GJournal of Microscopy, Vol. 107, Pt 2,July 1976, pp. 177-182. Revised paper accepted 10June 1976
Peculiar artefacts after fixation with glutaraldehyde and osmium tetroxide
by E N I K O K U T H Yand Z S O L T C S A P O , Department of Pathology, University of Medicine, 6701 Szeged, Pf.401, Hungary
SUMMARY
Peculiar electron dense granules were observed, in otherwise well preserved structures of the heart muscle and kidney of the rat after a 5-10 min perfusion, a 2 h fixation in the modified Karnovsky solution, followed by a postfixation in Millonig buffer containing 10; 0 ~ 0 4 The . granules appeared most frequently on membranes, the surface of the red blood cells and in pinocytotic vacuoles, in stained and unstained sections alike. No artefacts were produced when the same buffer was used for both pre- and postfixation, or when the postfixative contained Verona1 buffer. INTRODUCTION
Double fixation with glutaraldehyde and osmium tetroxide (Sabatini et al., 1963) is probably the most frequently used method in preparing specimens for electron microscopy. It has two advantages: the penetration of the cessation of blood circulation to a minimum; it preserves the activity of certain enzymes (Sabatini et al., 1963; Olah & Rolich, 1966; Ericson & Biberfeld, 1967) and so insures the structural integrity for the ultrastructural-histochemical investigations. As most of the fixatives, the aldehydes also produce artefacts. While the investigators are familiar with these artefacts and how to avoid them (Olah & Rohlich, 1966; Anderson, 1967; Reimer, 1967; Pearse, 1968; Robertson & Schultz, 1970), they can cause severe troubles during the introduction of new methods (Gil & Weibel, 1968; Gil, 1972), or even lead to erroneous eonclusions. During investigations made on heart muscle in this laboratory, a relatively rare but very disturbing artefact was observed. Following the double fixation of muscle, ovoid or globular electron dense granules of 200-1000 nm diameter appeared. They were homogenous with sharp or slightly rugged boundaries. They could be found in a great number throughout the specimen, especially along the membranes on the surfaces of the erythrocytes and in the pinocytotic vacuoles. We call attention to this artefact and the possibility of its elimination. MATERIALS A N D M E T H O D
Animals Inbred Wistar rats of both sexes were used throughout the experiments with no previous selection.
177
Eniko Kuthy and Zsolt Csapo Fixation method Laparatomy was made under Nembutal anaesthesia, 0.5 ml of 0.8’,, solution, injected parenterally. Perfusion of the coronary artery with modified Karnovsky I solution (Huttner et al., 1971) was made through a hypodermic needle inserted into the aorta from a retrograde direction. The pieces of tissue selected for electron microscopy were treated further in Karnovsky I1 solution (Huttner et aZ., 1971) for 2 h followed by washing in 0.1 M sodium cacodylate buffer p H 7.2 and postfixation with 1% oso4 in Millonig buffer, p H 7.4 (Millonig, 1961). The specimens were washed again in Millonig buffer, dehydrated with alcohol and embedded in Durcupan ACM (Fluka). Selections were made with an LKB ultrotome I apparatus and examined either after staining it according to Reynolds (1963), or without staining through an SEM 3-1 and JEOL JEM 100 B electron microscope. T o find the conditions essential for the appearance of the artefact, various steps in the method were combined systematically, keeping the rest unchanged. 1. Tissue and place of excision. (1) Heart: (a) atrial myocardium, (b) ventricular myocardium. (2) Kidney: (a) cortex, (b) medulla. 2. Prejixation. (1) Perfusion. (2) Immersion. 3. Glutaraldehyde. (1) Merck glutaraldehyde (perfusion) (a) ‘old’: It was kept at room temperature for 6 months, then at 4°C for several years. (b) ‘fresh’: after delivery it was kept at 4°C until used (3 months). (2) Schuchardt glutaraldehyde: sealed in ampoules (perfusion). (3) TAAB glutaraldehyde (immersion). (4) Polysciences glutaraldehyde, EM grade (immersion). 4. Karnovsky solution. Replaced by (1) Paraformaldehyde, Merck (immersion). (2) Glutaradehyde, Merck (immersion). 5. Prejixation buffer. (1) Sodium cacodylate without CaCle. (2) Millonig buffer. 6. Postjixation buffer. (1) Different samples of Millonig buffer prepared in different laboratories. (2) Veronal buffer, Palade (1952) fixative. (3) Sodium cacodylate buffer. (4) Postfixation omitted. 7. Washing buffer (after fixation). (1) Veronal buffer after Millonig postfixation. (2) Millonig buffer following Palade postfixation. 8. Prejixation omitted. (1) Fixation with oso4 dissolved in Millonig buffer. (2) Fixation with OsO4 dissolved in Veronal buffer. (3) Fixation with OsO4 dissolved in sodium cacodylate. 9. Staining omitted. The effect of purity of the chemicals used was also studied as follows. (1) Glutaraldehyde. Ultra-violet absorption was measured between 220-320 nm, in 4 n m steps, with a Beckman spectrophotometer at 25°C in a x 50 dilution in distilled water. Distilled water was used as a control (Anderson, 1967). Merck, Schuchardt and Polysciences glutaraldehyde samples were tested in this way as the amount available of TAAB glutaraldehyde was not enough for the purpose. The purification index was made according to Weakley (1974). (2) MiZZonig buffer. Contamination with calcium and heavy metal phosphates and NaOH was tested qualitatively with the standard methods of chemical analysis.
