Peutz-Jeghers syndrome (PJS) is a dominantly inherited disorder characterized by gastrointestinal hamartomatous polyps and mucocutaneous melanin pigmentation. Germ line mutations in LKB1 cause PJS. We have generated mice carrying an Lkb1 exon 2 to 8 deletion by gene targeting in embryonic stem cells. Heterozygotes develop gastric hamartomas that are histologically similar to those found in humans with PJS. LKB1 is also reportedly a mediator of p53-dependent apoptosis. To explore the potential combined effects of p53 and Lkb1 alterations on tumorigenesis, we carried out a series of matings with Lkb1+/− and p53 null mice to generate Lkb1+/−/p53+/− and Lkb1+/−/p53−/− mice. Similar to the Lkb1+/− mice, gastrointestinal hamartomas have also been detected in the mice with these two genotypes. The Lkb1+/−/p53+/− mice displayed a dramatically reduced life span and increased tumor incidence compared to the mice with either Lkb1 or p53 single gene knockout. The time to onset of polyposis in Lkb1+/−/p53−/− mice is ∼2 months earlier than Lkb1+/−/p53+/− and Lkb1+/− mice, whereas the latter two show a similar time to onset which is at ∼6 months of age. These results strongly suggested that mutations of p53 and Lkb1 gene cooperate in the acceleration of tumorigenesis. (Cancer Res 2005; 65(24): 11297-303)

Peutz-Jeghers syndrome (PJS) is a unique autosomal dominant syndrome characterized by cutaneous hypermelanocytic macules and hamartomatous polyp development in the gastrointestinal tract. Patients with PJS also have a dramatically increased risk for the development of a variety of different neoplasias (13). The syndrome is caused by germ line mutations in the LKB1/STK11 gene encoding serine threonine kinase 11 (46). Approximately 70% of clinically defined PJS families harbor germ line Lkb1 mutations (7, 8), and the majority of these mutations are inactivating. Also, somatic Lkb1 mutations have been reported in a variety of carcinomas, such as malignant melanoma (9), pancreatic cancer (10), and lung adenocarcinoma (11, 12).

Numerous studies have indicated a tumor-suppressive function for LKB1. This gene has been implicated in the regulation of multiple biological processes and signaling pathways. Restoring LKB1 activity into cancer cell lines defective for its expression results in a G1 cell cycle arrest (13) by inducing the expression of p21WAF1/CIP1 through a p53-dependent mechanism (14). The involvement of LKB1 in upstream signaling to AKT and the PI3 kinase pathway (15) indicates the importance of this gene as a regulator of cell growth. It has also been reported that LKB1 (XEEK1) regulates Wnt signaling in vertebrate development (16). Shaw et al. (17) showed that LKB1 directly activates AMP-activated kinase and regulates apoptosis in response to energy stress. Another study (18) shows that LKB1 regulates 13 kinases of the AMPK subfamily. Taken together, these data suggest that LKB1 plays an important role in tumorigenesis.

The p53 gene is the most intensively studied tumor suppressor gene, which produces a protein product that functions as a transcription factor. Close to half of all human cancers have p53 gene mutations, indicating that loss of its function is a critical event in tumor progression (19). Studies of p53 mouse models have revealed a key function of p53 in inhibition of cell proliferation and induction of cell apoptosis. A p53 null allele has been generated in mice by several groups and the conclusions are similar (2023). Jacks and colleagues (20) constructed a mouse strain on a C57BL/6J background carrying a p53 mutant allele, which deleted ∼40% of the p53 coding region and completely eliminated synthesis of the p53 protein. Mice homozygous for this mutant allele are viable but highly predisposed to malignancy and most of them die by 6 months of age. Mice heterozygous for this mutant allele are also susceptible to tumors, predominantly sarcoma, osteosarcoma, and lymphomas; ∼50% of these mice die by 18 months of age. Studies on these mice lacking one or two copies of the p53 gene have provided invaluable insight into the process of tumorigenesis.

There is evidence to suggest that LKB1 plays a key role in p53-mediated apoptosis (24). Loss of heterozygosity (LOH) at 17p as well as overexpression of p53, indicative of p53 gene mutation, was seen in adenocarcinomas associated with PJS (25). In order to study the interaction between LKB1 and p53, we generated heterozygous mice with a targeted inactivating mutation of Lkb1, and then crossed this line with the p53 null mice mentioned above (20) to obtain double knockout mice for Lkb1 and p53 genes. As our data show, gastrointestinal hamartomatous polyps occur with similar frequency and the onset of polyposis occurs at roughly the same age in both Lkb1+/− and Lkb1+/−/p53+/− mice. In contrast, the doubly heterozygous mice have a dramatically increased incidence of tumors characteristic of p53 knockout mice. Furthermore, the time to onset of polyposis in Lkb1+/−/p53−/− is ∼2 months earlier than Lkb1+/−/p53+/− and Lkb1+/− mice. All these results suggest that Lkb1 and p53 cooperate synergistically in tumorigenesis. Thus, our mouse models of human PJS should be useful in future studies aimed at revealing the mechanisms of action of LKB1, as well as developing novel preventive and therapeutic strategies.

Mice. Mice were housed at the Department of Veterinary Medicine and Surgery, The University of Texas M.D. Anderson Cancer Center. All experiments described in this article were conducted according to an approved Institutional Animal Care and Use Committee protocol. p53 null mice from a C57BL/6J background (20) were kindly provided by Dr. Guillermina Lozano (Department of Molecular Genetics, M.D. Anderson Cancer Center, Houston, TX).

