AP M I S 99: 257-261, 1991
Redistribution of myeloperoxidase and elastase in human neutrophil granulocytes L. NASSBERGER Department of Medical Microbiology Clinical Immunology University Hospital Lund, Lund, Sweden
Nassberger, L. Redistribution of myeloperoxidase and elastase in human neutrophil granulocytes. APMIS 99: 257-261, 1991. Both myeloperoxidase (MPO) and elastase were found to be redistributed in ethanol-and Bouin-fixed human neutrophil granulocytes. Air-dried and formalin-fixed cells, stained for MPO, appeared with a cytoplasmic fluorescence and unstained nuclei. In elastase air-dried cells a faint perinuclear staining was also detected. Otherwise the redistribution pattern was very similar for MPO and elastase. It was found that both elastase and MPO were most efficiently extracted with Bouin’s fixative. A discontinuous fluorescence was seen especially for MPO in ethanol-fixed cells. Nuclear membrane staining was verified with laser scan microscopy. This technique also visualized a scattered fluorescence in the nuclear matrix for MPO. Key words: Myeloperoxidase; elastase, redistribution; fixation; neutrophil granulocytes Lennart Nassberger, Department of Medical Microbiology, Solvegatan 23, S-223 62 Lund, Sweden
Human granulocytes are used a s substrate for detecting antineutrophil cytoplasmic antibodies (ANCA) (1 1) and granulocyte-specific A N A (GS-ANA) (12). GS-ANA may cause several different fluorescent patterns. These different patterns have earlier been stated t o reflect either common antibodies against antinuclear antigens (ANA), or antibodies directed against a specific antigen in neutrophil granulocytes. It has now recently been realized that a peripheral nuclear fluorescence may indicate antibodies against myeloperoxidase (4,6). This is probably due t o the fixation procedure used. Therefore, when a perinuclear staining is viewed, this may also indicate occurrence of non-nuclear antibodies due t o cytoplasmic redistribution of antigens. The aim of the present study was t o undertake an investigation of redistribution during differReceived April 9, 1990. Accepted August 28, 1990.
ent fixations of the two lysosomal enzymes, myeloperixodase and elastase, respectively.
MATERIALS A N D M E T H O D S
Chemicals All chemicals used were of analytical purity. Cell preparation Human neutrophil granulocytes from healthy subjects were isolated according to Boyum (2). Briefly, blood was drawn into tubes containing heparin (12.5 IE/ml) and 2 ml dextran and allowed to stand at room temperature for 1 h at a 45’’ angle. The supernatant was pipetted off and mixed with an equal volume of saline. 5 to 6 ml of the diluted supernatant was layered onto 5 ml lymfoprep. (Nyegaard & Co, Oslo, Norway) and centrifuged at 500 g,, for 18 minutes. Cell preparations were prepared using alcohol-washed slides on a cytocentrifuge (Shandon, England) 1000 rpm for 5 minutes and fixation carried out immediately.
251
NASSBERGER
Experimental design Redistribution of myeloperoxidase ( M P O ). Neutrophi1 granulocytes were fixed as follows: a/ air-drying (n= 12), b/ in 99% ethanol at 4°C for 5-10 minutes (n= 20), c/ by a spray histofix (Histolab Gothenburg, Sweden) (n = 10); histofix is a water-soluble fixative composed of ethanol, butanelpropane and polyethyleneglycol, d/ formalin lo%, 5-10 minutes (n= lo), e/ buffered formalin 4%, 5-10 minutes (n= 10) and f/ in Bouin’s fixative for 5-10 minutes (n= 18). Bouin’s fixative is an acid fixative composed of picric acid dissolved in water, formaldehyde and glacial acetic acid. 15:5:1. After washing ( x 3) with phosphatebuffered Saline (PBS) pH 7.2, cells were incubated with rabbit anti-human MPO diluted 1:lOO (Dakopatts A/S, Glostrup, Denmark) for 30 minutes. Washing was performed three times in PBS. Thereafter cells were incubated with swine-anti-rabbit IgG conjugated with FITC (Dakopatts A/S, Glostrup, Denmark) diluted 1:80 for 30 minutes. After washng ( x 3) in PBS the slides were mounted. Redistribution of elastase. The same fixation, number of experiments and washing procedures as for MPO were carried out except that acetone was also used as a fixative. The cells were incubated with sheep-anti-human elastase (kindly supplied by Dr A Heubner, E Merck, Darmstadt, West Germany diluted 1:200, for 30 minutes. Thereafter incubation with a FITC-conjugated rabbit-sheep immunoglobulin (Dakopatts A/S, Glostrup, Denmark) diluted 1:200 was carried out. Normal serum In control experiments normal sera diluted 1:80 were incubated instead of the primary antibody. Microscopy Fluorescence microscopy was performed with a Zeiss Axioskop and Laser scan fluorescence. with a Zeiss laser scan microscope (LSM).
