Pten and p53 Loss in the Mouse Lung Causes Adenocarcinoma and Sarcomatoid Carcinoma

Simple Summary Lung cancer is the world leading cause of cancer death. Therefore, a better understanding of the disease is needed to improve patient survival. In this work, we have deleted the tumor suppressor genes Pten and Trp53 in adult mouse lungs to analyze its impact on tumor formation. Double mutant mice develop Adenocarcinoma and Pulmonary Sarcomatoid Carcinoma, two different types of Non-Small Cell Carcinoma whose biological relationships are a matter of debate. The former is very common, with various models described and some therapeutic options. The latter is very rare with very poor prognosis, no effective treatment and lack of models reported so far. Interestingly, this study reports the first mouse model of pulmonary sarcomatoid carcinoma available for preclinical research. Abstract Lung cancer remains the leading cause of cancer deaths worldwide. Among the Non-Small Cell Carcinoma (NSCLC) category, Adenocarcinoma (ADC) represents the most common type, with different reported driver mutations, a bunch of models described and therapeutic options. Meanwhile, Pulmonary Sarcomatoid Carcinoma (PSC) is one of the rarest, with very poor outcomes, scarce availability of patient material, no effective therapies and no models available for preclinical research. Here, we describe that the combined deletion of Pten and Trp53 in the lungs of adult conditional mice leads to the development of both ADC and PSC irrespective of the lung targeted cell type after naphthalene induced airway epithelial regeneration. Although this model shows long latency periods and incomplete penetrance for tumor development, it is the first PSC mouse model reported so far, and sheds light on the relationships between ADC and PSC and their cells of origin. Moreover, human ADC show strong transcriptomic similarities to the mouse PSC, providing a link between both tumor types and the human ADC.


Introduction
Lung cancer remains the deadliest cancer condition worldwide [1]. Thus, improving patient survival is an unmet need, which emphasizes the absolute requirement to extend our current comprehension of the underlying mechanisms of the disease. Lung Adenocarcinoma (ADC) is the most common primary lung cancer seen and represents about 40% of all lung cancers [2]. It usually evolves from the mucosal glands. When feasible, complete tumoral resection is considered the best treatment option; however, targeted therapies (mainly tyrosine kinase inhibitors) and immunotherapies may induce clinical responses in suitable patient subgroups [3]. In sharp contrast to adenocarcinoma

Development of Adenocarcinoma and Pulmonary Sarcomatoid Carcinoma in Double-Deficient Lungs
Histopathological analyses of CMV-DKO and K5-DKO tumors led to the diagnosis of two main lung tumor types: ADC and PSC (Figure 2, Tables S1 and S2). Occasionally, lung squamous cell carcinoma was observed (Tables S1 and S2, Figure S3), and this was the primary tumor type for two Ad5-CMVcre-infected mice. Adenocarcinomas displayed glandular differentiation with characteristic acinar, papillary or lepidic patterns (Figure Mice were sacrificed when they showed signs of disease (which include weight loss, shortness of breath, lethargy, hunched posture or ruffled fur) and necropsy was performed ( Figure S1). CMV-DKO mice (n = 18) developed tumors with a latency of 7-22 months and a frequency of 39% ( Figure 1A-C, Tables S1 and S2). For the K5-DKO group (n = 19), the latency period was 20-26 months and the incidence was 26%. Nine uninfected Trp53 F/F ; Pten F/F littermates were followed up as controls and did not develop any sign of illness up to 26 months. Another three additional mice were treated with naphthalene (control + naphtha) and followed up without developing tumors or showing histological respiratory injury ( Figure 1B,C).
We treated DKO mice with naphthalene to explore the effects of naphthalene-induced lung injury on DKO-tumor development. Naphthalene-induced lung injury ablates Clara Cells ( Figure S2), exposes airways basal cells to external agents, increases Keratin 5 expressing tracheal basal cells ( Figure S2), is accompanied by proliferation of the principal lung cell types [26,27] and has been related to development of lung tumors [28]. DKO mice were pretreated with naphthalene and intratracheally injected with either Ad5-CMVcre (n = 22) or Ad5-K5cre (n = 19) to initiate tumorigenesis three days after lung injury. Three Trp53 F/F ; Pten F/F mice were treated only with naphthalene. Infected mice developed tumors with a latency of 5-22 months (CMV-DKO + Naphtha) or 10-25 months (K5-DKO + Naphtha) and an incidence of 59% or 47%, respectively ( Figure 1B,C).
Wide targeting of lung epithelial cells (by using Ad5-CMVcre) rendered a higher percentage of mice with tumors and reduced latency periods with respect to basal cell restricted targeting (Ad5-K5cre) of Pten and Trp53. DKO mice showed a higher incidence in the percentage of tumor-bearing mice depending whether or not naphthalene had been administered prior to adenovirus infection and irrespective of the targeted cell/adenovirus used ( Figure 1B,C). Ad5-CMVcre (but not Ad5-K5cre)-infected mice showed a significant difference in tumor-free survival depending upon naphthalene administration ( Figure 1B). Thus, naphthalene increased overall incidence of mice with lung tumors and accelerated tumor development after wide targeting loss of Pten and Trp53 in lung epithelial cells.

