1. Introduction
Lung cancer is a common type of cancer as well as a frequent cause of death in both men and women worldwide. In the UK in 2017, lung cancer was the third most common cancer (Cancer Research, UK. Available online:
https://about-cancer.cancerresearchuk.org/about-cancer/lung-cancer/about, accessed on 7 December 2021). Each year, more than a million new cases of lung cancer are diagnosed worldwide [
1]. Causes of lung cancer include smoking, air pollution, asbestos, and genetic elements [
2]. Lung cancer can be categorized into two types depending on its histological appearance. These types are non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). SCLC makes up about 20–25% of lung cancers [
3]. Moreover, the grading system (grades 1–3 on the basis of histology of the tumor) and the TNM staging (T-tumor size, N-node involvement, and M-distant metastases) are standardly used in the clinic to inform prognosis and therapy. Most patients are diagnosed at the stage of advanced disease and despite continued improvement in therapy, the five-year survival rate remains low [
4,
5].
The p53 tumor suppressor gene is frequently found mutated in lung cancer [
6,
7,
8]. Mutations of p53 lead to either inactivation or creation of p53 dominant-negative forms that often display oncogenic potential [
9,
10]. p53 is a transcription factor that mediates cellular response to stress controlling cell proliferation, promoting cellular response to DNA damage, DNA repair, and eliminating cells damaged beyond repair through apoptosis and other types of cell death [
7,
11]. If p53 or pathways involved in its regulation are altered by mutations, this results in uncontrolled cell division, increased mutations, and cancer. Given that p53 protein is pivotal for cancer progression, its function has been extensively researched [
12]. Numerous p53 mutations and its loss have been identified in lung cancer and suggested to contribute to poor survival rate [
13,
14,
15,
16].
p53 activity is regulated at several levels including protein stability and transcriptional control. P53 posttranslational modifications play an important role in this control and include phosphorylation, acetylation, ubiquitination, and several more. Phosphorylation of p53 plays a major role in its protein stability and transcriptional activation, notably controlling interaction with its negative regulator MDM2 that promotes p53 degradation [
17,
18,
19]. In particular, serine 46 phosphorylation directs the p53 protein to promoters of genes that regulate apoptosis [
20]. Transcriptional coactivators and histone acetyltransferases p300/CREB-binding protein (CBP) as well as MOZ were suggested to acetylate p53 at K382 [
21] and members of the sirtuin family including SIRT-1 were implicated in deacetylation [
22]. The lysine acetylation at these sites is linked to p53 protein stability, transcriptional activation, and the p53 mediated control of cell cycle arrest and apoptosis. Furthermore, p53 can be modified in many ways by various proteins at different locations within p53 in an independent manner, which contributes to the fine tuning of p53 protein activity [
23].
Another level of p53 activity control includes transcriptional cofactors. The histone acetyltransferase (HAT) p300 is a p53 coactivator that acts by acetylating p53 and thereby increasing p53 activity. TTC5/STRAP is a p300 interacting protein that is found in a complex with JMY and p300 and leads to p53 coactivation through p53 stabilization and increase in p53 mediated transcriptional activation [
24,
25,
26]. TTC5 is composed of six tandem tetratricopeptide repeat (TPR) motifs, which are 34 amino acid long motifs degenerate in nature, mediating protein–protein interaction and are found in a variety of proteins [
27]. TTC5/Strap is a stress responsive co-chaperone involved in the modulation of p53, HSF1, GR, and AP1 transcriptional activity as well as p53 and GR protein stability [
28,
29,
30]. TTC5 interacts with mitochondrial ATP synthase to downregulate ATP production, and it has been reported to increase the apoptotic effects of mitochondrial p53 [
31]. TTC5 augments p300 mediated acetylation [
26]. TTC5 is involved in the DNA damage response and is phosphorylated by ATM on S203, which increases its nuclear accumulation, whereas Chk2 phosphorylation on S221 augments protein stability [
26,
32]. JMY and TTC5 are involved in cell adhesion and control the tubulin mRNA stability, thereby inhibiting actin nucleation and regulation of autophagy [
33,
34,
35,
36,
37]. Interestingly, TTC5 prevents apoptosis of acute myeloid leukaemia cells [
38].
