Title: Activated Mutant p110α Causes Endometrial Carcinoma in the Setting of Biallelic Pten Deletion
Abstract: PTEN and PIK3CA mutations occur with high frequency in uterine endometrioid carcinoma (UEC). Although PTEN mutations are present in complex atypical hyperplasia and carcinoma, PIK3CA mutations are restricted to carcinoma. We generated mouse models harboring Pten loss and/or activated Pik3ca in the endometrial epithelium to investigate their respective roles in the pathogenesis of UEC. Presence of an activated mutant Pik3ca on the background of Pten loss led to aggressive disease, with 100% of mice exhibiting carcinoma. Expression of Pik3ca with E545K mutation alone was unable to cause hyperplasia or cancer in the uterus and did not activate Akt as effectively as Pten deletion in short-term cultures of mouse endometrial epithelium, likely explaining the lack of phenotype in vivo. We also report that nuclear localization of FOXO1 correlated with PTEN mutational status irrespective of the PIK3CA status in endometrial cancer cell lines. Furthermore, gene expression profiles resulting from Pten loss or activation of Pik3ca in primary mouse endometrial epithelial cells exhibit minimal overlap. Thus, Pten and Pik3ca have distinct consequences on the activation of the phosphatidylinositol 3-kinase pathway in endometrial epithelium and are likely to affect other nonoverlapping cellular mechanisms involved in the development and progression of the most common type of uterine cancer. PTEN and PIK3CA mutations occur with high frequency in uterine endometrioid carcinoma (UEC). Although PTEN mutations are present in complex atypical hyperplasia and carcinoma, PIK3CA mutations are restricted to carcinoma. We generated mouse models harboring Pten loss and/or activated Pik3ca in the endometrial epithelium to investigate their respective roles in the pathogenesis of UEC. Presence of an activated mutant Pik3ca on the background of Pten loss led to aggressive disease, with 100% of mice exhibiting carcinoma. Expression of Pik3ca with E545K mutation alone was unable to cause hyperplasia or cancer in the uterus and did not activate Akt as effectively as Pten deletion in short-term cultures of mouse endometrial epithelium, likely explaining the lack of phenotype in vivo. We also report that nuclear localization of FOXO1 correlated with PTEN mutational status irrespective of the PIK3CA status in endometrial cancer cell lines. Furthermore, gene expression profiles resulting from Pten loss or activation of Pik3ca in primary mouse endometrial epithelial cells exhibit minimal overlap. Thus, Pten and Pik3ca have distinct consequences on the activation of the phosphatidylinositol 3-kinase pathway in endometrial epithelium and are likely to affect other nonoverlapping cellular mechanisms involved in the development and progression of the most common type of uterine cancer. Endometrial carcinoma is a significant cause of morbidity and mortality worldwide, and in the United States, it is the eighth most common cause of cancer-related deaths in women.1Di Cristofano A. Ellenson L.H. Endometrial carcinoma.Annu Rev Pathol. 2007; 2: 57-85Crossref PubMed Scopus (179) Google Scholar Endometrial carcinoma is most often of endometrioid morphology, and is usually preceded by complex atypical hyperplasia (CAH). The development of both endometrial hyperplasia and uterine endometrioid carcinoma (UEC) is associated with unopposed estrogen stimulation and/or specific genetic alterations.1Di Cristofano A. Ellenson L.H. Endometrial carcinoma.Annu Rev Pathol. 