Chemical Biology Approaches to Uncovering Nuclear ROS Control

Abstract

Heightened concentrations of reactive metabolites, including reactive oxygen species (ROS), can damage all macromolecules leading to the erosion of cellular fidelity. In this regard, the control of nuclear ROS levels is essential for cellular homeostasis, and dysregulation of nuclear ROS has been attributed to multiple pathologies and the mechanism of action of certain chemotherapies. How nuclear ROS is generated, detoxified and sensed is poorly understood, and stems in part, from a historical lack of tools that allow for its precise generation and detection. Here, we summarize the latest advances in chemical biology inspired approaches that have been developed to study nuclear ROS and highlight how these tools have led to major breakthroughs in understanding its regulation. The continued development and application of chemical biology approaches to understand nuclear ROS promises to unlock fundamental insights into human physiology and disease.

Introduction

Metabolic compartmentalization is an essential feature of eukaryotic cells, providing a mechanism for resource allocation, detoxification and control¹. The compartmentalization of reactive metabolites, and in particular reactive oxygen species (ROS) is necessary for maintaining cellular homeostasis because these oxygen-based metabolites can severely damage all biological macromolecules resulting in their inactivation and eventual cell death^2,3. Much of the pathology of ROS-based damage has been attributed to the function of these reactive metabolites in the nucleus, given this organelle’s role in directing cellular function and as repository for hereditary material that is passed down to subsequent generations. ROS-based damage of nucleic acids results in both a change in the genetic code (most notably through the generation of 8-oxo-7,8-Dihydroguanine, 8-OxoG) and the generation of single stranded DNA lesions that ultimately result in genotoxic stress and a blockade in cell growth⁴. Not surprisingly, high levels of nuclear ROS are thought to underly multiple human pathologies, like immune deficiency (chronic granulomatous disease⁵), neurodegeneration (cockayne syndrome⁶) and premature aging (Hutchinson-Gilford progeria^7,8). In addition to its role in pathophysiology, increasing nuclear ROS and the resultant DNA damage is attributed to the mode of action of numerous chemotherapies and radiotherapy^9–11. Despite its clear biological importance, surprisingly little is known about how nuclear ROS is generated, detoxified and sensed. Underlying this knowledge-gap has been the historical lack of tools to study nuclear ROS at a necessary level of precision and resolution. However, the last decade has seen a remarkable growth in a set of ROS-centric chemical biology approaches that enable the detection and generation of ROS. In this perspective, we highlight recent technological breakthroughs useful for studying nuclear ROS and the new biological concepts that are emerging their deployment.

Chemical genomic approaches to generate nuclear ROS.

The traditional approach to manipulate nuclear ROS levels involved use of millimolar concentrations of H₂O₂¹² or other ROS generators (e.g., paraquat¹³), leading to pleiotropic cellular effects that made it difficult, if not impossible, to discern how nuclear ROS specifically impacts cell biology. The recent development of designer ROS generating enzymes and their nuclear localization has resulted in the ability to produce a spectrum of ROS in a nucleus-specific manner.

Photosensitizers.

Photosensitizing enzymes take advantage of photochemistry which occurs when a fluorescent protein chromophore is excited and results in the generation of a highly reactive and short-lived singlet oxygen (¹O2) or superoxide (O2⁻)¹⁴. Photoinducible ROS was first discovered using green fluorescent protein (GFP)¹⁵ and multiple derivatives have been developed over the last several years including KillerRed¹⁶, KillerOrange¹⁷, SuperNova¹⁸, miniSOG^19,20 and TagRFP²¹, which broaden the spectrum of ROS that can be generated. Because these photosenstizors are genetically encoded they can be localized to the nucleus and specific regions of chromatin using traditional cell biological tags. Using the KillerRed photosensitizoer, Li and colleagues have uncovered novel mechanisms by which singlet oxygen generated at telomeres damages DNA^22,23. This has led them to discover that RNA methyltransferase TRDMT1 is recruited to sites of ROS-induced lesions and that methylation of RNA in DNA:RNA hybrids increases the affinity for RAD52 binding and induction of homologous recombination and DNA repair²⁴.

