- Email: [email protected]
Thermolabile and Calcium-dependent Serum Factor Interferes with Polymerized Actin, and Impairs Anti-actin Antibody Detection
doi:10.1006/jaut.2001.0540, available online at http://www.idealibrary.com on
Journal of Autoimmunity (2001) 17, 223–228
Thermolabile and Calcium-dependent Serum Factor Interferes with Polymerized Actin, and Impairs Anti-actin Antibody Detection Eduardo Luiz Rachid Canc¸ado1,2, Clarice Pires Abrantes-Lemos2, Lucy Santos Vilas-Boas3, Neil Ferreira Novo4, Flair Jose´ Carrilho1 and Antonio Atı´lio Laudanna1 1
Department of Gastroenterology, University of Sao Paulo School of Medicine 2 Laboratory of Medical Investigation—Immunopathology of Schistosomiasis, Institute of Tropical Medicine, University of Sao Paulo School of Medicine 3 Laboratory of Medical Investigation—Virology, Institute of Tropical Medicine, University of Sao Paulo School of Medicine 4 Department of Preventive Medicine, Federal University of Sao Paulo School of Medicine, Sao Paulo, Brazil Received 11 May 2001 Accepted 6 August 2001
The detection of anti-actin (AAA) by immunofluorescence is hindered by the presence of a serum factor. To better understand how it interferes with AAA detection, we tested sera from 20 patients with autoimmune hepatitis, and from 21 healthy adults, diluted 1:10 and prepared as follows: (A) diluted with PBS; (B) inactivated at 56°C, and diluted with PBS; (C) diluted with 34 mM EDTA/PBS; (D) heated and diluted with EDTA/PBS. To reveal AAA, a fluorescein-labelled anti-human IgG was used in the process of indirect immunofluorescence. In a parallel assay, the substrate, acetone-fixed human fibroblasts, was preincubated with sera prepared as if it were to identify AAA, but instead, a rhodamine-phalloidin was used to identify F-actin, by direct immunofluorescence. All sera from patients were reactive to AAA when heat-inactivated and/or calcium-chelated, and 60% of them when diluted with unmodified sera (P=0.004). F-actin continued to be present after preincubation with heat-inactivated or calcium-chelated sera from patients and healthy controls, and in 41.5% of reactions with unmodified serum (P=0.0000001). The heat inactivation and the calcium chelation were both efficient procedures for maintaining the microfilament structure intact after serum incubation and, therefore, for identifying AAA. © 2001 Academic Press
Key words: anti-smooth antibody, autoantibodies, autoimmune hepatitis, microfilaments, gelsolin
Introduction
example as an anti-idiotype, preventing the antibody interaction with the antigen [5]. The ability of human serum to disassemble F-actin was first detected by two different groups of investigators. Norberg et al. [4] and Chaponnier et al. [6] described this factor for the purpose of improving AAA identification, and it was called F-actin depolymerizing (destabilizing) factor. Harris et al. [7] identified a protein with a similar activity, but they called it brevin, because this protein shortened actin filaments without depolymerizing them. A similar heat-labile protein capable of controlling the cytoplasmic actin gel-sol transformation was also identified in human macrophages and further in a variety of cells and tissues [8, 9], and it was called gelsolin. The characterization of a secretory form of gelsolin, as a serum protein with features similar to those of the F-actin depolymerizing factor and brevin [10], confirmed the presence of a normal system capable of inactivating F-actin released into the blood stream in the case of tissue injury [11]. In fact, brevin and the F-actin
The specificity to antigens on microfilaments, such as filamentous actin (F-actin), is an important adjuvant criterion for characterizing the anti-smooth muscle antibody (SMA), a marker of autoimmune hepatitis (AIH) [1, 2]. Indirect immunofluorescence using human fibroblasts is one of the preferred methods for anti-actin antibody identification (AAA). However, in low dilutions (<1:80), heat serum inactivation is a necessary procedure for eliminating a thermolabile serum factor which interferes with AAA reactivity [3]. This interfering factor could be interacting with F-actin, the most probable target antigen of SMA in AIH, through an actin-severing protein [4]. Theoretically, it could also be interacting with the antibody, for Correspondence to: Eduardo Luiz R. Canc¸ado, Instituto de Medicina Tropical, Rua Dr. Ene´as de Carvalho Aguiar—500, 2° andar, Cerqueira Ce´sar, Sa˜o Paulo (SP), CEP 05403/000, Brasil; Fax: 00 55 11 282-7599 or 00 55 11 3063-1559; Email: [email protected] 223 0896–8411/01/070223+06 $35.00/0
© 2001 Academic Press
224
depolymerizing factor proved to be the same plasma gelsolin [12], a high-affinity actin-binding protein which severs F-actin into short filaments, requiring free Ca2+ ions [9]. If the thermolabile factor were a serum gelsolin-like protein, serum dilution in a calcium-chelating solution could have an effect similar to that of heat serum inactivation on AAA identification. The phallotoxins are a family of bicyclic heptapeptides from Amanita phalloides. Their most important feature is the formation of tight complexes with F-actin from muscle and non-muscle cells [13, 14]. Fluorescent phallotoxins constitute an important tool for the visualization of cellular and tissular F-actin through direct immunofluorescence [15]. A double immunofluorescent experiment using rhodamine and fluorescein was performed to co-localize phalloidin and anti-cytoskeletal antibodies in eukariotic cells respectively [16]. In that experiment, the fluorochrome rhodamine, which was bound to the phallotoxin, stained F-actin directly, and the fluorescein-labelled anti-human immunoglobulin indirectly, through the anti-microfilament antibody. A similar double staining experiment could yield more details for explaining what occurs during AAA positive serum incubation with human fibroblasts, in the presence (and absence) of the F-actin destabilizing serum factor. The subjects of this study were to better identify the thermolabile serum factor in AAA detection, by a double-staining immunofluorescent assay, and to check if calcium chelation of serum has the similar effects as those of heat inactivation.
