Label-free super-resolution imaging enabled by VISTA (Vibrational Imaging of Swelled Tissue and Analysis)

Label-free super-resolution imaging enabled by VISTA

(

ibrational

maging of

welled

issue and

nalysis)

Kun Miao

Li-En Lin

Chenxi Qian

Lu Wei

Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena,

California 91125, United States

Abstract

The universal utilization of fluorescence microscopy, especially super-resolution microscopy,

has greatly advanced our knowledge about modern biology. Conversely, the requirement

of fluorophore labeling in fluorescent techniques poses significant challenges, such as

photobleaching and non-uniform labeling of fluorescent probes and prolonged sample processing.

In this protocol, we report the detailed working procedures, termed VISTA, to circumvent

obstacles associated with fluorophores and achieve label-free super-resolution volumetric imaging

in biological samples with spatial resolution down to 78 nm. VISTA is established by embedding

cells and tissues in hydrogel, isotropically expanding the hydrogel-sample hybrid, and visualizing

endogenous protein distributions by vibrational imaging with stimulated Raman scattering

microscopy. We demonstrated VISTA on both cells and mouse brain tissues and observed

highly correlative VISTA and immunofluorescence images, validating the protein origin of

imaging specificities. We exploited such correlation and trained a machine-learning based image-

segmentation algorithm to achieve multi-component prediction of nuclei, blood vessels, neuronal

cells and dendrites from label-free mouse brain images. We further adapted VISTA to investigate

pathologic poly-glutamine (polyQ) aggregates in cells and amyloid-beta (A

) plaques in brain

tissues with high throughput, justifying its potential for large-scale clinical samples.

SUMMARY:

By combining sample-expansion hydrogel chemistry with label-free chemical specific stimulated

Raman scattering microscopy, we described the protocol to achieve label-free super-resolution

volumetric imaging in biological samples. With additional machine-learning image segmentation

algorithm, we obtained protein specific multi-components images in tissues without antibody

labeling.

Corresponding author. lwei@caltech.edu.

A complete version of this article that includes the video component is available at

http://dx.doi.org/10.3791/63824

DISCLOSURES:

The authors declare no competing interests.

HHS Public Access

Author manuscript

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

INTRODUCTION:

The development of optical imaging methods has revolutionized our understanding of

modern biology because they provide unprecedented spatial and temporal information

of targets across different scales, from subcellular proteins to whole organs

. Among

them, fluorescence microscopy is the most well-established: with a large palette of

organic dyes with high extinction coefficients and quantum yields

, easy-to-use genetic

encoded fluorescent proteins

, and super-resolution methods such as STED, PALM and

STORM for imaging nanometer-scale structures

. In addition, recent advancement in

sample engineering and preservation chemistry, which expands specimens embedded in

swellable polymer hydrogels

, enables sub-diffraction limited resolution on conventional

fluorescence microscopes. For instance, typical expansion microscopy (ExM) effectively

enhances the image resolution by 4 times with fourfold isotropic sample expansion

Despite its advantages, super-resolution fluorescence microscopy shares limitations that

originate from fluorophore labeling. First, photobleaching and inactivation of fluorophores

compromises the capacity of repetitive and quantitative fluorescence evaluations.

Photobleaching is an inevitable event when light keeps pumping electrons into electronic

excited states

. Second, labeling the fluorophores to the desired targets is not always a

straightforward task. For instance, immunostaining demands a long and laborious sample

preparation process and hinders imaging throughput

. It could also introduce artifacts due

to inhomogeneous antibody-labeling, especially deep inside tissues

. Moreover, proper

labeling strategies that targets fluorophores to desired proteins might be underdeveloped. For

example, extensive screenings were required to find effective antibodies for A

plaques

Smaller organic dyes, such as Congo red, often have limited specificity, only staining

the core of the A

plaque. We aim to develop a label-free super-resolution modality that

circumvents the drawbacks from fluorophore-labeling and provides complementary high-

resolution imaging from cells to tissues, and even to large-scale human samples.

Raman microscopy provides label-free contrast for chemical-specific structures and maps

out the distribution of otherwise invisible chemical bonds by looking at the excited

vibrational transitions

. In particular, stimulated Raman scattering (SRS) imaging on label-

free or tiny-labeled samples has been demonstrated to have similar speed and resolution

to fluorescence microscopy

. For example, healthy brain region has been readily

differentiated from tumor infiltrated region in human and mouse tissues

. A

plaques

was also clearly imaged by targeting protein CH

vibration (2940 cm

−1

) and amide I (1660

−1

) on a fresh-frozen brain slice without any labeling

. Raman scattering therefore offers

robust label-free contrast that overcomes the limitations of fluorophores. The question then

became how we can accomplish super-resolution capacity using Raman scattering, which

could reveal nanoscale structural details and functional implications in biological samples.

Although extensive efforts have been made to achieve super-resolution for Raman

microscopy with elegant optic instrumentations, the resolution enhancement on biological

samples have been rather limited

. Here, based on our recent works

, we present

a protocol that combines a sample-expansion strategy with stimulated Raman scattering

for super-resolution label-free vibrational imaging, named

ibrational

maging of

welled

Miao et al.

Page 2

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

issues and

nalysis (VISTA). We embedded cells and tissues in hydrogel matrixes through

an optimized protein-hydrogel hybridization protocol. We then incubated hydrogel tissue

hybrids in the detergent-rich solutions for delipidation followed by expansion in water.

