Analysis Of The Proximal Tubular And Glomerulus Of The Kidney
Mar 15, 2022
Contact: joanna.jia@wecistanche.com / WhatsApp: 008618081934791
Click here for information about Part II (Introduction, materials, and methods )of this article.
Near-Single-Cell Proteomics Profiling of the Proximal Tubular and Glomerulus of the Normal Human Kidney
Tara K. Sigdel1, Paul D. Piehowski2, Sudeshna Roy1, Juliane Liberto1, Joshua R. Hansen2, Adam C. Swensen2, Rui Zhao3, Ying Zhu3, Priyanka Rashmi1, Andrew Schroeder1, Izabella Damm1, Swastika Sur1, Jinghui Luo4, Yingbao Yang4, Wei-Jun Qian2* and Minnie M. Sarwal1* for the Kidney Precision Medicine Project (KPMP) Consortium
1Division of MultiOrgan Transplantation, Department of Surgery, University of California, San Francisco, San Francisco, CA, United States
2Pacific Northwest National Laboratory, Biological Sciences Division, Richland, WA, United States
3Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, WA, United States
4Department of Pathology, University of Michigan, Ann Arbor, MI, United States
Keywords: glomerulus, mass spectrometry, single-cell analysis, proteomics, kidney
Molecular assessments at the single-cell level can accelerate biological research by providing detailed assessments of cellular organization and tissue heterogeneity in both disease and health. The human kidney has complex multi-cellular states with varying functionality, much of which can now be completely harnessed with recent technological advances in tissue proteomics at a near single-cell level. We discuss the foundational steps in the first application of this mass spectrometry (MS) based proteomics method for analysis of sub-sections of the normal human kidney, as part of the Kidney Precision Medicine Project (KPMP). Using ~30–40 laser captured micro-dissected kidney cells, we identified more than 2,500 human proteins, with specificity to the proximal tubular (PT; n = 25 proteins) and glomerular (Glom; n = 67 proteins) regions of the kidney and their unique metabolic functions. This pilot study provides the roadmap for application of our near-single-cell proteomics workflow for the analysis of other renal micro-compartments, on a larger scale, to unravel perturbations of renal sub-cellular function in the normal kidney as well as different etiologies of acute and chronic kidney disease.
cistanche can improve kidney function
Introduction
Recent advances in molecular profiling, specifically transcriptional analysis at a single-cell level, can uncover rare cell populations within heterogeneous clinical tissues, which is contributing to our understanding of kidney biology, and its molecular processes (1–7). There is an unmet need to develop methods that can provide comprehensive biological information at the RNA and DNA levels and also allow for single-cell proteomic tissue analysis.
Proteomic technologies, unlike genomics, can provide functional information on cellular states and regulatory networks (2). A technological challenge for mass spectroscopy (MS) based proteomics is the limitation of starting material. Unlike transcriptomics, proteomics does not allow for molecular amplification. This factoid has resulted in substantive efforts to enhance the analytical sensitivity of MS-based proteomics, inclusive of technology miniaturization and higher efficiencies at the electrospray ionization source (8, 9), such that the analytical sensitivity is now sufficient to detect proteins in single mammalian cells. Despite having high analytical sensitivity, proteomic applications to small sample volumes have introduced additional challenges, such as non-specific adsorption of proteins and peptides to the surfaces of reaction tubes, inefficient digestion kinetics, the need for cleanup, and challenges with delivery. Our group has recently reported on the near-single-cell proteomics (nscProteomics) method in HeLa cells, which provides highly innovative and sensitive technology for proteome measurements of samples with sub-nanogram amounts of protein from a small number of cells (10). This method requires highly customized sample processing equipment and is difficult to transfer out to other research labs in its current state of development. Processing human kidney tissues through this pipeline have required close attention to protocol development in order to optimize tissue preservation and cell capture.
