1. Stem Cells and Regenerative Medicine
Download icon

Pituitary stem cells produce paracrine WNT signals to control the expansion of their descendant progenitor cells

  1. John P Russell
  2. Xinhong Lim
  3. Alice Santambrogio
  4. Val Yianni
  5. Yasmine Kemkem
  6. Bruce Wang
  7. Matthew Fish
  8. Scott Haston
  9. Anaëlle Grabek
  10. Shirleen Hallang
  11. Emily J Lodge
  12. Amanda L Patist
  13. Andreas Schedl
  14. Patrice Mollard
  15. Roel Nusse
  16. Cynthia L Andoniadou  Is a corresponding author
  1. Centre for Craniofacial and Regenerative Biology, King’s College London, United Kingdom
  2. Skin Research Institute of Singapore, Agency for Science, Technology and Research, Singapore
  3. Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore
  4. Department of Medicine III, University Hospital Carl Gustav Carus, Technische Universität Dresden, Germany
  5. Institute of Functional Genomics (IGF), University of Montpellier, CNRS, France
  6. Howard Hughes Medical Institute, Stanford University School of Medicine, Department of Developmental Biology, Stanford University School of Medicine, United States
  7. Department of Medicine and Liver Center, University of California San Francisco, United States
  8. Developmental Biology and Cancer, Birth Defects Research Centre, UCL GOS Institute of Child Health, United Kingdom
  9. Université Côte d'Azur, Inserm, CNRS, France
Research Article
  • Cited 0
  • Views 480
  • Annotations
Cite this article as: eLife 2021;10:e59142 doi: 10.7554/eLife.59142

Abstract

In response to physiological demand, the pituitary gland generates new hormone-secreting cells from committed progenitor cells throughout life. It remains unclear to what extent pituitary stem cells (PSCs), which uniquely express SOX2, contribute to pituitary growth and renewal. Moreover, neither the signals that drive proliferation nor their sources have been elucidated. We have used genetic approaches in the mouse, showing that the WNT pathway is essential for proliferation of all lineages in the gland. We reveal that SOX2+ stem cells are a key source of WNT ligands. By blocking secretion of WNTs from SOX2+ PSCs in vivo, we demonstrate that proliferation of neighbouring committed progenitor cells declines, demonstrating that progenitor multiplication depends on the paracrine WNT secretion from SOX2+ PSCs. Our results indicate that stem cells can hold additional roles in tissue expansion and homeostasis, acting as paracrine signalling centres to coordinate the proliferation of neighbouring cells.

Introduction

How stem cells interact with their surrounding tissue has been a topic of investigation since the concept of the stem cell niche was first proposed (Schofield, 1978). Secreted from supporting cells, factors such as WNTs, FGFs, SHH, EGF, and cytokines regulate the activity of stem cells (Nabhan et al., 2018; Palma et al., 2005; Tan and Barker, 2014). Furthermore, communication is known to take place in a bidirectional manner (Doupé et al., 2018; Tata and Rajagopal, 2016).

The anterior pituitary (AP) is a major primary endocrine organ that controls key physiological functions including growth, metabolism, reproduction, and the stress response and exhibits tremendous capability to remodel its constituent hormone populations throughout life, in response to physiological demand. It contains a population of Sox2 expressing stem cells that self-renew and give rise to lineage-committed progenitors and functional endocrine cells (Andoniadou et al., 2013; Rizzoti et al., 2013). During embryonic development, SOX2+ undifferentiated precursor cells of Rathke’s pouch, the pituitary anlage (Arnold et al., 2011; Castinetti et al., 2011; Fauquier et al., 2008; Pevny and Rao, 2003), generate all committed endocrine progenitor lineages, defined by the absence of SOX2 and expression of either POU1F1 (PIT1), TBX19 (TPIT), or NR5A1 (SF1) (Bilodeau et al., 2009; Davis et al., 2011). These committed progenitors are proliferative and give rise to the hormone-secreting cells. Demand for hormone secretion rises after birth, resulting in dramatic organ growth and expansion of all populations by the second postnatal week (Carbajo-Pérez and Watanabe, 1990; Taniguchi et al., 2002). SOX2pituitary stem cells (PSCs) are most active during this period, but the bulk of proliferation and organ expansion during postnatal stages derives from SOX2 committed progenitors. The activity of SOX2+ PSCs gradually decreases and during adulthood is minimally activated even following physiological challenge (Andoniadou et al., 2013; Gaston-Massuet et al., 2011; Gremeaux et al., 2012; Zhu et al., 2015). By adulthood, progenitors carry out most of the homeostatic functions, yet SOX2+ PSCs persist throughout life in both mice and humans (Gonzalez-Meljem et al., 2017; Xekouki et al., 2019). The signals driving proliferation of committed progenitor cells are not known, and neither is it known if SOX2+ PSCs can influence this process beyond their minor contribution of new cells.

The self-renewal and proliferation of numerous stem cell populations rely on WNT signals (Basham et al., 2019; Lim et al., 2013; Takase and Nusse, 2016; Wang et al., 2015; Yan et al., 2017). WNTs are necessary for the initial expansion of Rathke’s pouch as well as for PIT1 lineage specification (Osmundsen et al., 2017; Potok et al., 2008). In the postnatal pituitary, the expression of WNT pathway components is upregulated during periods of expansion and remodelling. Gene expression comparisons between neonatal and adult pituitaries or in GH-cell ablation experiments (Gremeaux et al., 2012; Willems et al., 2016) show that the WNT pathway is upregulated during growth and regeneration.

Our previous work revealed that during disease, the paradigm of supporting cells signalling to the stem cells may be reversed; mutant stem cells expressing a degradation-resistant β-catenin in the pituitary promote cell non-autonomous development of tumours through their paracrine actions (Andoniadou et al., 2013; Gonzalez-Meljem et al., 2017). Similarly, degradation-resistant β-catenin expression in hair follicle stem cells led to cell non-autonomous WNT activation in neighbouring cells promoting new growth (Deschene et al., 2014). In the context of normal homeostasis, stem cells have been shown to influence daughter cell fate in the mammalian airway epithelium and the Drosophila gut via ‘forward regulation’ models, where the fate of a daughter cell is directed by a stem cell via juxtacrine Notch signalling (Ohlstein and Spradling, 2007; Pardo-Saganta et al., 2015). It remains unknown if paracrine stem cell action can also promote local proliferation in normal tissues.

Here, we used genetic approaches to determine if paracrine stem cell action takes place in the AP and to discern the function of WNTs in pituitary growth. Our results demonstrate that postnatal pituitary expansion, largely driven by committed progenitor cells, depends on WNT activation. Importantly, we show that SOX2+ PSCs are the key regulators of this process, acting through secretion of WNT ligands acting in a paracrine manner on neighbouring progenitors. Identification of this forward-regulatory model elucidates a previously unidentified function for stem cells during tissue expansion.

Results

WNT-responsive cells in the pituitary include progenitors driving major postnatal expansion

To clarify which cells respond to WNT signals in the postnatal AP, we first characterised the AP cell types activating the WNT pathway at P14, a peak time for organ expansion and a time point when a subpopulation of SOX2+ stem cells are proliferative. The Axin2-CreERT2 mouse line (van Amerongen et al., 2012) has been shown to efficiently label cells with activated WNT signalling in the liver, lung, breast, skin, testes, and endometrium among other tissues (Lim et al., 2013; Moiseenko et al., 2017; Syed et al., 2020; van Amerongen et al., 2012; Wang et al., 2015). Axin2 positive cells were labelled by GFP following tamoxifen induction in Axin2CreERT2/+;ROSA26mTmG/+ mice and pituitaries were analysed 2 days post-induction. We carried out double immunofluorescence staining using antibodies against uncommitted (SOX2), lineage committed (PIT1, TPIT, SF1), and hormone-expressing endocrine cells (GH, PRL, TSH, ACTH, or FSH/LH) together with antibodies against GFP labelling the WNT-activated cells. We detected WNT-responsive cells among all the different cell types of the AP including SOX2+ PSCs, the three committed populations and all hormone-secreting cells (Figure 1A, Figure 1—figure supplement 1A).

Figure 1 with 2 supplements see all
Axin2 expressing cells contribute to pituitary growth and expansion of all lineages.

(A) Immunofluorescence staining against GFP (green) with markers of pituitary stem cells (PSCs) or lineage commitment (magenta) in Axin2CreERT2/+; ROSA26mTmG/+ pituitaries harvested from mice induced at P14 and lineage traced for 2 days (top panel) and 14 days (bottom panel). Scale bar: 10 μm. (B) Quantification of lineage expansion between 2 and 14 days following induction at P14. Graph shows that the proportion of lineage committed cells (either PIT1+, TPIT+, or SF1+) and PSCs (SOX2+), that is, that are transcription factor (TF)+ cells that are GFP+ increases between 2 days (black bars) and 14 days (grey bars) post-induction. PIT1 p=0.000004, TPIT p=0.008 multiple t-tests. n = 4 animals per time point. (C) Immunofluorescence staining against GFP (green) in pituitaries harvested from Axin2CreERT2/+;ROSA26mTmG/+ mice induced at P14 and lineage traced for 2 days, 2 weeks, and 8 weeks. Bottom panel shows magnified fields of view of regions of interest indicated by white boxes in panels above. Scale bars: 50 μm. (D) Top panel showing the quantification of the proportion of all cells in Axin2CreERT2/+;ROSA26mTmG/+ pituitaries that are GFP+ at 2, 7, 14, 28, and 56 days post-induction as analysed by flow cytometry. Days 2–7 p<0.0001 unpaired t-test. Data points show individual measurements from biological replicates, n = 4–8 pituitaries per time point. (Bottom) Graph of the absolute number of GFP+ cells (green) and as a proportion of total cells (blue) at the time points indicated. (E) X-gal staining in Axin2CreERT2/+;ROSA26LacZ/+ pituitaries harvested from mice induced at P14 and lineage traced for 8 weeks (left) and 1 year (right). Scale bars: 500 μm. (F) Model summarising the major contribution of WNT-responsive progenitors of all lineages to pituitary growth, in addition to that of SOX2+ PSCs.

To confirm if the three committed lineages as well as uncommitted SOX2+ PSCs all expand in response to WNT, we further lineage traced Axin2-expressing cells for 14 days after tamoxifen administration at P14. Double labelling revealed an increase in all four populations between 2 and 14 days (Figure 1A,B). This increase reached significance for the PIT1 (13.7% at 2 days to 30.3% at 14 days, p=0.000004) and TPIT (3.78% to 11.03%, p=0.008) populations, but not SF1 (0.5% to 4%, n.s.). As this time course ends at P28 at the commencement of puberty, we repeated the analysis for SF1 cells to P42, which spans puberty and the expansion of gonadotrophs (Figure 1—figure supplement 1B). This reveals a significant expansion in WNT-responsive SF1+ cells as a proportion of the total SF1+ population (p=0.0048, n = 3). Lineage tracing of the PIT1-derivates (GH+ somatotrophs, PRL+ lactotrophs, and TSH+ thyrotrophs) reveals that there is a preferential expansion of somatotrophs and thyrotrophs (Figure 1—figure supplement 1C). Only a minority of SOX2+ PSCs were WNT-responsive at 2 days (0.57%) and this population expanded to 2% at 14 days (n.s.), suggesting that these are self-renewing. GFP+ cells were traced for a period of 8 weeks post-induction, which revealed that WNT-responsive descendants continued to expand at the same rate as the rest of the pituitary (n = 4–8 mice per time point at P16, P21, P28, P42, and P70) (Figure 1C,D). The time period between 2 and 7 days saw the greatest increase in GFP+ cells, during which the labelled population nearly tripled in size (Figure 1D). The persistence of labelled cells was evident in longer-term traces using the ROSA26lacZ/+ reporter (Axin2CreERT2/+;ROSA26lacZ/+), up to a year following induction at P14 (Figure 1E, n = 4). Clonal analysis using the Confetti reporter demonstrated that individual Axin2-expressing cells (Axin2CreERT2/+;ROSA26Confetti/+) gave a greater contribution after 4 weeks compared to lineage tracing from Sox2-expressing cells (Sox2CreERT2/+;ROSA26Confetti/+), in support of predominant expansion from WNT-responsive lineage-committed progenitors (Figure 1—figure supplement 1D).

