technical notes
http://www.kidney-international.org
© 2015 International Society of Nephrology
OPEN
Super-resolution stimulated emission depletion
imaging of slit diaphragm proteins in optically
cleared kidney tissue
David Unnersjö-Jess1, Lena Scott2, Hans Blom1 and Hjalmar Brismar1,2
1
Science for Life Laboratory, Department of Applied Physics, Royal Institute of Technology, Solna, Sweden and 2Science for Life
Laboratory, Department of Women´s and Children´s Health, Karolinska Institutet, Solna, Sweden
The glomerular filtration barrier, consisting of podocyte foot
processes with bridging slit diaphragm, glomerular basement
membrane, and endothelium, is a key component for renal
function. Previously, the subtlest elements of the filtration
barrier have only been visualized using electron microscopy.
However, electron microscopy is mostly restricted to
ultrathin two-dimensional samples, and the possibility to
simultaneously visualize multiple different proteins is limited.
Therefore, we sought to implement a super-resolution
immunofluorescence microscopy protocol for the study of
the filtration barrier in the kidney. Recently, several optical
clearing methods have been developed making it possible to
image through large volumes of tissue and even whole
organs using light microscopy. Here we found that
hydrogel-based optical clearing is a beneficial tool to study
intact renal tissue at the nanometer scale. When imaging
samples using super-resolution STED microscopy, the
staining quality was critical in order to assess correct
nanoscale information. The signal-to-noise ratio and
immunosignal homogeneity were both improved in optically
cleared tissue. Thus, STED of slit diaphragms in fluorescently
labelled optically cleared intact kidney samples is a new tool
for studying the glomerular filtration barrier in health and
disease.
Kidney International advance online publication, 7 October 2015;
doi:10.1038/ki.2015.308
KEYWORDS: glomerulus; podocyte; optical clearing; STED; super-resolution
Correspondence: Hjalmar Brismar, SciLifeLab, Box 1031, Solna 171 21,
Sweden. E-mail: brismar@kth.se
Received 7 May 2015; revised 23 June 2015; accepted 9 July 2015
Kidney International
Traditionally, electron microscopy has been the only method
available to visualize the glomerular filtration barrier, as the
dimensions of the podocyte foot processes and the slit
diaphragm are on the nanometer scale.1 In terms of spatial
resolution, electron microscopy is superior to light microscopy, due to the short effective wavelengths of electrons used
to reconstruct images.2 Even if the morphology of renal foot
processes has been dissected with electron microscopy, many
questions remain regarding the functional distribution of
proteins.1 The possibility to stain for proteins using
immunostaining or by expression of fluorescent proteins
would here be beneficial, if high enough spatial resolution
using light microscopy could be generated. With the use of
super-resolution stimulated emission depletion (STED)
microscopy,3 we demonstrate that the nanoscale localization
of proteins at the slit diaphragm is possible to study. Our
results show that super-resolution imaging in intact tissue
can be difficult due to unspecific/inhomogeneous staining
and autofluorescence background seen in paraformaldehyde
(PFA) fixed samples. These issues are probably reflected in the
limited number of high-resolution studies performed in renal
tissue to date. To the best of our knowledge, only two superresolution studies have been carried out,4,5 both performed
on ultrathin sections using a combination of transmission
electron microscopy and super-resolution single molecule
localization microscopy. High-resolution confocal microscopy has been used in one study of podocyte foot processes,
where genetic labelling was used to label a sparse subset of
podocytes.6 On this background, we have identified the need
for improved sample preparation protocols for superresolution microscopy in order to study functional structures
in kidneys using immunofluorescence at the nanoscale. By
applying an optical clearing protocol based on the
CLARITY7–9 and SeeDB10 methods, we show that background
as well as staining problems can be significantly decreased,
making it possible to study the spatial distribution of proteins
at the slit diaphragm in intact tissue.
