Presented here is a protocol that combines an in vitro neural-endothelial co-culture system and metabolic incorporation of sialoglycan with bioorthogonal functional groups to expand primary neural stem and progenitor cells and label their surface sialoglycoproteins for imaging or mass-spectrometry analysis of cell surface markers.
Neural stem and progenitor cells (NSPCs) are the cellular basis for the complex structures and functions of the brain. They are located in specialized niches in vivo and can be isolated and expanded in vitro, serving as an important resource for cell transplantation to repair brain damage. However, NSPCs are heterogeneous and not clearly defined at the molecular level or purified due to a lack of specific cell surface markers. The protocol presented, which has been previously reported, combines a neural-endothelial co-culture system with a metabolic glycan labeling method to identify the surface sialoglycoproteome of primary NSPCs. The NSPC-endothelial co-culture system allows self-renewal and expansion of primary NSPCs in vitro, generating a sufficient number of NSPCs. Sialoglycans in cultured NSPCs are labeled using an unnatural sialic acid metabolic reporter with bioorthogonal functional groups. By comparing the sialoglycoproteome from self-renewing NSPCs expanded in an endothelial co-culture with differentiating neural culture, we identify a list of membrane proteins that are enriched in NSPCs. In detail, the protocol involves: 1) set-up of an NSPC-endothelial co-culture and NSPC differentiating culture; 2) labeling with azidosugar per-O-acetylated N-azidoacetylmannosamine (Ac4ManNAz); and 3) biotin conjugation to modified sialoglycan for imaging after fixation of neural culture or protein extraction from neural culture for mass spectrometry analysis. Then, the NSPC-enriched surface marker candidates are selected by comparative analysis of mass spectrometry data from both the expanded NSPC and differentiated neural cultures. This protocol is highly sensitive for identifying membrane proteins of low abundance in the starting materials, and it can be applied to marker discovery in other systems with appropriate modifications
Neural stem cells are defined as a multipotent cell population that can self-renew to maintain a stem cell pool and differentiate into neurons and glia. They are the major cell types in the nervous system and may offer great therapeutic potential in regenerative medicine through cell transplantation into diseased and injured brains1,2. As development proceeds, the neural stem cell population becomes heterogenous3,4, and the proportion of neural stem cells in the brain gradually decreases5. Generally speaking, embryonic neural stem cells and other neural progenitor cells, collectively called neural stem and progenitor cells (NSPCs), are located in the germinal zones, the ventricular zone, and the subventricular zone in mice6. In the embryonic brain, neural stem cells generate neurons directly or indirectly through intermediate progenitor cells (IPCs), and in some species through the outer subventricular zone progenitors (oRGs)7,8. The specific molecular signature, morphology, location in the stem cell niche, and differentiation potential all determine the role of each subtype in brain organogenesis and clinical applications9. However, the currently available cell surface markers cannot unequivocally discriminate and purify different subtypes of NSPCs, limiting the understanding and utilization of these subtypes.
The identification of primary NSPCs surface markers is limited by three major hurdles. The first one is the limited cell number of NSPCs in the tissue, making it difficult to prepare cell surface protein samples for common mass spectrometry analysis. The second limitation is the difficulty in producing pure cell subtypes for generating subtype-specific membrane protein data. Finally, the third challenge is the low ratio of cell surface proteins in whole cell proteins, which hampers their detection sensitivities by mass spectrometry analysis.
To overcome these problems, we developed a chemoproteomic approach to selectively enrich and identify cell surface proteins in primary NSPCs by metabolically labeling the sialoglycoproteins10. To generate a sufficient number of NSPCs, we took advantage of an established protocol to expand and maintain primary embryonic NSPCs in undifferentiated states in vitro, by co-culturing NSPCs with mouse brain endothelial cell lines using a permeable support matrix insert (e.g., transwell) system11. In contrast, NPSCs cultured alone without endothelial cells generate differentiated progeny11,12. Thus, protein samples from these two culture systems can be comparatively analyzed to identify proteins that are differentially expressed in NSPCs and differentiated neurons. As most cell surface proteins are modified by sialic acid13, unnatural sialic acid precursor analog N-azidoacetylmannosamine-tetraacylated (Ac4ManNAz) was used to hijack the intrinsic metabolic pathway so that endogenous, newly synthesized sialoglycans are labeled with azido groups, generating a chemical handle14. Through azido-alkyne-mediated bioorthogonal reactions, which conjugate biotin to sialoglycans, cell surface proteins can be visualized and enriched for proteomic identification through a streptavidin-coupled fluorophore or matrix14.