RESULTS
The results are shown in Table 1 and the essential points can be summarized as follows. The appearance of the artefacts described depends on the buffer used for pre- and postfixation (Figs. 1,2). No artefacts are produced when the same buffer is used for both pre- and postfixation, or when the post-fixative contains Veronal buffer. No artefacts appeared if the prefixative contained paraformaldehyde or Poly-
178
Table 1. Prefixation :
Immersi Perfusion TAAB GA Polysciences Merck GA Merck GA + Schuchardt Na-cacodyla GA GA paraNa-cacodylate paraformalbuffer Na-cacodylate buffer dehyde formaldehyde buffer Na-cacodylate Na-cacodylate buffer buffer (Karnovsky (Karnovsky 1, 11) I, 11)
+
Postfixation: OsO4 Millonig buffer
*
*
*
t
*
Postfixation: oso4 Verona1 buffer
t
i
t
t
t
Postfixation : oso4 Na-cacodylate buffer
+
+
+ +
t
t
t
No postfixation
+
+
+
+
++
+
+
* Artefact appeared; t no artefact; 1 no data; GA glutaraldehyde.
Eniko Kuthy and Zsolt Csapo
Fig. 1. Heart. Fixation : perfusion with Schuchardt glutaraldehyde in Na-cacodylate buffer followed by OsO4 in Millonig buffer. Uranylacetate, lead citrate, x 50,000.
Fig. 2. Heart. Fixation : perfusion with Schuchardt glutaraldehyde in Na-cacodylate buffer followed by oso4 in Millonig buffer. Without staining, x 15,000.
180
Fixation artefacts sciences glutaraldehyde, EM grade. However, they appeared when glutaraldehyde from any other source was used, irrespective of the grade of purity. The U.V.spectrum of the Merck glutaraldehyde showed two absorption maxima at 235 nm and 280 run for the decomposition products and the dialdehyde, respectively. Both ‘old’ and ‘fresh’ samples of Merck glutaraldehyde showed the maximum at 235 nm. The purification index of ‘old’ glutaraldehyde was 5.14, ‘fresh‘ glutaraldehyde was 2.49. Schuchardt glutaraldehyde had only one maximum at 280 nm, its purification index was 0.304. In spite of this, there was no difference in the number of contaminating granules. The Polysciences glutaraldehyde, EM grade had a very low purification index: 0.236 (pH 4.3). Production of the artefactual granules was independent of the presence of CaClz, of staining (Figs. 1, 2), or of the kind of tissue. Summarizing the results, the contaminating granules in the heart and kidney preparations were formed only in the simultaneous presence of glutaraldehyde, except Polysciences, EM grade, Millonig buffer, sodium cacodylate buffer and Os04 (Table 1). The granules formed cannot be washed out of the tissues after the postfixation. DISCUSSION
From the experiments it can be seen the appearance and number of the artefact granules are not influenced by the nature of the fixed tissue, the mode of prefixation as well as washing after postfixation and staining the sections. They are stable, not removable by washing, and are formed during the fixation of the material. We cannot take up a definite position on the question of its nature and the mechanism of its origin. Undoubtedly, the granules do not originate from the contaminations in the chemical used. It is ruled out both by the controlled chemical purity of the materials and by the fact that the fixatives or their buffers did not cause any contamination when used separately or in combinations. However, the appearance of the granules is likely to be connected with the glutaraldehyde prefixation. They do not appear when glutaraldehyde was completely omitted (fixation with formaldehyde or Millonig 0 ~ 0 4 ) .Degree of purity of the glutaraldehyde can also be important as no artefact