Skeletal Muscle Fiber-Specific Green Autofluorescence: Potential for Stem Cell Engraftment Artifacts
http://www.100md.com
《干细胞学杂志》
a Center for Cell and Gene Therapy,
b Department of Molecular and Cellular Biology, Integrated Microscopy Core, and
c Department of Pediatrics, Baylor College of Medicine, Houston, Texas, USA
Key Words. Skeletal muscle ? Autofluorescence ? Stem cell plasticity
Margaret A. Goodell, Ph.D., Center for Cell and Gene Therapy and Department of Pediatrics, Baylor College of Medicine, One Baylor Plaza, BCM N1050, Houston, Texas 77030, USA. Telephone: 713-798-1269; Fax: 713-798-1230; e-mail: goodell@bcm.tmc.edu
ABSTRACT
The use of green fluorescent protein (GFP), instead of ?-galactosidase as a marker in transgenic mice is becoming increasingly common. While GFP provides some advantages over histo- or immunohistochemical marker systems, particularly in terms of live imaging, some tissues have high levels of autofluorescence that can confound analysis. Such autofluorescence can be particularly problematic when using sensitive imaging techniques to look for rare events in a tissue. We became acutely aware of this while investigating potential engraftment of hematopoietic stem cells (HSCs) into skeletal muscle.
The concept of "transdifferentiation" of adult stem cells, particularly HSCs, into nonautochthonous cell types has received a great deal of interest recently due to the potential offered for cellular therapy . Generation of skeletal muscle from HSCs was one of the first areas of interest, but reported efficiencies vary by two orders of magnitude from ~0.05% , to ~3% . Contribution of wild-type cells to dystrophic muscle could potentially be therapeutic in the range of 1%-10%. Considering the enormous disparities observed in engraftment levels, it is vital to identify the baseline engraftment potential of bone-marrow-derived cells into skeletal muscle in order to determine whether bone marrow transplantation may be useful for the treatment of skeletal muscle disorders in the near future.
Although engraftment levels vary from laboratory to laboratory, within a research group engraftment levels remain constant. This suggests that the differences observed in engraftment levels may be due to the techniques used to identify the transdifferentiation event. The presence of the Y chromosome , GFP expression , and nuclear-localized ?-galactosidase expression have all been used to identify skeletal muscle fibers containing bone-marrow-derived cells. Thus far, experiments performed using bone marrow cells expressing GFP have resulted in the highest levels of muscle engraftment. It has been suggested that the lower engraftment levels observed in experiments using ?-galactosidase as a marker may be due to silencing of the lacZ gene. Here, while performing preliminary experiments to determine whether we could detect engraftment from GFP-positive cells in skeletal muscle, we observed a high level of fiber-type-specific green autofluorescence that could potentially be mistaken for GFP fluorescence.
MATERIALS AND METHODS
Skeletal Muscle Autofluorescence
Prior to embarking on experiments to examine the prevalence of hematopoietic stem cell derived engraftment into skeletal muscle, we wanted to ensure that GFP could be readily detected in mature muscle fibers from GFP mice, and that muscle fibers produced upon regeneration with myogenic stem cells from the labeled mice could be detected in a background of negative cells. Using the TA muscle, a small muscle on the lower hind leg of mice, we observed bright green fluorescent muscle fibers in wild-type mice that expressed enhanced GFP fluorescence under a chicken ?-actin promoter (GFP mice, Fig. 1A). To test GFP fluorescence in regenerating muscles, we used a standard regeneration assay, wherein myogenic stem cells derived from GFP mice were injected into injured muscles of normal GFP-negative mice. The injury was induced by injection of cardiotoxin, derived from cobra venom, which destroys mature muscle fibers, allowing robust regeneration from myogenic stem cells to occur . Mononuclear muscle cell preparations that contained myogenic stem cells were prepared from GFP mice and injected in this assay, resulting in a subset of bright green fibers, as shown in Figure 1B. Note the variety of fiber sizes in this panel and the wide spacing between them, compared with the uninjured muscle (Fig. 1A), indicating the ongoing regeneration. However, the negative controls, involving uninjured, uninjected muscles of normal GFP-negative mice that were prepared under identical conditions also displayed a distinct subset of muscle fibers that were similarly bright green, with the clear appearance of GFP expression (Fig. 1C and Fig. 1D). These fibers were consistently of a smaller diameter than adjacent non-fluorescent fibers. Since regenerating fibers tend to be smaller, this autofluorescence could suggest correlation with regeneration, even though no regeneration should be occurring in these negative-control mice, and no GFP-positive cells are present. Even in the sections from GFP mice (Fig. 1A), one can observe a subset of fibers that are a brighter green than others, and that also have a smaller diameter. The brighter green of these fibers presumably reflects the addition of GFP and autofluorescence.
