Expression of muscle capillary is affected by hypoxia F. M. HANSEN-SMITH,
L. H. BLACKWELL,
alkaline phosphatase
AND
G. R. JOSWIAK
Department of Biological Sciences and Office of Computer Services, 4401; and Department of Pharmacology and Physiology, University Detroit, Michigan 48207 HANSEN-SMITH, F. M.,L.H. BLACKWELL,AND G. R.JosWIAK. Expressionof musclecapillary alkaline phosphataseis affected by hypoxia. J. Appl. Physiol. 73(2): 776-780,1992.-Hyp-
oxia stimulates angiogenesisin some microvascular beds, but no clear angiogenic effect of hypoxia has yet been demonstrated in adult skeletal muscle. In this study the distribution of alkaline phosphatase (APase) was compared with a novel microvascular marker, Griffonia simplicifolia I (GSI), to determine whether the respective markers were expressedby muscle capillaries during hypoxic conditions and to probe for the presence or absenceof angiogenesisin responseto short-term hypoxia. Mice were exposedto normobaric 8% oxygen for 7 or 21 days. Capillary density in the red and white areas of the gastrocnemius musclewasdetermined with the use of a double-labeling procedure for both APase and fluorescently tagged GSI. Little change in capillary density was found. Focal reductions in APase activity were observedwithin 1 wk of hypoxia, but no changeswere observedin GSI binding. In controls, 74 and 92% of red and white muscle capillaries, respectively, were APase positive. This percentage declined to 60% in red and 43% in white muscleafter 21 days of hypoxia. The results indicate that APase expression is labile under certain conditions and warrant a cautious approach to using the enzyme as a marker. Binding of the GSI lectin to musclecapillaries appearedto be unchanged by the exposure to hypoxia, indicating stability of this marker system. No significant change in the number of capillaries around individual muscle fibers was evident at 21 days when GSI was used to detect capillaries. These results confirm the absenceof hypoxia-induced angiogenesisin muscle capillaries during the time period studied. skeletal muscle; endothelium; lectins; Griffonia angiogenesis
simplicifolia
I;
MICROVASCULAR BED may be exposed to hypoxic conditions as a result of environmental conditions, hematologic changes, or hemodynamic alterations. Whereas hypoxia has been shown to stimulate angiogenesis directly or indirectly in embryos (1, 13, 18) and in vitro (21), it is less clear whether hypoxia has an angiogenic influence on the skeletal muscle microcirculation. The effects of hypoxia on the capillary density of skeletal muscle have been previously studied, with a number of different experimental models (2-4, 12, 20, 22, 23). SevTHE
776
Oakland University, Rochester of Detroit School of Dentistry,
48309-
era1 studies have suggested that the density of capillaries per unit area is increased, possibly as a result of angiogenesis. However, this is explained in part by the reduction in fiber diameter that may occur in some muscle fibers. The quantitation of intramuscular capillaries often relies on the detection of the capillaries by histochemical means. However, only those capillaries that express the marker will be detected by this method. Recent studies in this laboratory suggest that at least one histochemical endothelial cell marker, alkaline phosphatase (APase), is poorly expressed by capillaries during postnatal growth (8). We postulated that enzymes such as APase might not be expressed initially by endothelial cells during hypoxia-induced angiogenesis in the adult. If this were the case, new capillaries would be present but not histochemitally detectable until they were mature enough to express APase, giving the appearance of a delayed angiogenie response. We therefore have probed for the presence or absence of hypoxia-induced angiogenesis and examined the influence of hypoxia on the expression of APase in skeletal muscle of mice. The lectin Griffonia simplicifolia I (GSI) was employed as an alternative marker for capillaries because it has previously been shown to detect all capillaries, including developing capillaries in rat muscle (7-9). Binding sites for this lectin (terminal a-galactosyl groups) have been shown to be expressed by capillaries during postnatal growth, before the differentiation of the APase system in capillaries of skeletal muscle (8). In addition, the lectin has been shown to bind to the tips of sprouting capillaries during developmental angiogenesis in the retina (6). METHODS
Female C,,Bl, mice [30 t 1 (SE) g] from a colony at the University of Detroit School of Dentistry were placed in cages constructed with membranes semipermeable to oxygen (13). Previous studies have established that the ambient oxygen concentration can be adjusted by varying the total body weight of animals within the cage (13). We used five animals per cage, giving a total of 150 g per cage. This resulted in an oxygen concentration of 8%. Experi-
0161-7567192 $2.00 Copyright 0 1992 the American Physiological Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CAPILLARY
ALKALINE
PHOSPHATASE
mental groups of n = 5 were maintained in the cages for 7 or 21 days and were given standard lab chow and water ad libitum. Bedding in the cages was changed twice per week. Normoxic age controls were exposed to ambient air in standard polypropylene cages of the same size but lacking the semipermeable membrane. After 7 or 21 days the control and experimental mice were killed by cervical dislocation. Blood samples for hematocrit determinations were then taken from the axillary vein. The efficacy of the experimental protocol was monitored by the hematocrits, which were significantly elevated after 7 and 21 days (normoxic controls 44 f 0.3 and 43 + 1.1, respectively; hypoxic 55.2 + 0.9 and 67.0 f 0.9). The gastrocnemius muscle was removed and quickfrozen in a slurry of dry ice-isopentane. Transverse cryostat sections (12 pm thick) were made from the muscle at its widest girth. The following histochemical procedures were carried out: APase activity was demonstrated by the method employed in our previous study (8). The reaction buffer contained 1.7 g/l, MgSO, * 7H,O, and 3.8 g/l NazB,O, . lOH,O. Before the reaction, 6 mg 5bromo-4chloro-3-indoxyl phosphate, toluidine salt, and 30 mg nitroblue tetrazolium were added to 30 ml of the buffer. The pH was adjusted to 9.2 with 1% boric acid. Unfixed sections were incubated for 30 min at 37°C in the reaction mixture and then rinsed in phosphate-buffered saline. Succinate dehydrogenase (SDH) activity was demonstrated by the method of Nachlas et al. (17). Rhodamine-GSI lectin binding was carried out as described previously (7-g), except that GSI (Vector) was used instead of the GSI-B, isomer. The lectin was diluted to 25 rglml with physiological saline containing 0.001 M Nacetylgalactosamine to eliminate possible nonvascular binding sites. A double-staining procedure was used for colocalizing lectin binding and APase activity, with a concentration of 40 pg rhodamine-GSI/ml (9). The higher lectin concentration facilitated visualization of lectin binding sites when viewed with a low level of transmitted light in combination with