Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
[Preprint]. 2024 Apr 10:2023.10.23.563662.
doi: 10.1101/2023.10.23.563662.

CRISPR screens and lectin microarrays identify novel high mannose N-glycan regulators

C Kimberly Tsui  1 , Nicholas Twells  2 , Emma Doan  1 , Jacqueline Woo  1 , Noosha Khosrojerdi  1 , Janiya Brooks  1 , Ayodeji Kulepa  2 , Brant Webster  1 , Lara K Mahal  2 , Andrew Dillin  1
Affiliations

CRISPR screens and lectin microarrays identify novel high mannose N-glycan regulators

C Kimberly Tsui et al. bioRxiv. .

Update in

Abstract

Glycans play critical roles in cellular signaling and function. Unlike proteins, glycan structures are not templated from genes but the concerted activity of many genes, making them historically challenging to study. Here, we present a strategy that utilizes pooled CRISPR screens and lectin microarrays to uncover and characterize regulators of cell surface glycosylation. We applied this approach to study the regulation of high mannose glycans - the starting structure of all asparagine(N)-linked-glycans. We used CRISPR screens to uncover the expanded network of genes controlling high mannose surface levels, followed by lectin microarrays to fully measure the complex effect of select regulators on glycosylation globally. Through this, we elucidated how two novel high mannose regulators - TM9SF3 and the CCC complex - control complex N-glycosylation via regulating Golgi morphology and function. Notably, this method allowed us to interrogate Golgi function in-depth and reveal that similar disruption to Golgi morphology can lead to drastically different glycosylation outcomes. Collectively, this work demonstrates a generalizable approach for systematically dissecting the regulatory network underlying glycosylation.