Figure 1. GFP fluorescence and autofluorescence in skeletal muscle. A) Section of TA muscle from normal GFP transgenic mice (chicken ?-actin promoter). No image editing was performed for any of these images. All sections shown in this figure were from animals perfused with 1.5% PFA/0.5% glutaraldehyde prior to excision of the muscle and frozen sectioning. Sections are 10 μm. Photograph was taken for 5 milliseconds. B) Section of regenerating TA muscle injected with satellite cells derived from a GFP transgenic mouse (above). Photograph taken for 5 milliseconds. The TA of a normal GFP-negative wild-type mouse was injured by injection with cardiotoxin followed after 24 hours with GFP+ satellite cells. Two weeks later, the animals were sacrificed, perfused and the TA frozen sectioned. C) Cross-section of TA from a normal GFP-negative C57Bl/6 mouse after perfusion and frozen sectioning, performed identically as for the images in A and B. Photograph exposed for 75 milliseconds. D) Longitudinal sections of TA as in C. Exposure time 75 milliseconds.
To further characterize this autofluorescence phenomenon, we examined a variety of muscle types and found that a subset of muscle fibers consistently exhibited this bright green autofluorescence, although the prevalence varied among different muscle types. In the TA muscle, the average prevalence was around 6%, although this varied with the region of the muscle examined. This suggested to us that a specific biologic phenomenon underlied this autofluorescence, so we endeavored to determine what properties of the muscle fibers correlated with green autofluorescence and whether certain fixation conditions enhanced or diminished our ability to detect the fluorescence.
Regarding fixation conditions, we had initially followed protocols in the literature that suggested perfusion fixation in 1.5% PFA/0.5% glutaraldehyde . Muscles fixed under such conditions are shown in Figure 1. Fixation of muscle tissue after sectioning rather than by perfusion resulted in an overall increase in nonspecific background autofluorescence, which abolished the ability to distinguish a specific subset of autofluorescent fibers (not shown). In unfixed tissue, the distinct autofluorescence in specific fibers was still readily detectable, but the fluorescence was generally more punctate and concentrated at the edges of the fibers (Fig. 2E), making it slightly more difficult to identify. Glutaraldehyde was the essential component, as perfusion fixation with PFA alone resulted in staining most similar to unfixed tissue (not shown). Since perfusion fixation enhances the ability to detect the autofluorescent fibers, variations in the effectiveness of perfusion resulted in some animals appearing to have a higher prevalence of autofluorescence.
Figure 2. Type IIa muscle fibers exhibit autofluorescence. A through D and E through G are 10 μm serial sections through nonfixed TA muscle from a normal GFP-negative C57Bl/6 mouse. A) Section of TA muscle demonstrating readily identifiable autofluorescent fibers in a standard fluorescein channel. B) Adjacent section stained for NADH hydroxylase activity. The dark blue fibers have the highest levels of NADH hydroxylase, showing that autofluorescent fibers are oxidative. C) and D) Characterization of fiber type. A section adjacent to A/B was stained for myosin ATPase activity after preincubation at pH 10.4 (C), with dark black staining being characteristic of type IIa fibers, or after incubation at pH 4.0 (D), with dark staining indicative of type I fibers. Since no type I fibers were present in this section, another area of the TA muscle that contained these was analyzed in E-G. E) Section of TA muscle with autofluorescent fibers. F) Adjacent section stained for myosin ATPase activity at pH 10.4 for type IIa fibers. G) Adjacent section stained for myosin ATPase activity at pH 4.0. The dark fibers in this section are type 1, revealing that these fibers are not autofluorescent. Exposure time for all fluorescent panels was 75 milliseconds.
