Loss of RPS27a expression regulates cell cycle, apoptosis, and proliferation via the RPS27a-RPL11-MDM2-p53 pathway in lung adenocarcinoma cells

Some ribosomal proteins (RPs) might regulate the MDM2–p53 loop by binding to RPL5 or RPL11. This study aimed to explore whether ribosomal protein S27a (RPS27a) interacted with the ribosomal protein L11 (RPL11) to regulate p53 in lung adenocarcinoma (LUAD) cells. RPL11-interacting proteins were identied using a proteomics approach. Co-immunoprecipitation (co-IP), docking analysis, GST-fusion and in vitro ubiquitination assay were used to analyze the interaction of RPS27a and RPL11. Cell cycle, apoptosis, cell invasion, cell viability and colony-formation assay were analyzed by knocking down RPS27a. The RPS27a mRNA expression in LUAD was analyzed based on the TCGA dataset and the RPS27a expression was detected by immunohistochemistry in LUAD samples. At last, the RPS27a and p53 expression were analyzed by immunohistochemistry in xenograft tumors by blocking RPS27a.

The immunoprecipitation of RPL11 was performed as described previously [20]. Brie y, 10 µg of rabbit RPL11 antibodies (Abcam, Cambridge, UK) and the same amount of rabbit IgG (no. P2173, beyotime, Shanghai, China) were added in A549 cell lysate of the experimental and control groups, respectively, followed by overnight incubation at 4°C. After elution and puri cation, the immunoprecipitates were separated, then stained with silver. The immunoprecipitation was digested and then analyzed using an LC-MS/MS (TripleTOF, AB Sciex, Boston, MA, USA) instrument, and the results were evaluated. Credibility ≥ 95% and unique peptides ≥ 1 were the criteria to identify proteins [21].

GST-fusion assay
His-tagged RPL11 expression plasmids were transfected in Escherichia coli BL21. The His-tagged RPL11 was puri ed using an Ni 2 + -NTA column (Qiagen, Shanghai, China) after expression in E. coli. The GSTfusion assay was conducted as previously described. Then, 50 µg GST-RPS27a or GST was mixed with glutathione Sepharose 4B beads (Sigma) and incubated with 20 µg puri ed His-RPL11 proteins. anti-S-Tag and GST antibodies were used to analyze protein interactions by IB [22]. Transient transfection, IB, and co-immunoprecipitation (co-IP) analyses A549 cells were inoculated in six-well plates and cultured in complete medium. When the cells grew to 70% con uence, the cells were transfected with siRNA. The transfected cells for 48 h were collected and treated with lysis buffer. Equal amounts of clear cell lysate (50 mg) were used for IB and co-IP analyses [23].

Immuno uorescence
The immuno uorescence assay was performed as described previously [20]. Cells were then scanned and analyzed using a confocal laser microscope (LSM, Carl Zeiss AG, Germany).

Cell cycle and apoptosis analysis
After 48 h of transfection, the analysis of cell cycle and apoptosis was performed as described previously [21,22]. Data of DNA content were collected using CellQuest and analyzed using a ModFit software program. The apoptotic cells were analyzed with a Flowsight imaging ow cytometer (Amnis/Merck Millipore, Darmstadt, Germany).

In vitro ubiquitination assay
In vitro ubiquitination experiments refer to previous studies using Ni 2 + -NTA puri cation method [24]. The Ni 2+ -NTA pulldown used for ubiquitination experiment, the bead-bound proteins were analyzed using IB.

Dissociation of ribosomal subunits and measurement of the subunit ratio
The dissociation of ribosomal subunits was performed as described previously [26]. Samples measured at 254 nm absorbance (Biocomp, CA) and quantitative analysis of ribosome peaks. The area under the curve from the lowest point to the lowest point of the 40S, 60S, and 80S peaks was calculated by summing the digital measurements.

