| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
METHODS IN CELL PHYSIOLOGY
1Department of Physiology, Yamagata University School of Medicine; 2Department of Environmental Life Science, Graduate School of Medical Sciences, Yamagata University; 3Research Laboratory for Molecular Genetics, Yamagata University; and 4Central Laboratory for Research and Education, Yamagata University School of Medicine, Yamagata, Japan
Submitted 9 February 2006 ; accepted in final form 16 May 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODS AND RESULTSDISCUSSIONGRANTSREFERENCES |
|---|
heart; ischemia; hypoxia; molecular imaging
Endogenous chromophores that change their optical properties according to oxygen availability are particularly suitable for in vivo molecular imaging of oxygen. For instance, myoglobin occurs in the cytosol of red muscles and changes its light absorption according to the oxygen abundance, with a P50 of 
2 mmHg (33). Because the oxygen affinity of myoglobin overlaps the regulatory PO2 range of mitochondrial respiration (11), optical or NMR spectroscopy of myoglobin in situ is an excellent technique to define physiologically meaningful heterogeneities of tissue oxygen concentration in organs (14, 16, 28) and in isolated cells (27, 32). One apparent limitation of this technique is that it is only applicable to red muscles such as skeletal and cardiac muscles.
Other candidates for in vivo oxygen indicators are redox intermediates such as cytochromes, oxidized flavoproteins, and NADH. For example, NADH is oxidized in the respiratory chain to oxidized NAD (NAD+). Unlike NAD+, NADH produces strong fluorescence on ultraviolet excitation. Because NADH oxidation depends on oxygen availability at the terminal enzyme in the respiratory chain, cytochrome c oxidase, the oxygen concentration in mitochondria may be assessed by fluorometry of mitochondrial NADH in organs and in isolated cells (1, 4, 12, 26, 30). Although this technique is potentially capable of resolving heterogeneities of the oxygen demand-supply balance at the level of mitochondria, the NADH concentration in mitochondria is also affected by the rate of upstream substrate flow. Furthermore, the oxygen dependency of NADH oxidation in the respiratory chain is under physiological control and is not a singular function of oxygen concentration (10, 31). Thus, under in vivo conditions, the oxidation-reduction state of the mitochondrial redox intermediates is indirectly related to the oxygen availability in mitochondria.
Green fluorescent protein (GFP) has been widely used as a fluorescent marker for gene expression and for the intracellular localization of proteins in living cells. Several GFP variants have been generated in which the fluorescence spectrum ranges from blue to yellow, but the group of enhanced GFPs does not display significant red fluorescence. In 1997, Elowitz et al. (6) reported that in Escherichia coli transformed with a plasmid that expresses a GFP, the GFP fluorescence may turn reddish after a brief (2–60 s) illumination with 475–495 nm light if the bacteria are sealed between microscope slides and left for a prolonged time (25 min). The red shift of GFP fluorescence was also demonstrated in the transformed E. coli when photoactivation was conducted after addition of an oxygen scavenger to the suspension medium. These results indicate that the fluorescent characteristics of enhanced GFP may be modified according to the oxygen moiety of the molecule and, from the standpoint of oxygen measurement, suggest the possibility that GFP might be used as an in vivo oxygen indicator.
In the present study, we first determined the characteristics of enhanced GFP as an intracellular oxygen indicator in cultured COS-7 cells. Then, toward the goal of in vivo molecular imaging of oxygen in tissues, we attempted a high spatial resolution imaging of hypoxia in adult single cardiomyocytes freshly isolated from the GFP knock-in mouse (green mouse) and in organs (kidney, liver, and heart) in the green mouse. We photoactivated GFP in vivo to see whether the green fluorescence of GFP turns reddish under anaerobic conditions.
