Construction of the turboID systems
To identify the interacting proteome of PNO1 and NOB1 at the molecular level, we constructed fusion expression plasmids for PNO1-TurboID and NOB1-TurboID. The TurboID sequence was fused to the C-terminal of the PNO1 and NOB1 coding sequences (CDS), connected by a flexible linker consisting of 15 amino acids (GGGGS)3, to maximize the identification of proteins interacting with PNO1 and NOB1. The construction of the overexpression plasmids for PNO1-TurboID and NOB1-TurboID is illustrated (Fig. 1A). These plasmids include a tag protein HA, the target gene (PNO1, NOB1) CDS sequence, the flexible linker sequence, the biotin-labeling enzyme TurboID sequence, and the tag protein V5. They also include two different selection markers, AmpR and Puro, where AmpR is used to screen positive clones in E.coli, and Puro is used to select cells transfected with and stably expressing PNO1 and NOB1.
The workflow of the TurboID proximity labeling experiment is as follows (Fig. 1B-C). Using the gene PNO1 as an example, the constructed PNO1-TurboID plasmid was successfully transfected into 293T cells. After allowing the plasmid to express in the cells for 48 hours, the cells were incubated in an environment with an appropriate concentration of biotin for a suitable period to ensure complete biotin uptake by the cells. TurboID utilizes exogenously added biotin and intracellular ATP to convert biotin into biotin-AMP, which then biotinylates lysine residues of proximal proteins. Streptavidin magnetic beads were used to capture the biotinylated proteins, which were then subjected to protein affinity purification. The enriched proteins were eluted under denaturing conditions, followed by sample preparation for mass spectrometry analysis. The identification was performed using high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS).
PNO1 and NOB1 have different localizations in cells
To determine the expression levels of the constructed TurboID overexpression plasmids in cells (Fig. 2A), we transfected 293T cells with PNO1-TurboID and NOB1-TurboID overexpression plasmids. Western Blot analysis using anti-PNO1 and anti-NOB1 antibodies revealed clear bands at molecular weights of 70 kDa and 77 kDa, respectively (with TurboID at 35 kDa, PNO1 at 35 kDa, and NOB1 at 42 kDa). This indicates that the TurboID fusion expression plasmids are stably expressed in cells. As essential members of the ribosomal small subunit 40S, PNO1 and NOB1 are crucial for the processing, assembly, and maturation of the 40S ribosomal subunit. It is known that PNO1 inhibits the cleavage activity of NOB1 on ribosomal RNA, but whether NOB1's functions outside the ribosome in the cell also depend on PNO1 remains unknown. To understand the functions and localization of PNO1 and NOB1 on the 40S pre-ribosomal subunit, we used the PDB database to analyze their spatial distribution in the Cryo-EM structure of the late-stage human 40S pre-ribosomal subunit. The results showed that PNO1 and NOB1 physically interact and are in close spatial proximity (Fig. 2B). These findings suggest that these two proteins may work together in ribosome processing. However, their roles in regulating other cellular functions beyond ribosome maturation are still largely unexplored, leading to the hypothesis that they might also jointly participate in various other biological processes within the cell.
Due to the large size of the TurboID sequence, its fusion with target genes may affect the localization and biological function of the target genes in cells. To ensure that our constructed TurboID fusion expression plasmids are successful and do not affect the gene's localization and function within the cells, we transfected HeLa cells with GFP-PNO1 and GFP-NOB1 plasmids. Immunofluorescence experiments revealed that PNO1 is primarily localized in the nucleus and nucleolus, whereas NOB1 is mainly localized in the cytoplasm. Additionally, we transfected cells with HA-PNO1-TurboID and HA-NOB1-TurboID plasmids and used anti-HA antibodies to detect their localization. Immunofluorescence results demonstrated that the localization of the target genes fused with TurboID in cells matched the GFP localization (Fig. 2C-D), indicating that TurboID does not affect the function or proper localization of PNO1 and NOB1 in cells. This ensures the accuracy of subsequent mass spectrometry detection of proteins interacting with PNO1 and NOB1.
Moreover, immunofluorescence results revealed that PNO1 and NOB1 have different localizations, with PNO1 in the nucleus and NOB1 in the cytoplasm, yet they exhibit physical interactions in spatial structure. This suggests that, besides participating in ribosome biogenesis, PNO1 and NOB1 may have additional functions. Therefore, we also utilized TurboID technology to explore the additional biological functions of these two proteins.
Determination of biotin concentration and incubation time for the turboID system
Based on the results shown above, to further confirm the interaction between PNO1 and NOB1, we performed co-immunoprecipitation experiments (Fig. 3A-B), which indicated that PNO1 and NOB1 interact with each other. Before conducting the formal biotin proximity labeling experiments, we first needed to determine the optimal incubation concentration and time for the TurboID system to biotinylate proximal proteins.By setting incubation times of 0, 15, 30, 60, 120, and 240 minutes, we found that, with a constant biotin concentration, the number of proteins labeled by TurboID increased with longer biotin incubation times. Compared to the control group, TurboID could label proximal proteins after 60 minutes of biotin incubation. To prevent false positives caused by excessive biotin labeling due to prolonged incubation, we chose a moderate incubation time of 60 minutes (Fig. 3C). Therefore, we selected 60 minutes as the optimal labeling time for subsequent experiments.