The expanded samples were then imaged by a regular SRS microscope by targeting CH

vibrations from retained endogenous proteins. VISTA, owing its label-free imaging feature,

bypasses photobleaching and inhomogeneous labeling arising from fluorophore labeling

with much higher sample processing throughput. VISTA also for the first time enabled

sub-100 nm (down to 78 nm) label-free imaging with no change on SRS instrumental

setup

, making itself readily applicable. With correlative VISTA and immunofluorescence

images, we trained an established machine-learning image-segmentation algorithm

that

generates protein-specific multiplex images from single-channel VISTA images. We further

applied VISTA to investigate A

plaques in mouse brain tissues, providing a holistic image

suited for sub-phenotyping based on the fine views of the plaque core and peripheral

filaments surrounded by cell nuclei and blood vessels.

PROTOCOL:

All animal procedures performed in this study were approved by the California Institute of

Technology Institutional Animal Care and Use Committee (IACUC), and we have complied

with all relevant ethical regulations.

1. Preparation of stock solutions for fixation and sample expansion.

1.1. Prepare 40 ml of fixation solution by first dissolving 12g of acryl amide (30% w/v) solid

in 26 ml of nuclease-free water. Then add 10 ml of 16% PFA stock solution in the mixture.

Finally, add 4 ml of 10x phosphate-buffered saline (PBS, pH 7.4). Made solution can be

stored in 4 °C for 2 weeks.

Note: Acrylamide is hazardous so the step of dissolving acrylamide solid in water should be

handled in fume hood.

1.2. Prepare of gelation solution (stock X) by dissolving 70 mg of sodium acrylate (7% w/v),

200 mg of acryl amide (20% w/v), 50 μl N,N

′

-methylenebisacrylamide (0.1% w/v) in 420

μl ultrapure water, add 57 μl sterile-filtered 10x PBS (pH 7.4) in the end . Solution can be

stored in −20 °C up to 1 week.

Note: Avoid sodium acrylate solid that forms clumps, make sure sodium acrylate used is

dispersive powders. Stock X made will be a colorless liquid; if the liquid looks light yellow,

obtain a new sodium acrylate source.

1.3. Prepare the polymerization initiator solution by dissolving 1 g of ammonium persulfate

(APS, 10% w/w) in 9 ml nuclease-free water. Similarly, prepare polymerization accelerator

solution by dissolving 1 g of tetramethylethylenediamine (TEMED, 10% w/w) in 9 ml

nuclease-free water. The resulting solutions should be aliquot into small portions and stored

in −20 °C.

Miao et al.

Page 3

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

1.4. Denaturing buffer (50 ml) was prepared by dissolving 2.88 g of sodium dodecyl sulfate

(SDS, 200 mM), 0.584 g of sodium chloride (200 mM), and 2.5 ml of 1M Tris-HCl buffer

(50 mM, pH 8) in nuclease-free water.

Note: The concentration of SDS is close to its saturation condition. Some crystallization in

the buffer is normal. Slight warm up can make the solution clear and ready to use.

2. Preparation of mammalian cell samples.

2.1 Cultured HeLa cells were first seeded on to borosilicate coverslip (#1.5) and were

cultured in Dulbecco's Modified Eagle Medium with 10% fetal bovine serum and 1%

penicillin–streptomycin antibiotics (complete DMEM) under humidified atmosphere with

5% CO

at 37 °C.

2.2 To obtain cells in different mitotic stage, incubate the cells in DMEM without fetal

bovine serum for 20 h to synchronize the cell stages once they reached 60% confluency.

Switch the medium for cells to complete DMEM and incubate for another 2-6 hour to target

various stages of mitosis. Wash the HeLa cells on coverslips with sterile PBS and incubated

them with fixation solution (4% PFA 30% AA in PBS) at 37 °C for 7-8 h before further

processing.

2.3 For cells with polyQ aggregates: Cells were first grown in complete DMEM until

they reached 70–90% confluence. After changing to new complete DMEM, 1 μg plasmids

encoding mHtt97Q-GFP was transfected into the cells using transfection agent (details in

table of materials). 24–28 h post-transfection, the coverslips were harvested by fixing the

cells in fixation solution (4% PFA, 30% AA) at 37 °C for 6–8 h. The resulting cell samples

were ready for further processing

3. Preparation of mouse brain samples.

3.1 The normal mice and Alzheimer mice were purchased from commercial sources

(details in table of materials). Rodent euthanasia via carbon dioxide narcosis was performed

according to standard protocol

. In brief, place mice in a chamber and fill the chamber with

a flow of 100% CO

in the order of 30–70% of the volume of the chamber per minute and

maintain the flow for at least one minute after clinical death. After confirming the humane

euthanasia, post-mortem tissue collection of the brain was performed by trained veterinary

technical staff.

3.2 First, wash the fresh mouse brain with ice cold PBS and transfer the brain into fixation

buffer (4% PFA, 30% AA). Incubate the mouse brain tin fixation buffer first at 4 °C for

24-48h and then at 37 °C for overnight. After washing it with sterile PBS, cut the mouse

brain into 150 μm sections using the vibratome and stored in PBS at 4 °C.