Cloning of mouse Lkb1 gene fragments by PCR. A clone (clone 82) was obtained after screening the mouse library λ DASH-2 by hybridization with mouse Lkb1 exon 1 PCR product. Sequencing showed that clone 82 contains exon 1 and part of intron 1 of the mouse Lkb1 gene. Sequencing also allowed us to identify a forward primer (CJ1, 5′-TTGGTCTGGGCAAGTAGACACA-3′), which was used along with a reverse primer (exon 2R, 5′-CTGCTTCTCCTCATTGTACAGC-3′) to amplify the intron 1 fragments. Also, the regions of exons 2 to 8 (2-8F, 5′-GGCAGGGTCAGCTAAGCAATGT-3′; 2-8R, GGAGACACAGTAGCCCAAGGTC-3′) and intron 8 (3 arm F1, 5′-GAGGAGGAGGA GGAGGAAGACT-3′; 3 arm R, 5′-TCCATTCTGACCCACTTCCTCT-3′) were amplified separately by using the normal mouse genomic DNA as template. A TOPO XL PCR Cloning Kit (Invitrogen Corporation, San Diego, CA) was used to obtain a clone (clone 1) for the intron 1 fragment, a clone (clone B38) for the exon 2 to 8 fragment, and a clone (clone 13) for the intron 8 fragment, respectively. After sequencing, intron 1 was found to be 7894 bp in length (data not shown).

Generation of Lkb1 knockout mice. The 5′ homologous arm was cloned into the pBluescript II SK plasmid vector by ligating together the following three fragments one by one: a loxP sequence (5′-ATAACTTCGTATAGCATACATTATACGAAGTTAT-3′) attached with ClaI site, the PCR product of the primer pair of 5 arm 3′ F and 5 arm 3′ R (5 arm 3′ F, 5′-CGGAATTCGGCCATTGCCTGGGTATC-3′; 5 arm 3′ R, 5′-CCATCGATGCCTGCTGTGAGAAGCCTACAT-3′) digested with HindIII and ClaI, and the fragment resulting from clone 1 digested with KpnI and HindIII. Then the following fragments were inserted one by one into the resulting clone: a fragment of phosphoglycerate kinase-neo (neomycin resistance) gene cassette flanked by two loxP sites, a fragment of the thymidine kinase gene cassette as a negative selection marker, a fragment containing exons 2 to 8 resulting from digestion of clone B38 with EcoRI (originating from the polylinker) and AvaI, and a 3′ homologous arm fragment resulting from the digestion of clone 13 with AvaI and ApaI (Fig. 1A). The targeting vector was sequenced to verify that no errors had been produced in the process of cloning. AB-1 embryonic stem cells were transfected with the linearized targeting vector by electroporation. Cells were selected for resistance to G418 and 1-(2-deoxy-2-fluoro-d-arabinofuranosyl)-5-iodouracil. Southern blot analysis was done to identify clones in which the expected homologous recombination event had occurred at both 5′ and 3′ ends (Fig. 1C). These clones were then injected into C57BL/6J blastocysts. To obtain Lkb13loxP heterozygous mutants, male chimeras were mated with C57BL/6J females purchased from The Jackson Laboratory (Bar Harbor, ME). To produce null mutant mice in which the region containing exons 2 to 8 had been deleted, Lkb13loxP/+ mice were mated with CMV-Cre transgenic mice (26). The first generation resulting from this mating was mosaic, and the genotype was screened by PCR (data not shown). Then the mosaic mice were mated with C57BL/6J mice to separate the three different genotypes resulting from Cre-mediated recombination (Fig. 1B). Southern blot analysis and PCR were done on the DNA extracted from the tip of mice tails to screen the genotype.

Figure 1.

Mouse Lkb1 gene targeting. A, structure of the mouse Lkb1 gene (exons 1-10), targeting vector, and targeted allele (3loxP allele), respectively. In the targeting vector, the region containing exons 2 to 8 was flanked by two loxP sequences; a neomycin resistance gene cassette (neo) and a loxP sequence were inserted in intron 8. Solid boxes, exons; Neo, neomycin resistance gene cassette; TK, thymidine kinase gene cassette; ▹, loxP site; dashed line, probes used for Southern blot analysis; A, AflII; E, EcoRI; arrowhead; PCR primer positions. B, structure of the null allele, floxed neo allele, and 2loxP allele resulting from Cre-initiated recombination. C, Southern blot analysis of embryonic stem cell clone. DNA from each embryonic stem cell clone and normal mouse tissue were digested with AflII and hybridized with 5′ probe or digested with EcoRI and hybridized with 3′ probe to detect the correctly homologous recombination at both the 5′ and 3′ ends. W, wild-type allele; T, targeted allele (3loxP allele). D, Southern blot analysis of Lkb13loxP heterozygous mice. E, Southern analysis (left) and PCR analysis (right) of Lkb1+/− offspring resulting from mating with CMV-Cre transgenic mice and then with C57BL/6J mice. N, null allele.

Figure 1.

Mouse Lkb1 gene targeting. A, structure of the mouse Lkb1 gene (exons 1-10), targeting vector, and targeted allele (3loxP allele), respectively. In the targeting vector, the region containing exons 2 to 8 was flanked by two loxP sequences; a neomycin resistance gene cassette (neo) and a loxP sequence were inserted in intron 8. Solid boxes, exons; Neo, neomycin resistance gene cassette; TK, thymidine kinase gene cassette; ▹, loxP site; dashed line, probes used for Southern blot analysis; A, AflII; E, EcoRI; arrowhead; PCR primer positions. B, structure of the null allele, floxed neo allele, and 2loxP allele resulting from Cre-initiated recombination. C, Southern blot analysis of embryonic stem cell clone. DNA from each embryonic stem cell clone and normal mouse tissue were digested with AflII and hybridized with 5′ probe or digested with EcoRI and hybridized with 3′ probe to detect the correctly homologous recombination at both the 5′ and 3′ ends. W, wild-type allele; T, targeted allele (3loxP allele). D, Southern blot analysis of Lkb13loxP heterozygous mice. E, Southern analysis (left) and PCR analysis (right) of Lkb1+/− offspring resulting from mating with CMV-Cre transgenic mice and then with C57BL/6J mice. N, null allele.

