key: cord-0253141-wzm52sc8
authors: Rajagopalan, Anugraha; Venkatesh, Ishwarya; Aslam, Rabail; Kirchenbuechler, David; Khanna, Shreyaa; Cimbaluk, David; Gupta, Vineet
title: SeqStain using fluorescent-DNA conjugated antibodies allows efficient, multiplexed, spatialomic profiling of human and murine tissues
date: 2020-11-16
journal: bioRxiv
DOI: 10.1101/2020.11.16.385237
sha: 275e18b898859c4aab9dd15e249a61954f34a8df
doc_id: 253141
cord_uid: wzm52sc8

Spatial organization of molecules and cells in complex tissue microenvironments provides essential cues during healthy growth and in disease. Novel techniques are needed for elucidation of their spatial relationships and architecture. Although a few multiplex immunofluorescence based techniques have been developed for visualization of the spatial relationships of various molecules in cells and tissues, there remains a significant need for newer methods that are rapid, easy to adapt and are gentle during the cyclic steps of fluorescence staining and de-staining. Here, we describe a novel, multiplex immunofluorescence imaging method, termed SeqStain, that uses fluorescent-DNA labelled antibodies for immunofluorescence staining of cells and tissues, and nuclease treatment for de-staining that allows selective enzymatic removal of the fluorescent signal. SeqStain can be used with primary antibodies, secondary antibodies and antibody fragments, such as Fabs, to efficiently analyse complex cells and tissues in multiple rounds of staining and de-staining. Additionally, incorporation of specific endonuclease restriction sites in antibody labels allows for selective removal of fluorescent signals, while retaining other signals that can serve as marks for subsequent analyses. The application of SeqStain on human kidney tissue provided spatialomic profile of the organization of >25 markers in the kidney, highlighting it as a versatile, easy to use and gentle new technique for spatialomic analyses of complex microenvironments.

Understanding the molecular and cellular composition of tissues and their relative organization in three-dimensional space of a complex tissue microenvironment (spatialomic organization) is essential for obtaining fundamental insights in tissue biology, interplay between various molecules and cells and for more accurately determining changes due to disease or treatments. This is especially true in the case of cancer, where insights into the complex tumour microenvironment helps guide therapeutic choices (1) . Similar profiling of the cells infiltrating a transplanted tissue helps predict graft survival (2) .

Techniques to quantify multiple molecules in a single sample, such a transcriptomics, proteomics, flow cytometry and mass cytometry have revolutionized the field by providing deep characterization of tissue composition, yet they lack information about spatial organization of the various molecules and cells (3, 4) . Immunohistochemical (IHC) and immunofluorescence (IF) based imaging methods provide information about spatial distribution of molecules and cells in tissues, yet were limited to detecting only 3-6 analytes at a time in the past (5) (6) (7) (8) (9) . Thus, a great deal of recent effort has focused on increasing the multiplexing capabilities of such imaging-based techniques, and several methodologies for detecting multiple antigens in a single tissue section are now being developed (10) (11) (12) (13) (14) (15) (16) (17) (18) .

Given that methods using brightfield IHC carries a high translation potential (7) , techniques have been developed for detecting multiple antigens on a single tissue section by using cycles of IHC staining and de-staining using organic solvents (14, (19) (20) (21) . Similarly, methodologies using IF based imaging have been developed for multiplex imaging of tissues (15) (16) (17) (18) . For example, recently described CycIF, a cyclic immunofluorescence method, achieves multiplex staining of tissue sections by combining cycles of immunofluorescent staining with de-staining using fluorescence bleaching (17, 22) , whereas the 4i method achieves multiplex imaging by eluting antibodies after each round of staining (18) . These methods, while providing high multiplex IF capabilities, use harsh protocols to remove the stain in each cycle, which has the potential to harm sensitive tissues.

DNA tagged antibodies are also used widely in imaging and offer the combined benefits of the specificity of antibodies and the versatility of DNA oligonucleotides (23, 24) . Indeed, such reagents have redefined single-cell genomics by multiplexing cells from different samples (25, 26) . DNA tagged antibodies have also recently been used very elegantly for multiplexed IF imaging. The CODEX method combines DNA tagged antibodies and in-situ primer extension with fluorescently tagged nucleotides to achieve high-level multiplexing (15) . Although CODEX aids in deeper understanding of the tissue architecture, its complex experimental setup and design impedes wider accessibility of the technique. In contrast, Immuno-SABER uses cycles of in-situ DNA hybridization and removal (primer exchange reaction) for multiplexed IF imaging with DNA tagged antibodies (16) . However, it also requires a complex design of DNA oligonucleotides and set up for the exchange of DNA strands and concatemers. Thus, a need exists for rapid, mild, and easy to use method for multiplexed immunofluorescence imaging.