Specific Muscle Fibers Exhibit Autofluorescence
Since skeletal muscle is known to have several distinct fiber types, we sought to determine if the autofluorescence corresponded to one of these types. Oxidative fibers are distinguished from glycolytic fibers by their high level of NADH dehydrogenase activity , the first enzyme of the electron transport pathway. NADH dehydrogenase activity can be assessed by the addition of NADH and tetrazolium blue to skeletal muscle sections. The protons liberated from NADH reduce the tetrazolium blue, producing an intense purple stain. Figure 2 shows adjacent sections from unfixed skeletal muscle tissues for autofluorescence (Fig. 2A) and NADH activity (Fig. 2B). The autofluorescence is associated with the fibers containing the highest levels of NADH dehydrogenase activity, indicating that the autofluorescent fibers are oxidative.
Oxidative fibers can be subdivided into type IIa fast and type I slow twitch, which can be distinguished by the sensitivity of the myosin ATPase contained within these fibers to alkaline or acidic pH . Preincubation of skeletal muscle sections at an alkaline pH followed by staining for myosin ATPase activity results in a heavy staining of type IIa fibers, intermediate staining of type IIb fibers, and no staining of type I fibers. In contrast, preincubation at an acidic pH results in ATPase activity in only type I fibers. When TA muscle sections were preincubated at pH 10.4, the darkly stained type IIa fibers correlated with the green fluorescent fibers (compare Fig. 2A with 2C, and 2E with 2F). When sections were incubated at pH 4.0 (Fig. 2D and 2G), the dark-stained type I fibers were not autofluorescent (no type I fibers are present in Fig. 2D; compare Fig. 2E with 2G).
Strategies for Distinguishing GFP from Autofluorescence
Because autofluorescence closely resembles GFP fluorescence, as shown in Figure 1, we attempted to develop a standard method to distinguish autofluorescence from true GFP fluorescence. Variable levels of fluorescence between controls and GFP-expressing tissues were monitored during microscopy by recording the exposure times used for each photo as noted in the figure legends (pictures were not manipulated with image-editing software). As long as we utilized positive and negative controls, imaged alongside the experimental samples, this method of separating autofluorescence from GFP fluorescence was acceptable in experiments that used GFP-marked satellite cell incorporation into regenerating muscle fibers because satellite cells contribute numerous nuclei to the regenerating muscle fibers, thereby resulting in high levels of GFP expression. In contrast, it has been shown that bone marrow contributes many fewer nuclei to a single regenerating muscle fiber , rendering this method of distinguishing background fluorescence from GFP expression less reliable where contribution of true GFP-positive nuclei is rare.
Since all fluorophores have unique absorption and emission spectra, we investigated whether we could use confocal laser scanning microscopy (CLSM) to identify specific emission wavelengths with which to distinguish GFP fluorescence from autofluorescence. Our CLSM configuration excites a fluorophore using a 488-nm laser, and emissions are collected over a specified range in 11-nm increments. Thus, we scanned GFP+ and normal mouse TA muscle between 494 nm and 580 nm emission wavelengths. As shown in Figure 3A, the GFP fluorescence was brightest in a narrow range from 511 nm to 532 nm, whereas the autofluorescence appeared in this range, but also at higher wavelengths (Fig. 3B). This can be viewed graphically by measuring emission in a small area comprised of a single fiber and plotting the relative emission intensity versus wavelength, as shown in Figure 4. The overlay of the emission profiles of a bright GFP fiber and a bright autofluorescent fiber shows that both types of fluorescence are readily detected at around 520 nm, whereas the autofluorescence was much brighter above 560 nm. When this method was used to examine whether muscle fibers regenerated from GFP+ satellite cells (such as shown in Fig. 1B), we could readily distinguish GFP fluorescence from autofluorescence based on these typical profiles. These scans also demonstrate that judicious use of emission collection filters could at least reduce some background: a narrow bandpass filter optimized for GFP emission would reduce emissions coming from higher wavelengths derived from autofluorophores, whereas a longpass filter collecting a large range of wavelengths would be more prone to artifacts.
Figure 3. Scanning confocal microscopy of GFP+ and autofluorescent muscle. A) CLSM on GFP+ muscle was performed at the indicated wavelengths (numbers in white on each image). Sections from GFP-positive and negative mice were prepared as described for Figure 1. B) CLSM on GFP-negative cells. Note that the sensitivity of the detectors had to be increased to obtain sufficient signal, when scanning at such narrow wavelengths, on the GFP-negative tissue. This accounts for the somewhat grainy appearance in these unmanipulated images.