ROS-generating metabolic enzymes.

The generation of different ROS as product/byproduct of metabolic reactions is commonplace¹ and has been leveraged as a tool to study ROS biology. D-amino acid oxidases (DAAO) are peroxisome localized flavoproteins that convert D-amino acids to their imine form using FAD as an electron acceptor²⁵. During re-oxidation of FAD by O₂, H₂O₂ is generated as byproduct. Importantly, the levels of peroxide can be exquisitely controlled by manipulating the concentrations of D-amino acids. Much like photosensitizing enzymes, the location of DAAO can be easily directed with appropriate tags. By developing a cardiac specific DAAO, Michell and colleagues demonstrated that chronic cardiac generation of H₂O₂ induces dilated cardiomyopathy with significant systolic dysfunction, providing a powerful in vivo approach to dissect H₂O₂ signaling and damage²⁶.

Monitoring Nuclear ROS levels: small molecule and genetically encoded probes.

Many fluorescent probes have been developed to detect the intracellular H₂O₂, superoxide and hydroxyl measurement including Hydrocyanines, Xanthene and 2’,7’-dichlorodihydrofluorescein diacetate (H2DCFDA), respectively. While there has been much emphasis in measuring mitochondrial ROS levels by appending triphenylphosphine³⁰ to target small molecules to this compartment, most of these probes are not specific to a particular compartment and provide, with some accuracy, a holistic view of the redox potential within a cell. The recent development of nuclei-targetable H₂O₂ fluorescent probes³¹ and genetically encoded ROS reporters has been a game-changer in the study of nuclear ROS, enabling its real time dissection in vitro and in vivo.

Small molecule probes for nuclear H₂O₂.

Work by Chang and colleagues has focused on using boronate scaffolds which are responsive to H₂O₂. By taking advantage of SNAP-tag technology, they have successfully localized the cell permeable H₂O₂ probes to nucleus by conjugating them to nuclear-localized SNAP-tag fusion proteins³². Building on these initial findings, the Chang group developed the next-generation of nuclear targeted H₂O₂ probes (Nuclear Peroxy Emerald 1, NucPE1) with increased sensitivity for nuclear H₂O₂ levels that is compatible in both mammalian cell lines and C.elegans. Using these probes they find that over-expression of Sirtuin2 was sufficient to decrease nuclear H₂O₂ levels in C. elegans models of aging^33,34. Additionally, Yi and colleagues developed a ratiometric H₂O₂ fluorescent probe (NP1) based on a 1,8-naphthalimide scaffold that is targeted to the nucleus by linking a nuclear localization signal (NLS) to boric acid ester (pep-NP1), for nucleus-targeted H₂O₂ imaging³⁵.

Genetically encoded nuclear ROS reporters.

Much like their ROS-generating counterparts, genetically encoded fluorescent reporters have transformed the reactive metabolism field by enabling a new level of understanding of the heterogeneity and dynamics of compartmentalized ROS levels. Two families of H₂O₂ reporters are currently in wide use, reduction–oxidation sensitive GFP (roGFPs)-based H₂O₂ reporters and the Hydrogen Peroxide (HyPer) reporter a fusion of circularly permutated fluorescent protein (cpYFP) and the OxyR domain from the E. coli^14,36. roGFP and its analogs developed by Dick and colleauges were engineered by the introduction of two cysteine residues into the beta barrel surface of GFP enabling sensitivity to changes in redox level³⁷. The oxidation of the introduced cysteines within roGFP results in a conformational change and an alteration in GFP excitation from 400nm to 490 nm, providing a mechanism to distinguish the relative levels of oxidation versus reduction of the reporter³⁸. (2) The HyPer class of reporters, developed by Belousov and colleagues, leverages intramolecular disulfide bond formation within the DNA-binding domain OxyR which results in analogous excitation change in cpYFP upon oxidation³⁹. Although early versions of HyPer reporters were marked by sensitivity to intracellular pH, the latest iteration of the reporter (HyPer7), has overcome pH sensitivity providing an excellent dynamic range for measurement of peroxide levels in compartments throughout the cell⁴⁰. When paired with DAAO, the HyPer7 reporter has provided clear insights into how manipulation of mitochondrial H₂O₂ levels directly impact H₂O₂ levels in different organelles^41,42.