4. Hydrogel embedding, denaturation, and expansion of cell and tissue samples

4.1 Make sure the samples were incubated with fixation buffer for proper amount of time,

as indicated above. Thaw Stock X, free-radical initiator, and accelerator stock solutions and

keep them on ice during the whole process.

Miao et al.

Page 4

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

4.2 Pick up the coverslip with cells and place it on a glass slide by a tweezer. The brain

slices (150 μm thick) can be picked up and placed onto glass slide (make sure they are

flat) by a soft-wool paint brush. Get rid of excess liquid that come with samples. Stack 2

coverslips (# 1.5) on both sides of the sample on the glass slide to make a gelation chamber.

Place the glass slide with sample facing upward onto an ice-cold heat block and cool it to 4

°C for 1 min.

4.3 First, add 10% (w/w) TEMED to Stock X and then 10% APS solution. Quickly vortex

and slowly drop the resulting solution into the chamber onto samples to cover up the surface

of the sample, avoid air bubbles. Once the sample is fully immersed by the solution, place a

flat parafilm-covered coverslip on top of sample as a lid for the chamber. Leave the chamber

on ice or ice-cold heat block for another 1 min before incubating it in a humidify incubator

at 37 °C for 30 min.

Note: Sample hydrogel embedding must be done within 1 min after mixing stock X and

APS. The whole process should be kept on ice (or ice-cold heat block) as much as possible

to avoid gel formation (high degree of polymerization) before adding to the sample.

4.4 Take out the gelation chamber on the glass slide from the 37 °C incubator. A non-

transparent gel should be observed after removing the lid. Retrieve the gel by cutting with

razorblade or alternatively put the glass slide into denaturing buffer at room temperature

for 15 min. The gel will separate itself from the glass slide. Incubate the isolated gel in

denaturing buffer under heat condition to achieve further denaturation and delipidation. For

cell samples, denaturing at 95 °C for 1 hour would be sufficient. For 150 μm thick brains

slices, denaturing needs to be done at 70 °C for 3 h and 95 °C for 1 h.

Note: The gel could become curly when first incubated with denaturing buffer at room

temperature. It is normal and will become flat after high temperature denaturation. When

working with thicker tissue samples, longer denaturation time is needed.

4.5 After the heat treatment, wash the denatured sample 3 times with PBS for 10 min. At

this point, the gel should have expanded to about 1.5 times of the original size. Incubate the

washed gel with ultrapure water in a large container (at least 20 times the volume of the

gel) to achieve higher expansion ratio. Change the water every 1 hour 3 times and leave the

sample in darkness overnight. The resulting gel is ready to image.

Note: The expansion ratio largely depends on the mechanical properties of the original

sample. We obtained 4.2 times expansion for our cell samples and 3.4 times expansion for

brain tissue sample.

5. Label-free imaging of endogenous protein distribution in expanded cell and tissue

samples

5.1 Keep the expanded gel samples in H

O throughout entire imaging process. Place the

expanded sample-bearing gel onto a microscope slide (1 mm thick) with a microscope

spacer filled with water. Cover the spacer with a coverslip (#1.5) and ensure its properly

sealed to avoid sample movement. The spacer with appropriate opening sizes and depths was

used to hold the sample hydrogel hybrid and water.

Miao et al.

Page 5

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

Note: The gel is fragile after expansion. It needs to be handled with caution.

5.2 Place the sample with either the expanded cell or tissue onto the motorized stage with

the coverslip facing the objective. Use the bright field to find z and adjust the condenser

to proper position for Köhler illumination. Tune the pump wavelength to 791.3 nm on

laser panel control to target protein CH

vibration at 2940 cm

−1

. Switch the light of the

microscope from bright-field (eye piece) to laser scanning mode and open the SRS laser

shutter. Find the proper focus by changing the z while looking at the real-time SRS image

under short pixel-dwell time (12.5 μs/pixel) and low image resolution (256x256).

5.3 Once the ideal z-position is found, perform SRS imaging at higher resolution

(1024x1024) with a longer pixel-dwell time (80 μs/pixel) that matches the time constant

of the lock-in amplifier. Acquire volumetric images by collecting a z-stack with a step-size

of 1 μm in z-direction. Process and analyze the saved OIR file from the Olympus software

by ImageJ.

Note: In our SRS setup, a tunable picosecond laser with an 80-MHz repetition rate provides

the pump (770–990 nm) and Stokes (1031.2 nm) excitation lasers into a laser scanning

confocal microscope (detailed in table of materials). Temporal and spatial overlapping of the

two beams were optimized for signals with pure D

O. It takes some efforts to find the right

z for the expanded sample under bright field because refractive index is very homogenous

throughout the gel.

5.4 We determined the resolution of VISTA by taking an SRS image of polystyrene beads

(100 nm) using both a 25x objective and a 60x objective. The pump laser was set to 784.5

nm, corresponding to Raman shift of 3050 cm

−1

, characteristic of the aromatic C-H stretch

vibrations of polystyrene. The pump laser wavelength used here is also close to what we

use in VISTA for proteins. With the 60x objective, the experimental FWHM of the bead

image was 276.17 nm. We modeled the function of the bead object as a half circle; when the

PSF Gaussian function has a c (

) = 269 nm/2.35, the convoluted bead image would have a

measured FWHM of 276.17 nm. As a result, the resolution of our SRS system is 269 nm *

1.22 = 328 nm by the Rayleigh Criterion. As VISTA has an average of 4.2 times expansion

on cell samples, the effective resolution of VISTA is down to 328 nm / 4.2 = 78 nm.