In this report, the potential of total and modified p53, TTC5, and SIRT1 as biomarkers in lung cancer patients was analysed. Significant associations between clinicopathological features and expression of the studied proteins were identified and an unusual cytoplasmic location of acetylated p53 was observed. Results suggest that a combined analysis of p53 modification and cofactors such as TTC5 associations with clinicopathological features may provide new potential prognostic factors and drug targets in lung cancer.
3. Discussion
In this report, we demonstrate that p53, its modified versions, and the TTC5 cofactor are associated with several clinical parameters, suggesting that the p53 acetylation pathway has an important role in lung cancer development. Significant association was detected between the p53 total, S46, K382, and TTC5 protein expression levels and the grade of lung cancer. The SIRT1 protein expression was inversely associated with TTC5 expression and positively associated with p53 phosphorylated at S46. KM plots indicated that p53 was associated with worse survival, whereas TTC5 was associated with a better survival outcome.
Lung cancer is an important contributor to cancer related deaths in part due to a lack of biomarkers to detect early stages of this disease. In breast cancer, mutant p53 expression is detected in the early stages and signifies worse progression [
42]. In lung cancer, p53 tumour suppressor is often mutated and is an important regulator of response to DNA damage or elimination of damaged cells through the apoptotic process. Due to its pivotal role, it was extensively studied in lung cancer using the immunohistochemical method to detect p53 protein that was usually detected in the nucleus and suggestive of presence of mutations. There are over 60 studies investigating the prognostic value of p53 detection in NSCLC [
43]. However, the potential of p53 acetylation and TTC5 co-factors as new potential prognostic factors to improve patients’ stratification into high and low risk groups remains to be investigated.
Our results suggest that p53 total protein as well as its modified versions correlate with the grade of cancer, and total p53 positivity was associated with a markedly reduced patient survival. This is in line with other published reports indicating that mutated or overexpressed p53 detection was an indicator of poor prognosis for patients with adenocarcinoma [
43]. Unexpectedly, our report suggests that in some patients K382 acetylated p53 isoform was detected in the cytoplasm (
Supplementary Figure S1). This is surprising given the nuclear location of p53 protein in most reports. Our results suggest that studied cell lines show predominantly nuclear staining; in some cells, and under certain experimental conditions, residual cytoplasmic staining was observed when the K382 antibody was used in in the BEAS 2B cell line. Although the significance of this observation is not clear, acetylated p53 is known to have cytoplasmic functions, and its accumulation in the cytoplasm has been previously reported [
21,
44,
45,
46].
Many of the studies have examined p53 expression; however, the acetylation of mutant p53 has not been thoroughly investigated [
47]. This will be important to know as the vast majority of the tumours with high levels of p53 are actually expressing a mutant p53. Perhaps mutant p53 is more acetylated in the cytoplasm than the nucleus. Future experiments addressing mutant p53 status by sequencing and analysis of p53 posttranslational modifications should answer these questions. In addition, MDM2 and ARF14 have been shown to prevent cytoplasmic localization of acetylated p53, possibly suggesting that differential levels of MDM2 and ARF14 in these samples could determine cytoplasmic expression [
48]. It will be interesting to determine levels of MDM2 and ARF14 in parallel in future studies and to study the effect of acetylation on mutant p53.
Another unexpected finding was that more cases were positive for K382 p53 staining than for total p53 staining in males (139 vs. 127 cases,
Table 1). This discrepancy might be due to antibody specificity that potentially can recognize other antigens and not only p53. It is also possible the abovementioned observation is due to the difference in the DO-7 and K382 p53 specific antibodies affinities for the antigen. However, it could also point to the presence of N-terminal isoforms of p53. For total p53 levels, the well-established DO-7 antibody was used, which is a monoclonal antibody that recognizes a domain in the extreme N-terminus of p53, while K382 acetylation happens on the p53 C-terminus. N-terminal truncations of p53 (Δ40, or Δ133) have been reported to play a role in cancers [
49], but it is unknown whether they play a role in lung cancer specifically. Our results warrant further investigation into N-terminal truncated variants and acetylation in this cohort of patients.
An additional interesting finding in our manuscript is that acetylated p53 was proportionally higher in male patients than in female patients (
Table 1). Recently Haupt et al. demonstrated that there is sex-disparity in p53 mutations in cancers [
50]. In this work, it is suggested that negative p53 regulator genes expressed on the X-chromosome are involved in reduced survival of males from cancers. Perhaps some of these regulators are involved in p53 acetylation and could explain the findings in this manuscript.
