Methods for adjustment of refractive index
in kidney tissue
David Unnersjö-Jess (19850203-0151)
daviduj@kth.se
SA104X Master’s Thesis in Engineering Physics
Department of Cellular physics
Royal Institute of Technology (KTH)
Examiner: Hjalmar Brismar
Supervisor: Lena Scott
June 11, 2014
Abstract
Optical clearing of biological samples is of great interest as it opens up the opportunity to image large intact samples of tissue on a global level. Numerous methods for optical clearing have been presented such as CLAR-ITY [1], SeeDB [3] and Scale [4]. While these publications focus mostly on clearing brain tissue, this work investigates the effect of clearing protocols on kidney tissue by ap-plying a combination of the CLARITY and
1
Introduction
When imaging larger samples of tissue using techniques based on visible photons, imaging depth is limited due to attenuation of light. The attenuation is a result of both absorption and scattering, but the latter is the dominat-ing factor in biological tissue [8]. Scatterdominat-ing of light takes place at interfaces where there is a change in refractive index [12], and these changes has to be adjusted in order to increase light transparency. SeeDB and Scale are ex-amples of methods that obtain optical clearing by immersion in a medium with refractive in-dex matching that of biological tissue. These methods result in good transparency for em-bryonic and neonatal samples, but for adult samples the clearing is less successful (See ref. [3]) due to the difference in refractive in-dex still present at the lipid/protein interfaces [9] [10]. In 2013, the CLARITY method was presented [1], which provides a protocol for optical clearing of adult mouse brains. CLAR-ITY eliminates these residual differences in refractive index by removing lipids from the tissue without altering morphology [1]. In this work the effect of CLARITY and SeeDB on kidney tissue transparency is evalu-ated using confocal microscopy techniques.
2
Materials and methods
In all experiments kidneys from P20 rats (charles river, Germany) were used. All ex-periments were performed in accordance with animal welfare guidelines and regulations set fourth by Karolinska Institutet.
2.1
Clearing process
To clear samples, the CLARITY protocol [5] was followed with the exceptions mentioned below. For detailed protocol and mixing of the Hydrogel and ETC solutions, see ref. [5].
2.1.1 Hydrogel tissue embedding
The cardiac perfusion step was omitted, and kidneys were directly incubated in hydrogel solution at 4◦C for 10 days, as opposed to the 2-3 days suggested in ref. [5] for perfused samples. The oxygen removal step using a desiccator was also omitted and instead the presence of oxygen was minimized by filling the tubes with hydrogel solution all the way to the top prior to polymerization. Following polymerization, kidneys were washed in ETC solution for 3 days.
2.1.2 Electrophoretic Tissue Clearing (ETC)
Figure 1: ETC setup
Our setup for Electrophoretic Tissue Clearing is shown in Fig. 1. The pump used to cir-culate the ETC solution was a JBL ProFlow u500 fish tank pump. This was connected to a custom designed ETC chamber (see Appendix A) and a filter (Clas Ohlson Art. number 30-6630) with a 20 micron filter cartridge.
supply for the electrodes.
Kidneys were run in this setup for 10 days at 50◦C and 15V. Following ETC, kidneys were washed for 2-3 days in PBST as suggested in Ref. [5].
2.1.3 Staining
Samples were stained with anti-E-Cadherin antibodies (BD Transduction laboratories Cat.610182) to image the distal convoluted tubules. Slices of cleared tissue (≈1mm) were cut by hand using a razor. Slices were incu-bated at 37 degrees in primary antibody (dilu-tion 1:200), 1X PBST (0.1% TritonX) pH 7.5 for 48h, followed by washing in PBST for 24h. Samples were then incubated at 37 degrees in secondary antibody (AlexaFluor-633, dilution 1:200), 1X PBST for 48h, followed by wash-ing in PBST for 24h.
To image the glomeruli and distal convoluted tubules, samples were stained with Alexa-488 conjugated Wheat Germ Agglutin (WGA) (Molecular Probes W11261). 1 mm slices were incubated in 1X HBSS and 10 µg/ml WGA for 3 days, followed by 3 days of wash-ing in 1X PBST.
2.1.4 Refractive index matching
In our experiments the SeeDB clearing agent [3] was used for refractive index matching. The full protocol in Ref. [7] was followed for whole kidneys, but for 1mm slices it was suf-ficient to incubate the samples in SeeDB so-lution at 37◦C over night. It should be men-tioned that in Ref. [1] a clearing agent called FocusClear is used for this step of the proto-col. FocusClear was never used in these stud-ies, and nothing can be said on how it com-pares to the SeeDB agent.