Materials and Methods Patients and healthy controls: 20 serum samples from patients with AIH type 1 and 21 from healthy adult controls were used for the experiments. The diagnoses of AIH were done according to the suggested criteria of the International AIH Group [17]. Fifteen patients had a definite diagnosis and, the remaining five, a probable diagnosis of AIH, when taken into account the response to treatment of the liver disease. Immunofluorescent experiments: acetone-fixed human fibroblasts, from a primary cell culture of a child’s foreskin, were used as the substrate for AAA identification by indirect immunofluorescence. A confluent cellular monolayer was necessary in order to perform the reaction, after 24–72 h of cell culture directly on glass slides, in Dulbecco’s modified Eagle’s medium (D-5796, Sigma, St Louis, MO, USA), along with 10% bovine fetal serum, in a 5% CO2 humidified atmosphere [18]. The serum samples were tested in 1:10 dilution by four different techniques: (A) Serum diluted in 0.14 M saline solution buffered with 0.01 M phosphate (PBS), pH 7.2; (B) heated serum at 56°C, for 30 min, diluted in PBS (C) calcium-chelated serum (diluted in 34 mM EDTA/PBS), pH 7.2; (D) heated and calcium-chelated serum. From initial samples stored at −20°C, two identical ones were generated. One of them was heat-
E. L. R. Canc¸ ado et al.
inactivated for 30 min, at 56°C, and after that, all were coded, and blindly tested. The true identity of them was only reviewed at the end of the experiment. All samples from patients with AIH type 1 were reactive >1:40 to SMA (range 1:80 to 1:20,480, median 1:640), and two from controls (1:40), with the glomerular pattern, when tested by indirect immunofluorescence in non-fixed rodent tissue sections [19], with a different serum sample that had been used previously. F-actin and AAA were revealed by two different fluorochromes according to the following description. Human fibroblasts were incubated with serum samples, diluted as mentioned beforehand, for 30 min, at 37°C, and rinsed twice with PBS, for 10 min. In the second step, they were exposed for 30 min with fluorescein-labelled anti-human IgG (F-1641, Sigma, MO, USA) diluted in PBS in order to reveal the AAA. To demonstrate F-actin microfilaments, rhodaminelabelled phalloidin was prepared as follows: the vial content of a lyophilized tetramethylrhodamine isothiocyanate phalloidin (TRITC, P-1951, Sigma, St Louis, MO, USA) was dissolved in 1 ml methanol, and conserved frozen at −20°C [14]. This methanolic stock solution was diluted 1:40 in PBS, at the moment of the reaction, and applied to the substrate for 30 min at 37°C. After washing twice with PBS, glycerol/PBS solution was used to mount the coverslip to the slide. To check if the phalloidin and AAA bind to adjacent epitopes on the microfilaments, a simultaneous double immunofluorescent reaction was performed. In this case, the methanolic stock solution was diluted 1:40 in PBS together with fluorescein-labelled antihuman IgG. The substrate was examined under an epifluorescent photomicroscope (Axiophot, Zeiss D-7082, Oberkochen, Germany) equipped with a rhodamine filter (BP 546/12, FT 580, LP 590), and a fluorescein filter (BP 450–490, FT 510; LP 520). A definite result regarding AAA and F-actin reactivities was established with the 100× objective. The intensity of the reaction was graded for all positive samples in 1:10 dilution, according to the following criteria: 0, no reaction; 1+, reactivity in ≤1/3 of microscopic field; 2+, reactivity in >1/3 and <2/3 microscopic fields; 3+, reactivity in ≥2/3 of microscopic fields. This schedule was used not only for fluorescein-labelled AAA reactivity (in green) but also for rhodamine-labelled F-actin (in red). All sera reactive to AAA were titred to extinction dilution, except those which were reactive with unmodified serum (technique A), because of the well-known worse, and, even negative results at lower dilutions in this condition [3]. Statistical analysis The techniques were compared two by two, in the presence of only one variable. To study the effects of calcium chelation, the results of the techniques done and those of the two not done with heat serum inactivation (A×C and B×D), were compared. The inverse was performed to study the effects of heat
Interfering factors in anti-actin detection
225
Table 1. Effect of calcium-chelating on non-heated serum for detecting AAA in serum, and F-actin in fibroblast substrate Staining features
Unmodified serum (Technique A) AAA
Intensity in 1:10 0 +1 +2 +3 Total of positive AAA titration Range Median
Calcium-chelated serum (Technique C)
F-actin
AIH (n=20)
Controls (n=21)
AIH (n=20)
8 (40%) 11 (55%) 1 (5%) 0 12 (60%)
21 (100%) 0 0 0 0
AAA
Controls (n=21)
AIH (n=20)
10 (50%) 14 (66.6%) 8 (40%) 6 (28.6%) 2 (10%) 1 (4.8%) 0 0 10 (50%) 7 (33.3%)
Not performed Not performed
F-actin Controls (n=21)
AIH (n=20)
Controls (n=21)
0 1 (5%) 6 (30%) 13 (65%) 20 (100%)
21 (100%) 0 0 0 0 0 0 3 (15%) 4 (19.0%) 0 17 (85%) 17 (81%) 0 20 (100%) 21 (100%) All negative 1:80–1:20,480 in 1:10 1:320
Table 2. Effect of calcium-chelating on heated serum for detecting AAA in serum, and F-actin in fibroblast substrate Staining features
Heated serum (Technique B) AAA
Intensity in 1:10 0 +1 +2 +3 Total of positive AAA titration Range Median
Heated and calcium-chelated serum (Technique D)
F actin
AAA
F actin
AIH (n=20)
Controls (n=21)
AIH (n=20)
Controls (n=21)
AIH (n=20)
Controls (n=21)
AIH (n=20)
Controls (n=21)
0 2 (10%) 10 (50%) 8 (40%) 20 (100%)
20 (95.2%) 1 (4.8%) 0 0 1 (4.8%)
0 0 4 (20%) 16 (80%) 20 (100%)
0 0 4 (19.0%) 17 (81.0%) 21 (100%)
0 2 (10%) 5 (25%) 13 (65%) 20 (100%)
20 (95.2%) 1 0 0 1 (5%)
0 0 2 (10%) 18 (90%) 20 (100%)
0 0 2 (9.5%) 19 (90.5%) 21 (100%)
1:80–1:40,960 1:480
0–1:10 0
1:80–1:20,480 1:320
0–1:10 0
serum inactivation, i.e., the results of the two done and the two not done with calcium chelation (A×B, and C×D). The McNemar test was applied to evaluate the disagreement among the techniques, in studying the general reactivity defined as the total number of AAA positive sera or positive F-actin staining on the substrate in 1:10 dilution, without considering the intensity of the reaction. A P value of ≤0.05 was considered to be statistically significant.
Results The results of AAA detection in patients with AIH and healthy adult controls are displayed in Tables 1 and 2. When the serum samples were not heated or chelated, AAA was not observed in 1:10 serum dilution in 8 of 20 samples (40%) and was detected in 12 (60%) with 1+ intensity in 11 (55%), and 2+ in one (5%). In contrast, AAA (Figure 1) was detected in all of the patients when these procedures were performed separately or together (60 tests) with 1+ in 5 (8.3%), 2+ in
21 (35%); 3+ in 34 (56.7%). Therefore, the technique with unmodified serum (Technique A) was significantly less positive than the other techniques (P=0.004). After calcium chelation and/or heat inactivation, the results of AAA detection were statistically identical. Concerning F-actin detection, the results are also shown in Tables 1 and 2. The microfilaments remained significantly more intensively decorated by rhodamine after incubation of heated and/or calciumchelated serum in all 123 reactions (20 patients with AIH and 21 healthy controls tested with techniques B, C and D) with 2+ in 19 (15.4%) and +3 in 104 (84.6%). The distribution of microfilaments in fibroblasts remained easily identified by rhodamine-labelled phalloidin with heated or calcium-chelated sera. With these procedures, the microfilaments were uniformly identified (Figure 2). With unmodified serum, F-actin filaments were more weakly decorated in 17 of 41 reactions (41.5%, P=0.0000001), with 1+ intensity in 14 (34.1%), and 2+ in three (7.3%). When the fluorescence was not absent, it was patchy, with fields completely negative for microfilaments (Figure 3). The
226
Figure 1. Anti-actin and anti-nuclear antibodies by indirect immunofluorescence in human fibroblasts with fluoresceinlabelled anti-human IgG after incubation with 1:10 calciumchelated serum.