RESULTS
MPO distribution In unfixed granulocytes, as well as in formalin and buffered formalin, a strong cytoplasmic fluorescence was seen with unstained nuclei (Fig. 1). In ethanol and in cells fixed with histofix a clear nuclear fluorescent rim was found (Fig. 2). A discontinuous nuclear membrane fluorescence was seen in some cells. The same pattern was also noted for cells fixed with Bouin’s fixative. The fluorescence seemed t o be localized mainly to the nuclear membrane. This was verified by laser scan microscopy. A discon-
258
Fig. 1. Unfixed neutrophil granulocytes with cytoplasmic localization of myeloperoxidase and unstained nuclei.
tinuous fluorescence in the nuclear membrane was further visualized (Fig. 3). However, in several of the cells a scattered fluorescence was also detected within the nuclear matrix by laser scan microscopy (Fig. 4). Elastase distribution A cytoplasmic fluorescence was noted in formalin-fixed cells with unstained nuclei. In many of the air-fixed cells a moderate perinuclear staining was seen, in contrast t o myeloperoxidase staining. Ethanol fixation caused the perinuclear membrane t o become more prominent but a faint cytoplasmic fluorescence was also seen in several cells (Fig. 5 ) . Perinuclear staining was most pronounced in cells fixed with Bouin’s fixative and acetone, and had the same appearance as that seen for MPO. Normal serum Fluorescence was absent in cells incubated with normal sera.
DISCUSSION We have shown here that a redistribution of both MPO and elastase takes place when gra-
MY ELOPEROXIDASE AND ELASTASE IN GRANULOCYTES
nulocytes are fixed with ethanol, Bouin’s fixative and acetone for elastase. The redistribution pattern was very similar when MPO and elastase were compared. From these stuides Bouin’s fixative was found to cause the strongest extraction of MPO and elastase. Both MPO and elastase are lysosomal enzymes localized to azurophil granules. MPO has, however, also been demonstrated in nuclei of myeloid cells (5). Laser scan microscopy makes it possible to section through the cell and furthermore avoid stray light. Therefore laser scan microscopy gives an almost optimal picture near the limit of resolution. By this microscopic technique we were able to localize most of the fluorescence of ethanol and cells fixed with Bouin to the nuclear envelope. However, small amounts of MPO could also be detected within the nuclear matrix, which is in accordance with the findings of Murao and co-workers ( 5 ) .
Earlier immunofluorescence studies with MPO have shown a perinuclear staining (3) in ethanol fixation. The same phenomenon as reported here has also been found for lactoferrin (1). Briggs and co-workers found a positive nuclear staining reaction in unfixed cells. This is in accordance with our findings for elastase. This pattern is in contrast with MPO where a strict cytoplasmic fluorescence in unfixed cells was noted. This has also earlier been found for MPO (9). The occasional discontinuous fluorescence especially seen for MPO may be due to an unequal distribution of the enzyme from the cytoplasm or damage to the nuclear membrane. Alcohol-fixed granulocytes are used for screening of antibodies against cytoplasmic components. This study shows clearly that both myeloperoxidase and elastase give rise to a perinuclear staining pattern during fixation. There-
Fig. 2. MPO distribution in neutrophil granulocytes fixed with ethanol. A typical intensive membrane staining is seen.