Close modal

Histopathologic analysis of polyps and other tumors. Tissues were fixed with 10% formalin solution, embedded in paraffin, cut into 4- to 5-μm sections, stained with H&E, and examined microscopically.

Kaplan-Meier survival analysis. STATA software (StataCorp, College Station, TX) was used for survival curve analysis and tumor-free survival analysis. Statistical significance between genotypes was determined using built-in analysis for survival curves consisting of a log rank test yielding a P value.

Loss of heterozygosity analysis. Genomic DNA from paraffin embedded tissue was extracted using QIAamp DNA mini Kit (Qiagen Inc., Valencia, CA). Two sets of primers were used to detect the targeted Lkb1 allele (primer a, 5′-TGTCCTGGTGGTCCTTCTC-3′; and primer e, 5′-ACAGGGAGCTACACTAAAG-3′) and the wild-type Lkb1 allele (primer 3 arm F, 5′-GAGGAGGAGGAGGAGGAAGAC-3′ and primer e), respectively. The multiplex PCR was done for 10 minutes at 94°C followed by 30 cycles at 94°C for 30 seconds, 58°C for 45 seconds, and 72°C for 1 minute, and a final extension step at 72°C for 10 minutes.

Generation of Lkb1 conditional null and exon 2 to 8–deleted knockout mice. The Lkb13loxP allele with three loxP sites flanking the region of exons 2 to 8 and an adjacent neomycin resistance gene cassette (Fig. 1A), was introduced into mouse embryonic stem cells by homologous recombination. The chimeric mice derived from four recombinant embryonic stem cell clones transmitted the Lkb13loxP allele to their offspring (Fig. 1C). We established two independent mutant mouse lines from embryonic stem cell clones 5B and 5C, both of which were heterozygous for the Lkb13loxP allele (Fig. 1D). Mice of these two lines were bred and used for further analysis. To obtain Lkb1 exon 2 to 8–deleted (hereafter called knockout) mice, the Lkb13loxP/+ mouse was crossed with the C57BL/6J inbred CMV-Cre transgenic mouse. Three different genotypes such as Lkb1+/−, Lkb1floxed neo/+, and Lkb12loxP/+ resulted from the above mating (Fig. 1B). The Lkb1+/− genotypes of the offspring were screened by PCR and by Southern blot analysis (Fig. 1E). No homozygous mutant pups were obtained by intercrossing the heterozygote, indicating that Lkb1−/− mice were embryonic lethal. Further studies showed that the Lkb1−/− embryo was not viable after 9.5 days postcoitum. This is consistent with previous reports on Lkb1 knockout mice by other groups (2729).

Survival and polyposis in Lkb1+/− mice.Lkb1+/− mice were found to have reduced life spans with ∼50% dying by 14 months compared with their wild-type littermates (Fig. 2A). Their abdomens became increasingly distended as they aged. Complete necropsies were done on 46 Lkb1+/− mice that became moribund and were sacrificed at ages ranging from 6 to 18 months. At necropsy, gastrointestinal polyps were identified in Lkb1+/− mice as early as 6 months of age and in ∼75% of mice at age of 6.5 months. Most polyps were located at the junction of the pylorus and duodenum, and a small number of polyps were found in the antrum or fundus of the stomach. No polyps were found in other parts of the gastrointestinal tract including the large intestine (Table 1). In the majority of animals that we examined, polyps originating from the pylorus were large, whereas polyps located in the fundus and antrum were smaller. The large polyps in the pylorus protruded into the duodenum, resulting in gross distension and obstruction of the duodenum (Fig. 3A and B). These obstructions, which mimic the clinically acute obstruction seen in many cases of human PJS, seem to explain the high mortality of the Lkb1+/− mice.

Figure 2.

Effects of p53 mutation on the survival of Lkb1+/− mice. A, survival analysis of Lkb1+/− mice and littermate wild-type (WT) mice (P < 0.0001); dashed line, Lkb1+/− mice; dotted line, wild-type mice. B, survival analysis of Lkb1+/−/p53+/− mice and p53+/− mice (P < 0.0001); dashed line, Lkb1+/−/p53+/− mice; dotted line, p53+/− mice. C, survival analysis of Lkb1+/−/p53−/− mice and p53−/− mice (P > 0.05); dashed line, Lkb1+/−/p53−/− mice; dotted line, p53−/− mice. The curves represent the numbers of mice that either died or were euthanized when they become moribund. Statistical significance between curves was determined using log rank test.

Figure 2.

Effects of p53 mutation on the survival of Lkb1+/− mice. A, survival analysis of Lkb1+/− mice and littermate wild-type (WT) mice (P < 0.0001); dashed line, Lkb1+/− mice; dotted line, wild-type mice. B, survival analysis of Lkb1+/−/p53+/− mice and p53+/− mice (P < 0.0001); dashed line, Lkb1+/−/p53+/− mice; dotted line, p53+/− mice. C, survival analysis of Lkb1+/−/p53−/− mice and p53−/− mice (P > 0.05); dashed line, Lkb1+/−/p53−/− mice; dotted line, p53−/− mice. The curves represent the numbers of mice that either died or were euthanized when they become moribund. Statistical significance between curves was determined using log rank test.

Close modal
Table 1.

