Regenerative Capacity of the Adult Pituitary: Multiple Mechanisms of Lactotrope Restoration After Transgenic Ablation

Abstract

In a recent study, we showed that the adult pituitary gland is capable of regenerating transgenically ablated growth hormone-producing (GH⁺) somatotropes. Here, we investigated whether the gland's regenerative capacity is more general and also applies to the other major hormonal cell type, the prolactin-producing (PRL⁺) lactotropes. We set up the transgenic PRLCre/inducible diphtheria toxin receptor (iDTR) mouse model, in which the PRL promoter drives expression of Cre that induces DTR in lactotropes. Injection of female mice with DT for different periods causes a gradual ablation of PRL⁺ cells, reaching a maximum of 70% after 10-day DT treatment. During the following weeks, lactotropes progressively reappear achieving a 60% restoration after 6 weeks. The Sox2⁺ stem/progenitor cell compartment displays a prompt reaction to the DT-triggered cell ablation injury, including expansion of the marginal-zone niche and coexpression of PRL, the latter only very rarely observed in control pituitary. Throughout the regeneration period (2–6 weeks), Sox2⁺ as well as double Sox2⁺/PRL⁺ cells continue to be more abundant than in control pituitary. In addition to this stem cell reaction, surviving or newborn lactotropes increase their proliferative activity, and bihormonal PRL⁺/GH⁺ cells become detectable suggesting somatotrope-to-lactotrope transdifferentiation. In conclusion, the adult pituitary gland is capable of restoring lactotrope cells after destruction, further confirming its regenerative competence. Repair of lactotropes appears to be driven by a combination of mechanisms, including recruitment from stem cells, proliferation of lactotropes, and transdifferentiation of somatotropes.

Introduction

A s the master endocrine gland, the pituitary plays a crucial role in regulating vital processes, such as growth, metabolism, puberty, and reproduction, mainly through hormones secreted by dedicated cells in the anterior pituitary (AP) lobe, that is, somatotropes that produce growth hormone (GH); lactotropes that secrete prolactin (PRL); thyrotropes that release thyroid-stimulating hormone (TSH); corticotropes that generate adrenocorticotropic hormone (ACTH); and gonadotropes that make luteinizing hormone (LH) and follicle-stimulating hormone (FSH) [1]. Given this central role, dysfunction of the gland can cause severe morbidity through endocrine deficits. In our studies, we address the question whether the gland is capable of repairing damage when occurring at adult age. Only recently, stem/progenitor cells were identified in the pituitary based on efflux capacity [2,3], stemness marker expression, and multipotent differentiation potential [3 –6]. In particular, Sox2, a transcriptional regulator in neural and other stem cells [7], labels cells in putative stem cell niches of the pituitary, including the marginal zone around the cleft, a remnant of embryonic development [1,3 –6].

In many adult organs, stem cells participate in the repair process after damage [8 –11]. In a recent study, we demonstrated that the pituitary stem cells promptly react to injury inflicted at adult age through transgenic ablation of somatotropes, and that the destroyed GH⁺ cells are substantially replaced, most plausibly from the activated stem cells [12]. Thus, the mature pituitary is capable of regeneration, an issue unsettled and controversial until this study [11,13,14]. Here, we investigated whether the pituitary's regenerative capacity is not only limited to somatotropes, and used a comparable transgenic model in which the other main hormonal cell type, the PRL-producing lactotrope, is destroyed using the “diphtheria toxin (DT)/inducible DT-receptor” (iDTR) system [12,15,16]. We found that also lactotropes are regenerated, but that in contrast to the somatotrope restoration, not only stem cells, but also other mechanisms appear to be involved.

Materials and Methods

Transgenic mice

PRLCre mice express Cre recombinase under control of the rat 3.2-kb PRL promoter, and were kindly provided by Dr. LeTissier (Division of Molecular Neuroendocrinology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, UK) [17]. For genotyping, genomic DNA was subjected to PCR with 5′-TGACGGAAATAGATGATTGG-3′ as forward primer, and 5′-ACGCGTTGAAAGATGGCCTTTC-3′ (rat) and 5′-TTCTGGTCTTGAGTAGTCAC-3′ (mouse) as backward primers, yielding a transgene-specific amplicon of 1,480 bp and an endogenous (mouse) amplicon of 1,600 bp. iDTR mice harbor the DTR gene preceded by a STOP cassette, which is flanked by LoxP sites. Cre recombinase binds to these sites and excises the STOP sequence, thereby rendering the Cre-, thus the DTR-expressing cells susceptible to ablation by DT through apoptosis [12,15,16]. Homozygous iDTR/iDTR mice were crossed to hemizygous PRLCre/- animals, yielding PRLCre/iDTR mice (hemizygous for both transgenes) and −/iDTR littermates (controls).

