enzyme relative to Srx to deglutathionylate Cys^(52) but failed to deglutathionylate Cys83 and Cys^(173). In the third part, I demonstrated that in a deglutathionylation reaction Srx formed a glutathionylated intermediate through its single cysteine residue. The transfer of the GSH moiety from Prx I to Srx was confirmed using LC-MS analysis. In addition, using site-directed mutagenesis coupled with binding affinity and deglutathionylation activities showed that Pro^(174) and Pro^(179) of Prx I and Tyr^(92) of Srx were essential for both activities. These results indicate that Srx is a specific enzyme to deglutathionylate Prx I since they possess specific interacting sites for their complex formation. This notion was further confirmed with the findings showing that relative to Grx I, Srx exhibited negligible deglutathionylating activity for CSSG and BSA-SSG. In the fourth part, to investigate whether the observed glutathionylation of Prx I and its deglutathionylation by Srx is physiologically relevant, the BioGEE (biotinylated glutathione ethyl ester) method was used to monitor the glutathionylated Prx I in human lung carcinoma A549 cells. BioGEE is cell-permeable and once inside the cells, its ethyl ester is cleaved by cellular esterases and the biotin-labeled glutathionylated proteins can be isolated with streptavidin agarose. With this method, I showed that Prx I was glutathionylated in vivo in response to hydrogen peroxide treatment. However, when cellular Srx was depleted using the siRNA technique, it led to a substantial increase in glutathionylated Prx I, while the reverse was observed in Srx overexpressing cells. In contrast, glutathionylation of Prx V was not affected by the change of Srx expression level. In the fifth part of this study, the effects of Prx I glutathionylation on its oligomeric status and the chaperone activity were investigated. Prx I is known to exist in various oligomeric forms. It has been claimed that 2-Cys Prxs can undergo a functional change from peroxidase to molecular chaperone under oxidative stress. The structural change from low molecular weight species to high molecular weght complexes is known to be associated with the gain of chaperone activity. In this study analytical ultracentrifugation methods were used to monitor the oligomeric status of Prx I. The results revealed that glutathionylation totally destroyed the native decameric structure of reduced Prx I. In addition, using wild type and cysteine double mutants of Prx I (C52/173S), I have shown that glutathionylated Prx I inhibited chaperone activity of Prx I, consistent with the notions that glutathionylation of Prx I prevents its oligomerization to form decamers or higher multimers, and chaperone activity is associated with the decameric or higher multimeric structures of Prx I. In summary, based on biochemical, biophysical and molecular biological analysis of glutathionylation and deglutathionylation of human Prx I, I revealed that (i) Prx I can be glutathionylated on three out of four of its cysteine residues, which found to be Cys^(52), Cys^(83), and Cys^(173). (ii) Kinetic data are consistent with the notion that Cys83 and Cys^(173) are specifically deglutathionylated by Srx, with glutathionylated Srx as intermediate, while Grx I is more specific for deglutathionylating Cys^(52). (iii) Site-directed mutagenesis coupled with binding and enzymatic activities indicate that Srx is specific for deglutathionylating Prx I, a typical 2-Cys Prx, and not Prx V which does not bind to Srx. Thus, the observed specificity of Srx is derived from its favorable binding affinity to Prx I. (iv) Glutathionylation shifted Prx I from its decameric structure to a population consisting mainly of dimers, and concomittently inhibited its chaperone activity. (v) These observations are clearly biologically relevant since the state of Prx I glutathionylation is correlated with the cellular levels of Srx in A549 cells. Together, my study reveals that reversible protein glutathionylation mediated by Srx plays a key role in regulating the peroxidase and chaperone activity of Prx I.