One factor that connects p53 phosphorylation and acetylation is the p300 histone acetyltransferase. P53 phosphorylation status is important for p300 recruitment to p53 and p53 acetylation. This is through a multiprotein complex involving several other factors including JMY and TTC5 [
26]. TTC5 has been suggested to facilitate p300 mediated acetylation of p53. The results here show a significant and complex association of TTC5 protein expression with the grade, stage of cancer, and OS. TTC5 seems to be positively associated with tumor size and node involvement, whereas a potential association with metastasis is difficult to interpret due a difference in number of samples analysed in different categories. Furthermore, TTC5 was associated with a better overall survival. It is possible that this is due to different methods/histological analyses used in the survival and grading analysis indicating a need for larger sample investigations in future research. Several reports suggest that TTC5 controls elements of cytoskeleton actin and tubulin [
51], which are crucial components of migratory and metastatic potential of cancer cells potentially explaining association with metastatic and node status. TTC5 expression was also negatively associated with the intensity of SIRT1 protein expression. This inverse association with SIRT1 is in line with the role of TTC5 in facilitating p53 acetylation.
Overall, our results suggest that analysis of p53 posttranslational modifications, in combination with TTC5 expression, upon further testing on a larger number of patients, may provide further insight into the role of p53 pathway in lung cancer biology. These observations can be exploited to simultaneously target multiple pathways associated with p53 [
52,
53], including TTC5-mediated control of cytoskeleton, to overcome chemotherapy resistance to tubulin binding agents [
51,
54,
55]. In addition, given the potential association of TTC5 with better overall survival, upon confirmation of this finding on larger number of patients, this observation can be used to stratify patients into high or low risk group and inform future clinical decisions. Finally, TTC5 may have a role in the process of antigen presentation, and together with p53 posttranslational modifications, may be essential in the selection of patients for immunotherapy. Taken together, these observations will facilitate future clinical applications in the treatment of patients with lung cancer.
4. Materials and Methods
4.1. Patients and Samples
In total, 202 lung cancer patients’ TMA tissue samples were purchased from US Biomax, Inc., Derwood, MD, USA as formalin-fixed paraffin-embedded tissue sections. Additionally, another 48 lung cancer tissue sections were included in this study and obtained from Blackpool Teaching Hospitals NHS Foundation Trust (BTHNFT). Most of the 250 lung cancer tissue sections were adenocarcinomas and squamous cell carcinoma (93 adenocarcinoma (37.2%), 132 squamous cell carcinoma (52.8%), 1 carcinoid (0.4%), 3 atypical carcinoid (1.2%), 8 small cell carcinoma (3.2%), 3 large cell carcinoma (1.2%), 4 bronchioloalveolar carcinoma (1.6%), 3 mucinous adenocarcinoma (1.2%), and 3 adenosquamous carcinoma (1.2%)) (
Supplementary Table S5). In total, every tissue section was 4–5 µm thick with a 1.5mm diameter for the TMA cores on the slides. The tissue specimens in this study were used to identify the associations between the differentially expressed target proteins (total p53, p53 acetylated at K382, p53 phosphorylated at S46, SIRT1, and TTC5 proteins) and clinicopathological data. Clinical information was obtained by reviewing medical records, which included the patient’s age, gender, cancer histological grade, the TNM (primary tumor, lymph nodes, and metastasis) staging and stage groups I–IV. The 48 samples from BTHNFT had survival data. Ethical approval was granted by the University of Salford ethics committee and from the National Research Ethics Committee (NREC, LAMMA study).
4.2. Cell Lines
Human lung adenocarcinoma A549, an immortalized human lung epithelial cell line BEAS-2B, human squamous cell carcinoma H2170, and U2OS human osteosarcoma cancer cell lines were purchased from American Type Culture Collection (ATCC; Manassas, VA, USA); Mero-14 human mesothelioma cell line was a gift from Prof. Landi (University of Pisa). Cells were maintained at 5% CO
2 and 37°C in RPMI 1640 medium (Sigma Aldrich, Gillingham, UK), which contained 10% fetal calf serum (Gibco, Paisley. UK) and 1% penicillin/streptomycin 10,000 U/mL (Sigma Aldrich, Gillingham, UK). Cells were treated with 20 µM topoisomerase II inhibitor etoposide [
39] (Sigma Aldrich, Gillingham, UK) for 24 h.