epifluorescence. Double-labeled sections used for quantitation were reacted for 30 min. Additional sections of some muscles were also reacted for 60 min. All of the sections were mounted in a water-soluble mountant without prior dehydration. The sections were examined with a Nikon Optiphot microscope equipped with optics for both transmitted and epifluorescent illumination. All quantitative analyses were carried out on photomicrographs taken at a magnification of X200. The capillarity of the muscles was determined from preselected “red” and “white” regions of each muscle, with the sections reacted for SDH as a reference. The same regions of muscle were selected in both control and experimental samples. The deepest, or most medial, region, containing fibers that stained intensely for SDH, was classified as red, whereas the most peripheral region, containing larger fibers that were only faintly stained for SDH, was classified as white. The transitional intermediate region was not sampled. Capillary densities were determined from photomicrographs. Two photographs of capillaries were taken from nonoverlapping fields in the center of the red region. Three photographs were taken at equally spaced locations at the periphery of the muscle to sample the white
IN HYPOXIA
FIG. 1. Combined alkaline phosphatase (APase) staining and rhodamine Criffonio simplicifoh I (GSI) binding in red (A, B, E, and F) and white (C, 0, G, and H) gastrocnemius of normoxic control (A-D) and 21-day hypoxic mice (E-H) are shown with transmitted light only (A, C, E, and G) or transmitted light simultaneously with epifluorescence (I?, D, F, and H). Examples of capillaries positive for APase, showing no GSI, are noted by solid arrows. These were counted as APase-positive capillaries. APase-negative GSI-positive capillaries are noted by clear arrows. These were counted as APase-negative capillaries. Capillaries positive for both APase and GSI are noted by both solid and clear arrows. These were counted only as APase-positive capillaries. Bar, 20 pm.
region. Each field was photographed twice, first with transmitted light to reveal APase activity. The identical field was then rephotographed with epifluorescent illumination to reveal GSI binding sites. The APase reaction product was visible by epifluorescent illumination, and some of the GSI binding was obscured in capillaries having large amounts of APase reaction product. Some capillaries could be detected by both methods and some were only positive for GSI (see Fig. 1). All microvessels surrounding the muscle fibers were counted unless they were large enough to reveal a lumen in the APase method or were approximately double that of the capillaries in the area. Criteria to specifically select capillaries on the basis of size (i.e., -5 pm) were impossible to set because of the intrinsic differences in APase staining between the red and white regions of controls and because of the variations that occurred as a result of the experimental treatment. The total number of APase-positive capillaries per
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
778
CAPILLARY
ALKALINE
PHOSPHATASE
field was determined from the first photograph. In the epifluorescence photograph only those GSI-positive capillaries that were completely negative for APase were counted. All sites counted in the first photograph were marked on the second photograph to avoid duplication of counts. The total number of capillaries per field (0.18 mm’), an index of capillary density, was calculated as the sum of the number of APase-positive capillaries and GSI-positive (APase-negative) capillaries in each field. To detect whether angiogenesis occurred as a result of 21 days of hypoxia, the number of capillaries around individual muscle fibers (CAF) in the red and white region was determined from double-labeled samples and from adjacent sections exposed to rhodamine-GSI alone. The sampling protocol described previously (7, 8) was used. Briefly, this consisted of using a cross in the eyepiece to randomly select four nonadjacent fibers at a time. This was done with the rhodamine filter blocked to eliminate any bias. Then the rhodamine filter was replaced and the number of GSI-positive capillaries around each of the selected fibers was counted. Capillaries in a total of eight fields (32 fibers) each were counted in the red and white regions, respectively, for each muscle. The CAF was determined by the same protocol for the double-labeled samples. A low level of transmitted light was used in combination with the epifluorescence to better visualize all capillaries. Analysis of variance for these data was conducted with the SPSSX version 4.0 statistical package. Percentages were analyzed statistically with the arsine transformation to obtain a normal distribution (24). RESULTS
AND
DISCUSSION
In this study double-labeled sections were used to facilitate quantitation of capillaries at identical sites within the muscle, with two different histological methods to visualize capillaries. This is the first time that a histochemical method of detecting capillaries in muscle has been critically evaluated for its accuracy during an experimental protocol such as hypoxia. Capillary densities determined by the double-labeling technique were greater than those determined with APase alone, both in the control and experimental muscles (Table 1). APase-positive capillaries constituted 74 t 2 and 72 t 1% for 7- and 2Lday controls in the red region and 91 t 3 and 88 t 5% of the total capillaries per field for the white region of controls. These data in control muscle concur with our previously reported studies in rats, demonstrating that APase is not uniformly expressed by capillaries throughout the microvascular bed (8,9). Capillary density determined from the double-labeled sections was significantly increased in the white region of hypoxic muscle after 7 days (Table 1), but no statistically significant differences were found for capillary densities in the other groups. APase activity associated with endothelial cells, as demonstrated by the intensity and distribution of reaction product, was clearly affected by the hypoxic environment. An overall diminution of the density of reaction product was visually evident in some capillaries within the first week. This was more conspicuous in the white region than in the red region. In some areas
IN HYPOXIA
TABLE 1. Capillarity of gastrocnemius muscle from normoxic and hypoxic mice Normoxic Red
Control
Hypoxic
White
Red
White
7 days Total capillary density % APase-positive 2 capillaries
258.8k22.3 74k2
66.4t3.1
290.5k13.8
92k3
75t2
78.2t4.0* 67k4-f
21 days Total capillary density % APase-positive capillaries
257.8t13.4 72kl
56.4zk6.2 98t5
255.8k32.6 60+8$
73.3t5.9 43*10$
Values are means + SE. Total capillary density, total number of capillaries per field (0.18 mm2) = sum of APase-positive capillaries + GSI-positive (APase-negative) capillaries. Significantly different from normoxic age control white muscle: * P = 0.05; t P = 0.003 (arcsine transformation); § P = 0.008 (arcsine transformation). Significantly different from normoxic age control red muscle: $ P = 0.024 (arcsine transformation).
