Visualization of neutrophil extracellular traps in TEM Wolf Dietrich Krautgartner a , Ljubomir Vitkov b,c, * a Department of Light & Electron Microscopy, Organismic Biology, University of Salzburg, Hellbrunnerstraße 34, A-5020 Salzburg, Austria b Department of Operative Dentistry & Periodontology, Medical University of Innsbruck, Anichstraße 35, A-6020 Innsbruck, Austria c Mayburgerplatz 7, A-5204 Straßwalchen, Austria Received 25 February 2007; received in revised form 18 March 2007; accepted 19 March 2007 Abstract Neutrophil extracellular traps (NETs) have recently been described as an important innate defence mechanism in inflammation. However, routine electron microscopic staining techniques faintly stain NETs and are therefore insufficient for enabling a distinction between these and the host cell debris as well as proteins regularly present at the site of inflammation. In order to test suitable electron microscopic staining techniques, NETs induced ex vivo via phorbol myristate were absorbed on formvar. Four types of drop-on-grid positive staining were used: osmium tetroxide (Os), osmium tetroxide–uranyl acetate–lead citrate (Os–U–Pb), ruthenium red-osmium tetroxide (RR-OsO 4 ), and cuprolinic blue enhanced by sodium tungstate (CB-WO 4 ). We observed no staining of NETs using Os, faint staining with Os–U–Pb, better but still weak staining with CB-WO 4 and outstanding staining with RR-OsO 4 . Furthermore, RR-OsO 4 staining also enables the observation of bacterial fimbriae-mediated adhesion, which is possibly responsible for the ability of NETs to bind bacteria. Thus, the offered RR-OsO 4 staining technique may facilitate the study of the NETs-bacterial interactions. # 2007 Elsevier Ltd. All rights reserved. Keywords: Cationic dyes; Cuprolinic blue; Ruthenium red; Deoxyribonuclease; Globular domains 1. Introduction Recently, a ‘‘new’’ mechanism for bacterial clearance by neutrophil extracellular traps (NETs) has been reported (Brinkmann et al., 2004). NETs are extracellular net-like fibres generated by activated neutrophils and are able to disarm and kill bacteria extracellularly. They consist of a DNA backbone with embedded antimicrobial peptides and enzymes, e.g. histones and neutrophil elastase (Brinkmann et al., 2004; Fuchs et al., 2007). NETs bind Gram-positive as well as Gram- negative bacteria and are abundant in vivo in acute inflamma- tion. They appear to be a form of innate response that binds microorganisms, prevents them from spreading and ensures a high local concentration of antimicrobial agents to degrade virulence factors and kill bacteria (Brinkmann et al., 2004). Although NETs are of pronounced interest to the investigation of the host response to bacterial challenge, their visualization in inflamed tissues by routine transmission electron microscopy techniques is not reliable, as there is insufficient staining when applying routine techniques. Thus, the basic elements of NETs, the fibrils with a diameter of 15– 17 nm have not been visualized by transmission electron microscopy (TEM) until now, but only by high resolution scanning electron microscopy (Brinkmann et al., 2004; Fuchs et al., 2007). Consequently, the question arises, whether a selective staining for chromatin could enable the visualization and differentiation of NETs by TEM in the presence of extracellular host proteins regularly present at the site of inflammation. The aim of the present work was to offer a reliable method for visualization and differentiation of NETs by TEM. 2. Materials and methods 2.1. Neutrophil harvesting Mouse neutrophils were kindly supplied by Dr. J. Thalhammer and Dr. R. Weiss (Department of Molecular Biology, University of Salzburg, Austria). Peripheral blood mononuclear cells (PBMCs) were prepared from whole blood by erythrocyte lysis under gently hypotonic conditions. Two millilitre of heparin-treated whole mouse blood were mixed www.elsevier.com/locate/micron Micron 39 (2008) 367–372 * Corresponding author at: Mayburgerplatz 7, A-5204 Straßwalchen, Austria. Tel.: +43 676 4041225; fax: +43 6215 20088. E-mail address: [email protected](L. Vitkov). 0968-4328/$ – see front matter # 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.micron.2007.03.007
6
Embed
Visualization of Neutrophil Extra Cellular Traps in TEM
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
www.elsevier.com/locate/micron
Micron 39 (2008) 367–372
Visualization of neutrophil extracellular traps in TEM
Wolf Dietrich Krautgartner a, Ljubomir Vitkov b,c,*a Department of Light & Electron Microscopy, Organismic Biology, University of Salzburg, Hellbrunnerstraße 34, A-5020 Salzburg, Austria
b Department of Operative Dentistry & Periodontology, Medical University of Innsbruck, Anichstraße 35, A-6020 Innsbruck, Austriac Mayburgerplatz 7, A-5204 Straßwalchen, Austria
Received 25 February 2007; received in revised form 18 March 2007; accepted 19 March 2007
Abstract
Neutrophil extracellular traps (NETs) have recently been described as an important innate defence mechanism in inflammation. However,
routine electron microscopic staining techniques faintly stain NETs and are therefore insufficient for enabling a distinction between these and the
host cell debris as well as proteins regularly present at the site of inflammation. In order to test suitable electron microscopic staining techniques,
NETs induced ex vivo via phorbol myristate were absorbed on formvar. Four types of drop-on-grid positive staining were used: osmium tetroxide
(Os), osmium tetroxide–uranyl acetate–lead citrate (Os–U–Pb), ruthenium red-osmium tetroxide (RR-OsO4), and cuprolinic blue enhanced by
sodium tungstate (CB-WO4). We observed no staining of NETs using Os, faint staining with Os–U–Pb, better but still weak staining with CB-WO4
and outstanding staining with RR-OsO4. Furthermore, RR-OsO4 staining also enables the observation of bacterial fimbriae-mediated adhesion,
which is possibly responsible for the ability of NETs to bind bacteria. Thus, the offered RR-OsO4 staining technique may facilitate the study of the
The grids were washed with 0.1 M sodium cacodylate
buffered at pH 5.6. Subsequently, the grids were fixed with
1.2% glutaraldehyde and 0.05% ruthenium red (buffered at pH
5.6 with 0.1 M sodium cacodylate) for 15 min at room
temperature. Postfixation was performed with 1.2% osmium
tetroxide (buffered at pH 5.6 with 0.1 M sodium cacodylate)
and 0.05% ruthenium red for one min at room temperature.