RESULTS
Rat kidneys were optically cleared using a hydrogel-based
protocol,9 followed by immunostaining and mounting in
1
technical notes
Uncleared tissue
Hydrogel embedding
z =10 µm
b
Podocin
SDS lipid removal
z =100 µm
Fructose mounting
z =1,000 µm
Podocin
Podocin
Uncleared tissue
c
Attenuation
(µm–1)
a
D Unnersjö-Jess et al.: STED microscopy in cleared kidneys
Cleared tissue
0.04
0.03
0.02
0.01
0
Podocin
Podocin
Cleared tissue
d
Log(signal to
noise ratio)
Podocin
103
102
101
100
e
Uncleared Cleared
Uncleared Cleared
f
z
E-Cadherin Podocin
E-Cadherin Podocin
Figure 1 | Optical clearing of rat kidneys and millimeter-scale imaging. For confocal imaging, a Zeiss LSM780 system was used. (a) Overview
of the clearing process, with before and after images of 1 mm thick slices of a P20 rat kidney. (b) Confocal images acquired at three different
depths in cleared/uncleared 1 mm thick P20 rat kidney slices stained for podocin with Abberior STAR635P. Uncleared samples were
paraformaldehyde fixed, stained following the same immunostaining protocol as cleared kidneys, and mounted in phosphate-buffered saline
(PBS). Scale bars = 100 μm. (c) Attenuation coefficients in uncleared and cleared tissue. Cleared and uncleared kidney slices were stained for
E-cadherin with Alexa-647, and confocal z-stacks were acquired with an excitation laser of constant intensity. Mean fluorescence for each z-slice
was then plotted as a function of depth. Exponentials were fitted to the plots using MATLAB, and attenuation coefficients could then be
extracted. Error bars showing s.d. (n = 10). (d) Signal-to-noise ratios (SNRs) in cleared and uncleared kidney tissue stained for podocin, which is
expressed only in the glomerulus. The SNRs were measured as the ratio between mean fluorescence signal in the glomerulus and mean
fluorescence signal in an area with no glomerulus. Error bars showing s.d. (n = 7). (e) Three-dimensional (3D) rendering of a 1 mm thick P20
kidney slice stained for E-cadherin with Alexa-647 and podocin with Alexa-546. Inset showing an xz-view of the marked area. Confocal stacks
(tile scanning) acquired using a 5 × NA 0.25 air objective. Scale bar = 1 mm. (f) Substack of the same sample as in (e) acquired at a depth of
400–600 μm. Confocal stack acquired using a 10 × NA 0.45 air objective. Scale bar = 100 μm.
fructose solution.10 The optical transparency (Figure 1a) and
antibody penetration depth (Figure 1b and c) were sufficient
for imaging samples on the mm-scale using confocal
microscopy, allowing for a global view of kidney morphology
and protein expression (Figure 1e and f). Further, we show a
substantial increase in fluorescence contrast or signal-to-noise
ratio by a factor of ~ 100 (2.6 ± 0.44 to 230 ± 43) in cleared
samples by applying standard immunostaining protocols to
2
both PFA fixed and cleared samples (Figure 1d). For
super-resolution STED microscopy, samples were prepared
with the same protocol as for millimeter-scale imaging.
Immunostaining in cleared samples stained for podocin was
specific and highly localized (Figure 2d and e), which allows
elucidating the localization on both sides of the slit diaphragm
(Figure 2f). Comparison with non-cleared control samples
show that optical clearing was crucial to reveal the spatial
Kidney International
technical notes
D Unnersjö-Jess et al.: STED microscopy in cleared kidneys
distribution of podocin on the nanometer scale. Control
samples were stained following the same immunoprotocol as
cleared samples, but here the immunosignal was incomplete
(Figure 2g and h) and the nanoscale distribution of podocin
a
Podocyte foot
process
Slit
diaphragm
Endothelial cells
(capillary cell wall)
Urine
could not be revealed (Figure 2i). Furthermore, the resolution
as a function of depth is kept constant up to at least 30 μm
using oil immersion objectives (Figure 2j). This allows for
three-dimensional STED imaging, producing volumetric
b
Podocin
c
Podocin
Blood
Podocin
3D
STED
Nephrin
d
Podocin
e
Podocin
f
j
STED
Confocal
200
0
500
1,000
1,500
nm
Podocin
Podocin
h
i
Fluorescence
g
Cleared, confocal
150
Podocin
Nephrin
l
15
20
25
50
30
0
Uncleared, confocal
Fluorescence
k
10
STED
Confocal
100
0
Uncleared, STED
Width (nm)
0
50 100
5
100
0
Cleared, STED
0
Imaging depth (µm)
Fluorescence
300
500
1,000
nm
1,500
150
Podocin
Nephrin
100
50
0
0
500
nm
1,000
Figure 2 | Super-resolution stimulated emission depletion (STED) imaging in optically cleared kidney samples. All samples were stained
for podocin with Abberior STAR 635P. For STED/Confocal imaging, a Leica SP8 3X microscope with a 100 × NA 1.4 oil objective was used unless
other stated. (a) Schematic representation of the glomerular filtration barrier and the localization of podocin and nephrin at the slit diaphragm.