Here, we perform staining of SDS-PAGE gel analysis of the surface sialoglycoproteome from NSPCs expanded in an endothelial co-culture and differentiating cells in a non-co-culture system. We also selectively purify surface sialoglycoproteome in the two culture systems for proteomic comparison. Our protocol, compared with the traditional centrifugation-based cell surface purification protocols15, increases extraction efficacy by reducing the surface protein extraction procedures through specific tag conjugation and affinity purification. Meanwhile, it increases the extraction purity of cell surface proteins based on the premise that sialylation happens mostly at the cell surface proteins. Although endothelial factors cannot completely block differentiation of expanded NSPCs, the comparative study between a co-culture and differentiated culture provides a convenient method to pinpoint stem cell-enriched surface proteins without the need to analyze proteins from NPCs purified by FACS16. We believe this approach can be applied to studies of surface proteins in other systems with the appropriate modifications.
All animal protocols used in this study were approved by the IACUC (Institutional Animal Care and Use Committee) of Tsinghua University and performed in accordance with guidelines of the IACUC. The laboratory animal facility at Tsinghua University has been accredited by the AAALAC (Association for Assessment and Accreditation of Laboratory Animal Care International). For staging of embryos, mid-day of the vaginal plug identified was calculated as embryonic day 0.5 (E0.5).
NOTE: All cells are cultured in the cell incubator under conditions of 37 °C and 5% CO2.
1. Preparation of Mouse Endothelial Culture in Permeable Support Inserts
NOTE: BEND3 cells are maintained according to manufacturer's instructions.
2. Preparation of Mouse Primary Cortical NSPCs Culture
3. Set-up of Neural-endothelial Co-culture and Ac4ManNAz Labeling System
4. Immunofluorescent Staining of Sialoglycoproteins in Expanded Primary NSPCs and Differentiated Neurons
5. Purification of Sialoglycoproteins from Expanded Primary NSPCs and Differentiated Neurons
The whole procedure for in vitro expansion and metabolic labeling of primary embryonic NSPCs takes 6 days (Figure 1A). Quality of the BEND3 cell line and freshly isolated primary NSPCs are key to a successful experiment. BEND3 cells are the source of soluble factors that stimulate self-renewal and proliferation of NSPCs. It should be ensured that the BEND3 cells are free of any contamination and divide actively with minimal cell death before co-culturing with neural cells. The primary NSPCs must be carefully prepared to avoid excess damage during dissociation. Damaged NSPCs may still grow and differentiate; however, they are not able to respond to endothelial stimuli well to maintain stemness and expand. Extra caution should be taken to be aseptic during cell culturing, as the protocol does not suggest addition of antibiotics to the primary culture medium.
Successful endothelial co-culture will lead NSPCs to form large, sheet-like clones. Such featured clone shapes become evident at day 4 and are very typical at day 6. Within the clones, the cells maintain close contact with each other. Immunostaining with antibodies against the NSPC marker Nestin and the neuronal marker β-tubulin III should reveal that in the clone, most of the cells are Nestin+ NSPCs and very few are β-tubulin III+ neuronal cells. In contrast, the percentage of Nestin+ cells and β-tubulin III+ neuronal cells in clone formed in non-co-culture system are nearly the same (Figure 1B, 1D, and 1E).
The chemical reporter, Ac4ManNAz, is a metabolic analog and can be incorporated into the intrinsic protein sialylation pathway. High doses of Ac4ManNAz are toxic to cells. For each specific type of cell, the labeling concentration of Ac4ManNAz should be pre-tested to achieve the highest labeling efficiency without significant cytotoxicity. Here, the optimized labeling concentration of Ac4ManNAz for primary NSPC is 100 µM. Combinatory evaluation of cell death indicated by cellular and nuclei morphology suggests this labeling concentration does not cause obvious cytotoxic effects and is able to efficiently label NSPCs (Figure 1C and 1D). The clonal morphology, self-renewal, and differentiation potential of NSPCs in both the endothelial co-culture and non-co-culture system are not affected (Figure 1C, 1D, and 1E).
The successful labeling of NSPCs by Ac4ManNAz can be examined after conjugating biotin to a culture mediated by a bioorthogonal reaction between azide and alkyne. Every cell in the Ac4ManNAz-labeled culture is stained and visualized with Alexa Fluor 647-streptavidin. No cell is positive for Alexa Fluor 647-streptavidin staining in the DMSO control group. In addition, protein samples prepared from the Ac4ManNAz-labeled culture by biotin conjugation and streptavidin beads purification show strong Coomassie brilliant blue staining signal in SDS-PAGE gels. Meanwhile, there were only staining background and nonspecific binding signals in the lanes loaded with protein samples from the DMSO control group. This also indicates the efficient labeling of NSPCs by Ac4ManNAz (Figure 1F).