Figure 2. F-actin staining by direct immunofluorescence in human fibroblasts with rhodamine-labelled phalloidin, after incubation with 1:10 calcium-chelated solution. Absence of the nuclear staining.
simultaneous staining of F-actin and AAA provided the same immunofluorescent pattern in fibroblasts (Figure 4). AAA was detected in all of the patients’ sera, as expected, because all of them had been tested positive with high titres for anti-smooth muscle antibodies. It was detected in one control sample with technique B and one with D in 1:10 dilution, and the reactive samples were not the same, but both were weakly reactive to anti-smooth muscle antibodies in ≤1:40 dilutions.
Discussion The experiment using a double immunofluorescent assay for staining separately or simultaneously, the F-actin and its correspondent antibody was an appropriate method to show the serum factor’s interference, acting on microfilaments, in the detection of AAA. Initially, the four following staining possibilities were
E. L. R. Canc¸ ado et al.
Figure 3. Simultaneous double exposure using fluorescein and rhodamine filters. Absence of microfilament staining after serum incubation with unmodified serum diluted 1:10 in PBS. Antinuclear antibody is not influenced by the thermolabile and calcium-dependent factor.
Figure 4. Simultaneous double exposure using fluorescein and rhodamine filters. Presence of yellow microfilaments due to the superimposition of the rhodamine-labelled phalloidin (red) and the fluorescein-labelled anti-human IgG (green), after serum incubation with 1:10 calcium-chelated serum. Antinuclear antibody remained green.
considered concerning the technique developed in this study: (1) AAA stained by fluorescein-labelled anti-human immunoglobulin (in green) and F-actin by rhodamine-labelled phalloidin (in red), in the absence of an inhibitory effect of the serum factor; (2) lack or impairment of green and red reactivity for AAA and F-actin respectively, in the presence of a serum factor, if it hindered the binding of AAA and phalloidin on the microfilaments; (3) presence of red F-actin and absence of green AAA, if the serum factor interfered with AAA (for instance, an interaction with an antiidiotype), hampering its binding to the F-actin. Another explanation could take into account the interference of the serum factor only with the antibodybinding site, leaving the one for phalloidin intact; (4) reactivity for green AAA, and negativity for red F-actin, if the interference of the serum factor was on the phalloidin binding site.
Interfering factors in anti-actin detection
The results confirmed the existence of a thermolabile interfering serum factor in AAA identification in lower dilutions, since 40% of non-heated samples from patients with AIH were AAA negative, when diluted 1:10 PBS solution, while, in contrast, all of them were positive after incubation of heated serum. Furthermore, they also demonstrated that calcium serum chelation had the same effects as those achieved by heat inactivation, because in 1:10 PBSEDTA diluted samples, AAA was detected in all of them from patients with AIH, despite heat inactivation. In accordance with the four original hypotheses, the second proved to be the correct one. F-actin counterstaining with rhodamine produced definite evidence that the thermolabile and calcium-chelating factor interacts with F-actin and not with AAA. The immunofluorescent patterns of microfilaments (F-actin) and of the AAA showed to be the same. This finding means that the AAA binding to their epitope does not hinder the phalloidin binding on the cellular microfilaments. The absence of F-actin staining by rhodamine after unmodified serum exposition indicates that the structure of microfilaments changed, preventing both the binding of phalloidin and IgG AAA. These results were also demonstrated in HEp-2 cells, and the same procedures should be taken into account when using this substrate data not shown. It can be argued that the thermolabile or the calcium-dependent factor hampered the binding of the secondary antibody to the AAA/F-actin complex. Yet, the impairment of red F-actin staining after cell incubation with unmodified healthy control sera demonstrated that this phenomenon occurs in spite of the presence of AAA. Swedish researchers proposed that, besides heat inactivation, the treatment of smeared cells with chelating agents, for improving F-actin polymerization of the substrate, should be also performed to obtain the best results for AAA identification [20, 21]. In a previous study from our group using fibroblasts grown on glass slides, only heat serum inactivation was demonstrated to be efficient. A preincubation of fibroblasts with a calcium-chelating solution did not improve AAA identification, if heat serum inactivation was not concomitantly performed [3]. The discrepancy between the results of the Swedish studies and ours was probably due to a misinterpretation by the authors, because they had also diluted the serum samples in the same calciumchelating solution, in which they had preincubated the substrate. For routine SMA detection, heat serum inactivation or calcium chelation are not necessary, although it is not uncommon to see a better pattern of SMA in 1:80 than in 1:40 dilution, probably because the inhibitory effect of the serum factor is seen in lower dilutions. The reason why the thermolabile factor does not interfere in the same magnitude as with SMA and F-actin determination in tissue sections is not completely understandable. According to Norberg et al. [4], if the incubation time were prolonged for more than 150 min, all SMA patterns
227
would be abolished in a rodent substrate. One explanation that could be put forth is that actin filaments are arranged in tissue in a different form from that observed in cells [22, 23]. Indeed, platelet gelsolin was able to disrupt stress fibres in non-muscle cells permeabilized by either detergent or cold acetone, or in living cells microinjected with gelsolin. On the other hand, when skeletal and cardiac muscle cells were submitted to the same procedures, the organization of myofibrils remained unaltered [23]. Thus, although actins in stress fibres and in myofibrils share a sarcomeric structure, the former is much more labile than the latter. Actin makes up 10–20% of the whole cellular protein, representing the most abundant protein in mammalian cells [24]. Large amounts of actin are released into the blood stream after different tissue injuries, and this could be fatal to the organism owing to the risks of circulating free F-actin which could trigger vascular thrombosis. Therefore, a decline in plasma gelsolin concentration is observed due to its prompt action in scavenging F-actin, and its consequent consumption [11, 25]. Experimental AAA production is not an easy task. Actin has been a well-conserved protein throughout animal evolution, and rabbit immunization with homologous actin was not capable of inducing AAA. This was only possible after denaturation of actin [26]. The presence of this homeostatic F-actin depolymerizing system in the blood stream gives theoretical grounds for explaining the difficulty in producing experimental AAA. The reason why high levels of AAA are detected in AIH remains unknown. The immunosuppressor T cell defect associated with B cell hyperactivity could explain the high levels of immunoglobulins and autoantibodies observed in this disease [27]. Nevertheless, a disruption in the F-actin scavenger system could be an initial step in this process, since the persistence of circulating F-actin may act as an immunogenic antigen. The results of this experiment showed, for the first time, the effects of human sera on the microfilament network, hampering its visualization by a double immunofluorescent staining and, consequently, AAA detection. As calcium chelation of serum was as efficient as heat inactivation, a new immunofluorescent technique for detecting AAA is proposed as an option for its identification.