Fig. 3. Laser scan microscopy showing nuclear envelope staining with discontinuous fluorescence of
MPO.
2 59
NASSBERGER
Fig. 4. Laser scan microscopy showing scattered fluorescence of MPO within the nuclear matrix in ethanol-fixed cells.
fore, when a nuclear membrane fluorescence appears, antibodies against MPO or elastase have to be verified by ELISA technique. We have undertaken studies on the occurrence of anti-MPO in different systemic vasculitic disorders and found this typical nuclear membrane fluorescence in all cases (8, 7). Furthermore, antibodies against lactoferrin have recently been described in a patient with systemic vasculitis (10). Elastase redistribution has to our knowledge not been reported before. As a result of our study we recommend that antibodies directed against elastase also be considered when viewing this pattern ( 6 ) . This study was supported by grants from A . Osterlund’s Foundation, the Medical Research Council of the Swedish Life Insurance Companies, The Foundation of Aid to the disabled in Scandinavia, and King Gustav V’s 80th Birthday Fund. I am grateful to Dr H . Kapit-
260
Fig. 5 . Elastase distribution in neutrophil granulocytes fixed with ethanol. A perinuclear staining is seen. za at the Research Department, Zeiss, Oberkochen, West Germany, for help with the laser scan microscopy.
REFERENCES Briggs R . C., Glass II u! E . Montiel M . M . & Hnilica L. S.: Lactoferrin: Nuclear localization in the human neutrophilic granulocyte. I. Histochem. & Cytochem. 29: 1128-1136, 1981. B6yum A.: Separation of leucoytes from blood and bone marrow. Scand. J. Clin. Invest. Suppl. 977: 21. 1968. Deiman u!, Seitz M . , Gemsa D. & Fahiani H . D.: Endogenous peroxidase in the nuclear envelope and endoplasmic reliculum of human monocytes in vitro: association with arachidonic acid metabolism. Blood. 64: 491498, 1984. Falk R . I. & Jennette J . C.: Antineutrophil cytoplasmic autoantibodies with specificity for myeloperoxidase in patients with systemic vasculitis and idiopathic necrotizing and cresentic
MYELOPEROXIDASE AND ELASTASE IN GRANULOCYTES
5.
6. 7.
8.
glomerulonephritis. New Engl. J. Med. 318: 1651-57. 1988. Murao S. I., Stevens F: I., It0 A . & Huberman E.: Myeloperoxidase: A myeloid cell nuclear antigen with DNA-binding properties. Proc. Natl. Acad. Sci. USA. 85: 1232-1236, 1988. Nassberger L., Jonsson H . , Sjoholm A . G . , Sturfelt G . & Heubner A.: Circulating anti-elastase in systemic lupus erythematosus. Lancet i: 509, 1989. Nassberger L., Sjoholm A . G . .Jonsson H., Sturfelt G. & Akesson A,: Autoantibodies against neutrophi1 cytoplasm components in systemic lupus erythematosus and in hydralazine-induced lupus. Clin. & Exp. Immunology 81: 380-383, 1990. Nassberger L . . Sjoholm A . G. & Thysell H.: Antimyeloperoxidase antibodies in patients with extracapillary glomerulonephritis Nephron 56: 152-156, 1990.
9. Pryzwansky K . B., Martin L. E. & Spitznagel I . K.: Immunocytochemical localization of myeloperoxidase, lactoferrin, lysozyme and neutrophilic granulocytes. I. Reticuloendothel. SOC. 24: 295-310, 1978. 10. Thompson R . A . & Lee S. S.: Antineutrophil cytoplasmic antibodies. Lancet i: 67G671, 1989. 11. van der Woude F: I., Lobatto S., Permin H . , van der Giessen M . , Rasmussen U!, Wiik A . , van Es L. A , & van der Hem G . H.: Autoantibodies against neutrophils and monocytes: tool for diagnosis and marker of disease activity in Wegener’s granulomatosis. Lancet i: 423429, 1985. 12. Wiik A . & Munthe E.: Restrictions among heavy and light chain determinants of granulocyte-specific antinuclear factors. Immunology 23: 53-57, 1972.