Figure 1: Identification of cell surface markers for primary NSPCs assisted by endothelial co-culture system and metabolic sialoglycan labeling. (A) Schematic of the workflow for the protocol. This figure has been modified from Bai et al.10. The BEND3 cells are seeded into matrix inserts on D0. The preparation of primary cortical NSPCs and set up of co-culture system are performed on D1. Metabolic labeling of culture lasts from D2 to D6. Culture refeeding is carried on D3 and D5. (B) The immunofluorescent images for clones formed by primary NSPCs after 5-day culture with or without endothelial cells. Scale bar indicates 20 µm. (C) Bright-field images for clones formed by primary NSPCs after a 5-day culture with Ac4ManNAz or DMSO. The nuclei were counterstained by DAPI. The scale bar indicates 20 µm. The error bar indicates SEM (n.s. = not significant). (D) The immunofluorescent images for NSPC formed clones in the endothelial co-culture with Ac4ManNAz or DMSO. Dashed circle demarcates a single neural clone. The scale bar indicates 50 µm. (E) Quantification of NSPCs and differentiated neurons in clones formed by NSPCs in endothelial co-culture and non-co-culture system with Ac4ManNAz labeling or DMSO control. The error bar indicates SEM (***p < 0.0005; n.s. = not significant). (F) Coomassie brilliant blue staining of proteins purified by streptavidin beads from neural cells labeled with Ac4ManNAz or DMSO in endothelial co-culture and nonco-culture system. The 55 kD band in control labelling groups represents nonspecific binding proteins. (B, C, E and F) corresponding to this protocol have been adapted from Bai et al.10. Please click here to view a larger version of this figure.
Surface markers are commonly used to label and purify specific cell types in vitro and in vivo17,18. Discovery of surface markers contributes greatly to regenerative medicine and stem cell researches by providing molecular tools to selectively enrich a stem cell population from normal or pathological tissues and culture dishes, offering a purified cell resource for clinical use or study of biological properties. However, progress in developing surface markers for neural stem cell research has been slow due to the difficulty in isolating stem cells from primary tissues. The protocol described here is based on a simplified in vitro platform. By comparing primary NSPCs expanded by an endothelial co-culture to a differentiating neural culture, proteins differentially expressed in expanded NSPCs are highlighted and allow for further identification. Our protocol also provides an alternative strategy to purify cell surface proteins by hijacking the intrinsic metabolic pathway to label sialoglycan with bioorthogonal groups. Compared with traditional protocols for purifying cell surface proteins, the advantages of this protocol are underpinned by two specific features: 1) the prevalence of sialylation on cell surface proteins ensures maximal coverage of the cell surface proteome, and 2) the reaction specificity between the bioorthogonal group and its ligands grants purity of the acquired surface proteome. Thus, our protocol results in a more sensitive proteomic analysis in the case of less starting materials. We have demonstrated the feasibility of this protocol in primary NSPCs surface markers. With the appropriate modifications on expanding stem cells in vitro, this chemoproteomic approach can be compatible with identifying surface markers of other stem cell types. It is noteworthy that as Ac4ManNAz is per-O-acetylated it could lead to artificial S-glycosylation. The use of unacetylated unnatural sugars can avoid the artifact formation and improve the specificity and validity of metabolic glycan labeling in living cells19.
Preparation of primary cortical neural progenitor cells and endothelial cells are critical steps of the protocol. First, when digesting embryonic cortical tissues, the digestion time, amount of enzyme, and strength of handling must be carefully controlled. Excessive digestion and mechanical shearing forces will damage the integrity of plasma membrane and cell surface receptors that mediate signal transduction for cell survival and growth, and they will also disturb the responsiveness of NSPCs to the stimulation of endothelial cells and their self-renewal ability. To achieve proper digestion, experimenters must activate the papain fully and stop the digestion as soon as the tissue blocks disappear. Second, BEND3 cells must be maintained in a healthy state to support the secretion. It is recommended to use BEND3 cell batches with fewer passages and passage the cells before they reach 100% confluence. This will prevent cell cycle arrest and senescence caused by DNA damage accumulated during passaging or by overcrowded contact between cells.