the APase was significantly reduced but still detectable, whereas no reaction product was detectable in some areas. Extending the incubation times up to 60 min failed to generate detectable reaction product in the negative areas, suggesting that the APase activity was significantly reduced or absent in these areas of the hypoxic muscle (Fig. 1). In contrast to the results with APase, there was no detectable diminution of GSI binding to capillaries (Fig. 2). When the double-labeled sections were examined, it was clear that GSI-positive capillaries were present in regions in which APase activity was reduced or absent (Fig. 1). These double-labeled sections further revealed that the capillary APase in all regions of the muscle, not just the white regions, was affected by the hypoxia. However, in the red regions the diminution of APase was less striking visually because it did not occur in conspicuous large patches, as it did in the white region. The percentage of APase-positive capillaries in the red region declined significantly between 7 and 21 days of hypoxia (72 t 1 vs. 60 t 8%). The percentage of APase-positive capillaries in the white region was significantly lower than controls at 7 days (67 t 4%) and was reduced further at 21 days (43 t 10%). All counts of APase-labeled capillaries included any capillaries that could be detected, not just those that were strongly stained (see Fig. 1). The expression of APase by capillaries was thus diminished earlier and more extensively in the white region than in the red region of the muscle. These data indicate that the response of muscle capillaries to hypoxia is heterogeneous in different metabolic regions of the muscle. We examined the muscles for quantitative evidence of hypoxia-induced angiogenesis. Our original hypothesis predicted that, if angiogenesis occurred, new capillaries might fail to express APase initially. Capillary density is not a reliable index for detecting angiogenesis, because both capillary numbers and muscle fiber size may influence the resulting data. Therefore, we probed for an absolute change in CAF. No significant difference in CAF
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CAPIUARY
ALKALKNE
PHGSHIATASE
IN HYPOXIA
779
FIG. 2, Rhudamiae-GSI binding to Cepib&S irr red (A and 0 and white (B and D) gbrocnemius of normoxie controls (A and B) and 21.&y (C and D) hypoxic mice, shown with epifluoreecent illuminetiorr only. Bar, 20 pm.
was found in either the red or white region of the muscle after 21 days (Table 2), arguing against hypoxia-induced a~~o~e~~si~ within this time period. the same when C&I was used alone an beled samples. These results confirm those of most other with different models to induce hypoxia. eloping blood vessels may be stimulated by cle capillaries in adult animals appear to e remarkably resistant ti this putative angiogenic stimulus. y there was little quantitative change in or absolute number of CAF as a result of er, the enzymatic activity of the capillary
APase was changed, although we consider this an unlikely explanation for our observations. Our reaction em-
Red
CAP r(GSI-positive cap&&s)*
white
Red
Wbik
4.WM.11
2.6SM.09
4.&+0.11
2.79iO.09
4.39*0.10
2.71110.10
4.7&u.l4
2.63+0.15
CAF (total capillari&t
VaIues are means f SE. * Number of capilJariea around 32 rsndomly individual muscle fib in each retion Der szrou~ was determined from +Z?J-labeled cryostat t3ectiona. t-To&t nu&& of capillaries around 32 randomly selected individual muscle fibers was determin in each region per group. AFase- and GSI-positive (APase-negativej capillaries were counted from double-labeled cryostat s&ions,
ployed an indoxyl phosphate as a sub&rate. In the past, many different substrates, including ,&glycerophosphate used for demon-
es suggesting that APase activity histochemieal marker for quantitacapillarity (8,9). There is heterogeenzymes even under co study and our previous reports (8,9) illustrate. Growth and physiological status of the muscle may further affect the expression of this enzyme (and possibly other endothelial enzymes), Therefore, the further use of APase as a capillary marker should be dated by alternative markers OFby direct visualization in sections under both co ns (14). The present provide other useful information concerning endotbelial cell metabolism. Although further co~~rrna~~on is needed, the possibility al loss of APase activity may prove to be a identifying isebemic and/or bypoxic loci within skeletal or cardiac muscle. We gratefully acknowledge the technical assistance of Laura Morris and Paul Maximuke and the secretarial assistance of Rita Per+ This study was supp gmnt fmm merican Hea c&ion of Michigan, a at Sciences arch Grant. Research Committee at Oakland University. Address for reprint requeeta: F. M. Nan~n-Smith, Dept.. ofBiological Sciences, Oakland University, Rochester, MI 48309-4401, Received 21 December 1990; accepted in final form 5 February 1992, RENCES 1. AME, T. H., A, 6. GUYTON, J. P. hioNTANI, I-I. t. L&DsAY, AND K. A. STANEK. whole body structural vascular ad&@&ion to pro-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
780
CAPILLARY
ALKALINE
longed hypoxia in chick embryos. Am. J. Physiol.
