luorescent Proteins for Monitoring Intracellular Superoxide Production

Statement of Priority

M)Ol] This application claims priority to U.S. Provisional Patent Application Serial No.

[0/842,660, filed September 7, 2006 whose disclosure is hereby incorporated by reference herein. jvernment Interest

9002] The subject matter of this application was made with support from the United tates Government under Grant No. AR44657 from the National Institutes of Health. The United tates Government may retain certain rights.

ield of the Invention

)003] The present invention relates to methods of monitoring the real-time production

F superoxide in a cell or a compartment of a cell. The present invention also relates to modified oteins that are used to monitor the real-time superoxide production of a cell or a compartment ^? a cell.

ackground of the Invention

I] Reactive oxygen species (ROS) are produced by cells in response to stress and in

|e course of aerobic metabolism. ROS are capable of causing damage to almost all of the ilecular components of the cell, including lipids, fatty acids, amino acids, proteins and nucleic ids. Because of their ability to cause widespread damage, ROS are implicated in the :velopment of a variety of disorders including ischemia-reperfusion injury, neurodegeneration, sue inflammation, hypertension, atherosclerosis, diabetes and cancer. As changes in the llular redox state caused by ROS accompany such an eclectic assortment of different types of

an

the by of the need in

et al.,

of

Biol, with

DCFDA are non-ratiometric - meaning that ratios of emissions from different wavelengths cannot be compared - and exhibit substantial photobleaching and photocytoxicity. Further, neasuring the redox environment of cells with small molecule indicators is a labor intensive >rocess that typically requires that cells be harvested prior to obtaining readings. The time Selays and disruptions to the cell's environment that occur during cell harvesting make it difficult obtain an accurate reading of the in vivo redox environment, and make it impossible to onitor changes in the redox environment of a single cell over prolonged periods of time.

081 One solution to the problems associated with small molecule redox indicators has een to develop redox sensitive proteins. A green fluorescent protein (GFP) variant that is ensitive to the redox environment of cells is described in U.S. Patent Application Publication Fo. 2004/017112 to Remington, et al., which is hereby incorporated by reference herein, lthough the redox sensitive GFP proteins described by Remington are an advancement over the nail molecule based techniques described above, they have substantial disadvantages. One sadvantage is that the most significant signal changes indicated by the proteins described by emington are through a loss of signal during oxidation, making it difficult to distinguish ianges in redox environment because the signal to noise ratio is decreased. Further, the signal the redox sensitive proteins described by Remington develops over the course of minutes or nger, precluding the possibility of real-time monitoring and witnessing transient redox events.

)009J There remains the need in the art for redox sensing reagents that allow for facile al-time monitoring of the intracellular production of superoxide and other ROS.

»ummary of the Invention

0010] It is an object of the present invention to provide a method for the real-time onitoring of the formation of superoxide in a cell or specific cell compartment, The method of e invention uses a ratiometric protein probe for detection of formation of superoxide on a iillisecond timescale, making true real-time monitoring possible. The invention may be racticed with standard fluorescence microscopy techniques and equipment. The invention also

Hows the continuous monitoring of superoxide formation in cells while in culture.

)011] It is a further object of the present invention to provide proteins capable of acting s real-time superoxide detecting probes. These proteins may be modified by standard genetic jchniques to include targeting sequences that allow for their localization to a specific cell Dmpartment Upon localization of the superoxide sensing protein to the cell compartment, iperoxide formation the cell compartment may be monitored in real-time.

>O12] It is a still further object of the present invention to provide a method for testing itioxidant agents. As the in vivo formation of superoxide can be monitored by the method of e invention, potential antioxidant agents can be added to cells and their effect on the formation superoxide inside the cells can be monitored.

U3J It is a further object of the present invention to provide a biomarker for diagnosis a disease state. Proteins capable of monitoring superoxide formation within a cell can be tpressed in disease models, and variations in superoxide formation can be monitored during the

'regression of the disease. As such, specific patterns of superoxide formation within a cell can e developed and correlated to the onset of specific diseases, allowing for the early diagnosis of disease.

5014] It is yet a further object to provide a research animal, such as a transgenic mouse, xpressing a protein capable of monitoring changes in superoxide formation within a cell. These ϊsearch animals could be crossed with like animals modeling a specific disease state, such as ancer or neurological disease. The resultant offspring would then be a disease model that lowed for monitoring of superoxide formation within the animal. Such an animal model would low for in depth study of cellular changes in superoxide formation as a biomarker in the sllular environment during the progression of the disease.

rief Description of the Figures

)O15J Figure 1 : Circularly permuted yellow fluorescent protein (cpYFP) as a superoxide dicator. a, Excitation and emission spectra for fully reduced (10 mM reduced DTT) and fully cidized cpYFP (1 mM aldrithiol) purified using the E. CoIi. expression system. Ex: Excitation >ectra obtained at 515 nm emission; Em: Emission spectra at 488 ran excitation. A redox- dependent isosbestic point was identified near 405 nm excitation, permitting ratiometric easurement via dual wavelength excitation (488 nm/405 nm). b, The increase of cpYFP uorescence emission (at 488 nm excitation) and its partial reversal by Cu/Zn-SOD (600 U/ml) hen reduced cpYFP was exposed to xanthine (2 mM) plus xanthine oxidase (20 mU) under srobic conditions, c, cpYFP signal was insensitive to H ₂ O ₂ (0.1 and 10 mM)and slightly

Iecreased by -OH (produced by the Fenton reaction with 1 mM H ₂ O ₂ plus 0.1 mM FeS04 under

iaerobic conditions).