Figure 4. Graphical display of GFP fluorescence and autofluorescence intensity with wavelength. As indicated, the relative intensity of fluorescence from green fibers that were either GFP+ or autofluorescent is plotted against emission wavelength. The overlay shows that the minimum wavelength to detect autofluorescence is higher, as is the maximum range. Long-pass filters would, thus, allow substantially more autofluorescence to be detected than narrow band-pass filters.
DISCUSSION
This work was funded by Muscular Dystrophy Association grants to M.A.G. and K.A.J. M.A.G. is a Scholar of the Leukemia and Lymphoma Society, and K.A.J. was a Fellow of the Leukemia and Lymphoma Society. We thank Michael Mancini of the Baylor College of Medicine and Department of Molecular and Cellular Biology Integrated Microscopy Core for suggestions.
REFERENCES
Goodell MA. Stem-cell "plasticity": befuddled by the muddle. Curr Opin Hematol 2003;10:208–213.
Ferrari G, Cusella-De Angelis G, Coletta M et al. Muscle regeneration by bone marrow-derived myogenic progenitors. Science 1998;279:1528–1530.
Gussoni E, Soneoka Y, Strickland CD et al. Dystrophin expression in the mdx mouse restored by stem cell transplantation. Nature 1999;401:390–394.
LaBarge MA, Blau HM. Biological progression from adult bone marrow to mononucleate muscle stem cell to multinucleate muscle fiber in response to injury. Cell 2002;111:589–601.
Fukada S, Miyagoe-Suzuki Y, Tsukihara H et al. Muscle regeneration by reconstitution with bone marrow or fetal liver cells from green fluorescent protein-gene transgenic mice. J Cell Sci 2002;115:1285–1293.
Brazelton TR, Nystrom M, Blau HM. Significant differences among skeletal muscles in the incorporation of bone marrow-derived cells. Dev Biol 2003;262:64–74.
Guth L, Samaha FJ. Qualitative differences between actomyosin ATPase of slow and fast mammalian muscle. Exp Neuro 1969;25:138–152.
Jackson KA, Mi T, Goodell MA. Hematopoietic potential of stem cells isolated from murine skeletal muscle . Proc Natl Acad Sci USA 1999;96:14482–14486.
Allen DL, Harrison BC, Maass A et al. Cardiac and skeletal muscle adaptations to voluntary wheel running in the mouse. J Appl Physiol 2001;90:1900–1908.
Brooke MH, Kaiser KK. Muscle fiber types: how many and what kind? Arch Neurol 1970;23:369–379.
Billinton N, Knight AW. Seeing the wood through the trees: a review of techniques for distinguishing green fluorescent protein from endogenous autofluorescence. Anal Biochem 2001;291:175–197.
Weimann JM, Charlton CA, Brazelton TR et al. Contribution of transplanted bone marrow cells to Purkinje neurons in human adult brains. Proc Natl Acad Sci USA 2003;100:2088–2093.
Wagers AJ, Sherwood RI, Christensen JL et al. Little evidence for developmental plasticity of adult hematopoietic stem cells. Science 2002;297:2256–2259.
Brazelton TR, Rossi FM, Keshet GI et al. From marrow to brain: expression of neuronal phenotypes in adult mice. Science 2000;290:1775–1779.
Benson RC, Meyer RA, Zaruba ME et al. Cellular autofluorescence—is it due to flavins? J Histochem Cytochem 1979;27:44–48.
Orlic D, Kajstura J, Chimenti S et al. Bone marrow cells regenerate infarcted myocardium. Nature 2001;410:701–705.
Jackson KA, Majka SM, Wang H et al. Regeneration of ischemic cardiac muscle and vascular endothelium by adult stem cells. J Clin Invest 2001;107:1395–1402.
Quaini F, Urbanek K, Beltrami AP et al. Chimerism of the transplanted heart. N Engl J Med 2002;346:5–15.
Laflamme MA, Myerson D, Saffitz JE et al. Evidence for cardiomyocyte repopulation by extracardiac progenitors in transplanted human hearts. Circ Res 2002;90:634–640.
Castro RF, Jackson KA, Goodell MA et al. Failure of bone marrow cells to transdifferentiate into neural cells in vivo. Science 2002;297:1299.