In this study, we have focused on providing a roadmap for successful measurement of proteomic interrogation of various sub-compartments of kidney, at the near-single-cell level, with the following deliverables: (i) Assessment of optimal kidney tissue collection and storage for nscProteomics; (ii) Identification of the appropriate reference standard for nscProteomics; (iii) Optimization of the kidney tissue thickness for laser capture microdissection (LCM); (iv) Optimization of LCM parameters; (v) Description of the nscProteomics method using an LC-MS based, fully automated micro POTS method (μPOTS; Processing in One pot for Trace Samples) (11); (vi) Data analytics for nscProteomics.
To achieve the aforementioned tasks, we have processed 11 unique human kidney biopsies and developed protocols and methodologies to collect, process, and interrogate the proteomic expression of human kidney samples by μPOTS. When coupled with highly sensitive LC-MS, we exhibit a fully automated μPOTS method that enables reproducibility and provides quantitative proteomic measurements of ~3,000 proteins from 10 to 100 laser capture micro-dissected (LCM) kidney cells, a level of coverage only achieved previously for thousands of cells (12–17). This study will help for molecular characterization of tissue cellular heterogeneity and pathology in kidney biopsies from patients with different causes of acute and chronic kidney disease. This unique technology has the potential to unravel proteomic heterogeneity and unique protein markers in different kidney disease sub-types and sub-structures that will correlate to disease pathogenesis, prognosis, and risk stratification. The study is summarized in Figure 1.
Figure 1. The near single-cell proteomics (nscProteomics) workflow is optimized for processing human kidney tissues. The workflow includes the identification of optimal kidney tissue collection and storage, the quality controls for laser capture microdissection establishment of nscProteomics method for kidney cells.
Experimental Procedures
Study Design
The schematic representation of the proteomic studies undertaken on 11 unique human normal kidneys is shown in Figure 2A. A total of 28 mass spectrometry (MS) runs were performed for bulk and laser capture micro-dissected (LCM) nscProteomics on paired glomerular (Glom) and proximal convoluted tubular (PT) sections. Tissues from the first two kidneys were preserved as FFPE and in optimal cutting temperature compound (OCT) medium. Kidney tissues issues from 9 more kidneys were only processed in OCT (Figure 2A).
Figure 2. (A) A summary of study samples and assays. (B) The QC plot for reproducibility of the process using OCT tissue. Spectral count differences between proteins for 2 separate runs using OCT tissue were plotted against the average protein spectral count for 1,257 proteins. In the plot, the blue line depicts mean difference; the red and green lines depict 2 SD and 3 SD limits respectively. 97.96% of proteins are within the calculated reproducibility limit of 13.52 indicating a high degree of reproducibility. It is also seen that the inter-run variability of the process involving OCT tissue is lower than that of the process involving FFPE tissue (calculated reproducibility limit: 24.43). (C) Comparison of spectral count distributions of 374 common proteins in OCT and FFPE. A larger number of proteins from OCT tissue show high spectral counts while a higher preponderance of low spectral counts is seen for proteins from FFPE tissue. (D) Comparison of spectral count distributions of 76 unique proteins in FFPE and 244 unique proteins in OCT. The higher preponderance of low spectral count proteins is seen in OCT tissue. This may indicate that OCT tissue technique is better for detection of low-abundant proteins in the current scenario. The correlation coefficient between proteins identified from 2 OCT frozen tissues was 0.96 (P < 0.001).