0016] Figure 2: cpYFP responses to peroxynitrite (a), nitric oxide (b), Ca2+ (c) pH (d) nd several metabolites (e-g), including ADP (1 mM), ATP (10 mM), NAD+ (10 mM), NADH 1 mM), NADP+ (1OmM), and NADPH (1 mM). Peroxynitrite was produced by dissolve SIN-I 1 mM) in aerobic solution and nitric oxide was produced by dissolve SIN-I (1 mM) in naerobic solution.

3017] Figure 3: Superoxide flashes in single mitochondria, a, Confocal visualization of a ingle mitochondrion superoxide flash in a rat cardiac myocyte. Upper panel: Confocal image of mt-cpYFPexpressing cardiac myocyte. The enlarged view shows dual-wavelength excitation 188 and 405 run) imaging of the superoxide flash at 2 s intervals. The area shown is of 2.2x1.7 m ² . b, Time course of the superoxide flash shown in a. c-d, Depression of superoxide flash equency (c) and amplitude (δF/F ₀ , d) by the SOD mimetics, MnTMPyP (50 μM) and tiron (1 [M). n=12-64 flashes from 15-16 cells; *, PO.05; #, P<0.01 ; f, PO.001 versus control, e, uperoxide flashes in primary cultured hippocampal neurons. Arrows mark a spaghetti-shaped itochondrion undergoing repetitive superoxide flashes. Images correspond to the designated •ne points, f, Frequencies of spontaneous superoxide flash activity in different cell types. n=21- 5 cells.

H 8J Figure 4: Characteristics of superoxide flashes in four different cell types. δF/Fo, iplitude; T _p , time to peak; T ₅₀ , 50% decay time after the peak. n=21-53 cells.

0019] Figure 5: Effect of scanning laser intensity on superoxide flash incidence,

0023] Figure 9: Mitochondrial ETC activity is an intrinsic regulator of superoxide flash ncidence. a-f, Absence of superoxide flashes in 143B cells that are completely devoid of nitochondrial DNA (ρ° 143B). Superoxide flashes accompanying loss of TMRM signal were eadily observed in wild type 143B TK- human osteosarcoma cells (WT 143B) as shown by luorescent traces (a) and representative temporal diaries of superoxide flash incidence (c), but ot in p ⁰ 143B cells (b,d) in spite of the presence of δψ _m fluctuations (b). Atractyloside (20 μM) annot induce superoxide flashes in p° 143B cells (e). f, Statistics of superoxide flash frequency i WT and p° 143B cells. n=5- 10 cells, g, Attenuated superoxide flash activity in ETC-deficient ells. Insert; Treatment of PC 12 cells with ethidium bromide (EB, 200 ng/ml) to inhibit iitochondrial DNA replication for up to 60 days resulted in a time-dependent decrease in xpression of the mitochondrial DNA-encoded cytochrome C oxidase subunit I (COX-I) (hence, ;ferred to as p- PC 12 cells). R&A: rotenone (5 μM) and antimycin A (5 μg/ml). n=16-46 cells. , PO.05 versus wild type (WT PC 12) cells. #, P<0.01 versus cells without the ETC inhibitors, h, ihibition of superoxide flash activity by rotenone (Rot, 5 μM), antimycine A (AA, 5 μg/ml) or aCN (5 mM) in rat adult cardiac myocytes, n-7-16 cells, f , PO.001 versus control.

)024] Figure 10: Superoxide flashes in hypoxia and reoxygenation. a, Two dimensional ap of superoxide flashes in a cardiac cell. Light boxes mark locations of superoxide flashes ϊtected during a 1 OO s-scan during hypoxia, and dark boxes mark active sites ~5 min following ioxygenation. b, Temporal diaries of superoxide flash incidence in three representative cells αring hypoxia (left) and reoxygenation (right). Each vertical tick denotes a flash event; data in e top row correspond to the cell in a. c, Averaged data showing superoxide flash frequency ring hypoxia, 5 min, and 1 hr after reoxygenation in the absence or presence of diazoxide etreatment (30 μM for 20 min prior to hypoxia). n=10-16 cells. *, P<0.05 versus all other

jroups; #, PO.01 versus diazoxide group, d, Schematic model for the genesis of mitochondrial luperoxide flashes. In this model, the mPTP opens stochastically in response to physiological IOS levels set by constitutive ROS production by the ETC. Opening of the mPTP causes loss of ,ψ _m , dissipation of chemical gradients across the inner membrane, and mitochondrial swelling, vhich could permit exaggerated respiration and favor the diversion of more electrons to ROS ;eneration. This simple model accounts for superoxide flash properties (e.g., requiring both ETC nd mPTP activities, all-or-none behavior, sensitivity to SOD mimetics) and predicts that uperoxide flashes are a biomarker of oxidative stress. OMM: outer mitochondrial membrane; VlS: inter membrane space; IMM: inner mitochondrial membrane.

detailed Description of the Invention

⁾ 025] The present invention provides a method and protein probe for the facile real-time etection of superoxide formation within a cell or cellular compartment. The method allows for ie detection of changes in cellular superoxide formation on a millisecond timescale using -»mmon fluorescence microscopy techniques.