Development of Adenocarcinoma and Pulmonary Sarcomatoid Carcinoma in Double-Deficient Lungs
Histopathological analyses of CMV-DKO and K5-DKO tumors led to the diagnosis of two main lung tumor types: ADC and PSC (Figure 2, Tables S1 and S2). Occasionally, lung squamous cell carcinoma was observed (Tables S1 and S2, Figure S3), and this was the primary tumor type for two Ad5-CMVcre-infected mice. Adenocarcinomas displayed glandular differentiation with characteristic acinar, papillary or lepidic patterns ( Figure S3) [29]. PSCs showed the histological heterogeneity described in human patients [30] with pleomorphic, spindle, and giant cell carcinoma and carcinosarcoma variants were observed ( Figure S3). These histologic subtypes are useful for the recognition and pathology diagnoses of PSC, but they do not seem to have clinical or therapeutic value [4,31,32]. To our knowledge, this is the first time a pulmonary sarcomatoid carcinoma is reported in a mouse model.
Deletion of both Pten F/F and Trp53 F/F alleles in these tumors was observed by PCR ( Figure 3A), further confirmed by RT-qPCR ( Figure 3B), and consequently, no detection of the corresponding proteins PTEN and p53 was observed by immunohistochemistry ( Figure 3C). Tumors were further characterized by immunohistochemical staining (Figure 2). The Pten/Trp53-deficient ADCs expressed thyroid transcription factor1 (TTF-1), and Keratins K7 and K8 showed positive staining for pan-cytokeratin AE1-AE3 and were negative for the mesenchymal marker vimentin. Meanwhile, pulmonary sarcomatoid tumors developed upon inactivation of Pten and Trp53 expressed TTF-1 and vimentin, displaying weak staining of cytokeratins using the AE1-AE3 antibody and absence of keratins K7 and K8. Both tumors were negative for the epithelial marker p63 (marker for squamous cell carcinoma) and the neuroendocrine markers calcitonin gene related protein (CGRP) and Achaete-scute complex homolog-1 (MASH1/ASCL1), characteristic of neuroendocrine tumors ( Figure 2). S3) [29]. PSCs showed the histological heterogeneity described in human patients [30] with pleomorphic, spindle, and giant cell carcinoma and carcinosarcoma variants were observed ( Figure S3). These histologic subtypes are useful for the recognition and pathology diagnoses of PSC, but they do not seem to have clinical or therapeutic value [4,31,32]. To our knowledge, this is the first time a pulmonary sarcomatoid carcinoma is reported in a mouse model.  the mesenchymal marker vimentin. Meanwhile, pulmonary sarcomatoid tumors developed upon inactivation of Pten and Trp53 expressed TTF-1 and vimentin, displaying weak staining of cytokeratins using the AE1-AE3 antibody and absence of keratins K7 and K8. Both tumors were negative for the epithelial marker p63 (marker for squamous cell carcinoma) and the neuroendocrine markers calcitonin gene related protein (CGRP) and Achaete-scute complex homolog-1 (MASH1/ASCL1), characteristic of neuroendocrine tumors ( Figure 2). There was no difference between the tumor type developed regarding the adenovirus injected: both ADC and PSC arose from basal cells (Ad5K5cre-DKO) as well as from a variety of epithelial lung cells (Ad5CMVcre-DKO), such as Clara, Alveolar Type 1 and alveolar Type 2 cells [24,25], suggesting multiple cells of origin for both ADC and PSC. We observed a clear overlap in the histopathological features in the lungs following either Ad5-CMVcre or Ad5-K5cre administration only after naphthalene administration ( Figure  1D). DKO mice treated with naphthalene (irrespective of the adenovirus used to initiate tumorigenesis) had a higher incidence of PSCs carcinomas compared to non-treated mice There was no difference between the tumor type developed regarding the adenovirus injected: both ADC and PSC arose from basal cells (Ad5K5cre-DKO) as well as from a variety of epithelial lung cells (Ad5CMVcre-DKO), such as Clara, Alveolar Type 1 and alveolar Type 2 cells [24,25], suggesting multiple cells of origin for both ADC and PSC. We observed a clear overlap in the histopathological features in the lungs following either Ad5-CMVcre or Ad5-K5cre administration only after naphthalene administration ( Figure 1D). DKO mice treated with naphthalene (irrespective of the adenovirus used to initiate tumorigenesis) had a higher incidence of PSCs carcinomas compared to non-treated mice ( Figure 1D). The difference was found to be significant by a Fischer exact t-test, p = 0.001. Thus, naphthalene treatment favors the development of PSC over ADC in double-deficient lungs.
Metastatic lesions were present in 18% of Ad5-CMVcre and Ad5-K5cre naphthalene treated bearing-tumor mice (4 mice out of 22). Primary tumors in these mice (2 ADC and 4 PSC), were able to colonize distant organs, including the liver, and develop PSC. Metastatic lesions displayed histological features and vimentin expression characteristic of PSC ( Figure S4). This observation highlights the ability of Pten/Trp53 deficient mice after treatment with naphthalene to recapitulate the behavior of human PSC, which show metastasis to distal organs [33,34].