After staining, all of the grids were simply washed and air-
dried. The RR-OsO4 drop-on-grid specimens and particularly
the drop-on-grid cuprolinic blue specimens were unstable and
faded after a few days.
2.7. TEM processing
All samples were examined with a transmission electron
microscope LEO EM 910 (LEO Elektronenmikroskopie Ltd.,
Oberkochen, Germany) operating at 80 kV.
3. Results
Os-stained samples revealed no structures. Os–U–Pb
staining revealed well formed, but faintly blackened networks
consisting of a multitude of interwoven threads. They branched
into thinner threads, but no fibrils with a diameter 15–17 nm
could be distinguished (Fig. 1). CB staining revealed a similar
appearance of NETs, however, somewhat better staining was
achieved (Figs. 2–4). RR-OsO4 staining showed at low
magnification a similar appearance of NETs (Fig. 5). However,
the higher magnification revealed that the thinner threads
branched into fine fibres (Fig. 6). A sharp demarcation of
deeply blackened fibres with diameters of circa 15–17 nm and
less frequently of circa 6 nm was evident (Fig. 7), however, no
globular domains could be observed.
Two predominant patterns of thread alignment were
observed: a radiating one and a meshwork-like one. The
radiating thread alignment pattern was characterized by the
branching of a few thick threads into thinner ones, which on
their part branched into fibres (Figs. 5–7). The meshwork-like
pattern covered areas extending to more than 1000 mm2 (Fig. 8)
and was characterized by a relative uniformity of appearance of
dense interwoven threads and fibres with a diameter of circa
15–17 nm (Fig. 9). In contrast, fibres with a diameter of 6 nm
were not observed within this pattern.
The treatment by DNase prior to the fixation completely
disaggregated the NETs. In the RR-OsO4 stained samples,
however, a multitude of tiny particles, very probably NET
fragments, were observed (Fig. 10). The electron density of the
particles is the same as that of the NETs (Fig. 11).
4. Discussion
Morphologically, NETs have been characterized as DNase-
labile net-like formations of fibres with a diameter of 15–
17 nm, with so-called globular domains on them with a
diameter of approximately 25 nm (Brinkmann et al., 2004).
Figs. 1–4. (1) Os–U–Pb staining. NETs. Inset: The higher magnification shows lack of fine structural details. (2) CB staining. An overview of NETs. Some similarity
with light microscopy micrographs (Brinkmann et al., 2004) is evident. (3) CB staining (a detail of Fig. 2). The deeply blackened spots (arrows) are probably
phthalocyanine precipitates, as previously reported (Krautgartner et al., 2003, 2005). (4) CB staining (a detail of Fig. 3). The higher magnification shows lack of fine
structural details.
W.D. Krautgartner, L. Vitkov / Micron 39 (2008) 367–372 369
Besides the bactericidal effects caused by neutrophil elastase
and histones, the ability to trap bacteria is another key
characteristic of NETs. The bacterial binding by NETs
probably occurs by fimbriae-mediated adhesion. Indeed,
bacterial fimbriae-mediated adhesion to histones (Zhu et al.,
2005) and to DNA (van Schaik et al., 2005) has been
demonstrated. Both DNA and bacterial fimbriae are poorly
depicted by routine TEM staining techniques. As DNA and
microbial fimbriae are polyanions, cationic stains have been
used for their visualization by TEM. Thus, uranyl and lead ions
(Pearse, 1985; Trendelenburg and Puvion-Dutilleul, 1987), in
particular the combination of uranyl acetate and lead citrate, are
still constantly employed for this purpose. However, these
staining techniques have low specificity, feebly contrast the
DNA and therefore disable a good resolution. Two cationic
dyes, the phthalocyanine cuprolinic blue and ruthenium red
have also been used for chromatin staining.