was formed with Polysciences, EM grade glutaraldehyde. Sodium cacodylate and Millonig buffers have to be in some connection with each other because the granules appeared only in the case when both these buffers were used for pre- and postfixation. In the literature available for us, Gil (1972) reported similar artefacts (Gil & Weibel, 1968) and considered them as an ultrastructural appearance of the plasmal reaction (Pearse, 1961). Our results are inconsistent with the plasmal reaction because no artefact was formed after prefixation with paraformaldehyde or glutaraldehyde of very high purity. On the basis of these experiments we consider there is formation of a complex, i.e. a chelate compound which could be composed of the decomposition products of glutaraldehyde, substances dissolved from the tissue being fixed, Na-cacodylate, some component of the Millonig buffer and 0 ~ 0 4 T. o present the formation of this complex compound, one of the components should be eliminated. The easiest way to achieve this was to exchange one of the buffers for another one (e.g. Millonig buffer for Verona1 buffer) or to use the same buffer for both the pre- and postfixation. ACKNOWLEDGMENTS
The authors are grateful to P. Kasa, M.D. and G. Szepesy, Ph.D., Central
181
Eniko Kuthy and Zsolt Csapo Research Laboratory, University of Medicine, Szeged for the spectrophotometric investigations of glutaraldehyde samples.
References Anderson, P.J. (1967) Purification and quantitation of glutaraldehyde and its effect on several enzyme activities in skeletal muscle. 3. Histochem. Cytochem. 15, 652-661. Ericson, I.L.E. & Biberfeld, P. (1967) Studies on aldehyde fixation. Fixation rates and their relation to fine structure and some histochemical reactions in liver. Lab. Invest. 17, 28 1-298. Gil, I. (1972) Effect of tricomplex fixation on lung tissue. 3. Ultrastruct. Res. 40, 122-131. Gil, I. & Weibel, E.R. (1968) The role of buffers in lung fixation with glutaraldehyde and osmium tetroxide. 3. Ultrastruct. Res. 25, 331-348. Hiittner, I., Rona, G. & More, R.H. (1971) Fibrin deposition within cardiac muscle cells in malignant hypertension. An electron microscopic study. Arch. Pathol. 91, 19-28. Millonig, G. (1961) A modified procedure for lead staining of thin sections. 3. Biophys. Biochem. Cytol. 11, 736-739. Olah, J. & Rohlich, P. (1966) Peculiar membrane configurations after fixation in glutaraldehyde. Acta Biol. Acad. Sci. Hung. 17, 65-73. Palade, G. E. (1952) A study of fixation for electron microscopy. 3. Exp. Med. 95, 285-296. Pearse, A.G.E. (1968) Histochemistry Theoretical and Applied (3rd edn), pp. 74-76. Little Brown & Company, Boston, U.S.A. Pearse, A.G.E. (1961) Histochemistry Theoretical and Applied (2nd edn), pp. 347. Churchill, London. Reimer, L. (1967) Elekronenmikroskopische Untersuchungs- und Praparations-methoden, pp. 436-437. Zweite Auflage, Springer-Verlag, Berlin, Heidelberg, New York. Reynolds, E.S. (1963) The use of lead citrate at high p H as an electron-opaque stain in electron microscopy. 3. Cell Biol. 17, 208-213. Robertson, E.A. & Schultz, R.L. (1970) The impurities in commercial glutaraldehyde and their effect on the fixation of brain. J . Ultrastruct. Res. 30, 275-287. Sabatini, D., Bensch, K. & Barnett, R.J. (1963) The preservation of cellular ultrastructure and enzymatic activity by aldehyde fixation. 3. Cell Biol. 17, 19-58. Weakley, B. S. (1974) A comparison of three different electron microscopical grade glutaraldehydes used to fix ovarian tissue. 3. Microsc. 101, 127-141.