Effect of the p53 mutation on location of polyps in Lkb1+/− mice

GenotypeNumber of mice
TotalPolyp location
Stomach and duodenum*Large intestine
Wild-type 
Lkb1+/− 48 46 
p53+/− 
Lkb1+/−/p53+/− 38 35 
p53−/− 
Lkb1+/−/p53−/− 8§ 
GenotypeNumber of mice
TotalPolyp location
Stomach and duodenum*Large intestine
Wild-type 
Lkb1+/− 48 46 
p53+/− 
Lkb1+/−/p53+/− 38 35 
p53−/− 
Lkb1+/−/p53−/− 8§ 
*

Most of the polyps located at the junction of the pylorus and duodenum.

These mice were sacrificed at 6 to 18 months of age once they became moribund. No polyps were detected in 2 of 48 mice, which are 6.5 and 8 months old, respectively.

These mice were sacrificed at 7 to 14 months of age once they became moribund. No polyps were detected in 3 of 38 mice, which are 7, 8, and 8 months old, respectively.

§

These mice were sacrificed at 4 to 6 months of age once they became moribund. No polyps were detected in one of eight mice, which is 6 months old.

Figure 3.

Morphologic and histologic analysis of gastrointestinal polyps. A, stomach and duodenum of Lkb1+/− mice and wild-type (WT) littermates at 16 months of age. B to D, representative dissection micrographs of gastric polyps in Lkb1+/− mice (B), Lkb1+/−/p53+/− mice (C) and Lkb1+/−/p53−/− mice (D) showing small polyps (▸) in the stomach and large polyps (*) at the junction of the pylorus and duodenum. E to H, representative H&E staining of the polyps from a patient with PJS (E), Lkb1+/− mice (F), Lkb1+/−/p53+/− mice (G) and Lkb1+/−/p53−/− mice (H); original magnification, ×100.

Figure 3.

Morphologic and histologic analysis of gastrointestinal polyps. A, stomach and duodenum of Lkb1+/− mice and wild-type (WT) littermates at 16 months of age. B to D, representative dissection micrographs of gastric polyps in Lkb1+/− mice (B), Lkb1+/−/p53+/− mice (C) and Lkb1+/−/p53−/− mice (D) showing small polyps (▸) in the stomach and large polyps (*) at the junction of the pylorus and duodenum. E to H, representative H&E staining of the polyps from a patient with PJS (E), Lkb1+/− mice (F), Lkb1+/−/p53+/− mice (G) and Lkb1+/−/p53−/− mice (H); original magnification, ×100.

Close modal

Histologic evaluation indicated that all of the polyps were hamartomas (Fig. 3F). The histologic characteristics of these polyps had an extensive arborizing smooth muscle component and normal glandular epithelium. Histologic comparison of polyps derived from Lkb1+/− mice (Fig. 3F) with polyps resected from human patients with PJS (Fig. 3E) revealed striking similarities.

Out of 13 Lkb1+/− mice with complete histologic assessment of all of the organs, 9 (69%) exhibited organomegaly of spleen, lung, liver, stomach, and intestine. No other neoplasms or tumors were observed.

Survival and polyposis in Lkb1+/−/p53+/− and Lkb1+/−/p53−/− mice. As mentioned above, current evidence suggests that Lkb1 regulates p53-dependent apoptosis, and loss of p53 is an important factor in tumor development. To determine whether Lkb1 and p53 mutations act synergistically in tumorigenesis, we mated Lkb1 heterozygous knockout mice with p53 homozygous knockout mice (20) to generate a series of mice heterozygous for both mutations (Lkb1+/−/p53+/−). A population of 129 mice carrying a mutant allele of both Lkb1 and p53 were generated and monitored. Lkb1+/−/p53+/− mice were found to have reduced life spans with ∼50% dying by 10 months (Fig. 2B). The median age of survival of Lkb1+/−/p53+/− mice (10 months) was dramatically lower than that of mice heterozygous for the Lkb1 or p53 mutation alone (14 and 18 months, respectively).

Similar to Lkb1+/− mice, Lkb1+/−/p53+/− mice had enlarged abdomens due to hamartomatous polyps in the pylorus and duodenum (Fig. 3C). Mice that became moribund were euthanized and subjected to a complete necropsy. Ninety-two percent (35 out of 38) of the moribund Lkb1+/−/p53+/− mice ranging in age from 7 to 14 months showed polyps in their gastrointestinal tract. The majority of these mice had polyps in the stomach and/or duodenum (Table 1). One mouse also displayed colonic polyps. We compared the time to onset of polyposis observed in Lkb1+/− to that seen in Lkb1+/−/p53+/− mice. Necropsy was done on the randomly selected Lkb1+/−/p53+/− mice at different time points. Polyps could not be detected in the mice younger than 5.5 months (four out of four), but were detected in 60% of mice (three out of five) at 6.5 months. This result is similar to the time to onset of polyposis observed in Lkb1+/− mice. In addition, both intensity (data not shown) and the histologic appearance (Fig. 3G) of the polyps in Lkb1+/−/p53+/− mice were very similar to those in Lkb1+/− mice. Thus, it is very possible that loss of one allele of p53 is not enough to accelerate the polyposis. To further evaluate whether p53 plays a role in polyposis, a cohort of 20 mice nullizygous for p53 and containing one mutant Lkb1 allele (Lkb1+/−/p53−/−) mice were generated and monitored. The median age of survival of Lkb1+/−/p53−/− mice was ∼4.5 months which is very similar to their littermate p53−/− mice (Fig. 2C). Necropsy studies revealed that gastrointestinal hamartomatous polyps (Fig. 3D) in Lkb1+/−/p53−/− mice were identified as early as 4.3 months. This is ∼2 months earlier than those for Lkb1+/−/p53+/− mice, no time difference was observed between the Lkb1+/− and Lkb1+/−/p53+/− mice. Although the histologic appearance of these polyps from Lkb1+/−/p53−/− mice (Fig. 3H) are very similar to the polyps identified in the Lkb1+/− and Lkb1+/−/p53+/− mice, the difference in the polyps locations were observed as two out of seven mice also had polyps in their large intestines. These findings suggested that the knockout p53 gene in the Lkb1+/− mice increases the time to onset of polyposis and influence the distribution of polyps.