Mice were bred on the original C57Bl6/CBA genetic background and kept in the Animal Housing Facility of the University (KU Leuven, Belgium), providing constant temperature, humidity, day/night cycle, and water and food ad libitum. Animal experiments were approved by the KU Leuven Ethics Committee.

Treatment with DT

Adult (8- to 12-week-old) PRLCre/iDTR and −/iDTR mice were intraperitoneally (i.p.) injected with DT (4 ng/g body weight, twice a day; Sigma-Aldrich) or vehicle (PBS; Invitrogen) for the time periods mentioned (see Figs. 1A and 2A). Toxicity was not observed as assessed by overall health, including body weight, hair, posture, walk, and explorative behavior [12]. For independent experiments, 3–6 mice were grouped per condition. Since highest ablation was obtained in female mice (see Results), all experiments (except one mentioned at the beginning of the Results) were performed with this gender (random cycle). Control experiments were done as described before [12]; neither DTR nor Cre expression per se, nor DT treatment in the absence of DTR expression affected the pituitary architecture and the proportions of the different hormonal cell populations (data not shown).

FIG. 1.

Gradual ablation of lactotropes in DT-treated PRLCre/iDTR mice. (A) Time schedule of DT treatment and of pituitary analysis. (B) Vibratome sections (upper row) and dispersed cells (lower row) from APs of control (−/iDTR) and PRLCre/iDTR mice after different periods of DT treatment, immunostained for PRL (red). Representative pictures are shown. Scale bar, 20 μm for sections, 50 μm for dispersed cells. (C) Percent change in absolute PRL⁺ cell number in PRLCre/iDTR versus control −/iDTR mice as determined after different periods of DT injection using dispersed AP cells. Data are presented as mean±SEM (d4, n=3; d8, n=3; d11, n=7). *P<0.05 and **P<0.01 versus control mice, respectively. (D) Apoptosis within the PRL⁺ cell population in −/iDTR control and PRLCre/iDTR mice as analyzed after different periods of DT administration using dispersed AP cells. Data are shown as mean±SEM (n=3 for all conditions). **P<0.01 versus control mice. AP, anterior pituitary; iDTR, inducible diphtheria toxin receptor; PRL, prolactin. Color images available online at www.liebertpub.com/scd

FIG. 2.

Regeneration of lactotropes after DT-mediated ablation. (A) Time schedule of DT treatment and of pituitary analysis. (B) Vibratome sections (upper row) and dispersed cells (lower row) from APs of PRLCre/iDTR mice immunostained for PRL (red) the day after 10-day DT treatment and 2, 4, and 6 weeks later. Representative pictures are shown. Scale bar, 20 μm for sections, 50 μm for dispersed cells. (C) Percent difference in absolute PRL⁺ cell number in PRLCre/iDTR versus control (−/iDTR) mice, as analyzed at different time points after 10-day DT treatment using dispersed AP cells. Data are presented as mean±SEM (d11, n=6; 2 weeks, n=3; 4 weeks, n=3; 6 weeks, n=4). **P<0.01 versus analysis at d11. Color images available online at www.liebertpub.com/scd