PubMed Disclaimer

Figures

Extended Data Fig. 1
Extended Data Fig. 1
a, Dox-titration of XBP1s-induction. Cell surface HHL-binding is measured by flow cytometry. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results. b, Representative flow cytometry results of lectin binding for HHL as related to Fig. 1c. c, Representative flow cytometry results of lectin binding for GRFT as related to Fig. 1C. d, HHL staining on cells treated with Endoglycosidase H for 1 hour at 37°C. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results. e, Representative flow cytometry results of lectin binding for MBL2 as related to Fig. 1C. f, Cell surface HHL-binding on cells treated with increasing concentration of IXA4. Flow cytometry results are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results. g, Proportional changes of high mannose structures in A549 cells with or without XBP1s-induction, as related to Fig. 1d. The experiment was performed in triplicate, and data is presented as mean ± s.e.m. h, Cell surface lectin binding of XBP1s-induced cells normalized to basal control cells. Data are representative of two independent experiments performed in triplicate.
Extended Data Fig. 2
Extended Data Fig. 2
a, Flow cytometry result of HHL binding in A549 wild type control and cells expressing XBP1-targeting sgRNA (XBP1-KO). Data are representative of two independent experiments performed in triplicate. b, Reproducibility plot of casTLE scores for the CRISPR-KO genome-wide screen. c, Validation of hits in basal A549s using competitive HHL binding assays: Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results.
Extended Data Fig. 3
Extended Data Fig. 3
a, Reproducibility plot of casTLE scores for the CRISPRi targeted screen under basal condition. b, Reproducibility plot of casTLE scores for the CRISPRi targeted screen under XBP1s-indcued condition. c, Clustered heat map representing the results of competitive cell surface lectin binding assay for KD of top screen hits under basal conditions. Lectin binding specificities are indicated. Each column is a replicate and each row represents a different gene knock down.
Extended Data Fig. 4
Extended Data Fig. 4
a, Competitive HHL binding assay in A549s with expressing three independent sgRNAs targeting TM9SF3. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate. b, Competitive HHL binding assay in A549s under basal conditions expressing three independent sgRNAs targeting TM9SF3. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results. c, RT-qPCR for TM9SF family members. Gene expression is normalized to housekeeping genes GAPDH and HRPT1. d, Flow cytometry quantification of intracellular staining of GM130 and TGN46 in TM9SF3 knockdown cells under basal and XBP1s-induced conditions compared to wildtype control. Data are presented as mean ± s.e.m. and are representative of three independent experiments performed in triplicate with consistent results. e, Schematic for microscopy analysis. Wildtype control cells (mcherry-positive) were cocultured with TM9SF3 knockdown cells (mcherry-negtaive) as in-well internal controls. f and h, Representative confocal microscopy images of TM9SF3 knockdown and wildtype control cells, stained with cis-/medial-Golgi marker GM130 (f) or TGN marker TGN46 (h). Wildtype cells are outlined in dotted white lines. Magnified view of the red boxed areas are shown in the right-most column, with red arrows denoting KD cells. Scale bars, 10 μm. Images are representative of two independent experiments performed in triplicate. g and i, Quantification of average cis-/medial-Golgi or TGN distances from the nucleus in Extended Data Fig. 4f and 4h, respectively. Each data point represents the mean distance of a single cell. P-value is calculated by Mann-Whiteney U test. j, Volcano plot for lectin microarray results of XBP1s-induced A549 cells with TM9SF3 knocked down compared to wild type control. Lectins are color-coded by their glycan-binding specificities. k, Competitive cell surface lectin binding assay for TM9SF3 knocked down A549s compared to wildtype control under XBP1s-induced conditions. Lectin binding specificities and the location of where the modification predominately occurs are indicated. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results.
Extended Data Fig. 5
Extended Data Fig. 5
a, Competitive HHL binding assays on A549s under XBP1s-induced conditions for all knock down of all members of the CCC complex. Each gene is knocked down by co-expression of two independent sgRNAs. Data are presented as mean ± s.e.m. and are representative of three independent experiments performed in triplicate with consistent results. b, Western blot of CCDC22 knockdown cell line showing expected reduction in CCDC22 protein levels. c, Flow cytometry quantification of intracellular staining of GM130 and TGN46 in CCDC22 knockdown cells under basal and XBP1s-induced conditions compared to wildtype control. Data are presented as mean ± s.e.m. and are representative of three independent experiments performed in triplicate with consistent results. d and f, Representative confocal microscopy images of CCDC22 knockdown and wildtype control cells, stained with cis-/medial-Golgi marker GM130 (d) or TGN marker TGN46 (f). Wildtype cells are outlined in dotted white lines. Magnified view of the red boxed areas are shown in the right-most column, with red arrows denoting KD cells. Scale bars, 10 μm. Images are representative of two independent experiments performed in triplicate. e and g, Quantification of average cis-/medial-Golgi or TGN distances from the nucleus in Extended Data Fig. 5d and 5f, respectively. Each data point represents the mean distance of a single cell. P-value is calculated by Mann-Whiteney U test. h, Volcano plot for lectin microarray results of XBP1s-induced A549 cells with CCDC22 knocked down compared to wild type control. Lectins are color-coded by their glycan-binding specificities. i, Competitive cell surface lectin binding assay for CCDC22 knocked down A549s compared to wildtype control under XBP1s-induced conditions. Lectin binding specificities and the location of where the modification predominately occurs are indicated. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results. j, Model for how glycosylation is altered in TM9SF3-KD cells: High mannose glycans are converted into oligomannose glycans in the cis- and medial-Golgi despite changes in morphology. However, the final stages of complex glycan synthesis, such as elongation by LacNAc motifs and capping with sialic acids, are inhibited. k, Model for how glycosylation is altered in CCDC22-KD cells: Despite the fragmented Golgi morphology, glycans are able to be remodeled from high mannose into complex glycans, perhaps at an enhanced efficiency. Resulting in a more complex N-glycan repertoire at the expense of high mannose glycans.
Fig. 1:
Fig. 1:
XBP1s-induction upregulates high mannose N-glycans on and within cells a, Schematic for dox-inducible XBP1s upregulating high mannose N-glycans. b, RT-qPCR for targets of general UPRER and XBP1s. Gene expression is normalized to housekeeping genes GAPDH and HRPT1. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results. c, Fluorescent HHL and GRFT binding on A549 cells with or without dox-induction of XBP1s. Cells were treated with 2 μg/mL dox for 48 hours to overexpress XBP1s. The flow cytometry data is representative of three independent experiments performed in triplicate. d, Fluorescent MBL2 binding on A549 cells with or without dox-induction of XBP1s. Cells were treated with 2 μg/mL dox for 48 hours to overexpress XBP1s. The flow cytometry data is representative of three independent experiments performed in triplicate. e, UPLC quantification of high mannose N-glycan structures of A549 cells with or without XBP1s-induction. Levels of each high mannose structure are normalized to the protein amount of each replicate. The experiment was performed in triplicate, and data is presented as mean ± s.e.m. f, Schematic for lectin microarray analysis of A549s under basal or XBP1s-induced conditions. g, Volcano plot of lectin microarray data. Median normalized log2 ratios (sample /reference) of the A549 samples are presented. Lectins are color-coded by their glycan-binding specificities.
Fig. 2:
Fig. 2:
Genome-wide CRISPR screen uncovers the expanded network of genes regulating high mannose. a, Schematic for FACS-based CRISPR screen. Cas9-expressing A549s were lentivirally transduced with a genome-wide CRISPR-deletion sgRNA library. Resulting cells were dox-treated to induce XBP1s overexpression for 48 hours. Cells were then gently lifted with Accutase, fixed, and stained with FITC-labeled HHL. The top and bottom 25% of HHL stained cells were isolated by FACS. The resulting populations were subjected to deep sequencing and analysis. The screen was performed in duplicate. b, Volcano plot of all genes indicating effect and confidence scores for the genome-wide screen performed in duplicate. Effect and P values were calculated by casTLE. c, Schematic for initial steps of N-glycan mannose-trimming and remodeling. All three enzymes indicated are hits in genome-wide screen. d, Disruption of tail-anchored protein insertion pathway by ASNA1 inhibitor Retro-2 in wild type A549s also upregulates cell surface high mannose glycan levels. A549s were treated with treated with 2 μg/mL dox, 100 μM of Retro-2, both, or left untreated for 48 hours. Resulting cells were lifted with Accutase and stained with FITC-labeled HHL, followed by flow cytometry analysis. Data are presented as mean ± s.e.m. of median of each replicate and are representative of two independent experiments performed in triplicate with consistent results. e, Schematic for competitive binding assays for measuring changes in high mannose levels. Cells expressing sgRNAs for CRISPRi-mediated knockdown (KD) and miRFP and cells expressing a control sgRNA and BFP were cocultured in 1:1 ratio. Cells were either treated with dox to induce XBP1s or left untreated for 48 hours. Resulting cells were lifted and stained with HHL-FITC, and log2 ratio of HHL intensity of KO: control was determined using flow cytometry. f, Validation of hits in XBP1s-induced A549s using competitive HHL binding assays: Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results.
Fig. 3:
Fig. 3:
Targeted CRISPRi screen uncovers additional novel regulators of high mannose glycans under basal and UPRER induced conditions a, Schematic for MACS-based CRISPR screen. A549 cells stably expressing CRISPRi machinery and the targeted sgRNA sublibrary were either dox-treated to induce XBP1s or left untreated for 48 hours. Cells were lifted and incubated with HHL coupled to magnetic beads. The cells were then placed on a magnet in which high HHL-binding cells would be retained on the magnet, whereas the low HHL-binding cells were removed from the population. This separation was repeated twice more on each high and low HHL binding cells to improve the purity of the populations. Finally, each resulting population were subjected to deep sequencing and analysis to identify hits. The screen was performed in duplicate. b, The maximum effect size (center value) estimated by CasTLE from both basal and XBP1s-induced conditions with five independent sgRNA per gene. The bars represent the 95% credible interval, with red representing XBP1s and blue representing basal conditions. Only genes considered to be a hit in at least one condition are shown. Genes are ordered in descending order of estimated maximum effect size of XBP1s-induced condition. The top 30 positive and negative hits are shown in the expanded panels. c, Top 30 regulators for high mannose N-glycans with their reported subcellular localization. d, Validation of hits in A549 under basal conditions using competitive HHL binding assays. Each gene is knocked down by co-expression of two independent sgRNAs. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results. e, Validation of hits in A549 under XBP1s-induced conditions using competitive HHL binding assays. Each gene is knocked down by co-expression of two independent sgRNAs. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate with consistent results.
Fig. 4:
Fig. 4:
TM9SF3 regulates the Golgi organization and is required for formation of complex N-glycans a, Competitive HHL binding assay in A549s with each TM9SF family member knocked down. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate. b, Flow cytometry quantification of intracellular staining of GM130 and TGN46 in A549s. Data are presented as mean ± s.e.m. and are representative of three independent experiments performed in triplicate. c, Representative confocal microscopy images of TM9SF3 knockdown and wildtype control cells, co-stained with cis-/medial-Golgi marker GM130 and TGN marker TGN46. Magnified views of the red boxed areas are shown in the right-most column. Scale bars, 10 μm. Images are representative of two independent experiments performed in triplicate. d, Percent area of each Golgi compartment co-localized with the other compartment. Data are presented as mean ± s.e.m., from 12 images each from wildtype or TM9SF3 knockdown of two independent experiments, with >20 cells per image. e, Volcano plot for lectin microarray results of A549 cells with TM9SF3 knocked down compared to wild type control. Lectins are color-coded by their glycan-binding specificities. f, Competitive cell surface lectin binding assay for TM9SF3 knocked down A549s compared to wild type control under basal conditions. Lectin binding specificities and location of where the modification predominately occurs are indicated. Data are presented as mean ± s.e.m. and are representative of three independent experiments performed in triplicate.
Fig. 5:
Fig. 5:
CCDC22 regulates elongation and sialylation of glycans through modulating Golgi expansion a, Competitive HHL binding assays on A549s for all knock down of all members of the CCC complex. Each gene is knocked down by co-expression of two independent sgRNAs. Data are presented as mean ± s.e.m. and are representative of three independent experiments performed in triplicate. b, Flow cytometry quantification of intracellular staining of GM130 and TGN46 in A549s in CCDC22 knockdown and wild type control cells. Data are presented as mean ± s.e.m. and are representative of three independent experiments performed in triplicate with consistent results. c, Representative confocal microscopy images of CCDC22 knockdown and wildtype control cells, co-stained with cis-/medial-Golgi marker GM130 and TGN marker TGN46. Magnified views of the red boxed areas are shown in the right-most column. Scale bars, 10 μm. Images are representative of three independent experiments performed in triplicate. d, Percent area of each Golgi compartment co-localized with the other compartment. Data are presented as mean ± s.e.m., from 12 images each from wildtype or CCDC22 knockdown of two independent experiments, with >20 cells per image. e, Volcano plot for lectin microarray results of basal A549 cells with CCDC22 knocked down compared to wildtype control. Lectins are color-coded by their glycan-binding specificities. f, Competitive cell surface lectin binding assay for CCDC22 knocked down A549s compared to wildtype control under basal conditions. Lectin binding specificities and the location of where the modification predominately occurs are indicated. Data are presented as mean ± s.e.m. and are representative of two independent experiments performed in triplicate.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Reily C., Stewart T. J., Renfrow M. B. & Novak J. Glycosylation in health and disease. Nat. Rev. Nephrol. 15, 346–366 (2019). - PMC - PubMed
    1. Varki A. Biological roles of glycans. Glycobiology 27, 3–49 (2017). - PMC - PubMed
    1. Ohtsubo K. & Marth J. D. Glycosylation in cellular mechanisms of health and disease. Cell 126, 855–867 (2006). - PubMed
    1. Pinho S. S. & Reis C. A. Glycosylation in cancer: mechanisms and clinical implications. Nat. Rev. Cancer 15, 540–555 (2015). - PubMed
    1. Qin R. & Mahal L. K. The host glycomic response to pathogens. Curr. Opin. Struct. Biol. 68, 149–156 (2021). - PubMed

Publication types