Endonucleases are enzymes that selectively cleave oligonucleotides either non-specifically (such as DNase I) or in a sequence specific fashion (such as restriction endonucleases). These enzymes are routinely used in molecular and cell biology laboratories (27) to cleave oligonucleotide sequences rapidly and under mild conditions that do not harm the cells and tissues. Here, we describe a multiplex immunofluorescence imaging technique, termed SeqStain (for Sequential

SeqStain Fabs were mixed with unlabelled primary antibodies at equal molar ratio (3:1 weight ratio) in PBS, pH 7.2, and incubated at room temperature for 2 hours. Any unbound, excess Fab was removed by filtration using 100 kDa molecular weight cut-off size exclusion Amicon ultra filtration columns (Millipore, # UFC510096). Subsequently, the complex was used for staining of cells and tissue sections.

Secondary antibodies were conjugated to the linker oligo using the maleimide-sulfhydryl chemistry and hybridized to the fluorescent oligo complex as detailed for primary antibodies. SeqStain secondary antibodies were mixed with unlabelled primary antibodies at 2:1 molar ratio at 0.1uM concentration with respect to the primary antibody to avoid formation of superclusters of primary and secondary antibodies. This reaction was incubated at room temperature for 15 minutes. Excess secondary antibody was then blocked by adding the corresponding normal serum at 1:20 v/v ratio and the mixture was incubated for 5 minutes at room temperature (Thermo Fisher Scientific, #10410 and #10510). The complex was used immediately for staining of cells and tissue sections.

To determine linker oligo conjugation efficiency of antibodies, the reaction mixture was analysed using 10% SDS PAGE gel (Thermo Fisher Scientific, NW04122BOX) following manufacturer's instructions.

Controls, including unmodified antibodies and a protein ladder (Bio-rad, #1610375) were also used.

Subsequently, the gel was Coomassie stained (Thermo Fisher Scientific, #24617) to visualize the protein bands. To determine efficiency of hybridization reaction, the annealing reaction mixture was analysed using a 4% agarose gel containing a DNA stain (Thermo Fisher Scientific, #S33102). The bands were visualized using a UV transilluminator (Bio-Rad laboratories, Hercules, California)

All animal studies were performed in compliance with the Institutional Animal Care and Use Committee (IACUC) at Rush University Medical Center. Murine spleen tissues were from mice bearing orthotopically transplanted LLC cells, as previously described (34) . Briefly, LLC cells were passaged at least 3 times before inoculating into mice. Cells were routinely checked for mycoplasma using the MycoAlert assay (Lonza) and found negative prior to their use in animals. Wild type C57Bl/6 mice (6-8 week old) were obtained from The Jackson Laboratory. LLC cells (1.0 x 10 6 ) in 100 μL cold phosphate-buffered saline (PBS) were subcutaneously inoculated into the mouse rear flank. Once the tumours grew to 1000mm 3 in volume, the animals were sacrificed, and the harvested spleen tissue was immediately embedded in Tissue Tex OCT. Embedded tissue was cryopreserved in liquid nitrogen and transferred to -80˚C for long term storage. Cryopreserved normal human kidney and tonsil tissue blocks were purchased from Origene (Rockville, MD).

Single cell suspensions of cells in culture were prepared in PBS containing 1% BSA. Non-specific antibody recognition sites on cells were blocked by adding Fc Block (Biolegend, #101301) to the suspension for ten minutes at 4°C. Next, cells were stained with either unlabelled primary antibodies (conventional staining) or with SeqStain antibodies for 20 min at 4°C. Subsequently, cells were washed with PBS containing 1% BSA and analysed using LSR-Fortessa flow cytometer. Data was analysed using FlowJo software version 10.2.

Buffers and reagents. The following buffers and reagents were used in the assays. containing the cells or tissue sections were prepared as described before and mounted onto the perfusion chamber. The mounted samples were stained with primary antibodies after blocking with block-1 solution for 1 hour at room temperature. The chamber was perfused with wash buffer for 5 minutes to remove non-specific staining and then incubated with the corresponding secondary antibodies and stained at room temperature for 30 minutes. The samples were washed by perfusing wash buffer for 5 minutes and imaged after nuclear staining with DAPI.

Perfusion set-up. Cover glass was mounted on the closed bath chamber (Warner instruments, #RC-43C) following manufacturer's instructions. This, in-turn, was mounted on the quick exchange platform (Warner instruments, #64-0375 (QE-1)) with built-in perfusion and suction holders. Buffers and solutions were perfused in and suctioned out via the inlet and outlet ports, respectively. A stage adapter (Warner instruments, #64-2415) was used to place the platform on the microscope to perform the iterative rounds of staining and de-staining.