Kidney Tissue Sample Acquisition
Human kidney tissues were collected from a total of 11 de-identified partial nephrectomies, obtained from the University of California San Francisco (UCSF) and University of Michigan (UM), with approved IRB, as part of a pilot technology feasibility study in the NIH Kidney Precision Medicine Project (KPMP). Only the healthy area of the kidney as determined by a trained pathologist was used for this purpose. Samples were anonymized with the only caveat being that the tissues were collected from adults. To assess the optimal sample collection protocol for the technology, initially, ~100 mg of tissue from 2 kidneys was divided equally and either prepared as formalin-fixed paraffin-embedded (FFPE) tissue sections or was frozen in the Tissue-Plus™ O.C.T. Compound (Fisher Scientific). One 5 μm and three 10 μm thick consecutive sections were cut from each OCT block with a cryo-microtome and mounted on 1.0 polyethylene terephthalate (PET) membrane slides (Zeiss) for glomerular and proximal tubule dissection. The 5 μm thick section was H&E stained for visualization of major histological structures. The remaining 10 μm thick sections were left unstained. All slides were stored at −80°C until dissection.
Cistanche for improving kidney dysfunction
Assessment of Impact of Tissue Shipment and Processing on Proteomic Output
To assess the impact of delays on the proteomic analysis of OCT frozen tissue, kidney tissue samples were collected at one center, shipped to a second site >4 h travel away, and processed at the second center, and vice versa. Bulk proteomics was performed on two 2 unique kidneys (kidney #3, #4), procured at the University of Michigan, and shipped to 2 different locations (Ohio State University and UCSF) for MS analysis, using the same protocol and MS parameters at both sites. Results of the MS runs were then compared between the 2 sites.
Protein Extraction and Precipitation
This was conducted on both the OCT and the FFPE tissue prep methods for the selection of the optimal method for tissue proteomics. OCT tissue was sectioned into 10 μm thick sections using a cryomicrotome. To select the minimal amount of tissue needed for protein extraction for MS, three curls (a section that is put in an Eppendorf tube instead of spread on the slide), and 5 curls of frozen tissue were cut, thawed, and transferred to microcentrifuge tubes containing a 5 mm stainless steel bead (Qiagen), exposed to Mammalian Cell Lysis Buffer (including Benzonase® Nuclease and Protease Inhibitor Solution; Qiagen), and homogenized in a TissueLyser LT (Qiagen) at 50 r/s for 3 min. Protein concentration was quantified (Bradford assay; Thermo Scientific) and stored at −20°C until use. Tissue stored for <2 months in FFPE was sectioned into ten 10 μm thick sections using a microtome. Three and five curls of tissue were cut, deparaffinized in xylene, centrifuged, and incubated in Extraction Buffer EXB1 (Qiagen) supplemented with β-Mercaptoethanol. Samples were incubated on ice, followed by incubation at 80°C for 2 h. Protein was quantified (Bradford assay; ThermoScientific) and stored at −20°C until further use. Protein was precipitated by the addition of cold acetone, followed by centrifugation to pellet and dry the precipitated protein. The pellet was re-suspended in 6 M Urea/100 mM Tris-HCl pH 7.8, and protein concentration was quantified (Bradford assay; ThermoScientific) and stored at −20°C until trypsin digestion.
Trypsin Digestion
Approximately 150 μg of total protein (~15% of the available input) was exposed to reducing agents (200 mM DTT and 100 mM Tris-HCl) and alkylating reagents (200 mM iodoacetamide and 100 mM Tris-HCl) and then diluted with RNAse/DNAse free water to reduce Urea concentration to 0.6 M. Four μg of porcine Trypsin (Sigma) was added and the sample digested at 37°C. Protein concentration was quantified (Bradford assay; ThermoScientific) and 10 μg of protein was used for MS.
MS for Bulk Proteomics to Assess the Optimal Kidney Tissue Processing Method
The tryptic peptide mixture was acidified with formic acid, cleaned with C18 Monospin columns, and analyzed on the Orbitrap Q Exactive HF-X (Thermo Scientific). MS/MS was performed using Higher-Energy C-Trap Dissociation (HCD). Mass spectra were analyzed using Byonic v2.14.27 (Protein Metrics) allowing for 12 ppm mass tolerances. Variable post-translational modifications were allowed for, including oxidation, methylation, carbamylation, and phosphorylation. Proteins were limited to those scoring better than a 1% false discovery rate (FDR). Comparisons of protein abundance and MS data were made to select the optimal method for kidney tissue collection between FFPE and OCT.