252 (Heart
PHOSPHATASE Circ.
Physiol. 21): H1228-H1234, 1987. 2. BANCHERO, N., S. R. KAYAR, AND
A. J. LECHNER. Increased capillarity in skeletal muscle of growing guinea pigs acclimated to cold and hypoxia. Respir. Physiol. 62: 245-255, 1985. 3. BENDER, P. R., A. TUCKER, AND T. A. TWIETMEYER. Skeletal muscle capillarity during recovery from chronic hypoxic exposure. Microcirc. Endothelium Lymphatics 4. CASSIN, S., R. D. GILBERT, C.
1: 71-85,
1984.
E. BUNNELL, AND E. M. JOHNSON. Capillary development during exposure to chronic hypoxia. Am. J.
Physiol. 220: 448-451, 1971. 5. CHRISTIE, K. N., AND C. THOMSON.
brane-bound
Effect of ischaemia on memhydrolases in rat soleus capillaries. Acta Histochem.
Cytochem. 22: 47-65,1989. 6. HANSEN-SMITH, F. M. Postnatal development sels (Abstract). FASEB J. 4: A1897, 1990. 7. HANSEN-SMITH, F. M., K. BANKER, L. MORRIS,
of retinal microves-
AND G. JOSWIAK. Alternative histochemical markers for skeletal muscle capillaries: a statistical comparison among .three muscles. Microuasc. Res. In press. 8. HANSEN-SMITH, F. M., L. WATSON, AND G. R. JOSWIAK. Postnatal changes in capillary density of rat sternomastoid muscle. Am. J. Physiol. 257 (Heart Circ. Physiol. 26): H344-347, 1989. 9. HANSEN-SMITH, F. M., L. WATSON, D. Y. Lu, AND I. GOLDSTEIN. Grifjonia simpZicijoZia I: fluorescent tracer for microcirculatory vessels in nonperfused thin muscles and sectioned muscle. Microuasc. Res. 36: 199-215, 1988. 10. HUDLICKA, O., AND S. PRICE. The role of blood flow and/or muscle
hypoxia in capillary growth in chronically stimulated fast muscles. Pfluegers Arch. 417: 67-72, 1990. 11. HUDLICKA, O., AND K. R. TYLER. Angiogenesis: The Growth of the Vascular System. New York: Academic, 1986. 12. JUARBE, C., AND A. H. SILLAU. Muscle capillarity in rats with increased blood oxygen affinity. Respir. Physiol. 72: 83-94, 1988.
IN
HYPOXIA
R. D., M. L. SIMMONS, AND T. P. MCDONALD. Use of silicone rubber membrane enclosures for preparation of erythropoietin assay in mice. Ann. NY Acad. Sci. 149: 34-39, 1968. 14. MATHIEU-COSTELLO, 0. Muscle capillary tortuosity in high altitude mice depends on sarcomere length. Respir. Physiol. 76: 28913. LANGE,
302, 1989. 15. MCCOMB, R. B., G. N. BOWERS, AND S. POSEN. AZkaZine Phosphatase. New York: Plenum, 1979. A. T., AND D. M. HALE. Increased vascularity of brain, 16. MILLER, heart, and skeletal muscle of polycythemic rats. Am. J. Physiol. 219: 702-704, 1970. 17. NACHLAS, M. M., K. C. Tsou, E. D. SOUZA, C. S. CHENG, AND A. M. SELIGMAN. Cytochemical demonstration of succinic dehydrogenase by the use of a new p-nitrophenyl substituted ditetrazole. J. Histochem. Cytochem. 5: 420-36, 1957. 18. RIBATTI, D., L. MANCINI, L. RONCALI, B. NICO, AND M. BERTOSSI.
Morphometric analysis of the effects of hypoxia during the central nervous system vasculogenesis. Int. J. Microcirc. CZin. Exp. 8: 135142,1989. 19. SILLAU, A. H., L. AQUIN,
M. V. BUI, AND N. BANCHERO. Chronic hypoxia does not affect guinea pig skeletal muscle capillarity.
Pfluegers Arch. 386: 39-45, 1980. 20. SILLAU, A. H., AND N. BANCHERO.