26 1
Redistribution of myeloperoxidase and elastase in human neutrophil granulocytes L. NASSBERGER Department of Medical Microbiology Clinical Immunology University Hospital Lund, Lund, Sweden
Nassberger, L. Redistribution of myeloperoxidase and elastase in human neutrophil granulocytes. APMIS 99: 257-261, 1991. Both myeloperoxidase (MPO) and elastase were found to be redistributed in ethanol-and Bouin-fixed human neutrophil granulocytes. Air-dried and formalin-fixed cells, stained for MPO, appeared with a cytoplasmic fluorescence and unstained nuclei. In elastase air-dried cells a faint perinuclear staining was also detected. Otherwise the redistribution pattern was very similar for MPO and elastase. It was found that both elastase and MPO were most efficiently extracted with Bouin’s fixative. A discontinuous fluorescence was seen especially for MPO in ethanol-fixed cells. Nuclear membrane staining was verified with laser scan microscopy. This technique also visualized a scattered fluorescence in the nuclear matrix for MPO. Key words: Myeloperoxidase; elastase, redistribution; fixation; neutrophil granulocytes Lennart Nassberger, Department of Medical Microbiology, Solvegatan 23, S-223 62 Lund, Sweden
Human granulocytes are used a s substrate for detecting antineutrophil cytoplasmic antibodies (ANCA) (1 1) and granulocyte-specific A N A (GS-ANA) (12). GS-ANA may cause several different fluorescent patterns. These different patterns have earlier been stated t o reflect either common antibodies against antinuclear antigens (ANA), or antibodies directed against a specific antigen in neutrophil granulocytes. It has now recently been realized that a peripheral nuclear fluorescence may indicate antibodies against myeloperoxidase (4,6). This is probably due t o the fixation procedure used. Therefore, when a perinuclear staining is viewed, this may also indicate occurrence of non-nuclear antibodies due t o cytoplasmic redistribution of antigens. The aim of the present study was t o undertake an investigation of redistribution during differReceived April 9, 1990. Accepted August 28, 1990.
ent fixations of the two lysosomal enzymes, myeloperixodase and elastase, respectively.
MATERIALS A N D M E T H O D S
Chemicals All chemicals used were of analytical purity. Cell preparation Human neutrophil granulocytes from healthy subjects were isolated according to Boyum (2). Briefly, blood was drawn into tubes containing heparin (12.5 IE/ml) and 2 ml dextran and allowed to stand at room temperature for 1 h at a 45’’ angle. The supernatant was pipetted off and mixed with an equal volume of saline. 5 to 6 ml of the diluted supernatant was layered onto 5 ml lymfoprep. (Nyegaard & Co, Oslo, Norway) and centrifuged at 500 g,, for 18 minutes. Cell preparations were prepared using alcohol-washed slides on a cytocentrifuge (Shandon, England) 1000 rpm for 5 minutes and fixation carried out immediately.
251
NASSBERGER
Experimental design Redistribution of myeloperoxidase ( M P O ). Neutrophi1 granulocytes were fixed as follows: a/ air-drying (n= 12), b/ in 99% ethanol at 4°C for 5-10 minutes (n= 20), c/ by a spray histofix (Histolab Gothenburg, Sweden) (n = 10); histofix is a water-soluble fixative composed of ethanol, butanelpropane and polyethyleneglycol, d/ formalin lo%, 5-10 minutes (n= lo), e/ buffered formalin 4%, 5-10 minutes (n= 10) and f/ in Bouin’s fixative for 5-10 minutes (n= 18). Bouin’s fixative is an acid fixative composed of picric acid dissolved in water, formaldehyde and glacial acetic acid. 15:5:1. After washing ( x 3) with phosphatebuffered Saline (PBS) pH 7.2, cells were incubated with rabbit anti-human MPO diluted 1:lOO (Dakopatts A/S, Glostrup, Denmark) for 30 minutes. Washing was performed three times in PBS. Thereafter cells were incubated with swine-anti-rabbit IgG conjugated with FITC (Dakopatts A/S, Glostrup, Denmark) diluted 1:80 for 30 minutes. After washng ( x 3) in PBS the slides were mounted. Redistribution of elastase. The same fixation, number of experiments and washing procedures as for MPO were carried out except that acetone was also used as a fixative. The cells were incubated with sheep-anti-human elastase (kindly supplied by Dr A Heubner, E Merck, Darmstadt, West Germany diluted 1:200, for 30 minutes. Thereafter incubation with a FITC-conjugated rabbit-sheep immunoglobulin (Dakopatts A/S, Glostrup, Denmark) diluted 1:200 was carried out. Normal serum In control experiments normal sera diluted 1:80 were incubated instead of the primary antibody. Microscopy Fluorescence microscopy was performed with a Zeiss Axioskop and Laser scan fluorescence. with a Zeiss laser scan microscope (LSM).
