STAINTKCHNOLOGY Copyright 0 1976 by The Williams & Wilkins Co.
Val. 51, No.3 Printed in U.S.A.
STUDIES OF OSMIUM DEPOSITS IN THE CAROTID BODY OF THE CAT
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
FADHIL AL-LAMI,Department of Anatomy and Physiology and School of Optometry, Indiana University, Bloomington, Indiana 4740 I ABSTRACT. Doubk aldehyde fixed carotid bodies and small p t a s of vagus nerve of cat8 were incubated in 3 mM copper sulfate and 0.5 mM potassium krricyanide in 0.05 M aatate buffer (pH 5.6) for 30 minutes at room temperature. Several modifications of this procedure were also attempted. Tissues were then pastosmicatcd with 2% unbufkred osmium tetroxidc and heated to 50-55 C for ten minutes.Under the electron microscope carotid body alla exhibited fine osmium deposits within cisternae of endoplasmic reticulum, socculo and vesicles of Golgi complex, and cristae of mitochondria. Intense osmium precipitation wpll also noted in the mitochondria of nerve endings. In addition, much more intense, more conspicuous and more localized reaction wan noted in the intrapcriod lines of the myelin sheath of m~eryeir. Deposits here were rod-shaped, displaying considerable variation in length. These results are discussed in the light of previous findings on osmium deposits in various tissues. It was concluded that the osmium reaction is unspecific, and that histochemical methods employing hot osmium tetroxide to amplify enzymatic activities may therefore not be reliable.
Prolonged fixation (38-75 hours) of a variety of tissues with osmium tetroxide (OsO3 at an elevated temperature (40-45 C) produces deposits at specific locations in the cytoplasm (Friend and Murray 1965, Friend 1969, Friend and Brassil 1970, Windhorn and Selig 1974, Pourcho and Bernstein 1975). Such locations include the cisternae of Golgi apparatus, the cisternae of endoplasmic reticulum and sometimes the cristae of mitochondria. This report describes a method whereby a brief exposure (10 minutes) of the carotid body to hot (50-55 C) 0sO6 in combination with several other reagents, produces within the glomus cells of the carotid body fine osmium deposits at locations corresponding to the above-mentioned sites. In addition, it demonstrates hitherto undescribed, rather consistent, and extensive deposits at the intraperiod lines of the myelin sheath of carotid body nerves and the vagus nerve.
MATERIALS AND METHODS Six healthy, full-grown cats were used in this study. The animals, male and female, were anesthetized with intraperitoneal injections of sodium pentobarbital (40 mg/kg) and perfused through the heart with 3% glutaraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.3) at room temperature. About five minutes later, when satisfactory perfusion was attained, the carotid bodies, as well as several small pieces of the vagus nerve, were dissected out and placed in fresh cold fixative for an additional two hours. After fixation, tissues were washed with 0.88 M sucrose in 1% gum acacia, and incubated for 30 minutes at room temperature with one of the variants of the agent described below. They were then washed in distilled water, immersed in 2% unbuffered OsO and heated to 50-55 C in closed jars for ten minutes. Some tissues, after aldehyde fixation and sucrose washing, were directly immersed in OsO without prior incubation. All tissues were then cooled, rinsed in distilled water, dehydrated
,
Supported by Grant number NS 07472 from the Institute of Neurological Diseases and Stroke, United States Public Health Service. 163
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
164
STAIN TECHNOLOGY
FIGS.1-4.
OSMIUM DEPOSITS IN CAROTID BODY
165
TABLE 1. INCUBATION MEDIAUSED EXPERIMENTALLY BEFORETISSUE OSMICATION, ON
Modification Number
Acetate buffer (P.05M)
(3mM)
1
X X
X X
2
cus04
X
3
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
4
5 6 7
8
A N D THEIR %ECTS
TISSUE OSMIUM DEPOSITS
X
:flE:;*
OSo' Temp. C
X X
50-55 50-55 50-55 50-55 50-55 20-23 20-23 50-55
X X
X
pHof
bufFer
7.0 5.6
-
5.6-7.0 -
-
Results
+ + + + + -
in alcohols and embedded in Araldite 502. Sections were examined with the electron microscope without staining. T h e incubation solution used was 3 m M copper sulphate and 0.5 m M potassium ferricyanide in 0.05 M acetate buffer (pH 5.6). Several modifications, including varying the p H and/or omitting various elements of the procedure, were used, the results of which will be described.