Image acquisition. Immunofluorescence images were acquired using the Zeiss 700 LSM confocal microscope and Zen software (Carl Zeiss Group, Hartford, Connecticut). The destaining videos were acquired using the time-lapse image acquisition option in the Zen software. DNase I was added after the first image was acquired (-20 sec), following which images were acquired every 20 seconds for 5 minutes to monitor the rate of destaining. For the whole slide imaging, the images were acquired using the Nikon Eclipse TI2-E inverted confocal microscope (Nikon Corporation, Tokyo, Japan) with a motorized stage. The image tiles were acquired with the perfect focus system (PFS) to correct for any focal drifts between the tiles and then processed with the NIS Elements software.

Image analysis. Cell profiler version 3.1.9 was used to identify individual cells and their nuclei for both immunofluorescent staining of cells and tissue sections according to published protocols (35) (36) (37) .

Briefly, a working pipeline was created to run the required analysis. A set of modules were designed to identify objects and their relative intensities. For each image set, nuclei were considered as starting points and was identified using identify primary objects module. Identify secondary objects module was used to identify objects such as cells based on objects identified in previous module. Measure object intensity module was used to measure the intensity of each identified object. Overlay Object module created a color-coded label of the previously identified objects. Overlay Outline was used to place outlines of an object on a desired image as to create segmentation in the cellular region.

Intensity measured was exported to excel files for further analysis.

For the whole slide images obtained from the multiplex imaging experiment, to align the images, DAPI images from each round of staining and de-staining were rigid registered (translation + rotation) to each other using the Register Virtual Stack Slices plugin in Fiji (38) . For the other channels in each round, the same transformation as the corresponding DAPI channel was applied with the Transform Virtual Stack Slices plugin in Fiji. To measure distances of various cell types from the B-cell cluster, a region of interest was selected, and the cluster was traced manually using the signal from the DAPI and B220 channels. Euclidean Distance Map to the boarder of the cluster was calculated with Fiji. To measure signal intensities of each channel, the nuclei were first segmented using the DAPI channel. Segmentation was performed using the StarDist plugin, which uses an already trained model for convolution neural network (DSB 2018 dataset) (39) . Once segmented, intensities for each channel was calculated on a per cell basis. Alongside, the EDM measurements for each segmented cell was calculated. In order to generate the plot profiles of the same markers from distinct rounds, first a stack of the corresponding images was generated. From the stack, a random region was cropped, and stack was converted to individual images. Plot profiles were calculated by drawing an arbitrary yellow line at the same position in both the images using Fiji. The overlay plots were generated using GraphPAD PRISM 9.

Statistical analyses were performed using Excel (Microsoft, Bothell, WA) and Prism 8.2 software (GraphPad Software). Student's t-test were utilized when the data were normally distributed. Mann-Whitney test was used for comparison between two groups when data was not normally distributed.

A sequential combination of staining with fluorescent-DNA conjugated antibodies and destaining with nuclease provides a novel, easy-to-use, multiplexed imaging platform -SeqStain DNA-tagged antibodies have recently been utilized in various techniques to perform multiplexed immunofluorescence imaging of cells and tissues (15-17, 22, 40, 41) . These techniques rely on different methods for applying and removing fluorescent tags from the antibodies immobilized on substrates in an antigen specific fashion (15, 16, 42, 43) . However, methods for cyclical removing or quenching of fluorescent labels and application of new ones is currently cumbersome and laborious, which also necessitates specialized equipment and setup. There is a current need for techniques that can perform rapid and selective cleavage of fluorescent labels from antibodies, that are also easy to use in a standard laboratory setting. Here, we describe an easy to use, rapid, multiplex imaging technique (termed "SeqStain") that uses antibodies conjugated with fluorescently labelled DNA oligonucleotides (termed "SeqStain antibodies"). The method relies on sequential steps of immunofluorescent labelling with a set of such antibodies, gentle removal of the fluorescent labels post imaging, followed by another round of labelling with a new set of fluorescently labelled antibodies ( Figure 1A) . Importantly, the SeqStain methodology utilizes efficient and selective enzymatic processes, such as treatment with DNase I or restriction enzymes, for rapidly cleaving fluorescently labelled oligonucleotides (oligos) off of the antibodies, thereby removing the fluorescent labels from immune-fluorescently labelled substrates in the de-staining steps. Subsequently, the cleaved labels are washed off prior to initiation of the next round of staining. The enzymatic removal of fluorescent oligonucleotides also offers flexibility in the oligo sequence design, the length and complexity of the oligos and the types of oligos that can be used. Thus, cyclic steps of staining tissues with SeqStain antibodies and de-staining with nuclease treatment allows for efficient and rapid multiplex immunofluorescent imaging of cells and tissues.