Immunofluorescence, AMCA labeling, and TUNEL analysis of dissociated AP cells

Mice were euthanized with CO₂, and the anterior pituitary (AP) lobe carefully isolated and dissociated into single cells using trypsin as described before [2,3,12]. Cells were resuspended in a serum-free pituitary-optimized cell culture medium (Invitrogen) [2,3,12] and spun down on cytospin slides (Superfrost Plus; Thermo Scientific). Cells were fixed with paraformaldehyde (4% in PBS), permeabilized with saponin (Sigma-Aldrich; 0.5% in PBS), and incubated overnight with the following antibodies: rabbit anti-mouse GH (1/10,000), guinea pig anti-rat GH (1/5,000), rabbit anti-rat PRL (1/10,000), guinea pig anti-rat PRL (1/2,500), rabbit anti-rat ACTH (1/5,000), guinea pig anti-rat αGSU (1/1,000) (all obtained from Dr. A. F. Parlow, NHPP, Harbor-UCLA Medical Center, Torrance, CA), goat anti-human Sox2 (1/250–1/750; Immune Systems), and rabbit anti-Ki67 (1/50; Thermo Scientific). Subsequently, cells were incubated during 1 h with Alexa Fluor 488-, 555- (Invitrogen), and/or FITC- (Jackson Immunoresearch Europe) labeled secondary antibodies (at 1/1,000), and then 4′, 6-diamidino-2-phenylindole (DAPI; 0.5 μg/mL; Sigma-Aldrich). Images were captured using an Olympus IX81 microscope equipped with Cell^M software (Olympus). Cells were counted using ImageJ (http://rsb.info.nih.gov/ij/) in at least 5 random cytospin fields (1,000–2,000 cells per antigen per group in each experiment). The proportions of immunopositive cells on total AP cells were determined. Reliability of quantification using dispersed AP cells was validated by comparing hormonal cell proportions in cytospin cell samples (see above) and in pituitary sections (paraffin-embedded and immunostained as described before, see [12]; and counted in 5 slices spaced at regular intervals). No significant differences were observed (Supplementary Fig. S1A; Supplementary Data are available online at www.liebertpub.com/scd). Moreover, proportions were comparable to data reported in other studies [18 –20].

Because of the reduction of the total cells within the AP when lactotropes are destroyed, unaffected hormonal cell types logically increase in proportion. Therefore, proportions do not correctly reflect the real abundance of the cell types, and absolute cell numbers must be determined for appropriate interpretation and comparison. The absolute cell number of each hormonal cell population was calculated by multiplying its proportion by the total number of AP cells (as obtained after AP dispersion).

Figures shown were prepared with ImageJ and assembled with Microsoft PowerPoint (2007). Negative controls for immunostaining (without primary antibodies) did not show any signal [2,3,12; and data not shown].

Folliculostellate (FS) cells were identified by the uptake of the fluorescent dipeptide β-Ala-Lys-Nɛ-7-amino-4-methylcoumarin-3-acetyl (further referred to as AMCA; Biotrend) as described before [12]. In the AP, uptake of AMCA is a specific property of FS cells [21]. Briefly, dispersed AP cells were incubated with AMCA (40 μM) for 2 h at 37°C, and then fixed and immunostained for Sox2 and Ki67 as described above.

Apoptosis was assessed by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) analysis in cytospin preparations of fixed AP cells using the FragEL kit (Calbiochem) according to the manufacturer's protocol [12].

Immunofluorescence of pituitary vibratome sections

Coronal pituitary vibratome sections were obtained and immunostained as described in detail before [2,3,12,22]. Briefly, pituitaries were isolated, fixed in 4% paraformaldehyde, embedded in 2% agarose (MP Biomedicals), and coronally sectioned to 45-μm slices using a vibratome (Microm HM 650V; Prosan). After permeabilisation with 0.4% Triton X100, sections were incubated with antibodies as described above for cells, followed by nuclear labeling with ToPro-3 (0.01 mM; Invitrogen) and scanning using a confocal laser scanning microscope (Zeiss LSM 510; Zeiss). Images were prepared with Zeiss LSM Image Brower and Microsoft PowerPoint (2007).

Statistical analysis

One-way analysis of variance (ANOVA) was applied (SPSS, Version 13.0; IBM). For comparison of 3 groups, ANOVA was followed by Least Significant Difference Post Hoc Multiple Comparison. Statistical significance was defined as P<0.05.

Results

Gradual ablation of lactotropes in DT-treated PRLCre/iDTR mice

Adult PRLCre/iDTR and control (−/iDTR) mice were injected with DT for different periods (3, 7, or 10 days), and pituitaries were analyzed 1 day later (d4, d8, or d11) (Fig. 1A). Lactotrope ablation was not very efficient in male mice (35% reduction of absolute PRL⁺ cell number after 10-day DT treatment; n=5). Because we aimed at the maximal cell obliteration to study the regenerative capacity of the gland, all further experiments were performed with female mice.