RESULTSAND DISCUSSION The carotid body in the cat, as well as in other mammals, is composed of two cell types, enclosing cells, which are characterized by their long and tortuous processes, and the enclosed cells, which these processes surround (Al-Lami and Murray 1968). T h e more numerous enclosed cells contain characteristic dark-cored vesicles, welldeveloped Golgi zones, and distinct rough-surfaced endoplasmic reticulum (Figs. 1,2). Nerve endings partly surrounded by the enclosing cells are in contact with the enclosed cells. Except for precipitate in the occasional mitochondria, the cytoplasm of the enclosing cells, unlike that of the enclosed, exhibited virtually no osmium deposit. This report will deal only with the enclosed cells. Various combinations of reagents were tried, as listed in Table 1 . However, all tissues, regardless of differences in incubation agents or pH, exhibited finely granulated osmium deposits at similar intracellular locations. The quality of tissue preservation varied. It was most favorable when both copper sulfate and potassium ferricyanide in acetate buffer (pH 5.6) were used as a n incubation agent. This was more evident in the carotid body than in the vagus nerve, which seemed to be well FIGS.1-4. Electron micrographs of sections of carotid bodies which were fixed with double aldehyde, incubated for 30 minutes with potassium ferricyanide and copper sulfate in acetate buffer (pH 5.6), and postosmicated for 10 minutes in 2% unbuffered OsO, at 50-55 C, as described in the section on Materials and Methods. FIG. 1. Cytoplasm of a carotid body cell showing osmium precipitates in the cisternae of endoplasmic reticulum (arrows), in the cristae of mitochondria (M) and more intense in the mitochondria of nerve endings x 30,000. FIG. 2. Golgi apparatus (G)in a carotid body cell with osmium deposits in its cisternae (arrows) and its vesicles. Reaction product can also be seen in the cristae of mitochondria (lmlabelcd). x 30,000. FIG.3. Portions of two myelinatel nerve fibers (F)in the neighborhood of the carotid body. Osmium precipitates can be seen in the myelin sheath (unlatxled); (S)Schwann cells. x 75,000. FIG.4. High magnification of the myelin of a nerve fiber, similar to that of Figure 3. Arrows indicate major dense lines. Osmium deposit can be seen between these lines in the intraperiod lines. x 450,000.
(a.
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
I66
STAIN TECHNOLOGY
FIGS.5 and 6 . Sections of nerve fiber myelin in vagus nerve immersed directly in 50-55 C OsO, after double aldehyde fixation without incubation. It may be seen that the intraperiod lines are free of deposit. 100,000 x and 450,000 x respectively.
preserved in virtually every case. T h e figures presented are all from material preincubated in this medium. In such materials osmium deposits were seen within the enclosed cells, in the cisternae of endoplasmic reticulum, in the cisternae and vesicles of Golgi zones, and in the cristae but not the matrix of mitochondria (Figs. 1 and 2). Often, mitochondria1 cristae of nerve endings displayed heavier deposits than those of the enclosed cells (Fig. 1). Occasional lysosome-like structures also exhibited some osmium pre.cipitate (not demonstrated). However, more intense and more consistent reaction was noted in the myelin of the nerve fibers of the carotid body and of the vagus nerve (Figs. 3 and 4). Such deposits were invariably present in every section of myelin we have encountered, even when carotid body cells in the same section showed little or no reaction. As can be seen in Fig. 4, these thread-like deposits, which varied considerably in length but less so in width (30-70 A wide), were confined mainly to the intraperiod lines of the myelin. As shown in Table 1, osmication of preincubated materials with cold (4 C) or room temperature OsO, showed no deposit. Similarly, when tissues were directly treated with hot OsO, without prior incubation, the results were also negative (Figs. 5 and 6), and the myelin structure was similar to that which could be seen in conventional micrographs. Brief treatment with hot OsO, therefore, represented one of the two factors required for the induction of such deposits. T h e other