|
METHODS AND RESULTS |
|---|
|
TOPABSTRACTMETHODS AND RESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Cultured COS-7 cells were transfected with expression vector pCMX-SAH/Y145F (pCMX-humanized GFP-S65A/Y145F) by using a linear polyethylenimine reagent (jetPEI; Polyplus-transfection, Illkirch, France), and cultured for 48 h on a 13.5-mm-diameter low-fluorescence plastic disk (Cell Disk LF; Sumilon, Tokyo, Japan). Cultured cells were washed in PBS three times and placed in an airtight measuring cuvette at 37°C and superfused with humidified gas at various oxygen concentrations. A 16-bit charge-coupled device (CCD) camera (model SV512; PixelVision, Tigard, OR) captured green and red fluorescent images of cells by using NIBA (excitation 470–490 nm, emission 515–550 nm) and WIG (excitation 520–550 nm, emission >580 nm) filter sets, respectively, in an inverted microscope with a x40 long-working-distance objective lens (model IX70; Olympus, Tokyo, Japan). The exposure duration was 2 s while a 6% neutral density filter was inserted in the light path. Photoactivation was conducted by illuminating the cell with 470–490 nm light (neutral density filter removed) for 60 s using a 100-W mercury arc lamp in the microscope.
As shown in Fig. 1, red fluorescence was negligible (<1.7% of the green fluorescence level) either before or after photoactivation in normoxic cells (superfusion with 10% oxygen). In contrast, in hypoxic cells (superfusion with <1% oxygen), red fluorescence appeared following photoactivation (Fig. 1). The magnitude of the anaerobic red fluorescence increased with an increase in photoactivation duration, while it stayed essentially constant at >150 s photoactivation. Thus the red shift of GFP fluorescence was demonstrated in anaerobic COS-7 cells.
|
R) was normalized to the green fluorescence level before photoactivation (G) and designated as R/G. Figure 2 shows that a significant red shift of GFP fluorescence was induced by photoactivation when the oxygen concentration was lowered to <2%. The magnitude of the red fluorescence reached 30% of the green fluorescence level under anaerobic conditions (Fig. 2).
|
|
The intensity of the green fluorescence was 7 ± 2 (in arbitrary units, mean ± SD, n = 10) and 269 ± 70 (n = 10) in single cardiomyocytes isolated from wild-type and green mice, respectively. Although the 38-fold increase in the green fluorescence certainly indicates stable expression of GFP in the green mouse heart, it was substantially lower than the intensity of the green fluorescence in transiently GFP-expressing cultured COS-7 cells (19,600 ± 12,000, n = 9).
Because of the considerably lower expression of GFP in cardiomyocytes in the green mouse, the red autofluorescence in cardiomyocytes now could not be ignored (Fig. 4). The levels of the red autofluorescence under our experimental conditions were 86 ± 30 (n = 10) and 125 ± 20 (n = 10) in single cardiomyocytes isolated from wild-type and green mice, respectively. Thus the level of nonspecific red fluorescence was comparable to 46% of the GFP green fluorescence. To abolish the effects of the red autofluorescence, the difference in the red fluorescence images before and after photoactivation (R) was calculated, with the assumption that photoactivation per se does not affect the red autofluorescence. Finally, the image was normalized to the corresponding green fluorescence image (R/G) to compensate for variations in fluorescence arising from the geometry of the cell (for instance, cell thickness). These image-related calculations were performed using IPLab software (Scanalytics, Rockville, MD). As a result, the red fluorescence, with a magnitude of 20% of the green GFP fluorescence, could be demonstrated (Fig. 4).
|
|
5 mm illuminated the surface of the organ. Ten minutes after ligation of the hepatic artery and the portal vein, the first fluorescence measurements were conducted. Photoactivation was then conducted and was immediately followed by the second fluorescence measurements. Figure 6 shows the appearance of the red fluorescence in the ischemic liver after a photoactivation. It must be noted that the red shift of the GFP fluorescence was consistently observed only in regions where the surface of the liver was covered with Saran Wrap. These changes in the red fluorescence were not observed in the normoxic blood-perfused liver (data not shown). Because the hepatic artery and the portal vein were occluded, the oxygen supply to the tissue should have been almost completely cut off. Therefore, the finding that the red shift was not observed in regions without Saran Wrap suggests that a significant amount of oxygen might have diffused from the atmosphere into the tissue. Similar observations were made in the blood-perfused kidney (data not shown).
