MAPK pathway control of stem cell proliferation and differentiation in the embryonic pituitary provides insights into the pathogenesis of papillary craniopharyngioma

Abstract

Despite the importance of the RAS-RAF-MAPK pathway in normal physiology and disease of numerous organs, its role during pituitary development and tumourigenesis remains largely unknown. Here, we show that the over-activation of the MAPK pathway, through conditional expression of the gain-of-function alleles BrafV600E and KrasG12D in the developing mouse pituitary, results in severe hyperplasia and abnormal morphogenesis of the gland by the end of gestation. Cell-lineage commitment and terminal differentiation are disrupted, leading to a significant reduction in numbers of most of the hormone-producing cells before birth, with the exception of corticotrophs. Of note, Sox2+ stem cells and clonogenic potential are drastically increased in the mutant pituitaries. Finally, we reveal that papillary craniopharyngioma (PCP), a benign human pituitary tumour harbouring BRAF p.V600E also contains Sox2+ cells with sustained proliferative capacity and disrupted pituitary differentiation. Together, our data demonstrate a crucial function of the MAPK pathway in controlling the balance between proliferation and differentiation of Sox2+ cells and suggest that persistent proliferative capacity of Sox2+ cells may underlie the pathogenesis of PCP.

Keywords: Mouse; Papillary craniopharyngioma; Pituitary development; Sox2; Tumour.

Conflict of interest statement

Competing interestsT.F. is a co-founder, shareholder and manager of Inivata. Inivata is a company focused on developing assays for circulating tumour DNA analysis. All the other authors declare no competing interests.

© 2017. Published by The Company of Biologists Ltd.

Figures

Fig. 1.

Abnormal pituitary morphogenesis in Hesx1Cre/+;KrasG12D/+ and Hesx1Cre/+;BrafV600E/+ mutants. Haematoxylin and Eosin staining of sagittal (A-I) or transverse (J-L) histological sections of the developing pituitary gland in control and mutant embryos; genotypes and stages are indicated. (A-C) At 10.5 dpc, Rathke's pouch (RP) is morphologically comparable between genotypes. (D-I) The developing pituitary is enlarged and dysmorphic in the mutant compared with the control pituitary at 12.5 and 14.5 dpc (arrowheads). (J-L) At 18.5 dpc, the cleft is expanded and ramified in the mutant pituitaries (arrowheads in K,L) compared with the control (J). The posterior pituitary (PP) is comparable between genotypes. AL, anterior lobe; IL, intermediate lobe. (M) Quantification of total numbers of cells in the control, Hesx1Cre/+;KrasG12D/+ and Hesx1Cre/+;BrafV600E/+ pituitaries at 18.5 dpc, showing a significant increase in the mutants (**P≤0.01, Student's t-test). Data are mean±s.e.m. Scale bars: 200 μm.

Fig. 2.

Braf and Kras mRNA is detected in the developing hypothalamus and pituitary gland.In situ hybridisation using Braf and Kras antisense riboprobes on histological sections of developing wild-type pituitaries. (A,D) At 10.5 dpc, weak overall expression of Braf and Kras is detected in the prospective hypothalamus (Hyp) and Rathke's pouch (RP). (B,E) At 14.5 dpc, strong expression is detected in the infundibulum (Inf), hypothalamus (Hyp), periluminal epithelium (PE) and developing anterior lobe (AL). (C,F) At 18.5 dpc, expression is detected in specific cells within the posterior pituitary (PP), intermediate lobe (IL) and the parenchyma of the anterior lobe (AL), including the marginal zone (MZ). Insets in C,F show the IZ and MZ at higher magnification. Scale bars: 200 μm.

Fig. 3.

Braf mRNA and BRAF-V600E protein are expressed in the developing pituitary in Hesx1Cre/+;BrafV600E/+ mutants.In situ hybridisation using Braf antisense riboprobes (A-C) and immunohistochemistry against BRAF-V600E mutant protein (D-F) on histological sections of Hesx1Cre/+;BrafV600E/+ embryos at the indicated stages. (A-C) Braf mRNA expression is detected in the hypothalamic neuroepithelium (Hyp) and developing pituitary at all stages analysed, with a similar expression pattern to the wild-type embryos (Fig. 2A-C). (D-F) At 10.5 dpc, only a few cells express the BRAF-V600E protein in the developing Rathke's pouch (RP), but numbers increased between 12.5 and 18.5 dpc. Inset in C shows the IL and MZ at higher magnification. Scale bars: 200 μm.