SeqStain antibodies were designed such that they could carry multiple fluorophores on each of the conjugated oligonucleotides (oligos) on the antibody ( Figure S1A) . Additionally, the oligo chains were designed so that the fluorescent dye molecules on the DNA were spaced apart by >5 nm dye-to-dye distance (at least 15 nucleotides apart) to prevent any unwanted dye-dye interactions (such as selfquenching) (44) . SeqStain antibodies were generated in a two-step process ( Figure S1B ). In step 1, a 5'-amine modified 29-nucleotide long single-stranded DNA chain (linker oligo), containing a prespecified 15-mer sequence for hybridization with a complementary docking oligo, was covalently attached to the antibody using published protocols and purified (23, 31, 45, 46) . Conjugation efficiency was determined using SDS-PAGE gels, which showed an average of 2-5 linker oligos chemically conjugated to an antibody during a typical labelling experiment ( Figure S1C ). In step 2, a fluorescently labelled double-stranded DNA (dsDNA) complex, containing a docking oligo and multiple fluorescently labelled oligos was hybridized to the antibody-DNA conjugate. This design of the dsDNA also allows for using many copies of fluorescently labelled DNA oligos to tune the signal intensity on the antibody. Subsequently, fluorescent-DNA labelled antibodies (SeqStain antibodies) were purified using the 100 kDa size-exclusion column. Formation of the fluorescent-DNA antibody complex was confirmed using 4% agarose gels ( Figure S1D ). This method of using fluorophores on the complementary oligonucleotides also avoids chemical quenching of the fluorophores during antibody-DNA conjugation step.

Flow cytometric analyses showed comparable staining of cell surface proteins CD45 and Figure S3) . Furthermore, as chemical conjugation of linker oligo to antibody can be performed using a variety of methods, we tested two additional orthogonal chemistries to generate SeqStain antibodies (Figures S4A-S4B ) (31, 47) . We found that these different methods had no effect on the immunofluorescent staining and de-staining steps of SeqStain ( Figures S4C-S4D) , suggesting that modification of antibodies with fluorescent dsDNA can be successfully accomplished using multiple different methods.

Finally, to evaluate SeqStain technique for multiplex staining, a 6-plex panel with a combination of immune markers and structural markers was used to iteratively stain and de-stain the immobilized RAW cells in 3 rounds of staining and de-staining ( Figure 1E) . Staining was performed with a set of two unique antibodies, each with a unique fluorophore-labelled DNA tag, per round. Cell nuclei were stained using DAPI, to visualize nuclear boundary of each cells and to count cells.

Fluorescent signal from the antibodies was removed using DNase I. Quantification showed complete removal of both fluorophores after each round, with no effect on DAPI staining, suggesting that this treatment does not affect the integrity of cells, nuclei or nuclear materials. A concern for the antibody based multiplex staining methods is that application of so many antibodies, either together or sequentially, could crowd the antigens, making multiplexing difficult. To address this concern in SeqStain, whether earlier round of staining may mask the antigens for probing at the later rounds, we stained and de-stained immobilized cells with the same SeqStain antibody for five cycles ( Figure S5) and found no loss in immunofluorescent labelling capability of these samples or sample integrity.

Moreover, the samples were further stained, de-stained and re-stained with a second set of two antibodies for five more rounds, without any loss of labelling capability or sample integrity. Together, these data show that SeqStain offers a rapid and robust multiplex immunofluorescence imaging platform.

Conventional immunofluorescence imaging methods utilize antigen recognition using an unlabelled primary antibody followed by antibody recognition using a fluorescently labelled secondary antibody.

However, extension of such methodology in a multiplex environment is significantly limited by the number of species (such as mouse, rat, goat, sheep) available to generate unique secondary antibodies. Pre-mixing primary antibodies with fluorescently labelled secondary antibodies or purified Fabs (the regions of antibodies with antigen binding capacity (48, 49) ) prior to their use in staining could potentially overcome some of these limitations. Here, as an alternative to using primary SeqStain antibodies for multiplex staining, we tested if fluorescent-DNA labelled secondary antibodies (SeqStain secondary Ab) or Fabs of secondary antibodies (SeqStain Fabs) (Figure 2A-2B ) could be applied in the SeqStain protocol. Such reagents have at least two major advantages. One, by precomplexing these reagents with primary antibodies, we can achieve signal amplification for targets with low level of expression. Two, these reagents may be quickly pre-complexed with many different primary antibodies, thus circumventing the need for modifying primary antibodies with fluorescent-DNA.