Akt Pathway Activation and Epithelial-Mesenchymal Transition (EMT) Occur in Pten and Trp53 Double-Deficient Mouse Tumors
We assessed the impact of Trp53/Pten deletion on Akt signaling. As expected for Ptendeficient tumors, Akt-pathway activity was elevated in both ADC and PSC Trp53/Pten double-mutant tumors ( Figure 4A). Consistently, p-Akt, mTOR, p70S6K and pS6 proteins were clearly detected in histological sections of the tumors, while control lungs were negative for Akt-P and showed weak expression of mTOR, p70S6K and pS6 proteins in pneumocytes ( Figure 4A). An increase in Akt and p-Akt proteins in tumors was further confirmed by Western blot ( Figure 4B).
Given the characteristics observed in the tumors obtained and as EMT processes have been involved in the carcinogenic mechanisms and evolution of PSC [35][36][37], we tested the expression of EMT transcription factors in lungs, ADCs and PSCs ( Figure 4D). The hallmark epithelial marker E-cadherin increased in ADCs while it decreased in PSCs. Opposite to this finding and consistent with an EMT process, Vimentin, Snai1 and Snai2 mRNA levels were decreased in ADCs and increased in PSCs. Similar to the mRNA levels, E-cadherin protein was seen in lung epithelial cells (mainly bronchiole) and detected in adenocarcinomas, while vimentin (positive in myoepithelial cells of control lungs) and snail proteins (negative in lung) were detected in PSC tumors by Immunohistochemistry ( Figure 4C). Thus, PSCs exhibit characteristics of EMT, including a reduction of E-cadherin and an increase of Vimentin, Snail and Slug expression, supporting that PSCs have undergone EMT.