2007; 2: 57-85Crossref PubMed Scopus (179) Google Scholar Approximately 30% to 80% of primary UECs harbor PTEN mutations,2Tashiro H. Blazes M.S. Wu R. Cho K.R. Bose S. Wang S.I. Li J. Parsons R. Ellenson L.H. Mutations in PTEN are frequent in endometrial carcinoma but rare in other common gynecological malignancies.Cancer Res. 1997; 57: 3935-3940PubMed Google Scholar, 3Salvesen H.B. Stefansson I. Kretzschmar E.I. Gruber P. MacDonald N.D. Ryan A. Jacobs I.J. Akslen L.A. Das S. Significance of PTEN alterations in endometrial carcinoma: a population-based study of mutations, promoter methylation and PTEN protein expression.Int J Oncol. 2004; 25: 1615-1623PubMed Google Scholar, 4Risinger J.I. Hayes K. Maxwell G.L. Carney M.E. Dodge R.K. Barrett J.C. Berchuck A. PTEN mutation in endometrial cancers is associated with favorable clinical and pathologic characteristics.Clin Cancer Res. 1998; 4: 3005-3010PubMed Google Scholar with a similar frequency detected in CAH. In contrast, PIK3CA mutations occur in 20% to 40% of UECs and are rarely, if ever, found in CAH. Phosphatase and tensin homolog (PTEN) and PIK3CA are main components of the phosphatidylinositol 3-kinase (PI3K)/AKT/PTEN pathway with opposing actions. PIK3CA encodes the p110α catalytic subunit of the PI3K complex and is considered an oncogene. As a component of PI3K, it is activated downstream of growth factor signaling, resulting in phosphorylation of phosphatidylinositol-4,5-phosphate (PIP2) to generate phosphatidlyinositol-3,4,5-phosphate (PIP3). PIP3 activates AKT, a protein kinase that regulates several downstream pathways that impinge on cell proliferation, cell growth, and apoptosis. In contrast, PTEN is a tumor-suppressor gene encoding a dual-specificity phosphatase, capable of dephosphorylating both proteins and lipids. Its most well-described activity is the conversion of PIP3 to PIP2, indirectly inhibiting the action of PI3K. Thus, by preventing AKT activation, PTEN inhibits cell proliferation.5Cantley L.C. The phosphoinositide 3-kinase pathway.Science. 2002; 296: 1655-1657Crossref PubMed Scopus (4780) Google Scholar, 6Song M.S. Salmena L. Pandolfi P.P. The functions and regulation of the PTEN tumour suppressor.Nat Rev Mol Cell Biol. 2012; 13: 283-296Crossref PubMed Scopus (469) Google Scholar Most PTEN mutations in CAH and UEC are localized to exons encoding the lipid phosphatase domain, resulting in loss of its ability to dephosphorylate PIP3.2Tashiro H. Blazes M.S. Wu R. Cho K.R. Bose S. Wang S.I. Li J. Parsons R. Ellenson L.H. Mutations in PTEN are frequent in endometrial carcinoma but rare in other common gynecological malignancies.Cancer Res. 1997; 57: 3935-3940PubMed Google Scholar, 7Risinger J.I. Hayes A.K. Berchuck A. Barrett J.C. PTEN/MMAC1 mutations in endometrial cancers.Cancer Res. 1997; 57: 4736-4738PubMed Google Scholar, 8Kanamori Y. Kigawa J. Itamochi H. Shimada M. Takahashi M. Kamazawa S. Sato S. Akeshima R. Terakawa N. Correlation between loss of PTEN expression and Akt phosphorylation in endometrial carcinoma.Clin Cancer Res. 2001; 7: 892-895PubMed Google Scholar PIK3CA mutations, E542K E545K (both in exon 9) and H1047R (in exon 20), are hotspots, found in endometrial carcinoma, as well as several other cancers and lead to constitutive activation of p110α.5Cantley L.C. The phosphoinositide 3-kinase pathway.Science. 2002; 296: 1655-1657Crossref PubMed Scopus (4780) Google Scholar Because loss of PTEN or activation of PIK3CA lead to activation of the pathway, it is reasonable to assume that they have similar downstream effects. However, as noted above, PTEN mutations are found in both CAH and UEC, whereas PIK3CA mutations are almost exclusively found in UEC, suggesting unique and nonoverlapping functions in endometrial carcinoma pathogenesis. Pten+/− female mice develop CAH by 32 weeks of age, and at approximately 52 weeks of age, 25% of the Pten+/− female mice develop UEC. Both CAH and UEC exhibit complete loss of Pten expression, as a result of biallelic Pten inactivation by either loss of heterozygosity or intragenic mutations in the wild-type allele.9Wang H. Douglas W. Lia M. Edelmann W. Kucherlapati R. Podsypanina K. Parsons R. Ellenson L.H. DNA mismatch repair deficiency accelerates endometrial tumorigenesis in Pten heterozygous mice.Am J Pathol. 