To test, we prepared SeqStain Fabs using commercially available secondary antibodies or with affinity purified Fc-specific Fab fragments from secondary antibodies (mouse, rat and rabbit) ( Figure S6 ). Primary rabbit anti-mouse antibody against a-Tubulin was pre-complexed with anti-rabbit SeqStain secondary Ab or with anti-rabbit secondary SeqStain Fab, respectively. Next, we stained immobilized mouse podocytes (Figure 2C ) or HeLa cells (Figure 2D ) with them. Immunofluorescence imaging showed high level of staining similar to the staining with conventional primary-secondary antibodies. Expectedly, treatment of these SeqStain secondary Ab or the SeqStain Fab stained cells with nuclease DNase I resulted in rapid loss of the fluorescence signal. As a control, staining with SeqStain Secondary Ab or SeqStain Fab alone, in the absence of the primary rabbit anti-mouse antibody against a-Tubulin did not show any staining, highlighting the specificity of this approach.

Furthermore, use of a 6-plex panel of primary antibodies pre-complexed with SeqStain Fabs to iteratively stain and de-stain immobilized RAW cells also showed efficient labelling and clearing of fluorescent signal from these cells (Figure S7 ), further suggesting that SeqStain methodology can be applied with a variety of reagents for rapid multiplex immunofluorescence imaging.

To test the feasibility of the SeqStain approach on complex tissues, we tested it on murine (spleen) and human tissues (kidney, tonsil) (details in Methods and Materials section). We selected a set of commercially available, well-characterized antibodies ( Table S1 ) that recognize several different cell types for these assays. Figures 3D and 3E . Furthermore, we found that the type of fluorophore used to label the dsDNA complex on SeqStain antibodies did not have any material effect on their performance, suggesting that most commonly available fluorophores can be used to fluorescently tag the DNA on the antibodies ( Figure S8 ). Together, these data suggest that SeqStain antibodies and Fabs are highly efficient in immunostaining of a variety of complex tissues and provide staining results like those obtained with conventional immunofluorescence staining methods.

Subsequently, we evaluated the feasibility of using enzymatic approach for removing the DNA-linked fluorophore in tissues immunostained with SeqStain antibodies. Surprisingly, we found that nucleases were equally efficient in removing the fluorescent signal from SeqStain antibody stained tissue sections as they were with immobilized cells, even though the tissues provide a highly complex microenvironment. Human kidney tissue (, Figure 4) 

To test spatialomic profiling via multiplex immunofluorescence staining of a single tissue section using SeqStain, in a proof-of-concept experiment, we developed a panel of nine unique antibodies against various immune markers, along with pan-nuclear marker DAPI, and used it in five rounds of sequential staining and de-staining steps on mouse spleen tissue. Each round of staining used two unique SeqStain antibodies followed by imaging of the whole slide and de-staining with nuclease DNase I. Again, DAPI-stained nuclei provided a guidepost for aligning the various image sets at the end of the experiment. Results in Figure 6A show high level of fluorescence staining by each SeqStain antibody, followed by complete and rapid removal of fluorescent label from both channels after DNase I treatment ( Figure S13 ). After completion of the imaging rounds, images were stacked and aligned using DAPI stained nuclei using ImageJ (38, 54) . Cell Profiler based image analyses showed that the individual cells can be efficiently segmented computationally for cell-based fluorescent signal analyses of SeqStain stained tissues ( Figure S14 ).

Subsequently, select composites were generated from the aligned images to show the spatial organization of different markers in the spleen tissue (Figure 6B-6D) . Images of the entire tissue section clearly showed no changes in overall tissue morphology or integrity during due to the repeated cycles of staining and de-staining. Staining of different myeloid cells using CD169, CD68

and CD11b revealed their spatial relationship with respect to B220+ B cells. Expectedly, the CD169+ macrophages were found in the marginal zone area lining B220+ B cell clusters, in close contact with the B-cells as is typical for these cell population whereas the CD68+ macrophages were predominantly found in the red pulp region of the spleen. Additionally, the CD11b+ cells were found scattered outside the red pulp region (53, 55, 56) . Thus SeqStain can be used to profile the heterogenous myeloid populations in the spleen whose distinct spatial location and relationship with other cell types help orchestrate the immune response during an infection (57, 58) .

We also characterized the MHC II expression and on various cell types in the spleen tissue and their respective spatial relationship. Quantification of MHC II levels revealed that CD68+ and CD11b+ macrophages in the spleen have low levels of MHC II expression (59) . While CD169+ macrophages residing in the marginal zone area had higher expression of MHC II (60, 61) .