Tumorigenesis in Lkb1+/−/p53+/− and Lkb1+/−/p53−/− mice. In addition to the gastrointestinal polyps, a variety of tumors were also found in the Lkb1+/−/p53+/− and Lkb1+/−/p53−/− mice (Table 2; Fig. 4). Twenty-three (61%) of 38 Lkb1+/−/p53+/− mice developed a total of 30 tumors. Among them, 13 (43%) of these tumors were osteosarcomas, 3 (10%) were sarcomas, and 6 (20%) were lymphomas. These tumor types are the characteristic tumor types detected in the mice heterozygous for p53. In Lkb1+/−/p53+/− mice, osteosarcoma is the predominant one, which is very similar to the dominant tumor type reported previously (2023) as well as in their p53+/− littermate presented here. Also, we did a tumor-free survival analysis on the Lkb1+/−/p53+/− and p53+/− mice that underwent necropsy. The median age of tumor-free survival for Lkb1+/−/p53+/− mice was 11 months, whereas for p53+/− mice, it was ∼17 months (Fig. 5A), this difference was statistically significant (P < 0.01) suggesting that Lkb1+/−/p53+/− mice had a significant increase in tumor incidence compared with the p53+/− mice.

Table 2.

Tumor spectrum from Lkb1+/−/p53+/−, Lkb1+/−/p53−/−, and control mice

GenotypeAge (mo)*Tumor type
p53+/− 10.8, 12.0, 16.8, 12.7 osteosarcoma 
 18.2 lymphoma (spleen) 
Lkb1+/−/p53+/− 9.2, 8.3, 8.6, 9.9, 9.8, 12.0, 7.1, 10.2, 9.9, 11.0, 8.8, 12.6, 7.6 osteosarcoma 
 10.8, 12.3, 12.0, 12.4,§ 12.6 lymphoma (spleen) 
 7.6 lymphoma (mediastinum) 
 7.6 lymphoma (lung) 
 12.4§ lymphoma (small intestine) 
 11.2, 7.6, 12.0 sarcoma 
 12.6 adenocarcinoma (small intestine) 
 11.0 anaplastic carcinoma (pancreas) 
 11.0 anaplastic metastatic lung 
 10.9, 12.6 adenoma (small intestine) 
 9.9 adenocarcinoma mammary gland 
p53−/− 6.0 lymphoma (mediastinum) 
 8.0¥ osteosarcoma vertebra 
 8.0¥ osteosarcoma metastatic lung 
 8.0¥ osteosarcoma metastatic liver 
 8.0¥ osteosarcoma metastatic spleen 
 8.0¥ osteosarcoma metastatic kidney 
Lkb1+/−/p53−/− 6.0 glioblastoma (brain) 
 6.0, 4.5, 5.7£ lymphoma (spleen) 
 6.0, 4.5, 5.7 lymphoma (lung) 
 4.5, 5.7, 5.7£ lymphoma (liver) 
 6.0, 5.7 lymphoma (heart) 
 6.0, 5.7 lymphoma (kidney) 
 3.5 sarcoma (unclassified) 
 4.0 fibrosarcoma 
 5.3 anaplastic sarcoma (unclassified) 
GenotypeAge (mo)*Tumor type
p53+/− 10.8, 12.0, 16.8, 12.7 osteosarcoma 
 18.2 lymphoma (spleen) 
Lkb1+/−/p53+/− 9.2, 8.3, 8.6, 9.9, 9.8, 12.0, 7.1, 10.2, 9.9, 11.0, 8.8, 12.6, 7.6 osteosarcoma 
 10.8, 12.3, 12.0, 12.4,§ 12.6 lymphoma (spleen) 
 7.6 lymphoma (mediastinum) 
 7.6 lymphoma (lung) 
 12.4§ lymphoma (small intestine) 
 11.2, 7.6, 12.0 sarcoma 
 12.6 adenocarcinoma (small intestine) 
 11.0 anaplastic carcinoma (pancreas) 
 11.0 anaplastic metastatic lung 
 10.9, 12.6 adenoma (small intestine) 
 9.9 adenocarcinoma mammary gland 
p53−/− 6.0 lymphoma (mediastinum) 
 8.0¥ osteosarcoma vertebra 
 8.0¥ osteosarcoma metastatic lung 
 8.0¥ osteosarcoma metastatic liver 
 8.0¥ osteosarcoma metastatic spleen 
 8.0¥ osteosarcoma metastatic kidney 
Lkb1+/−/p53−/− 6.0 glioblastoma (brain) 
 6.0, 4.5, 5.7£ lymphoma (spleen) 
 6.0, 4.5, 5.7 lymphoma (lung) 
 4.5, 5.7, 5.7£ lymphoma (liver) 
 6.0, 5.7 lymphoma (heart) 
 6.0, 5.7 lymphoma (kidney) 
 3.5 sarcoma (unclassified) 
 4.0 fibrosarcoma 
 5.3 anaplastic sarcoma (unclassified) 

NOTE: , §, , , ¥, , , , £ symbols represent individual mice where multiple tumors were detected in the same mouse, respectively.