factor seemed to be exposure to one of perhaps a number of apparently unrelated compounds. Of interest was the fact that osmium deposition was no more intense in tissues incubated in copper sulfate and potassium ferricyanide combined than in those which were incubated in either one alone. This indicates a lack of additive effect. It seemed an “all or none” reaction. T h e nature of this deposit and the reason it forms at particular sites remain open questions. It could be that substances such as potassium ferricyanide or copper sulfate react with specific lipids and/or lipoprotein molecules in myelin and in cytoplasmic membrane of carotid body cells, altering the composition of such molecules, possibly in different ways, thereby increasing their affinity for hot OsO .,
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
OSMIUM DEPOSITS IN CAROTID BODY
167
The lack of a common characteristic among the inductive agents (potassium ferricyanide, copper sulfate or acetate buffer) suggests that the osmium reaction is nonspecific and that the osmium deposits represent more than a single substance (see Hanker et al. 1967, Friend and Brassil 1970, Windborn and Selig 1974). As can be seen in Figure 4, osmium deposits occurred mainly in the intraperiod lines of the myelin. These lines correspond to the outer leaflets of the plasma membrane of Schwann cells, and it is not known why osmium deposits favor these sites. However, it has been reported that in plasma membranes unsaturated carbon-carbon bonds and the hydrophilic ends of the lipid molecules selectively react with osmium tetroxide (Bahr 1954, Stoeckenius and Mahr 1965). Accordingly, in myelin, hydrophilic groups and double bonds may be more abundant in the intraperiod lines than in the major dense lines. No explanation of these results with regard to the carotid body cells as chemoreceptors has been advanced. In the adrenal cortex, however, osmium deposits had been correlated with the functional activity of the cell (Friend and Brassil 1970). Work is under way to study osmium deposits in stimulated carotid bodies. The choice of the components of the incubation solutions in this study was based solely on their common use in histochemical procedures. Based on the results reported here, caution should be exercised when such reagents are used to amplify cytochemical reactivity, as in the amplification of Hatchett’s brown deposits advanced by Hanker et al. (1972, 1975).
Ac KNOWLEDGMENT The author is grateful to Professor R. G. Murray for his interest, advice and criticism. REFERENCES Al-Lami, F., and Murray, R. G. 1968. Fine structure of the carotid body of normal and anoxic cats. Anat. Rec. 160: 697-718. Bahr, G. F. 1954. Osmium tetroxide and ruthenium tetroxide and their reactions with biologically important substances. Exp. Cell Res. 7: 457-479. Friend, D. S., and Murray, M. J. 1965. Osmium impregnation of the Golgi apparatus. Am. J. Anat. 777: 135-150. Friend, D. S. 1969. Cytochemical staining of multivesicular body and Golgi vesicles. J. Cell Bi01. 41: 269-279. Friend, D. S., and Brassil, G. E. 1970. Osmium staining ofendoplasmic reticulum and mitochondria in the rat adrenal cortex. J. Cell Biol. 46: 252-266. Hanker, J. S., Kasler, F., Bloom, M. G., Copeland, J. S., Seligman, A. M. 1967. Coordination polymer of osmium; the nature of osmium black. Science 156: 1737-1738. Hanker, J. S., Yeates, P. E., Clap, D. H., and Anderson, W. A. 1972. New methods for the demonstration of lywsomal hydrolases by the formation of osmium blacks. Histochemie 30: 201-214. Hanker, J. S., Thornhurg, L. P., Yates, P. E., and Romanovicz, D. K. 1975. The demonstration of arylsufatases with 4-nitro-l,2-benzenediol mono (hydrogen sulfate) by the formation of osmium blacks at the sites of copper capture. Histochemie 47: 207-225. Pourcho, R.G., and Bernstein, M. H. 1975. Light dependence of osmium reactivity in mouse photoreceptor cells. Am. J. Anat. 743: 371-385. Stoeckenius, W., and Mahr, S. C. 1965. Studies on the reaction of osmium tetroxide with lipids and related compounds. Lab. Invest. 74: 1196-1207. Windhorn, W. B., and Selig, L. L. Jr. 1974. Pattern of osmium deposition in the parietal cells of the stomach. J. Cell Biol. 63: 99-108.