4.3. Immunohistochemistry Approach (IHC)
Immunohistochemical staining was performed using a two-step indirect immunohistochemistry protocol. Histo-clear1 and 2 solutions were used to remove the paraffin wax followed by rehydration. For antigen retrieval, slides were placed in the microwave for 15–20 min at 310–440 W and then covered with the Tris-EDTA Buffer pH 9.0 (10 mM Tris Base, 1 mM EDTA, 0.05% Tween-20) or Tri-sodium citrate buffer PH 6.0 (10 mM Sodium Citrate, 0.05% Tween-20, pH 6.0), depending on the antibody used in the experiment. The blocking of endogenous enzymes was performed by incubating the slides with the 0.3% of hydrogen peroxide (H2O2) for 15 min. Samples that underwent the same procedure but with the addition of the IgG antibody were considered negative control.
The antibody for p53 (DO-7, 1:100 dilution) was purchased from DAKO (CA, USA), Phosphorylated p53 (Ser46 (1:100 dilution) was obtained from Abnova (Taipei, Taiwan), acetylated p53 (K382) (1:50 dilution) was from Abcam, Cambridge, UK. Antibodies for TTC5 (1:100 dilution) and SIRT1 (1:100 dilution) were ordered from Abcam (Cambridge, UK). The immunohistochemical staining patterns were reviewed by two researchers including a pathologist and calculated as staining intensities (0–3).
4.4. Immunofluorescence
The A549 cells were cultured on the slides as described previously [
26] and treated with 20 µM Etoposide for 24hrs. Cells were first washed 3 times with cold PBS, then fixed with 4% formaldehyde for 15 min, followed by 5 min incubation in 1% Triton/PBS, and then washed 3 times with PBS. Then, 1% BSA (Bovine Serum Albumin) was used for blocking procedure, for one hour. TTC5 rabbit polyclonal antibody and mouse monoclonal antibody against p53 acetylated at K382 were used. The TTC5 antibody (1:200 dilution) was mixed for one hour with K382 antibody (1:1000 dilution) in the 1% BSA. The slides were washed with 1% BSA, then incubated for an hour with secondary antibodies (green fluorescence for the anti-TTC5 antibody and red fluorescence for the anti-K382 antibody (dilution 1:1000)). DAPI staining was used to stain nuclei and slides were analysed using the Leica microscope.
4.5. TTC5 and p53 Coexpression Analysis
Data were obtained from the UCSC Xena browser (
https://xenabrowser.net/, accessed on the 7 December 2021). Expression data for TTC5 and TP53 were downloaded from the TCGA LUNG database. Using R, the expression levels of TP53 and TTC5 mRNA were plotted on a scatter graph using scatter.smooth. Using the same dataset, TTC5 expression was subdivided by samples with mutant or wild-type TP53. The expression data of TTC5 for each subgroup were then displayed as a boxplot using boxplot.
4.6. Statistical Analysis
SPSS version 25 and R version 3.6.2. were used for statistical analysis. The following R packages were used in R script for the survival analysis and further PH hypothesis verification: “survival”, “surviminer”, and “rcompanion”. Chi-square tests and Fisher’s exact tests were utilized to determine the association of expression of proteins with clinical parameters such as age, gender, tumor grade and disease stage, primary tumor size, and metastases to other organs and lymph nodes. Post-hoc pairwise tests of TTC5 expression at different stages were adjusted for multiple comparisons using Bonferroni corrected p-values. Statistical significance was determined based on a significance level of p-value ≤ 0.05.
4.7. Survival Analysis
All survival analyses were performed using the survival and survminer packages on the R platform [
56,
57]. Kaplan-Meier estimates of survival probability for sub-groups of parameters including p53, K382, S46, SIRT1, and TTC5 expression level, status, and K382 localization were calculated, and sub-group differences tested using a log-rank test. Kaplan-Meier curves were also plotted for visualization of these differences.
Univariate Cox Proportional Hazards models were fitted to each of the status/expression parameters under investigation to estimate the marginal strength of each association with survival, and those found to be significantly associated with survival (with p-value < 0.05) were further included in a multivariate Cox PH model. The assumption of proportional hazards was satisfied; hazard ratios for each parameter are presented along with their 95% confidence interval.