PRL-immunocreactive (PRL⁺) cells gradually decline in PRLCre/iDTR pituitaries, the ablation becoming more pronounced with a longer DT treatment period (Fig. 1B). Some of the surviving lactotropes become hypertrophic, and some are connected and organized in clusters (data not shown). To determine the ablation grade, PRL-immunoreactive (PRL⁺) cells were counted after AP cell dispersion, and absolute cell numbers were calculated (see Materials and Methods) to take into account the reduced total AP cell number as observed from d8 (∼30% reduction, corresponding to a decline in pituitary weight; see Supplementary Fig. S1B, C). PRL⁺ cell ablation reaches ∼70% after 10-day DT treatment (Fig. 1C), which does not further increase (data not shown). This ablation grade is concordant with the reported activity of the PRLCre transgene in 70%–80% of the lactotropes in female mice [17]. In analogy, apoptotic PRL⁺ cells are highest between d4–d8 and return to basal (control) levels at d11 (Fig. 1D), indicating that all susceptible lactotropes are killed by that time. We further examined whether the other hormonal cell lineages are influenced. ACTH-immunopositive (ACTH⁺) corticotropes and α-glycoprotein subunit-immunoreactive (αGSU⁺) thyrotropes, gonadotropes, and αGSU-only cells are not affected (Supplementary Fig. S1D–F). GH⁺ somatotropes, however, slightly, but significantly decline in absolute number (∼20% at d11; Supplementary Fig. S1D–F). Possible explanations for this decay are given in the Discussion. There were no signs of inflammation (data not shown), neither infiltration of mononuclear immune cells (lymphocytes, monocytes, macrophages), nor of polymorphonuclear immune cells (neutrophils, basophils, eosinophils). Absence of inflammatory infiltrate is in accordance with our previous study [12] and other studies using DT-mediated cell ablation through apoptosis [15, 23 –25].

Taken together, we set up a transgenic mouse model to conditionally obliterate lactotropes, which allowed us to study whether and how the pituitary gland repairs this injury and restores the destroyed cells.

Gradual regeneration of lactotropes after DT-induced ablation in PRLCre/iDTR mice

Female mice were injected with DT for 10 days to obtain maximal lactotrope ablation, and pituitaries were analyzed at d11 as well as 2, 4, and 6 weeks later (Fig. 2A). From in situ examination of pituitary sections, it is clear that PRL⁺ cells gradually reappear (Fig. 2B, upper row), which is confirmed by cell counting; the ∼70% decline in absolute PRL⁺ cell number in PRLCre/iDTR mice (as compared to control mice) progressively decreases to ∼30% after 6 weeks (Fig. 2C). Further regeneration was not observed (as analyzed after 12–16 weeks; data not shown). Of note, the absolute GH⁺ cell number is restored to control levels after 4 weeks (Supplementary Fig. S2A–C), while ACTH⁺ and αSU⁺ cells do not change (Supplementary Fig. S2B, C and data not shown).

Taken together, the pituitary gland is capable of regenerating lactotropes to a considerable extent within weeks, as well as recovering the slightly declined somatotropes.

Multiple mechanisms of lactotrope regeneration

Lactotropes may be regenerated after destruction through several mechanisms, including recruitment from stem cells, proliferation of surviving or newborn PRL⁺ cells, and/or transdifferentiation from somatotropes. In our previous study, we advanced development from stem cells as a major mechanism in the restoration of GH⁺ cells after somatotrope ablation [12].