2002; 160: 1481-1486Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar, 10Wang H. Joshi A. Iaconis L. Solomon G.J. Xiang Z. Verhage H.G. Douglas W. Ronnett B.M. Ellenson L.H. Oviduct-specific glycoprotein is a molecular marker for invasion in endometrial tumorigenesis identified using a relevant mouse model.Int J Cancer. 2009; 124: 1349-1357Crossref PubMed Scopus (7) Google Scholar In this mouse model, the frequency of biallelic inactivation is similar in CAH and UEC, and we have previously shown that mismatch repair deficiency hastens inactivation of the remaining Pten allele and promotes the progression of hyperplastic lesions to UEC.9Wang H. Douglas W. Lia M. Edelmann W. Kucherlapati R. Podsypanina K. Parsons R. Ellenson L.H. DNA mismatch repair deficiency accelerates endometrial tumorigenesis in Pten heterozygous mice.Am J Pathol. 2002; 160: 1481-1486Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar These observations suggest that, although biallelic inactivation of Pten is an early event in endometrial tumorigenesis, it may not be sufficient for progression to invasive carcinoma. Given the genetic analyses of primary human tumors, it seems likely that mutations in PIK3CA may promote progression to carcinoma. To determine the effect of loss of PTEN and/or activated PIK3CA in the development of UEC, we have developed several genetic mouse models. The results reported herein demonstrate distinct roles of these two genes in the pathogenesis of endometrial carcinoma. PtenloxP/loxP mice on a Balbc/129SvJ background and Ksp1.3-Cre+/− mice were obtained from The Jackson Laboratory (Bar Harbor, ME). To investigate the role of Pik3ca, a mouse line with a Cre-inducible mutant form of Pik3ca (designated as Pik3caE545K/+) was constructed in the laboratory of Dr. Suzanne Baker (St. Jude Children's Research Hospital, Memphis, TN). The Pik3caE545K/+ mouse contains a lox-STOP-lox cassette upstream of the Pik3ca gene containing the E545K mutation. In the presence of Cre, the stop codon is excised such that the mutant allele is expressed from its endogenous promoter. These mice were bred with PtenloxP/loxP strain to generate PtenloxP/loxP;Pik3caE545K/+ (Ptenf/f;Pik3caE545K), PtenloxP/loxP;Pik3ca+/+ (Ptenf/f), Pten+/+; Pik3caE545K/+ (Pik3caE545K), and PtenloxP/+;Pik3caE545K/+ (Ptenf/+;Pik3caE545K) mice. Mice expressing Cre under the Wnt7a promoter (Wnt7a-Cre+/−) were obtained from Dr. Kenneth S. Korach (National Institute of Environmental Health Science, Research Triangle Park, NC). Ptenf/f;Pik3caE545K mice were crossed with the Wnt7a-Cre and Ksp1.3-Cre mice to generate the described genotypes. All animal experiments were performed in accordance with Institutional Animal Care and Use Committee guidelines. The uteri harvested from all mice at indicated time points were fixed in formalin and embedded in paraffin. Sections (5 μm thick) were processed for hematoxylin and eosin staining and immunohistochemical (IHC) analysis. For IHC, the slides were deparaffinized, rehydrated, and subjected to antigen retrieval by boiling in the microwave in 10 mmol/L sodium citrate. The Vectastain ABC reagent (Vector Labs, Burlingame, CA) was then used according to the manufacturer's protocol. Antibodies used were Pten (1:100), phospho-Akt S473 (1:100), β-catenin (1:800), p-Gsk3β (1:50), p-Pras40 (1:800), Stathmin (1:50), and FoxO1 (1:100), all from Cell Signaling Technologies (Beverly, MA). Two human tissue microarrays (TMAs), one containing 100 cases of normal cycling endometrium and the second with 96 cases of UEC (24 International Federation of Gynecology and Obstetrics grade 1, 37 International Federation of Gynecology and Obstetrics grade 2, and 35 International Federation of Gynecology and Obstetrics grade 3 cases) generated at Weill Cornell Medical College (New York, NY) under Institutional Review Board approval were also stained with FOXO1 antibody, as per the protocol described above. Areas corresponding to CAH, UEC, or normal epithelium were microdissected from deparaffinized, rehydrated, hematoxylin-stained sections (5 μm thick) under light microscope visualization with a 26-gauge needle. The DNA was extracted using previously published protocols.11Hayes M.P. Wang H. Espinal-Witter R. Douglas W. Solomon G.J. Baker S.J. Ellenson L.H. PIK3CA and PTEN mutations in uterine endometrioid carcinoma and complex atypical hyperplasia.Clin Cancer Res. 2006; 12: 5932-5935Crossref PubMed Scopus (194) Google Scholar, 12Hayes M.P. Ellenson L.H. Molecular alterations in uterine serous carcinoma.Gynecol Oncol. 2010; 116: 286-289Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar Epithelial cells from uteri of Ptenf/f;Pik3caE545K, Ptenf/f, and Pik3caE545K mice were prepared as described previously.13Joshi A. Ellenson L.H. Adenovirus mediated homozygous endometrial epithelial Pten deletion results in aggressive endometrial carcinoma.Exp Cell Res. 2011; 317: 1580-1589Crossref PubMed Scopus (18) Google Scholar Briefly, the horns were slit lengthwise and digested in a solution containing 0.25% trypsin in Hanks' balanced salt solution for 1 hour at 4°C, followed by another incubation for 1 hour at 22°C. The tissue pieces were transferred to ice-cold Hanks' balanced salt solution and vortex mixed to release sheets of epithelial cells. The cell suspension was filtered through a 20 μm nylon mesh, and the epithelial sheets (retained on the filter) were collected and resuspended in Dulbecco's modified Eagle's medium/F12 (1:1), 10% charcoal-stripped fetal bovine serum, 20 mmol/L HEPES, 100 μg/mL streptomycin, 100 U/mL penicillin, and 2 mmol/L l-glutamine and insulin-transferrin-selenium supplement (1×). Cells were plated in 6-well dishes coated with 1:10 diluted Matrigel (Life Technologies, Grand Island, NY) and cultured at 37°C in an incubator with 5% CO2. After reaching 80% to 90% confluence, the cells were infected with adenoviruses expressing Cre or green fluorescent protein (GFP; control), as described previously,13Joshi A. Ellenson L.H. Adenovirus mediated homozygous endometrial epithelial Pten deletion results in aggressive endometrial carcinoma.Exp Cell Res. 2011; 317: 1580-1589Crossref PubMed Scopus (18) Google Scholar for 48 hours. Cells were harvested for RNA or protein extraction. The RNA was used for RNA sequencing analysis and for subsequent validation of differentially expressed genes (DEGs) by real-time quantitative PCR (qPCR). Whole cell protein lysates were used for immunoblot analysis using Pten, phospho-Akt, and Akt antibodies (Cell Signaling Technologies), all at 1:1000 dilution. Actin antibody (Sigma, St. Louis, MO) was used for normalization. Endometrial cancer cell lines Hec1A parental clone (Hec1A WT) and PTEN-deleted Hec1A (Hec1A clone 16) have been described previously.14Joshi A. Wang H. Jiang G. Douglas W. Chan J.S. Korach K.S. Ellenson L.H. Endometrial tumorigenesis in Pten(+/-) mice is independent of coexistence of estrogen and estrogen receptor alpha.Am J Pathol. 