Stably knockdown of RPS27a cells constructed
A stable knockdown of the RPS27a (RPS27a knockdown) cell line was achieved by lentiviral infection and drug screening. In brief, a 7.5 × 10 4 cells/mL suspension was made from A549 cells, and 2 mL of the suspension per well was seeded in a six-well plate. The virus was added 20 h after seeding the cells. The fresh medium was replaced with 2 µg/mL puromycin every 2-3 days..

Transwell cell invasion assay
Transwell chambers (Costar, USA were used for Transwell invasion assays, as previously described [27]. The IC 50 of doxorubicin (Dox) on A549 cells and its effect on the survival of A549 cell clones are shown in Supplementary File 3. The incubation was continued for 24 h. The cells were xed with 4% paraformaldehyde and stained with crystal violet solution..

Colony-formation assay
The colony-formation assay of A549 cells was performed as described previously [28]. The colony formation rate was calculated by dividing the number of colonies/number of cells × 100% by the control.

Docking analysis
Brie y, the protein-protein interaction module of Schrodinger software (Schrodinger 2015 suit) was used for analyzing RPL11 and PRS27a interactions. The three-dimensional crystal structures of the human 80S ribosome (PDBID: 4v6x) were extracted from the PDB database. The small-molecule 3D structures were docked from the x-ray crystal structure of RPL11 and PRS27a, and two proteins were extracted from the 80S ribosome. The ubiquitin and water molecules were removed from the two protein structures to simulate the interaction [29].

Tumor xenografts
The management and handling of animals complied with the administrative regulations of the Laboratory Animal Affairs Administration of the Ministry of Science and Technology of China (1988.11.14). The research on experimental animals was approved by the Ethics Committee of the Institute of Modern Physics, Chinese Academy of Sciences, and the Institutional Animal Care and Use Committee. Four-ve-old nude mice (female, weight 16-17 g, SPF level) were obtained from the Laboratory Animal Center of obio biotechnology (Shanghai, China). NC and RPS27a knockdown cells in the logarithmic growth phase were injected (2 × 10 6 ) subcutaneously into mice to establish a cell xenograft model. Based on the equation of the at spheroid [tumor volume = (short diameter) 2 × large diameter × π/6], the average tumor volume of each group was calculated and expressed in mm 3 [30].

Immunohistochemistry analysis
The LUAD samples and xenograft tumors were analyzed by immunohistochemistry was performed as described previously [30]. The sections were examined with an 80i Nikon optical microscope (Nikon, Tokyo, Japan). Image-Pro Plus software was also used to analyze the optical density of protein expression.

Statistical analysis
Prism 8 software (GraphPad Software, CA, USA) was used to analyze the data. Statistical differences were analyzed based on the Student's t test and on one-way analysis of variance test. p value of < 0.05 was represented statistically signi cant.

Results
RPS27a is a potential binding protein with RPL11 The untreated A549 cell lysate was used to identify endogenous RPL11-binding proteins by IP/MS to discover the potential RPL11-binding ribosomal proteins. The silver staining image of the binding proteins showed that a band was detected at about 18 kDa (Fig. 1A). The number of proteins identi ed in the IP protein sample of the band was 133. The protein-related information is shown in Supplementary File 4, among which 43 interactors were RPs (Fig. 1B). The combined degree of RPS27a was the highest. Next, we detected the expression of RPS27a in BEAS-2B, A549 and H460 cells (Fig. 1C), the results showed in expression of RPS27a in H460 and A549 cells was higher than BEAS-2B cell, and was highest in A549 cell (Fig. 1D). This results indicated that the high expression of RPS27a is associated with the tumor growth of non-small cell lung cancer. Moreover, a previous study showed the RPS27a was involved in the regulation of p53 levels. Therefore, the present study focused on the hypothesis that the RPS27a-RPL11 interaction might involve in p53 activation.