6. Correlative VISTA and Fluorescent imaging of immuno-labeled and expanded tissue

samples

6.1 After denaturation, pre-incubate the hydrogel embedded samples (e.g. 150 μm brain

coronal section) in 1% (v/v) Triton X-100 (PBST) for 15min. Then switch the incubation

buffer to PBST with primary antibody at 1:100 dilution. If multiple protein targets are

needed for multiplex imaging, dilute respective primary antibodies for different targets to

proper concentrations in the same cocktail and incubated with the samples simultaneously.

Incubate the gels with diluted primary antibodies at 37 °C with gentle shaking (RPM 80) for

16-18 h, followed by extensive washing (3 times) with PBST for 1-2 h at 37 °C.

6.2 Incubate the samples with secondary antibodies of corresponding species targets at

1:100 dilution with PBST at 37 °C for 12-16 h, protected from the light. Wash labeled

Miao et al.

Page 6

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

hydrogel samples 3 times by PBST for 1-2 h at 37 °C with gentle shaking. Expand the

immuno-labeled gel samples by incubating with a large volume of double deionized H

Change the water every 1 h 3 times and incubate samples in H

O overnight, protected from

light.

Note: During labeling, the gel should be in 1.5 times expansion as it is in PBS buffer.

Flipping over the gel during the incubation would help prevent inhomogeneous antibody

labeling. Double-check antibody cross-reactivities to avoid crosstalk between fluorescence

channels. Cell nuclei are stained by DAPI.

6.3 Prepare the imaging sample as described section 5.1. Place the imaging sample onto our

laser scanning confocal microscope with the 25x, 1.05 NA, water immersion objective for

fluorescent imaging. Change the laser (405-, 488-, 561-, and 640-nm) and PMT pair to the

proper wavelength according to the targeted antibodies. Adjust focus position, laser power,

image acquisition time, and PMT gain according to real-time fluorescence signal to avoid

dim signal or over saturation

6.4 Acquire correlative SRS and fluorescent images by first performing volumetric

fluorescence imaging on immuno-labeled samples. Switch the microscope light path to IR

transparent condition. Open the shutter for SRS laser and perform SRS volumetric imaging

on the field of view of same sample with the same range of z.

Note: There is slight chromatic aberration that causes shift in z position between the

fluorescent images and the SRS images because of the wavelength difference in the

excitation lasers. Manual side-by-side comparison between the SRS and fluorescence

z-stack images is needed. We seek the exact matching features from both SRS and

fluorescence channel to match the z position precisely.

7. Construction, training, and validation of U-Net architecture

Note: Installing on Linux is recommended. An graphics card with >10GB of ram is required.

7.1 Setting up environment

7.1.1 Install Anaconda or Miniconda (we use 3-5.3.0-linux-x86_64).

7.1.2 Clone or download

https://github.com/Li-En-Good/VISTA

7.1.3 Create a Conda environment for the platform.

Command line:

conda env create -f environment.yml

7.2. Training the prediction model

7.2.1 Pair the directories of corresponding SRS images with ground truth images in a csv

file. Place the directories of SRS images under path_signal and ground truth images under

path_target columns.

7.2.2 Put the csv file in the folder data/csvs

Miao et al.

Page 7

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

7.2.3 Modify the configuration in scripts/train_model_2d.sh if needed

7.2.4 Activate the environment

Command line:

conda activate fnet

7.2.5 Initiate model training

Command line:

./scripts/train_model_2d.sh <file name of the csv file> 0

The training will then start. The losses for each iteration will be shown in the command line

and saved with the model in the folder save_models/<file name for the csv file>.

7.3. Validate the images in the training set and test set.

Command line:

./scripts/train_model_2d.sh <file name of the csv file> 0

The prediction results will be saved in the folder results/<file name of the csv file>.

8. VISTA combined with U-Net predictions enables protein specific multiplexity in label-

free images.

8.1 Modify the csv file in data/csvs/<file name of the csv file>/test.csv. Replace the

directories of both path_signal and path_target with the directory of new SRS images.

8.2 Remove the prediction results from the training, which is the folder results/<file name of

the csv file>.

8.3 Run predictions

Command line:

./scripts/train_model_2d.sh <file name of the csv file> 0

The prediction results will be saved in the folder results/<file name of the csv file>/test.

REPRESENTATIVE RESULTS:

After establishing the working principle of VISTA, we employed image registration to

evaluate the expansion ratio of VISTA and to ensure isotropic expansion during the sample

processing (Fig. 1a, b). We imaged both untreated and VISTA samples targeting bond

vibration at 2940 cm

−1

, which originates from CH

of endogenous proteins. In untreated

sample, the protein rich structures like nuclei are dark due to the overwhelming lipid

content from surrounding tissues

(Fig. 1a). After VISTA processing that includes the

delipidation treatment, the resulting image showed exactly the same feature with an inversed

contrast (Fig. 1b). The shapes and the relative positions of nuclei and vessels are completely

unaltered (Fig. 1a, b; numbered structures), confirming VISTA treatment is an isotropic

process. By comparing the sizes of the corresponding nuclei, we concluded that VISTA

achieves 3.4 times expansion in brain tissue samples

Knowing the expansion ratio in brain tissue, we can now resolve new features in our label-

free SRS images that are previously unresolvable. First, we showed that we could capture

Miao et al.