Transcriptomic Analysis of Double-Deficient Pten and Trp53 Mouse Lung Tumors
To further characterize these lung tumors, we performed microarray analysis of both tumor types, mouse ADC (moADC, n = 7) and mouse PSC (moPSC, n = 6), from mice treated with both adenoviral vectors and with/without naphthalene, and lung tissue from untreated and uninfected littermates (moLUNG, n = 6) (Tables S3 and S4). The principal component analysis (PCA) grouped the samples into three sets, according to the histopathologic sample type (i.e., lung, ADC and PSC) ( Figure 5A). PSC was the group that displayed the highest intragroup heterogeneity. Naphthalene treatment or the initial targeted cell type did not seem to have an impact in the transcriptional characteristics of the tumors ( Figure 5B,C).
We compared the gene expression profiles of each tumor type with normal lung, to select for genes that were specifically deregulated in PSC or ADC ( Figure 5D). Next, we analyzed the enrichment in molecular pathways in the upregulated genes that were either shared or distinctive of moADC or moPSC. Gene ontology pathways significantly enriched in the upregulated genes common to moADC and moPSC were glycolytic processes, nucleoside diphosphate phosphorylation and cell proliferation, indicating higher metabolic activity and proliferation of these tumors versus normal lung ( Figure 5E; Table S5). In agreement, Gene Set Enrichment Analysis (GSEA) showed that both tumor types were enriched in glycolysis, E2F targets and G2M checkpoint hallmark signatures (Molecular Signatures Database, MSigDB, hallmark gene sets) compared to lung ( Figure S5A). Genes specifically upregulated in moADCs were consistent with a glandular phenotype, while genes specifically upregulated in moPSC revealed a more aggressive/undifferentiated phenotype ( Figure 5E). In fact, when we directly compared moADC with moPSC using GSEA, moPSC were significantly enriched in hallmark gene sets of epithelial-mesenchymal transition, inflammatory response, angiogenesis and interferon-gamma response and depleted in fatty acid metabolism genes ( Figure S5B).