High throughput sequencing technology boosts the identification of cell surface markers through analyzing RNA expression, especially for cell types including tissue stem cells, which are often present in vivo in amounts too small to perform proteome analysis by mass spectrometry. Even though RNA-seq analysis can identify genes specifically expressed in NSPCs, it may not truly reflect protein expression levels, because RNA expression is not always consistent with protein expression20. In addition, non-protein biomolecules that can work as surface markers are not able to be detected by transcriptomic studies. For example, oligosaccharide Lewis X is a well-known surface maker widely used to label human embryonic stem cells and NSPCs, even though it can be associated with multiple proteins21. Therefore, direct mass spectrometry analysis is not yet substitutable, and the development of methods that can make mass spectrometry analysis more feasible and convenient is of great interest for future studies.
In addition to sialylation, other types of post-translational protein modifications play an important role in regulating functions of modified proteins. These modifications affect protein properties such as the conformation, half-life, and subcellular localization22,23. Several protein modifications have cell type specificity24,25,26. With the growing contents of the chemical toolbox, more modification types are amenable to metabolic labelling with chemical reporters27. Hence, the chemical approach described here can be used for studying other differences in protein modification between stem cells and differentiated cells, illustrating the molecular mechanisms behind maintenance of stem cell properties and differentiation regulation.
The authors have nothing to disclose.
Figure 1B, 1C, 1E and 1F are reproduced from Bai et al.10 with permission from the Royal Society of Chemistry. We thank Yi Hao in X. C.'s lab for figure editing. This work is supported by the National Natural Science Foundation of China (No. 91753206 to Q. S. and X. C., No. 31371093 to Q. S., and Nos. 21425204 and 21672013 to X. C.).
BEND3 | ATCC | CRL-229 | |
DMEM | Gibco | 11960044 | |
L-glutamine | Gibco | 25030081 | 1% |
Sodium pyruvate | Sigma | P5280 | 1% |
N2 supplement | Gibco | 17502048 | 1 to 100 |
N-acetyl-L-cysteine | Sigma | A7250 | 1 mM |
Papain | Worthington | LS003726 | 10 U/mL |
B27 supplement | Gibco | 17504044 | 1 to 50 |
Poly-L-lysine | Sigma | P4707 | |
basic Fibroblast growth factor | Gibco | PHG0261 | 10 ng/mL |
Penicillin-Streptomycin | Gibco | 15140122 | 1% |
Fetal bovine serum | Gibco | 10099141 | 10% |
HBSS | Gibco | 14175095 | |
Tripsin-EDTA, 0.25% | Gibco | 25200056 | |
DPBS | Gibco | 14190094 | |
Transwell | Corning | 3450 | |
Paraformaldehyde | Sigma | 158127 | 4% |
Sucrose | Sangon | A100335 | |
DAPI | Gibco | 62248 | |
RIPA buffer | Thermo Scientific | 89900 | |
SDS-PAGE loading buffer 2X | Solarbio | P1018 | |
6-well plate | Corning | 3335 | |
Tris-Glycine protein gel | invitrogen | xp00100box | |
mouse monoclonal anti-Nestin | Developmental Study Hybridoma Bank | Rat-401 | 1 to 20 |
mouse monoclonal anti-beta-tubulin III | Sigma | T8860 | 1 to 1000 |
Alexa Fluor 488 goat anti-mouse IgG1 | invitrogen | A-21121 | 1 to 1000 |
Alexa Fluor 546 goat anti-mouse IgG2b | invitrogen | A-21143 | 1 to 1000 |
Albumin Bovine V | Amresco | 0332 | |
Triton X-100 | Amresco | 0694 | |
BCA assay kit | Thermo Scientific | 23225 | |
dimethyl sulfoxide | Sigma | D2650 | |
Brij97 | Aladdin | B129088 | |
CuSO4 | Sigma | 209198 | |
alkyne-biotin | Click Chemistry Tools | TA105 | |
BTTAA | Click Chemistry Tools | 1236 | |
Ac4ManNAz | Click Chemistry Tools | 1084 | 100 µM |
9AzSia | synthesized in lab | ||
sodium ascorbate | Sigma | A4034 | |
Methanol | Sigma | 34860 | |
EDTA | Sangon | A100322 | |
NaCl | Sangon | A100241 | |
SDS | Sangon | A100227 | |
Alexa Flour 647-conjugated streptavidin | invitrogen | S21374 | 1 to 1000 |
Triethanolamine | Sigma | V900257 | |
Dynabeads M-280 Streptavidin | invitrogen | 60210 | |
ammonium bicarbonate | Sigma | 9830 | |
Coomassie Brilliant Blue R-250 | Thermo Scientific | 20278 | |
Isoflurane | RWD Life Science Co. | 970-00026-00 | |
DNase I | Sigma | DN25 | 12 µg/mL |
urea | Sigma | U5378 |