Page 8

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

features from healthy mouse cortexes down to 150 nm (Fig. 1c, d). Based on the dispersive

patterns around neuronal dendrites, the observed small structures are likely dendritic spine

heads

, which have a size of 146 nm (Fig. 1d). In addition, we applied VISTA to image

the fibrillar structures in A

plaques which are believed to have thickness around 100

. Indeed, we demonstrate that we could resolve ~ 130 nm fibrillar structures in a

representative diffusive A

plaque by VISTA (Fig. 1e, f).

As VISTA enables effective protein retention and protein imaging

, we can clearly

distinguish protein rich nucleoli in nuclei and the ribbon-like cytoskeleton structures in

the cytosols of cultured HeLa cells (Fig. 2a, arrowheaded). We further applied VISTA

to study poly-glutamine (polyQ) aggregates that are transiently expressed in mammalian

cell (Fig. 2b, c). We confirmed that the aggregates, as an expectedly densely packed

structure, were expanded isotropically by comparing the same aggregate structures before

and after expansion across multiple replicate samples

. We then indeed obtained high-

resolution structures that are absent/blurred in the normal-resolution SRS images. Our

VISTA-aggregates images revealed fibril-like protrusions on the peripheral of the polyQ

aggregates and with a hollow structure in the center (Fig. 2b, arrowheaded). The observation

that protrusions seamlessly attach to cytosolic contents might suggest that aggregates engage

functional proteins in cytosol. With hindsight, the capacity to expand dense aggregates also

becomes plausible because fixation reagents formaldehyde and hydrogel monomers acryl

amide, sodium acrylates are all small molecules that can diffuse in and out of protein

aggregates. Once the aggregate was co-polymerized with monomers into hydrogel, the

expansion process should proceed as normal.

We then applied VISTA to mouse brain tissue to further extend the scope of our method.

Although tissue samples pose challenges like reduced permeability, increased thickness,

and heterogeneous mechanical strength, we successfully imaged a mouse brain sample with

VISTA (Fig. 2d). Similar to cell samples, protein rich structures including cell nuclei, blood

vessels, and neuronal processes are observed (Fig. 2d, arrowheaded). The limitation of brain

tissues is that we only achieved 3.4-times expansion which makes the effective the resolution

of VISTA in brain samples to be 99 nm

. We validated the structural origin of SRS signals

by correlative dye and antibody staining, in which DAPI stains for nuclei and lectin stains

for blood vessels (Fig. 2e, f). Neuronal cell bodies and processes were also delineated

by immunofluorescence from NeuN and MAP2

. With the trained convolutional neural

network (CNN) algorithm utilizing the correlative fluorescence images as ground truth, the

single-channel VISTA images were then segmented into specific protein-structure channel

for multiplex VISTA images

Finally, we aim to interrogate pathologic A

plaques in brains of 5xFAD mice, a well-

known animal model for Alzheimer’s disease

. After standard VISTA procedures, we

acquired a 3-dimensional SRS image of amyloid plaques deposition in brain tissues (Fig.

2g). Puncta with high protein concentration was observed (Fig. 2g, orange arrowed),

representing the core of the A

plaque. Such image also reveals peripheral A

plaques (Fig.

2g, magenta arrowheaded), which is often neglected by conventional congo-red staining

that only targets the A

core. When combined with the trained segmentation algorithm, the

label-free VISTA image could be transformed into target-specific multiplex image (Fig. 2f)

Miao et al.

Page 9

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

and can be performed in joint with immunofluorescence

to study plaque-astrocyte and

plaque-microglia microenvironment interactions

, in a comprehensive and high-throughput

manner.

DISCUSSION:

In summary, we present the protocol for VISTA, which is a label-free modality to image

protein-rich cellular and subcellular structures of cells and tissues. By targeting endogenous

from proteins in hydrogel embedded cell and tissues, VISTA achieves an effective

imaging resolution down to 78 nm in biological samples and resolves minor extrusion

in Huntingtin aggregates and fibrils in A

plaques. VISTA is the first instance to report

sub-100 nm resolution for label-free imaging modalities

. Compared to existing expansion

methods

, VISTA inherits the merit of label-free SRS imaging and hence is free

from photobleaching, inactivation, or quenching caused by laser-illuminations. In addition,

as a label-free method, VISTA circumvents the demanding, inefficient, and potentially

artifact-causing antibody-labeling that is always involved in methods such as DISCO

and ExM

and thus offers high throughput sample preparation and uniform imaging

throughout tissues. To address the lack of multiplexity in the label-free approach, VISTA,

implemented with CNN-based image segmentation algorithm

, provides protein-specific

multi-component images without any labels in brain tissues

. We further applied VISTA

on 5xFAD mouse brains and enabled a holistic volumetric view of aggregates core and

periphery fibrils, nuclei and blood vessels

. We envision that VISTA would scale up well

for larger samples such as primate or human brain slices and could ultimately be useful for

clinical investigations.