Fig. 4.

Temporal and spatial regulation of pERK1/2 expression developing pituitary. Immunofluorescent staining against pERK1/2 on histological sections of control and mutant pituitaries; genotypes and stages are indicated. (A-D) Abundant pERK1/2+ cells are detected in the hypothalamus (Hyp) and Rathke's pouch (RP) at 10.5 dpc, but only very few cells express pERK1/2 at 12.5 and 14.5 dpc. At 18.5 dpc, most of the signal is restricted to the intermediate lobe (IL) and marginal zone (MZ) of the anterior lobe. (E-L) A similar expression pattern is observed in both Hesx1Cre/+;KrasG12D/+ (E-H) and Hesx1Cre/+;BrafV600E/+ (I-L) mutant pituitaries. However, the pERK1/2 signal is markedly increased in 18.5 dpc pituitaries (H,L) relative to the control (D). Insets show the outlined areas at higher magnification. Scale bars: 200 μm.

Fig. 5.

Cell-lineage commitment is disrupted in Hesx1Cre/+;KrasG12D/+ and Hesx1Cre/+;BrafV600E/+ mutants. Immunostaining (green) against the commitment markers PIT1, SF1 and TPIT on sagittal histological sections of mutant and control embryos at 14.5 dpc. (A-I) The expression of TPIT (A,D,G) and SF1 (B,E,H) are markedly reduced in the mutant pituitary, especially in the Hesx1Cre/+;BrafV600E/+ mutant, relative to the control. In contrast, TPIT expression is elevated in the mutant pituitary, more apparently in the Hesx1Cre/+;BrafV600E/+ genotype (C,F,I). Blue staining is DAPI. (J) Quantification analyses demonstrate the significant reduction of PIT1+ and SF1+ cells in both mutants relative to the control pituitary, but the increase of TPIT+ cells is significant only in the Braf mutants (*P≤0.05, **P≤0.01, ***P≤0.001, Student's t-test). Data are mean±s.e.m. Scale bar: 200 μm.

Fig. 6.

Terminal differentiation of hormone-producing cells is impaired in Hesx1Cre/+;KrasG12D/+ and Hesx1Cre/+;BrafV600E/+ mutants. Immunostaining against pituitary hormones (green) on transverse histological sections of mutant and control embryos at 18.5 dpc. (A-R) Numbers of GH+ (somatotrophs) (A-C), PRL+ (lactotrophs) (D-F), TSH+ (thyrotrophs) (G-I), as well as FSH+ and LH+ (gonadotrophs) cells (J-O) appear reduced in the Braf and Kras mutants relative to the control pituitary. Blue staining is DAPI. (P-R) ACTH+ cell numbers (corticotrophs and melanotrophs) look increased in the mutant pituitaries compared with the control. (S) Quantitative analyses demonstrate a significant reduction of all of the hormone-producing cells in the Hesx1Cre/+;BrafV600E/+ mutant pituitary, except for ACTH+ cells, which are markedly increased relative to controls. Hesx1Cre/+;KrasG12D/+ mutant pituitaries show a significant decrease in the number of only PRL+, TSH+ and FSH+ cells, and ACTH+ cell number is also increased. (T) Absolute quantitative RT-PCR analysis of Gh and Acth mRNA expression in mutant and control pituitaries at 18.5 dpc. Significance is only reached for Gh expression, but there is a trend towards an increase in Acth expression (*P≤0.05, **P≤0.01, ***P≤0.001, Student's t-test). Data are mean±s.e.m. Scale bar: 200 μm.

Fig. 7.