Laser Capture Microdissection
Unstained 5, 10, and 20 μM thick sections mounted on PET slides were placed on the stage of the Zeiss PALM MicroBeam Laser Microdissection system (Zeiss) to test optimal cutting thickness. Glomerular and proximal tubule regions were selected using the freehand tool and dissections were performed at 10X objective. Sections with ~4,580 μm2 area and 10 uM thickness were captured, which accounted for ~40 glomerular cells based on the previously published formula for estimation of glomerular cells in a healthy adult kidney (18). For the proximal tubular cells, we captured ~2,900 μm2 of proximal tubular cells, which accounted for ~36 cells. Cut sections were catapulted into a 200 μL opaque adhesive cap (Zeiss) (LPC Energy, 66; LPC Focus, 79) and stored at −80°C.
Near-Single-Cell-Proteomics (nscProteomics)
To solubilize proteins collected in capture caps with LCM, a 10 μL droplet of 0.1% DDM 50 mM Tris pH 8 was added directly to the cap. Samples were incubated for 30 min at 37°C in an Eppendorf ThermoTop thermomixer to reduce evaporation, followed by centrifugation at 2,000 × g for 1 min to transfer the lysate to the vial. Proteins were reduced with 5 mM dithiothreitol and incubated at 37°C for 30 min. Alkylation was carried out at 10 mM iodoacetamide with a 45 min incubation in the dark at 25°C. Next, a 10 ng aliquot of Ls-C was added followed by a 3 h incubation at 25°C. A 10 ng aliquot of trypsin was then added followed by overnight digestion at 25°C. The digested peptides were mixed with a 15 μL aliquot of 18 MΩ water.
LC-MS Platform for the Ultra-Small Samples
Peptide samples (25 μL) were separated using a 60 cm column, with a 50 μm inner diameter and an integrated emitter (New Objective). The columns were packed in-house with 1.7 μm diameter Waters BEH media. A 100-min gradient was produced using a Dionex Ultimate 3000 RSLC nano pump (Thermo Scientific). This system was coupled to a QExactive Plus mass spectrometer (Thermo Scientific). Mass spectra were collected from 300 to 1,800 m/z, using a top 12 data-dependent acquisition method. MS1 spectra were collected with a mass resolution of 35K and MS2 with 17.5K. A 100 ms maximum IT was used to increase identifications from low abundance ions.
Proteomic Data Extraction
All raw files were processed using MaxQuant (version 1.5.3.30) for feature detection, database searching and protein/peptide quantification (19) following settings described previously (10). Tandem mass spectra were searched in the UniProtKB/Swiss-Prot human database (downloaded on 29 Dec 2018 and containing 20,417 reviewed entries). The MS proteomics data for nscProteomics has been deposited to the ProteomeXchange Consortium via the PRIDE (20) partner repository with the dataset identifier PXD015058 and 10.6019/PXD015058.
Data Analysis
MS data were obtained from runs performed at PNNL (nscProteomics) and Stanford University (bulk proteomics). For bulk proteomics. The following statistical methods were used for selecting the optimal kidney tissue collection method, either frozen in OCT or processed as FFPE, as assessed by quantitative tissue bulk proteomics, (i) Protein yield/mg of equivalent input tissue was calculated; (ii) MS reproducibility was tested on MS output from FFPE and OCT sections from each of 2 kidneys, prepared by the same individual using the same protocol, analyzed on the same MS instrument using a set protocol, a few days apart (21). A reproducibility limit (22) was used to provide an approximate bound on measurement differences between successive runs of a single process as well as an estimate of precision for comparison between processes. The correlation coefficient, which provided the degree of agreement between successive runs (using the same tissue), was calculated; (iii) Protein abundance and peptide fragment size distribution were calculated to assess variations originated from protein degradation, crosslinking due to formalin, etc. Other aspects of QC, included (iv) assessment of process drift over replicates and time by using a common pool kidney reference and (v) calculation of bias. Stringent quantitative analysis and identification of unique proteins utilized data only with proteins with spectral counts ≥5 (23). Fisher's exact test was used to assess the enrichment of common and unique proteins mapped between two normal human kidneys, (#1 and #2). Based on the above analysis (see results), OCT frozen tissue was selected as the collection method for all subsequent analyses.