There are three essential steps that ensure the successful implementation of VISTA

method. First, maximum protein retention in hydrogel sample hybrid is crucial and

required for VISTA

. To achieve this goal, we modified our fixation condition to contain

high concentration of acrylamide

and replaced the protein digestion procedure with

high-concentration detergent delipidation that saves significant protein loss from protein

digestion. The addition of AA quenches intermolecular crosslinking of proteins and

enables the isotropic expansion without protein digestions

. Second, proper correlations

between SRS and immuno-labeling and distinctions between different protein targets

need to be established. As VISTA relies on image-segmentation algorithms to add

multiplexity to monochromatic SRS images, crosstalk between different protein targets

in immunofluorescence will significantly compromise the quality of VISTA images. We

meticulously selected protein rich structures that are obvious in SRS images and validated

their corresponding immunofluorescence features. Third, before using the model to predict

fluorescence patterns from new SRS data sets, the validity and reliability of trained ML

model should be testified. Distinct features that are not included in the training sets will

likely cause issues in prediction. If the prediction results are not satisfying, the user should

try to include more data for training and avoid predicting patterns that are not included in

the training sets. Pearson's correlations of the testing sets and validation sets should also be

monitored to ensure the quality of the prediction

. We suggest the user to have at least

100 corresponding image sets for training.

Miao et al.

Page 10

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

While VISTA has immense potential for biological studies, there are certain limitations

awaiting creative solutions. First and foremost, the sensitivity of VISTA needs further

improvement. The detection limit of label-free simulated Raman scattering is in low

millimolar range and the isotropic expansion of samples in three dimensions significantly

dilutes chemical bonds and weakens VISTA signal. We hence are limited to imaging the

total ensemble of endogenous proteins that lacks specificity and multiplexity. Combining

VISTA with ultrasensitive SRS

, we could possibly extend VISTA to image low abundance

proteins and study aggregate structures and compositions at super-resolution level by

targeting orthogonal chemical bonds

. Second, the current 3.4 times expansion ratio in

brain tissues only gives moderate resolution improvement. Although we have already

resolved minor extrusions in A

plaques that are previously indistinguishable, higher

resolution is always desirable. In this case, innovations in protein-anchoring and hydrogel

chemistry would greatly benefit VISTA. For example, different gel formulation could

enable larger expansion ratios for even higher image resolution

. New procedures in

sample processing would allow VISTA to be applied with widely available FFPE histology

samples

, making VISTA even better suited for large-scale clinical studies.

Supplementary Material

Refer to Web version on PubMed Central for supplementary material.

ACKNOWLEDGMENTS:

We acknowledge the Caltech Biological Imaging Facility for software support. L.W. acknowledges the support of

the National Institutes of Health (NIH Director’s New Innovator Award, DP2 GM140919-01), Amgen (Amgen

Early Innovation Award), and the start-up funds from the California Institute of Technology.

REFERENCES:

1. Ntziachristos V Going Deeper than Microscopy: The Optical Imaging Frontier in Biology. Nature

Methods 2010, 7 (8), 603–614. [PubMed: 20676081]

2. Lavis LD; Raines RT Bright Ideas for Chemical Biology. ACS Chemical Biology 2008, 3 (3),

142–155. [PubMed: 18355003]

3. Tsien RY Constructing and Exploiting the Fluorescent Protein Paintbox (Nobel Lecture).

Angewandte Chemie International Edition 2009, 48 (31), 5612–5626. [PubMed: 19565590]

4. Huang B; Bates M; Zhuang X Super-Resolution Fluorescence Microscopy. Annual Review

Biochemistry 2009, 78, 993–1016.

5. Sahl SJ; Hell SW; Jakobs S Fluorescence Nanoscopy in Cell Biology. Nature Review Molecular

Cell Biology 2017, 18 (11), 685–701. [PubMed: 28875992]

6. Wassie AT; Zhao Y; Boyden ES Expansion Microscopy: Principles and Uses in Biological Research.

Nature Methods 2019, 16 (1), 33–41. [PubMed: 30573813]

7. Chen F; Tillberg PW; Boyden ES Expansion Microscopy. Science 2015, 347 (6221), 543–548.

[PubMed: 25592419]

8. Gambarotto D; Zwettler FU; Le Guennec M; Schmidt-Cernohorska M; Fortun D; Borgers S;

Heine J; Schloetel J-G; Reuss M; Unser M; Boyden ES; Sauer M; Hamel V; Guichard P Imaging

Cellular Ultrastructures Using Expansion Microscopy (U-ExM). Nature Methods 2019, 16 (1), 71–

74. [PubMed: 30559430]

9. Demchenko AP Photobleaching of Organic Fluorophores: Quantitative Characterization,

Mechanisms, Protection. Methods and Applications in Fluorescence. 2020, 8 (2), 022001. [PubMed:

32028269]

Miao et al.

Page 11

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

10. Murray E; Cho JH; Goodwin D; Ku T; Swaney J; Kim S-Y; Choi H; Park Y-G; Park J-Y; Hubbert

A; McCue M; Vassallo S; Bakh N; Frosch MP; Wedeen VJ; Seung HS; Chung K Simple, Scalable

Proteomic Imaging for High-Dimensional Profiling of Intact Systems. Cell 2015, 163 (6), 1500–

1514. [PubMed: 26638076]

11. Kim S-Y; Cho JH; Murray E; Bakh N; Choi H; Ohn K; Ruelas L; Hubbert A; McCue M; Vassallo

SL; Keller PJ; Chung K Stochastic Electrotransport Selectively Enhances the Transport of Highly

Electromobile Molecules. Proceedings of the National Academy of Sciences 2015, 112 (46),

E6274–E6283.