Nuclear ROS Generation and Sensing.

The utility of any tool is its applicability to discover new biology. Here we provide two examples where chemical biology tools specifically designed for the study of ROS were instrumental in uncovering how nuclear ROS is controlled.

Mitochondria ROS production controls nuclear H₂O₂ levels.

The source of nuclear ROS remains controversial. To date, few studies have pinpointed a nucleus-specific process that results in heightened ROS levels. Lysine demethylase 1 (LSD1) is a flavin-containing amine oxidases which catalyzes amine oxidation by oxidative cleavage of the α-carbon bond of the substrate to form an imine intermediate, which in turn, is hydrolyzed to form an aldehyde and amine via a nonenzymatic process. Recent studies have suggested that nuclear LSD1 demethylates histone H3 using FAD as a cofactor which upon reoxidation by O₂ leads to localized H₂O₂ and 8-OxoG generation^48,49. Additionally, NOX4 and MICAL1 are thought to contribute to nuclear H₂O₂ levels, albeit through poorly defined mechanisms^50,51. A leading hypothesis for the source of nuclear ROS is the mitochondria given its predominant role as a cellular ROS generator^52,53. Adding experimental validity to this hypothesis, Gillespie and colleagues found that during hypoxia a cellular stress known to increase mitochondrial ROS, mitochondria cluster around the nucleus providing the correct proximity for transfer of H₂O₂ to the nucleus. This finding was borne out of a clever combination of nuclear localized roGFP and disruption of mitochondrial trafficking, providing one mechanism by which mitochondrial positioning impacts nuclear ROS levels⁵⁴. In later studies, Riemer and colleagues demonstrated that disabling cytoplasmic peroxiredoxins PRDX1/2 was a requirement for a rise in nuclear H₂O₂ levels following heightened mitochondrial ROS production⁴¹ A key finding emerging from these studies is that for mitochondrial-based ROS to reach the nucleus there must be an inactivation of cytosolic antioxidant defense systems or the perinuclear mitochondrial localization, given the robust reductive capacity of the cytosol^41,42. More studies are need to understand the precise mechanisms by which ROS is transferred from the mitochondria through the cytoplasm to the nucleus.

Nuclear ROS sensing.