2.2
Imaging
The sample was placed on a glass bottom dish along with the immersion medium (SeeDB), and a cover glass was placed on top. The sam-ples were then imaged using a Zeiss LSM510 confocal microscopy setup. AlexaFluor-633 was excited using a 633 nm HeNe laser and detected at >650 nm using a LP650 filter. Alexa-488 was excited using a 488 nm Argon laser and detected using a BP 505-550 filter.
3
Results
(a) No CLARITY/PBS (b) CLARITY/PBS
(c) CLARITY/SeeDB
0 100 200 300 400 500 600 0 10 20 30 40 50 60 70 80 90 100 No CLARITY/PBS Depth(µm)
Mean Intensity (au)
0 100 200 300 400 500 600 0 20 40 60 80 100 120 140 No CLARITY/SeeDB Depth(µm)
Mean Intensity (au)
0 100 200 300 400 500 600 0 10 20 30 40 50 60 70 80 90 CLARITY/PBS Depth(µm)
Mean Intensity (au)
0 100 200 300 400 500 600 0 5 10 15 20 25 30 35 40 45 50 CLARITY/SeeDB Depth(µm)
Mean Intensity (au)
z = 100 µm z = 200 µm z = 500 µm No CLARITY + PBS No CLARITY + SeeDB CLARITY + PBS CLARITY + SeeDB
(a) Glomerulus
(b) Distal convoluted tubule
Figure 2: Increase in transparency of kidney at different steps of the protocol. Figure 2(a): PFA fixed kidney in PBS. Figure 2(b): Kid-ney after 3 days of passive clearing and 10 days of ETC in PBS. Figure 2(c): Kidney after 3 days of passive clearing and 10 days of ETC immersed in SeeDB and prepared for imaging.
Figure 3: Decrease in fluorescence intensity as a function of depth at different steps of the protocol. Samples were stained with anti-E-Cadherin antibodies and the fluorescence sig-nal was detected with a 10X air objective at a constant 633 HeNe excitation laser intensity. The mean intensity for each optical slice was calculated using ImageJ and plotted against depth. Since the dependence of fluorescence intensity on excitation laser intensity is lin-ear (if the fluorophore is not saturated) [13], the fluorescence signal can be used to measure laser attenuation. Laser intensity as a func-tion of depth will follow the Beer-Lambert law [14]:
I I0
= e−µz,
where I is the intensity at depth z, I0 is the
laser intensity before entering the sample and µ is the attenuation coefficient. Since there is attenuation of both the excitation and fluores-cence signals, the detected fluoresfluores-cence will be
I I0
= e−µze−γµz = e−(1+γ)µz,
where γ is a constant representing the change in attenuation due to the red shift of the fluo-rescence signal. By fitting each curve in Fig. 3 to an exponential function using MATLAB, the product (1 + γ)µ for each case could be determined, and the relative change in attenua-tion coefficients upon clearing shown in Fig. 3
was given by µ µ0 = (1 + γ)µ (1 + γ)µ0 ,
where µ0 is the attenuation coefficient for the
non-cleared kidney immersed in PBS. Fig-ure 4: Image quality at different depths in renal cortex for E-Cadherin stained samples. The effect of CLARITY and SeeDB on image quality are shown separately and in combina-tion (10X air objective, Scale bar 100 µm). Figure 5(a): Glomerulus stained with Alexa-488 conjugated WGA and imaged with a 63X oil immersion objective (Scale bar 50 µm). Figure 5(b): E-Cadherin/Alexa-633 stained convoluted distal tubule imaged with a 63X oil immersion objective (Scale bar 50 µm).
4
Discussion
Kidneys were successfully cleared and flu-orescently stained, showing optical clearing methods to be applicable to kidney studies. The protocol for both CLARITY and antibody staining had to be somewhat prolonged to obtain good results, and whether this had to do with methods used in the studies or was due to the structural difference of kidney tissue compared to brain tissue has to be investigated further. A possible cause of the slower clearing could be the exclusion of cardiac perfusion and extension of the hydrogel embedding step.
From Fig. 4 it is apparent that the image qual-ity at larger depths is improved for cleared tissue, showing that scattering effects due to changes in refractive index has been highly reduced.
From Fig. 2, Fig. 3 and Fig. 4 it is clear that the removal of lipids using the CLARITY protocol as well as refractive index matching using SeeDB were both important to render the tissue transparent. CLARITY without refractive index matching (imaged in PBS) as well as immersion in SeeDB without CLAR-ITY had some effect on the transparency, but it was not until CLARITY and SeeDB were combined that dramatic effects on laser penetration and image quality at larger depths were observed.