Transcriptomic Analysis of Double-Deficient Pten and Trp53 Mouse Lung Tumors
To further characterize these lung tumors, we performed microarray analysis of both tumor types, mouse ADC (moADC, n = 7) and mouse PSC (moPSC, n = 6), from mice treated with both adenoviral vectors and with/without naphthalene, and lung tissue from untreated and uninfected littermates (moLUNG, n = 6) (Tables S3 and S4). The principal  (D)). The Venn diagram shows genes common to both tumor types (133) or specific of moADC (533) or moPSC (498). Hypergeometric test was used to assess the statistical significance of the overlap (p < 10 −100 ). The rectangular boxes contain the main signaling pathways enriched in the indicated groups (gene ontology biological processes). p-values in brackets.
To compare the gene expression profiles of the lung tumors from our DKO mouse model with human lung adenocarcinomas (huLADC), we developed two huLADC gene signatures based on the RNAseq data available from Gillette et al. [38] and the Tumor Cancer Genome Atlas (TCGA Lung Adenocarcinoma). Gillete et al.'s study included 110 treatment-naive human LADC tumors and 101 paired normal adjacent tissue (NAT). TCGA data included 517 LADC and 59 NAT (see materials and Methods section). Within each study, huLUAD samples were compared to NAT and significantly upregulated genes were selected ( Figure 6A, Table S6). We compared these two huLADC gene sig-natures with two other signatures for human LADC form the MSigDB, FALVELLA and HP_LUNG_ADENOCARCINOMA, to look for common genes ( Figure 6B). Gillette and TCGA huLADC signatures shared > 60% of their genes; however, the overlap of the MSigDB huLADC signatures with Gillette and TCGA was less than <0.1% and was not significant. genes were selected ( Figure 6A, Table S6). We compared these two huLADC gene signatures with two other signatures for human LADC form the MSigDB, FALVELLA and HP_LUNG_ADENOCARCINOMA, to look for common genes ( Figure 6B). Gillette and TCGA huLADC signatures shared > 60% of their genes; however, the overlap of the MSigDB huLADC signatures with Gillette and TCGA was less than <0.1% and was not significant. Next, we selected the common genes within the top 150 upregulated genes in TCGA and Gillette huLADC signatures to create a TCGA-Gillete_huLADC gene signature (Figure 6A). We used this signature, the FALVELLA and HP_LUNG_ ADENOCARCINOMA, and other signatures for small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC) from the MSiGDB (Table S6, Figure S6) to analyze enrichment in the mouse tumors (moADC and moPSC) versus normal lung (moLUNG). The human LADC signature developed here, TCGA-Gillete_huLADC, but not signatures for other lung cancer types Next, we selected the common genes within the top 150 upregulated genes in TCGA and Gillette huLADC signatures to create a TCGA-Gillete_huLADC gene signature ( Figure 6A). We used this signature, the FALVELLA and HP_LUNG_ ADENOCARCINOMA, and other signatures for small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC) from the MSiGDB (Table S6, Figure S6) to analyze enrichment in the mouse tumors (moADC and moPSC) versus normal lung (moLUNG). The human LADC signature developed here, TCGA-Gillete_huLADC, but not signatures for other lung cancer types (small cell lung cancer), was significantly enriched in our mouse tumors, indicating that the expression characteristics of the mouse tumors developed by the inactivation of the Trp53 and Pten genes are similar to human LADC ( Figure 6C). The MSigDB HP_LUNG_ ADENOCAR-CINOMA signature showed negative enrichment in mouse tumors. This could be partly due to the small gene size (n = 17) of the signature and the fact that it includes genes commonly mutated and/or deleted in lung cancer such as TP53 and KRAS. Similarly, we did not find enrichment in our mouse tumors of WP (WikiPathways) or KEGG (Kyoto Encyclopedia of Genes and Genomes) non-small cell lung cancer signatures, which include ADCs together with SCCs and large cell carcinoma. These are curated gene signatures that also include genes, such as KRAS, TP53 and RB1. As for the FALVELLA signature, it is the only one that was created using analysis of gene expression data; however, it used fewer samples (24 human LADC and 24 NAT) than TCGA-Gillette and was designed to distinguish smokers from non-smokers in addition to tumor versus non-tumor [40]. Interestingly, the TCGA-Gillette human ADC gene signature was significantly enriched in the mouse PSC tumors compared to mouse ADC tumors ( Figure 6C,D), highlighting the similarity between moPSC tumors and human LADC.