Involvement of stem cells in lactotrope regeneration

The Sox2⁺ pituitary stem cells promptly react to the DT-provoked lesion in the PRLCre/iDTR mice. At some locations, in particular, the marginal-zone niche and the wedge region (where the anterior and intermediate lobes meet; see [12,22]) Sox2⁺ cells are clearly expanded at d8–d11 (Fig. 3A, B). At some sites, a multilayer Sox2⁺ marginal zone is observed, whereas a unicellular layer is generally seen in control mice (Fig. 3A). The absolute number of Sox2⁺ cells within the total AP cells is increased by ∼2-fold (Fig. 3C). As observed before [3,12,22], the Sox2⁺ cell compartment does not only include cells with Sox2 in the nucleus, but also some cells with Sox2-immunoreactive signal in the cytoplasm, proposed to represent cells in which the nuclear action of Sox2 is stopped, thereby allowing differentiation [3,12,22]. Exclusion of Sox2 from the nucleus by, for instance, nucleocytoplasmic shuttling, has also been shown to permit differentiation in other stem cells [26]. These Sox2-cytoplasmic cells are more frequently encountered—mostly grouped—, in the anterior lobe's parenchyma of PRLCre/iDTR mice, particularly at the end of the peak period of lactotrope ablation (d8; Fig. 3A), which is also confirmed by counting (Fig. 3C). The expansion of the Sox2⁺ stem/progenitor cells is, at least partially, explained by increased proliferative activity at d4 and d8 (as assessed by Ki67 immunostaining), returning to the basal control level of practically not observable dividing Sox2⁺ cells at d11 (Fig. 3D, E). Also the FS cells, previously proposed as pituitary stem/progenitor cells and significantly overlapping with the Sox2⁺ stem/progenitor cell compartment [3 –6,11,12], increase in number (as visualized by AMCA uptake), although to a lower extent than the Sox2⁺ cells (Fig. 3D and Supplementary Fig. S3A). In analogy, proliferative activity of the AMCA⁺ cells increases, but not as prominently as in the Sox2⁺ cell compartment (and not statistically significant) (Fig. 3D and Supplementary Fig. S3B). Of note, the proportion of double Sox2⁺/AMCA⁺ cells within the AMCA⁺ cell population shows a tendency (P>0.05) of increase from ∼80% (control) to ∼90% at d4 and d8, and a subsequent decrease at d11 (Supplementary Fig. S3C). Together, the AMCA⁺ cells may also contribute to the enlargement of the Sox2⁺ cell compartment, by first expanding (either as AMCA⁺/Sox2⁺, or as AMCA⁺ cells that then start to express Sox2), and subsequently loosing the AMCA uptake ability.

FIG. 3.

Acute response of the Sox2⁺ stem/progenitor cell compartment to DT-induced lactotrope ablation. (A) Sox2 immunofluorescence (red) in vibratome sections (upper rows) and dispersed cells (lower row) from pituitaries of control (−/iDTR) and PRLCre/iDTR mice as analyzed the day after 3, 7, and 10 days of DT administration. Representative pictures of different regions in sections (marginal zone; wedge; AP lobe parenchyma) are shown. Sox2-cytoplasmic cells increase at peak of injury (arrow). Scale bar, 20 μm for sections, 50 μm for dispersed cells. IL, intermediate lobe. (B) Fold change (relative to −/iDTR control mice) of Sox2⁺ cell number in the AP's marginal-zone and wedge region of PRLCre/iDTR mice as analyzed after different periods of DT injection (d4-d11) using immunostained sections. Bars represent mean±SEM (n=5 for all conditions). *P<0.05 versus control mice. (C) Absolute number of Sox2⁺ cells within the total AP cells of −/iDTR control mice (white bar) and PRLCre/iDTR mice (black bar) as determined after different periods of DT injection using dipsersed AP cells. Bars represent mean±SEM (n=3 for all conditions), with indication of the Sox2-nuclear cells (white and black part, respectively), and Sox-cytoplasmic cells (gray part). **P<0.01 versus control mice (for total Sox2⁺ cell number). (D) Sox2 (red) and Ki67 (green) immunostaining with AMCA uptake (blue) in dispersed AP cells of −/iDTR control and PRLCre/iDTR mice the day after 3 and 7 days of DT injection. Representative pictures are shown. The PRLCre/iDTR pituitary contains dividing Sox2⁺ cells (filled arrow), proliferating AMCA⁺ cells (empty arrow), and cycling Sox2⁺/AMCA⁺ cells (arrowhead). Large part of the cells are double Sox2⁺/AMCA⁺, although there are also cells containing only Sox2 or AMCA. The lower row shows the light-microscopic pictures visualizing all cells. Scale bar, 50 μm. (E) Proportion of dividing (Ki67⁺) Sox2⁺ cells within the proliferating population in the AP of DT-treated −/iDTR control mice (no bar; value=0 at all time points analyzed) and PRLCre/iDTR mice as determined after different periods of DT treatment using dispersed AP cells. Bars represent mean±SEM (n=5 for all conditions). *P<0.05 versus control mice. (F) Vibratome sections of the AP lobe from −/iDTR control and PRLCre/iDTR mice immunostained for Sox2 (red) and PRL (green) as analyzed after different periods of DT injection. Representative pictures are shown. Some Sox2-cytoplasmic cells coexpress PRL, more readily detected in the PRLCre/iDTR mice (arrows; separate colors and merged are shown; blue, ToPro-3 DNA stain) than in the control mice (merged colors shown). Scale bar, 20 μm. AMCA, β-Ala-Lys-Nɛ-7-amino-4-methylcoumarin-3-acetyl. Color images available online at www.liebertpub.com/scd