|
200 beats/min and fluorescence measurements required prolonged durations of exposure, motion blurring in the fluorescent image was a serious problem. To avoid the motion artifact associated with spontaneous heart beating, the Langendorff perfusion was temporarily discontinued (for <1 min) after a photoactivation. After verifying that the heartbeat was spontaneously diminished and finally stopped, the fluorescence measurement for the red shift was conducted. This protocol was based on the finding that the red shift persists as long as the heart is kept hypoxic (Fig. 3). Finally, to compensate for heterogeneities in GFP fluorescence in the heart, the red fluorescence was normalized to the GFP (green) fluorescence. The red shift was evident in regions where the blood supply appeared to have depended on the ligated coronary artery (Fig. 7), while significant macroscopic heterogeneities were observed. Thus visualization of hypoxic regions by the red fluorescence of GFP was possible in the beating crystalloid-perfused heart.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODS AND RESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Chance (3) was the first to establish the theoretical basis for the detection of tissue oxygen heterogeneities. He pointed out that the oxygen indicator must be uniformly distributed in the tissue and its oxygen affinity must be appropriate for the respiratory chain of mitochondria. If spatial heterogeneity is concerned, the measurement should be conducted with a spatial resolution as high as the order of the intercapillary distances (tens of micrometers).
The present technique using GFP as an oxygen indicator may fulfill these requirements. First, GFP is expressed in the cytoplasm of the green mouse. The expression level of GFP within a particular organ appeared to be relatively uniform, whereas the GFP fluorescence in our green mouse varied considerably according to the organ (in saline-washed organs, the GFP fluorescence levels normalized by the fluorescence in the heart were 2.8, 2.1, 34, 2.6, and 4.7 in the kidney, the spleen, the liver, the lung, and the brain, respectively). Furthermore, variations in the GFP fluorescence within an organ arising from variations in tissue thickness and heterogeneous illumination by the excitation light are effectively compensated for by introducing an image equalization technique (R/G). Second, as shown in Fig. 2, the magnitude of the red shift abruptly increased when the COS-7 cells were superfused with <1% (7 Torr) oxygen. This result certainly indicates that GFP cannot be used for quantitative determination of oxygen concentrations ranging from arterial blood PO2 to anoxia. Rather, it may be suitable for the detection of dysoxia, because this threshold oxygen concentration overlaps the critical PO2 of the cell at which mitochondrial oxygen metabolism is compromised (9, 20). Third, in the green mouse, GFP is endogenously expressed in the cytoplasm, as myoglobin is in red muscles. Thus GFP reports the intracellular oxygen level at a molecular spatial resolution, and GFP can therefore be used as a probe for imaging tissue hypoxia optimal for the physiological PO2 range.