In Fig. 5 it is shown that the morphological characteristics of kidney tissue on a global level are preserved after optical clearing. However, antibody staining for the WT-1 protein in podocytes was attempted using several different protocols, but resulted in no visible fluorescence signal. This suggests that the clearing process has made the epi-tope undetectable for the antibody. Possible reasons for this could be overfixation of the tissue, molecular alterations of the protein or problems with the antibody itself. As the unsuccessful staining could indicate tissue damage caused by the clearing process, it has to be a subject for further studies whether the morphology is preserved not only on a global level, but also on a molecular level.
Improvements can be made on the imaging procedure as there were difficulties with imaging at depths larger than about 500 µm. When imaging with the 63X oil immersion objective, the imaging depth was restricted by the working distance of the objective (≈ 100 µm). When using the 10X air
objec-tive, axial resolution got poor around 500 µm, as seen in Fig. 4. A possible reason for this could be spherical aberrations introduced by the refractive index mismatch between air and SeeDB [11]. Another reason could be that there are still small inhomogeneities in refrac-tive index, causing blurring of deeper parts of samples. Both the spherical aberrations and the limited working distance could to a large extent be avoided by using optics optimized for SeeDB, as in ref. [3].
References
[1] C. Deisseroth et al., Structural and molecular interrogation of intact bio-logical systems, Nature 497, 332-337 (16 May 2013).
[2] K. Chung and C. Deisseroth, CLARITY for mapping the nervous system, Nature Methods 10, 508-513 (20 June 2013).
[3] M-T. Ke, S. Fujimoto and T. Imai, SeeDB: a simple and morphology-preserving optical clearing agent for neuronal circuit reconstruction, Nature Neuroscience 16, 1154-1161 (23 June 2013).
[4] H. Hama et al., Scale: a chemical ap-proach for fluorescence imaging and reconstruction of transparent mouse brain, Nature Neuroscience 14, 1481-1488 (30 August 2011).
[5] http://
[6] http:// clarityresourcecenter. org/pdfs/ETC_chamber_ instructions.pdf [7] http://bio-protocol.org/ wenzhang.aspx?id=1042# .U2t7JFeBglI
[8] W-F. Cheong, S.A. Prahl and A.J. Welch, A Review of the Optical Properties of Biological Tissues, IEEE Journal of Quantum Electronics 26, 2166-2185 (December 1990).
[9] A. Lacourt and N. Delande, Thermo mi-croidentification of amino acids: Re-fractive index, Microchimica Acta 52, 547-560 (1964).
[10] E. Anibal Disalvo and S.A. Simon, Per-meability and Stability of Lipid Bilay-ers, p204 (1994).
[11] S. Hell, G. Reiner, C. Cremer and E.H.K. Stelzer Aberrations in confo-cal fluorescence microscopy induced by mismatches in refractive index, Journal of Microscopy 169, 391-405 (March 1993).
[12] M. Kerker The Scattering of Light, Aca-demic Press (1969)
[13] J.M. Seitzman, R.K. Hanson, P.A. De-Barber and C.F. Hess Application of quantitative two-line OH planar laser-induced fluorescence for temporally re-solved planar thermometry in reacting flows, Applied Optics 33, 4000-4012 (1994).
A
Custom designed ETC chamber
The ETC chamber described in ref. [6] was tried out at first, but this design showed severe leakage problems as well as problems with electrolytic effects at the electrodes. To solve these issues, a custom designed ETC chamber was fabricated in cooperation with the mechanical workshop at the Royal Institute of Technology. The design is shown in Fig. 6, Fig. 7, Fig. 8 and Fig. 9.
Figure 6: Chamber as viewed from the top, photo
Figure 8: Chamber as viewed from the side, photo
Figure 9: Chamber as viewed from the side, blueprint
Figure 11: Lid design
a two chambers in series-design with the inlets placed at the bottom of each chamber and the outlets placed at the top to prevent bubble formation (as suggested in Ref. [2]).
Figure 10: Lid design
The lid was made of a solid 0.5 cm thick plexi plate of the same size as the chamber block. A polypropylene net was glued to the lid using epoxy (See Fig. 12) to create a basket for the sample. Four electrode inlets (0.4 mm diameter) were drilled on each side of the basket and platinum electrodes (Sigma GF95639901 0.4 mm diameter) were glued into these holes using epoxy (Fig. 12). The resulting lid design is shown in Fig. 10 and Fig. 11.
Figure 12: Lid design