12. Liebmann T; Renier N; Bettayeb K; Greengard P; Tessier-Lavigne M; Flajolet M Three-

Dimensional Study of Alzheimer’s Disease Hallmarks Using the IDISCO Clearing Method. Cell

Report 2016, 16 (4), 1138–1152.

13. Min W; Freudiger CW; Lu S; Xie XS Coherent Nonlinear Optical Imaging: Beyond Fluorescence

Microscopy. Annual Review of Physical Chemistry 2011, 62 (1), 507–530.

14. Saar BG; Freudiger CW; Reichman J; Stanley CM; Holtom GR; Xie XS Video-Rate Molecular

Imaging in Vivo with Stimulated Raman Scattering. Science 2010, 330 (6009), 1368–1370.

[PubMed: 21127249]

15. Cheng J-X; Xie XS Vibrational Spectroscopic Imaging of Living Systems: An Emerging Platform

for Biology and Medicine. Science 2015, 350 (6264), 1054–1063

16. Ji M; Orringer DA; Freudiger CW; Ramkissoon S; Liu X; Lau D; Golby AJ; Norton I; Hayashi

M; Agar NYR; Young GS; Spino C; Santagata S; Camelo-Piragua S; Ligon KL; Sagher O; Xie

XS Rapid, Label-Free Detection of Brain Tumors with Stimulated Raman Scattering Microscopy.

Science Translational Medicine 2013, 5 (201), 201ra119–201ra119.

17. Wei M; Shi L; Shen Y; Zhao Z; Guzman A; Kaufman LJ; Wei L; Min W Volumetric Chemical

Imaging by Clearing-Enhanced Stimulated Raman Scattering Microscopy. Proceedings of the

National Academy of Sciences 2019, 116 (14), 6608–6617.

18. Ji M et al. Label-free imaging of amyloid plaques in Alzheimer’s disease with stimulated Raman

scattering microscopy. Science Advance 4, eaat7715 (2018).

19. Silva WR; Graefe CT; Frontiera RR Toward Label-Free Super-Resolution Microscopy. ACS

Photonics 2016, 3 (1), 79–86.

20. Gong L; Zheng W; Ma Y; Huang Z Higher-Order Coherent Anti-Stokes Raman Scattering

Microscopy Realizes Label-Free Super-Resolution Vibrational Imaging. Nature Photonics 2020,

14 (2), 115–122.

21. Watanabe K; Palonpon AF; Smith NI; Chiu L; Kasai A; Hashimoto H; Kawata S; Fujita K

Structured Line Illumination Raman Microscopy. Nature Communication 2015, 6 (1), 10095.

22. Qian C; Miao K; Lin L-E; Chen X; Du J; Wei L Super-Resolution Label-Free Volumetric

Vibrational Imaging. Nature Communication 2021, 12 (1), 3648.

23. Lin L-E; Miao K; Qian C; Wei L High Spatial-Resolution Imaging of Label-Free in Vivo Protein

Aggregates by VISTA. Analyst 2021, 146 (13), 4135–4145. [PubMed: 33949430]

24. Ounkomol C; Seshamani S; Maleckar MM; Collman F; Johnson GR Label-Free Prediction of

Three-Dimensional Fluorescence Images from Transmitted-Light Microscopy. Nature Methods

2018, 15 (11), 917–920. [PubMed: 30224672]

25. Falk T; Mai D; Bensch R; Çiçek Ö; Abdulkadir A; Marrakchi Y; Böhm A; Deubner J; Jäckel

Z; Seiwald K; Dovzhenko A; Tietz O; Dal Bosco C; Walsh S; Saltukoglu D; Tay TL; Prinz M;

Palme K; Simons M; Diester I; Brox T; Ronneberger O U-Net: Deep Learning for Cell Counting,

Detection, and Morphometry. Nature Methods 2019, 16 (1), 67–70. [PubMed: 30559429]

26. Yang B; Treweek JB; Kulkarni RP; Deverman BE; Chen C-K; Lubeck E; Shah S; Cai L; Gradinaru

V Single-Cell Phenotyping within Transparent Intact Tissue through Whole-Body Clearing. Cell

2014, 158 (4), 945–958. [PubMed: 25088144]

27. Mlodzianoski MJ; Cheng-Hathaway PJ; Bemiller SM; McCray TJ; Liu S; Miller DA; Lamb BT;

Landreth GE; Huang F Active PSF Shaping and Adaptive Optics Enable Volumetric Localization

Microscopy through Brain Sections. Nature Methods 2018, 15 (8), 583–586. [PubMed: 30013047]

28. Querol-Vilaseca M; Colom-Cadena M; Pegueroles J; Nuñez-Llaves R; Luque-Cabecerans J;

Muñoz-Llahuna L; Andilla J; Belbin O; Spires-Jones TL; Gelpi E; Clarimon J; Loza-Alvarez

Miao et al.

Page 12

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

P; Fortea J; Lleó A Nanoscale Structure of Amyloid-

Plaques in Alzheimer’s Disease. Scientific

Report 2019, 9 (1), 5181.