(b) Eighty micrometer deep confocal stack of a whole glomerulus in a cleared slice of kidney tissue acquired using a Zeiss LSM780 system and a
63 × NA 1.4 oil objective. Scale bar = 20 μm. (c) Three-dimensional (3D) renderings (Amira software) of a 3D STED z-stack display foot processes
at the surface of a glomerular capillary. Scale bar = 2 μm. (d–i) Deconvolved (SVI Huygen's) STED/Confocal images of glomerular foot processes
in cleared (d–e) and uncleared (g–h) kidney samples with intensity profiles (f, i) along marked arrows in STED/Confocal images, showing that
podocin expression at both sides of the slit diaphragm is revealed only in cleared samples. Scale bars = 500 nm. (j) Plot of measured slit
diaphragm width as a function of imaging depth. STED images were acquired at six different depths, and for each image the minimum
resolvable slit diaphragm width was measured at eight different positions. Error bars show s.d. of measured values. (k) Two-color STED image of
a cleared kidney sample stained for podocin (green) with Abberior STAR635P and nephrin (magenta) with Alexa-594. (l) Intensity profiles along
the marked arrow, showing the two-sided expression of podocin and the central expression of nephrin at the slit diaphragm.
Kidney International
3
technical notes
D Unnersjö-Jess et al.: STED microscopy in cleared kidneys
a
b
Podocin
Control, 3D STED
Podocin
Effacement fraction
1
*
0.8
0.6
0.4
0.2
0
Control
Podocin
PHN, STED
Control, STED
e
d
PHN
f
4
3
2
*
1
0
Control
PHN
Figure 3 | Superresolved glomerular pathology in passive
Heymann nephritis (PHN) rats. All samples were stained for podocin
with Atto 594 and imaged using a Leica SP8 3X microscope with a
100 × NA 1.4 oil objective. All images were deconvolved using SVI
Huygens software. (a–b) Three-dimensional (3D) renderings (Amira
software) of 3D stimulated emission depletion (STED) z-stacks
showing inside views of glomerular capillaries from control
(a) and PHN (b) rats. (c–d) STED images of glomerular foot processes
in control (c) and PHN (d) rats. (e) Quantification of slit diaphragm
effacement in control and PHN rats. The total length where slit
diaphragms could not be resolved was measured and divided by the
total length of the filtration slit. This was carried out along five
randomly selected segments, showing a significant *(Po0.0001)
increase in effacement for PHN rats. Error bars show s.d.
(f) Quantification of filtration slit coverage in control and PHN rats.