182
Peculiar artefacts after fixation with glutaraldehyde and osmium tetroxide
by E N I K O K U T H Yand Z S O L T C S A P O , Department of Pathology, University of Medicine, 6701 Szeged, Pf.401, Hungary
SUMMARY
Peculiar electron dense granules were observed, in otherwise well preserved structures of the heart muscle and kidney of the rat after a 5-10 min perfusion, a 2 h fixation in the modified Karnovsky solution, followed by a postfixation in Millonig buffer containing 10; 0 ~ 0 4 The . granules appeared most frequently on membranes, the surface of the red blood cells and in pinocytotic vacuoles, in stained and unstained sections alike. No artefacts were produced when the same buffer was used for both pre- and postfixation, or when the postfixative contained Verona1 buffer. INTRODUCTION
Double fixation with glutaraldehyde and osmium tetroxide (Sabatini et al., 1963) is probably the most frequently used method in preparing specimens for electron microscopy. It has two advantages: the penetration of the cessation of blood circulation to a minimum; it preserves the activity of certain enzymes (Sabatini et al., 1963; Olah & Rolich, 1966; Ericson & Biberfeld, 1967) and so insures the structural integrity for the ultrastructural-histochemical investigations. As most of the fixatives, the aldehydes also produce artefacts. While the investigators are familiar with these artefacts and how to avoid them (Olah & Rohlich, 1966; Anderson, 1967; Reimer, 1967; Pearse, 1968; Robertson & Schultz, 1970), they can cause severe troubles during the introduction of new methods (Gil & Weibel, 1968; Gil, 1972), or even lead to erroneous eonclusions. During investigations made on heart muscle in this laboratory, a relatively rare but very disturbing artefact was observed. Following the double fixation of muscle, ovoid or globular electron dense granules of 200-1000 nm diameter appeared. They were homogenous with sharp or slightly rugged boundaries. They could be found in a great number throughout the specimen, especially along the membranes on the surfaces of the erythrocytes and in the pinocytotic vacuoles. We call attention to this artefact and the possibility of its elimination. MATERIALS A N D M E T H O D
Animals Inbred Wistar rats of both sexes were used throughout the experiments with no previous selection.
177
Eniko Kuthy and Zsolt Csapo Fixation method Laparatomy was made under Nembutal anaesthesia, 0.5 ml of 0.8’,, solution, injected parenterally. Perfusion of the coronary artery with modified Karnovsky I solution (Huttner et al., 1971) was made through a hypodermic needle inserted into the aorta from a retrograde direction. The pieces of tissue selected for electron microscopy were treated further in Karnovsky I1 solution (Huttner et aZ., 1971) for 2 h followed by washing in 0.1 M sodium cacodylate buffer p H 7.2 and postfixation with 1% oso4 in Millonig buffer, p H 7.4 (Millonig, 1961). The specimens were washed again in Millonig buffer, dehydrated with alcohol and embedded in Durcupan ACM (Fluka). Selections were made with an LKB ultrotome I apparatus and examined either after staining it according to Reynolds (1963), or without staining through an SEM 3-1 and JEOL JEM 100 B electron microscope. T o find the conditions essential for the appearance of the artefact, various steps in the method were combined systematically, keeping the rest unchanged. 1. Tissue and place of excision. (1) Heart: (a) atrial myocardium, (b) ventricular myocardium. (2) Kidney: (a) cortex, (b) medulla. 2. Prejixation. (1) Perfusion. (2) Immersion. 3. Glutaraldehyde. (1) Merck glutaraldehyde (perfusion) (a) ‘old’: It was kept at room temperature for 6 months, then at 4°C for several years. (b) ‘fresh’: after delivery it was kept at 4°C until used (3 months). (2) Schuchardt glutaraldehyde: sealed in ampoules (perfusion). (3) TAAB glutaraldehyde (immersion). (4) Polysciences glutaraldehyde, EM grade (immersion). 4. Karnovsky solution. Replaced by (1) Paraformaldehyde, Merck (immersion). (2) Glutaradehyde, Merck (immersion). 5. Prejixation buffer. (1) Sodium cacodylate without CaCle. (2) Millonig buffer. 6. Postjixation buffer. (1) Different samples of Millonig buffer prepared in different laboratories. (2) Veronal buffer, Palade (1952) fixative. (3) Sodium cacodylate buffer. (4) Postfixation omitted. 7. Washing buffer (after fixation). (1) Veronal buffer after Millonig postfixation. (2) Millonig buffer following Palade postfixation. 8. Prejixation omitted. (1) Fixation with oso4 dissolved in Millonig buffer. (2) Fixation with OsO4 dissolved in Veronal buffer. (3) Fixation with OsO4 dissolved in sodium cacodylate. 9. Staining omitted. The effect of purity of the chemicals used was also studied as follows. (1) Glutaraldehyde. Ultra-violet absorption was measured between 220-320 nm, in 4 n m steps, with a Beckman spectrophotometer at 25°C in a x 50 dilution in distilled water. Distilled water was used as a control (Anderson, 1967). Merck, Schuchardt and Polysciences glutaraldehyde samples were tested in this way as the amount available of TAAB glutaraldehyde was not enough for the purpose. The purification index was made according to Weakley (1974). (2) MiZZonig buffer. Contamination with calcium and heavy metal phosphates and NaOH was tested qualitatively with the standard methods of chemical analysis.