Val. 51, No.3 Printed in U.S.A.
STUDIES OF OSMIUM DEPOSITS IN THE CAROTID BODY OF THE CAT
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
FADHIL AL-LAMI,Department of Anatomy and Physiology and School of Optometry, Indiana University, Bloomington, Indiana 4740 I ABSTRACT. Doubk aldehyde fixed carotid bodies and small p t a s of vagus nerve of cat8 were incubated in 3 mM copper sulfate and 0.5 mM potassium krricyanide in 0.05 M aatate buffer (pH 5.6) for 30 minutes at room temperature. Several modifications of this procedure were also attempted. Tissues were then pastosmicatcd with 2% unbufkred osmium tetroxidc and heated to 50-55 C for ten minutes.Under the electron microscope carotid body alla exhibited fine osmium deposits within cisternae of endoplasmic reticulum, socculo and vesicles of Golgi complex, and cristae of mitochondria. Intense osmium precipitation wpll also noted in the mitochondria of nerve endings. In addition, much more intense, more conspicuous and more localized reaction wan noted in the intrapcriod lines of the myelin sheath of m~eryeir. Deposits here were rod-shaped, displaying considerable variation in length. These results are discussed in the light of previous findings on osmium deposits in various tissues. It was concluded that the osmium reaction is unspecific, and that histochemical methods employing hot osmium tetroxide to amplify enzymatic activities may therefore not be reliable.
Prolonged fixation (38-75 hours) of a variety of tissues with osmium tetroxide (OsO3 at an elevated temperature (40-45 C) produces deposits at specific locations in the cytoplasm (Friend and Murray 1965, Friend 1969, Friend and Brassil 1970, Windhorn and Selig 1974, Pourcho and Bernstein 1975). Such locations include the cisternae of Golgi apparatus, the cisternae of endoplasmic reticulum and sometimes the cristae of mitochondria. This report describes a method whereby a brief exposure (10 minutes) of the carotid body to hot (50-55 C) 0sO6 in combination with several other reagents, produces within the glomus cells of the carotid body fine osmium deposits at locations corresponding to the above-mentioned sites. In addition, it demonstrates hitherto undescribed, rather consistent, and extensive deposits at the intraperiod lines of the myelin sheath of carotid body nerves and the vagus nerve.
MATERIALS AND METHODS Six healthy, full-grown cats were used in this study. The animals, male and female, were anesthetized with intraperitoneal injections of sodium pentobarbital (40 mg/kg) and perfused through the heart with 3% glutaraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.3) at room temperature. About five minutes later, when satisfactory perfusion was attained, the carotid bodies, as well as several small pieces of the vagus nerve, were dissected out and placed in fresh cold fixative for an additional two hours. After fixation, tissues were washed with 0.88 M sucrose in 1% gum acacia, and incubated for 30 minutes at room temperature with one of the variants of the agent described below. They were then washed in distilled water, immersed in 2% unbuffered OsO and heated to 50-55 C in closed jars for ten minutes. Some tissues, after aldehyde fixation and sucrose washing, were directly immersed in OsO without prior incubation. All tissues were then cooled, rinsed in distilled water, dehydrated
,
Supported by Grant number NS 07472 from the Institute of Neurological Diseases and Stroke, United States Public Health Service. 163
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
164
STAIN TECHNOLOGY
FIGS.1-4.
OSMIUM DEPOSITS IN CAROTID BODY
165
TABLE 1. INCUBATION MEDIAUSED EXPERIMENTALLY BEFORETISSUE OSMICATION, ON
Modification Number
Acetate buffer (P.05M)
(3mM)
1
X X
X X
2
cus04
X
3
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
4
5 6 7
8
A N D THEIR %ECTS
TISSUE OSMIUM DEPOSITS
X
:flE:;*
OSo' Temp. C
X X
50-55 50-55 50-55 50-55 50-55 20-23 20-23 50-55
X X
X
pHof
bufFer
7.0 5.6
-
5.6-7.0 -
-
Results
+ + + + + -
in alcohols and embedded in Araldite 502. Sections were examined with the electron microscope without staining. T h e incubation solution used was 3 m M copper sulphate and 0.5 m M potassium ferricyanide in 0.05 M acetate buffer (pH 5.6). Several modifications, including varying the p H and/or omitting various elements of the procedure, were used, the results of which will be described.
