-
Ligandomics was recently developed as the only technology to globally map cell-wide ligands with simultaneous binding or functional activity quantification in the absence of receptor information. Here we applied this new approach to globally profile LEC ligands by performing three rounds of cell-based binding selection in PM2.5-treated or control cells (Figure 1A). After three rounds of selection, enriched phages showed ≥ 30-fold increase in cell-binding activity (Figure 2).
Figure 1. Schematics of comparative ligandomics to globally identify PM2.5-selective cellular ligands. (A) Open reading frame phage display (OPD) selection. OPD cDNA library was incubated with human lung epithelial cells (LECs) pretreated with or without PM2.5 for multiple rounds of binding selection to enrich clones displaying cellular ligands. (B) Global identification of all enriched ligands by NGS. The cDNA inserts of enriched clones were amplified by PCR and identified by NGS. The copy number of the cDNA inserts is the equivalent of their clone numbers or the binding activity of the displayed ligands. (C) Comparative ligandomics data analysis. Quantitative comparison of the entire ligandome profiles for PM2.5-treated vs. untreated LECs systematically identified PM2.5-selective ligands.
Figure 2. Enrichment of LEC-binding phages by cell-based phage binding selection. The library was incubated with A549 cells with or without PM2.5 treatment in 6-well plates. After washing, bound phages were eluted by 3C protease cleavage. Eluted phages were quantified by plaque assay.
Instead of performing a labor-intensive manual screening of enriched clones individually, the cDNA inserts of all enriched clones were analyzed by NGS. A total of 975,632 and 1,132,564 valid sequence reads for PM2.5-treated and control cells, respectively, were identified. Identified sequences were aligned to 4,140 (PM2.5-treated cells) and 4,091 (control cells) proteins in the NCBI CCDS database (Figure 3A and 3B).
Figure 3. Ligandome profile. (A) Binding activity profile of the entire ligandome for control A549 human LECs. (B) Binding activity profile of the entire ligandome for PM2.5-treated A549 cells. (C) Binding activity ratio of the entire ligandomes for PM2.5-treated vs. control cells.
The copy numbers of cDNA inserts quantified by NGS represent the relative binding activity of the identified ligands (see Discussion), most of which had background binding activity with less than 10 copies of detected cDNAs. Only 1,034 and 686 ligands had ≥ 10 copies that bound to control and PM2.5-treated LECs, respectively (Figure 3A and 3B). Quantitative comparison of the entire ligandome profiles for PM2.5-treated vs. control A549 LECs by Chi-square test systematically identified 143 ligands with increased binding (PM2.5-high) to PM2.5-exposed A549 LECs and 404 ligands with decreased binding (PM2.5-low) (Figure 3C, Table 1).
CCDS_ID Protein Binding activity Activity ratio Control PM2.5 CCDS40232 Gas6 4,904 3,364 1.5X ↓ CCDS16777 Notch2* 1,020 55 18.5X ↓ CCDS20912 ApoE* 79 1,824 23.1X ↑ CCDS28578 Tulp1 20,012 27,450 1.3X ↑ CCDS28713 Abcf1 3,658 3,957 1.1X ↑ Total identified sequences 1,132,564 975,632 Total identified ligandsa 1,034 686 PM2.5-related ligands* 404 ↓ 143 ↑ Note. *P < 0.001, control vs. PM2.5, χ2 test. a, bound ligand count ≥ 10. Table 1. PM2.5-related LEC ligands identified by comparative ligandomics
The binding activity of the entire ligandome profile for PM2.5-treated vs. untreated A549 LECs is plotted in Figure 4A. Ligands represented by green dots within the bottom-right circle have decreased binding activity to PM2.5-treated cells, whereas ligands represented by red dots within the upper-left circle have increased binding activity. Ligands represented by blue dots around the diagonal line have minimal change in binding activity. Proteins represented by bottom-left circles have low background binding activity to A549 cells and indicate non-specific binding. The Pearson correlation coefficient for the entire ligandomes of PM2.5-treated and control LECs in the binding activity plot was calculated as r = 0.9942 (Figure 4A).
Figure 4. Global mapping of PM2.5-specific cellular ligands by comparative ligandomics. (A) Binding activity plot for PM2.5-treated vs. untreated A549 cells. Ligands with increased or decreased binding to PM2.5-treated cells are classified as PM2.5-high or PM2.5-low, respectively. PM2.5-unchanged ligands and background binding showed similar binding activities in both conditions. PM2.5-high ApoE, PM2.5-low Notch2 and unchanged Gas6 are indicated. The Pearson correlation coefficient was calculated as r = 0.9942. (B) PM2.5 treatment altered ligand binding activity. ApoE preferentially bound to PM2.5-treated LECs (1,824:79 for PM2.5-treated vs. control cells). Notch2 binding to LECs was reduced in PM2.5-treated cells (55:1,020). Gas6 had minimal binding activity changes to PM2.5-treated cells (3,364:4,904).
-
The up- and down-regulation of ligandome binding activity profiles to PM2.5-treated cells implied that the particles could alter the expression of cell surface receptors to modulate various cellular responses. One of the important biological responses of the lung epithelium is phagocytosis or engulfment of PM2.5 as a part of innate defense mechanisms for debris clearance[1,20]. Phagocytosis is mediated through cell surface phagocytic receptors[25], whose altered expression can be detected by comparative ligandomics.
Apolipoprotein E (ApoE), a well-known phagocytosis ligand[26,27], was uncovered with up-regulated binding activity. Comparative ligandomics analysis revealed that ApoE binding to PM2.5-treated cells increased by 23.1-fold (P < 0.001) (Figure 4B, Table 1). However, not all phagocytosis ligands showed increased binding activity. For example, growth arrest-specific 6 (Gas6), a well-characterized phagocytosis ligand, had a 1.5-fold decrease in binding to PM2.5-exposed cells (Figure 4B, Table 1). Other phagocytosis ligands, such as Tubby-like protein 1 (Tulp1) and ATP-binding cassette subfamily F member 1 (Abcf1), showed negligible binding activity changes in PM2.5-treated cells (1.3- and 1.1-fold increase, respectively) (Figure 4B, Table 1). However, neurogenic locus notch homolog protein 2 (Notch2), with a potential to regulate phagocytosis, had an 18.5-fold increase in binding to particle-exposed cells (Figure 4B, Table 1). These data suggested that PM2.5 differentially promoted the induction of some phagocytic receptors but not others on human LECs.
Comparative Ligandomic Analysis of Human Lung Epithelial Cells Exposed to PM2.5
doi: 10.3967/bes2020.023
- Received Date: 2019-07-23
- Accepted Date: 2019-12-16
-
Key words:
- PM2.5 /
- Ligandomics /
- Lung epithelial cell /
- Comparative ligandomics /
- Ligand
Abstract:
Citation: | TIAN Hong, Akhalesh SHAKYA, WANG Feng, WU Wei Dong, LI Wei. Comparative Ligandomic Analysis of Human Lung Epithelial Cells Exposed to PM2.5[J]. Biomedical and Environmental Sciences, 2020, 33(3): 165-173. doi: 10.3967/bes2020.023 |