Discussion
We disrupted Pten and Trp53 with two different cre-deleter lines: CMVcre (which targets all types of lung epithelial cells) [24] and K5cre (which targets specifically airway basal cells) [25]. The histology, immunohistochemistry and transcriptomic analysis revealed that tumors driven by these two promoter-cre lines were indistinguishable, even after naphthalene induced lung injury. This is in contrast to our previous work in which the targeted cell initiating tumorigenesis determines the type of high-grade neuroendocrine lung tumor developed when Pten and Trp53 along with Rb1 and Rbl1 are ablated [20,25]. However, it sheds light on the hypothesized multiple cells of origin of PSCs. Recently, Yang et al. [11], in their molecular characterization of a good-sized cohort of PSC patients, inferred that the cell origin of this tumor could be similar to that of adenocarcinoma, which has been reported to originate from Clara cells, alveolar epithelial cells and basal cells [13,21,41]. The work described here further supports their hypothesis as: (i) mouse models develop PSC (and ADC) arisen from basal cells (Ad5K5cre-DKO) as well as from other lung epithelial cells (Ad5-CMVcre-DKO); (ii) comparative genomics with a signature of human ADC clearly identified both mouse tumors as ADC.
PSC is a term comprising different histological subtypes with different morphology, suggesting heterogeneity [30]. Patients often exhibit tumors with combined carcinomatous and sarcomatoid components [42]. The coexistence of (well-differentiated) epithelial and (poorly differentiated) sarcomatoid cells has led to the hypothesis that PSC might represent an epithelial neoplasia undergoing divergent tissue differentiation [6,11,36]. PSC cells are likely derived from the epithelial-mesenchymal transition. The current work shows that, after combined deletion of Pten and Trp53 in lung epithelial cells, the PSC tumors developed undergo an EMT process, unlike the ADC tumors (arisen when the same set of genes are deleted in the same targeted cells) that preserve their epithelial nature. In fact, analysis of the transcriptome profiles of the mouse tumors revealed that PSC show hallmark features of EMT. These data are in line with an increasing number of studies based on the hypothesis that pulmonary sarcomatoid cells may be derived from carcinoma cells through the activation of an EMT process that leads to sarcomatous transformation of the carcinoma cells [11,36,43,44]. This also highlights the potential of targeting EMT in the treatment strategies for PSC.
The adeno-cre intratracheal infection has proven to be a robust method of modeling lung cancer in mice [19,21,45]. Both Ad5-CMVcre and Ad5-K5cre viral vectors have been successfully used in the generation of lung tumor in mice [20,25,46]. Probably, Ad5-CMVcre is the most widely adenoviral vector used for this purpose [21,47]. Using this viral vector for wide targeting of lung epithelial cells, the Pten-p53 deletion gives rise to ADC and PSC proving that this combination of tumor suppressors acts as genetic drivers of both tumoral types. However, tumors develop after long latency periods with incomplete penetrance. Naphthalene treatment slightly accelerates the process, suggesting that additional mutations are needed. Lung ADCs are frequently characterized by different oncogenic driver mutations that affect a variety of kinases and their downstream signaling pathways [48][49][50][51]. In fact, both Pten and Trp53 have been reported to accelerate Kras lung ADC formation [45,[52][53][54]. Unfortunately, very little is known about the molecular events underlying development of PSC and the potential driver mutations characterizing this tumor. As an example, given the frequency of actionable MET gene mutations described for PSC patients [7,55], it could be interesting to generate a model approaching loss of Pten and Trp53 along with the reported MET mutation. Other novel mutations identified as potential candidates in the molecular pathogenesis of PSC, such as CDH4, CDH7, LAMB4, SCAF1 and LMTK2 [7] are worth considering. These aspects could be relevant in the context of human PSC tumor characterization and approaching novel preclinical therapies and would deserve future investigations. Interestingly, ablation of Pten and Trp53 along with Rb1 [56,57] or Rb-family members [34] gives rise to high grade neuroendocrine tumors, supporting the role of pRb in neuroendocrine differentiation in a context of Pten and Trp53 loss.
The role of Pten and Trp53 in tumorigenesis has been analyzed in diverse tissues through different genetic strategies in genetically engineered mouse models. Loss of Pten and Trp53 rendering adenocarcinoma progression to sarcomatoid carcinoma due to EMT transformation has been described in a murine cancer prostate model [58]. Combined inactivation of Pten and Trp53 induces sarcomatoid Triple Negative Breast Cancer with enhanced features of EMT. A lower proportion of these tumors exhibit differentiated adenocarcinoma or mixed sarcomatoid plus adenocarcinoma tumors [59]. These similarities could represent a common model to explain the role of EMT in the evolution to sarcomatoid characteristics rendering a highly aggressive form of cancer. However, cell cycle regulation was found to be the driving force of liposarcoma formation or thymic lymphomas when Pten and Trp53 were deleted in adipose tissue [60] or thymus [61], respectively, indicating differences in tissue susceptibility. Deregulation of mTOR (bladder) or activation of Notch signaling (smooth muscle) have also been described as mechanisms underlying tumor formation in the absence of Pten and Trp53 [62,63]. Notably, basal-but not non-basal-cell type-restricted deletion of these genes in urothelial cells gives rise to muscle-invasive bladder tumors [64], allowing progression of bladder cancer in the context of inactivation of Pten and Trp53. Collectively, these data point to combined Pten and p53 exerting diverse functions/activating diverse pathways in tumor progression in a tissue-specific manner.
Mouse models of lung cancer provide critical insights into disease mechanisms [19,[21][22][23], and while a number of different ADC models have been described [21,65], there is a critical need for translational PSC models that recapitulate human disease and provide opportunities for tumor characterization and pre-clinical testing. In addition, primary tumor cells isolated from these PSC tumors constitute a high valuable tool for preclinical use. It has been suggested that PSCs are transformed or dedifferentiated variants of conventional Non-Small Cell Lung Carcinoma (NSCLC) [4]. However, little is known about the biological relationships between both tumor types, and they are a current matter of debate. The Pten/p53 double-deficient mice described here develop both ADC and PSC, providing an excellent tool for this purpose, and suggests a relationship between both tumor types, supported by human-mouse transcriptomic analyses. Furthermore, as far as we know, no models of PSC have been previously reported. Although the CMV and K5-DKO mice have incomplete penetrance and long latency periods for tumor development even after naphthalene treatment, the model described here shows the role of Pten and Trp53 as gene drivers of this type of tumor and the potential cells of origin from which PSC arises. Thus, the development of a murine model of PSC is significant, given the lack of models for this disease.