To investigate whether the activated Sox2⁺ stem/progenitor cell compartment is involved in lactotrope regeneration, double Sox2/PRL immunostaining was performed. During the acute reaction (d4–d11), PRL is found in Sox2-cytoplasmic cells in the PRLCre/iDTR mice (Fig. 3F). Although at low frequency, these cells are more readily encountered (1.6%±0.6% of the Sox2⁺ cells; mean±SEM of n=4) than in the control mice (<0.1%). GH was virtually not found together with Sox2 (Supplementary Fig. S4A). In agreement with our previous studies, some sporadic Sox2-cytoplasmic cells contain ACTH or αGSU, however, not different from control pituitary (Supplementary Fig. S4A), considered to represent normal turnover events in the slowly renewing adult gland [1,3,12,22].

During the regeneration period (2 to 6 weeks after the 10-day DT injection), Sox2⁺ cells remain more abundant in the PRLCre/iDTR than the control pituitary (Fig. 4A–C). Sox2-cytoplasmic cells are most numerous after 2 weeks, and then return to control levels (Fig. 4C). This picture may mirror the restorative attempt at d4–d8, but in contrast, this response can now develop because DT has been cleared [12,15,23 –25] and nascent lactotropes are no longer killed (in contrast to d4–d8). During the regeneration period, PRL remains to be found together with Sox2 more easily in the PRLCre/iDTR than the control mice, although their occurrence remains low (Fig. 4D; 2.1%±0.9% of the Sox2⁺ cells at 2 weeks after injection versus <0.1% in control mice; n=4). GH remains virtually absent from Sox2⁺ cells, whereas ACTH and αGSU are present in some sporadic Sox2-cytoplasmic cells, not different from control mice (Supplementary Fig. S4B). Both in control and PRLCre/iDTR mice, ACTH (not the other hormones) is sometimes found in cells that have Sox2 in the nucleus (Supplementary Fig. S4B, arrowheads), as also reported before [22]. At present, we do not know the meaning of this observation.

FIG. 4.

Phenotype of the Sox2⁺ stem/progenitor cell compartment during the regeneration process. (A) Sox2 immunofluorescence (red) in vibratome sections (upper rows) and dispersed cells (lower row) from pituitaries of control (−/iDTR) and PRLCre/iDTR mice as analyzed 2, 4, and 6 weeks after 10-day DT injection (see Fig. 2A). Representative pictures of different regions in sections (marginal zone; wedge; AP lobe parenchyma) are shown. Scale bar, 20 μm for sections, 50 μm for dispersed cells. (B) Fold change (relative to −/iDTR control mice) of Sox2⁺ cell number in the AP's marginal-zone and wedge region of PRLCre/iDTR mice as analyzed 2, 4, and 6 weeks after 10-day DT injection using immunostained sections. Bars represent mean±SEM (n=5 for all conditions). *P<0.05 versus control mice. (C) Absolute number of Sox2⁺ cells within the AP of −/iDTR control mice (white bar) and PRLCre/iDTR mice (black bar) as analyzed 2, 4, and 6 weeks after 10-day DT injection using dispersed AP cells. Bars represent mean±SEM (n=3 for all conditions), with indication of the Sox2-nuclear cells (white and black part, respectively), and Sox-cytoplasmic cells (gray part). *P<0.05 and **P<0.01 versus control mice, respectively (for total Sox2⁺ cell number). (D) A representative vibratome section of the AP lobe from a PRLCre/iDTR mouse immunostained for Sox2 (red) and PRL (green) as analyzed 4 weeks after 10-day DT injection. Some Sox2-cytoplasmic cells coexpress PRL, sporadically, but still more readily detected in the PRLCre/iDTR mice (arrow; separate colors and merged are shown; blue, ToPro-3 DNA stain) than in the control mice (not shown). Scale bar, 20 μm. Color images available online at www.liebertpub.com/scd

Taken together, these data indicate that stem/progenitor cells promptly react to the lactotrope-ablation injury by expanding, and suggest that they are involved in the regeneration process, at least partially, by initiating PRL expression.