Expectedly, B220+ B cells had the highest expression of MHC II among the cell types analysed (62, 63) . Thus, by quantifying the relative spatial location of cells and their co-expression of markers, SeqStain offers a simple yet robust method to understand the spatial organization of various cell types in the tissue and an ability to generate spatial relationship maps (SRMs).

Next, we expanded antibody set to profile human kidney tissue and also tested the feasibility of using three-color antibody mixtures for staining in each round of SeqStain. We developed a 20color panel (19 unique antibodies + DAPI) and used it to stain a human kidney tissue section in 9

cycles of staining and de-staining (Figures 7 and S15) . This time we used three unique SeqStain antibodies in each round of staining, followed by imaging of the whole tissue section, de-staining with nuclease DNase I and re-imaging. Additionally, to confirm staining, we used a few of the antibodies twice, to obtain a 25-plex image of the kidney tissue at high resolution ( Figure S15) . As above, images were stacked and aligned using DAPI stained nuclei in ImageJ for data quantification (38, 54) . Figure 8A ) (64, 65) . We were also able to identify the collecting duct (AQP2+ AQP3+ EpCAM+) (Cyto7+ Cyto8+) (Figure 8A and 8C ) (66) (67) (68) , Distal convoluted tubules (AQP2-AQP3-EpCAM+), Thick Ascending Loop of Henle's (EpCAM+ Uromodulin+) and their spatial location with respect to one another ( Figure 8A ) (69) . Similarly, in the glomerulus, we were able to discern the three components of the glomerular capillary filter-podocytes, glomerular basement membrane and glomerular endothelial cells (69) . Podocytes can be identified by the co-expression of WT1 and Vimentin, the glomerular basement membrane can be visualized by Collagen IV staining and the glomerular endothelial cells by staining for CD31 ( Figure 8B ). In addition, the specialized contractile mesenchymal cells of the glomerulus can be discerned by Vimentin staining in the absence of WT1

staining. In the renal interstitium (70) (71) (72) (73) , Collagen IV delineates the tubular basement membrane forming a network throughout the entire tissue (74) . Peritubular capillaries visualized by CD31 staining can be seen as a disconnected network of cells within this tubular basement membrane (75) . The renal interstitium also contains stromal cells which can be visualized by Vimentin and a-SMA staining ( Figure 8B ) (76) . Angiotensin Converting Enzyme (ACE2), the functional receptor for the SARS coronaviruses, was found to be highly expressed in the proximal tubular cells identified by ACE2 coexpression with AQP1 (77, 78) . The normal human kidney had resident immune cells which were identified by CD45 staining while the renal macrophages were identified by co-expression of CD45 and CD68 ( Figure 8C ) (79) . In summary, these experiments show that SeqStain is highly applicable for multiplexed spatial profiling of various types of tissues and allows for rapid generation of spatial relationship maps.

These data also suggest that, although the experiments here utilized either a 10-plex or a 25plex SeqStain panel to evaluate multiplexed staining, the method is easily scalable to tens of different markers. Signal amplification, if necessary for markers expressed at low levels, can be done using a 

Spatial profiling of cells in tissues provides critical insights into disease pathogenesis and can be diagnostic. The importance of such insights is appreciated especially for diseases like cancer where spatial heterogeneity often leads to poor clinical outcomes (1, 85, 86) . Indeed, to address such needs, the recent years have seen a surge of innovative multiplex staining techniques at both transcriptomic and proteomic levels (10-19, 87, 88) . However, the existing protein multiplexing methods either use harsh de-staining conditions or require complex experimental set-up. The SeqStain spatialomic analysis methodology presented here is a gentle, easy-to-use, and efficient multiplex imaging technique that provides a unique platform for obtaining such spatialomic insights. The method applies fluorescent DNA-labelled antibodies with gentle, enzymatic method for removing fluorescence signal after each cycle of staining. In particular, we demonstrate multiplex staining of immobilized cells and tissue sections, where de-staining gently removes the fluorescent signal to pre-staining levels on the whole slide. Strikingly, de-staining using the SeqStain method was also rapid. We observed that treatment with nucleases removed >99% of the signal in <1 minute, without affecting sample integrity or tissue morphology. Additionally, by engineering specific restriction sites into DNA during antibody modification, we show that selective de-staining is possible with SeqStain. Retention of selective markers for subsequent rounds may be important for spatial aligning of tissues when it is not possible to perform whole slide scanning or for measuring information about multiple neighbouring antigens via FRET or other such methods, where keeping fluorophore fixed on one antigen might be helpful, while changing the second or third fluorophores on different antigens. SeqStain is thus a highly configurable, multiplex staining method which uses the familiar laboratory essentials such as the endonucleases system to achieve rapid multiplexing of cells and tissues. Furthermore, the enzymatic approach for the removal of fluorescent labels offers flexibility in the design of oligo sequence used for conjugation, the length and complexity of the oligos, the types of fluorophores that can be included, and the types of oligo-based higher order structures that can be used.