RESULTS
The results are shown in Table 1 and the essential points can be summarized as follows. The appearance of the artefacts described depends on the buffer used for pre- and postfixation (Figs. 1,2). No artefacts are produced when the same buffer is used for both pre- and postfixation, or when the post-fixative contains Veronal buffer. No artefacts appeared if the prefixative contained paraformaldehyde or Poly-
178
Table 1. Prefixation :
Immersi Perfusion TAAB GA Polysciences Merck GA Merck GA + Schuchardt Na-cacodyla GA GA paraNa-cacodylate paraformalbuffer Na-cacodylate buffer dehyde formaldehyde buffer Na-cacodylate Na-cacodylate buffer buffer (Karnovsky (Karnovsky 1, 11) I, 11)
+
Postfixation: OsO4 Millonig buffer
*
*
*
t
*
Postfixation: oso4 Verona1 buffer
t
i
t
t
t
Postfixation : oso4 Na-cacodylate buffer
+
+
+ +
t
t
t
No postfixation
+
+
+
+
++
+
+
* Artefact appeared; t no artefact; 1 no data; GA glutaraldehyde.
Eniko Kuthy and Zsolt Csapo
Fig. 1. Heart. Fixation : perfusion with Schuchardt glutaraldehyde in Na-cacodylate buffer followed by OsO4 in Millonig buffer. Uranylacetate, lead citrate, x 50,000.
Fig. 2. Heart. Fixation : perfusion with Schuchardt glutaraldehyde in Na-cacodylate buffer followed by oso4 in Millonig buffer. Without staining, x 15,000.
180
Fixation artefacts sciences glutaraldehyde, EM grade. However, they appeared when glutaraldehyde from any other source was used, irrespective of the grade of purity. The U.V.spectrum of the Merck glutaraldehyde showed two absorption maxima at 235 nm and 280 run for the decomposition products and the dialdehyde, respectively. Both ‘old’ and ‘fresh’ samples of Merck glutaraldehyde showed the maximum at 235 nm. The purification index of ‘old’ glutaraldehyde was 5.14, ‘fresh‘ glutaraldehyde was 2.49. Schuchardt glutaraldehyde had only one maximum at 280 nm, its purification index was 0.304. In spite of this, there was no difference in the number of contaminating granules. The Polysciences glutaraldehyde, EM grade had a very low purification index: 0.236 (pH 4.3). Production of the artefactual granules was independent of the presence of CaClz, of staining (Figs. 1, 2), or of the kind of tissue. Summarizing the results, the contaminating granules in the heart and kidney preparations were formed only in the simultaneous presence of glutaraldehyde, except Polysciences, EM grade, Millonig buffer, sodium cacodylate buffer and Os04 (Table 1). The granules formed cannot be washed out of the tissues after the postfixation. DISCUSSION
From the experiments it can be seen the appearance and number of the artefact granules are not influenced by the nature of the fixed tissue, the mode of prefixation as well as washing after postfixation and staining the sections. They are stable, not removable by washing, and are formed during the fixation of the material. We cannot take up a definite position on the question of its nature and the mechanism of its origin. Undoubtedly, the granules do not originate from the contaminations in the chemical used. It is ruled out both by the controlled chemical purity of the materials and by the fact that the fixatives or their buffers did not cause any contamination when used separately or in combinations. However, the appearance of the granules is likely to be connected with the glutaraldehyde prefixation. They do not appear when glutaraldehyde was completely omitted (fixation with formaldehyde or Millonig 0 ~ 0 4 ) .Degree of purity of the glutaraldehyde can also be important as no artefact was