>026] The present invention measures superoxide formation within a cell or cellular

>mpartment. It is to be understood that all methods described herein for measuring superoxide >rmation within a cell are also applicable for measuring superoxide formation within a cellular )mpartment, such as the mitochondria, the endoplasmic reticulum, or the nucleus. Superoxide >rmation within a specific compartment can be measured by targeting the protein probes of the vention to that specific compartment. Such targeting can be accomplished by the addition of calization sequences.

B*rotein Probes of the Invention

0027] The invention describes protein probes for detection and monitoring superoxide ormation within a cell. A preferred embodiment the protein probe of the invention is the protein robe represented by SEQ ID NO. 1. SEQ ID NO. 1 is a modification of the circularly permuted ellow fluorescent protein (YFP) described as ratiometric pericam in U.S. patent application 0050208624 to Miyawaki et al. and Nagai et al. (Proc. Natl. Acad. ScL, 98:3197-3202, 2001), hich are hereby incorporated by reference herein. The YFP described in US 20050208624 is a ircularly permuted version of the yellow fluorescent protein described by Miyawaki et al. (Proc. ^J atl. Acad. Sci,, 96: 2135 - 2140, 1999), which is hereby incorporated by reference herein. The alcium binding (calmodulin) and transduction (M 13 - calmodulin binding domain from myosin ght chain) domains were removed from the protein desribed in US 20050208624 to form the Dvel superoxide sensing probe of the invention.

>028] The embodiment of the invention represented by SEQ ID NO. 1 is a protein φeroxide probe referred to as cpYFP and having the following properties:

1) a superoxide sensitive excitation maximum wavelength of 488 nm;

2) a superoxide insensitive excitation wavelength (isobestic point) of 405 nm; and

3) an emission maximum wavelength of 515 nm.

)029] The embodiment of the invention set forth in SEQ ID NO. 1 is circularly luted and otherwise modified from the wild type GFP (wtGFP) sequence described by Tsien \nnual Rev. Biochem., 67:509-44, 1998) which is presented here as SEQ ID NO. 2. For the

purposes of describing the invention, specific residues will be referred to as they are numbered in 3EQ ID NO. 1. To illustrate the function of various residues within SEQ ID NO. 1, these esidues are compared with residues in wtGFP and mutants thereof, such as YFP mutants. When liscussing the function of a residue within the sequence of non-circularly permuted fluorescent roteins, residues will be numbered as they are in Tsien (Annual Rev. Biochem., 67:509-44, 998) and the residue numbering system will be referred to as wtGFP (SEQ ID NO. 2).

30] Many modifications, mutations, deletions and additions to SEQ ID NO. 1 can be iade without detracting from the function of the protein probe. However, it is preferred that pecific residues be unchanged in certain embodiments of the protein probes. Preferred residues lclude, but are not limited to: Dl 3, A28, G40, F68, Ll 58, C 160, Gl 77, Yl 78, Gl 79, Ll 80, .181 and Cl 82. Other embodiments of the protein probe of the invention may have variations the residues listed, non-limiting examples of which are described below. It should be derstood that substituting residues in the protein probe cpYFP (SEQ ID NO. 1) may cause ianges in the emission and excitation properties of the probe listed above.

)031] Residue Dl 3 of SEQ ID NO. 1 may contribute to the ratiometric properties of the otein probe. This aspartic acid substitution was introduced by Nagai (Proc. Natl. Acad. ScL ^T SA, 98:3197-202, 2001) in the development of the "ratiometric pericam" Ca ²⁺ sensing protein at is the basis for SEQ ID NO. 1. It is also contemplated that residue 13 of SEQ ID NO. 1 may ϊ other residues that allow the probe to retain its superoxide sensing properties, for example, stidine.

»032] Residues A28 and G40 of SEQ ID NO. 1 may improve the folding properties of e protein. These residues correspond to residues 163 and 175 in wtGFP (SEQ ID NO. 2),

Iwhich were found by Nagai et al. {Nature Biotechnology, 20:87- 90, 2002) to improve the folding of the fluorescent protein at 37 ⁰ C. It is also contemplated that residues 28 and 40 of SEQ D NO. 1 may be other residues that allow the probe to retain its superoxide sensing properties, or example, residue 28 may be valine and residue 40 may be serine,

0033] Residue F68 of SEQ ID NO. 1 may be important for determination of the luorescence wavelength. Residue 68 of SEQ ID NO. 1 corresponds to residue 203 in the wtGFP SEQ ID NO. 2). Various substituions at residue 203 in wtGFP (SEQ ID NO. 2) cause a red shift i the fluoresce of the protein from the green region to the yellow region of the visible light trum, forming a YFP. YFPs described in the literature have either a histidine, tyrosine or henylalanine residue at position 203 of the wild type sequence (see Tsien, Annual Rev. iochem., 67:509-44, 1998). It is preferred that residue 68 of SEQ ID NO. 1 be phenylalanine, owever, it may also be tyrosine or histidine or another residue that allows for the protein probe ) retain its superoxide sensing properties. For example, F68 may be mutated to threonine to :>rm a green fluorescing protein.