Effect of hypoxia on the capillarity of guinea pig skeletal muscle. Proc. Sot. Exp. BioZ. Med. 160:
368-373,1979. 21. SMITH, P. Effect
of hypoxia upon the growth and sprouting activity of cultured aortic endothelium from the rat. J. CeZZ Sci. 92: 505-512,
1989. 22. SNYDER, G. K. Muscle capillarity in chicks following hypoxia. Comp. Biochem. Physiol. A Comp. Physiol. 87: 819-822, 1987. 23. SNYDER, G. K., E. E. WILCOX, AND E. W. BURNHAM. Effects of hypoxia on muscle capillarity in rats. Respir. Physiol. 62: 135-140, 1985. 24. SOKAL, ROBERT R., AND F. J. ROHLF. Biometry (2nd ed.). New York: Freeman, 1981, p. 859.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
L. H. BLACKWELL,
alkaline phosphatase
AND
G. R. JOSWIAK
Department of Biological Sciences and Office of Computer Services, 4401; and Department of Pharmacology and Physiology, University Detroit, Michigan 48207 HANSEN-SMITH, F. M.,L.H. BLACKWELL,AND G. R.JosWIAK. Expressionof musclecapillary alkaline phosphataseis affected by hypoxia. J. Appl. Physiol. 73(2): 776-780,1992.-Hyp-
oxia stimulates angiogenesisin some microvascular beds, but no clear angiogenic effect of hypoxia has yet been demonstrated in adult skeletal muscle. In this study the distribution of alkaline phosphatase (APase) was compared with a novel microvascular marker, Griffonia simplicifolia I (GSI), to determine whether the respective markers were expressedby muscle capillaries during hypoxic conditions and to probe for the presence or absenceof angiogenesisin responseto short-term hypoxia. Mice were exposedto normobaric 8% oxygen for 7 or 21 days. Capillary density in the red and white areas of the gastrocnemius musclewasdetermined with the use of a double-labeling procedure for both APase and fluorescently tagged GSI. Little change in capillary density was found. Focal reductions in APase activity were observedwithin 1 wk of hypoxia, but no changeswere observedin GSI binding. In controls, 74 and 92% of red and white muscle capillaries, respectively, were APase positive. This percentage declined to 60% in red and 43% in white muscleafter 21 days of hypoxia. The results indicate that APase expression is labile under certain conditions and warrant a cautious approach to using the enzyme as a marker. Binding of the GSI lectin to musclecapillaries appearedto be unchanged by the exposure to hypoxia, indicating stability of this marker system. No significant change in the number of capillaries around individual muscle fibers was evident at 21 days when GSI was used to detect capillaries. These results confirm the absenceof hypoxia-induced angiogenesisin muscle capillaries during the time period studied. skeletal muscle; endothelium; lectins; Griffonia angiogenesis
simplicifolia
I;
MICROVASCULAR BED may be exposed to hypoxic conditions as a result of environmental conditions, hematologic changes, or hemodynamic alterations. Whereas hypoxia has been shown to stimulate angiogenesis directly or indirectly in embryos (1, 13, 18) and in vitro (21), it is less clear whether hypoxia has an angiogenic influence on the skeletal muscle microcirculation. The effects of hypoxia on the capillary density of skeletal muscle have been previously studied, with a number of different experimental models (2-4, 12, 20, 22, 23). SevTHE
776
Oakland University, Rochester of Detroit School of Dentistry,
48309-
era1 studies have suggested that the density of capillaries per unit area is increased, possibly as a result of angiogenesis. However, this is explained in part by the reduction in fiber diameter that may occur in some muscle fibers. The quantitation of intramuscular capillaries often relies on the detection of the capillaries by histochemical means. However, only those capillaries that express the marker will be detected by this method. Recent studies in this laboratory suggest that at least one histochemical endothelial cell marker, alkaline phosphatase (APase), is poorly expressed by capillaries during postnatal growth (8). We postulated that enzymes such as APase might not be expressed initially by endothelial cells during hypoxia-induced angiogenesis in the adult. If this were the case, new capillaries would be present but not histochemitally detectable until they were mature enough to express APase, giving the appearance of a delayed angiogenie response. We therefore have probed for the presence or absence of hypoxia-induced angiogenesis and examined the influence of hypoxia on the expression of APase in skeletal muscle of mice. The lectin Griffonia simplicifolia I (GSI) was employed as an alternative marker for capillaries because it has previously been shown to detect all capillaries, including developing capillaries in rat muscle (7-9). Binding sites for this lectin (terminal a-galactosyl groups) have been shown to be expressed by capillaries during postnatal growth, before the differentiation of the APase system in capillaries of skeletal muscle (8). In addition, the lectin has been shown to bind to the tips of sprouting capillaries during developmental angiogenesis in the retina (6). METHODS
Female C,,Bl, mice [30 t 1 (SE) g] from a colony at the University of Detroit School of Dentistry were placed in cages constructed with membranes semipermeable to oxygen (13). Previous studies have established that the ambient oxygen concentration can be adjusted by varying the total body weight of animals within the cage (13). We used five animals per cage, giving a total of 150 g per cage. This resulted in an oxygen concentration of 8%. Experi-
0161-7567192 $2.00 Copyright 0 1992 the American Physiological Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CAPILLARY
ALKALINE
PHOSPHATASE
mental groups of n = 5 were maintained in the cages for 7 or 21 days and were given standard lab chow and water ad libitum. Bedding in the cages was changed twice per week. Normoxic age controls were exposed to ambient air in standard polypropylene cages of the same size but lacking the semipermeable membrane. After 7 or 21 days the control and experimental mice were killed by cervical dislocation. Blood samples for hematocrit determinations were then taken from the axillary vein. The efficacy of the experimental protocol was monitored by the hematocrits, which were significantly elevated after 7 and 21 days (normoxic controls 44 f 0.3 and 43 + 1.1, respectively; hypoxic 55.2 + 0.9 and 67.0 f 0.9). The gastrocnemius muscle was removed and quickfrozen in a slurry of dry ice-isopentane. Transverse cryostat sections (12 pm thick) were made from the muscle at its widest girth. The following histochemical procedures were carried out: APase activity was demonstrated by the method employed in our previous study (8). The reaction buffer contained 1.7 g/l, MgSO, * 7H,O, and 3.8 g/l NazB,O, . lOH,O. Before the reaction, 6 mg 5bromo-4chloro-3-indoxyl phosphate, toluidine salt, and 30 mg nitroblue tetrazolium were added to 30 ml of the buffer. The pH was adjusted to 9.2 with 1% boric acid. Unfixed sections were incubated for 30 min at 37°C in the reaction mixture and then rinsed in phosphate-buffered saline. Succinate dehydrogenase (SDH) activity was demonstrated by the method of Nachlas et al. (17). Rhodamine-GSI lectin binding was carried out as described previously (7-g), except that GSI (Vector) was used instead of the GSI-B, isomer. The lectin was diluted to 25 rglml with physiological saline containing 0.001 M Nacetylgalactosamine to eliminate possible nonvascular binding sites. A