*

Each age corresponds to a different mouse.

Figure 4.

Representative H&E staining of the tumor detected in the Lkb1+/−/p53+/− mice. A to E, original magnification, ×250. A, adenocarcinoma in small intestine; B, adenocarcinoma in mammary gland; C, anaplastic carcinoma in pancreas; D, lymphoma in spleen; E, osteosarcoma.

Figure 4.

Representative H&E staining of the tumor detected in the Lkb1+/−/p53+/− mice. A to E, original magnification, ×250. A, adenocarcinoma in small intestine; B, adenocarcinoma in mammary gland; C, anaplastic carcinoma in pancreas; D, lymphoma in spleen; E, osteosarcoma.

Close modal
Figure 5.

Effects of Lkb1 mutation on the tumorigenesis of p53+/− mice. A, tumor-free survival analysis of Lkb1+/−/p53+/− and p53+/− mice (P < 0.01); dashed line, Lkb1+/−/p53+/− mice; dotted line, p53+/− mice. B, Lkb1 allele-specific PCR analysis on DNA from normal mouse tail (N), adenoma in small intestine (1), adenocarcinoma in small intestine (2), adenocarcinoma in mammary gland (3), anaplastic carcinoma in pancreas (4), lymphoma in spleen (5) and osteosarcoma (6). Lanes 2 to 6 correspond to the samples describe in Fig. 4A-E, respectively; WT, wild-type allele; KO, knockout allele.

Figure 5.

Effects of Lkb1 mutation on the tumorigenesis of p53+/− mice. A, tumor-free survival analysis of Lkb1+/−/p53+/− and p53+/− mice (P < 0.01); dashed line, Lkb1+/−/p53+/− mice; dotted line, p53+/− mice. B, Lkb1 allele-specific PCR analysis on DNA from normal mouse tail (N), adenoma in small intestine (1), adenocarcinoma in small intestine (2), adenocarcinoma in mammary gland (3), anaplastic carcinoma in pancreas (4), lymphoma in spleen (5) and osteosarcoma (6). Lanes 2 to 6 correspond to the samples describe in Fig. 4A-E, respectively; WT, wild-type allele; KO, knockout allele.

Close modal

Most importantly, we observed a pancreatic carcinoma, a lung carcinoma, a breast carcinoma, a small intestinal adenocarcinoma, and two intestinal adenomas in these Lkb1+/−/p53+/− mice (Fig. 4). It was reported that PJS patients are predisposed to develop cancer of the pancreas, intestine, breast, and other organs. Our results suggest that heterozygosity for p53 increases tumor incidence in these organs. Thus, Lkb1+/−/p53+/− mice probably provide a better model for PJS. To test whether there is LOH of Lkb1 involved in these PJS-related tumors in Lkb1+/−/p53+/− mice, we analyzed the status of the wild-type copy of Lkb1 gene in three carcinomas from pancreas, small intestine, and mammary gland, and one adenoma in the small intestine, as well as an osteosarcoma and a lymphoma by allele-specific PCR analysis. We found that two of the six tumors (Fig. 5B , lanes 2 and 5) displayed a decrease in intensity of staining for the PCR product of the wild-type allele relative to the knockout allele suggesting LOH. Whereas in the rest of the samples, both wild-type and knockout alleles are comparably amplified, suggesting that the wild-type allele was retained.

The Lkb1+/−/p53−/− mice were highly susceptible to tumors at an early age, particularly malignant lymphoma (Table 2). Carcinoma of the pancreas, mammary gland, and small intestines, observed in Lkb1+/−/p53+/− mice at >10 months of age, were not observed in the Lkb1+/−/p53−/− mice, probably due to the severely compromised life span.

Together, these findings showed that the loss of the p53 gene accelerates the time to onset of polyposis in Lkb1+/− mice and the added deletion of the Lkb1 gene dramatically increased tumor incidence in p53+/− mice and strongly suggests that mutation of both Lkb1 and p53 genes did have a cooperative effect on tumorigenesis.

In the present study, we developed an Lkb1+/− mouse line. It serves as a model for PJS, as hamartomatous polyps were found in the gastrointestinal tract of heterozygous mutant mice, and histologically, the polyps of Lkb1+/− mice are strikingly similar to those of polyps found in human patients with PJS. However, the distribution of polyps is different in PJS patients compared with Lkb1+/− mice. In human patients with PJS, polyps can arise anywhere in the gastrointestinal tract (3032). In the present study, the polyps in Lkb1+/− mice were located predominantly at the pylorus. Interestingly, mice with either Lkb1+/−/p53+/− or Lkb1+/−/p53−/− genotypes showed a shift in the distribution of polyps compared with the Lkb1+/− genotype, suggesting that defective p53 can influence the distribution of polyps. Furthermore, it has been shown here that loss of both alleles of p53 accelerate the polyposis by ∼2 months, although the time to onset of polyposis (∼6 months) in Lkb1+/−/p53+/− mice was not very different from that in Lkb1+/− mice. Taken together, these findings suggest that the interaction between these two tumor suppressor genes in the gastrointestinal tract had a cooperative effect on polyposis.

LKB1 has been reported as a mediator of p53-dependent apoptosis and a player in cell cycle arrest. Therefore, to evaluate the joint effects of the Lkb1 and p53 genes on tumorigenesis, Lkb1+/−/p53+/− and Lkb1+/−/p53−/− mice were generated. Hamartomatous polyps in the gastrointestinal tracts as well as various other types of tumors have been detected in mice with these two genotypes. Compared with the Lkb1+/− mice present here as well as the Lkb1+/− mice reported by other groups (28, 29, 33), Lkb1+/−/p53+/− mice showed a dramatically increased incidence of various types of tumors including malignant cancers of the pancreas, mammary gland, and intestinal organs, which are the most common tumor types seen in PJS. Thus, the Lkb1+/−/p53+/− mouse line provide a model for tumor formation for PJS. Although the tumor spectrum in Lkb1+/−/p53+/− mice is not identical to that observed in human patients with PJS, this is the first mouse model of PJS in which a significant incidence of various types of tumors was observed.