To utilize GFP as a molecular oxygen probe, the protein must be expressed in cells/tissues. Albeit the required level of expression differs according to the sensitivity of the optical system, a high level of GFP expression in a tissue is of particular importance. In cultured COS-7 cells, the conventional transient expression technique using a linear polyethylenimine reagent resulted in sufficient GFP expression. With the conventional fluorescence microscope, the green fluorescence was strong and the spectrum shift was evident after a brief irradiation with the activation light, enabling us to easily define the basic characteristics of the anaerobic red shift in the mammalian cell (Figs. 1–3). In contrast, efficient expression of the protein was difficult to achieve in vivo, which is a common issue concerning the in vivo introduction of exogenous genes. We attempted in vivo transient GFP expression in the rat using a linear polyethylenimine reagent specifically designed for in vivo use but were unsuccessful at that time. We then used a genetically engineered mouse that stably expresses enhanced GFP (green mouse). Although the organs of our green mouse exhibited significant green fluorescence, the level in single ventricular myocytes was only 1/73 that in the transiently GFP-expressing COS-7 cells. Thus we had to prolong the exposure duration of the CCD camera until sufficient GFP fluorescence was detected. This procedure, at the same time, elevated the level of nonspecific red fluorescence, probably originating from endogenous flavoproteins. In fact, under our experimental conditions, the level of the nonspecific red fluorescence was almost of the same order as the GFP green fluorescence in cardiac myocytes of the green mouse. The presence of the strong background red fluorescence caused significant deterioration of the signal-to-noise ratio in the red shift detection. For this reason, we failed to detect the red shift in single red fluorescence images alone, although it was possible to do so in the COS-7 cells (Fig. 1), and somewhat complicated calculations using data from the images (R/G) were indispensable (Fig. 4). Furthermore, in our macroscopic imaging device used for the organs, the magnitude of the photoactivation light was only one-third compared with that for the microscope with high numerical aperture optics. Therefore, we had to prolong the photoactivation duration to 10–15 min. This significantly restricted the time resolution of the imaging in the time-lapse measurement. For the in vivo imaging of hypoxia in organs to be practical, more efficient photoactivation should be realized by using, for example, laser scanning photoactivation optics.
Recently, various fluorescent proteins that show a photoactivatable shift in the fluorescence wavelength, such as Kaede (2), or fluorescence augmentation, such as PA-GFP (18) and KFP1 (5), have been reported (see Table 1 in Ref. 18). Because distinct fluorescence can be artificially induced, these proteins have been used as endogenous molecular markers in living cells. The molecular mechanism of the red shift of GFP fluorescence demonstrated in the present study may not be identical to that in these photoactivatable proteins, because the red shift can be achieved and is stable only under anaerobic conditions (Fig. 3). The precise mechanism of the anaerobic red shift after photoactivation thus remains to be elucidated.
Irradiated light penetrates into tissues. Gandjbakhche et al. (8), using Monte Carlo simulation of photon migration in tissues, estimated that the mean penetration depth of 510–590 nm light within the epicardium of blood-free heart tissue is 0.33–0.40 mm. In fact, Seiyama et al. (22) demonstrated in blood-free liver tissue a penetration depth of 0.6 mm at 605 nm. On the other hand, atmospheric oxygen diffuses into tissues. Stücker et al. (23) reported that atmospheric air penetrates to a depth of 0.25–0.40 mm in the skin. Thus the volume of tissue subjected to the epifluorescence measurement of oxygen is also sensitive to the oxygen concentration of surrounding medium. In the majority of physiological experiments in isolated organs, including the present one, organs are exposed and the surface is in contact with the atmospheric air. Therefore, it is not surprising that the red shift of GFP fluorescence was consistently observed only in regions where the surface was covered with Saran Wrap (Fig. 6). Thus special care should be taken in hypoxia imaging experiments using isolated organs so that atmospheric oxygen does not disturb oxygen distribution in the tissue.
The red shift occurs at <2% oxygen. It may be more relevant experimentally and clinically if GFP could detect higher level of oxygen such as 6% oxygen (venous blood oxygen level). Because the complete amino acid sequence of GFP has been determined, it would be possible to manipulate the sensitivity of the fluorescence shift to oxygen using genetic engineering technology. Then, the oxygen gradients from the capillary blood to mitochondria could be imaged in the tissue.
In addition to the in vivo molecular imaging of hypoxia, the present technique may potentially permit further applications in hypoxia research in vivo. For example, if genes responsible for adaptation to decreases in oxygen, such as HIF-1
, are cotransfected with the GFP gene (17), then the GFP-HIF-1 fusion protein allows one to determine, from the red shift of the fluorescence, whether the spatial heterogeneities in HIF-1 inductions depend on those of oxygen distribution within the tissue.