Correlation of A549 cell apoptosis with RPS27a expression
Inducing tumor cell apoptosis is a common strategy to inhibit tumor development. We have demonstrated that carbon ion radiation (CIR) induced nucleolar stress leads to reduction of RPL27a expression and promote spermatogonia apoptosis. Therefore, 4 Gy CIR were used to induce apoptosis of A549 cells. Then, the apoptosis of A549 induced by CIR (Fig. 1E) and expression of RPS27a was detected ( Fig. 1G). The results showed that the increased apoptotic ratio (Fig. 1F) and the decreased expression of RPS27a were time dependent (Fig. 1H). Therefore, the A549 cell apoptosis has correlation with the reduction of RPS27a expression.
Knockdown of RPS27a activated p53, promoted cell apoptosis, and inhibited cell proliferation Small RNA interference and overexpression plasmids were used to knock down and induce overexpression of RPS27a, respectively, so as to explore the relationship between RPS27a and RPL11, and p53 activation. The protein levels of p53, MDM2, p21, RPL11 and RPS27a increased in RPS27a siRNA-treated cells compared to NC cells ( Fig. 2A). Similar to the immunoblotting results, the immuno uorescence results also showed that the uorescence signal of RPL11 ( Fig. 2B) was enhanced in the nucleus and cytoplasm after knocking down RPS27a (Fig. 2C). The mRNA expression levels of genes increased in knockdown of RPS27a cells compared to NC cells (Fig. 2D). In addition, the knockdown of RPS27a promoted cell apoptosis, inhibited cell proliferation and accelerated G1/S cell cycle progression (Fig. 2E-G). Moreover, the increased p53 expression was relatively stable with an increased half-life (Fig. 2H). The aforementioned ndings suggested that the knockdown of RPS27a stabilized and activated p53, resulting in A549 cell apoptosis. Representative images of the ow cytometer are shown in Figures S1 and S2.

Knockdown of RPS27a induced nucleolar stress
The immuno uorescence technology was used to analyze the distribution of nucleolar integrity marker proteins, nucleolin (NCL) (Fig. 2J) and nucleophosmin (B23) (Fig. 2K) to further determine whether RPS27a knockdown would destroy the nucleolar structure of A549 cells. The results showed that RPL27a knockdown induced the movement of NCL and B23 into the nucleoplasm, indicating that the nucleolar stress was generated. Further, polysome pro les were compared to study the effect of RPS27a knockdown on the ratios between small and large ribosomal subunits in A549 cells (Fig. 2L). The 80S:60S ratio decreased and 80S:40S increased was observed in the RPS27a knockdown cells. Thus, the knockdown of RPS27a impaired the ribosomal pro les in A549 cells, which might be related to the changes in the nucleolus. These results indicated that RPS27a knockdown could change the ribosomal subunit ratio and generate nucleolar stress.
Knockdown of RPS27a induced RPL11-dependent p53 activation and p53-dependent cell cycle arrest The co-transfection experiment of RPS27a and p53 showed that the knockdown of p53 eliminated the increase in MDM2 and p21 protein levels (Fig. 3A), G1-phase arrest (Fig. 3B), and apoptosis ( Fig. 3C) induced by the knockdown of RPS27a. In addition, the knockdown of p53 eliminated the inhibition of cell growth and proliferation (Fig. 3D). In H1299 cells, the knockdown of RPS27a inhibited cell proliferation (Fig. 3D) and induced G1 phase arrest (Fig. 3B). Representative images of the ow cytometer are shown in Figures S3 and S4.
Deleting some ribosomal proteins induces ribosomal stress to activate p53, which is mainly mediated by RPL11. The RPS27a and RPL11 co-transformation experiments were performed to prove that knocking down RPS27a to activate p53 was also regulated by this mechanism. The results showed that the knockdown of RPL11 eliminated the increase in MDM2 and p21 protein levels (Fig. 3E), G1-phase arrest (Fig. 3F), and apoptosis (Fig. 3G) induced by the knockdown of RPS27a. In addition, the knockdown of RPL11 eliminated the inhibition of cell growth and proliferation (Fig. 3H). Representative images of the ow cytometer are shown in Figures S5 and S6. Therefore, the knockdown of RPS27a requires RPL11 to induce p53 upregulation and a decrease in cell proliferation and growth.