The technique offers significant new advantages for deeper understanding of complex tissue microenvironments. This novel methodology uses commercially available primary and secondary antibodies and Fab fragments that are chemically conjugated with fluorescently labelled doublestranded DNA (SeqStain antibodies and Fabs) that are easy to modify in any laboratory. We also show that such modifications do not affect their function in any way by testing in a variety of systems.

To avoid cross-reactivity, pre-complexation of SeqStain Fabs or SeqStain secondary antibodies with primary antibodies can also be used during staining. This paves way to build a multiplex panel that in each panel is also presented. Graphs show the mean ± standard deviation. Immunofluorescence images showing human kidney tissue sections stained using SeqStain antibodies (as indicated in the panel). The antibodies were labelled using either the AF488 fluorophore (shown in green), the Cy3 fluorophore (shown in red) or the Cy5 fluorophore (shown in yellow). Immunofluorescence images of these tissue sections after de-staining with DNase I treatment are shown below each panel. All images are representative of at least three replicates. Scale bar is 100µm. Graphs showing quantification of fluorescence intensity after staining (red bars, green bars or yellow bars) and de-staining (brown bars) in each panel is also presented on the right. Graphs show the mean ± standard deviation. 

Understanding the tumor immune microenvironment (TIME) for effective therapy

Multiplexed mass cytometry profiling of cellular states perturbed by small-molecule regulators

A deep profiler's guide to cytometry

Layered peptide array for multiplex immunohistochemistry

Aqueous two-phase system-mediated antibody micropatterning enables multiplexed immunostaining of cell monolayers and tissues

Patterns of cancer invasion revealed by QDs-based quantitative multiplexed imaging of tumor microenvironment

Simultaneous visualization of multiple antigens with tyramide signal amplification using antibodies from the same species

Ultra-fast and automated immunohistofluorescent multistaining using a microfluidic tissue processor

Multiplexed immunohistochemistry, imaging, and quantitation: A review, with an assessment of Tyramide signal amplification, multispectral imaging and multiplex analysis

Recent developments in multiplexing techniques for immunohistochemistry

Multispectral imaging of clinically relevant cellular targets in tonsil and lymphoid tissue using semiconductor quantum dots

Highly multiplexed single-cell analysis of formalinfixed, paraffinembedded cancer tissue

In-depth tissue profiling using multiplexed immunohistochemical consecutive staining on single slide

Deep Profiling of Mouse Splenic Architecture with CODEX Multiplexed Imaging

Immuno-SABER enables highly multiplexed and amplified protein imaging in tissues

Highly multiplexed imaging of single cells using a high-throughput cyclic immunofluorescence method

Multiplexed protein maps link subcellular organization to cellular states

Multiplexed Immunohistochemical Consecutive Staining on Single Slide (MICSSS): Multiplexed Chromogenic IHC Assay for High-Dimensional Tissue Analysis

A colorful future of quantitative pathology: Validation of Vectra technology using chromogenic multiplexed immunohistochemistry and prostate tissue microarrays

SIMPLE: A sequential immunoperoxidase labeling and erasing method

Highly multiplexed immunofluorescence imaging of human tissues and tumors using t-CyCIF and conventional optical microscopes

Sitespecific DNA-antibody conjugates for specific and sensitive immuno-PCR

Immuno-PCR: Very sensitive antigen detection by means of specific antibody-DNA conjugates

Cell Hashing with barcoded antibodies enables multiplexing and doublet detection for single cell genomics

Nuclei multiplexing with barcoded antibodies for single-nucleus genomics

How restriction enzymes became the workhorses of molecular biology

Small molecule-mediated activation of the integrin CD11b/CD18 reduces inflammatory disease

Podocytes in culture: Past, present, and future. Kidney Int

Synthesis of peptide-oligonucleotide conjugates using a heterobifunctional crosslinker

Simple Method to Prepare Oligonucleotide-Conjugated Antibodies and Its Application in Multiplex Protein Detection in Single Cells

Addition of Sulfhydryl Groups to Escherichia coli Ribosomes by Protein Modification with 2-Iminothiolane (Methyl 4-Mercaptobutyrimidate)