Acknowledgements We thank Professor Diego Vergani (University College of London) for his helpful advice, and Molecular Probes (Europe BV, Leiden, The Netherlands) for their kind gift of a vial of rhodamine-phalloidin (R-415) during the period of delineating the project of this study. ELR Canc¸ ado was supported by a grant of FAPESP (Fundac¸ a˜ o de Amparo a` Pesquisa do Estado de Sa˜ o Paulo, Brazil, process number 95/02651-9), during the period of delineating the project of this study.
228
References 1. Gabbiani G., Graeme B.R., Lamelin J.P., Vassali P., Majno G., Bouvier C.A., Cruchaud A., Lu¨ scher E.F. 1973. Human smooth muscle autoantibody. Its identification as antiactin antibody and a study of its binding to ‘nonmuscular’ cells. Am. J. Pathol. 72: 473–488 2. Lidman K., Biberfeld G., Fagraeus A., Norberg R., Torstensson R., Utter G. 1976. Anti-actin specificity of human smooth muscle antibodies in chronic active hepatitis. Clin. Exp. Immunol. 24: 266–272 3. Canc¸ ado E.L.R., Vilas-Boas L.S., Abrantes-Lemos C.P., Novo N.F., Porta G., Da Silva L.C., Laudanna A.A. 1996. Heat serum inactivation as a mandatory procedure for antiactin antibody detection in cell culture. Hepatology 23: 1098–1104 4. Norberg R., Thorstensson R., Utter G., Fagraeus A. 1979. F-actin-depolymerizing activity of human serum. Eur. J. Bioch. 100: 575–583 5. Woodruff T., Cook A., Rowley M., Mackay I. 1994. Antibodies to type II collagen in human serum: Storage effects and role of IgM antiidiotype in regulation. J. Rheumatol. 21: 197–202 6. Chaponnier C., Borgia R., Rungger-Brandle E., Weil R., Gabbiani A. 1979. An actin-destabilizing factor is present in human plasma. Experientia 35: 1039–1041 7. Harris D.A., Schwartz J.H. 1981. Characterization of brevin, a serum protein that shortens actin filaments. Proc. Natl. Acad. Sci. USA 78: 6798–6802 8. Yin H.L., Stossel T.P. 1979. Control of cytoplasmic actin gel-sol transformation by gelsolin, a calciumdependent regulatory protein. Nature 281: 583–586 9. Yin H.L., Albrecht J.H., Fattoum A. 1981. Identification of gelsolin, a Ca2+ -dependent regulatory protein of actin gel-sol transformation, and its intracellular distribution in a variety of cells and tissues. J. Cell Biol. 91: 901–906.23 10. Nodes B.R., Shackelford J.E., Lebherz H.G. 1987. Synthesis and secretion of serum gelsolin by smooth muscle tissue. J. Biol. Chem. 262: 5422–5427 11. Ito H., Kambe H., Kimura Y., Nakamura H., Hayashi E., Kishimoto T., Kishimoto S., Yamamoto H. 1992. Depression of plasma gelsolin levels during acute liver injury. Gastroenterology 102: 1686–1692 12. Janmey P.A., Lind S.E. 1987. Capacity of human serum to depolymerize actin filaments. Blood 70: 524–530 13. Huang Z.J., Haugland R.P., Weimin Y., Haugland R.P. 1992. Phallotoxin and actin binding assay by fluorescence enhancement. Anal. Biochem. 200: 199–204 14. Molecular Probes Europe BV. 1995. Phallotoxins: revised protocol. 1–4. Tel. 31-71-236850, Fax 31-71-23319
E. L. R. Canc¸ ado et al.
15. Wulf E., Deboben A., Bautz F.A., Faulstich H., Wieland T. 1979. Fluorescent phallotoxin, a tool for the visualization of cellular actin. Proc. Natl. Acad. Sci. USA 76: 4498–4502 16. Sene´ ca J-L., Fortin S., Roussin A., Joyal F. 1987. Anticytoskeletal autoantibody to microfilament anchorage sites recognizes novel focal contact proteins. J. Clin. Invest. 80: 778–785 17. Alvarez F., Berg P.A., Bianchi F.B., Bianchi L., Burroughs A.K., Cancado E.L., et al. 1999. International autoimmune hepatitis group report: review of criteria for diagnosis of autoimmune hepatitis. J. Hepatol. 31: 929–938 18. Sharon N. 1976. The use of human foreskin cell cultures for isolation of Herpes virus group in the diagnosis laboratory. Am. J. Clin. Pathol. 66: 65–72 19. Bottazzo G,F., Florin-Christensen A., Fairfax A., Swana G., Doniach D., Groeschel-Stewart U. 1976. Classification of smooth muscle autoantibodies detected by immunofluorescence. J. Clin. Pathol. 29: 403–410 20. Fagraeus A., Lidman K., Norberg R. 1975. Indirect immunofluorescence staining of contractile proteins in smeared cells by smooth muscle antibodies. The influence of chelating agents. Clin. Exp. Immunol. 20: 469–477 21. Fagraeus A., Norberg R., Thorstensson R., Utter G., ¨ rvell C. 1983. Auto-antibodies to actin: recent O findings. Ann. N.Y. Acad. Sci. 420: 297–301 22. Sanger J.M., Mittal B., Wegner A., Jockusch B.M., Sanger J.W. 1987. Differential response of stress fibers and myofibrils to gelsolin. Eur. J. Cell Biol. 43: 421–428 23. Gonsior S., Hinssen H. 1995. Exogenous gelsolin binds to sarcomeric thin filaments without severing. Cell Motil Cytosk. 31: 196–206 24. Lee W.M., Galbraith R.M. 1992. The extracellular actin-scavenger system and actin toxicity. N. Engl. J. Med. 326: 1335–1341 25. Lind S.E., Smith D.B., Janmey P.A., Stossel T.P. 1988. Depression of gelsolin levels and detection of gelsolin-actin complexes in plasma of patients with acute lung injury. Am. Rev. Resp. Dis. 138: 429–434 26. Nishioka M., Aikibi T., Shirai M., Terada S., Kagawa H., Watanabe S. 1986. Rabbit autoantibodies to actin induced by immunization with modified homologous actins. Microbiol. Immunol. 30: 1291–1297 27. Galbraith R.M., Eddleston A.L.W.F., Williams R., Webster A.D.B., Pattison J., Doniach D., Kennedy L.A., Batchelor J.R. 1976. Enhanced antibody responses in active chronic hepatitis: relation to HLA-B8 and HLA-B12 and portosystemic shunting. Lancet 1: 930–934
Journal of Autoimmunity (2001) 17, 223–228
Thermolabile and Calcium-dependent Serum Factor Interferes with Polymerized Actin, and Impairs Anti-actin Antibody Detection Eduardo Luiz Rachid Canc¸ado1,2, Clarice Pires Abrantes-Lemos2, Lucy Santos Vilas-Boas3, Neil Ferreira Novo4, Flair Jose´ Carrilho1 and Antonio Atı´lio Laudanna1 1
Department of Gastroenterology, University of Sao Paulo School of Medicine 2 Laboratory of Medical Investigation—Immunopathology of Schistosomiasis, Institute of Tropical Medicine, University of Sao Paulo School of Medicine 3 Laboratory of Medical Investigation—Virology, Institute of Tropical Medicine, University of Sao Paulo School of Medicine 4 Department of Preventive Medicine, Federal University of Sao Paulo School of Medicine, Sao Paulo, Brazil Received 11 May 2001 Accepted 6 August 2001