Knockdown of RPS27a stabilized p53
In addition, the co-immunoprecipitation assay showed that the knockdown of RPS27a enhanced the binding interaction between RPL11 and MDM2 (Fig. 4A). Therefore, the reduction of RPS27a was likely to enhance the interaction between RPL11 and MDM2, thereby inhibiting MDM2 and stabilizing p53. The A549 cells transfected with plasmid pIRES2-ZsGreen1-Hum-Ub-His and knocked-down RPS27a were constructed to analyze p53 ubiquitination to demonstrate that the RPS27a interacts with RPL11 and stabilizes p53. Figure 4B shows that RPS27a partially rescued the degradation of p53, and the expression of MDM2 and p53 increased in the combinations of His-Ub and knockdown of RPS27a, suggesting that RPS27a knockdown inhibited MDM2-mediated degradation of p53.

RPS27a interacted with RPL11
The small-molecule 3D structures were docked from the x-ray crystal structure of the RPS27a and RPL11 (Fig. 4C). RPL11 and PRS27a interactions were simulated using the protein-protein interaction module of Schrodinger software (Schrodinger 2015 suit). The 3D crystal structures of the human 80S ribosome (PDBID: 4v6x) were extracted from the PDB database. The structure and function of proteins were closely related to the hydrogen bonding between amino acids. The well-known nucleotide-binding residues are shown in (Fig. 4C).
His-RPL11 and GST-RPS27a deletion fusion proteins puri ed from bacteria to further prove that RPS27a directly interacted with RPL11 in vitro. Figure 4D shows that puri ed Tag-RPL11 and puri ed Tag-MDM2 were bound by puri ed GST-RPS27a protein, but not GST alone. These results demonstrated that RPS27a directly bound to RPL11 in cells.
Knockdown of RPS27a increases the sensitivity of A549 cells to actinomycin D (ActD) and doxorubicin (Dox) treatment The expression of proteins were detected in RPS27a-siRNA and then treated with ActD cells (Fig. 5A). p53 had similar half-life changes in control siRNA-and RPS27a siRNA-treated cells after 24 h of ActD treatment (Fig. 5B). The knockdown of RPS27a had the least effect on the level and stability of p53 protein (Fig. 5C). These results indicated that under the condition of nucleolar stress, the effect of RPS27a on p53 might be masked by other RPs.
This study evaluated the effect of knocking down RPS27a on DNA damage-induced p53 activation. The selection of Dox concentration was shown in Supplementary File 4. Figure 5D shows that the knockdown of RPS27a did not affect Dox-induced p53 activation. The half-life measurement showed that p53 induced small difference in control siRNA-and RPS27a siRNA−treated cells after 24 h of Dox treatment (Fig. 5E, F). The PCR results showed that knockdown of RPS27a had the least effect on the level of p53 target genes MDM2 and p21 (Fig. 5G). Therefore, these observations indicated that RPS27a was not involved in DNA damage-induced p53 stabilization. In addition, the study also analyzed cell apoptosis and cycle. The knockdown of RPS27a aggravated the Dox-induced G1 phase arrest (Fig. 5H). It also aggravated the suppression of the colony-forming (Fig. 5I, J), invasion (Fig. 5K, L), and migration abilities (Fig. 5M, N) in A549 cells.