Potent and specific antitumor effects of an anti-CD22-targeted cytotoxic ribonuclease: Potential for the treatment of non-Hodgkin lymphoma

Integrin CD11b activation drives anti-tumor innate immunity

CellProfiler: Image analysis software for identifying and quantifying cell phenotypes

CellProfiler TM : Free, versatile software for automated biological image analysis

CellProfiler 3.0: Next-generation image processing for biology

Consistent and elastic registration of histological sections using vector-spline regularization

Cell detection with star-convex polygons

DNA-barcoded labeling probes for highly multiplexed Exchange-PAINT imaging

High-Plex Spatially Resolved RNA and Protein Detection Using Digital Spatial Profiling: A Technology Designed for Immuno-oncology Biomarker Discovery and Translational Research

Superresolution microscopy with DNA-PAINT

Multiplexed 3D cellular super-resolution imaging with DNA-PAINT and Exchange-PAINT

Sub-100-nm metafluorophores with digitally tunable optical properties self-assembled from

Oligonucleotide conjugates for therapeutic applications

Antibody-Oligonucleotide Conjugates as Therapeutic, Imaging, and Detection Agents

Simultaneous epitope and transcriptome measurement in single cells

Simplified preparation of rabbit fab fragments

A new enzymatic method to obtain high-yield F(ab)2 suitable for clinical use from mouse IgGl

Sphingosine-1-phosphate receptors control B-cell migration through signaling components associated with primary immunodeficiencies, chronic lymphocytic leukemia, and multiple sclerosis

B cell regulation in cancer and anti-tumor immunity

CD83 influences cell-surface MHC class II expression on B cells and other antigen-presenting cells

Splenic macrophage subsets and their function during blood-borne infections

NIH Image to ImageJ: 25 years of image analysis

Identification, activation, and selective in vivo ablation of mouse NK cells via NKp46

Experimental stroke differentially affects discrete subpopulations of splenic macrophages

Structure and function of the spleen

Protective and pathogenic functions of macrophage subsets

Most Tissue-Resident Macrophages Except Microglia Are Derived from Fetal Hematopoietic Stem Cells

CD169 + macrophages at the crossroads of antigen presentation

Antigen targeting reveals splenic CD169 + macrophages as promoters of germinal center B-cell responses

Boosting the MHC class II-restricted tumor antigen presentation to CD4+ T helper cells: A critical issue for triggering protective immunity and re-orienting the tumor microenvironment toward an anti-tumor state

Splenic CD19-CD35+B220+ cells function as an inducer of follicular dendritic cell network formation

Development of a 2-dimensional atlas of the human kidney with imaging mass cytometry

Aquaporins in the kidney: From molecules to medicine

Spatial and temporal expression pattern of N-cadherin cell adhesion molecules correlated with morphogenetic processes of chicken embryos

Distribution of cytokeratins and vimentin in adult renal neoplasms and normal renal tissue: Potential utility of a cytokeratin antibody panel in the differential diagnosis of renal tumors

The Urinary Excretion of Uromodulin is Regulated by the Potassium Channel ROMK

What is the glomerular ultrafiltration barrier?

WT1 regulates the expression of the major glomerular podocyte membrane protein Podocalyxin

Expression of smooth muscle cell phenotype by rat mesangial cells in immune complex nephritis: α-Smooth muscle actin is a marker of mesangial cell proliferation

Glomerular endothelial mitochondrial dysfunction is essential and characteristic of diabetic kidney disease susceptibility

The glomerular basement membrane

Renal basement membrane components

Regardless of etiology, progressive renal disease causes ultrastructural and functional alterations of peritubular capillaries

The renal (myo-)fibroblast: A heterogeneous group of cells

Renal ACE2 expression in human kidney disease

Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first step in understanding SARS pathogenesis

Specific macrophage subtypes influence the progression of rhabdomyolysis-induced kidney injury

Rolling circle amplification: A versatile tool for chemical biology, materials science and medicine

Specific Synthesis of DNA in Vitro via a Polymerase-Catalyzed Chain Reaction

Enzymatic signal amplification of molecular beacons for sensitive DNA detection

A branched DNA signal amplification assay for quantification of nucleic acid targets below 100 molecules/ml

2020) A 6-nm ultra-photostable DNA FluoroCube for fluorescence imaging

Tumour heterogeneity and resistance to cancer therapies

Quantitative spatial profiling of PD-1/PD-L1 interaction and HLA-DR/IDO-1 predicts improved outcomes of anti-PD-1 therapies in metastatic melanoma

Spatial reconstruction of single-cell gene expression data

Single-cell spatial reconstruction reveals global division of labour in the mammalian liver