Mezey E, Chandross KJ, Harta G et al. Turning blood into brain: cells bearing neuronal antigens generated in vivo from bone marrow. Science 2000;290:1779–1782.(Kathyjo A. Jacksona,c, D.)
b Department of Molecular and Cellular Biology, Integrated Microscopy Core, and
c Department of Pediatrics, Baylor College of Medicine, Houston, Texas, USA
Key Words. Skeletal muscle ? Autofluorescence ? Stem cell plasticity
Margaret A. Goodell, Ph.D., Center for Cell and Gene Therapy and Department of Pediatrics, Baylor College of Medicine, One Baylor Plaza, BCM N1050, Houston, Texas 77030, USA. Telephone: 713-798-1269; Fax: 713-798-1230; e-mail: goodell@bcm.tmc.edu
ABSTRACT
The use of green fluorescent protein (GFP), instead of ?-galactosidase as a marker in transgenic mice is becoming increasingly common. While GFP provides some advantages over histo- or immunohistochemical marker systems, particularly in terms of live imaging, some tissues have high levels of autofluorescence that can confound analysis. Such autofluorescence can be particularly problematic when using sensitive imaging techniques to look for rare events in a tissue. We became acutely aware of this while investigating potential engraftment of hematopoietic stem cells (HSCs) into skeletal muscle.
The concept of "transdifferentiation" of adult stem cells, particularly HSCs, into nonautochthonous cell types has received a great deal of interest recently due to the potential offered for cellular therapy . Generation of skeletal muscle from HSCs was one of the first areas of interest, but reported efficiencies vary by two orders of magnitude from ~0.05% , to ~3% . Contribution of wild-type cells to dystrophic muscle could potentially be therapeutic in the range of 1%-10%. Considering the enormous disparities observed in engraftment levels, it is vital to identify the baseline engraftment potential of bone-marrow-derived cells into skeletal muscle in order to determine whether bone marrow transplantation may be useful for the treatment of skeletal muscle disorders in the near future.
Although engraftment levels vary from laboratory to laboratory, within a research group engraftment levels remain constant. This suggests that the differences observed in engraftment levels may be due to the techniques used to identify the transdifferentiation event. The presence of the Y chromosome , GFP expression , and nuclear-localized ?-galactosidase expression have all been used to identify skeletal muscle fibers containing bone-marrow-derived cells. Thus far, experiments performed using bone marrow cells expressing GFP have resulted in the highest levels of muscle engraftment. It has been suggested that the lower engraftment levels observed in experiments using ?-galactosidase as a marker may be due to silencing of the lacZ gene. Here, while performing preliminary experiments to determine whether we could detect engraftment from GFP-positive cells in skeletal muscle, we observed a high level of fiber-type-specific green autofluorescence that could potentially be mistaken for GFP fluorescence.
MATERIALS AND METHODS
Skeletal Muscle Autofluorescence
Prior to embarking on experiments to examine the prevalence of hematopoietic stem cell derived engraftment into skeletal muscle, we wanted to ensure that GFP could be readily detected in mature muscle fibers from GFP mice, and that muscle fibers produced upon regeneration with myogenic stem cells from the labeled mice could be detected in a background of negative cells. Using the TA muscle, a small muscle on the lower hind leg of mice, we observed bright green fluorescent muscle fibers in wild-type mice that expressed enhanced GFP fluorescence under a chicken ?-actin promoter (GFP mice, Fig. 1A). To test GFP fluorescence in regenerating muscles, we used a standard regeneration assay, wherein myogenic stem cells derived from GFP mice were injected into injured muscles of normal GFP-negative mice. The injury was induced by injection of cardiotoxin, derived from cobra venom, which destroys mature muscle fibers, allowing robust regeneration from myogenic stem cells to occur . Mononuclear muscle cell preparations that contained myogenic stem cells were prepared from GFP mice and injected in this assay, resulting in a subset of bright green fibers, as shown in Figure 1B. Note the variety of fiber sizes in this panel and the wide spacing between them, compared with the uninjured muscle (Fig. 1A), indicating the ongoing regeneration. However, the negative controls, involving uninjured, uninjected muscles of normal GFP-negative mice that were prepared under identical conditions also displayed a distinct subset of muscle fibers that were similarly bright green, with the clear appearance of GFP expression (Fig. 1C and Fig. 1D). These fibers were consistently of a smaller diameter than adjacent non-fluorescent fibers. Since regenerating fibers tend to be smaller, this autofluorescence could suggest correlation with regeneration, even though no regeneration should be occurring in these negative-control mice, and no GFP-positive cells are present. Even in the sections from GFP mice (Fig. 1A), one can observe a subset of fibers that are a brighter green than others, and that also have a smaller diameter. The brighter green of these fibers presumably reflects the addition of GFP and autofluorescence.