RPS27a is a oncogene in LUAD
The RPS27a mRNA expression in LUAD was analyzed based on the TCGA dataset to determine the role of RPS27a expression in the tumorigenesis of LUAD. The expression of RPS27a showed a signi cant difference between 483 LUAD tissues and 59 normal tissues (Fig. 6A), but no difference in four different LUAD stages (Fig. 6B). More dismal overall survival with high RPS27a expression of LUAD patients showed in the Kaplan-Meier survival analysis (Fig. 6C) and disease-free survival (Fig. 6D). More dismal survival with high RPS27a expression of LUAD patients showed in the plotter survival analysis (Fig. 6E). A total of 16 LUAD specimens and 10 normal tissue clinical specimens were collected from Gansu Cancer Hospital to determine the correlation of RPS27a expression in clinical LUAD. Representative images of immunohistological analysis are shown in Fig. 6F. As shown in Fig. 6G, the RPS27a expression were signi cantly upregulated in LUAD tissues compared with normal tissues. The ndings proved for the rst time that the overexpression of RPS27a in patients with LUAD might contribute to the development and short survival of LUAD.
The stable knockdown of RPS27a cells were injected into the left forelimb muscle of BALB/c nude mice to explore the effects of RPS27a in cell proliferation and apoptosis, and tumor nodules were harvested 53 days after injection (Fig. 7A). Silencing RPS27a inhibited tumor growth in vivo (Fig. 7B, C). It showed a relatively weak intensity of RPS27a, MMP-9, Ki67, and E-cadherin staining in knockdown of RPS27a cells (Fig. 7D-K), and a strong intensity of p53 in knockdown of RPS27a cells (Fig. 7L, M). These results indicated that p53 increased apoptosis by blocking RPS27a, and resulted in the inhibition of A549 xenograft growth in nude mice.

Discussion
RPS27a is a ribosomal protein constituting the 40S small subunit of the ribosome and plays an important role in ribosome biogenesis [31] and overexpressed in chronic myeloid leukemia, colon, renal, breast cancers and lung adenocarcinoma [32]. This study found that ablation of RPS27a expression could induce cell cycle arrest and apoptosis of A549 cells, which indicated that RPS27a is involve in regulator of the RPL11-MDM2-p53 pathway. The knockdown of RPS27a increased the expression of RPL11, promoted the binding of RPL11 to MDM2, leading to p53 activation. Therefore, RPS27a was a key factor in maintaining the normal level of p53 in LUAD. Also, it was crucial in negatively regulating the apoptosis of LUAD. p53 is critical for regulating cell apoptosis and proliferation [33][34][35]. The activation of p53 is strictly regulated by its target gene product, an E3 ubiquitin ligase MDM2, forming an MDM2-p53 feedback loop [36,37]. Previous studies have shown that RPs can regulate p53 activation by inhibiting MDM2 activity, thereby affecting cell cycle progression and apoptosis [38]. This process is involved in regulating MDM2 binding by RPs to indirectly affect the negative feedback loop of MDM2-p53 [39]. Overexpressed RPs have similar functions as RPL11 and RPL5. They can bind to the central acid domain of MDM2, such as RPL23 and RPL26 [13], and RPS7 [11], which is followed by the inhibition of MDM2-mediated p53 ubiquitination and degradation. In fact, some low-expressed RPs, such as RPL22 [14], RPS19 [16], RPS14 [17], and RPL4 [18] can also activate p53. This process must be completed with the participation of RPL11 and RPL5 [19]. The RPL5/RPL11-MDM2-p53 ternary complex was the most classic model of RPs and p53 binding because RPL11 and RPL5 were nucleolar stress effectors and sensors; RPL5 and RPL11 could bind to MDM2 alone or to MDM2 [19]. Moreover, 5S rRNA formed the 5S ribonucleoprotein complex (5S RNP) to bind to MDM2. Many RPs indirectly induced p53 activation through RPL11 and RPL5 [40]. Therefore, the combination of RPs with RPL11 and RPL5 cannot be ignored. The low-expressed RPS27a may have the aforementioned function. RPS27a have been found overexpressed in renal, breast and colon carcinomas [41,42], a dramatic increase in the expression of RPS27a gene in an oncomouse model of hepatocellular carcinoma [43], it is essential role in the activation of cellular checkpoints via p53.