By setting biotin incubation concentrations of 0, 50, 100, 150, 200, and 250 µM, we observed that the number of proteins labeled by TurboID did not increase with higher biotin concentrations. This may be because the TurboID system is limited, and excess exogenous biotin does not enhance TurboID activity. Compared to the control group, adding 50 µM of biotin allowed TurboID to label proximal proteins. To avoid non-specific binding caused by excess free biotin competing with streptavidin magnetic beads, which would reduce the specificity of the biotin-labeled proteins isolated by magnetic separation, we decided to use a biotin concentration of 50 µM for subsequent experiments (Fig. 3D).Through preliminary experiments, we confirmed the correct localization of PNO1-TurboID and NOB1-TurboID plasmids in cells, and determined the optimal biotin incubation concentration to be 50 µM and the optimal incubation time to be 60 minutes.
Capture and identification of biotinylated proteins
To demonstrate that the amount of endogenous biotin in cells is very low and that TurboID has low activity without the addition of a large amount of exogenous biotin, we set up a control group without added biotin. We performed Western blot analysis on the cell lysates. The results showed that, with equal sample loading, the control group without exogenous biotin had very few proteins labeled by TurboID, whereas the group with additional exogenous biotin had a large number of proximal proteins labeled by biotin (Fig. 4A).Using streptavidin magnetic beads, we enriched the biotinylated proteins and identified them through silver staining. The results indicated that the proximal proteins labeled by TurboID, PNO1-TurboID, and NOB1-TurboID in the presence of exogenous biotin were successfully enriched (Fig. 4B). We identified the enriched proximal proteins from these three groups using mass spectrometry, with the TurboID group serving as the control group. Each group was repeated three times. We processed the raw mass spectrometry data through database searches, initially screening for proteins with Unique Peptide ≥ 2 to obtain a preliminary protein set. Subsequently, we compared the overexpression groups to the control group (overexpression group/control group), conducting a secondary screening for proteins with Unique Peptide ≥ 2 to obtain the final protein set. These proteins were then subjected to enrichment analysis using KEGG, Reactome, and GO (Fig. 4C).
Functional analysis of PNO1 and NOB1 protein interaction groups
After identifying the proteins using high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS), we filtered the identified proteins based on criteria such as Unique Peptide, PSMs, protein abundance, and biotin modification to obtain a final protein set. We identified 2054 proteins interacting with PNO1 and 2363 proteins interacting with NOB1, with 1437 proteins common to both. KEGG enrichment analysis of proteins enriched by PNO1-TurboID and NOB1-TurboID revealed that the proteins were mainly involved in cell cycle, endocytosis, spliceosome, and ribosome functions (Fig. 5A-B). Reactome enrichment analysis showed that the proteins enriched by PNO1 and NOB1 were primarily involved in GTPase signaling, RNA metabolism, cell cycle, translation, and ribosomal RNA processing (Fig. 5C-D). Biological process analysis indicated that PNO1 and NOB1 were mainly involved in RNA metabolic processes, translation, and the assembly of protein-containing complexes. The enriched proteins were primarily localized in the nucleolus, microtubule cytoskeleton, and centrosome. The molecular functions of these proteins included adenosine nucleotide binding, mRNA binding, rRNA binding, and snRNA binding (Fig. 5E-F).
Functional analysis of the proteins enriched by PNO1 and NOB1 revealed significant functional similarities between the two. Despite their different localizations, PNO1 and NOB1 shared many common regulatory functions, such as mRNA translation, which are crucial for cell growth.
Protein interaction network of PNO1 and NOB1
The protein-protein interaction network is fundamental to understanding the intricate interactions and functional relationships between proteins within a cell. These networks illustrate how proteins interact physically and functionally, providing valuable insights into cellular processes and pathways. By mapping these interactions, researchers can elucidate the roles of proteins in various biological functions, such as signal transduction, metabolic pathways, and structural assembly.In this study, we constructed a protein-protein interaction network for the proteins interacting with both PNO1 and NOB1 using the CPDB database. The results revealed that PNO1 and NOB1 are primarily enriched in the pre-rRNA complex, mRNA translation complex, and IGF2BP1 complex(Fig. 6A).
PNO1 and NOB1 may be involved in mRNA translation regulation
mRNA translation is a crucial biochemical process within cells that converts genetic information from mRNA into proteins, which are the primary executors of cellular functions. The translation process itself is tightly regulated. By controlling the translation of specific mRNAs, cells can rapidly respond to endogenous and exogenous signals, enabling physiological changes and maintaining homeostasis. Through the biological functional analysis of mass spectrometry data from PNO1-TurboID and NOB1-TurboID, we surprisingly discovered that PNO1 and NOB1 share significant functional similarities and may be involved in the process of mRNA translation. Based on our mass spectrometry results, we identified translation-related proteins such as EIF4B and EIF4G2 (Fig. 7A-B). Analyzing the peptide segments of EIF4B and EIF4G2 through mass spectrometry database searches, we obtained the M/Z spectra for these proteins (Fig. 7C-D).To validate our mass spectrometry results, we performed co-immunoprecipitation experiments, which demonstrated that PNO1 and NOB1 interact with EIF4B and EIF4G2 (Fig. 7E-F). However, the mechanism by which PNO1 and NOB1 regulate the translation process remains unknown, and we look forward to further experimental investigations to explore this process.