The detection of anti-actin (AAA) by immunofluorescence is hindered by the presence of a serum factor. To better understand how it interferes with AAA detection, we tested sera from 20 patients with autoimmune hepatitis, and from 21 healthy adults, diluted 1:10 and prepared as follows: (A) diluted with PBS; (B) inactivated at 56°C, and diluted with PBS; (C) diluted with 34 mM EDTA/PBS; (D) heated and diluted with EDTA/PBS. To reveal AAA, a fluorescein-labelled anti-human IgG was used in the process of indirect immunofluorescence. In a parallel assay, the substrate, acetone-fixed human fibroblasts, was preincubated with sera prepared as if it were to identify AAA, but instead, a rhodamine-phalloidin was used to identify F-actin, by direct immunofluorescence. All sera from patients were reactive to AAA when heat-inactivated and/or calcium-chelated, and 60% of them when diluted with unmodified sera (P=0.004). F-actin continued to be present after preincubation with heat-inactivated or calcium-chelated sera from patients and healthy controls, and in 41.5% of reactions with unmodified serum (P=0.0000001). The heat inactivation and the calcium chelation were both efficient procedures for maintaining the microfilament structure intact after serum incubation and, therefore, for identifying AAA. © 2001 Academic Press
Key words: anti-smooth antibody, autoantibodies, autoimmune hepatitis, microfilaments, gelsolin
Introduction
example as an anti-idiotype, preventing the antibody interaction with the antigen [5]. The ability of human serum to disassemble F-actin was first detected by two different groups of investigators. Norberg et al. [4] and Chaponnier et al. [6] described this factor for the purpose of improving AAA identification, and it was called F-actin depolymerizing (destabilizing) factor. Harris et al. [7] identified a protein with a similar activity, but they called it brevin, because this protein shortened actin filaments without depolymerizing them. A similar heat-labile protein capable of controlling the cytoplasmic actin gel-sol transformation was also identified in human macrophages and further in a variety of cells and tissues [8, 9], and it was called gelsolin. The characterization of a secretory form of gelsolin, as a serum protein with features similar to those of the F-actin depolymerizing factor and brevin [10], confirmed the presence of a normal system capable of inactivating F-actin released into the blood stream in the case of tissue injury [11]. In fact, brevin and the F-actin
The specificity to antigens on microfilaments, such as filamentous actin (F-actin), is an important adjuvant criterion for characterizing the anti-smooth muscle antibody (SMA), a marker of autoimmune hepatitis (AIH) [1, 2]. Indirect immunofluorescence using human fibroblasts is one of the preferred methods for anti-actin antibody identification (AAA). However, in low dilutions (<1:80), heat serum inactivation is a necessary procedure for eliminating a thermolabile serum factor which interferes with AAA reactivity [3]. This interfering factor could be interacting with F-actin, the most probable target antigen of SMA in AIH, through an actin-severing protein [4]. Theoretically, it could also be interacting with the antibody, for Correspondence to: Eduardo Luiz R. Canc¸ado, Instituto de Medicina Tropical, Rua Dr. Ene´as de Carvalho Aguiar—500, 2° andar, Cerqueira Ce´sar, Sa˜o Paulo (SP), CEP 05403/000, Brasil; Fax: 00 55 11 282-7599 or 00 55 11 3063-1559; Email: [email protected] 223 0896–8411/01/070223+06 $35.00/0
© 2001 Academic Press
224
depolymerizing factor proved to be the same plasma gelsolin [12], a high-affinity actin-binding protein which severs F-actin into short filaments, requiring free Ca2+ ions [9]. If the thermolabile factor were a serum gelsolin-like protein, serum dilution in a calcium-chelating solution could have an effect similar to that of heat serum inactivation on AAA identification. The phallotoxins are a family of bicyclic heptapeptides from Amanita phalloides. Their most important feature is the formation of tight complexes with F-actin from muscle and non-muscle cells [13, 14]. Fluorescent phallotoxins constitute an important tool for the visualization of cellular and tissular F-actin through direct immunofluorescence [15]. A double immunofluorescent experiment using rhodamine and fluorescein was performed to co-localize phalloidin and anti-cytoskeletal antibodies in eukariotic cells respectively [16]. In that experiment, the fluorochrome rhodamine, which was bound to the phallotoxin, stained F-actin directly, and the fluorescein-labelled anti-human immunoglobulin indirectly, through the anti-microfilament antibody. A similar double staining experiment could yield more details for explaining what occurs during AAA positive serum incubation with human fibroblasts, in the presence (and absence) of the F-actin destabilizing serum factor. The subjects of this study were to better identify the thermolabile serum factor in AAA detection, by a double-staining immunofluorescent assay, and to check if calcium chelation of serum has the similar effects as those of heat inactivation.
Materials and Methods Patients and healthy controls: 20 serum samples from patients with AIH type 1 and 21 from healthy adult controls were used for the experiments. The diagnoses of AIH were done according to the suggested criteria of the International AIH Group [17]. Fifteen patients had a definite diagnosis and, the remaining five, a probable diagnosis of AIH, when taken into account the response to treatment of the liver disease. Immunofluorescent experiments: acetone-fixed human fibroblasts, from a primary cell culture of a child’s foreskin, were used as the substrate for AAA identification by indirect immunofluorescence. A confluent cellular monolayer was necessary in order to perform the reaction, after 24–72 h of cell culture directly on glass slides, in Dulbecco’s modified Eagle’s medium (D-5796, Sigma, St Louis, MO, USA), along with 10% bovine fetal serum, in a 5% CO2 humidified atmosphere [18]. The serum samples were tested in 1:10 dilution by four different techniques: (A) Serum diluted in 0.14 M saline solution buffered with 0.01 M phosphate (PBS), pH 7.2; (B) heated serum at 56°C, for 30 min, diluted in PBS (C) calcium-chelated serum (diluted in 34 mM EDTA/PBS), pH 7.2; (D) heated and calcium-chelated serum. From initial samples stored at −20°C, two identical ones were generated. One of them was heat-