Increased proliferation in the Hesx1Cre/+;KrasG12D/+ and Hesx1Cre/+;BrafV600E/+ mutant pituitaries. (A-I) Immunofluorescent staining against Ki67 (green) on histological sections of mutant and control embryos; genotypes and stages are indicated. (J) Quantitative analyses of the number of Ki67+ cells out of the total DAPI+ cells (blue) demonstrate a significant increase in the proliferation index in the Hesx1Cre/+;KrasG12D/+ at 18.5 dpc and Hesx1Cre/+;BrafV600E/+ at 14.5 and 18.5 dpc (*P≤0.05, **P≤0.01, ***P≤0.001, Student's t-test). Data are mean±s.e.m. Scale bars: 200 μm.

Fig. 8.

The Sox2+ stem cell compartment is increased in Hesx1Cre/+;KrasG12D/+ and Hesx1Cre/+;BrafV600E/+ mutant pituitaries. (A-I) Immunofluorescence staining revealing the presence of SOX2+ cells (green) in the developing pituitary; genotypes and stages are indicated. The overall numbers appear elevated in the mutant pituitary relative to the controls at all stages analysed. Blue staining is DAPI. (J) Quantitative analyses demonstrate that the number of SOX2+ cells is significantly increased in the mutant pituitaries compared with controls at 14.5 and 18.5 dpc. (K) Absolute quantitative RT-PCR analysis of Sox2 mRNA expression in mutant and control pituitaries at 18.5 dpc. (L,M) Culture of dissociated cells reveals a significantly higher clonogenic potential of the mutant pituitaries relative to the controls (*P≤0.05, **P≤0.01, ***P≤0.001, Student's t-test). Data are mean±s.e.m. Scale bars: 200 μm.

Fig. 9.

Abnormal balance between self-renewal and differentiation in the Hesx1Cre/+;BrafV600E/+ mutant pituitaries. The nucleotide analogue EdU was administered once to pregnant females at 14.5 dpc and embryos were analysed at 16.5 dpc. (A-C) Numbers of SOX2;EdU double-positive cells are increased in the mutant pituitary relative to the control. (D-F) In contrast, the proportion of PIT1;EdU double-positive cells is dramatically reduced in the mutant gland. (G-I) The proportion of TPIT;EdU double-positive cells remained constant between genotypes. (J-L) The percentage of cycling cells within the initially EdU-labelled population is significantly higher in the Hesx1Cre/+;BrafV600E/+ mutants compared with controls (*P≤0.05, ***P≤0.001; NS, not significant; Student's t-test). Data are mean±s.e.m. Scale bars: 200 μm.

Fig. 10.

Human PCP tumours contain a population of cycling SOX2+ cells. Analysis of histological sections of human PCP tumours. (A,A′) Haematoxylin and Eosin staining showing the presence of fibrovascular cores (black asterisks) lined by a layer of basal cells (arrowheads) and large areas of suprabasal squamous epithelium (white asterisks). (B,B′) Immunohistochemistry revealing the expression of BRAF-V600E in a PCP tumour. Staining is present (light brown) in the vast majority of tumour cells. Black asterisks indicate the presence of fibrovascular cores. (C,C′) Immunohistochemical detection using a pan-cytokeratin antibody. Most of the tumour cells show positive staining (dark brown) with the exception of the basal cell layer (arrowheads) surrounding the fibrovascular cores (black asterisks). (D,D′) Specific staining (light brown) against cytokeratin 19 showing absent staining in the basal layer cells (arrowheads) surrounding fibrovascular cores (asterisks). The difference in the staining, from light to dark brown, is due to technical reasons (e.g. antibody used, manual or automatized immunohistochemistry). (E,E′) Immunofluorescence staining of pERK1/2 showing expression restricted to the cells surrounding the fibrovascular cores (white arrowheads), with some positive cells within the fibrovascular cores (white asterisk). Outlined areas in A-E are shown in A′-E′. Outlined areas in A′-E′ are shown in the insets in A′-E′. (F-H) Double immunofluorescent staining shows the co-expression of SOX2 and SOX9 in cells surrounding the fibrovascular cores, with stronger signal in the basal cells. (I-K) Double immunofluorescence staining reveals the co-expression of SOX2 and Ki67 in basal cells. (L) Quantification of SOX2:Ki67 double-positive cells as a proportion of either the SOX2+ or Ki67+ populations in human PCP. Data are mean±s.e.m. of SOX2+ KI67+ and KI67+ SOX2+ cell quantifications. Scale bars: 200 μm; 100 μm in insets.