To establish if signalling mediated by β-catenin is necessary for organ expansion we carried out deletion of Ctnnb1 in the Axin2+ population from P14 during normal growth (Axin2CreERT2/+;Ctnnb1lox(ex2-6)/lox(ex2-6) hereby Axin2CreERT2/+;Ctnnb1LOF/LOF). Due to morbidity, likely due to Axin2 expression in other organs, we were limited to analysis up to 5 days post-induction. Deletion of Ctnnb1 resulted in a significant reduction in the number of dividing cells, marked by pH-H3 (40% reduction, Figure 1—figure supplement 2A, p=0.0313, n = 3), confirming that activation of the WNT pathway is necessary for expansion of the pituitary populations. This deletion did not result in significant differences in overall numbers among the three lineages, as determined by the numbers of PIT1+, SF1+, or ACTH+ cells among the targeted population (Figure 1—figure supplement 2B, n = 4 controls, two mutants). The number of SOX2+ stem cells and cells undergoing cell death also remained unaffected during the 5-day period (Figure 1—figure supplement 2C and D). Taken together, these results confirm that postnatal AP expansion depends on WNT-responsive progenitors across all lineages, in addition to SOX2+ PSCs (Figure 1F).

WNT/β-catenin signalling is required for long-term AP expansion from SOX2PSCs

We further explored the role of WNT pathway activation in postnatal SOX2+ stem cells. To permanently mark WNT-responsive cells and their descendants whilst simultaneously marking SOX2+ PSCs, we combined the tamoxifen-inducible Axin2CreERT2/+;ROSA26tdTomato/+ with the Sox2Egfp/+ strain, where cells expressing SOX2 are labelled by enhanced green fluorescent protein (EGFP) (Axin2CreERT2/+;Sox2Egfp/+;ROSA26tdTomato/+). Following tamoxifen administration from P21, tdTomato- and EGFP-labelled cells were analysed by flow sorting after 72 hr, by which point all induced cells robustly express tdTomato (Figure 2A, Figure 2—figure supplement 1). Double-labelled cells comprised 23.4% of the SOX2+ population (n = 5 individual pituitaries) (Figure 2A, arrowheads), with the majority of tdTomato+ cells found outside of the SOX2+ compartment. It was previously shown that only around 2.5–5% of SOX2+ PSCs has clonogenic potential through in vitro assays (Andoniadou et al., 2012; Andoniadou et al., 2013; Pérez Millán et al., 2016). To determine if WNT-responsive SOX2+ cells are stem cells capable of forming colonies, we isolated double-positive tdTomato+;EGFP+ cells (i.e. Axin2+;Sox2+) as well as the single-expressing populations and plated these in equal numbers in stem cell-promoting media at clonal densities (Figure 2B). Double-positive tdTomato+;EGFP+ cells showed a significant increase in the efficiency of colony formation compared to single-labelled EGFP+ cells (average 9% compared to 5%, n = 5 pituitaries, p=0.0226, Mann–Whitney U-test [two-tailed]), demonstrating WNT-responsive SOX2+ PSCs have a greater clonogenic potential under these in vitro conditions, confirming in vivo data in Figure 1B. As expected from previous work, none of the single-labelled tdTomato+ cells (i.e. SOX2 negative) was able to form colonies (Andoniadou et al., 2012).

Figure 2 with 2 supplements see all
Activation of WNT signalling in SOX2pituitary stem cells (PSCs) and their descendants is necessary for long-term growth.

(A) Schematic of the experimental timeline used in panels A and B. Endogenous expression of tdTomato (magenta, Axin2 targeted cells) and EGFP (green, Sox2 expressing cells) in Axin2CreERT2/+;Sox2Egfp/+;ROSA26tdTomato/+ pituitaries harvested at P24 sectioned in the frontal plane. Nuclei are counterstained with Hoechst in the merged panel. Scale bar: 50 μm. (B) A representative culture plate showing colonies derived from Tomato+, EGFP+, or Tomato+;EGFP+ cells that were isolated from Axin2CreERT2/+;Sox2Egfp/+;ROSA26tdTomato/+ pituitaries by fluorescence-activated cell sorting (FACS) plated in stem cell promoting media at clonogenic densities and stained with crystal violet (left panel). The proportion of colony-forming cells in each subpopulation was quantified by counting the number of colonies per well (right panel). Each data point indicates individual wells, n = 5 separate pituitaries. p=0.0226, Mann–Whitney U-test (two-tailed). Scale bar: 10 mm. (C) Immunofluorescence staining against SOX2 (green) and Ki-67 (magenta) in Sox2+/+Ctnnb1LOF/LOF (control) and Sox2CreERT2/+Ctnnb1LOF/LOF (mutant) pituitaries from mice induced at P14 and analysed 22 weeks after induction (at P168) (bottom panel). Scale bar: 50 μm. (D) Dorsal view of whole mount pituitaries of Sox2+/+;Ctnnb1LOF/LOF (control) and Sox2CreERT2/+;Ctnnb1LOF/LOF (mutant), 22 weeks after induction (i.e. P168). Scale bars: 1 mm. (E) Model summarising the effect of Ctnnb1 deletion in SOX2+ PSCs. PL, posterior lobe; IL, intermediate lobe; AL, anterior lobe.

To confirm that PSCs with active WNT signalling through β-catenin have a greater propensity to form colonies in vitro, we analysed postnatal pituitaries from TCF/Lef:H2B-EGFP mice, reporting the activation of response to WNT signals. This response is detected through expression of an EGFP-tagged variant of histone H2B, which is incorporated into chromatin and diluted in descendants with cell division (Ferrer-Vaquer et al., 2010). Therefore, cells responding to, or having recently responded to, WNT as well as their immediate descendants will be EGFP+. At P21, EGFP+ cells were abundant in all three lobes and particularly in the marginal zone harbouring SOX2+ stem cells (Figure 2—figure supplement 2A). Through double mRNA in situ hybridisation against Egfp and Sox2 in TCF/Lef:H2B-EGFP pituitaries, we confirmed that Sox2-expressing cells activate H2B-EGFP expression at this time point (Figure 2—figure supplement 2B). Isolation by fluorescence-activated cell sorting (FACS) and in vitro culture of the postnatal EGFP+ compartment revealed an enrichment of cells with clonogenic potential in the EGFPHigh fraction compared to EGFPLow or negative cells (Figure 2—figure supplement 2C, n = 5 pituitaries). Together these results reveal that a proportion of postnatal SOX2+ stem cells respond to WNTs through downstream β-catenin/TCF/LEF signalling and that these cells have greater clonogenic capacity in vitro.

To further address the role of the canonical WNT response in the activity of SOX2+ PSCs in vivo, we expressed a loss-of-function allele of β-catenin specifically in Sox2-expressing cells (Sox2CreERT2/+;Ctnnb1lox(ex2-6)/lox(ex2-6) hereby Sox2CreERT2/+;Ctnnb1LOF/LOF) from P14. Twenty-two weeks following induction, at P168, there was a substantial drop in the number of cycling cells in the pituitary of Sox2CreERT2/+;Ctnnb1LOF/LOF mutants compared to Sox2+/+;Ctnnb1LOF/LOF controls (Figure 2C, n = 2 pituitaries per genotype). This was accompanied by AP hypoplasia following the loss of Ctnnb1 in SOX2+ PSCs (Figure 2D). Therefore, in this small sample size, the proliferative capacity of Ctnnb1-deficient SOX2+ PSCs and of their descendants was impaired long term, leading to reduced growth. In vivo genetic tracing of targeted cells over the 22-week period (Sox2CreERT2/+;Ctnnb1LOF/+;ROSA26mTmG/+ compared to Sox2CreERT2/+;Ctnnb1LOF/LOF;ROSA26mTmG/+ pituitaries) revealed that targeted (Ctnnb1-deficient) SOX2+ PSCs were capable of giving rise to the three committed lineages PIT1, TPIT, and SF1 (Figure 2—figure supplement 2D), indicating that the loss of Ctnnb1 does not prevent differentiation of SOX2+ PSCs into the three lineages. Downregulation of β-catenin was confirmed by immunofluorescence in SOX2+ (mGFP+) derivatives (Figure 2—figure supplement 2E). Although limited by a small sample size, we conclude that WNT/β-catenin signalling is likely required cell-autonomously in SOX2+ stem cells and their descendants (Figure 2E).

SOX2+ stem cells express WNT ligands

Having established that WNT activation is responsible for promoting proliferation in the AP, we next focused on identifying the source of WNT ligands. Axin2 expressing cells from Axin2CreERT2/+;ROSA26mTmG/+ mice were labelled at P14 by tamoxifen induction. Cells expressing Axin2 at the time of induction are labelled by GFP expression in the membrane. Double immunofluorescence staining for GFP together with SOX2 revealed that Axin2 expressing cells (mGFP+) are frequently located in close proximity to SOX2+ PSCs (Figure 3A). Two-dimensional quantification of the two cell types revealed that over 50% of mGFP+ cells were in direct contact with SOX2+ nuclei (n = 3 pituitaries, >500 SOX2+ cells per gland, Figure 3A). The analysis did not take into account the cellular processes of SOX2+ cells. These results led us to speculate that SOX2+ PSCs may be a source of key WNT ligands promoting proliferation of lineage-committed cells.

Figure 3 with 1 supplement see all
SOX2pituitary stem cells (PSCs) are as a source of WNT ligands in the pituitary.

(A) Immunofluorescence staining against GFP (green) and SOX2 (magenta) in Axin2CreERT2/+; ROSA26mTmG/+ pituitaries 48 hr post-induction. Graph representing a quantification of the proximity of individual GFP+ cells to the nearest SOX2+ cell as quantified by the number of nuclei separating them. Plotted data represents the proportion of GFP+ cells that fall into each category of the total GFP+ cells, taken from n = 3 separate pituitaries. Scale bars: 50 μm. (B) Experimental paradigm for RNA Seq analysis of Sox2 positive and negative cells. (C) Graphs representing the FPKM values of Wls and Porcupine in Sox2 positive and negative cells (black and grey bars, respectively). mRNA in situ hybridisation for Sox2 and for Wls on wild-type sagittal pituitaries at P14, demonstrating strong Wls expression in the marginal zone epithelium. Scale bars: 250 μm. (D) Bar chart showing the FPKM values of Wnt genes in the Sox2+ and Sox2 fractions. Double mRNA in situ hybridisation against Wnt2, Wnt5a, and Wnt9a (blue) together with Sox2 (red) validating expression in the Sox2+ population. Boxed regions through the marginal zone epithelium are magnified. Scale bars: 100 μm and 50 μm in boxed inserts.