RESULTS
MPO distribution In unfixed granulocytes, as well as in formalin and buffered formalin, a strong cytoplasmic fluorescence was seen with unstained nuclei (Fig. 1). In ethanol and in cells fixed with histofix a clear nuclear fluorescent rim was found (Fig. 2). A discontinuous nuclear membrane fluorescence was seen in some cells. The same pattern was also noted for cells fixed with Bouin’s fixative. The fluorescence seemed t o be localized mainly to the nuclear membrane. This was verified by laser scan microscopy. A discon-
258
Fig. 1. Unfixed neutrophil granulocytes with cytoplasmic localization of myeloperoxidase and unstained nuclei.
tinuous fluorescence in the nuclear membrane was further visualized (Fig. 3). However, in several of the cells a scattered fluorescence was also detected within the nuclear matrix by laser scan microscopy (Fig. 4). Elastase distribution A cytoplasmic fluorescence was noted in formalin-fixed cells with unstained nuclei. In many of the air-fixed cells a moderate perinuclear staining was seen, in contrast t o myeloperoxidase staining. Ethanol fixation caused the perinuclear membrane t o become more prominent but a faint cytoplasmic fluorescence was also seen in several cells (Fig. 5 ) . Perinuclear staining was most pronounced in cells fixed with Bouin’s fixative and acetone, and had the same appearance as that seen for MPO. Normal serum Fluorescence was absent in cells incubated with normal sera.
DISCUSSION We have shown here that a redistribution of both MPO and elastase takes place when gra-
MY ELOPEROXIDASE AND ELASTASE IN GRANULOCYTES
nulocytes are fixed with ethanol, Bouin’s fixative and acetone for elastase. The redistribution pattern was very similar when MPO and elastase were compared. From these stuides Bouin’s fixative was found to cause the strongest extraction of MPO and elastase. Both MPO and elastase are lysosomal enzymes localized to azurophil granules. MPO has, however, also been demonstrated in nuclei of myeloid cells (5). Laser scan microscopy makes it possible to section through the cell and furthermore avoid stray light. Therefore laser scan microscopy gives an almost optimal picture near the limit of resolution. By this microscopic technique we were able to localize most of the fluorescence of ethanol and cells fixed with Bouin to the nuclear envelope. However, small amounts of MPO could also be detected within the nuclear matrix, which is in accordance with the findings of Murao and co-workers ( 5 ) .
Earlier immunofluorescence studies with MPO have shown a perinuclear staining (3) in ethanol fixation. The same phenomenon as reported here has also been found for lactoferrin (1). Briggs and co-workers found a positive nuclear staining reaction in unfixed cells. This is in accordance with our findings for elastase. This pattern is in contrast with MPO where a strict cytoplasmic fluorescence in unfixed cells was noted. This has also earlier been found for MPO (9). The occasional discontinuous fluorescence especially seen for MPO may be due to an unequal distribution of the enzyme from the cytoplasm or damage to the nuclear membrane. Alcohol-fixed granulocytes are used for screening of antibodies against cytoplasmic components. This study shows clearly that both myeloperoxidase and elastase give rise to a perinuclear staining pattern during fixation. There-
Fig. 2. MPO distribution in neutrophil granulocytes fixed with ethanol. A typical intensive membrane staining is seen.