3034] Residue Ll 58 of SEQ ID NO. 1 may improve the maturation of the protein probe to a fluorescent protein. Residue Ll 58 of SEQ ID NO. 1 corresponds to residue 46 of wtGFP JEQ ID NO. 2), which was shown by Nagai et al. (Nature Biotechnology, 20:87- 90, 2002) to iprove the formation of the fluorophore. It is contemplated that residue 158 of SEQ ID NO. 1 ay be other residues that allow the probe to retain its superoxide sensing properties, for ample, phenylalanine.

k)35] Residues C 160 and C 182 of SEQ ID NO. 1 may form the redox center of the fotein probe. Substitution of both of these residues to either alanine (C160A / C182A) or

■ 0038] Along with being circularly permuted, the protein probe of SEQ ID NO. 1 also ncludes linker amino acid sequences not present in standard GFP or YFP sequences. In a referred embodiment of the invention, these linker sequences are from residues 2 to 9 RSGIGSAGY) and 104 to 112 (VDGGSGGTG), as shown in SEQ ID NO. 1. It is also υntemplated that the linker sequences may be varied in any manner that retains the superoxide ensing properties of the protein probe. For example, the linker sequences may be shorter or Miger. Further, it is contemplated that the size and relative hydrophobicity index of the amino cids in the linkers could be varied. Varying the types of the amino acids in the linker region iay affect the flexibility of the protein and may cause other solvent effects or changes in the >cal pH surrounding the linker. For example, glycine linkers have been used to allow for reater flexibility in protein linkers (Mori et al., Science, 304:432 - 5, 2005). Even further, it is Dntemplated that one of the linker sequences may not be present at all. The amino acid jquence of the linker sequences can also vary greatly, as long as the superoxide sensing roperties of the protein are maintained.

)039] Preferably, the protein probe of the invention is a circularly permuted variant of

FP. However, in another embodiment of the invention, the protein probe may be the non- rcularly permuted variant as provided in SEQ ID NO. 3, which may also be referred to as 3YFP (non-permuted YFP).

)040I Many modifications, mutations, deletions and additions to SEQ ID NO. 3 can be ade without detracting from the function of the protein probe. However, it is preferred that ecific residues be unchanged in embodiments of the protein probes. Preferred residues include, t are not limited to: Dl 77, Al 92, G204, F232, L75, C77, G94, Y95, G96, L97, K98 and C99.

[ ⁾ ther embodiments of the protein probe of the invention may have variations in the residues isted, non-limiting examples of which are described below. It should be understood that ubstituting residues in the protein probe npYFP may cause changes in the emission and ixcitation properties of the probe.

0041] The preferred residues of npYFP (SEQ ID NO. 3) correspond to the preferred esidues of cpYFP (SEQ ID NO. 1) described above. The corresponding residues are:

D 177 of SEQ ID NO. 3 corresponds to D 13 of SEQ ID NO. 1.

Al 92 of SEQ ID NO. 3 corresponds to A28 of SEQ ID NO. 1.

G204 of SEQ ID NO. 3 corresponds to G40 of SEQ ID NO. 1.

F232 of SEQ ID NO. 3 corresponds to F68 of SEQ ID NO. 1.

L75 of SEQ ID NO. 3 corresponds to Ll 58 of SEQ ID NO. 1.

C77 of SEQ ID NO. 3 corresponds to C 160 of SEQ ID NO. 1.

G94 of SEQ ID NO. 3 corresponds to G177 of SEQ ID NO. 1.

Y95 of SEQ ID NO. 3 corresponds to Y 178 of SEQ ID NO. 1.

G96 of SEQ ID NO. 3 corresponds to Gl 79 of SEQ ID NO. 1.

L97 of SEQ ID NO. 3 corresponds to Ll 80 of SEQ ID NO. 1.

K98 of SEQ ID NO. 3 corresponds to Kl 81 of SEQ ID NO. 1.

C99 of SEQ ID NO. 3 corresponds to CI 82 of SEQ ID NO. 1.

>421 The residues listed above may have essentially the same function as their

(>rrespondmg residues in SEQ ID NO. 1. Further, the non-limiting example mutations of the

'referred residues of SEQ ID NO. 1 may also be substituted to for the preferred residues of SEQ D NO. 3. In other words, as a non-limiting example, Dl 77 of SEQ ID NO. 3 may also be istidine.