In order to determine if SOX2+ PSCs express WNT ligands, we carried out gene expression profiling of SOX2+ and SOX2 populations at P14, through bulk RNA-sequencing. Pure populations of Sox2-expressing cells excluding lineage-committed populations were isolated from Sox2Egfp/+ male and female pituitaries at P14 based on EGFP expression as previously shown (Andoniadou et al., 2012; Figure 3B, Figure 3—figure supplement 1A). Analysis of global gene expression signatures using ‘gene set enrichment analysis’ (GSEA) (Subramanian et al., 2005) identified a significant enrichment of molecular signatures related to epithelial-to-mesenchymal transition, adherens, and tight junctions in the EGFP+ fraction, characteristic of the SOX2+ population (Figure 3—figure supplement 1B). The SOX2+ fraction also displayed enrichment for genes associated with several signalling pathways known to be active in these cells, including epidermal growth factor receptor (EGFR) (Iwai-Liao et al., 2000), Hippo (Lodge et al., 2016; Lodge et al., 2019; Xekouki et al., 2019), MAPK (Haston et al., 2017), FGF (Higuchi et al., 2017), Ephrin (Yoshida et al., 2015; Yoshida et al., 2017), and p53 (Gonzalez-Meljem et al., 2017; Figure 3—figure supplement 1C, Supplementary file 1). Additionally, PI3K, TGFβ, and BMP pathway genes were significantly enriched in the SOX2+ population (Figure 3—figure supplement 1C, Supplementary file 1). Query of the WNT-associated genes did not suggest a global enrichment in WNT targets (e.g. enrichment of Myc and Jun, but not of Axin2 or Lef1) (Figure 3—figure supplement 1D, Supplementary file 1). Instead, SOX2+ PSCs expressed a unique transcriptomic fingerprint of key pathway genes including Lgr4, Znrf3, Rnf43 capable of regulating WNT signal intensity in SOX2+ PSCs, as well as enriched expression of the receptors Fzd1, Fzd3, Fzd4, Fzd6, and Fzd7 (Figure 3—figure supplement 1D). The predominant R-spondin gene expressed in the pituitary was Rspo4, specifically by the EGFP-negative fraction (Figure 3—figure supplement 1D). The gene profiling revealed that Wls expression encoding Gpr177/WLS, a necessary mediator of WNT ligand secretion (Carpenter et al., 2010; Takeo et al., 2013; Wang et al., 2015), is enriched in SOX2+ PSCs (Figure 3C). Analysis of Wnt gene expression confirmed enriched expression of Wnt2, Wnt5a, and Wnt9a in SOX2+ PSCs, and the expression of multiple additional Wnt genes by both fractions at lower levels (SOX2+ fraction: Wnt5b, Wnt6, Wnt16; SOX2 fraction: Wnt2, Wnt2b, Wnt3, Wnt4, Wnt5a, Wnt5b, Wnt9a, Wnt10a, Wnt16) (Figure 3D). These results reveal that SOX2+ PSCs express the essential components to regulate activation of the WNT pathway and express Wnt genes as well as the necessary molecular machinery to secrete WNT ligands.

Paracrine signalling from SOX2+ stem cells promotes WNT activation

We sought to conclusively determine if WNT secretion specifically from SOX2+ PSCs drives proliferation of surrounding cells in the postnatal pituitary gland. We proceeded to delete Wls only in the Sox2-expressing population (Sox2CreERT2/+;Wlsfl/fl) from P14 by a series of tamoxifen injections. Due to morbidity, we limited analyses to 1 week following induction. Pituitaries appeared mildly hypoplastic at P21 along the medio-lateral axis (Figure 4—figure supplement 1, n = 4 controls and n = 5 mutants). To determine if this is a result of reduced proliferation, we carried out immunofluorescence using antibodies against Ki-67 and SOX2. This revealed significantly fewer cycling cells in the SOX2 population of Sox2CreERT2/+;Wlsfl/fl mutant pituitaries compared to Sox2+/+;Wlsfl/fl controls (10.326% Ki-67 in control [n = 4] compared to 3.129% in mutant [n = 5], p=0.0008, unpaired t-test) (Figure 4A). Additionally, we observed a reduction of cycling cells within the SOX2+ population (5.582% Ki-67 in control compared to 2.225% in induced Sox2CreERT2/+;Wlsfl/fl mutant pituitaries, p=0.0121, unpaired t-test) (Figure 4A), resulting in a smaller SOX2+ cell pool in mutants (23.425% SOX2+/total AP cells in Sox2+/+;Wlsfl/fl controls compared to 19.166% SOX2+/total AP cells in induced Sox2CreERT2/+;Wlsfl/fl mutant pituitaries, p=0.0238, Student’s t-test, n = 5 mutants, four controls). To determine if reduced levels of WNT activation accompanied this phenotype, we carried out double mRNA in situ hybridisation using specific probes against Lef1 and Sox2. There was an overall reduction in Lef1 expression in mutants compared to controls (n = 4 per genotype), in which we frequently observed robust expression of Lef1 transcripts in close proximity to cells expressing Sox2 (arrows, Figure 4B). Together, our data support a paracrine role for SOX2PSCs in driving the expansion of committed progeny through the secretion of WNT ligands (Figure 4C).

Figure 4 with 1 supplement see all
Paracrine secretion of WNTs from SOX2pituitary stem cells (PSCs) is necessary for expansion of committed cells.

(A) Immunofluorescence staining against SOX2 (green) and Ki-67 (magenta) in Sox2+/+;Wlsfl/fl (control) and Sox2CreERT2/+;Wlsfl/fl (mutant) pituitaries induced from P14 and analysed after 1 week. Nuclei were counterstained with Hoechst. (i and ii) represent magnified fields of view of regions indicated by white boxes in top panels. Scale bars: 50 μm. Graph of quantification of cycling cells marked by Ki-67 among cells negative for SOX2. Values represent mean ± SEM, p=0.0008, unpaired t-test. Graph of quantification of cycling cells marked by Ki-67 among SOX2-positive cells. Values represent mean ± SEM, p=0.0121, unpaired t-test. Each data point shows the mean of one biological replicate, n = 4 pituitaries from controls and five pituitaries from mutants. (B) Double mRNA in situ hybridisation using specific probes against Lef1 (blue) and Sox2 (red) in control and mutant pituitaries following tamoxifen induction from P14 and tracing for 7 days. Scale bars: 250 μm and 50 μm in boxed regions. (C) Model summarising paracrine WNT secretion from SOX2+ PSCs to lineage-committed progenitors and the effects of abolishing WNT secretion from SOX2+ PSCs through the deletion of Wls.

Discussion

Emerging disparities between the archetypal stem cell model, exhibited by the haematopoietic system, and somatic stem cells of many organs have led to the concept that stem cell function can be executed by multiple cells not fitting a typical stem cell paradigm (Clevers and Watt, 2018). In organs with persistent populations possessing typical functional stem cell properties yet contributing minimally to turnover and repair, the necessity for such classical stem cells is questioned. Here we show that WNT signalling is required for postnatal pituitary growth by both SOX2+ PSCs and SOX2 committed progenitors. We identify an additional discreet function for SOX2+ PSCs, where these signal in a feedforward manner by secreting WNT ligands as cues to stimulate proliferation and promote tissue growth.

Consistent with previous reports, our data support that SOX2+ PSCs contribute, but do not carry out the majority of tissue expansion during the postnatal period (Zhu et al., 2015); instead, new cells primarily derive from more committed progenitors, which we show to be WNT-responsive. We demonstrate that this population of lineage-restricted WNT-responsive cells rapidly expands and contributes long-lasting clones from postnatal stages. It remains to be shown if cells among the SOX2 lineage-committed populations may also fall under the classical definition of a stem cell. Preventing secretion of WNT ligands from SOX2+ PSCs reveals that far from being dispensable, paracrine actions of the SOX2+ population that are inactive in their majority are necessary for anterior lobe expansion from lineage-committed populations. In the adrenal gland, R-spondins are necessary for cortical expansion and zonation, where deletion of Rspo3, expressed by the capsule that contains adrenocortical stem cells, results in reduced proliferation of the underlying steroidogenic cells (Vidal et al., 2016). Corroborating a model where committed pituitary progenitors depend on the paracrine actions of SOX2+ PSCs, Zhu and colleagues observed that in pituitaries with reduced numbers of PSCs, proliferation among PIT1+ cells was significantly impaired (Zhu et al., 2015). It would be intriguing to see if there is a reduction in WNT signalling in this model, or following genetic ablation of adult SOX2+ PSCs (Roose et al., 2017).

We show that a subpopulation of SOX2+ PSCs in the postnatal gland are also WNT-responsive and have greater in vitro colony-forming potential under defined conditions. This colony-forming potential is normally a property of a minority of SOX2+ PSCs at any given age and reflects their in vivo proliferative capacity (Andoniadou et al., 2012; Rizzoti et al., 2013). A role for the WNT pathway in promoting SOX2+ cell activity is supported by studies showing that pathogenic overexpression of β-catenin promotes their colony-forming ability (Sarkar et al., 2016) and their in vivo expansion (Andoniadou et al., 2012). Additionally, elevated WNT pathway activation has been described for pituitary side-population cells, enriched for SOX2+ stem cells from young, compared to old pituitaries (Gremeaux et al., 2012). This is in line with our findings that the WNT pathway has an important function in promoting the activation of SOX2+ PSCs. It remains to be shown if this response relies on autocrine WNT-signalling as for other stem cells (Lim et al., 2013); however, our results reveal reduced proliferation among SOX2+ PSCs and reduced SOX2+ cell numbers when WNT secretion from these cells is abolished, supportive of either autocrine signalling or paracrine signalling between different subsets of the SOX2+ population.

The mechanism preventing the majority of SOX2+ PSCs from responding to WNT signals remains elusive but points to heterogeneity among the population. Such regulation could occur at the level of receptor signalling; we have shown by bulk transcriptomic profiling that SOX2+ PSCs express the receptors required to respond to the WNT pathway, but also express high levels of the frizzled inhibitor Znrf3, and the R-spondin receptor Lgr4. One conceivable scenario is that high levels of Znrf3 inhibit frizzled receptors in the absence of R-spondin under normal physiological conditions, supressing a WNT response. In support of this, R-spondins have been shown to promote pituitary organoid formation (Cox et al., 2019). Whether the R-spondin/LGR/ZNRF3 module is active under physiological conditions needs to be determined. Furthermore, well-described factors expressed in PSCs are known to have inhibitory effects on β-catenin-mediated transcription, such as YAP/TAZ (Azzolin et al., 2014; Gregorieff et al., 2015) and SOX2 itself (Alatzoglou et al., 2011; Kelberman et al., 2008).