double-staining procedure was used for colocalizing lectin binding and APase activity, with a concentration of 40 pg rhodamine-GSI/ml (9). The higher lectin concentration facilitated visualization of lectin binding sites when viewed with a low level of transmitted light in combination with epifluorescence. Double-labeled sections used for quantitation were reacted for 30 min. Additional sections of some muscles were also reacted for 60 min. All of the sections were mounted in a water-soluble mountant without prior dehydration. The sections were examined with a Nikon Optiphot microscope equipped with optics for both transmitted and epifluorescent illumination. All quantitative analyses were carried out on photomicrographs taken at a magnification of X200. The capillarity of the muscles was determined from preselected “red” and “white” regions of each muscle, with the sections reacted for SDH as a reference. The same regions of muscle were selected in both control and experimental samples. The deepest, or most medial, region, containing fibers that stained intensely for SDH, was classified as red, whereas the most peripheral region, containing larger fibers that were only faintly stained for SDH, was classified as white. The transitional intermediate region was not sampled. Capillary densities were determined from photomicrographs. Two photographs of capillaries were taken from nonoverlapping fields in the center of the red region. Three photographs were taken at equally spaced locations at the periphery of the muscle to sample the white
IN HYPOXIA
FIG. 1. Combined alkaline phosphatase (APase) staining and rhodamine Criffonio simplicifoh I (GSI) binding in red (A, B, E, and F) and white (C, 0, G, and H) gastrocnemius of normoxic control (A-D) and 21-day hypoxic mice (E-H) are shown with transmitted light only (A, C, E, and G) or transmitted light simultaneously with epifluorescence (I?, D, F, and H). Examples of capillaries positive for APase, showing no GSI, are noted by solid arrows. These were counted as APase-positive capillaries. APase-negative GSI-positive capillaries are noted by clear arrows. These were counted as APase-negative capillaries. Capillaries positive for both APase and GSI are noted by both solid and clear arrows. These were counted only as APase-positive capillaries. Bar, 20 pm.
region. Each field was photographed twice, first with transmitted light to reveal APase activity. The identical field was then rephotographed with epifluorescent illumination to reveal GSI binding sites. The APase reaction product was visible by epifluorescent illumination, and some of the GSI binding was obscured in capillaries having large amounts of APase reaction product. Some capillaries could be detected by both methods and some were only positive for GSI (see Fig. 1). All microvessels surrounding the muscle fibers were counted unless they were large enough to reveal a lumen in the APase method or were approximately double that of the capillaries in the area. Criteria to specifically select capillaries on the basis of size (i.e., -5 pm) were impossible to set because of the intrinsic differences in APase staining between the red and white regions of controls and because of the variations that occurred as a result of the experimental treatment. The total number of APase-positive capillaries per
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
778
CAPILLARY
ALKALINE
PHOSPHATASE
field was determined from the first photograph. In the epifluorescence photograph only those GSI-positive capillaries that were completely negative for APase were counted. All sites counted in the first photograph were marked on the second photograph to avoid duplication of counts. The total number of capillaries per field (0.18 mm’), an index of capillary density, was calculated as the sum of the number of APase-positive capillaries and GSI-positive (APase-negative) capillaries in each field. To detect whether angiogenesis occurred as a result of 21 days of hypoxia, the number of capillaries around individual muscle fibers (CAF) in the red and white region was determined from double-labeled samples and from adjacent sections exposed to rhodamine-GSI alone. The sampling protocol described previously (7, 8) was used. Briefly, this consisted of using a cross in the eyepiece to randomly select four nonadjacent fibers at a time. This was done with the rhodamine filter blocked to eliminate any bias. Then the rhodamine filter was replaced and the number of GSI-positive capillaries around each of the selected fibers was counted. Capillaries in a total of eight fields (32 fibers) each were counted in the red and white regions, respectively, for each muscle. The CAF was determined by the same protocol for the double-labeled samples. A low level of transmitted light was used in combination with the epifluorescence to better visualize all capillaries. Analysis of variance for these data was conducted with the SPSSX version 4.0 statistical package. Percentages were analyzed statistically with the arsine transformation to obtain a normal distribution (24). RESULTS
AND
DISCUSSION
In this study double-labeled sections were used to facilitate quantitation of capillaries at identical sites within the muscle, with two different histological methods to visualize capillaries. This is the first time that a histochemical method of detecting capillaries in muscle has been critically evaluated for its accuracy during an experimental protocol such as hypoxia. Capillary densities determined by the double-labeling technique were greater than those determined with APase alone, both in the control and experimental muscles (Table 1). APase-positive capillaries constituted 74 t 2 and 72 t 1% for 7- and 2Lday controls in the red region and 91 t 3 and 88 t 5% of the total capillaries per field for the white region of controls. These data in control muscle concur with our previously reported studies in rats, demonstrating that APase is not uniformly expressed by capillaries throughout the microvascular bed (8,9). Capillary density determined from the double-labeled sections was significantly increased in the white region of hypoxic muscle after 7 days (Table 1), but no statistically significant differences were found for capillary densities in the other groups. APase activity associated with endothelial cells, as demonstrated by the intensity and distribution of reaction product, was clearly affected by the hypoxic environment. An overall diminution of the density of reaction product was visually evident in some capillaries within the first week. This was more conspicuous in the white region than in the red region. In some areas
IN HYPOXIA
TABLE 1. Capillarity of gastrocnemius muscle from normoxic and hypoxic mice Normoxic Red
Control
Hypoxic
White
Red
White
7 days Total capillary density % APase-positive 2 capillaries
258.8k22.3 74k2
66.4t3.1
290.5k13.8
92k3
75t2
78.2t4.0* 67k4-f
21 days Total capillary density % APase-positive capillaries
257.8t13.4 72kl
56.4zk6.2 98t5
255.8k32.6 60+8$
73.3t5.9 43*10$
Values are means + SE. Total capillary density, total number of capillaries per field (0.18 mm2) = sum of APase-positive capillaries + GSI-positive (APase-negative) capillaries. Significantly different from normoxic age control white muscle: * P = 0.05; t P = 0.003 (arcsine transformation); § P = 0.008 (arcsine transformation). Significantly different from normoxic age control red muscle: $ P = 0.024 (arcsine transformation).