0043] The protein probe of SEQ ID NO. 3 also includes similar linker amino acid equences to those in SEQ ID NO. 1. In a preferred embodiment of the invention, these linker equences are from residues 13 to 20 (RSGIGSAG) and 21 to 29 (VDGGSGGTG), as shown in EQ ID NO. 3. It is also contemplated that the linker sequences may be varied in any manner iat retains the superoxide sensing properties of the protein probe. For example, the linker equences may be shorter or longer. Further, it is contemplated that the size and relative ydrophobicity index of the amino acids in the linker could be varied. Varying the types of the Tiino acids in the linker region may affect the flexibility of the protein and may cause other )lvent effects or changes in the local pH surrounding the linker. For example, glycine linkers ave been used to allow for greater flexibility in protein linkers (Mori et al., Science, 304:432 - 5, 005). Even further, it is contemplated that one of the linker sequences may not be present at all. he amino acid sequence of the linker sequences can also vary greatly, as long as the superoxide :nsing properties of the protein are maintained.

⁾ 044] It should be noted that, although the non- circularly permuted version of a iodified YFP is a functional superoxide sensing protein, this function is not inherent in other FPs and YFPs. When a commercially available, mitchondrially targeted, non-circularly ϊrmuted YFP {Calbiochem, Mountain View, CA - catalog number 632347 (discontinued - now Ltalog number 632432)) was tested, it was found to have no superoxide sensing properties (data rt shown).

0045] Protein tags known in the art may be added to the protein probes to effect argeting, purification and/or location of the probes, One or more tags may be added to either the 4- or C- terminus, or both termini, as required.

0046] Various localization signals and targeting sequences that are well known in the art iay be added to the probes as targeting tags. Targeting tags may be selected based on the itracellular compartment inside of which superoxide is to be monitored. For example, targeting igs may be added to probes to effect their targeting to the cytoplasm, the Golgi, the ndoplasmic/sarcoplasmic reticulum, mitochondria, peroxisome and the nucleus, along with ther cellular compartments. Non-limiting examples of sequences that may be used as targeting igs in the present invention are disclosed in Wickner and Schekman (Science, 310:1452 - 6, 005) and Shaner et al. (Nature Methods, 2:905 - 09, 2005) which are hereby incorporated by ϊference herein.

)047] Specific protein tags may be added to the probes of the invention to allow for ieir purification. Examples of protein tags that may be added to effect purification of the probes lclude, hexahistidine (Hisg) tags, maltose binding protein (MBP) tags, glutathione-S-transferase 3ST) tags, the IgG domain from protein A, and the like.

>048] Specific protein tags may also be added to the probes of the invention to allow for ieir purification and/or localization after they are expressed inside a cell or cellular jmpartment. Examples of tags that may be added to effect location of the probes include smagglutin (HA) tags, FLAG-tags, Myc-tags and the like. Protein probes bearing these tags in then be purified and/or identified using antibodies to the tags, as is well known in the art. ucleic Acids of the Invention

0049] The protein probes of the invention may be expressed from a nucleic acid equence encoding the amino acid sequence of the probe, A preferred nucleic acid sequence of he invention is encoded by the nucleic acid sequence SEQ ID NO. 4, which is one of the ^p ossible nucleic acid sequences encoding the protein probe of SEQ ID NO. 1. Other nucleic acid equences are contemplated by the invention, including other nucleic acid sequences encoding be probes of SEQ ID NO. 1 and SEQ ID NO. 3, along with nucleic acid sequences encoding ther variants of protein probes, as described above.

3050] The nucleic acid sequences of the invention may be incorporated into larger ucleic acids, such as a vector, to allow for their transformation into cells for expression of the rotein probes. For example, the nucleic acid sequences of the invention may be incorporated ito a vector that allows for transformation of the protein probes into mammalian cells, fungal Ms or bacterial cells. The nucleic acid sequences may also be incorporated into viral vectors iat allow for the transfection of mammalian or other types of cells.

⁾ 051] If a protein tag is to be added to the probe, the nucleic acid sequence encoding the rotein tag can be linked upstream or downstream from the nucleic acid sequence of the ivention. As such, probes expressed from these nucleic acid sequences will contain the desired gs for targeting, localization, and the like. Further, it is also contemplated that the probe could s tagged to another cellular protein, such as xanthine oxidase or superoxide dimutase, predicted i influence superoxide production or degredation within the cell.

'ell Lines and Organisms of the Invention

)052] The invention contemplates cell lines stably or transiently expressing protein robes capable of monitoring intracellular superoxide formation. Nucleic acids encoding

nibodiments of the protein probe described above may be transfected or otherwise delivered to ells using methods known in the art. The nucleic acids encoding the protein probe will then be xpressed during the regular growth of the cell line. Cell lines of the invention may be modified ersions of mammalian, fungal, bacterial, insect, fish and plant cell lines. Non limiting examples f mammalian cells lines which may be modified include HeLa cells, MDCK cells, CHO cells, 4CF-7 cells, U87 cells, Al 72 cells, HL60 cells, A549 cells, Vero cells, GH3 cells, 9L cells, 1C3T3 cells, C3H-10T1/2 cells, C2C12 cells, PC12 cells, 143B cells and NIH-3T3 cells. Real- me changes in an intracellular superoxide formation in these cells can then be monitored by andard fluorescence techniques.