29. Oakley H; Cole SL; Logan S; Maus E; Shao P; Craft J; Guillozet-Bongaarts A; Ohno M; Disterhoft

J; Van Eldik L; Berry R; Vassar R Intraneuronal Beta-Amyloid Aggregates, Neurodegeneration,

and Neuron Loss in Transgenic Mice with Five Familial Alzheimer’s Disease Mutations: Potential

Factors in Amyloid Plaque Formation. Journal of Neuroscience 2006, 26 (40), 10129–10140.

[PubMed: 17021169]

30. Bartels T, Schepper SD and Hong S. Microglia modulate neurodegeneration in Alzheimer's and

Parkinson's diseases. Science. 370, 66–69 (2020). [PubMed: 33004513]

31. Chen F; Wassie AT; Cote AJ; Sinha A; Alon S; Asano S; Daugharthy ER; Chang J-B; Marblestone

A; Church GM; Raj A; Boyden ES Nanoscale Imaging of RNA with Expansion Microscopy.

Nature Methods 2016, 13 (8), 679–684. [PubMed: 27376770]

32. Tillberg PW; Chen F; Piatkevich KD; Zhao Y; Yu C.-C. (Jay); English BP; Gao L; Martorell A;

Suk H-J; Yoshida F; DeGennaro EM; Roossien DH; Gong G; Seneviratne U; Tannenbaum SR;

Desimone R; Cai D; Boyden ES Protein-Retention Expansion Microscopy of Cells and Tissues

Labeled Using Standard Fluorescent Proteins and Antibodies. Nature Biotechnol 2016, 34 (9),

987–992. [PubMed: 27376584]

33. Ku T; Swaney J; Park J-Y; Albanese A; Murray E; Cho JH; Park Y-G; Mangena V; Chen J; Chung

K Multiplexed and Scalable Super-Resolution Imaging of Three-Dimensional Protein Localization

in Size-Adjustable Tissues. Nature Biotechnol 2016, 34 (9), 973–981. [PubMed: 27454740]

34. Renier N; Wu Z; Simon DJ; Yang J; Ariel P; Tessier-Lavigne M IDISCO: A Simple, Rapid

Method to Immunolabel Large Tissue Samples for Volume Imaging. Cell 2014, 159 (4), 896–910.

[PubMed: 25417164]

35. Zhuge M; Huang K-C; Lee HJ; Jiang Y; Tan Y; Lin H; Dong P-T; Zhao G; Matei D; Yang Q;

Cheng J-X Ultrasensitive Vibrational Imaging of Retinoids by Visible Preresonance Stimulated

Raman Scattering Microscopy. Advanced Science 2021, 8 (9), 2003136. [PubMed: 33977045]

36. Miao K; Wei L Live-Cell Imaging and Quantification of PolyQ Aggregates by Stimulated Raman

Scattering of Selective Deuterium Labeling. ACS Central Science 2020, 6 (4), 478–486. [PubMed:

32341997]

37. Klimas A, et al. Nanoscale imaging of biomolecules using molecule anchorable gel-enabled

nanoscale in-situ fluorescence microscopy. Nature Portfolio, 10.21203/rs.3.rs-858006/v1, (2021).

38. Shi L et al. Super-resolution vibrational imaging using expansion stimulated

Raman scattering microscopy.

http://biorxiv.org/lookup/doi/10.1101/2021.12.22.473713

(2021)

doi:10.1101/2021.12.22.473713.

39. Zhao Y; Bucur O; Irshad H; Chen F; Weins A; Stancu AL; Oh E-Y; DiStasio M; Torous V; Glass

B; Stillman IE; Schnitt SJ; Beck AH; Boyden ES Nanoscale Imaging of Clinical Specimens Using

Pathology-Optimized Expansion Microscopy. Nature Biotechnol 2017, 35 (8), 757–764. [PubMed:

28714966]

40. M’Saad O; Kasula R; Kondratiuk I; Kidd P; Falahati H; Gentile JE; Niescier RF; Watters K;

Sterner RC; Lee S; Liu X; Camilli PD; Rothman JE; Koleske AJ; Biederer T; Bewersdorf J

All-Optical Visualization of Specific Molecules in the Ultrastructural Context of Brain Tissue.

bioRxiv April 5, 2022, p 2022.04.04.486901. 10.1101/2022.04.04.486901.

Miao et al.

Page 13

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript

Fig. 1: Sample expansion strategy enables super-resolution label-free imaging in mouse brain

tissues.

a) An SRS image at CH

frequency of mouse hippocampus. b) A VISTA image at the same

field of view of mouse hippocampus. Labeled area show the corresponding features before

and after VISTA treatments. c). A VISTA image of a normal mouse cortex that shows finer

features. Inset shows region of interest. d) Resolution quantification for the fine structure

observed in expanded samples. FWHM of 497 nm corresponds to effective resolution of 146

nm with 3.4 times expansion. e) A VISTA image of an amyloid beta plaque in mouse brain

tissues. Inset shows enlarged region of interest. f) Resolution quantification for the extrusion

fiber structure of the expanded amyloid beta plaque. FWHM of 442 nm corresponds to

effective resolution of 130 nm with 3.4 times expansion. Scale bars: 20 μm.

Miao et al.

Page 14

J Vis Exp

. Author manuscript; available in PMC 2022 October 10.

Author Manuscript