the APase was significantly reduced but still detectable, whereas no reaction product was detectable in some areas. Extending the incubation times up to 60 min failed to generate detectable reaction product in the negative areas, suggesting that the APase activity was significantly reduced or absent in these areas of the hypoxic muscle (Fig. 1). In contrast to the results with APase, there was no detectable diminution of GSI binding to capillaries (Fig. 2). When the double-labeled sections were examined, it was clear that GSI-positive capillaries were present in regions in which APase activity was reduced or absent (Fig. 1). These double-labeled sections further revealed that the capillary APase in all regions of the muscle, not just the white regions, was affected by the hypoxia. However, in the red regions the diminution of APase was less striking visually because it did not occur in conspicuous large patches, as it did in the white region. The percentage of APase-positive capillaries in the red region declined significantly between 7 and 21 days of hypoxia (72 t 1 vs. 60 t 8%). The percentage of APase-positive capillaries in the white region was significantly lower than controls at 7 days (67 t 4%) and was reduced further at 21 days (43 t 10%). All counts of APase-labeled capillaries included any capillaries that could be detected, not just those that were strongly stained (see Fig. 1). The expression of APase by capillaries was thus diminished earlier and more extensively in the white region than in the red region of the muscle. These data indicate that the response of muscle capillaries to hypoxia is heterogeneous in different metabolic regions of the muscle. We examined the muscles for quantitative evidence of hypoxia-induced angiogenesis. Our original hypothesis predicted that, if angiogenesis occurred, new capillaries might fail to express APase initially. Capillary density is not a reliable index for detecting angiogenesis, because both capillary numbers and muscle fiber size may influence the resulting data. Therefore, we probed for an absolute change in CAF. No significant difference in CAF
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CAPIUARY
ALKALKNE
PHGSHIATASE
IN HYPOXIA
779
FIG. 2, Rhudamiae-GSI binding to Cepib&S irr red (A and 0 and white (B and D) gbrocnemius of normoxie controls (A and B) and 21.&y (C and D) hypoxic mice, shown with epifluoreecent illuminetiorr only. Bar, 20 pm.
was found in either the red or white region of the muscle after 21 days (Table 2), arguing against hypoxia-induced a~~o~e~~si~ within this time period. the same when C&I was used alone an beled samples. These results confirm those of most other with different models to induce hypoxia. eloping blood vessels may be stimulated by cle capillaries in adult animals appear to e remarkably resistant ti this putative angiogenic stimulus. y there was little quantitative change in or absolute number of CAF as a result of er, the enzymatic activity of the capillary
APase was changed, although we consider this an unlikely explanation for our observations. Our reaction em-
Red
CAP r(GSI-positive cap&&s)*
white
Red
Wbik
4.WM.11
2.6SM.09
4.&+0.11
2.79iO.09
4.39*0.10
2.71110.10
4.7&u.l4
2.63+0.15
CAF (total capillari&t
VaIues are means f SE. * Number of capilJariea around 32 rsndomly individual muscle fib in each retion Der szrou~ was determined from +Z?J-labeled cryostat t3ectiona. t-To&t nu&& of capillaries around 32 randomly selected individual muscle fibers was determin in each region per group. AFase- and GSI-positive (APase-negativej capillaries were counted from double-labeled cryostat s&ions,
ployed an indoxyl phosphate as a sub&rate. In the past, many different substrates, including ,&glycerophosphate used for demon-
es suggesting that APase activity histochemieal marker for quantitacapillarity (8,9). There is heterogeenzymes even under co study and our previous reports (8,9) illustrate. Growth and physiological status of the muscle may further affect the expression of this enzyme (and possibly other endothelial enzymes), Therefore, the further use of APase as a capillary marker should be dated by alternative markers OFby direct visualization in sections under both co ns (14). The present provide other useful information concerning endotbelial cell metabolism. Although further co~~rrna~~on is needed, the possibility al loss of APase activity may prove to be a identifying isebemic and/or bypoxic loci within skeletal or cardiac muscle. We gratefully acknowledge the technical assistance of Laura Morris and Paul Maximuke and the secretarial assistance of Rita Per+ This study was supp gmnt fmm merican Hea c&ion of Michigan, a at Sciences arch Grant. Research Committee at Oakland University. Address for reprint requeeta: F. M. Nan~n-Smith, Dept.. ofBiological Sciences, Oakland University, Rochester, MI 48309-4401, Received 21 December 1990; accepted in final form 5 February 1992, RENCES 1. AME, T. H., A, 6. GUYTON, J. P. hioNTANI, I-I. t. L&DsAY, AND K. A. STANEK. whole body structural vascular ad&@&ion to pro-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
780
CAPILLARY
ALKALINE
longed hypoxia in chick embryos. Am. J. Physiol.