Involvement of lactotrope proliferation in the regeneration process

We further examined the proliferative activity of surviving (or newborn) PRL⁺ cells as a possible mechanism of lactotrope regeneration. During the acute period (d4–d11), proliferating (Ki67⁺) lactotropes increase 4- to 7-fold when compared to control mice (Fig. 5A, B). During the further regeneration period (2–6 weeks), even a higher proportion (2%–4%) of the PRL⁺ cells is dividing in the PRLCre/iDTR mice (Fig. 5A, B), which is likely explained by the fact that proliferating PRL⁺ cells (surviving or newborn) are not killed anymore after clearance of DT from the system.

FIG. 5.

Lactotrope proliferation in DT-treated PRLCre/iDTR mice. (A) PRL (green) and Ki67 (red) immunostaining of dispersed AP cells from −/iDTR control and PRLCre/iDTR mice acutely after different periods of DT injection (d4-d11; upper row) and during the regeneration period (i.e., 2, 4, and 6 weeks after 10-day treatment; lower row). Representative pictures are shown. The PRLCre/iDTR pituitary contains more dividing PRL⁺ cells (arrows) than the control pituitary. Scale bar, 50 μm. (B) Proportion of dividing (Ki67⁺) PRL⁺ cells within the lactotrope cell population in the AP of DT-treated −/iDTR control and PRLCre/iDTR mice as determined acutely after different periods of DT injection (d4-d11) and during the regeneration period (i.e., 2, 4, and 6 weeks after 10-day treatment) using dispersed AP cells. Bars represent mean±SEM (n=3 for all conditions). *P<0.05 and **P<0.01 versus control mice, respectively. Color images available online at www.liebertpub.com/scd

Thus, the pituitary acutely attempts to compensate the loss of lactotropes, at least, in part, through proliferation of surviving PRL⁺ cells. During the further regeneration period, lactotrope restoration also appears partly driven by proliferation of (surviving or newborn) PRL⁺ cells.

Involvement of transdifferentiation in lactotrope regeneration

Transdifferentiation between somatotropes and lactotropes, reported to take place under certain circumstances (see eg, [27,28]), may also be involved in the regeneration process. Therefore, we investigated the presence of transitory bihormonal lactosomatotropes. At all time points analyzed (d4 till 6 weeks), double GH⁺/PRL⁺ cells are more abundant in the pituitary of DT-treated PRLCre/iDTR mice when compared to control mice (Fig. 6). Thus, transdifferentiation may also, at least, in part, be responsible for the lactotrope restoration process. Proliferating GH⁺ cells, not detected in control mice, but in PRLCre/iDTR mice at 2, 4, and 6 weeks after DT (although rare and estimated 0.01%–0.05% of the total AP cells; data not shown), may provide the source of the transdifferentiating cells while keeping the number of somatotropes stable.

FIG. 6.

Analysis of double GH⁺/PRL⁺ cells (lactosomatotropes) in DT-treated PRLCre/iDTR mice. (A) Vibratome sections (upper row) and dispersed cells (lower row) from APs of −/iDTR control and PRLCre/iDTR mice, immunostained for PRL (green) and GH (red) acutely after different periods of DT injection. Representative pictures are shown. The PRLCre/iDTR pituitary contains more lactosomatotropes (arrows) than the control pituitary. Scale bar, 20 μm for sections, 50 μm for dispersed cells. (B) PRL (green) and GH (red) immunostaining of dispersed AP cells from −/iDTR control and PRLCre/iDTR mice during the regeneration period (i.e., 2, 4, and 6 weeks after 10-day treatment). Representative pictures are shown. The PRLCre/iDTR pituitary contains more lactosomatotropes (arrows) than the control pituitary. Scale bar, 50 μm. (C) Absolute number of double GH⁺/PRL⁺ cells within the AP of −/iDTR control mice (white bar) and PRLCre/iDTR mice (black bar) as analyzed acutely during DT injection (d4) and during the regeneration period (4 weeks after 10-day treatment) using dipsersed AP cells. Bars represent mean±SEM (n=3 for all conditions). *P<0.05 versus control mice. GH, growth hormone. Color images available online at www.liebertpub.com/scd

In conclusion, our data show that lactotropes are regenerated after transgenic ablation and support the involvement of multiple processes, including recruitment from stem cells, from proliferating lactotropes, and from transdifferentiating somatotropes. Lineage-tracing studies are required for further proof, but the necessary transgenic models are at present not available (see Discussion).