RESULTSAND DISCUSSION The carotid body in the cat, as well as in other mammals, is composed of two cell types, enclosing cells, which are characterized by their long and tortuous processes, and the enclosed cells, which these processes surround (Al-Lami and Murray 1968). T h e more numerous enclosed cells contain characteristic dark-cored vesicles, welldeveloped Golgi zones, and distinct rough-surfaced endoplasmic reticulum (Figs. 1,2). Nerve endings partly surrounded by the enclosing cells are in contact with the enclosed cells. Except for precipitate in the occasional mitochondria, the cytoplasm of the enclosing cells, unlike that of the enclosed, exhibited virtually no osmium deposit. This report will deal only with the enclosed cells. Various combinations of reagents were tried, as listed in Table 1 . However, all tissues, regardless of differences in incubation agents or pH, exhibited finely granulated osmium deposits at similar intracellular locations. The quality of tissue preservation varied. It was most favorable when both copper sulfate and potassium ferricyanide in acetate buffer (pH 5.6) were used as a n incubation agent. This was more evident in the carotid body than in the vagus nerve, which seemed to be well FIGS.1-4. Electron micrographs of sections of carotid bodies which were fixed with double aldehyde, incubated for 30 minutes with potassium ferricyanide and copper sulfate in acetate buffer (pH 5.6), and postosmicated for 10 minutes in 2% unbuffered OsO, at 50-55 C, as described in the section on Materials and Methods. FIG. 1. Cytoplasm of a carotid body cell showing osmium precipitates in the cisternae of endoplasmic reticulum (arrows), in the cristae of mitochondria (M) and more intense in the mitochondria of nerve endings x 30,000. FIG. 2. Golgi apparatus (G)in a carotid body cell with osmium deposits in its cisternae (arrows) and its vesicles. Reaction product can also be seen in the cristae of mitochondria (lmlabelcd). x 30,000. FIG.3. Portions of two myelinatel nerve fibers (F)in the neighborhood of the carotid body. Osmium precipitates can be seen in the myelin sheath (unlatxled); (S)Schwann cells. x 75,000. FIG.4. High magnification of the myelin of a nerve fiber, similar to that of Figure 3. Arrows indicate major dense lines. Osmium deposit can be seen between these lines in the intraperiod lines. x 450,000.
(a.
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
I66
STAIN TECHNOLOGY
FIGS.5 and 6 . Sections of nerve fiber myelin in vagus nerve immersed directly in 50-55 C OsO, after double aldehyde fixation without incubation. It may be seen that the intraperiod lines are free of deposit. 100,000 x and 450,000 x respectively.
preserved in virtually every case. T h e figures presented are all from material preincubated in this medium. In such materials osmium deposits were seen within the enclosed cells, in the cisternae of endoplasmic reticulum, in the cisternae and vesicles of Golgi zones, and in the cristae but not the matrix of mitochondria (Figs. 1 and 2). Often, mitochondria1 cristae of nerve endings displayed heavier deposits than those of the enclosed cells (Fig. 1). Occasional lysosome-like structures also exhibited some osmium pre.cipitate (not demonstrated). However, more intense and more consistent reaction was noted in the myelin of the nerve fibers of the carotid body and of the vagus nerve (Figs. 3 and 4). Such deposits were invariably present in every section of myelin we have encountered, even when carotid body cells in the same section showed little or no reaction. As can be seen in Fig. 4, these thread-like deposits, which varied considerably in length but less so in width (30-70 A wide), were confined mainly to the intraperiod lines of the myelin. As shown in Table 1, osmication of preincubated materials with cold (4 C) or room temperature OsO, showed no deposit. Similarly, when tissues were directly treated with hot OsO, without prior incubation, the results were also negative (Figs. 5 and 6), and the myelin structure was similar to that which could be seen in conventional micrographs. Brief treatment with hot OsO, therefore, represented one of the two factors required for the induction of such deposits. T h e other factor seemed to be exposure to one of perhaps a number of apparently unrelated compounds. Of interest was the fact that osmium deposition was no more intense in tissues incubated in copper sulfate and potassium ferricyanide combined than in those which were incubated in either one alone. This indicates a lack of additive effect. It seemed an “all or none” reaction. T h e nature of this deposit and the reason it forms at particular sites remain open questions. It could be that substances such as potassium ferricyanide or copper sulfate react with specific lipids and/or lipoprotein molecules in myelin and in cytoplasmic membrane of carotid body cells, altering the composition of such molecules, possibly in different ways, thereby increasing their affinity for hot OsO .,
Biotech Histochem Downloaded from informahealthcare.com by University of Otago on 01/04/15 For personal use only.