CB has been used for visualizing DNA in light microscopy
(Tas et al., 1983; Mendelson et al., 1983), however, enhancing
the phthalocyanine staining by sodium tungstate enables the
attainment of a sufficient contrast in TEM (Scott et al., 1981).
Despite the good staining specificity, the DNA contrast attained
by this technique was insufficient to achieve high resolution.
Nevertheless, the possibility to visualize the microbial adhesion
via phthalocyanines at TEM level (Fassel et al., 1992;
Krautgartner et al., 2003, 2005) makes this method, albeit
restricted by its limited resolution, suitable to study the NETs-
bacterial interactions.
Ruthenium red shows a strong affinity for chromatin
(Stockert and Pelling, 1992; Engelhardt, 2000). However, as
chromatin and also NETs consist of two main components,
DNA and proteins, it remains unclear which component(s)
really stain by RR-OsO4. Thus, it has been hypothesized by
Engelhardt (Engelhardt, 2000) that the chromatin staining by
RR-OsO4 could be caused by a reaction with the chromatin
glycoproteins (Reeves et al., 1981; Turner et al., 1990).
Although RR alone produces some degree of electron opacity
(Stockert and Pelling, 1992), RR-OsO4 creates a much more
intense contrasting. Indeed, the presently employed RR-OsO4
staining technique enabled the outstanding staining of the
Figs. 5–7. (5) RR-OsO4 staining. An overview of NETs. Some similarity with light microscopic micrographs (Brinkmann et al., 2004) is evident. (6) RR-OsO4
staining (a detail of Fig. 5). The alignment of the threads is clearly distinguishable. (7) RR-OsO4 staining (a detail of Fig. 6). Fibres with diameter of 15–17 nm form a
meshwork. Arrows: some fibres with a diameter of 6 nm are also evident.
W.D. Krautgartner, L. Vitkov / Micron 39 (2008) 367–372370
NETs, despite the superimposition by cellular debris and
proteins from the nutritive solution. The RR-positive NETs
mainly consisted of fibres with a diameter of circa 15–17 nm,
which were interwoven in net-like formations. RR-OsO4
staining revealed also fibres with a diameter of circa 6 nm,
which has not been observed with scanning electron micro-
scopy. RR-OsO4 stains microbial fimbriae (Fassel et al., 1992;
Fassel and Edmiston, 1999; Vitkov et al., 2001, 2002, 2005a,b),
Figs. 8–11. (8) RR-OsO4 staining. Huge areas are covered by meshwork of a relatively uniform density. (9) RR-OsO4 staining. The higher magnification reveals no
basic differences to the loose networks. Inset: Interwoven fibres with similar dimension as shown in Fig. 7. (10) RR-OsO4 staining. The sample is treated by
deoxyribonuclease prior to the staining. No networks, but multitudes of tiny particles were observed. (11) RR-OsO4 staining (a detail from Fig. 10). Many deeply
contrasted particles, probably NET fragments, are evident. The blackening of the particles is the same as that of the NETs.
W.D. Krautgartner, L. Vitkov / Micron 39 (2008) 367–372 371
proteoglycans (Chan and Wong, 1992; Rodgers et al., 1995) and
chromatin. Consequently, the DNase digestion of the NET
fibres verified their DNA character. By contrast, no globular
domains were visualized via RR-OsO4; consequently, they lack
chromatin. Two modes of NET spreading were observed:
radiating ones and web-like ones. In the radiating NETs, the
threads were branched into fibres with a diameter of 15–17 nm.
The threads have been suggested to consist of aggregated fibres
(Brinkmann et al., 2004). In the web-like NETs, the spread was
more homogenous and fewer threads with smaller diameters
were observed. In our opinion, the web-like appearance was a
consequence of the superimposition of a multitude of better
spread NETs. As the adhesion of NETs to the formvar causes a
transformation of the three-dimensional NETs into two-
dimensional images, deductions concerning the three-dimen-
sional relationship between the NET elements by virtue of the
presented results are restricted. The ability of RR staining to
visualize the microbial adhesion makes this method particu-
larly appropriate for studying the NETs-bacterial interactions.
The deeply blackened DNase-sensitive smooth stretches, with
their characteristic diameter, are the assertive criterion to
identify NETs.
5. Conclusions
The employed RR-OsO4 staining technique enables the
visualization of both NETs and fimbriae-mediated bacterial
adhesion (Vitkov et al., 2001, 2002, 2005a,b). It thus facilitates
the examination of NETs-bacterial interactions at TEM level.
The combination of the following two characteristics guarantees
the high specificity of this staining procedure: the blackening by
RR-OsO4 and the disintegration of the RR-OsO4-positive net-
like structures by DNase. Additionally, the combination of the
specific fibre thickness (15–17 nm) and of the deep blackening of
the fibres is a very strong indication of NETs.
Acknowledgements
The authors thank Professor Josef Thalhammer and Dr.
Robert Weiss for supplying the mouse neutrophils, Dr. Karin
W.D. Krautgartner, L. Vitkov / Micron 39 (2008) 367–372372
Oberascher, Mrs. Adda Maenhardt and Mrs. Michaela
Klappacher for the technical assistance as well as Mr. Andreas