In summary, we demonstrate an alternative mechanism for stem cell contribution to homeostasis, whereby these can act as paracrine signalling hubs to promote local proliferation. Applicable to other organs, this missing link between SOX2+ PSCs and committed cell populations of the AP is key for basic physiological functions and renders stem cells integral to organ expansion.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or
reference
IdentifiersAdditional
information
Genetic reagent (Mus musculus)Axin2CreERT2/+Roel Nusse, Stanford University
The Jackson Laboratory
JAX:018867, RRID:IMSR_JAX:018867
Genetic reagent (Mus musculus)Sox2CreERT2/+(Andoniadou et al., 2013)
PMID:24094324 DOI: 10.1016/j.stem.2013.07.004
MGI:5512893
Genetic reagent (Mus musculus)ROSA26mTmG/mTmGThe Jackson LaboratoryJAX:007676, RRID:IMSR_JAX:007676
Genetic reagent (Mus musculus)ROSA26Confetti/ConfettiThe Jackson LaboratoryJAX:017492, RRID:IMSR_JAX:017492
Genetic reagent (Mus musculus)ROSA26tdTomato/tdTomatoThe Jackson LaboratoryJAX:007909, RRID:IMSR_JAX:007909
Genetic reagent (Mus musculus)Ctnnb1fl(ex2-6)/ fl(ex2-6) (CtnnbLOF/LOF)The Jackson LaboratoryJAX:004152, RRID:IMSR_JAX:004152
Genetic reagent (Mus musculus)Wlsfl/flThe Jackson LaboratoryJAX:012888, RRID:IMSR_JAX:012888
Genetic reagent (Mus musculus)Sox2eGFP/+Ellis et al., 2004
PMID:15711057 DOI: 10.1159/000082134
MGI:3589809
Genetic reagent (Mus musculus)TCF/Lef:H2B-GFPThe Jackson LaboratoryJAX:013752, RRID:IMSR_JAX:013752
Cell line (Mus musculus)Primary anterior pituitary cellsThis paperN/AFreshly isolated from Mus musculus.
AntibodyAnti-GFP (Chicken Polyclonal)Abcamab13970, RRID:AB_300798IF(1:400)
AntibodyAnti-SOX2 (Goat Polyclonal)Immune Systems LtdGT15098, RRID:AB_2195800IF(1:200)
AntibodyAnti-SOX2(Rabbit Monoclonal)Abcamab92494, RRID:AB_10585428IF(1:100)
AntibodyAnti-SOX9(Rabbit Monoclonal)Abcamab185230, RRID:AB_2715497IF(1:500)
AntibodyAnti-POU1F1 (PIT1) (Rabbit Monoclonal)Gifted by Dr S. J. Rhodes (IUPUI, USA)422_Rhodes, RRID:AB_2722652IF(1:500)
AntibodyAnti-SF1 (NR5A1, clone N1665)(Mouse Monoclonal)Thermo Fisher Scientific434200, RRID:AB_2532209IF(1:300)
AntibodyAnti-TBX19 (TPIT), (Rabbit Polyclonal)Gifted by Dr J. Drouin (Montreal Clinical Research Institute, Canada)Ac1250 #71, RRID:AB_2728662IF(1:200)
AntibodyAnti-Ki67 (Rabbit Monoclonal)Abcamab15580, RRID:AB_443209IF(1:100)
AntibodyAnti-pH-H3 (Rabbit Polyclonal)Abcamab5176, RRID:AB_304763IF(1:500)
AntibodyAnti-GH (Rabbit Polyclonal)National Hormone and Peptide Program (NHPP)AFP-5641801IF(1:1000)
AntibodyAnti-TSH (Rabbit Polyclonal)National Hormone and Peptide Program (NHPP)AFP-1274789IF(1:1000)
AntibodyAnti-PRL (Rabbit Polyclonal)National Hormone and Peptide Program (NHPP)AFP-4251091IF(1:1000)
AntibodyAnti-ACTH(Mouse Monoclonal)Fitzgerald10C-CR1096M1, RRID:AB_1282437IF(1:400)
AntibodyAnti-LH (Rabbit Polyclonal)National Hormone and Peptide Program (NHPP)AFP-697071PIF(1:300)
AntibodyAnti-FSH (Rabbit Polyclonal)National Hormone and Peptide Program (NHPP)AFP-HFS6IF(1:300)
AntibodyAnti-ZO-1 (Rat Monoclonal)Santa CruzSC33725, RRID:AB_628459IF(1:300)
AntibodyAnti-E-CADHERIN(Rabbit Monoclonal)Cell Signaling3195S, RRID:AB_2291471IF(1:300)
AntibodyAnti-Rabbit 488 (Goat Polyclonal)Life TechnologiesA11008, RRID:AB_143165IF(1:400)
AntibodyAnti-Rabbit 555 (Goat Polyclonal)Life TechnologiesA21426, RRID:AB_1500929IF(1:400)
AntibodyAnti-Rabbit 633 (Goat Polyclonal)Life TechnologiesA21050, RRID:AB_141431IF(1:400)
AntibodyAnti-Goat 488 (Donkey Polyclonal)Abcamab150133, RRID:AB_2832252IF(1:400)
AntibodyAnti-Chicken 488 (Goat Polyclonal)Life TechnologiesA11039, RRID:AB_142924IF(1:400)
AntibodyAnti-Chicken 647 (Goat Polyclonal)Life TechnologiesA21449, RRID:AB_1500594IF(1:400)
AntibodyAnti-Rat 555 (Goat Polyclonal)Life TechnologiesA21434, RRID:AB_141733IF(1:400)
AntibodyAnti-Mouse 555 (Goat Polyclonal)Life TechnologiesA21426, RRID:AB_1500929IF(1:400)
AntibodyAnti-Rabbit Biotinylated (Donkey Polyclonal)Abcamab6801, RRID:AB_954900IF(1:400)
AntibodyAnti-Rabbit Biotinylated (Goat Polyclonal)Abcamab207995IF(1:400)
AntibodyAnti-Mouse Biotinylated (Goat Biotinylated)Abcamab6788, RRID:AB_954885IF(1:400)
Sequence-based reagentRNAscope probe M. musculus Axin2Advanced Cell Diagnostics400331
Sequence-based reagentRNAscope probe M. musculus Lef1Advanced Cell Diagnostics441861
Sequence-based reagentRNAscope probe M. musculus WlsAdvanced Cell Diagnostics405011
Sequence-based reagentRNAscope probe M. musculus Rspo1Advanced Cell Diagnostics401991
Sequence-based reagentRNAscope probe M. musculus Rspo2Advanced Cell Diagnostics402001
Sequence-based reagentRNAscope probe M. musculus Rspo3Advanced Cell Diagnostics402011
Sequence-based reagentRNAscope probe M. musculus Rspo4Advanced Cell Diagnostics402021
Sequence-based reagentRNAscope probe M. musculus Lgr4Advanced Cell Diagnostics318321
Sequence-based reagentRNAscope probe M. musculus Wnt9aAdvanced Cell Diagnostics405081
Sequence-based reagentRNAscope probe M. musculus Wnt2Advanced Cell Diagnostics313601
Sequence-based reagentRNAscope probe M. musculus Wnt5aAdvanced Cell Diagnostics316791
Sequence-based reagentRNAscope probe eGFPAdvanced Cell Diagnostics400281
Sequence-based reagentRNAscope probe M. musculus JunAdvanced Cell Diagnostics453561
Sequence-based reagentRNAscope probe M. musculus Axin2 (Channel 2)Advanced Cell Diagnostics400331-C2
Sequence-based reagentRNAscope probe M. musculus Sox2 (Channel 2)Advanced Cell Diagnostics401041-C2
Sequence-based reagentRNAscope probe eGFP (Channel 2)Advanced Cell Diagnostics400281-C2
Sequence-based reagentRNAscope probe M. musculus Sox2Advanced Cell Diagnostics401041
Sequence-based reagentRNAscope probe M. musculus Pou1f1Advanced Cell Diagnostics486441
Sequence-based reagentRNAscope probe Duplex Positive Control Ppib-C1, Polr2a-C2Advanced Cell Diagnostics321641
Sequence-based reagentRNAscope probe Duplex Negative Control DapB (both channels)Advanced Cell Diagnostics320751
Sequence-based reagentRNAscope probe Singleplex Positive Control PpibAdvanced Cell Diagnostics313911
Sequence-based reagentRNAscope probe: Singleplex Negative Control DapBAdvanced Cell Diagnostics310043
Peptide, recombinant proteinStreptavidin 488Life TechnologiesS11223IF(1:400)
Peptide, recombinant proteinStreptavidin 555Life TechnologiesS32355IF(1:400)
Peptide, recombinant proteinStreptavidin 633Life TechnologiesS21375IF(1:400)
Commercial assay or kitRNAScope 2.5 HD Assay-REDAdvanced Cell Diagnostics322350
Commercial assay or kitRNAScope 2.5 HD Duplex AssayAdvanced Cell Diagnostics322430
Commercial assay or kitLIVE/DEAD Fixable Near IR-Dead Cell Stain KitLife TechnologiesL34975
Commercial assay or kitFIX and PERM Cell Permeabilization KitLife TechnologiesGAS003
Chemical compound, drugTamoxifenSigmaT5648
Chemical compound, drugCorn OilSigmaC8267
Chemical compound, drugCollagenase Type 2Worthington4178C
Chemical compound, drug10× TrypsinSigma59418C
Chemical compound, drugDeoxyribonuclease IWorthingtonLS002172
Chemical compound, drugFungizoneGibco15290
Chemical compound, drugHank’s Balanced Salt Solution (HBSS)Gibco14170
Chemical compound, drugFoetal Bovine SerumSigmaF2442
Chemical compound, drugHEPESThermo Fisher15630
Chemical compound, drugbFGFR&D Systems233-FB-025
Chemical compound, drugCholera ToxinSigmaC8052
Chemical compound, drugDMEM-F12Thermo Fisher31330
Chemical compound, drugPenicillin/StreptomycinGibco15070-063
Chemical compound, drugNeutral Buffered FormalinSigmaHT501128
Chemical compound, drugHoechst 33342Thermo FisherH35701:1000
Chemical compound, drugDeclereSigmaD3565
Chemical compound, drugNeo-ClearSigma65351-M
Software, algorithmFlowJoFlowJo, LLChttps://www.flowjo.com/
RRID:SCR_008520
Software, algorithmPrism 7GraphPad Softwarehttps://www.graphpad.com/
Software, algorithmImage LabBio-Rad Laboratorieshttp://www.bio-rad.com/
Software, algorithmNDP ViewHamamatsu Photonicshttps://www.hamamatsu.com/
Software, algorithmHISAT v2.0.3Kim et al., 2015https://github.com/infphilo/hisat2
RRID:SCR_015530
Software, algorithmDESeq2 v2.11.38Love et al., 2014https://github.com/Bioconductor-mirror/DESeq2
RRID:SCR_015687
Software, algorithmfeatureCounts v1.4.6p5Liao et al., 2014http://subread.sourceforge.net/
RRID:SCR_012919
Software, algorithmThe Galaxy PlatformAfgan et al., 2016; Blankenberg et al., 2010; Goecks et al., 2010https://usegalaxu.org
RRID:SCR_006281
Software, algorithmGene Set Enrichment Analysis (GSEA)Subramanian et al., 2005software.broadinstitute.org/gsea/index.jsp
RRID:SCR_003199
Software, algorithmCufflinksTrapnell et al., 2012https://github.com/cole-trapnell-lab/cufflinks
RRID:SCR_014597
OtherDeposited Data, RNA-SeqBioProject (NCBI)PRJNA421806

Mice

All procedures were performed under compliance of the Animals (Scientific Procedures) Act 1986, Home Office License (P5F0A1579). KCL Biological Services Unit staff undertook daily animal husbandry. Genotyping was performed on ear biopsies taken between P11 and P15 by standard PCR using the indicated primers. These experiments were not conducted at random and the experimenters were not blind while conducting the animal handling and assessment of tissue. Images are representative of the respective genotypes. For all studies, both male and female animals were used and results combined.