252 (Heart
PHOSPHATASE Circ.
Physiol. 21): H1228-H1234, 1987. 2. BANCHERO, N., S. R. KAYAR, AND
A. J. LECHNER. Increased capillarity in skeletal muscle of growing guinea pigs acclimated to cold and hypoxia. Respir. Physiol. 62: 245-255, 1985. 3. BENDER, P. R., A. TUCKER, AND T. A. TWIETMEYER. Skeletal muscle capillarity during recovery from chronic hypoxic exposure. Microcirc. Endothelium Lymphatics 4. CASSIN, S., R. D. GILBERT, C.
1: 71-85,
1984.
E. BUNNELL, AND E. M. JOHNSON. Capillary development during exposure to chronic hypoxia. Am. J.
Physiol. 220: 448-451, 1971. 5. CHRISTIE, K. N., AND C. THOMSON.
brane-bound
Effect of ischaemia on memhydrolases in rat soleus capillaries. Acta Histochem.
Cytochem. 22: 47-65,1989. 6. HANSEN-SMITH, F. M. Postnatal development sels (Abstract). FASEB J. 4: A1897, 1990. 7. HANSEN-SMITH, F. M., K. BANKER, L. MORRIS,
of retinal microves-
AND G. JOSWIAK. Alternative histochemical markers for skeletal muscle capillaries: a statistical comparison among .three muscles. Microuasc. Res. In press. 8. HANSEN-SMITH, F. M., L. WATSON, AND G. R. JOSWIAK. Postnatal changes in capillary density of rat sternomastoid muscle. Am. J. Physiol. 257 (Heart Circ. Physiol. 26): H344-347, 1989. 9. HANSEN-SMITH, F. M., L. WATSON, D. Y. Lu, AND I. GOLDSTEIN. Grifjonia simpZicijoZia I: fluorescent tracer for microcirculatory vessels in nonperfused thin muscles and sectioned muscle. Microuasc. Res. 36: 199-215, 1988. 10. HUDLICKA, O., AND S. PRICE. The role of blood flow and/or muscle
hypoxia in capillary growth in chronically stimulated fast muscles. Pfluegers Arch. 417: 67-72, 1990. 11. HUDLICKA, O., AND K. R. TYLER. Angiogenesis: The Growth of the Vascular System. New York: Academic, 1986. 12. JUARBE, C., AND A. H. SILLAU. Muscle capillarity in rats with increased blood oxygen affinity. Respir. Physiol. 72: 83-94, 1988.
IN
HYPOXIA
R. D., M. L. SIMMONS, AND T. P. MCDONALD. Use of silicone rubber membrane enclosures for preparation of erythropoietin assay in mice. Ann. NY Acad. Sci. 149: 34-39, 1968. 14. MATHIEU-COSTELLO, 0. Muscle capillary tortuosity in high altitude mice depends on sarcomere length. Respir. Physiol. 76: 28913. LANGE,
302, 1989. 15. MCCOMB, R. B., G. N. BOWERS, AND S. POSEN. AZkaZine Phosphatase. New York: Plenum, 1979. A. T., AND D. M. HALE. Increased vascularity of brain, 16. MILLER, heart, and skeletal muscle of polycythemic rats. Am. J. Physiol. 219: 702-704, 1970. 17. NACHLAS, M. M., K. C. Tsou, E. D. SOUZA, C. S. CHENG, AND A. M. SELIGMAN. Cytochemical demonstration of succinic dehydrogenase by the use of a new p-nitrophenyl substituted ditetrazole. J. Histochem. Cytochem. 5: 420-36, 1957. 18. RIBATTI, D., L. MANCINI, L. RONCALI, B. NICO, AND M. BERTOSSI.
Morphometric analysis of the effects of hypoxia during the central nervous system vasculogenesis. Int. J. Microcirc. CZin. Exp. 8: 135142,1989. 19. SILLAU, A. H., L. AQUIN,
M. V. BUI, AND N. BANCHERO. Chronic hypoxia does not affect guinea pig skeletal muscle capillarity.
Pfluegers Arch. 386: 39-45, 1980. 20. SILLAU, A. H., AND N. BANCHERO.
Effect of hypoxia on the capillarity of guinea pig skeletal muscle. Proc. Sot. Exp. BioZ. Med. 160:
368-373,1979. 21. SMITH, P. Effect
of hypoxia upon the growth and sprouting activity of cultured aortic endothelium from the rat. J. CeZZ Sci. 92: 505-512,
1989. 22. SNYDER, G. K. Muscle capillarity in chicks following hypoxia. Comp. Biochem. Physiol. A Comp. Physiol. 87: 819-822, 1987. 23. SNYDER, G. K., E. E. WILCOX, AND E. W. BURNHAM. Effects of hypoxia on muscle capillarity in rats. Respir. Physiol. 62: 135-140, 1985. 24. SOKAL, ROBERT R., AND F. J. ROHLF. Biometry (2nd ed.). New York: Freeman, 1981, p. 859.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