Figure 1. GFP fluorescence and autofluorescence in skeletal muscle. A) Section of TA muscle from normal GFP transgenic mice (chicken ?-actin promoter). No image editing was performed for any of these images. All sections shown in this figure were from animals perfused with 1.5% PFA/0.5% glutaraldehyde prior to excision of the muscle and frozen sectioning. Sections are 10 μm. Photograph was taken for 5 milliseconds. B) Section of regenerating TA muscle injected with satellite cells derived from a GFP transgenic mouse (above). Photograph taken for 5 milliseconds. The TA of a normal GFP-negative wild-type mouse was injured by injection with cardiotoxin followed after 24 hours with GFP+ satellite cells. Two weeks later, the animals were sacrificed, perfused and the TA frozen sectioned. C) Cross-section of TA from a normal GFP-negative C57Bl/6 mouse after perfusion and frozen sectioning, performed identically as for the images in A and B. Photograph exposed for 75 milliseconds. D) Longitudinal sections of TA as in C. Exposure time 75 milliseconds.
To further characterize this autofluorescence phenomenon, we examined a variety of muscle types and found that a subset of muscle fibers consistently exhibited this bright green autofluorescence, although the prevalence varied among different muscle types. In the TA muscle, the average prevalence was around 6%, although this varied with the region of the muscle examined. This suggested to us that a specific biologic phenomenon underlied this autofluorescence, so we endeavored to determine what properties of the muscle fibers correlated with green autofluorescence and whether certain fixation conditions enhanced or diminished our ability to detect the fluorescence.
Regarding fixation conditions, we had initially followed protocols in the literature that suggested perfusion fixation in 1.5% PFA/0.5% glutaraldehyde . Muscles fixed under such conditions are shown in Figure 1. Fixation of muscle tissue after sectioning rather than by perfusion resulted in an overall increase in nonspecific background autofluorescence, which abolished the ability to distinguish a specific subset of autofluorescent fibers (not shown). In unfixed tissue, the distinct autofluorescence in specific fibers was still readily detectable, but the fluorescence was generally more punctate and concentrated at the edges of the fibers (Fig. 2E), making it slightly more difficult to identify. Glutaraldehyde was the essential component, as perfusion fixation with PFA alone resulted in staining most similar to unfixed tissue (not shown). Since perfusion fixation enhances the ability to detect the autofluorescent fibers, variations in the effectiveness of perfusion resulted in some animals appearing to have a higher prevalence of autofluorescence.
Figure 2. Type IIa muscle fibers exhibit autofluorescence. A through D and E through G are 10 μm serial sections through nonfixed TA muscle from a normal GFP-negative C57Bl/6 mouse. A) Section of TA muscle demonstrating readily identifiable autofluorescent fibers in a standard fluorescein channel. B) Adjacent section stained for NADH hydroxylase activity. The dark blue fibers have the highest levels of NADH hydroxylase, showing that autofluorescent fibers are oxidative. C) and D) Characterization of fiber type. A section adjacent to A/B was stained for myosin ATPase activity after preincubation at pH 10.4 (C), with dark black staining being characteristic of type IIa fibers, or after incubation at pH 4.0 (D), with dark staining indicative of type I fibers. Since no type I fibers were present in this section, another area of the TA muscle that contained these was analyzed in E-G. E) Section of TA muscle with autofluorescent fibers. F) Adjacent section stained for myosin ATPase activity at pH 10.4 for type IIa fibers. G) Adjacent section stained for myosin ATPase activity at pH 4.0. The dark fibers in this section are type 1, revealing that these fibers are not autofluorescent. Exposure time for all fluorescent panels was 75 milliseconds.