2012; 180: 2536-2547Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar KLE and RL95 cells were purchased from ATCC (Manassas, VA) and were cultured as per the ATCC protocol. Nuclear and cytoplasmic extracts were prepared using the NE-PER Nuclear Protein Extraction Kit from Thermo Fisher Scientific (Waltham, MA), as per the manufacturer's protocol. The extracts were resolved by 4% to 20% SDS-PAGE and transferred to a polyvinylidene difluoride membrane, and immunoblot analysis was performed using p-FOXO1 and FOXO1 antibodies. Lamin A/C and glyceraldehyde-3-phosphate dehydrogenase antibodies were used to ensure purity of the nuclear and cytoplasmic extracts, respectively. All antibodies were purchased from Cell Signaling Technologies and used at a dilution of 1:1000. Total RNA was extracted from the cells using the RNeasy Plus Universal Micro kit (Qiagen, Valencia, CA), as per the manufacturer's instructions. After isolation, total RNA integrity was checked using a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA), and samples with RNA integrity number value greater than eight were used for library preparation and sequencing. Library construction, template hybridization, and cluster amplification were performed using the TruSeq RNA Sample Preparation Kit version 2 (Illumina Inc., San Diego, CA). After cluster generation, sequencing by synthesis was performed on the HiSeq2000/1000 machine (Illumina Inc.). The RNA sequencing data were analyzed using Genesifter software (Geospiza, Seattle, WA). Raw reads from an Illumina HiSeq2000 in fastq format were filtered to remove any reads that contained primarily adaptor/primer sequences. Reads were then aligned using Burrows-Wheeler Aligner against the genomic reference sequence for Mus musculus (Build 37.2). The Mus musculus (Build 37.2) reference sequence and annotation were pulled from RefSeq database (http://www.ncbi.nlm.nih.gov/refseq; last accessed November 12, 2014), and reads mapping to the genome were characterized as exon, intron, or intergenic (outside any annotated gene). These aligned data were then used to calculate gene expression by taking the total of exon and known splice reads for each annotated gene to generate a count value per gene. For each gene, a normalized expression value was generated in two ways: reads per mapped million (RPMM; which is calculated by taking the count value and dividing it by the number of million mapped reads) and RPMM per kilobase (RPMM value divided by the kilobase length of the longest transcript for each gene). The RPMM values were used for comparing gene expression across samples to remove the bias of different numbers of reads mapped per sample. RPMM per kilobase values were used for comparing relative expression of genes to one another to remove the bias of different numbers of mapped reads and different transcript lengths. Samples were initially analyzed in a pairwise fashion with normalization by the number of million mapped reads (not including rRNA, tRNA, snRNA, or mitochondrial RNA), followed by likelihood ratio test, and Benjamini and Hochberg correction testing for significance via Bioconductor. After the statistical qualification of expression values, the data were filtered for genes that were threefold differentially expressed. A multiple sample analysis was also performed wherein GFP-treated (control) cells were compared to Cre-treated cells from Ptenf/f;Pik3caE545K, Ptenf/f, and Pik3caE545K genotypes. For this analysis, the data were filtered for genes that were also threefold differentially regulated, and a Kyoto Encyclopedia of Genes and Genomes pathway analysis was performed. We initially focused on genes in the endometrial cancer and Wnt signaling pathways. Validation of DEGs was done by qPCR. We first used cDNA prepared from the same RNA used for RNA