The Lkb1+/−/p53−/− mice presented here also showed various types of tumors which are characteristic tumor types detected in p53-deficient mice. However, no PJS-predisposed cancer type was observed, which was most likely due to the severely compromised longevity. Meanwhile, the survival curve of the Lkb1+/−/p53−/− and p53−/− are very similar to each other, suggesting that LKB1 and p53 may function in the same apoptotic pathway. Recently, Jones et al. (34) reported that AMPK activation induces phosphorylation of p53 on serine 15 and this phosphorylation is required to initiate AMPK-dependent cell cycle arrest in response to the glucose deprivation. Meanwhile, Shaw et al. (17) showed that LKB1 directly activates AMPK and regulates apoptosis in response to energy stress. These findings suggest that regulation of p53-dependent apoptosis by LKB1 is likely to occur through the activation of AMPK. Further studies need to be done to elucidate the possible linkage between the Lkb1, AMPK, and p53. Several mouse cross-breeding studies have shown cooperation between p53 and other tumor suppressor genes or oncogenes (35, 36). For example, mutation of both Rb and p53 (Rb+/−/p53+/−) produces mice that have shorter life spans and develop a variety of tumors not found in mice with single gene mutations. This also happened in Lkb1+/−/p53+/− mice. Furthermore, it is very possible that LKB1 is in the p53-dependent pathway, loss of one wild-type allele of Lkb1 and p53 both contribute to the tumorigenesis because the p53-dependent pathway is not entirely inactive and the efficiency of the pathway could be decreased through partial loss of LKB1. Together, our data suggests that Lkb1 and p53 cooperate in the acceleration of tumorigenesis.

LOH contributes to tumor formation when cells are heterozygous for a mutant allele of a tumor suppressor gene. To further understand how LKB1 is involved in tumor formation, LOH analysis was done on the genomic DNA extracted from tumors observed in Lkb1+/−/p53+/− mice as well as in mouse tail tissue. In one adenocarcinoma of the small intestine and one lymphoma, the PCR product of wild-type alleles displayed a decrease in intensity compared with the knockout alleles, suggesting that LOH had occurred for Lkb1. Although in the rest of the samples, both wild-type and knockout alleles are comparably amplified, suggesting that the wild-type allele was retained. For the samples where LOH was not observed, we cannot rule out the possibility that the Lkb1 wild-type allele was inactivated by a localized mutation (i.e., small insertion/deletion or point mutations) because we did not sequence the wild-type gene.

In conclusion, we have described different Lkb1+/− mouse lines with or without p53 mutation for human PJS and have shown that the Lkb1 and p53 genes cooperate in tumorigenesis. Our findings also suggest that our Lkb1+/− and Lkb1+/−/p53+/− mice will be an important and useful model in which to elucidate the molecular mechanisms underlying PJS polyposis and tumorigenesis. Crossing the Lkb1 knockout mouse with other cancer-prone mouse strains will also help to establish the mechanism of action of Lkb1 by revealing genetic interactions between Lkb1 and other genes important in tumor development.

Grant support: RO3 (CA 89843-1), NIH Cancer Center Support grant CA16672 (M. Frazier) and the Janis David Gordon Memorial Fellowship from the Cancer Prevention Division at M.D. Anderson Center (C. Wei).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Haidee Chancoco for DNA extraction and PCR analysis, and Jingyun Chen for mice maintenance assistance.