Specific Muscle Fibers Exhibit Autofluorescence
Since skeletal muscle is known to have several distinct fiber types, we sought to determine if the autofluorescence corresponded to one of these types. Oxidative fibers are distinguished from glycolytic fibers by their high level of NADH dehydrogenase activity , the first enzyme of the electron transport pathway. NADH dehydrogenase activity can be assessed by the addition of NADH and tetrazolium blue to skeletal muscle sections. The protons liberated from NADH reduce the tetrazolium blue, producing an intense purple stain. Figure 2 shows adjacent sections from unfixed skeletal muscle tissues for autofluorescence (Fig. 2A) and NADH activity (Fig. 2B). The autofluorescence is associated with the fibers containing the highest levels of NADH dehydrogenase activity, indicating that the autofluorescent fibers are oxidative.
Oxidative fibers can be subdivided into type IIa fast and type I slow twitch, which can be distinguished by the sensitivity of the myosin ATPase contained within these fibers to alkaline or acidic pH . Preincubation of skeletal muscle sections at an alkaline pH followed by staining for myosin ATPase activity results in a heavy staining of type IIa fibers, intermediate staining of type IIb fibers, and no staining of type I fibers. In contrast, preincubation at an acidic pH results in ATPase activity in only type I fibers. When TA muscle sections were preincubated at pH 10.4, the darkly stained type IIa fibers correlated with the green fluorescent fibers (compare Fig. 2A with 2C, and 2E with 2F). When sections were incubated at pH 4.0 (Fig. 2D and 2G), the dark-stained type I fibers were not autofluorescent (no type I fibers are present in Fig. 2D; compare Fig. 2E with 2G).
Strategies for Distinguishing GFP from Autofluorescence
Because autofluorescence closely resembles GFP fluorescence, as shown in Figure 1, we attempted to develop a standard method to distinguish autofluorescence from true GFP fluorescence. Variable levels of fluorescence between controls and GFP-expressing tissues were monitored during microscopy by recording the exposure times used for each photo as noted in the figure legends (pictures were not manipulated with image-editing software). As long as we utilized positive and negative controls, imaged alongside the experimental samples, this method of separating autofluorescence from GFP fluorescence was acceptable in experiments that used GFP-marked satellite cell incorporation into regenerating muscle fibers because satellite cells contribute numerous nuclei to the regenerating muscle fibers, thereby resulting in high levels of GFP expression. In contrast, it has been shown that bone marrow contributes many fewer nuclei to a single regenerating muscle fiber , rendering this method of distinguishing background fluorescence from GFP expression less reliable where contribution of true GFP-positive nuclei is rare.
Since all fluorophores have unique absorption and emission spectra, we investigated whether we could use confocal laser scanning microscopy (CLSM) to identify specific emission wavelengths with which to distinguish GFP fluorescence from autofluorescence. Our CLSM configuration excites a fluorophore using a 488-nm laser, and emissions are collected over a specified range in 11-nm increments. Thus, we scanned GFP+ and normal mouse TA muscle between 494 nm and 580 nm emission wavelengths. As shown in Figure 3A, the GFP fluorescence was brightest in a narrow range from 511 nm to 532 nm, whereas the autofluorescence appeared in this range, but also at higher wavelengths (Fig. 3B). This can be viewed graphically by measuring emission in a small area comprised of a single fiber and plotting the relative emission intensity versus wavelength, as shown in Figure 4. The overlay of the emission profiles of a bright GFP fiber and a bright autofluorescent fiber shows that both types of fluorescence are readily detected at around 520 nm, whereas the autofluorescence was much brighter above 560 nm. When this method was used to examine whether muscle fibers regenerated from GFP+ satellite cells (such as shown in Fig. 1B), we could readily distinguish GFP fluorescence from autofluorescence based on these typical profiles. These scans also demonstrate that judicious use of emission collection filters could at least reduce some background: a narrow bandpass filter optimized for GFP emission would reduce emissions coming from higher wavelengths derived from autofluorophores, whereas a longpass filter collecting a large range of wavelengths would be more prone to artifacts.
Figure 3. Scanning confocal microscopy of GFP+ and autofluorescent muscle. A) CLSM on GFP+ muscle was performed at the indicated wavelengths (numbers in white on each image). Sections from GFP-positive and negative mice were prepared as described for Figure 1. B) CLSM on GFP-negative cells. Note that the sensitivity of the detectors had to be increased to obtain sufficient signal, when scanning at such narrow wavelengths, on the GFP-negative tissue. This accounts for the somewhat grainy appearance in these unmanipulated images.