The Sox2CreERT2/+ and Sox2Egfp/+ strains were kept on a CD-1 background. Axin2CreERT2/+ animals were kept on a mixed background of C57BL/6 backcrossed onto CD-1 for five generations and were viable and fertile, with offspring obtained at the expected Mendelian ratios. ROSA26mTmG/mTmG, ROSA26Confetti/Confetti, ROSA26tdTomato/tdTomato, Wlsfl/fl, Ctnnb1fl(ex2-6)/ fl(ex2-6), and TCF/LEF:H2B-EGFP mice were kept on a mixed background of 129/Sv backcrossed onto CD-1 for at least three generations. For lineage tracing studies, male Axin2CreERT2/+ or Sox2CreERT2/+ mice were bred with homozygous ROSA26mTmG/mTmG or ROSA26Confetti/Confetti dams to produce the appropriate allele combinations on the reporter background. Pups were induced at P14 or P15 with a single dose of tamoxifen (resuspended to 20 mg/ml in Corn Oil with 10% ethanol) by intraperitoneal injection, at a concentration of 0.15 mg/g of body weight. Pituitaries were harvested at the indicated time points post-induction and processed for further analysis as described below. Mice were harvested from different litters for each time point at random. For litters in which there was a surplus of experimental mice, multiple samples were harvested for each required time point.

For Wntless deletion studies, Sox2CreERT2/+;Wlsfl/+;ROSA26mTmG/mTmG males were bred with Wlsfl/fl;ROSA26mTmG/mTmG dams, to produce Sox2CreERT2/+;Wlsfl/+;ROSA26mTmG/mTmG, Sox2CreERT2/+;Wlsfl/fl;ROSA26mTmG/mTmG, and Wlsfl/fl;ROSA26mTmG/mTmG offspring. Pups of the indicated genotypes received intraperitoneal injections of 0.15 mg of tamoxifen per gram body weight on four consecutive days, beginning at P14, and harvested 3 days after the final injection.

For the β-catenin loss-of-function experiments, either Sox2CreERT2/+;Ctnnb1fl(ex2-6)/+;ROSA26mTmG/mTmG or Axin2CreERT2/+;Ctnnb1fl(ex2-6)/+;ROSA26mTmG/mTmG males were crossed with Ctnnb1fl(ex2-6)/fl(ex2-6);ROSA26mTmG/mTmG dams. Axin2CreERT2/+;Ctnnb1fl(ex2-6)/fl(ex2-6);ROSA26mTmG/mTmG and Axin2CreERT2/+;Ctnnb1fl(ex2-6)/+;ROSA26mTmG/mTmG pups were induced with a single dose of tamoxifen, at a concentration of 0.15 mg/g of body weight and kept alive for 7 days before harvesting. Sox2CreERT2/+;Ctnnb1fl(ex2-6)/+;ROSA26mTmG/mTmG and Sox2CreERT2/+;Ctnnb1fl(ex2-6)/fl(ex2-6);ROSA26mTmG/mTmG pups received two intraperitoneal injections of tamoxifen, at a concentration of 0.15 mg/g of body weight, on two consecutive days and were kept alive for the indicated length of time before harvesting.

TCF/LEF:H2B-EGFP mice were culled and the pituitaries harvested at the indicated ages for the respective experiments. For FACS experiments, mice were harvested at 21 days of age. Axin2CreERT2/+;Sox2eGFP/+ males were crossed with ROSA26tdTomat/tdTomato dams to produce Axin2CreERT2/+;Sox2eGFP/+;ROSA26tdTomato/+ that were induced with single doses of tamoxifen at 21 and 22 days of age and harvested 3 days after the first injection for FACS experiments.

Flow cytometry analysis of lineage traced pituitaries

Request a detailed protocol

For the quantification of cells by flow cytometry, anterior lobes of Axin2CreERT2/+;ROSA26mTmG/+ mice dissected at the indicated time points. The posterior and intermediate lobes were dissected from the anterior lobes under a dissection microscope. Untreated ROSA26mTmG/+ and wild-type pituitaries from age-matched litters were used as tdTomato only and negative controls, respectively. Dissected pituitaries were incubated in Enzyme Mix (0.5% w/v collagenase type 2 [Lorne Laboratories], 0.1× Trypsin [Gibco], 50 μg/ml DNase I [Worthington], and 2.5 μg/ml Fungizone [Gibco] in Hank’s Balanced Salt Solution [HBSS] [Gibco]) in a cell culture incubator for up to 3 hr; 850 ml of HBSS was added to each Eppendorf in order to quench the reaction. Pituitaries were dissociated by agitation, pipetting up and down 100× at first with a 1 ml pipette, followed by 100× with a 200 μl pipette. Cells were transferred to a 15 ml Falcon tube and resuspended in 9 ml of HBSS and spun down at 200 g for 5 min. The supernatant was aspirated, leaving behind the cell pellet that was resuspended in PBS and spun down at 1000 rpm for 5 min before being resuspended in a Live/Dead cell stain (Life Technologies, L34975) prepared to manufacturer’s instructions, for 30 min in the dark. Cells were washed in PBS as above. The pellet was resuspended in FIX and PERM Cell Permeabilization Kit (Life Technologies, GAS003) prepared as per manufacturer’s instructions for 10 min at room temperature. Cells were washed as above, and the pellet was resuspended in 500 µl of FACS buffer (1% foetal calf serum [Sigma], 25 mM HEPES in PBS) and filtered through 70 μm filters (BD Falcon), into 5 ml round bottom polypropylene tubes (BD Falcon). One minute prior to analysis, 1 µl of Hoechst was added to the suspended cells and incubated. Samples were analysed on a BD Fortessa and gated according to negative and single fluorophore controls. Single cells were gated according to SSC-A and SSC-W. Dead cells were excluded according to DAPI (2 ng/ml, incubated for 2 min prior to sorting). GFP+, tdTomato+, and GFP+;tdTomato+ cells were gated according to negative controls in the PE-A and FITC-A channels.

FACS for sequencing or colony forming assays

Request a detailed protocol

For FACS, the anterior lobes from Sox2eGFP/+, TCF/LEF:H2B-GFP, or Axin2CreERT2/+;Sox2eGFP/+;ROSA26tdTomato/+ and their respective controls were dissected and dissociated as above. After dissociation cells were spun down at 200 g in HBSS and the pellet was resuspended in 500 µl FACS buffer. Using an Aria III FACs machine (BD systems), samples were gated according to negative controls, and where applicable single fluorophore controls. Experimental samples were sorted according to their fluorescence, as indicated, into tubes containing either RNAlater (Qiagen) for RNA isolation or 1 ml of Pit Complete Media for culture (Pit Complete: 20 ng/ml) bFGF and 50 ng/ml of cholera toxin in ‘Pit Basic’ media (DMEM-F12 with 5% foetal calf serum, 100 U/ml penicillin, and 100 μg/ml streptomycin). Cells were plated in 12-well plates at clonal density, approximately 500 cells/well. Colonies were incubated for a total of 7 days before being fixed in 10% neutral buffered formalin (NBF) (Sigma) for 10 min at room temperature, washed for 5 min, three times, with PBS and stained with crystal violet in order for the number of colonies to be quantified.

RNA-sequencing

Request a detailed protocol

Total RNA was isolated from each sample and following poly-A selection, cDNA libraries were generated using TruSeq (Clontech, 634925). Barcoded libraries were then pooled at equal molar concentrations and sequenced on an Illumina Hiseq 4000 instrument in a 75 base pair, paired-end sequencing mode, at the Wellcome Trust Centre for Human Genetics (Oxford, United Kingdom). Raw sequencing reads were quality checked for nucleotide calling accuracy and trimmed accordingly to remove potential sequencing primer contaminants. Following QC, forward and reverse reads were mapped to GRCm38/mm10 using Hisat2 (Kim et al., 2015). Using a mouse transcriptome specific GTF as a guide, FeatureCounts (Liao et al., 2014) was used to generate gene count tables for every sample. These were utilised within the framework of the Deseq2 (Love et al., 2014) and FPKM values (generated by FPKM count Wang et al., 2012) were processed using the Cufflinks (Trapnell et al., 2012) pipelines that identified statistically significant gene expression differences between the sample groups. Following identification of differentially expressed genes (at an FDR < 0.05) we focused on identifying differentially expressed pathways using a significance threshold of FDR < 0.05 unless otherwise specified. The gene lists used for GSEA were as found on the BROAD institute GSEA MSigDBv.7 ‘molecular signatures database’. The deposited data set (BioProject, accession PRJNA421806) can be accessed through the following link: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA421806.

Immunofluorescence and microscopy

Request a detailed protocol

Freshly harvested pituitaries were washed in PBS for 10 min before being fixed in 10% NBF for 18 hr at room temperature. In short, embryos and whole pituitaries were washed in PBS three times, before being dehydrated through a series of 1 hr washes in 25%, 50%, 70%, 80%, 90%, 95%, and 100% ethanol. Tissues were washed in Neo-Clear (Sigma) at room temperature for 10 min, then in fresh preheated Neo-Clear at 60°C for 10 min. Subsequently, tissues were incubated in a mixture of 50% Neo-Clear:50% paraffin wax at 60°C for 15 min followed by three changes of pure wax for a minimum of 1 hr washes at 60°C, before being orientated to be sectioned in the frontal plane. Embedded samples were sectioned at 5 µm and mounted on to Super Frost+ slides.

For immunofluorescence, sections were deparaffinised in Neo-Clear by three washes of 10 min, washed in 100% ethanol for three times 5 min, and rehydrated in a series of 5-min ethanol washes up to distilled water (95%, 90%, 80%, 70%, 50%, 25%, H2O). Heat induced epitope retrieval was performed with 1× DeClear Buffer (citrate pH 6) in a Decloaking chamber NXGEN (Menarini Diagnostics) for 3 min at 110°C. Slides were left to cool to room temperature before proceeding to block for 1 hr at room temperature in blocking buffer (0.2% BSA, 0.15% glycine, 0.1% TritonX in PBS) with 10% serum (sheep or donkey, depending on secondary antibodies). Primary antibodies were diluted in blocking buffer with 1% of the appropriate serum and incubated overnight at 4°C. Slides were washed three times for 10 min with PBST. Slides were incubated with secondary antibodies diluted 1:400 in blocking buffer with 1% serum for 1 hr at room temperature. Slides were washed three times with PBST as above. Where biotinylated secondary antibodies were used, slides were incubated with streptavidin diluted 1:400 in blocking buffer with 1% serum for 1 hr at room temperature. Finally, slides were washed with PBST and mounted using Vectashield Antifade Mounting Medium (Vector Laboratories, H-1000).