sequencing as technical validation. For biological validation, RNA was extracted from adenovirus-Cre–treated primary cultures of epithelial cells from at least two independent experiments, as described in Materials and Methods. RNA (1.0 μg) was reverse transcribed using the High Capacity RNA-to-cDNA Kit (Life Technologies), as per the manufacturer's instructions. Volume of reverse transcriptase reaction corresponding to 25 ng cDNA was used to set up qPCRs using TaqMan primer-probe sets from Life Technologies in the StepOnePlus PCR machine, and fold differences in gene expression were calculated using the −ΔΔCT method. The epithelium-restricted expression levels of the Wnt7a15Winuthayanon W. Hewitt S.C. Orvis G.D. Behringer R.R. Korach K.S. Uterine epithelial estrogen receptor alpha is dispensable for proliferation but essential for complete biological and biochemical responses.Proc Natl Acad Sci U S A. 2010; 107: 19272-19277Crossref PubMed Scopus (180) Google Scholar and Ksp1.316Wild P.J. Ikenberg K. Fuchs T.J. Rechsteiner M. Georgiev S. Fankhauser N. Noske A. Roessle M. Caduff R. Dellas A. Fink D. Moch H. Krek W. Frew I.J. p53 Suppresses type II endometrial carcinomas in mice and governs endometrial tumour aggressiveness in humans.EMBO Mol Med. 2012; 4: 808-824Crossref PubMed Scopus (51) Google Scholar, 17Frew I.J. Minola A. Georgiev S. Hitz M. Moch H. Richard S. Vortmeyer A.O. Krek W. Combined VHLH and PTEN mutation causes genital tract cystadenoma and squamous metaplasia.Mol Cell Biol. 2008; 28: 4536-4548Crossref PubMed Scopus (36) Google Scholar, 18Shao X. Somlo S. Igarashi P. Epithelial-specific Cre/lox recombination in the developing kidney and genitourinary tract.J Am Soc Nephrol. 2002; 13: 1837-1846Crossref PubMed Scopus (247) Google Scholar promoters in the endometrium have been described previously. Therefore, Ptenf/f;Pik3caE545K mice were crossed with two mouse strains expressing Cre under Wnt7a (Wnt-Cre) or Ksp1.3 (Ksp-Cre) promoters to delete Pten or express mutant activated Pik3ca in the uterine epithelium. Uteri of Wnt-Cre+/−;Ptenf/+;Pik3caE545K, Wnt-Cre+/−;Ptenf/+, and Wnt-Cre+/−;Pik3caE545K mice were analyzed at 5 months of age. This time point was determined due to the development of lymphoma and breast cancers in many of the Wnt-Cre+/−;Ptenf/+;Pik3caE545K mice at this age. The Wnt-Cre and Ptenf/f;Pik3caE545K crosses never produced Wnt-Cre+/−;Ptenf/f;Pik3caE545K or Wnt-Cre+/−;Ptenf/f mice, likely due to embryonic expression of Wnt7a with subsequent lethality. Ten of 15 Wnt7a-Cre+/−;Ptenf/+;Pik3caE545K and 12 of 17 Wnt-Cre+/−;Ptenf/+ female mice developed CAH (Figure 1A and Table 1). Only two Wnt-Cre+/−;Ptenf/+;Pik3caE545K mice exhibited in situ carcinoma and/or myoinvasive carcinoma. We did not observe significant differences between the number of CAH foci and severity of disease between Wnt-Cre+/−;Ptenf/+ and Wnt7a-Cre+/−;Ptenf/+;Pik3caE545K genotypes. Interestingly, uteri of Wnt-Cre+/−;Ptenf/+;Pik3caE545K mice exhibited lymphatic dilation and hemorrhagic cysts (Figure 1B). Of 15 Wnt7a-Cre+/−;Ptenf/+;Pik3caE545K mice, 9 also developed lymphoma and breast lesions, likely due to expression of Wnt-Cre in these tissues. None of the 17 Wnt-Cre+/−;Ptenf/+ mice developed hemorrhagic cysts, lymphomas, or breast lesions. All Wnt-Cre+/−;Pik3caE545K mice (n = 7) exhibited normal uterine histological features at 5 months of age and also lacked hemorrhagic cysts, lymphoma, and breast lesions.Table 1Endometrial Lesions in Wnt-Cre+/−;Ptenf/+;Pik3caE545K Mice and LittermatesGenotypeNo. of miceMice with endometrial pathological featuresMice with CAHAverage No. of CAH fociMice with CA/InvBreast/lymph node pathological