E. L. R. Canc¸ ado et al.
inactivated for 30 min, at 56°C, and after that, all were coded, and blindly tested. The true identity of them was only reviewed at the end of the experiment. All samples from patients with AIH type 1 were reactive >1:40 to SMA (range 1:80 to 1:20,480, median 1:640), and two from controls (1:40), with the glomerular pattern, when tested by indirect immunofluorescence in non-fixed rodent tissue sections [19], with a different serum sample that had been used previously. F-actin and AAA were revealed by two different fluorochromes according to the following description. Human fibroblasts were incubated with serum samples, diluted as mentioned beforehand, for 30 min, at 37°C, and rinsed twice with PBS, for 10 min. In the second step, they were exposed for 30 min with fluorescein-labelled anti-human IgG (F-1641, Sigma, MO, USA) diluted in PBS in order to reveal the AAA. To demonstrate F-actin microfilaments, rhodaminelabelled phalloidin was prepared as follows: the vial content of a lyophilized tetramethylrhodamine isothiocyanate phalloidin (TRITC, P-1951, Sigma, St Louis, MO, USA) was dissolved in 1 ml methanol, and conserved frozen at −20°C [14]. This methanolic stock solution was diluted 1:40 in PBS, at the moment of the reaction, and applied to the substrate for 30 min at 37°C. After washing twice with PBS, glycerol/PBS solution was used to mount the coverslip to the slide. To check if the phalloidin and AAA bind to adjacent epitopes on the microfilaments, a simultaneous double immunofluorescent reaction was performed. In this case, the methanolic stock solution was diluted 1:40 in PBS together with fluorescein-labelled antihuman IgG. The substrate was examined under an epifluorescent photomicroscope (Axiophot, Zeiss D-7082, Oberkochen, Germany) equipped with a rhodamine filter (BP 546/12, FT 580, LP 590), and a fluorescein filter (BP 450–490, FT 510; LP 520). A definite result regarding AAA and F-actin reactivities was established with the 100× objective. The intensity of the reaction was graded for all positive samples in 1:10 dilution, according to the following criteria: 0, no reaction; 1+, reactivity in ≤1/3 of microscopic field; 2+, reactivity in >1/3 and <2/3 microscopic fields; 3+, reactivity in ≥2/3 of microscopic fields. This schedule was used not only for fluorescein-labelled AAA reactivity (in green) but also for rhodamine-labelled F-actin (in red). All sera reactive to AAA were titred to extinction dilution, except those which were reactive with unmodified serum (technique A), because of the well-known worse, and, even negative results at lower dilutions in this condition [3]. Statistical analysis The techniques were compared two by two, in the presence of only one variable. To study the effects of calcium chelation, the results of the techniques done and those of the two not done with heat serum inactivation (A×C and B×D), were compared. The inverse was performed to study the effects of heat
Interfering factors in anti-actin detection
225
Table 1. Effect of calcium-chelating on non-heated serum for detecting AAA in serum, and F-actin in fibroblast substrate Staining features
Unmodified serum (Technique A) AAA
Intensity in 1:10 0 +1 +2 +3 Total of positive AAA titration Range Median
Calcium-chelated serum (Technique C)
F-actin
AIH (n=20)
Controls (n=21)
AIH (n=20)
8 (40%) 11 (55%) 1 (5%) 0 12 (60%)
21 (100%) 0 0 0 0
AAA
Controls (n=21)
AIH (n=20)
10 (50%) 14 (66.6%) 8 (40%) 6 (28.6%) 2 (10%) 1 (4.8%) 0 0 10 (50%) 7 (33.3%)
Not performed Not performed
F-actin Controls (n=21)
AIH (n=20)
Controls (n=21)
0 1 (5%) 6 (30%) 13 (65%) 20 (100%)
21 (100%) 0 0 0 0 0 0 3 (15%) 4 (19.0%) 0 17 (85%) 17 (81%) 0 20 (100%) 21 (100%) All negative 1:80–1:20,480 in 1:10 1:320
Table 2. Effect of calcium-chelating on heated serum for detecting AAA in serum, and F-actin in fibroblast substrate Staining features
Heated serum (Technique B) AAA
Intensity in 1:10 0 +1 +2 +3 Total of positive AAA titration Range Median
Heated and calcium-chelated serum (Technique D)
F actin
AAA
F actin
AIH (n=20)
Controls (n=21)
AIH (n=20)
Controls (n=21)
AIH (n=20)
Controls (n=21)
AIH (n=20)
Controls (n=21)
0 2 (10%) 10 (50%) 8 (40%) 20 (100%)
20 (95.2%) 1 (4.8%) 0 0 1 (4.8%)
0 0 4 (20%) 16 (80%) 20 (100%)
0 0 4 (19.0%) 17 (81.0%) 21 (100%)
0 2 (10%) 5 (25%) 13 (65%) 20 (100%)
20 (95.2%) 1 0 0 1 (5%)
0 0 2 (10%) 18 (90%) 20 (100%)
0 0 2 (9.5%) 19 (90.5%) 21 (100%)
1:80–1:40,960 1:480
0–1:10 0
1:80–1:20,480 1:320
0–1:10 0
serum inactivation, i.e., the results of the two done and the two not done with calcium chelation (A×B, and C×D). The McNemar test was applied to evaluate the disagreement among the techniques, in studying the general reactivity defined as the total number of AAA positive sera or positive F-actin staining on the substrate in 1:10 dilution, without considering the intensity of the reaction. A P value of ≤0.05 was considered to be statistically significant.
Results The results of AAA detection in patients with AIH and healthy adult controls are displayed in Tables 1 and 2. When the serum samples were not heated or chelated, AAA was not observed in 1:10 serum dilution in 8 of 20 samples (40%) and was detected in 12 (60%) with 1+ intensity in 11 (55%), and 2+ in one (5%). In contrast, AAA (Figure 1) was detected in all of the patients when these procedures were performed separately or together (60 tests) with 1+ in 5 (8.3%), 2+ in
21 (35%); 3+ in 34 (56.7%). Therefore, the technique with unmodified serum (Technique A) was significantly less positive than the other techniques (P=0.004). After calcium chelation and/or heat inactivation, the results of AAA detection were statistically identical. Concerning F-actin detection, the results are also shown in Tables 1 and 2. The microfilaments remained significantly more intensively decorated by rhodamine after incubation of heated and/or calciumchelated serum in all 123 reactions (20 patients with AIH and 21 healthy controls tested with techniques B, C and D) with 2+ in 19 (15.4%) and +3 in 104 (84.6%). The distribution of microfilaments in fibroblasts remained easily identified by rhodamine-labelled phalloidin with heated or calcium-chelated sera. With these procedures, the microfilaments were uniformly identified (Figure 2). With unmodified serum, F-actin filaments were more weakly decorated in 17 of 41 reactions (41.5%, P=0.0000001), with 1+ intensity in 14 (34.1%), and 2+ in three (7.3%). When the fluorescence was not absent, it was patchy, with fields completely negative for microfilaments (Figure 3). The
226
Figure 1. Anti-actin and anti-nuclear antibodies by indirect immunofluorescence in human fibroblasts with fluoresceinlabelled anti-human IgG after incubation with 1:10 calciumchelated serum.