The next step of the data analysis focus was to identify, quantify, and validate protein sets enriched in or specific to the ~10–100 cells obtained from each of two selected kidney sub-compartments—Glom and PT regions, obtained from the same kidney, by bulk and LCM of OCT frozen kidney tissue (n = 9; kidneys #3–11) and run by nscProteomics.
A 2-step semi-conservative filtering approach was adopted to deal with issues of missing/incomplete data, and the G-test (24) was used to identify if the data was Missing At Random (MAR) or Missing Not At Random (MNAR). Additional data reliability assessment was done by stringently selecting proteins observed in ≥50% of all samples. T-test was used to identify for proteins significantly enrichment in glomming or PT and multiple testing correction (Benjamini-Hochberg FDR) with a significance threshold of 0.1 was applied. For purposes of the analysis, the relative intensity MS data (log 2 transformed) obtained from Glom, PT, and bulk fractions from a single kidney were considered to be paired. We identified proteins enriched in Glom (with low levels in PT/ bulk), and unique to Glom (absent in PT) as well as proteins enriched in PT (with low levels in Glom/ bulk) and unique to PT (absent in Glom). Enrichment analysis was done only on data without missing values in each group to increase confidence. The correlation between the abundance values of common proteins was used to assess the degree of agreement between enriched sub-compartment proteins between separate runs of the same experiment. We also performed cross-validation of significantly enriched proteins in each compartment by parsing these comparisons from unique sample sets between 2 different runs, Run 1 and Run 2. Biological validation for enriched sub-compartment (Glom and PT) sets of proteins was undertaken by querying if known sub-compartment enriched proteins could be localized back to their specific regions in public data from the Human Protein Atlas (https://www.proteinatlas.org/) (25).
Cistanche for kidney failure symptoms
Cross-Omics Data Comparison/Integration
To assess the concordance of protein expression at the RNA level, data generated from single-cell RNA-sequencing (scRNA Seq) performed on 10 native nephrectomies was utilized, 9 of which were overlapping with the kidneys used in nscProteomics. For this, 10x Genomics Chromium platform with v3 chemistry was used using the method optimized in the Sarwal lab as a Tissue Interrogation Site (TIS) of KPMP Consortium (a manuscript detailing the method and results is under preparation). For this cross-omics analysis, we used transcriptome data from 45,411 kidney cells resolved into nine major cell types: Podocytes, mesangial cells, glomerular endothelium, stroma/interstitium, immune cells, proximal tubule, thick ascending limb of the loop of Henle, distal/connecting tubule, and collecting duct. Analysis of single-cell data was done using the Seurat R package version 2.3.4. More in-depth analysis of this data is in development (26).
From: 'Near-Single-Cell Proteomics Profiling of the Proximal Tubular and Glomerulus of the Normal Human Kidney' by Tara K. Sigdel1, Paul D. Piehowski2, Sudeshna Roy1, Juliane Liberto1, Joshua R. Hansen2, Adam C. Swensen2, Rui Zhao3, Ying Zhu3, Priyanka Rashmi1, Andrew Schroeder1, Izabella Damm1, Swastika Sur1, Jinghui Luo4, Yingbao Yang4, Wei-Jun Qian2* and Minnie M. Sarwal1* for the Kidney Precision Medicine Project (KPMP) Consortium
---ORIGINAL RESEARCH article Front. Med., 17 September 2020 |