Fig. 3. Laser scan microscopy showing nuclear envelope staining with discontinuous fluorescence of
MPO.
2 59
NASSBERGER
Fig. 4. Laser scan microscopy showing scattered fluorescence of MPO within the nuclear matrix in ethanol-fixed cells.
fore, when a nuclear membrane fluorescence appears, antibodies against MPO or elastase have to be verified by ELISA technique. We have undertaken studies on the occurrence of anti-MPO in different systemic vasculitic disorders and found this typical nuclear membrane fluorescence in all cases (8, 7). Furthermore, antibodies against lactoferrin have recently been described in a patient with systemic vasculitis (10). Elastase redistribution has to our knowledge not been reported before. As a result of our study we recommend that antibodies directed against elastase also be considered when viewing this pattern ( 6 ) . This study was supported by grants from A . Osterlund’s Foundation, the Medical Research Council of the Swedish Life Insurance Companies, The Foundation of Aid to the disabled in Scandinavia, and King Gustav V’s 80th Birthday Fund. I am grateful to Dr H . Kapit-
260
Fig. 5 . Elastase distribution in neutrophil granulocytes fixed with ethanol. A perinuclear staining is seen. za at the Research Department, Zeiss, Oberkochen, West Germany, for help with the laser scan microscopy.
REFERENCES Briggs R . C., Glass II u! E . Montiel M . M . & Hnilica L. S.: Lactoferrin: Nuclear localization in the human neutrophilic granulocyte. I. Histochem. & Cytochem. 29: 1128-1136, 1981. B6yum A.: Separation of leucoytes from blood and bone marrow. Scand. J. Clin. Invest. Suppl. 977: 21. 1968. Deiman u!, Seitz M . , Gemsa D. & Fahiani H . D.: Endogenous peroxidase in the nuclear envelope and endoplasmic reliculum of human monocytes in vitro: association with arachidonic acid metabolism. Blood. 64: 491498, 1984. Falk R . I. & Jennette J . C.: Antineutrophil cytoplasmic autoantibodies with specificity for myeloperoxidase in patients with systemic vasculitis and idiopathic necrotizing and cresentic
MYELOPEROXIDASE AND ELASTASE IN GRANULOCYTES
5.
6. 7.
8.
glomerulonephritis. New Engl. J. Med. 318: 1651-57. 1988. Murao S. I., Stevens F: I., It0 A . & Huberman E.: Myeloperoxidase: A myeloid cell nuclear antigen with DNA-binding properties. Proc. Natl. Acad. Sci. USA. 85: 1232-1236, 1988. Nassberger L., Jonsson H . , Sjoholm A . G . , Sturfelt G . & Heubner A.: Circulating anti-elastase in systemic lupus erythematosus. Lancet i: 509, 1989. Nassberger L., Sjoholm A . G . .Jonsson H., Sturfelt G. & Akesson A,: Autoantibodies against neutrophi1 cytoplasm components in systemic lupus erythematosus and in hydralazine-induced lupus. Clin. & Exp. Immunology 81: 380-383, 1990. Nassberger L . . Sjoholm A . G. & Thysell H.: Antimyeloperoxidase antibodies in patients with extracapillary glomerulonephritis Nephron 56: 152-156, 1990.
9. Pryzwansky K . B., Martin L. E. & Spitznagel I . K.: Immunocytochemical localization of myeloperoxidase, lactoferrin, lysozyme and neutrophilic granulocytes. I. Reticuloendothel. SOC. 24: 295-310, 1978. 10. Thompson R . A . & Lee S. S.: Antineutrophil cytoplasmic antibodies. Lancet i: 67G671, 1989. 11. van der Woude F: I., Lobatto S., Permin H . , van der Giessen M . , Rasmussen U!, Wiik A . , van Es L. A , & van der Hem G . H.: Autoantibodies against neutrophils and monocytes: tool for diagnosis and marker of disease activity in Wegener’s granulomatosis. Lancet i: 423429, 1985. 12. Wiik A . & Munthe E.: Restrictions among heavy and light chain determinants of granulocyte-specific antinuclear factors. Immunology 23: 53-57, 1972.
26 1