Figure 2. F-actin staining by direct immunofluorescence in human fibroblasts with rhodamine-labelled phalloidin, after incubation with 1:10 calcium-chelated solution. Absence of the nuclear staining.
simultaneous staining of F-actin and AAA provided the same immunofluorescent pattern in fibroblasts (Figure 4). AAA was detected in all of the patients’ sera, as expected, because all of them had been tested positive with high titres for anti-smooth muscle antibodies. It was detected in one control sample with technique B and one with D in 1:10 dilution, and the reactive samples were not the same, but both were weakly reactive to anti-smooth muscle antibodies in ≤1:40 dilutions.
Discussion The experiment using a double immunofluorescent assay for staining separately or simultaneously, the F-actin and its correspondent antibody was an appropriate method to show the serum factor’s interference, acting on microfilaments, in the detection of AAA. Initially, the four following staining possibilities were
E. L. R. Canc¸ ado et al.
Figure 3. Simultaneous double exposure using fluorescein and rhodamine filters. Absence of microfilament staining after serum incubation with unmodified serum diluted 1:10 in PBS. Antinuclear antibody is not influenced by the thermolabile and calcium-dependent factor.
Figure 4. Simultaneous double exposure using fluorescein and rhodamine filters. Presence of yellow microfilaments due to the superimposition of the rhodamine-labelled phalloidin (red) and the fluorescein-labelled anti-human IgG (green), after serum incubation with 1:10 calcium-chelated serum. Antinuclear antibody remained green.
considered concerning the technique developed in this study: (1) AAA stained by fluorescein-labelled anti-human immunoglobulin (in green) and F-actin by rhodamine-labelled phalloidin (in red), in the absence of an inhibitory effect of the serum factor; (2) lack or impairment of green and red reactivity for AAA and F-actin respectively, in the presence of a serum factor, if it hindered the binding of AAA and phalloidin on the microfilaments; (3) presence of red F-actin and absence of green AAA, if the serum factor interfered with AAA (for instance, an interaction with an antiidiotype), hampering its binding to the F-actin. Another explanation could take into account the interference of the serum factor only with the antibodybinding site, leaving the one for phalloidin intact; (4) reactivity for green AAA, and negativity for red F-actin, if the interference of the serum factor was on the phalloidin binding site.
Interfering factors in anti-actin detection
The results confirmed the existence of a thermolabile interfering serum factor in AAA identification in lower dilutions, since 40% of non-heated samples from patients with AIH were AAA negative, when diluted 1:10 PBS solution, while, in contrast, all of them were positive after incubation of heated serum. Furthermore, they also demonstrated that calcium serum chelation had the same effects as those achieved by heat inactivation, because in 1:10 PBSEDTA diluted samples, AAA was detected in all of them from patients with AIH, despite heat inactivation. In accordance with the four original hypotheses, the second proved to be the correct one. F-actin counterstaining with rhodamine produced definite evidence that the thermolabile and calcium-chelating factor interacts with F-actin and not with AAA. The immunofluorescent patterns of microfilaments (F-actin) and of the AAA showed to be the same. This finding means that the AAA binding to their epitope does not hinder the phalloidin binding on the cellular microfilaments. The absence of F-actin staining by rhodamine after unmodified serum exposition indicates that the structure of microfilaments changed, preventing both the binding of phalloidin and IgG AAA. These results were also demonstrated in HEp-2 cells, and the same procedures should be taken into account when using this substrate data not shown. It can be argued that the thermolabile or the calcium-dependent factor hampered the binding of the secondary antibody to the AAA/F-actin complex. Yet, the impairment of red F-actin staining after cell incubation with unmodified healthy control sera demonstrated that this phenomenon occurs in spite of the presence of AAA. Swedish researchers proposed that, besides heat inactivation, the treatment of smeared cells with chelating agents, for improving F-actin polymerization of the substrate, should be also performed to obtain the best results for AAA identification [20, 21]. In a previous study from our group using fibroblasts grown on glass slides, only heat serum inactivation was demonstrated to be efficient. A preincubation of fibroblasts with a calcium-chelating solution did not improve AAA identification, if heat serum inactivation was not concomitantly performed [3]. The discrepancy between the results of the Swedish studies and ours was probably due to a misinterpretation by the authors, because they had also diluted the serum samples in the same calciumchelating solution, in which they had preincubated the substrate. For routine SMA detection, heat serum inactivation or calcium chelation are not necessary, although it is not uncommon to see a better pattern of SMA in 1:80 than in 1:40 dilution, probably because the inhibitory effect of the serum factor is seen in lower dilutions. The reason why the thermolabile factor does not interfere in the same magnitude as with SMA and F-actin determination in tissue sections is not completely understandable. According to Norberg et al. [4], if the incubation time were prolonged for more than 150 min, all SMA patterns
227
would be abolished in a rodent substrate. One explanation that could be put forth is that actin filaments are arranged in tissue in a different form from that observed in cells [22, 23]. Indeed, platelet gelsolin was able to disrupt stress fibres in non-muscle cells permeabilized by either detergent or cold acetone, or in living cells microinjected with gelsolin. On the other hand, when skeletal and cardiac muscle cells were submitted to the same procedures, the organization of myofibrils remained unaltered [23]. Thus, although actins in stress fibres and in myofibrils share a sarcomeric structure, the former is much more labile than the latter. Actin makes up 10–20% of the whole cellular protein, representing the most abundant protein in mammalian cells [24]. Large amounts of actin are released into the blood stream after different tissue injuries, and this could be fatal to the organism owing to the risks of circulating free F-actin which could trigger vascular thrombosis. Therefore, a decline in plasma gelsolin concentration is observed due to its prompt action in scavenging F-actin, and its consequent consumption [11, 25]. Experimental AAA production is not an easy task. Actin has been a well-conserved protein throughout animal evolution, and rabbit immunization with homologous actin was not capable of inducing AAA. This was only possible after denaturation of actin [26]. The presence of this homeostatic F-actin depolymerizing system in the blood stream gives theoretical grounds for explaining the difficulty in producing experimental AAA. The reason why high levels of AAA are detected in AIH remains unknown. The immunosuppressor T cell defect associated with B cell hyperactivity could explain the high levels of immunoglobulins and autoantibodies observed in this disease [27]. Nevertheless, a disruption in the F-actin scavenger system could be an initial step in this process, since the persistence of circulating F-actin may act as an immunogenic antigen. The results of this experiment showed, for the first time, the effects of human sera on the microfilament network, hampering its visualization by a double immunofluorescent staining and, consequently, AAA detection. As calcium chelation of serum was as efficient as heat inactivation, a new immunofluorescent technique for detecting AAA is proposed as an option for its identification.