Mice and Adenoviral Infections
The Trp53 F/F ; Pten F/F mice were generated by breeding Rb1 F/F , Rbl1 −/− , Pten F/F and Trp53 F/F mice [25,66] with FVB/NJ mice (purchased to the Jackson Laboratory, Strain #001800). All animal experiments were approved by the Animal Ethical Committee and conducted in compliance with the CIEMAT guidelines. Specific procedures were approved by Comunidad Autónoma de Madrid (ProEX 208/15; ProEX 111.1/21).
Ablation of Trp53 and Pten in pulmonary cells was achieved by intratracheal administration of 10 8 plaque-forming units of Ad5-CMVcre and Ad5-K5cre to 8-10-week-old mice [45]. Adenoviruses Ad5-CMVcre and custom-made Ad5-K5cre were obtained from the Viral Vector Production Unit of the Autónoma University of Barcelona [25]. As control animals, uninfected Trp53 F/F ; Pten F/F littermates were used. Mice were sacrificed 5 to 29 months after the adenoviral infection. Mice were sacrificed when they showed any symptom of respiratory disease or sign of illness (labored breathing, lethargy, hunched back, ruffled hair or 10-15% loss of median body weight).

Naphthalene Treatment
Naphthalene solution (20 mg/mL) was prepared dissolving naphthalene (Sigma-Aldrich, St. Louis, MO, USA) in corn oil (Sigma-Aldrich, St. Louis, MO, USA) by gentle rocking at room temperature for 60 min and passed through a 0.2 mm filter to remove any undissolved solute. A single dose of naphthalene was delivered to adult mice by intraperitoneal injection (200 mg naphthalene per kg body weight) three days before intratracheal administration of adenovirus. As control animals, Trp53 F/F ; Pten F/F littermates were given a dose of naphthalene.

Genotyping
Genomic DNA was isolated from Trp53 F/F ; Pten F/F control lungs and tumors using DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA, USA). Primers sequences, amplified fragments and PCR amplification product sizes are in Table S7. Fabpi gene was used as loading control of samples.

Histology and Immunostaining/Immunohistochemistry
At necropsy, lungs were perfused with 4% formaldehyde. Samples were fixed in 4% buffered formalin and embedded in paraffin wax. Sections (5 µm) were stained with hematoxylin and eosin (H/E) for histological analysis or processed for immunohistochemistry. Immunohistochemical analyses were performed essentially as in previously described standard protocols [67,68]. Antibodies used are listed in Table S8.