In summary, utilizing the red shift of GFP fluorescence originally described by Elowitz et al. (6), we have demonstrated that hypoxia can be detected with a subcellular spatial resolution in single cells and organs in vivo. Although the precise molecular mechanism for the red shift in anaerobic GFP is not known at present, the combination of genetic and fluorescence imaging techniques permits construction of high spatial resolution oxygen maps in organs, which should reveal the impact of oxygen heterogeneities within tissues on the physiology and pathophysiology of gene expression and organ function in hypoxia.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODS AND RESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODS AND RESULTSDISCUSSIONGRANTSREFERENCES |
|---|
2. Ando R, Hama H, Yamamoto-Hino M, Mizuno H, and Miyawaki A. An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc Natl Acad Sci USA 99: 12651–1256, 2002.
3. Chance B. Metabolic heterogeneities in rapidly metabolizing tissues. J Appl Cardiol 4: 207–221, 1989.[ISI]
4. Chance B. Pyridine nucleotide as an indicator of the oxygen requirements for energy-linked functions of mitochondria. Circ Res 38, Suppl 1: I31–I38, 1976.[Medline]
5. Chudakov DM, Belousov VV, Zaraisky AG, Novoselov VV, Staroverov DB, Zorov DB, Lukyanov S, and Lukyanov KA. Kindling fluorescent proteins for precise in vivo photolabeling. Nat Biotechnol 21: 191–194, 2003.[CrossRef][ISI][Medline]
6. Elowitz MB, Surette MG, Wolf PE, Stock J, and Leibler S. Photoactivation turns green fluorescent protein red. Curr Biol 7: 809–812, 1997.[CrossRef][ISI][Medline]
7. Ereci
ska M and Silver IA. Tissue oxygen tension and brain sensitivity to hypoxia. Respir Physiol 128: 263–276, 2001.[CrossRef][ISI][Medline]
8. Gandjbakhche AH, Bonner RF, Arai AE, and Balaban RS. Visible-light photon migration through myocardium in vivo. Am J Physiol Heart Circ Physiol 277: H698–H704, 1999.
9. Gnaiger E. Bioenergetics at low oxygen: dependence of respiration and phosphorylation on oxygen and adenosine diphosphate supply. Respir Physiol 128: 277–297, 2001.[CrossRef][ISI][Medline]
10. Gnaiger E, Méndez G, and Hand SC. High phosphorylation efficiency and depression of uncoupled respiration in mitochondria under hypoxia. Proc Natl Acad Sci USA 97: 11080–11085, 2000.
11. Gnaiger E. Oxygen conformance of cellular respiration. A perspective of mitochondrial physiology. In: Hypoxia: Through the Lifecycle, edited by Roach RC, Wagner PD, and Hackett PH. New York: Kluwer Academic/Plenum, 2003, p. 1–17.
12. Ince C, Ashruf JF, Avontuur JA, Wieringa PA, Spaan JA, and Bruining HA. Heterogeneity of the hypoxic state in rat heart is determined at capillary level. Am J Physiol Heart Circ Physiol 264: H294–H301, 1993.
13. Intaglietta M, Johnson PC, and Winslow RM. Microvascular and tissue oxygen distribution. Cardiovasc Res 32: 632–643, 1996.[CrossRef][ISI][Medline]
14. Jue T and Anderson S. 1H NMR observation of tissue myoglobin: an indicator of cellular oxygenation in vivo. Magn Reson Med 13: 524–528, 1990.[ISI][Medline]
15. Kietzmann T, Dimova EY, Flugel D, and Scharf JG. Oxygen: modulator of physiological and pathophysiological processes in the liver. Z Gastroenterol 44: 67–76, 2006.[CrossRef][ISI][Medline]
16. Mancini DM, Wilson JR, Bolinger L, Li H, Kendrick K, Chance B, and Leigh JS. In vivo magnetic resonance spectroscopy measurement of deoxymyoglobin during exercise in patients with heart failure. Demonstration of abnormal muscle metabolism despite adequate oxygenation. Circulation 90: 500–508, 1994.