M ⁾ 53] The invention also contemplates organisms that contain cells expressing protein robes capable of monitoring intracellular superoxide formation. Nucleic acids encoding mbodiments of the protein probe described above may be incorporated into the DNA of the rganism or delivered to cells as an extra-chromosomal element. After the nucleic acid encoding protein probe is provided to at least some of the cells of an organism, these cells of the rganism will express a superoxide sensitive protein probe. Any research model organism can 3 modified to express the protein probe of the invention, including, rats, mice, zebrafish, aenorhabditis elegans, yeasts such as Saccharomyces cerevisiae, Schizosaccharomyces pombe id Pichia pastoris and bacteria such as Escherichia coli.

K)54] The modified organisms of the invention can then be used for monitoring tracellular superoxide formation under standard growth and development conditions. These ^■ ganisms may also be exposed to a variety of agents, both therapeutic and toxic, to determine e effect of these agents on intracellular superoxide formation. Further, the modified organisms

f the invention may be crossed with known disease organism models. As the progeny of these rosses will both develop the disease in question and express superoxide sensitive protein probes, iey may be used to monitor the change in intracellular superoxide formation during the rogression of the disease.

lethods for Monitoring Intracellular Superoxide

3055] The methods of monitoring superoxide formation in a cell or cellular ompartment of the invention can be carried out using the standard techniques for expression and sualization of fluorescent proteins known in the art. Non-limiting examples of such techniques an be found in Silver (J. Biol. Chem., 277 :34042 - 7, 2002) and Weiss et al. (Am. J. Physiol. "ell Physiol., 287:C1094 - U 02, 2004), which are hereby incorporated by reference herein. lonitoring the Effect of an Agent

)056] Using the methods described above and other methods known in the art, the cell nes and organisms of the invention maybe used to monitor the effect of an agent on itracellular superoxide formation. Agents that may be tested include therapeutic agents, such as iarmaceuticals and biologies, known toxic agents and agents with unknown effect. Such agents ay be administered at levels previously known from pharmacological or toxicological studies.

)057] After an agent is administered, the changes in superoxide formation may be onitored. Such changes will be indicative of the effects of the agent, and may be correlated ith the development of a specific disease state by analyzing the pattern of change.

iagnosis of a Disease State

zommercially prepared reagents, for example, nuclear cDNA injection as described by Weiss et il., (Am. J. Physiol. Cell Physiol, 287:C1094 - 1 102, 2004).

0061] The examples set forth below are meant to provide non-limiting examples of nethods of the invention. It should be apparent that there are variations of the invention not resented in the examples below that fall within the scope and the spirit of the invention as laimed.

Examples

xample 1 - Materials and Methods

DNA Constructs

)062] mt-cpYFP was constructed from mitochondrial targeted ratiometric peri cam pericamMT) cloned into pcDNA3 (Nagai et al., Proc. Natl. Acad. ScL USA, 98: 3197 - 3202, 001) by removing nucleotide sequences encoding calmodulin (nt 886-1323) and M13 (nt 49- 26) using the gene splicing by overlap extension (SOE) technique (Horton et al, Gene, 77:6\ - 8, 1989). The final PCR product was digested with Hindlll/Xbal and cloned into the 5352 bp indlll/Xbal fragment of ρcDNA3, cpYFP was constructed from mt-cpYFP by removing jcleotide sequences encoding the 11 amino acid (LSLRQSIRFFK) mitochondrial targeting :quence of cytochrome oxidase subunit IV (nt 4-36) using gene-SOEing. The final PCR product as digested with Hindlll/Xbal and cloned into the 5352 bp Hindlll/Xbal fragment of pcDNA3. ouble cysteine-to-alanine and cysteine-to-methionine substitutions in mt-cpYFP "171 A/C193A, and C171M/C193M) were constructed using a standard two-step sitedirected

or

an

cell

--SM 510 confocal microscope equipped with a 63x, 1.3NA oil immersion objective and a ampling rate of 0.7 s/frame was used. Dual wavelength excitation imaging of mt-cpYFP was ichieved by alternating excitation at 405 and 488 nm and collecting emission at >505 nm. Tri- vavelength excitation imaging of mt-cpYFP and TMRM (20 nM) or rhod-2 was achieved by andem excitation at 405, 488, and 543 nm, and the emission was collected at 515-550, 515-550 nd >560 nm, respectively. To increase mitochondrion retention of rhod-2, the indicator loading trotocol described by Hajnoczky G et al. was used with modification (Hajnoczky et al., Cell, 82: 15 - 424, 2000). Briefly, cells were loaded with 4 μM rhod-2 AM (after NaBH ₄ quenching) at ⁰ C for 1 hr, and then changed to normal culture medium for 4 hrs. The standard extracellular erfusion solution contained (in mM): NaCl 137, KCl 4.9, CaCl ₂ 1, MgSO ₄ 1.2, NaH ₂ PO ₄ 1.2, lucose 15, and HEPES 20 (pH 7.4). Digital image processing was performed using IDL oftware (Research Systems) and customer-devised programs.