The following antibodies, along with their dilutions and detection technique, were used: GFP (1:400, Alexa Fluor-488 or −647 secondary), SOX2 raised in goat (1:200, Alexa Fluor-488 secondary), SOX2 raised in rabbit (1:100, biotinylated secondary), SOX9 (1:500, biotinylated secondary), PIT1 (1:500, biotinylated secondary), SF1 (1:300, biotinylated secondary), TPIT (1:200, biotinylated secondary), Ki-67 (1:100, biotinylated secondary), pH-H3 (1:500, biotinylated secondary), GH (1:1000, biotinylated secondary), TSH (1:1000, biotinylated secondary), PRL (1:1000, biotinylated secondary), ACTH (1:400, Alexa Fluor-555 secondary), LH/FSH (1:300, biotinylated secondary), ZO-1 (1:300, Alexa Fuor-488), and E-Cadherin (1:300, Alexa Fluor-488). Nuclei were visualised with Hoechst (1:1000). Images were taken on a TCS SPS Confocal (Leica Microsystem) with a 20× objective for analysis.

mRNA in situ hybridisation

Request a detailed protocol

All mRNA in situ hybridisations were performed using the RNAscope singleplex or duplex chromogenic kits (Advanced Cell Diagnostics) on formalin fixed paraffin embedded sections processed as described in the above section. The protocol followed the manufacturer’s instructions with slight modifications. ImmEdge Hydrophobic Barrier PAP Pen (Vector Laboratories, H-4000) was used to draw a barrier around section while air-drying following the first ethanol washes. Pretreatment followed the standard length of time for pituitaries (12 min), while embryos were boiled for 10 min. For singleplex, the protocol proceeded to follow the instructions exactly. For duplex, Amplification nine was extended to 1 hr and the dilution of the Green Detection reagent was increased to 1:30. For both protocols, sections were counterstained with Mayer’s Haematoxylin (Vector Laboratories, H-3404), left to dry at 60°C for 30 min before mounting with VectaMount Permanent Mounting Medium (Vector Laboratories, H-5000). Slides were scanned using a Nanozoomer-XR Digital Slide Scanner (Hamamatsu) and processed using Nanozoomer Digital Pathology View (Hamamatsu).

Quantification of cells

Request a detailed protocol

Cell numbers were quantified in ImageJ using the cell counter plugin (Schindelin et al., 2012). At a minimum, three sections per pituitary were quantified, spaced no less than 100 µM apart in the tissue.

Statistics

All statistical analyses were performed in GraphPad Prism. Data points in graphs represent the mean values of recordings from a single biological replicate unless otherwise stated.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
    The relationship between the spleen colony-forming cell and the haemopoietic stem cell
    1. R Schofield
    (1978)
    Blood Cells 4:7–25.
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67

Decision letter

  1. Marianne E Bronner
    Senior and Reviewing Editor; California Institute of Technology, United States
  2. Elaine Emmerson
    Reviewer; The University of Edinburgh, United Kingdom
  3. Sally Camper
    Reviewer; University of Michigan Medical School, United States
  4. Lori T Raetzman
    Reviewer; University of Illinois at Urbana-Champaign, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This study uses a number of sophisticated genetic models to test the novel hypothesis that the pituitary stem cells produce WNT as a paracrine signal to drive proliferation of neighboring cells. The strengths of this study are that the authors use both in vitro and in vivo models to robustly test their hypothesis in a highly quantitative fashion. This study is of relevance to understanding development, disease, regeneration, aging and cancer of the pituitary gland, as well as stem cells, paracrine signalling and regeneration in other glandular and non-glandular organs.

Decision letter after peer review:

Thank you for submitting your article "Pituitary stem cells produce paracrine WNT signals to control the expansion of their descendant progenitor cells" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Marianne Bronner as the Senior and Reviewing Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Elaine Emmerson (Reviewer #1); Sally Camper (Reviewer #2); Lori T Raetzman (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, when editors judge that a submitted work as a whole belongs in eLife but that some conclusions require a modest amount of additional new data, as they do with your paper, we are asking that the manuscript be revised to either limit claims to those supported by data in hand, or to explicitly state that the relevant conclusions require additional supporting data.

Our expectation is that the authors will eventually carry out the additional experiments and report on how they affect the relevant conclusions either in a preprint on bioRxiv or medRxiv, or if appropriate, as a Research Advance in eLife, either of which would be linked to the original paper.

Summary:

This is a well-designed study that addresses an important question regarding paracrine signalling in the adult mouse pituitary gland. It fills a critical gap in knowledge regarding pituitary progenitors and also answers a longstanding question as to the role of WNT signaling in pituitary development. It is well written and most of the conclusions are well supported by the data.

Essential revisions:

1) Stating n numbers throughout to be transparent.

2) A more thorough analysis of the Axin2CreERT2/+; Ctnnb1LOF/LOF loss of function with regard to proliferation (to also look at the effect of tamoxifen), cell death and lineage distribution of GH, PRL and TSH cells would be important (e.g. apoptosis; SOX2; PIT1, TPIT, SF-1).

3) If it is not possible to increase n numbers where low (e.g. Figure 2C where n=2) the implications of this for data interpretation and statistics (or lack thereof) should be addressed and discussed in the manuscript.

4) The major point to support that the secretion of WNT from SOX2 cells is the key factor would require looking at the number of stem cells to make sure they are not changed with the WIS deletion. If this can't be done, the authors should tone down the conclusion from this experiment.

5) Clarifying whether both male and female mice were used throughout the study and if not, which sexes were used for which experiments.

The complete reviews are included below for your reference.

Reviewer #1:

This original manuscript demonstrates that SOX2+ pituitary stem cells are a key source of WNT ligands in the adult mouse pituitary gland, and the production of WNT from these cells is essential for paracrine signalling to regulate neighbouring cell proliferation. The approach taken is highly appropriate to address the experimental question, the experiments that were performed are elegant and logical and the conclusions drawn accurately represent the data presented. Appropriate statistical tests have been used and the figures are well presented and provide all relevant data. I have the following comments:

1) I appreciate that breeding sufficient numbers of complex genetic models can often be challenging, but can the authors be confident of the statistical power of experiments where they had n=2 (such as Figure 2C)?

2) It may not be possible, especially under the current circumstances, but have the authors investigated changes in any other markers in the Axin2CreERT2/+; Ctnnb1LOF/LOF mice, for example cell death (e.g. apoptosis); SOX2; or any of the differentiated lineages (PIT1, TPIT, SF-1)? It would be really interesting to see what else changes in the gland in absence of β-catenin signalling.

Reviewer #2:

This is an outstanding manuscript that uses multiple genetically engineered mice, stem cell cultures, and both gain and loss of function models to define the role of WNT signaling from pituitary stem cells to committed hormone-producing cell lineage progenitors in regulation of proliferation. Paracrine WNT secretion from stem cells is essential for the stimulation of proliferation in progenitors that have activated expression of lineage-specific transcription factors but not yet begun to produce hormones. This is an important step in understanding the molecular basis for craniopharyngiomas, which can be caused by gain of function mutations in β catenin, which cause non-cell autonomous acceleration of proliferation in neighboring cells. As such, it is an important advance for the field.

The paper is well-written and scholarly, drawing comparisons between the role of WNTs in pituitary development to the roles in other organ systems. For this reason, the work should be of broad interest for developmental biologists.

The authors conduct lineage tracing to determine which cells are responsive to WNT signaling using Axin2creERT2/+ mice. The choice of postnatal day 14 for induction seems somewhat arbitrary. Is there any information about when WNT signaling is active during development? This may not change the conclusions, but it would be quite interesting to know when it is active. The timing may also bear on finding whether SF1 cells are significantly induced. Is there an expansion of gonadotrope cells around puberty? If so, then inducing at P14 and harvesting at P28, as puberty is commencing, may not be enough time to reach significance for SF1 cells. The authors noted that morbidity necessitated a limitation in the length of time permitted for analysis.

The authors focus on three lineage specific transcription factors: PIT1, TPIT and SF1. Since PIT1 specifies three hormone producing lineages (GH, PRL, and TSH), it could be worthwhile to assess whether there are any differential effects on the final differentiation into the three different cell types.

The authors demonstrate that the most WNT-responsive SOX2 expressing cells have the highest clonogenic potential, and by RNA sequencing of sorted cells they define the WNT ligands and frizzled receptors that the stem cells express. This is a valuable dataset as it also reveals expression of genes in other signaling pathways including Hippo, EGFR, ephrin, Mapk, FGF and p53. Noting the expression of Wls, which is necessary for WNT secretion, the authors deleted it in stem cells and discovered a reduction in proliferating progenitors, clearly establishing the need for WNT secretion in order to expand the stem cell compartment.

Can the authors clarify why a large proportion of cells appear to be tdTomato+ after 72 hrs when it appears in Figure 1D that ~30% are positive after 7 days?

The data presented are quite clear, quantified, and support the conclusions drawn from the data.

Reviewer #3:

This study is an important examination of the role of WNT signaling in postnatal pituitary expansion. The authors have a novel hypothesis that the pituitary stem cells produce WNT as a paracrine signal to drive proliferation of neighboring cells. In fact, paracrine signaling is the main role of the stem cells as opposed to being the primary source of new cells in the postnatal pituitary. It is a bit complicated that the WNT responsive, Axin2 positive cells include both the stem cells and the lineage restricted cells. Additionally, WNTs are made in both stem and lineage restricted cells. This makes it hard to definitively say that WNT released from stem cells impacts the proliferation of the other lineage restricted cells. The experiment deleting Wls from SOX2 positive stem cells was an important way to address the role of paracrine WNT signaling. More details on this experiment are necessary to fully support the assertion that paracrine WNT signaling is an important driver of lineage restricted cell expansion.

1) For the Sox2-Wls conditional deletion experiment, there seems to be far fewer SOX2+ cells in the anterior lobe that in the control, based on the images in Figure 4A and B. If so, is it the loss of SOX2+ cells that leads to the reduced proliferation of the lineage committed cells and not the loss of WNT from the SOX2 cells? Confirming the number of SOX2 cells and contrasting this result to studies in which SOX2 cells are ablated at this time point would help to address this.

There are some additional comments to help clarify the manuscript:

2) The axin2 conditional knockout experiments were only briefly described and the data presented in a supplementary figure. This could be highlighted more in that it shows WNT signaling is necessary for postnatal pituitary expansion. Although outside the scope of what is presented here, a pituitary specific temporally controlled Ctnnb1 knockout would have been helpful to look more long term at postnatal loss of Ctnnb1.

3) At the end of the subsection “WNT/β-catenin signalling is required for long-term anterior pituitary expansion from Sox2+ pituitary stem cells”, it may be better to say the conclusion from those experiments is that WNT signaling is required in SOX2 positive stem cells to promote expansion of all pituitary populations.

4) The PTU-induced hypothyroidism experiment seems out of context with the story. Axin2 cells aren't analyzed in detail. Although presented as a supplementary figure, it might be better to remove this data.

And a few experimental questions:

5) Does the tamoxifen administration alter proliferation in the pituitary? For example, could that have influenced the large expansion in GFP positive cells between PND2 and 7 in the experiments depicted in Figure 1. A control experiment showing proliferation during that time period with and without tamoxifen would help to allay this concern.

6) Why did you switch from analysis at p14 for the experiments in Figure 1 to p21 for the experiments in the first part of Figure 2? Are there any fundamental differences in the pituitary expansion at these ages?

7) For the Ctnnb1 loss of function in Sox2 cells, did you look after 5 days like was done for the Ctnnb1 loss of function in all axin2 positive cells? This may have helped compare directly what the contribution of Ctnnb1 in stem cells is to the overall role of Ctnnb1 in pituitary expansion at that age.

https://doi.org/10.7554/eLife.59142.sa1

Author response

Essential revisions:

1) Stating n numbers throughout to be transparent.