RNA Extraction and RT-qPCR
RNA was isolated from whole mouse lungs in control mice and tumors using RNALater (Ambion Inc., Austin, TX, USA) and miRNeasy Mini Kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's instructions (control lungs n = 3; ADC n = 5; PSC n = 4). Genomic DNA was eliminated from the samples by a DNase treatment (Rnase-Free Dnase Set, Qiagen GmbH, Hilden, Germany). The Omniscript RT kit (Qiagen GmbH, Hilden, Germany) and oligo dT primers were used to prepare cDNA from RNA of the mouse samples, using 2 µg of total RNA. Real-time quantitative PCR was done on a 7500 Fast Real-Time PCR system (Applied Biosystems, Foster, CA, USA) with the GoTaq qPCR Master Mix (Promega, Madison, WI, USA), using 1 µL of cDNA (as a template). Each sample was normalized using the values for the TATA binding protein gene (Tbp). The sequences of the specific oligonucleotides used are listed in Table S9. Discrimination between samples showing increased or decreased relative expression was made using the Mean ± SEM.

Transcriptome Analyses
Total RNA was isolated from normal lungs and tumors as described above. RNA yield and quality were determined using Agilent 2100 Bioanalyzer (RNA Integrity Number RIN > 8) (Agilent, Santa Clara, CA, USA). A total of six normal lung (from Trp53 F/F ; Pten F/Funinfected mice), seven ADC and six PSC samples (Table S3)  . TAC was also used to identify genes differentially expressed ADC and PSC versus normal lung, selecting genes with a false discovery rate (FDR) threshold of ≤0.05 and an expression fold change (FC) of ± 2. A Venn diagram (https://bioinformatics.psb.ugent.be/webtools/Venn/ (accessed on 27 May 2021)) was used to represent the significantly upregulated genes in ADC or PSC tumors versus control lung. DAVID, annotation database software (http: //david.abcc.ncifcrf.gov/home.jsp (accessed on 3 June 2021)) [69], was used to identify Gene Ontology Biological Process (GOBP) functional categories of genes upregulated only in ADC, only in PSC or in both tumor types.

Statistical Analyses
Comparisons between two groups were performed using Student's unpaired t-test or Mann-Whitney test depending on the normal distribution of the data. Tumor-free survival analyses were performed using the Kaplan-Meier method and statistical differences between the two groups were tested by the log-rank test. Contingency analyses were performed using Fisher's exact test. Statistical significance was accepted at p < 0.05. GraphPad Prism 6.0, 9.0 software was used.

Conclusions
This work shows that combined deletion of Pten and Trp53 in mouse lung leads to the development of ADC and PSC, irrespective of the targeted cell type in which the gene alterations initially occur at least after naphthalene treatment. Naphtalene-induced epithelial airway injury prior to the inactivation of Pten and Trp53 favors the development of PSC. The tumors originated show transcriptomic similarities to the human ADC reported data, supporting a relationship between both tumor types.
To our knowledge, no PSC models have been described so far. Given the paucity of human material and the lack of models for this disease, the development of a model for PSC provides a valuable tool for the understanding of the disease and the development of novel therapeutic approaches.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cancers14153671/s1, Figure S1: I Representative images of DKO mice; Figure S2: Naphthalene induced lung injury in Trp53 F/F ; Pten F/F -uninfected mice; Figure S3: Histology of tumors developed in DKO mice; Figure S4: Development of metastases in doubledeficient mice; Figure S5: Hallmark Gene Set Enrichment Analysis (GSEA) of mouse tumors and normal lung; Figure S6: Venn diagram of human lung cancer signatures; Table S1: Lung tumors in conditional mutant Trp53 F/F ; Pten F/F mutant mice; Table S2: Type and number of tumors per mouse; Table S3: Samples used in the transcriptome analysis of lung, ADC and PSC; Table S4: Gene expression data of mouse lung, moADC and moPSC samples; Table S5: Genes differentially regulated in moADC tumors versus moLung and in moPSC tumors versus moLung; Table S6: Human lung ADC signatures; Table S7: Genotyping primers, amplified fragments and PCR amplification product sizes; Table S8: Primary antibodies used in the immunohistochemistry analyses; Table S9: Primers for quantitative RT-PCR analysis of gene expression; File S1: Original image of PCR and western blot.

Informed Consent Statement: Not applicable.
Data Availability Statement: All the microarray data are available at Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/; accession number GSE199905 (access release date 30 September 2022)).