litochondrial DNA-deleted or deficient (ρ° or p-) cells

}065] p° 143B TK- human osteosarcoma cells and its wild type control were a generous ift from Dr. Nadja C. de Souza-Pinto (National Institute on Aging, NIH). Wild type and p° 43B cells were cultured under identical conditions, in DMEM medium supplemented with 10% S, 100 μg/ml pyruvate, 100 μg/ml bromodeoxyuridine and 50 μg/ml uridinel7. Mitochondria ρ° 143B cells completely lack mitochondrial respiration, due to the loss of critical ETC oteins including constituents of complex I (NDl -6, ND4L), complex III (cytochrome b) and >mplex IV (COX Mil) encoded by mitochondrial DNA. To partially deplete mitochondrial ¹ NA and allow partial disruption of mitochondrial respiration, PC 12 pheochromocytoma cells

ivere cultured in DMEM medium with 10% FBS, 200 ng/ml ethidium bromide, 100 μg/ml ^p yruvate and 50 μg/ml uridine for up to 60 days. Depletion of mitochondrial DNA was ϊvidenced by western blot analysis of cytochrome C oxidase subunit I.

ϊypoxia and reoxygenation treatment of cardiac myocytes

066] Cardiac myocytes expressing mt-cpYFP were cultured in a hypoxia chamber

Billups-Rothenberg) at 37 ^C C and ventilated with 95% N2 plus 5% CO ₂ for 6 hours. At the end f hypoxia treatment, culture dishes were sealed with a plastic cover and immediately transferred nto the stage of a confocal microscope. After recording superoxide flashes under hypoxic ondition, reoxygenation was achieved by removing the seal and superfusing cells with standard xygenated extracellular solution.

tatistics

)067J Data were reported as mean ± SEM. Paired and unpaired Student's t test and

NOVA with repeated measurements were applied, when appropriate, to determine statistical gnificance of the differences. P<0.05 was considered statistically significant.

cample 2 - Spectral analysis of cpYFF

J] Unexpectedly, it was found that a circularly permuted yellow fluorescent protein

3 YFP), previously used to construct the Ca2+ indicator pericam (Nagai et al., Proc. Natl. Acad,

^■ ci USA, 98: 3197 - 3202, 2001), can serve as a novel biosensor for superoxide anions (CV),

tie primal ROS from the electron transfer chain (ETC) in mitochondria, via a redox dependent lechanism. Using cpYFP purified from an E, coli expression system, excitation and emission luorescence spectra were measured in response to reducing (10 mM reduced DTT) and xidizing manipulations (1 mM aldrithiol). The oxidized cpYFP was about five times brighter ian the fully reduced species when excited at 488 nm (Fig. Ia), indicative of a good signal-to- ackground in contrast to recently reported redox-sensitive GFP probes (Hanson et al., J. Biol, "hem. 279: 13044-13053, 2004; Ostergaard et al., EMBOJ., 20: 5836 - 5862, 2001). Extensive i vitro experiments were performed to determine the selectivity of cpYFP among hysiologically relevant oxidants and metabolites. It was found that compared to the fully

ϊduced state, cpYFP fluorescence displayed a 420% increase in response to O ₂ - ^" produced by

ie xanthine / xanthine oxidase (2 mM / 20 mU) system under aerobic conditions; addition of WZn-superoxide dismutase (600 U/ml) partially inhibited this response (Fig. Ib). The cpYFP gnal, however, was insensitive to hydrogen peroxide (H ₂ O ₂ ) over a wide range of Micentrations (0.1-10 mM) (Fig. Ic) and peroxynitrite (Fig. 2), and was decreased by hydroxyl

idicals (-OH) (Fig. Ic) and nitric oxide (Fig. 2). Other metabolites tested, including ATP, ADP,

AD(P)+, NAD(P)H and Ca ²⁺ at physiological concentrations, exerted negligible or only arginal effects (Fig. 2). As would be expected of a fluorescent protein-based indicator (Nagai et ., Proc. Natl. Acad. Sci USA, 98: 3197 - 3202, 2001; Bel ousov, et al., Nat, Methods, 3:281 - 16, 2006) cpYFP was brighter in basic environments such as those found within the itochondrial matrix (pH -8.0) (Fig. 2).

example 3 - Expression of cpYFP in cardiac myocytes

0069] Adenoviral gene transfer was employed to express cpYFP targeted to the iitochondria of cardiac myocytes via a cytochrome C oxidase subunit IV (COX IV) targeting equence (mt-cpYFP).