The cell cycle arrest and apoptosis caused by RPS27a knockdown were found to be RPL11 and p53 dependent. Co-transfection experiments proved that the knockdown of p53 eliminated the apoptosis and cell cycle arrest caused by the decreased expression of RPS27a, which was also observed in H1299 cells lacking p53, indicating that the apoptosis and cell cycle arrest caused by knocking down RPS27a was p53 dependent. In addition, the knockdown of RPS27a to activate p53 was RPL11 dependent because the knockdown of RPL11 eliminated the p53 activation caused by the knockdown of RPS27a, and attenuated cell cycle arrest and apoptosis induced by RPS27a knockdown.
This study found that the knockdown of RPS27a enhanced the binding of MDM2 and RPL11, and leading to the accumulation and activation of p53. The protein structures of RPL11 and RPS27a were constructed using homology modeling methods. The analysis of the results of protein-protein docking revealed that RPS27a and RPL11 formed a stable composite structure. GST experiments also con rmed that RPS27a had a binding relationship with RPL11. Taken together, these results suggested that RPS27a bound to RPL11, forming a complex with p53.
Studies exploring the role of RPS27a in stress-induced p53 activation have shown that RPS27a expression was regulated by the DNA damage response. ActD treatment showed that the knockdown of RPS27a had minimal effects on p53 stability (Fig. 5A-C). Although knocking down RPS27a could regulate p53 transcriptional activity in response to ribosomal stress, this effect was limited. Many nucleolar resident proteins were released to the nucleoplasm because the low dose of ActD destroyed the nucleolus and induced nucleolar stress. Therefore, the effect of RPS27a on p53 activity might be masked by other ribosomal proteins with the same effect. Dox treatment showed that the knockdown of RPS27a did not affect Dox-induced p53 activation, but had an impact on the degree of p53 activation. The half-life measurement showed that the half-life of p53 was similar in control and RPS27a siRNA-treated cells after 24 h Dox treatment (Fig. 5D-F). Therefore, the results indicated that RPS27a did not participate in DNA damage-induced p53 activation. The aforementioned results were different from previous ndings showing that RPS26 regulated p53 transcriptional activity under DNA damage [44]. This discrepancy needs further investigation.
This study found that the knockdown of RPS27a changed the ribosomal subunit ratio and induced nucleolar stress. It also enhanced the RPL11 expression and interaction between RPL11 and MDM2. This might be because the disruption of small subunits of ribosomes led to an enhanced translation of RPL11 through a 5 , -top-mediated mechanism [45]. The knockdown of RPS27a caused the upregulation of RPL11 mRNA expression, thus, proving the aforementioned speculation.
RPS27a also had the ability to bind to MDM2, and inhibited the activity of MDM2 and leading to p53 stabilization [46]. In addition, RPS27a was not only an inducer of p53 but also a direct transcription target of p53 [47]. The DNA damage response induced the overexpression of RPS27a to active p53 [47]. In this study, another important pathway by which RPS27a regulated p53 activation was added, namely, RPS27a-RPL11-MDM2-p53. Together with RPS27a-MDM2-p53, the study provided the two modes of action of RPS27a in regulating p53. This study and previous studies together con rmed that nucleolar stress was monitored by an RPs-RPL5-RPL11-mediated p53 surveillance system.

Conclusions
In summary, this study showed that RPS27a is a new regulator of the RPL11-MDM2-p53 complex. This study was novel in demonstrating that RPs bound to RPL11 to regulate the MDM2-p53 feedback loop (Fig. 8). This study con rmed that the RPS27a-RPL11-MDM2-p53 signaling pathway was important in regulating the growth of LUAD cells, suggesting that RPS27a might be a potential target in LUAD treatment.

Consent for publication
All authors have read and agreed to publish this manuscript.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.       size was calculated every 7 days from 5 days after implantation, ***P < 0.001 with Student's t-test analysis (n=6). (D, F, H, J, L) Immunohistochemistry analysis of tumors showed relatively weak staining of RPS27a, Ki67, MMP-9, and E-cadherin, but a higher expression of p53 in stable knockdown of RPS27a cell xenograft tumors, scale bars = 50 μm (×400 magni cation). (E, G, I, K, M) The quanti cation proteins (IOD/area) by digital image analysis, *P < 0.05, **P < 0.01, ***P < 0.001 with Student's t-test analysis (n=6). LUAD, Lung adenocarcinoma; NC, negative control; KD, knockdown.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.