To maintain homeostasis, cells have evolved metabolic sensing pathways that are activated by changes in particular metabolites and regulate corresponding metabolic circuits to counter a systemic imbalance^55,56,57. Thus, it stands that nuclear ROS sensing pathways would be activated prior to the accrual of DNA damage and regulate pathways that directly contribute to the control of ROS. Recent studies have provided compelling evidence for this hypothesis, finding that bona-fide nuclear ROS sensors are embedded within canonical DNA-damage response (DDR) pathways. Accordingly, upon heightened ROS levels these sensors control pathways to squelch these ROS. Recently, Zhang et al. have discovered cysteine 408 in the DDR kinase CHK1 functions as a nuclear H₂O₂ sensor. When CHK1 is activated by nuclear ROS, it downregulates mitochondrial translation and corresponding nuclear H₂O₂ levels. By integrating cysteine-focused chemical proteomics and genome wide CRISPRi screening, they profiled the functional cysteine reactive targets of chemotherapies which are thought to mediate cytotoxicity by increasing ROS levels. This roadmap of chemotherapy ROS targets revealed both distinct and common protein targets of chemotherapy and revealed that CHK1•C408 was sensitive to nuclear H₂O₂ levels. By deploying a modified nitrosobenzoate based sulfinic acid probe (NO-BIO⁵⁸), they discovered that nuclear H₂O₂ oxidizes CHK1•C408 to its sulfinic acid form, resulting a conformational change whereby the interaction between the CHK1 autoinhibitory and kinase domains are disrupted, leading to kinase activation. Interestingly, disruption of CHK1 signaling by either a clinical-grade inhibitor (MK-8776^59,60) or genetic disruption, increased nuclear H₂O₂ levels, suggesting that CHK1 not only is modified by nuclear H₂O₂ but may control a cellular network that regulates nuclear ROS levels. By comparing CHK1 substrates to their corresponding sensitivity in the CRISPRi screen, mitochondrial DNA binding protein SSBP1 was identified as a compelling candidate. CHK1 phosphorylates SSBP1 at S67 in a H₂O₂ dependent manner which results in the relocation of SSBP1 from the mitochondria to the cytosol. The localization of SSBP1 is directly responsible for nuclear H₂O₂ levels. When phosphorylated by CHK1, SSBP1 accumulates in the cytosol and is rendered inactive resulting in a decrease in mitochondrial translation, which is necessary to mediate nuclear ROS levels through ETC complex I. In addition, the CHK1-SSBP1 circuit controls nuclear H₂O₂ levels in response to treatment with multiple chemotherapies including cisplatin and arsenic trioxide. Intriguingly, SSBP1 levels were downregulated in ovarian cancer tumors with decreased platinum free intervals and following the onset of chemoresistance. Further depleting SSBP1, mediated resistance to cisplatin in a panel of ovarian cancer models. These results suggest that SSBP1 may function as a biomarker for a tumors responsiveness to chemotherapies which result in heightened levels of nuclear ROS, providing an exciting example of nuclear-to-mitochondria communication through ROS-based signaling⁶¹.

Summary.

The exploration of nuclear ROS levels is still in its infancy and dissecting the regulation of this compartmentalized metabolite promises to unlock significant underlying physiology in both normal and diseases states. This exploration will benefit tremendously from tool development that allows the field to both control and monitor nuclear ROS levels. Looking into the future, as redox proteomic^62–64 analysis of nuclear proteins sheds light on additional ROS-modified targets, the development of corresponding antibodies against these sites will be critical not only for validation will serve as a critical tool to easily monitor compartmentalized ROS. We envision that these ROSdirected antibodies will accelerate research in this field akin to how the development of phosphoantibodies have revolutionized the study of kinase signaling.

Table 1.

ROS-generating enzymes which can be targeted to nucleus.

Enzymes	EX/Em	ROS type	Ref.
GFP	489/511	1O2	15
KillerRed	585/610	O2•−	16
KillerOrange	512/555	O2•−	17
SOPP	439/488(515)	1O2	27
SuperNova	579/610	1O2/ O2•−	18
SuperNova Green	440/510	O2•−	28
miniSOG	448/500	1O2	19,20
mKate2	588/633	1O2/ O2•−	29
TagRFP	555/584	1O2	21
DAAO	N/A	H2O2	25

Ex excitation wavelength, Em emission wavelength, ROS reactive oxygen species, Ref. references, GFP green fluorescent protein, SOPP singlet-oxygen photosensitizing protein, SOG single-oxygen generator, DAAO D-amino acid oxidase.

Table 2.

Probe and protein reporters for nuclear ROS measurement.

Probe/ Reporter	EX /Em (nm)	ROS type	Ref.
SNAP-PG1/PG2	488/500–550	H2O2	32
NucPE1	505/530	H2O2	33
pep-NP1	353/403 or 551	H2O2	35
roGFP2- Orp1	390 or 480/510–530	H2O2	43
roGFP2- Tsa2	405 or 488/510	H2O2	38
roGFP2- Prx1	405–415 or 474–490/510–550	H2O2	44
roGFP2- Tpx1	400 or 475–490/510	H2O2	45
HyPer	420 or 500/516	H2O2	39
HyPer2	420 or 497/516	H2O2	46
HyPer3	420 or 500/516	H2O2	47
HyPer7	400 or 490/516	H2O2	40

SNAP-PG SNAP-Peroxy-Green, NucPE1 Nuclear Peroxy Emerald 1, roGFP reduction–oxidation sensitive GFP, HyPer Hydrogen Peroxide reporter.