featuresWnt-Cre+/−;Ptenf/+;Pik3caE545K151086–729Wnt-Cre+/−;Ptenf/+1712127–800Wnt-Cre+/−;Pik3caE545K700000CA, carcinoma; CAH, complex atypical hyperplasia; Inv, invasion. Open table in a new tab CA, carcinoma; CAH, complex atypical hyperplasia; Inv, invasion. Mating of Ptenf/f;Pik3caE545K and Ksp1.3-Cre mouse strains generated Ksp-Cre+/−;Ptenf/+;Pik3caE545K, Ksp-Cre+/−;Ptenf/+, Ksp-Cre+/−;Ptenf/f, Ksp-Cre+/−;Ptenf/f;Pik3caE545K, and Ksp-Cre+/−;Pik3caE545K/+ genotypes. Previously published work from our laboratory demonstrated that Ptenf/f mice injected with adeno-Cre virus developed carcinoma between 2 and 4 months after injection.13Joshi A. Ellenson L.H. Adenovirus mediated homozygous endometrial epithelial Pten deletion results in aggressive endometrial carcinoma.Exp Cell Res. 2011; 317: 1580-1589Crossref PubMed Scopus (18) Google Scholar We, therefore, chose to sacrifice mice of the five genotypes mentioned above at 4 months of age (Table 2). Eleven of 12 Ksp-Cre+/−;Ptenf/+;Pik3caE545K and 12 of 16 Ksp-Cre+/−;Ptenf/+ mice exhibited CAH only. The number of foci of CAH was comparable in Ksp-Cre+/−;Ptenf/+;Pik3caE545K and Ksp-Cre+/−;Ptenf/+ genotypes, to those observed in the Wnt7a-Cre+/−;Ptenf/+;Pik3caE545K mice and the Wnt-Cre+/−;Ptenf/+ mice. However, in the Wnt7a-Cre+/−;Ptenf/+;Pik3caE545K genotype, two mice developed invasive carcinomas that were not observed in the Ksp-Cre+/−;Ptenf/+;Pik3caE545K mice. All Ksp-Cre+/−;Ptenf/f mice (n = 8) developed extensive endometrial disease, with CAH involving the entire luminal and glandular epithelium (Figure 1C). In five of eight mice, CAH exhibited squamous metaplasia (Figure 1D). Only one mouse (12%) showed carcinoma with myometrial invasion. Hematoxylin and eosin analysis of Ksp-Cre+/−;Ptenf/f;Pik3caE545K uteri (n = 6) showed CAH with squamous metaplasia and carcinoma with myometrial invasion (Figure 1, E and F) in all of the mice. In five mice, the gross examination of the uterus demonstrated enlarged uteri with ovaries engulfed by cysts that could not be distinguished from the oviduct (Figure 1G) as compared to Ksp-Cre+/−;Ptenf/f mice (Figure 1H). The carcinoma was well to moderately differentiated and extended into and completely surrounded the ovaries (Figure 1G). Thus, the presence of a mutant Pik3ca on the background of biallelic Pten deletion resulted in extensive invasive carcinoma with 100% penetrance. None of the Ksp-Cre+/−;Pik3caE545K mice developed endometrial disease (Figure 1, I and J), similar to Wnt-Cre+/−;Pik3caE545K mice.Table 2Endometrial Lesions in Ksp-Cre+/-;Ptenf/f;Pik3caE545K Mice and LittermatesGenotypeNo. of miceMice with endometrial pathological featuresMice with CAHAverage No. of CAH fociMice with CA/InvKsp-Cre+/−;Ptenf/+;Pik3caE545K12111180Ksp-Cre+/−;Ptenf/+1612129–100Ksp-Cre+/−;Pik3caE545K60000Ksp-Cre+/−;Ptenf/f888Extensive1Ksp-Cre+/−;Ptenf/f;Pik3caE545K666Extensive6CA, carcinoma; CAH, complex atypical hyperplasia; Inv, invasion. Open table in a new tab CA, carcinoma; CAH, complex atypical hyperplasia; Inv, invasion. Carcinoma and/or CAH did not develop in mice regardless of the promoter used to activate Pik3ca. This lack of phenotype could be due to inefficient DNA recombination of the mutant Pik3ca allele or lack of clonal selection in endometrial epithelial cells expressing mutant Pik3ca. To determine the reason, we used PCR analysis to ascertain excision of the STOP codon upstream of exon 1 in the Pik3ca mutant allele by Cre recombinase. Presence of a recombined allele leads to amplification of a 735-bp band, whereas the nonrecombined allele amplifies a 637-bp band. Areas of CAH and carcinoma were microdissected from tissue sections of Ksp-Cre+/−;Ptenf/f;Pik3caE545K, Ksp-Cre+/−;Ptenf/f, and Ksp-Cre+/−;Ptenf/+;Pik3caE545K uteri. DNA was extracted and