We have ensured that n numbers are stated throughout the manuscript.

2) A more thorough analysis of the Axin2CreERT2/+; Ctnnb1LOF/LOF loss of function with regard to proliferation (to also look at the effect of tamoxifen), cell death and lineage distribution of GH, PRL and TSH cells would be important (e.g. apoptosis; SOX2; PIT1, TPIT, SF-1).

We have investigated the effect of tamoxifen on anterior pituitary proliferation in the Axin2CreERT2/+ model and do not find any differences between control and tamoxifen-injected animals. Please see our response to point 5 of reviewer 3 and Author response image 1.

Author response image 1

We have more thoroughly analysed the Axin2CreERT2/+;Ctnnb1LOF/LOF model as requested. We carried out cell death analysis by immunofluorescence against cleaved caspase-3, as well as analyses of the three committed lineages (by staining against PIT1, SF1 and ACTH) and SOX2 on 2 mutants and 3 control pituitaries. We have not observed significant differences in any of these analyses. The new data are provided as Figure 1—figure supplement 2B-D.

3) If it is not possible to increase n numbers where low (e.g. Figure 2C where n=2) the implications of this for data interpretation and statistics (or lack thereof) should be addressed and discussed in the manuscript.

Thank you for giving us the possibility to discuss the implications of the low numbers for the data interpretation and statistics. Unfortunately we have not been able to generate any further homozygous Sox2CreERT2/+;Ctnnb1LOF/LOF (or Axin2CreERT2/+;Ctnnb1LOF/LOF for new supplementary figures) animals during the pandemic in order to increase sample sizes. Upon lockdown the colonies had to be reduced down to the minimum and our Ctnnb1-LOF colony has suffered consequences, where we are trying to rescue this line. We hope that the reviewers are understanding of this situation. We have highlighted the limitations in the subsection “WNT/β-catenin signalling is required for long-term anterior pituitary expansion from SOX2+ pituitary stem cells”.

4) The major point to support that the secretion of WNT from SOX2 cells is the key factor would require looking at the number of stem cells to make sure they are not changed with the WIS deletion. If this can't be done, the authors should tone down the conclusion from this experiment.

We have analysed the stem cell compartment following Wls deletion. The number of SOX2 cells are reduced in Sox2CreERT2/+;Wlsfl/fl mutants compared to Wlsfl/fl controls (see below). These data support that WNT activation is also required by the SOX2+ cell compartment, a minority of which are also shown to be WNT responsive. We have included these new findings in the Results and refer to them in the Discussion.

Sox2CreERT2/+;Wlsfl/fl mutants: 19.166% SOX2+/Total cells

Wlsfl/fl controls: 23.425% SOX2+/Total cells

P = 0.0238, Student’s t-test, 5 mutants, 4 controls, over 2000 SOX2+ cells counted per genotype.

5) Clarifying whether both male and female mice were used throughout the study and if not, which sexes were used for which experiments.

Both male and female mice have been used throughout the study and the data pooled throughout. We do not expect sexual dimorphism before puberty (beginning around P26). Inductions have been carried out prior to this age. We have clarified this in the Materials and methods.

The complete reviews are included below for your reference.

Reviewer #1:

[…] I have the following comments:

1) I appreciate that breeding sufficient numbers of complex genetic models can often be challenging, but can the authors be confident of the statistical power of experiments where they had n=2 (such as Figure 2C)?

We see an overwhelming reduction in dividing cells throughout the gland following long-term deletion of Ctnnb1 in SOX2+ cells and descendants (n=2), accompanied by pituitary hypoplasia. We do agree with the reviewer that the statistical power of this would be low, and given that we have not been able to increase the sample number, we have indicated the sample size limitations in the subsection “WNT/β-catenin signalling is required for long-term anterior pituitary expansion from SOX2+ pituitary stem cells”.

2) It may not be possible, especially under the current circumstances, but have the authors investigated changes in any other markers in the Axin2CreERT2/+; Ctnnb1LOF/LOF mice, for example cell death (e.g. apoptosis); SOX2; or any of the differentiated lineages (PIT1, TPIT, SF-1)? It would be really interesting to see what else changes in the gland in absence of β-catenin signalling.

Thank you for this suggestion. As per our response to the Editor’s summary comment 2., we have carried out the requested analyses in existing samples to look at apoptosis, SOX2, PIT1, TPIT and SF1 staining by immunofluorescence. We did not observe differences between controls and mutants. The new data are provided as Figure 1—figure supplement 2B-D.

Reviewer #2:

[…] The authors conduct lineage tracing to determine which cells are responsive to WNT signaling using Axin2creERT2/+ mice. The choice of postnatal day 14 for induction seems somewhat arbitrary. Is there any information about when WNT signaling is active during development? This may not change the conclusions, but it would be quite interesting to know when it is active. The timing may also bear on finding whether SF1 cells are significantly induced. Is there an expansion of gonadotrope cells around puberty? If so, then inducing at P14 and harvesting at P28, as puberty is commencing, may not be enough time to reach significance for SF1 cells. The authors noted that morbidity necessitated a limitation in the length of time permitted for analysis.

There is a robust WNT activation in Rathke’s pouch from 11.5dpc. Thank you to the reviewer for bringing up this insightful point. The timeline on Figure 1B does indeed end at P28, before full commencement of puberty, which misses the expansion of SF1+ cells during puberty. We have therefore analysed the expansion of targeted WNT-responsive SF1+ cells at 28 days following induction at P14 (i.e. at P42), which spans puberty. This reveals a significant expansion of WNT-responsive SF1+ cells. The new data are included as Figure 1—figure supplement 1B and described in the subsection “WNT-responsive cells in the pituitary include progenitors driving major postnatal expansion”.

The authors focus on three lineage specific transcription factors: PIT1, TPIT and SF1. Since PIT1 specifies three hormone producing lineages (GH, PRL, and TSH), it could be worthwhile to assess whether there are any differential effects on the final differentiation into the three different cell types.

Thank you for the suggestion. We have analysed expansion of the three PIT1-derivatives and these data are presented as new Figure 1—figure supplement 1C. There is a preferential expansion of somatotrophs and thyrotrophs but not of lactotrophs. The data are described in the subsection “WNT-responsive cells in the pituitary include progenitors driving major postnatal expansion”.

The authors demonstrate that the most WNT-responsive SOX2 expressing cells have the highest clonogenic potential, and by RNA sequencing of sorted cells they define the WNT ligands and frizzled receptors that the stem cells express. This is a valuable dataset as it also reveals expression of genes in other signaling pathways including Hippo, EGFR, ephrin, Mapk, FGF and p53. Noting the expression of Wls, which is necessary for WNT secretion, the authors deleted it in stem cells and discovered a reduction in proliferating progenitors, clearly establishing the need for WNT secretion in order to expand the stem cell compartment.Can the authors clarify why a large proportion of cells appear to be tdTomato+ after 72 hrs when it appears in Figure 1D that ~30% are positive after 7 days?

The tdTomato reporter more closely resembles Axin2 mRNA expression as seen by RNAScope. This is likely due to increased sensitivity of the R26-tdTomato reporter, recombining in more cells with lower Axin2 expression than the R26-mTmG reporter. Additionally, the dosage of tamoxifen in Figure 2A (tdTomato) is double the amount used in Figure 1C (mTmG). For clarity, we have added a timeline for the mTmG experiments in Figure 1—figure supplement 1A.

The data presented are quite clear, quantified, and support the conclusions drawn from the data.

Reviewer #3:

This study is an important examination of the role of WNT signaling in postnatal pituitary expansion. The authors have a novel hypothesis that the pituitary stem cells produce WNT as a paracrine signal to drive proliferation of neighboring cells. In fact, paracrine signaling is the main role of the stem cells as opposed to being the primary source of new cells in the postnatal pituitary. It is a bit complicated that the WNT responsive, Axin2 positive cells include both the stem cells and the lineage restricted cells. Additionally, WNTs are made in both stem and lineage restricted cells. This makes it hard to definitively say that WNT released from stem cells impacts the proliferation of the other lineage restricted cells. The experiment deleting Wls from SOX2 positive stem cells was an important way to address the role of paracrine WNT signaling. More details on this experiment are necessary to fully support the assertion that paracrine WNT signaling is an important driver of lineage restricted cell expansion.

1) For the Sox2-Wls conditional deletion experiment, there seems to be far fewer SOX2+ cells in the anterior lobe that in the control, based on the images in Figure 4A and B. If so, is it the loss of SOX2+ cells that leads to the reduced proliferation of the lineage committed cells and not the loss of WNT from the SOX2 cells? Confirming the number of SOX2 cells and contrasting this result to studies in which SOX2 cells are ablated at this time point would help to address this.

We show that SOX2 cells proliferate less in the absence of secreted WNTs (2.225% in mutants compared to 5.582% in controls). In addition based on the reviewer’s suggestion we show that they are fewer in number (19.166% in mutants compared with 23.425% in controls, see point 4 in response to the Editor’s summary, results in the subsection “Paracrine signalling from SOX2+ stem cells promotes WNT activation”). However, cycling SOX2+ cells are still present, at higher numbers than in other studies (Zhu et al., 2015, SOX2 cell reduction; Roose et al., 2017, SOX2 cell ablation). Even so, following Wls deletion, the reduction in surrounding cycling cells in the mutant is overwhelming, after only 7 days, where this cannot be explained by a reduction in proliferation of direct descendants of mutant SOX2 cells. Our confetti traces suggest that over a longer period, the majority of traced Sox2 cells only appear to undergo one or two divisions (Figure 1—figure supplement 1D). This argues that the reduced proliferation we see is not due to commitment and expansion of SOX2 cells to a proliferative progenitor.

We do support the view that the mild reduction in SOX2 cells may be additive to the phenotype stemming from loss of WNT secretion following the deletion of Wls, reducing WNT levels even further. In the Zhu et al., 2015 study, where SOX2 cell numbers are reduced in Rbp-J mutants, proliferation of surrounding cells is also reduced. This points to a paracrine effect, especially given the low contribution of SOX2 cells to homeostasis. This hypothesis is discussed in the Discussion.

There are some additional comments to help clarify the manuscript:

2) The axin2 conditional knockout experiments were only briefly described and the data presented in a supplementary figure. This could be highlighted more in that it shows WNT signaling is necessary for postnatal pituitary expansion. Although outside the scope of what is presented here, a pituitary specific temporally controlled Ctnnb1 knockout would have been helpful to look more long term at postnatal loss of Ctnnb1.

Thank you for the suggestion. We did not place further emphasis on the Axin2-driven Ctnnb1 deletions, since the analysis is short-term and limited to 5 days post-induction due to morbidity. We believe that the key deletions of Ctnnb1 in Sox2-expressing cells and their descendants illustrate this point with clarity (Figure 2C). Carried out over 22 weeks, the near-complete loss of Ki-67 positive cells and evident pituitary hypoplasia show that b-catenin is essential for pituitary expansion. This is consolidated in Figure 4, which reveals that without WNT signals from SOX2+ cells there is a reduction in cycling cells throughout the anterior pituitary.