9070] Confocal imaging revealed that mt-cpYFP stained bundle-like subcellular

Eructures that were punctuated at Z-lines of the sarcomere, in agreement with spatial rganization of cardiac mitochondria (Fig. 3a; Ramesh et al., Ann. N. Y. Acad. Sci., 853:341 - 344, 998). Strikingly, it was found that localized flashes of mt-cpYFP fluorescence occur ;ochastically in a quiescent background (Figs. 3a-b). A typical flash rose abruptly, peaked in 5±0.1 s, and then dissipated with a half time of 8.6±0.2 s (n=409) (Figs. 3b and 4). The veraged fold-increase of mt-cpYFP fluorescence in a flash was 0.41 ± 0.02 (δF/FO); the top 0% brightest events, which were most likely located on or close to the confocal imaging plane, splayed a δF/FO of 1.0 ± 0.1 (n=41). While randomly distributed throughout the myocyte, dividual flashes were sharply confined to tiny elliptical areas each spanning 0.94 ± 0.01 μm terally and 1.68 ± 0.03 μm longitudinally (n=409 flashes from 53 cells), while mitochondria in eir immediate vicinity remained quiescent.

K)71] Since a 5-fold increase of the scanning laser intensity did not significantly alter e rate of flash production (Fig. 5), flashes described above were unlikely a phenomenon duced by photostimulation.

Example 4 - Spontaneous mt-cpYFP fluorescent flashes reflect bursts of matrix O ₂ - ^* in

ingle mitochondria (superoxide flashes) under physiological conditions

0072] Experiments using mitochondrially targeted-EYFP as a pH biosensor (Takahashi it al., Biotechniques, 30: 804 - 808, 2001) failed to detect transient mitochondrial alkalinisation rith a similar frequency and time course as flashes (Fig. 6), excluding mitochondrial alkalosis as n explanation for flashes. Importantly, application of MnTMPyP (50 μM), an SOD mimetic, ihibited flash activity by 83% and halved flash amplitude (δF/F ₀ =0.18±0.02, n=12; p<0.01 vs ontrol) (Figs. 3c-d); tiron (1 raM), a superoxide radical scavenger, similarly diminished the

equency and amplitude of flashes (Figs. 3c-d), supporting their O ₂ - ^" origin. Substituting the

nly two cysteine residues in cpYFP with either alanine (Cl 71 A/C193A) or methionine

171M/C193M) diminished basal fluorescence and made the indicator redox-insensitive (Fig. ). Mitochondrial flash activity was never detectable in cardiac cells expressing either of the two ysteine-null, redox-insensitive cpYFP variants (n=15 cells).

xamplc 5 - Superoxide flashes are not unique to cardiac cells

)073J Superoxide flashes were not unique to cardiac cells, but appeared to be universal nong a wide diversity of cell types examined, including skeletal myotubes, neurons, ξuroendocrine cells, fibroblasts and osteosarcoma cells. Fig. 3e shows superoxide flashes :solved in spaghetti-shaped mitochondria in primary cultured hippocampal neurons. Careful ispection revealed that multiple superoxide flashes can occur within one mitochondrion. The ite of superoxide flash occurrence, however, varied widely across cell type, ranging from 3,8 ±

idence for physiological mPTP activity in quiescent cells.

espectively, all virtually abolished the occurrence of superoxide flashes in cardiac myocytes as ^y ell as p- PC 12 cells (Figs. 9g-h), It is noteworthy that AA-induced decreased superoxide lashes in sharp contrast to the previous finding that AA increases cellular ROS production when ieasured with the H ₂ O ₂ indicator dichlorodihydrofluorescein (DCF, Aon et al., J. Biol, Chem. 78: 44735 - 44744, 2003). This apparent discrepancy may be reconciled by the fact that DCF oes not discriminate between intra- and extra-mitochondrial matrix ROS signals, and AA ihibits electron transfer from the outer (Qo) to inner (Qi) center of complex III, decreasing

latrix O ₂ - ^' production while facilitating production of O ₂ - ^" (membrane-impermeable) and H ₂ θ ₂

nembranepermeable but cpYFP-insensitive) toward the cytosol (Turrens, J. Physiol, 552: 335 344, 2003).

xample 8 - Frequency-dependent modulation of superoxide flashes in cardiac myocytes uring hypoxia and reoxygenation.

)0791 Oxidative stress and aggravated ROS production contribute to the pathogenesis of number of clinically distinct disorders including neurodegeneration (e.g. Alzheimer's disease), ssue inflammation, hypertension, atherosclerosis, diabetes, and cancer (Andersen, Nat. Med., 3: S 18- S25, 2004; Dhalla et al., J. Hypertens., 18: 655 - 673, 2000; Klaunig and Kamendulis, nnu. Rev. Pharmacol. Toxicol., 44:239 - 267, 2004). Since flashes are triggered by mPTP itivity that is itself sensitive to ROS (Vercesi et al., Biosci. Rep., 17: 43 - 52, 1997; Turrens, J. hysiol., 552: 335 - 344, 2003), the frequency of superoxide flashes may vary during stress or sease, and may therefore serve as a biomarker of oxidative stress such as those in ischemia- perfusion. Sustained hypoxic treatment (95% N ₂ and 5% CO ₂ for 6 hrs) depressed

ψd chemical gradients across the inner membrane, the further activation of the ETC, and

>erhaps mitochondrial swelling due to water movement. This gives rise to a burst of matrix O ₂ -

iroduction that is visualized as a superoxide flash in a single mitochondrion. (Fig, 1Od).