formed with Polysciences, EM grade glutaraldehyde. Sodium cacodylate and Millonig buffers have to be in some connection with each other because the granules appeared only in the case when both these buffers were used for pre- and postfixation. In the literature available for us, Gil (1972) reported similar artefacts (Gil & Weibel, 1968) and considered them as an ultrastructural appearance of the plasmal reaction (Pearse, 1961). Our results are inconsistent with the plasmal reaction because no artefact was formed after prefixation with paraformaldehyde or glutaraldehyde of very high purity. On the basis of these experiments we consider there is formation of a complex, i.e. a chelate compound which could be composed of the decomposition products of glutaraldehyde, substances dissolved from the tissue being fixed, Na-cacodylate, some component of the Millonig buffer and 0 ~ 0 4 T. o present the formation of this complex compound, one of the components should be eliminated. The easiest way to achieve this was to exchange one of the buffers for another one (e.g. Millonig buffer for Verona1 buffer) or to use the same buffer for both the pre- and postfixation. ACKNOWLEDGMENTS
The authors are grateful to P. Kasa, M.D. and G. Szepesy, Ph.D., Central
181
Eniko Kuthy and Zsolt Csapo Research Laboratory, University of Medicine, Szeged for the spectrophotometric investigations of glutaraldehyde samples.
References Anderson, P.J. (1967) Purification and quantitation of glutaraldehyde and its effect on several enzyme activities in skeletal muscle. 3. Histochem. Cytochem. 15, 652-661. Ericson, I.L.E. & Biberfeld, P. (1967) Studies on aldehyde fixation. Fixation rates and their relation to fine structure and some histochemical reactions in liver. Lab. Invest. 17, 28 1-298. Gil, I. (1972) Effect of tricomplex fixation on lung tissue. 3. Ultrastruct. Res. 40, 122-131. Gil, I. & Weibel, E.R. (1968) The role of buffers in lung fixation with glutaraldehyde and osmium tetroxide. 3. Ultrastruct. Res. 25, 331-348. Hiittner, I., Rona, G. & More, R.H. (1971) Fibrin deposition within cardiac muscle cells in malignant hypertension. An electron microscopic study. Arch. Pathol. 91, 19-28. Millonig, G. (1961) A modified procedure for lead staining of thin sections. 3. Biophys. Biochem. Cytol. 11, 736-739. Olah, J. & Rohlich, P. (1966) Peculiar membrane configurations after fixation in glutaraldehyde. Acta Biol. Acad. Sci. Hung. 17, 65-73. Palade, G. E. (1952) A study of fixation for electron microscopy. 3. Exp. Med. 95, 285-296. Pearse, A.G.E. (1968) Histochemistry Theoretical and Applied (3rd edn), pp. 74-76. Little Brown & Company, Boston, U.S.A. Pearse, A.G.E. (1961) Histochemistry Theoretical and Applied (2nd edn), pp. 347. Churchill, London. Reimer, L. (1967) Elekronenmikroskopische Untersuchungs- und Praparations-methoden, pp. 436-437. Zweite Auflage, Springer-Verlag, Berlin, Heidelberg, New York. Reynolds, E.S. (1963) The use of lead citrate at high p H as an electron-opaque stain in electron microscopy. 3. Cell Biol. 17, 208-213. Robertson, E.A. & Schultz, R.L. (1970) The impurities in commercial glutaraldehyde and their effect on the fixation of brain. J . Ultrastruct. Res. 30, 275-287. Sabatini, D., Bensch, K. & Barnett, R.J. (1963) The preservation of cellular ultrastructure and enzymatic activity by aldehyde fixation. 3. Cell Biol. 17, 19-58. Weakley, B. S. (1974) A comparison of three different electron microscopical grade glutaraldehydes used to fix ovarian tissue. 3. Microsc. 101, 127-141.
182