17. Nomura Y, Tanaka H, Poellinger L, Higashino F, and Kinjo M. Monitoring of in vitro and in vivo translation of green fluorescent protein and its fusion proteins by fluorescence correlation spectroscopy. Cytometry 44: 1–6, 2001.[CrossRef][ISI][Medline]
18. Patterson GH and Lippincott-Schwartz J. A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297: 1873–1877, 2002.
19. Rumsey WL, Pawlowski M, Lejavardi N, and Wilson DF. Oxygen pressure distribution in the heart in vivo and evaluation of the ischemic "border zone". Am J Physiol Heart Circ Physiol 266: H1676–H1680, 1994.
20. Schenkman KA. Cardiac performance as a function of intracellular oxygen tension in buffer-perfused hearts. Am J Physiol Heart Circ Physiol 281: H2463–H2472, 2001.
21. Schroeder T, Yuan H, Viglianti BL, Peltz C, Asopa S, Vujaskovic Z, and Dewhirst MW. Spatial heterogeneity and oxygen dependence of glucose consumption in R3230Ac and fibrosarcomas of the Fischer 344 rat. Cancer Res 65: 5163–5171, 2005.
22. Seiyama A, Chen SS, Kosaka H, and Shiga T. Microspectroscopic measurement of the optical properties of rat liver in the visible region. J Microsc 175: 84–89, 1994.[Medline]
23. Stücker M, Struk A, Altmeyer P, Herde M, Baumgartl H, and Lubbers DW. The cutaneous uptake of atmospheric oxygen contributes significantly to the oxygen supply of human dermis and epidermis. J Physiol 538: 985–994, 2002.
24. Swartz HM and Dunn JF. Measurements of oxygen in tissues: overview and perspectives on methods. Adv Exp Med Biol 530: 1–12, 2003.[ISI][Medline]
25. Takahashi E and Asano K. Mitochondrial respiratory control can compensate for intracellular O2 gradients in cardiomyocytes at low PO2. Am J Physiol Heart Circ Physiol 283: H871–H878, 2002.
26. Takahashi E, Endoh H, and Doi K. Intracellular gradients of O2 supply to mitochondria in actively respiring single cardiomyocyte of rats. Am J Physiol Heart Circ Physiol 276: H718–H724, 1999.
27. Takahashi E, Sato K, Endoh H, Xu Z-L, and Doi K. Direct observation of radial intracellular PO2 gradients in a single cardiomyocyte of the rat. Am J Physiol Heart Circ Physiol 275: H225–H233, 1998.
28. Tamura M, Oshino N, Chance B, and Silver IA. Optical measurements of intracellular oxygen concentration of rat heart in vitro. Arch Biochem Biophys 191: 8–22, 1978.[CrossRef][ISI][Medline]
29. Tsai AG, Friesenecker B, Mazzoni MC, Kerger H, Buerk DG, Johnson PC, and Intaglietta M. Microvascular and tissue oxygen gradients in the rat mesentery. Proc Natl Acad Sci USA 95: 6590–6595, 1998.
30. White RL and Wittenberg BA. NADH fluorescence of isolated ventricular myocytes: effects of pacing, myoglobin, and oxygen supply. Biophys J 65: 196–204, 1993.
31. Wilson DF, Ereciska M, Drown C, and Silver IA. The oxygen dependence of cellular energy metabolism. Arch Biochem Biophys 195: 485–493, 1979.[CrossRef][ISI][Medline]
32. Wittenberg BA and Wittenberg JB. Oxygen pressure gradients in isolated cardiac myocytes. J Biol Chem 260: 6548–6554, 1985.
33. Wittenberg JB. Myoglobin-facilitated oxygen diffusion: role of myoglobin in oxygen entry into muscle. Physiol Rev 50: 559–636, 1970.
34. Zuurbier CJ, van Iterson M, and Ince C. Functional heterogeneity of oxygen supply-consumption ratio in the heart. Cardiovasc Res 44: 488–497, 1999.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