At present there are no drivers to temporally control the Ctnnb1 deletion only in the pituitary. Hesx1Cre/+;Ctnnb1LOF/- mice do not survive and neither do Hesx1Cre/+;Wlsfl/fl, which was our alternative approach. We do not know of any inducible drivers solely targeting the pituitary postnatally or any loss of function mutations in Ctnnb1 on reversible systems e.g. TetOn.

3) At the end of the subsection “WNT/β-catenin signalling is required for long-term anterior pituitary expansion from SOX2+ pituitary stem cells”, it may be better to say the conclusion from those experiments is that WNT signaling is required in SOX2 positive stem cells to promote expansion of all pituitary populations.

We have changed this sentence to: “In conclusion, WNT/b-catenin signalling is required cell-autonomously in SOX2+ stem cells and their descendants (Figure 2E).”

4) The PTU-induced hypothyroidism experiment seems out of context with the story. Axin2 cells aren't analyzed in detail. Although presented as a supplementary figure, it might be better to remove this data.

Upon the reviewer’s recommendation, we have removed these data on elevation of Axin2 expression following PTU-induced hypothyroidism.

And a few experimental questions:

5) Does the tamoxifen administration alter proliferation in the pituitary? For example, could that have influenced the large expansion in GFP positive cells between PND2 and 7 in the experiments depicted in Figure 1. A control experiment showing proliferation during that time period with and without tamoxifen would help to allay this concern.

From numerous past studies at the doses administered, tamoxifen administration has not been show to alter proliferation in the pituitary. To allay any concerns of the reviewer we carried out a control experiment on the Axin2CreERT2/+;ROSA26mTmG/+ model and provide the data in Author response image 1. We do not see any difference in cycling cells marked by Ki-67, between corn oil or tamoxifen-injected pituitaries at P14, analysed 48 hours later.

6) Why did you switch from analysis at p14 for the experiments in Figure 1 to p21 for the experiments in the first part of Figure 2? Are there any fundamental differences in the pituitary expansion at these ages?

We do not detect any fundamental differences in clonogenic capacity at between P17 (P14+3 days) and P24 (P21+3 days). As the percentage of double positive (tdTomato+;EGFP+) cells is low, we opted for the experiments starting at P21 due to the larger pituitaries. This ensured we had sufficient numbers of double positive cells per experiment to increase the accuracy of colony counting.

7) For the Ctnnb1 loss of function in Sox2 cells, did you look after 5 days like was done for the Ctnnb1 loss of function in all axin2 positive cells? This may have helped compare directly what the contribution of Ctnnb1 in stem cells is to the overall role of Ctnnb1 in pituitary expansion at that age.

We did not look after 5 days as we tried to maximise the tracing periods whilst animals remain healthy. We agree with the reviewer that this would help a direct comparison and would be worthwhile to pursue in future studies.

https://doi.org/10.7554/eLife.59142.sa2

Article and author information

Author details

  1. John P Russell

    Centre for Craniofacial and Regenerative Biology, King’s College London, London, United Kingdom
    Contribution
    Conceptualization, Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
  2. Xinhong Lim

    1. Skin Research Institute of Singapore, Agency for Science, Technology and Research, Singapore, Singapore
    2. Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore, Singapore
    Contribution
    Resources, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4725-5161
  3. Alice Santambrogio

    1. Centre for Craniofacial and Regenerative Biology, King’s College London, London, United Kingdom
    2. Department of Medicine III, University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany
    Contribution
    Formal analysis, Investigation
    Competing interests
    No competing interests declared
  4. Val Yianni

    Centre for Craniofacial and Regenerative Biology, King’s College London, London, United Kingdom
    Contribution
    Software, Formal analysis, Investigation
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-9857-7577
  5. Yasmine Kemkem

    Institute of Functional Genomics (IGF), University of Montpellier, CNRS, Montpellier, France
    Contribution
    Resources, Investigation
    Competing interests
    No competing interests declared
  6. Bruce Wang

    1. Howard Hughes Medical Institute, Stanford University School of Medicine, Department of Developmental Biology, Stanford University School of Medicine, Stanford, United States
    2. Department of Medicine and Liver Center, University of California San Francisco, San Francisco, United States
    Contribution
    Resources
    Competing interests
    No competing interests declared
  7. Matthew Fish

    Howard Hughes Medical Institute, Stanford University School of Medicine, Department of Developmental Biology, Stanford University School of Medicine, Stanford, United States
    Contribution
    Resources, Investigation
    Competing interests
    No competing interests declared
  8. Scott Haston

    Developmental Biology and Cancer, Birth Defects Research Centre, UCL GOS Institute of Child Health, London, United Kingdom
    Contribution
    Investigation
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-3928-4808
  9. Anaëlle Grabek

    Université Côte d'Azur, Inserm, CNRS, Nice, France
    Contribution
    Resources
    Competing interests
    No competing interests declared
  10. Shirleen Hallang

    Centre for Craniofacial and Regenerative Biology, King’s College London, London, United Kingdom
    Contribution
    Investigation
    Competing interests
    No competing interests declared
  11. Emily J Lodge

    Centre for Craniofacial and Regenerative Biology, King’s College London, London, United Kingdom
    Contribution
    Investigation
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-0932-8515
  12. Amanda L Patist

    Centre for Craniofacial and Regenerative Biology, King’s College London, London, United Kingdom
    Contribution
    Investigation
    Competing interests
    No competing interests declared
  13. Andreas Schedl

    Université Côte d'Azur, Inserm, CNRS, Nice, France
    Contribution
    Resources, Supervision
    Competing interests
    No competing interests declared
  14. Patrice Mollard

    Institute of Functional Genomics (IGF), University of Montpellier, CNRS, Montpellier, France
    Contribution
    Resources, Supervision, Funding acquisition, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-2324-7589
  15. Roel Nusse

    Howard Hughes Medical Institute, Stanford University School of Medicine, Department of Developmental Biology, Stanford University School of Medicine, Stanford, United States
    Contribution
    Resources, Supervision, Funding acquisition, Methodology, Writing - review and editing
    Competing interests
    Reviewing editor, eLife
  16. Cynthia L Andoniadou

    1. Centre for Craniofacial and Regenerative Biology, King’s College London, London, United Kingdom
    2. Department of Medicine III, University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany
    Contribution
    Conceptualization, Supervision, Funding acquisition, Investigation, Methodology, Writing - original draft, Writing - review and editing
    For correspondence
    cynthia.andoniadou@kcl.ac.uk
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-4311-5855

Funding

Medical Research Council (MR/L016729/1)

  • Cynthia Lilian Andoniadou

Medical Research Council (MR/T012153/1)

  • Cynthia Lilian Andoniadou

Deutsche Forschungsgemeinschaft (314061271 - TRR 205)

  • Cynthia Lilian Andoniadou

Howard Hughes Medical Institute

  • Roel Nusse

Agence Nationale de la Recherche (ANR-18-CE14-0017)

  • Patrice Mollard

Fondation pour la Recherche Médicale (DEQ20150331732)

  • Patrice Mollard

Lister Institute of Preventive Medicine

  • Cynthia Lilian Andoniadou

Dianna Trebble Endowment Fund

  • John P Russell

King’s Bioscience Institute and the Guy’s and St Thomas’ Charity Prize

  • Emily J Lodge

British Society for Neuroendocrinology

  • Yasmine Kemkem

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This study has been supported by the Medical Research Council (MR/L016729/1, MR/T012153/1) (CLA), The Lister Institute of Preventive Medicine (CLA), the Deutsche Forschungsgemeinschaft (DFG German Research Foundation) (Project Number 314061271 – TRR 205) (CLA), the Howard Hughes Medical Institute (RN), the Agence Nationale de la Recherche (ANR-18-CE14-0017), and Fondation pour la Recherche Médicale (DEQ20150331732) (PM). JPR was supported by a Dianna Trebble Endowment Fund Dental Institute Studentship, EJL by the King’s Bioscience Institute and the Guy’s and St Thomas' Charity Prize PhD Programme in Biomedical and Translational Science, and YK by a Project Support Grant from the British Society for Neuroendocrinology. We thank Dr AF Parlow and the National Hormone and Peptide Program (Harbor–University of California, Los Angeles Medical Center) for providing some of the antibodies used in this study and Prof. J Drouin and Prof. S Rhodes for TPIT and PIT1 antibodies respectively. We thank the High-Throughput Genomics Group at the Wellcome Trust Centre for Human Genetics (funded by Wellcome Trust grant reference 090532/Z/09/Z) for the generation of the sequencing data. For flow sorting and analysis, this research was supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London. We thank Marie Isabelle Garcia, Juan Pedro Martinez-Barbera, and Paul Le Tissier for useful discussions and critical comments on the manuscript.

Ethics

Animal experimentation: This study was performed under compliance of the Animals (Scientific Procedures) Act 1986, Home Office License (P5F0A1579) and KCL Biological Safety approval for project 'Function and Regulation of Pituitary Stem Cells in Mammals'.

Senior and Reviewing Editor

  1. Marianne E Bronner, California Institute of Technology, United States

Reviewers

  1. Elaine Emmerson, The University of Edinburgh, United Kingdom
  2. Sally Camper, University of Michigan Medical School, United States
  3. Lori T Raetzman, University of Illinois at Urbana-Champaign, United States

Publication history

  1. Received: May 20, 2020
  2. Accepted: January 4, 2021
  3. Accepted Manuscript published: January 5, 2021 (version 1)
  4. Version of Record published: January 12, 2021 (version 2)

Copyright

© 2021, Russell et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 480
    Page views
  • 85
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Neuroscience
    2. Stem Cells and Regenerative Medicine
    Laura J Wagstaff et al.
    Research Article

    After nerve injury, myelin and Remak Schwann cells reprogram to repair cells specialized for regeneration. Normally providing strong regenerative support, these cells fail in aging animals, and during chronic denervation that results from slow axon growth. This impairs axonal regeneration and causes significant clinical problems. In mice, we find that repair cells express reduced c-Jun protein as regenerative support provided by these cells declines during aging and chronic denervation. In both cases, genetically restoring Schwann cell c-Jun levels restores regeneration to control levels. We identify potential gene candidates mediating this effect and implicate Shh in the control of Schwann cell c-Jun levels. This establishes that a common mechanism, reduced c-Jun in Schwann cells, regulates success and failure of nerve repair both during aging and chronic denervation. This provides a molecular framework for addressing important clinical problems, suggesting molecular pathways that can be targeted to promote repair in the PNS.

    1. Cancer Biology
    2. Stem Cells and Regenerative Medicine
    Madeline Sandoval et al.
    Research Article Updated

    Skin epithelium can accumulate a high burden of oncogenic mutations without morphological or functional consequences. To investigate the mechanism of oncogenic tolerance, we induced HrasG12V in single murine epidermal cells and followed them long term. We observed that HrasG12V promotes an early and transient clonal expansion driven by increased progenitor renewal that is replaced with an increase in progenitor differentiation leading to reduced growth. We attribute this dynamic effect to emergence of two populations within oncogenic clones: renewing progenitors along the edge and differentiating ones within the central core. As clone expansion is accompanied by progressive enlargement of the core and diminishment of the edge compartment, the intraclonal competition between the two populations results in stabilized oncogenic growth. To identify the molecular mechanism of HrasG12V-driven differentiation, we screened known Ras-effector in vivo and identified Rassf5 as a novel regulator of progenitor fate choice that is necessary and sufficient for oncogene-specific differentiation.