The total length of the filtration slit was divided by the total area of
interest. This was carried out for five randomly selected areas of at
least 14 μm2, showing a significant *(Po0.00001) decrease for PHN
rats. Error bars show s.d.
representations of foot processes on the nanometer scale
(Figure 2b and c). Kidney samples were additionally counterstained for nephrin, an abundant anchoring protein at the
filtration slit. Localizations of podocin (at podocyte membranes proximal to slit diaphragms) and nephrin (spanning
the slit diaphragm; Figure 2a) could clearly be resolved
(Figure 2k and l). The protocol was also applied to
quantitative analysis of glomerular pathology. Podocin
expression was analyzed in kidney samples from rats with
passive Heymann nephritis (PHN), a rat model of human
membranous nephropathy.11 We observed a distinct alteration of foot process morphology in PHN compared with
control rats (Figure 3a–d, Supplementary Movies S1–S2
4
online). These alterations were quantified in terms of the
effacement fraction and the foot process coverage (Figure 3e
and f). We found a significant increase in the effacement
fraction and a significant decrease in the foot process coverage
in PHN rats, in line with what has previously been observed
using electron microscopy.11
DISCUSSION
PHN, 3D STED
Foot process
coverage (µm/µm2)
c
Podocin
We demonstrate that hydrogel-based optical clearing enables
volumetric high-resolution fluorescence imaging of thick
kidney samples.12 Further, we show that optical clearing is an
important procedure in order to reveal the finest structures of
the filtration barrier. Conventionally prepared PFA fixated
samples give staining quality that is insufficient at the detailed
level now available in super-resolution microscopy. The
improved staining quality is due to the removal of lipids,
which results in an increased antibody penetration and
exposure of binding epitopes. Even if samples can be
mechanically sectioned for studies of fine structures, sample
transparency is beneficial in order to give access to large
imaging volumes. A large imaging volume facilitates identification of regions of interest, and large quantities of imaging
data can be extracted from the same sample. Also, the risk of
mechanically damaging the sample by physical sectioning is
eliminated. Concern has been raised that clearing tissue by
removing lipids may introduce artifacts where the fine
localization of proteins is lost. By the use of STED
microscopy, we found that this is not an issue in the
presented protocol. Tissue morphology is well preserved even
at the nanoscale, as demonstrated by the organization of
podocin and nephrin in the slit diaphragm.
To conclude, the presented sample preparation protocol
adds a novel tool for high- and super-resolution microscopy
studies of protein localization with an impact for understanding pathologies occurring at the glomerular filtration
barrier.
MATERIALS AND METHODS
Rats
Rats from Charles River, Germany, were used in all experiments.
PHN was induced in 40-day-old rats as described by Saran et al.11
16 weeks before euthanization. Animals were anesthetized (intraperitoneal injection of pentobarbital), the aorta cut, and kidneys
dissected. All experiments were performed in accordance with
animal welfare guidelines set forth by Karolinska Institutet and were
approved by Stockholm North Ethical Evaluation Board for Animal
Research.
Optical clearing
Kidneys were dissected and immediately incubated at 4 °C in
hydrogel solution (1–2% v/v acrylamide, 0.0125–0.025% v/v bisacrylamide, 0.25% w/v VA-044 initiator, 4% PFA, 1 × phosphatebuffered saline (PBS)) for 5 days. The gel was polymerized at 37 °C
for 3 h, and the presence of oxygen was minimized by filling
tubes to the top with hydrogel solution. Samples were removed
from the hydrogel solution and immersed in clearing solution
(200 mmol/l boric acid, 4% SDS, pH 8.5) and incubated for
Kidney International
technical notes
D Unnersjö-Jess et al.: STED microscopy in cleared kidneys
1 day. Kidneys were cut in 0.8–1.5 mm thick slices using a
Vibratome (myNeurolab, St Louis, MO) and incubated at 50 °C for
2 weeks with clearing solution changed every 3 days. Before
immunolabelling, samples were incubated in PBST (0.1% Triton-X
in 1 × PBS) for 1 day. For PHN and PHN-negative control, kidneys
were fixed in Duboscq-Brasil solution for 6 h, then hydrated and
immersed in hydrogel solution (4% acrylamide, 0% bis-acrylamide) without PFA, and then the above protocol was followed.
Supplementary material is linked to the online version of the paper at
http://www.nature.com/ki
REFERENCES
1.
2.
3.