1
Giardiello FM, Welsh SB, Hamilton SR, et al. Increased risk of cancer in the Peutz-Jeghers syndrome.
N Engl J Med
1987
;
316
:
1511
–4.
2
Boardman LA, Thibodeau SN, Schaid DJ, et al. Increased risk for cancer in patients with the Peutz-Jeghers syndrome.
Ann Intern Med
1998
;
128
:
896
–9.
3
Giardiello FM, Brensinger JD, Tersmette AC, et al. Very high risk of cancer in familial Peutz-Jeghers syndrome.
Gastroenterology
2000
;
119
:
1447
–53.
4
Hemminki A, Tomlinson I, Markie D, et al. Localization of a susceptibility locus for Peutz-Jeghers syndrome to 19p using comparative genomic hybridization and targeted linkage analysis.
Nat Genet
1997
;
15
:
87
–90.
5
Hemminki A, Markie D, Tomlinson I. A serine/threonine kinase gene defective in Peutz-Jeghers syndrome.
Nature
1998
;
391
:
184
–7.
6
Jenne DE, Reimann H, Nezu J, et al. Peutz-Jeghers syndrome is caused by mutations in a novel serine threonine kinase.
Nat Genet
1998
;
18
:
38
–43.
7
Amos CI, Keitheri-Cheteri MB, Sabripour M, et al. Genotype-phenotype correlations in Peutz-Jeghers syndrome.
J Med Genet
2004
;
41
:
327
–33.
8
Schumacher V, Vogel T, Leube B, et al. STK11 genotyping and cancer risk in Peutz-Jeghers syndrome.
J Med Genet
2005
;
42
:
428
–35.
9
Guldberg P, thor Straten P, Ahrenkiel V, et al. Somatic mutation of the Peutz-Jeghers syndrome gene, LKB1/STK11, in malignant melanoma.
Oncogene
1999
;
18
:
1777
–80.
10
Su GH, Hruban RH, Bansal RK, et al. Germ line and somatic mutations of the STK11/LKB1 Peutz-Jeghers gene in pancreatic and biliary cancers.
Am J Pathol
1999
;
154
:
1835
–40.
11
Sanchez-Cespedes M, Parrella P, Esteller M, et al. Inactivation of LKB1/STK11 is a common event in adenocarcinomas of the lung.
Cancer Res
2002
;
62
:
3659
–62.
12
Carretero J, Medina PP, Pio R, et al. Novel and natural knockout lung cancer cell lines for the LKB1/STK11 tumor suppressor gene.
Oncogene
2004
;
23
:
4037
–40.
13
Tiainen M, Ylikorkala A, Makela TP. Growth suppression by LKB1 is mediated by a G(1) cell cycle arrest.
Proc Natl Acad Sci U S A
1999
;
96
:
9248
–51.
14
Tiainen M, Vaahtomeri K, Ylikorkala A, Makela TP. Growth arrest by the LKB1 tumor suppressor: induction of p21(WAF1/CIP1).
Hum Mol Genet
2002
;
11
:
1497
–504.
15
Jimenez AI, Fernandez P, Dominguez O, et al. Growth and molecular profile of lung cancer cells expressing ectopic LKB1: down-regulation of the phosphatidylinositol 3′-phosphate kinase/PTEN pathway.
Cancer Res
2003
;
63
:
1382
–8.
16
Spicer J, Rayter S, Young N, et al. Regulation of the Wnt signalling component PAR1A by the Peutz-Jeghers syndrome kinase LKB1.
Oncogene
2003
;
22
:
4752
–6.
17
Shaw RJ, Kosmatka M, Bardeesy N, et al. The tumor suppressor LKB1 kinase directly activates AMP-activated kinase and regulates apoptosis in response to energy stress.
Proc Natl Acad Sci U S A
2004
;
101
:
3329
–35.
18
Lizcano JM, Goransson O, Toth R, et al. LKB1 is a master kinase that activates 13 kinases of the AMPK subfamily, including MARK/PAR-1.
EMBO J
2004
;
23
:
833
–43.
19
Greenblatt MS, Bennett WP, Hollstein M, Harris CC. Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis.
Cancer Res
1994
;
54
:
4855
–78.
20
Jacks T, Remington L, Williams BO, et al. Tumor spectrum analysis in p53-mutant mice.
Curr Biol
1994
;
4
:
1
–7.
21
Donehower LA, Harvey M, Slagle BL, et al. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumors.
Nature
1992
;
356
:
215
–21.
22
Clarke AR, Purdie CA, Harrision DJ, et al. Thymocyte apoptosis induced by p53-dependent and independent pathways.
Nature
1993
;
362
:
849
–52.
23
Tsukada T, Tomooka Y, Takai S, et al. Enhanced proliferative potential in culture of cells from p53-deficient mice.
Oncogene
1993
;
8
:
3313
–22.
24
Karuman P, Gozani O, Odze RD. The Peutz-Jegher gene product LKB1 is a mediator of p53-dependent cell death.
Mol Cell
2001
;
7
:
1307
–19.
25
Gruber SB, Entius MM, Petersen GM, et al. Pathogenesis of adenocarcinoma in Peutz-Jeghers syndrome.
Cancer Res
1998
;
58
:
5267
–70.
26
Arango NA, Lovell-Badge R, Behringer RR. Targeted mutagenesis of the endogenous mouse Mis gene promoter: in vivo definition of genetic pathways of vertebrate sexual development.
Cell
1999
;
99
:
409
–19.
27
Ylikorkala A, Rossi DJ, Korsisaari N, et al. Vascular abnormalities and deregulation of VEGF in LKB1-deficient mice.
Science
2001
;
293
:
1323
–6.
28
Jishage K, Neuz J, Kawase Y, et al. Role of LKB1, the causative gene of Peutz-Jegher's syndrome, in embryogenesis and polyposis.
Proc Natl Acad Sci U S A
2002
;
99
:
8903
–8.
29
Miyoshi H, Nakau M, Ishikawa T, et al. Gastrointestinal hamartomatous polyposis in Lkb1 heterozygous knockout mice.
Cancer Res
2002
;
62
:
2261
–6.
30
Utsunomiya J, Gocho H, Miyanaga T, et al. Peutz-Jeghers syndrome: its natural course and management.
Johns Hopkins Med J
1975
;
136
:
71
–82.
31
Bartholomew LG, Dearing WH, Baggenstoss AH. Jaundice associated with lipochrome pigmentation of the liver (Dubin-Johnson syndrome).
Gastroenterology
1957
;
33
:
302
–7.
32
Bartholomew LG, Moore CE, Dahlin DC, Waugh JM. Intestinal polyposis associated with mucocutaneous pigmentation.
Surg Gynecol Obstet
1962
;
115
:
1
–11.
33
Rossi DJ, Ylikorkala A, Korsisaari N, et al. Induction of cyclooxygenase-2 in a mouse model of Peutz-Jeghers polyposis.
Proc Natl Acad Sci U S A
2002
;
99
:
12327
–32.
34
Jones RG, Plas DR, Kubek S, et al. AMP-activated protein kinase induces a p53-dependent metabolic checkpoint.
Mol Cell
2005
;
18
:
283
–93.
35
Williams BO, Remington L, Albert DM, et al. Cooperative tumorigenic effects of germline mutations in Rb and p53.
Nat Genet
1994
;
7
:
480
–4.
36
Harvey M, Vogel H, Lee EYH-P, et al. Mice deficient in both p53 and Rb develop tumors primarily of endocrine origin.
Cancer Res
1995
;
55
:
1146
–51.