analyzed by PCR. DNA from Ksp-Cre+/−;Ptenf/f;Pik3caE545K and Ksp-Cre+/−;Ptenf/+;Pik3caE545K uteri amplified both 637- and 735-bp bands (Figure 2A), proving that the mutant allele had undergone recombination in the lesions from these mice. DNA extracted from Ksp-Cre+/−;Ptenf/f uteri amplified only a 637-bp band, as expected (Figure 2A). DNA was also extracted from normal epithelium of Ksp-Cre+/−;Pik3caE545K mice and PCR amplified a 637-bp nonrecombined band, but a faint 735-bp band from the recombined allele was detected (Figure 2A). This observation suggested that even though recombination occurred, epithelial cells with the recombined allele do not have a significant growth advantage in this setting. We also performed IHC analysis on the tissue sections using p-Akt antibody. Although areas of CAH and carcinoma in all of the genotypes with endometrial pathological features (Figure 2B) exhibited characteristic membrane staining for p-Akt, the epithelium or glands in Ksp-Cre+/−;Pik3caE545K mice (Figure 2C) were completely negative for p-Akt. The uteri of Wnt-Cre+/−;Pik3caE545K mice were also negative for p-Akt staining (data not shown). Pik3ca was activated in vitro by adeno-Cre treatment of primary epithelial cells isolated from Ptenf/f and Ptenf/f;Pik3caE545K mice. Adeno-GFP–treated cells were used as control cells (GFP). We have previously demonstrated that in vitro adeno-Cre–mediated Pten deletion in epithelial cells results in activation of Akt.13Joshi A. Ellenson L.H. Adenovirus mediated homozygous endometrial epithelial Pten deletion results in aggressive endometrial carcinoma.Exp Cell Res. 2011; 317: 1580-1589Crossref PubMed Scopus (18) Google Scholar In cells prepared from Pik3caE545K mice, Akt was activated but was severalfold lower than that observed in Ptenf/f or Ptenf/f;Pik3caE545K cells (Figure 2D). Thus, mutant Pik3ca alone results in a much less robust activation of Akt, compared to loss of Pten, in endometrial epithelial cells. Quantitation of the immunoblot analysis is depicted in Figure 2E. Next, we performed IHC analysis on uterine sections from Ksp-Cre+/−;Ptenf/+, Ksp-Cre+/−;Ptenf/+;Pik3caE545K, Ksp-Cre+/−;Ptenf/f, and Ksp-Cre+/−;Ptenf/f;Pik3caE545K mice with antibodies against p-Gsk3β, p-Pras40, Stathmin, and FoxO1, which are known substrates of p-Akt. IHC analysis was also performed using antibodies against β-catenin because cytoplasmic and nuclear accumulation of β-catenin has been associated with squamous metaplasia. In the genotypes with CAH and carcinoma, the lesions expressed p-Gsk3β and p-Pras40 (Figure 3, A and B), whereas normal epithelium was negative (Figure 3, A and B), correlating with activation of the pathway. Stathmin and β-catenin expression was not altered, and β-catenin was localized to cell membranes in normal and neoplastic cells, even in areas of squamous metaplasia (Supplemental Figure S1). FoxO1 expression was also significantly altered. FoxO1 localization was nuclear in normal epithelium and stromal cells (Figure 3C), but areas of CAH (Figure 3C) and carcinoma (Figure 3D) exhibited complete lack of nuclear FoxO1 with reduced cytoplasmic expression, irrespective of the genotypes. Pras40, Gsk3β, and FoxO1, therefore, are likely downstream targets of Pten inactivation in the uterine epithelium. IHC analysis with FOXO1 antibody was also performed on TMAs of human endometrial carcinomas and normal cycling (proliferative and secretory) endometria. Intense nuclear expression was observed only in the normal secretory epithelium but was absent in normal proliferative epithelium, as reported previously.19Ward E.C. Hoekstra A.V. Blok L.J. Hanifi-Moghaddam P. Lurain J.R. Singh D.K. Buttin B.M. Schink J.C. Kim J.J. The regulation and function of