Acknowledgements We thank Professor Diego Vergani (University College of London) for his helpful advice, and Molecular Probes (Europe BV, Leiden, The Netherlands) for their kind gift of a vial of rhodamine-phalloidin (R-415) during the period of delineating the project of this study. ELR Canc¸ ado was supported by a grant of FAPESP (Fundac¸ a˜ o de Amparo a` Pesquisa do Estado de Sa˜ o Paulo, Brazil, process number 95/02651-9), during the period of delineating the project of this study.
228
References 1. Gabbiani G., Graeme B.R., Lamelin J.P., Vassali P., Majno G., Bouvier C.A., Cruchaud A., Lu¨ scher E.F. 1973. Human smooth muscle autoantibody. Its identification as antiactin antibody and a study of its binding to ‘nonmuscular’ cells. Am. J. Pathol. 72: 473–488 2. Lidman K., Biberfeld G., Fagraeus A., Norberg R., Torstensson R., Utter G. 1976. Anti-actin specificity of human smooth muscle antibodies in chronic active hepatitis. Clin. Exp. Immunol. 24: 266–272 3. Canc¸ ado E.L.R., Vilas-Boas L.S., Abrantes-Lemos C.P., Novo N.F., Porta G., Da Silva L.C., Laudanna A.A. 1996. Heat serum inactivation as a mandatory procedure for antiactin antibody detection in cell culture. Hepatology 23: 1098–1104 4. Norberg R., Thorstensson R., Utter G., Fagraeus A. 1979. F-actin-depolymerizing activity of human serum. Eur. J. Bioch. 100: 575–583 5. Woodruff T., Cook A., Rowley M., Mackay I. 1994. Antibodies to type II collagen in human serum: Storage effects and role of IgM antiidiotype in regulation. J. Rheumatol. 21: 197–202 6. Chaponnier C., Borgia R., Rungger-Brandle E., Weil R., Gabbiani A. 1979. An actin-destabilizing factor is present in human plasma. Experientia 35: 1039–1041 7. Harris D.A., Schwartz J.H. 1981. Characterization of brevin, a serum protein that shortens actin filaments. Proc. Natl. Acad. Sci. USA 78: 6798–6802 8. Yin H.L., Stossel T.P. 1979. Control of cytoplasmic actin gel-sol transformation by gelsolin, a calciumdependent regulatory protein. Nature 281: 583–586 9. Yin H.L., Albrecht J.H., Fattoum A. 1981. Identification of gelsolin, a Ca2+ -dependent regulatory protein of actin gel-sol transformation, and its intracellular distribution in a variety of cells and tissues. J. Cell Biol. 91: 901–906.23 10. Nodes B.R., Shackelford J.E., Lebherz H.G. 1987. Synthesis and secretion of serum gelsolin by smooth muscle tissue. J. Biol. Chem. 262: 5422–5427 11. Ito H., Kambe H., Kimura Y., Nakamura H., Hayashi E., Kishimoto T., Kishimoto S., Yamamoto H. 1992. Depression of plasma gelsolin levels during acute liver injury. Gastroenterology 102: 1686–1692 12. Janmey P.A., Lind S.E. 1987. Capacity of human serum to depolymerize actin filaments. Blood 70: 524–530 13. Huang Z.J., Haugland R.P., Weimin Y., Haugland R.P. 1992. Phallotoxin and actin binding assay by fluorescence enhancement. Anal. Biochem. 200: 199–204 14. Molecular Probes Europe BV. 1995. Phallotoxins: revised protocol. 1–4. Tel. 31-71-236850, Fax 31-71-23319
E. L. R. Canc¸ ado et al.
15. Wulf E., Deboben A., Bautz F.A., Faulstich H., Wieland T. 1979. Fluorescent phallotoxin, a tool for the visualization of cellular actin. Proc. Natl. Acad. Sci. USA 76: 4498–4502 16. Sene´ ca J-L., Fortin S., Roussin A., Joyal F. 1987. Anticytoskeletal autoantibody to microfilament anchorage sites recognizes novel focal contact proteins. J. Clin. Invest. 80: 778–785 17. Alvarez F., Berg P.A., Bianchi F.B., Bianchi L., Burroughs A.K., Cancado E.L., et al. 1999. International autoimmune hepatitis group report: review of criteria for diagnosis of autoimmune hepatitis. J. Hepatol. 31: 929–938 18. Sharon N. 1976. The use of human foreskin cell cultures for isolation of Herpes virus group in the diagnosis laboratory. Am. J. Clin. Pathol. 66: 65–72 19. Bottazzo G,F., Florin-Christensen A., Fairfax A., Swana G., Doniach D., Groeschel-Stewart U. 1976. Classification of smooth muscle autoantibodies detected by immunofluorescence. J. Clin. Pathol. 29: 403–410 20. Fagraeus A., Lidman K., Norberg R. 1975. Indirect immunofluorescence staining of contractile proteins in smeared cells by smooth muscle antibodies. The influence of chelating agents. Clin. Exp. Immunol. 20: 469–477 21. Fagraeus A., Norberg R., Thorstensson R., Utter G., ¨ rvell C. 1983. Auto-antibodies to actin: recent O findings. Ann. N.Y. Acad. Sci. 420: 297–301 22. Sanger J.M., Mittal B., Wegner A., Jockusch B.M., Sanger J.W. 1987. Differential response of stress fibers and myofibrils to gelsolin. Eur. J. Cell Biol. 43: 421–428 23. Gonsior S., Hinssen H. 1995. Exogenous gelsolin binds to sarcomeric thin filaments without severing. Cell Motil Cytosk. 31: 196–206 24. Lee W.M., Galbraith R.M. 1992. The extracellular actin-scavenger system and actin toxicity. N. Engl. J. Med. 326: 1335–1341 25. Lind S.E., Smith D.B., Janmey P.A., Stossel T.P. 1988. Depression of gelsolin levels and detection of gelsolin-actin complexes in plasma of patients with acute lung injury. Am. Rev. Resp. Dis. 138: 429–434 26. Nishioka M., Aikibi T., Shirai M., Terada S., Kagawa H., Watanabe S. 1986. Rabbit autoantibodies to actin induced by immunization with modified homologous actins. Microbiol. Immunol. 30: 1291–1297 27. Galbraith R.M., Eddleston A.L.W.F., Williams R., Webster A.D.B., Pattison J., Doniach D., Kennedy L.A., Batchelor J.R. 1976. Enhanced antibody responses in active chronic hepatitis: relation to HLA-B8 and HLA-B12 and portosystemic shunting. Lancet 1: 930–934