Figure 4. Graphical display of GFP fluorescence and autofluorescence intensity with wavelength. As indicated, the relative intensity of fluorescence from green fibers that were either GFP+ or autofluorescent is plotted against emission wavelength. The overlay shows that the minimum wavelength to detect autofluorescence is higher, as is the maximum range. Long-pass filters would, thus, allow substantially more autofluorescence to be detected than narrow band-pass filters.
DISCUSSION
This work was funded by Muscular Dystrophy Association grants to M.A.G. and K.A.J. M.A.G. is a Scholar of the Leukemia and Lymphoma Society, and K.A.J. was a Fellow of the Leukemia and Lymphoma Society. We thank Michael Mancini of the Baylor College of Medicine and Department of Molecular and Cellular Biology Integrated Microscopy Core for suggestions.
REFERENCES
Goodell MA. Stem-cell "plasticity": befuddled by the muddle. Curr Opin Hematol 2003;10:208–213.
Ferrari G, Cusella-De Angelis G, Coletta M et al. Muscle regeneration by bone marrow-derived myogenic progenitors. Science 1998;279:1528–1530.
Gussoni E, Soneoka Y, Strickland CD et al. Dystrophin expression in the mdx mouse restored by stem cell transplantation. Nature 1999;401:390–394.
LaBarge MA, Blau HM. Biological progression from adult bone marrow to mononucleate muscle stem cell to multinucleate muscle fiber in response to injury. Cell 2002;111:589–601.
Fukada S, Miyagoe-Suzuki Y, Tsukihara H et al. Muscle regeneration by reconstitution with bone marrow or fetal liver cells from green fluorescent protein-gene transgenic mice. J Cell Sci 2002;115:1285–1293.
Brazelton TR, Nystrom M, Blau HM. Significant differences among skeletal muscles in the incorporation of bone marrow-derived cells. Dev Biol 2003;262:64–74.
Guth L, Samaha FJ. Qualitative differences between actomyosin ATPase of slow and fast mammalian muscle. Exp Neuro 1969;25:138–152.
Jackson KA, Mi T, Goodell MA. Hematopoietic potential of stem cells isolated from murine skeletal muscle . Proc Natl Acad Sci USA 1999;96:14482–14486.
Allen DL, Harrison BC, Maass A et al. Cardiac and skeletal muscle adaptations to voluntary wheel running in the mouse. J Appl Physiol 2001;90:1900–1908.
Brooke MH, Kaiser KK. Muscle fiber types: how many and what kind? Arch Neurol 1970;23:369–379.
Billinton N, Knight AW. Seeing the wood through the trees: a review of techniques for distinguishing green fluorescent protein from endogenous autofluorescence. Anal Biochem 2001;291:175–197.
Weimann JM, Charlton CA, Brazelton TR et al. Contribution of transplanted bone marrow cells to Purkinje neurons in human adult brains. Proc Natl Acad Sci USA 2003;100:2088–2093.
Wagers AJ, Sherwood RI, Christensen JL et al. Little evidence for developmental plasticity of adult hematopoietic stem cells. Science 2002;297:2256–2259.
Brazelton TR, Rossi FM, Keshet GI et al. From marrow to brain: expression of neuronal phenotypes in adult mice. Science 2000;290:1775–1779.
Benson RC, Meyer RA, Zaruba ME et al. Cellular autofluorescence—is it due to flavins? J Histochem Cytochem 1979;27:44–48.
Orlic D, Kajstura J, Chimenti S et al. Bone marrow cells regenerate infarcted myocardium. Nature 2001;410:701–705.
Jackson KA, Majka SM, Wang H et al. Regeneration of ischemic cardiac muscle and vascular endothelium by adult stem cells. J Clin Invest 2001;107:1395–1402.
Quaini F, Urbanek K, Beltrami AP et al. Chimerism of the transplanted heart. N Engl J Med 2002;346:5–15.
Laflamme MA, Myerson D, Saffitz JE et al. Evidence for cardiomyocyte repopulation by extracardiac progenitors in transplanted human hearts. Circ Res 2002;90:634–640.
Castro RF, Jackson KA, Goodell MA et al. Failure of bone marrow cells to transdifferentiate into neural cells in vivo. Science 2002;297:1299.
Mezey E, Chandross KJ, Harta G et al. Turning blood into brain: cells bearing neuronal antigens generated in vivo from bone marrow. Science 2000;290:1779–1782.(Kathyjo A. Jacksona,c, D.)