OSMIUM DEPOSITS IN CAROTID BODY
167
The lack of a common characteristic among the inductive agents (potassium ferricyanide, copper sulfate or acetate buffer) suggests that the osmium reaction is nonspecific and that the osmium deposits represent more than a single substance (see Hanker et al. 1967, Friend and Brassil 1970, Windborn and Selig 1974). As can be seen in Figure 4, osmium deposits occurred mainly in the intraperiod lines of the myelin. These lines correspond to the outer leaflets of the plasma membrane of Schwann cells, and it is not known why osmium deposits favor these sites. However, it has been reported that in plasma membranes unsaturated carbon-carbon bonds and the hydrophilic ends of the lipid molecules selectively react with osmium tetroxide (Bahr 1954, Stoeckenius and Mahr 1965). Accordingly, in myelin, hydrophilic groups and double bonds may be more abundant in the intraperiod lines than in the major dense lines. No explanation of these results with regard to the carotid body cells as chemoreceptors has been advanced. In the adrenal cortex, however, osmium deposits had been correlated with the functional activity of the cell (Friend and Brassil 1970). Work is under way to study osmium deposits in stimulated carotid bodies. The choice of the components of the incubation solutions in this study was based solely on their common use in histochemical procedures. Based on the results reported here, caution should be exercised when such reagents are used to amplify cytochemical reactivity, as in the amplification of Hatchett’s brown deposits advanced by Hanker et al. (1972, 1975).
Ac KNOWLEDGMENT The author is grateful to Professor R. G. Murray for his interest, advice and criticism. REFERENCES Al-Lami, F., and Murray, R. G. 1968. Fine structure of the carotid body of normal and anoxic cats. Anat. Rec. 160: 697-718. Bahr, G. F. 1954. Osmium tetroxide and ruthenium tetroxide and their reactions with biologically important substances. Exp. Cell Res. 7: 457-479. Friend, D. S., and Murray, M. J. 1965. Osmium impregnation of the Golgi apparatus. Am. J. Anat. 777: 135-150. Friend, D. S. 1969. Cytochemical staining of multivesicular body and Golgi vesicles. J. Cell Bi01. 41: 269-279. Friend, D. S., and Brassil, G. E. 1970. Osmium staining ofendoplasmic reticulum and mitochondria in the rat adrenal cortex. J. Cell Biol. 46: 252-266. Hanker, J. S., Kasler, F., Bloom, M. G., Copeland, J. S., Seligman, A. M. 1967. Coordination polymer of osmium; the nature of osmium black. Science 156: 1737-1738. Hanker, J. S., Yeates, P. E., Clap, D. H., and Anderson, W. A. 1972. New methods for the demonstration of lywsomal hydrolases by the formation of osmium blacks. Histochemie 30: 201-214. Hanker, J. S., Thornhurg, L. P., Yates, P. E., and Romanovicz, D. K. 1975. The demonstration of arylsufatases with 4-nitro-l,2-benzenediol mono (hydrogen sulfate) by the formation of osmium blacks at the sites of copper capture. Histochemie 47: 207-225. Pourcho, R.G., and Bernstein, M. H. 1975. Light dependence of osmium reactivity in mouse photoreceptor cells. Am. J. Anat. 743: 371-385. Stoeckenius, W., and Mahr, S. C. 1965. Studies on the reaction of osmium tetroxide with lipids and related compounds. Lab. Invest. 74: 1196-1207. Windhorn, W. B., and Selig, L. L. Jr. 1974. Pattern of osmium deposition in the parietal cells of the stomach. J. Cell Biol. 63: 99-108.