Immunolabelling
For all steps PBST was used as dilutant. Samples were incubated in
primary antibody for 24 h at 37 °C and then washed in PBST
for 8 h at 37 °C followed by secondary antibody incubation for
24 h at 37 °C and washed for 8 h at 37 °C prior to mounting. To
stain for E-cadherin, a mouse anti-E-cadherin primary antibody
(BD Biosciences, San Jose, CA, 610182, 1:50) and a goat anti-mouse
Alexa-647-conjugated secondary antibody (1:100) were used. To
stain for podocin, a rabbit anti-podocin primary antibody (SigmaAldrich, St Louis, MO, P0372, 1:50) and a goat anti-rabbit Alexa-546
(1:200), Abberior STAR635P (1:100), or Atto-594 (1:100) secondary
antibody were used. To stain for nephrin, a goat anti-nephrin
primary antibody (Santa Cruz Biotechnology, Santa Cruz, CA,
sc19000 1:100) and a donkey anti-goat Alexa-594 secondary antibody
(1:200) were used.
Mounting
Samples were immersed in 40% (w/v) fructose with 0.5% (v/v)
1-thioglycerol for 2 h and then transferred to 80% (w/w) fructose
with 0.5% (v/v) 1-thioglycerol for 24 h.
Samples were mounted in a glass bottom dish (MatTek, Ashland,
MA, P35G-1.5-14-C) before imaging.
DISCLOSURE
All the authors declared no competing interests.
ACKNOWLEDGMENTS
This study was supported by grants from the Swedish Research
Council (VR 2013-6041). We thank Evgeniya Burlaka for support with
PHN induction.
SUPPLEMENTARY MATERIAL
Movie S1. Foot processes in healthy rat.
Movie S2. Foot processes in rat with PHN renal disease.
Kidney International
4.
5.
6.
7.
8.
9.
10.
11.
12.
Grahammer F, Schnell C, Huber TB. The podocyte slit diaphragm –
from a thin grey line to a complex signaling hub. Nat Rev Nephrol 2013; 9:
587–598.
Rose HH. Optics of high performance electron microscopes. Sci Technol
Adv Mater 2008; 9: 014107.
Hell S, Wichmann J. Breaking the diffraction resolution limit by stimulated
emission: stimulated emission depletion microscopy. Opt Lett 1994; 19:
780–782.
Suleiman H, Zhang L, Roth R et al. Nanoscale protein architecture
of the kidney glomerular basement membrane. Elife 2013; 2:
e01149.
Yu H, Suleiman H, Kim AH et al. Rac1 Activation in podocytes induces
Rapid Foot Process Effacement and Proteinuria. Mol Cell Biol 2013; 33:
4755–4764.
Grgic I, Brooks CR, Hofmeister AF et al. Imaging of podocyte foot
processes by fluorescence microscopy. J Am Soc Nephrol 2012; 23:
785–791.
Chung K, Wallace J, Kim SY et al. Structural and molecular interrogation of
intact biological systems. Nature 2013; 497: 332–337.
Chung K, Deisseroth K. CLARITY for mapping the nervous system. Nat
Methods 2013; 10: 508–513.
Tomer R, Ye L, Hsueh B et al. Advanced CLARITY for rapid
and high resolution imaging of intact tissues. Nat Protoc 2014; 9:
1682–1697.
Ke MT, Fujimoto S, Imai T. SeeDB: a simple and morphology-preserving
optical clearing agent for neuronal circuit reconstruction. Nat Neurosci
2013; 16: 1154–1161.
Saran AM, Yuang H, Takeuchi E et al. Compliment mediates nephrin
redistribution and actin dissociation in experimental membranous
nephropathy. Kidney Int 2003; 64: 2072–2078.
Yang B, Treweek JB, Kulkarni RP et al. Single-cell phenotyping within
transparent intact tissue through whole-body clearing. Cell 2014; 158:
945–958.
This work is licensed under a Creative Commons
Attribution-NonCommercial-ShareAlike
4.0
International License. The images or other third party
material in this article are included in the article’s Creative
Commons license, unless indicated otherwise in the credit
line; if the material is not included under the Creative
Commons license, users will need to obtain permission from
the license holder to reproduce the material. To view a copy of
this license, visit http://creativecommons.org/licenses/by-ncsa/4.0/
5