Discussion

The adult pituitary gland can restore somatotropes after transgenic ablation [12]. Here, we show that also lactotropes can be substantially renewed after destruction, thereby supporting a more general regenerative capacity of the mature gland. In both pituitary injury models, stem cells promptly react by expanding (most obvious at the marginal-zone and wedge niches) and by coexpressing the ablated hormone, thus supporting their differentiation toward the affected cell type. However, there are also noticeable differences between the somatotrope (see [12]) and lactotrope regeneration process (as described here). Whereas surviving somatotropes do not proliferate in response to the cell destruction within their population, lactotropes clearly react by increasing their proliferative activity. Along the same line, lactosomatotropes, generally considered as transitory stages during transdifferentiation, are virtually absent in the somatotrope-ablated pituitary, but clearly present in response to lactotrope obliteration. Thus, mechanisms of repair appear to depend on the cell type that is affected. Indeed, whereas scarce for somatotropes, there are several indications that lactotropes represent a plastic, dynamic cell population in the pituitary [28 –30]. The dissimilar plasticity of both cell types, and the inherent link to either a single or multiple mechanisms involved, may further explain the different pace of regeneration, being weeks for lactotropes and months for somatotropes, as well as the final restoration grade, being more efficient for lactotropes. On the other hand, the final ablation of somatotropes was higher (90%; see [12]) compared with lactotropes here (70%), which may also explain the superior restoration of the lactotropes. Moreover, the different degrees of ablation may also explain the different regeneration strategies employed. In pancreas, it has been shown that only extensive ablation activates the resident stem cells for regeneration, whereas lower degrees of injury can be repaired by self-replication of β-cells [25,31].

Also, within the stem cell response, differences are noted between both models. The expansive reaction is faster and depicts as more prominent in the somatotrope ablation model (already after 1–2 days of DT injection; see [12]), which may be accounted for by the swifter and more vigorous ablation process (injury) than in the PRLCre/iDTR model described here. In addition, peculiar small Sox2-cytoplasmic cells immediately start to coexpress GH after somatotrope ablation (from d3; [12]), whereas in the PRLCre/iDTR model here, such small cells are not observed and the coexpression (in larger Sox2-cytoplamic cells) starts later (d4). The meaning of these observations requires further investigation.

Further support and direct demonstration of development of new lactotropes through the different mechanisms proposed requires multiple lineage-tracing studies, in which, for instance, Sox2⁺ stem cells or GH⁺ somatotropes are marked with a label that remains present in the progeny. However, the Cre-driven ablation in our model precludes the use of a Cre-based approach for lineage tracing (such as Cre-induced reporter activation regulated by tamoxifen). At present, no appropriate (pituitary) models using other tactics (such as Flp/Frt) are available. In addition, we can so far not exclude that (partial) dedifferentiation of surviving and/or DT-affected lactotropes or of somatotropes plays a role; these cells may regain a stem cell phenotype (with or without reinitiation of Sox2 expression), and subsequently restart lactotrope differentiation. Finally, proliferating PRL⁺ cells may represent surviving lactotropes, but also newborn PRL⁺ cells that just recently developed from Sox2⁺ stem cells. Again, complex and compound transgenic models are needed, but not available (e.g., strictly time-controlled and inducible PRL, GH, and/or Sox2 lineage-tracing systems, not using Cre/LoxP).

Full recovery was not attained, neither after somatotrope [12], nor after lactotrope ablation, even following longer lag periods. Possibly, no full restoration is needed to attain satisfactory hormone functions, or the mature organ, in contrast to the embryonic tissue, is not competent anymore for full restoration, as also found in other organs (see eg, [32 –34]). In analogy, also other mature organs (e.g., thyroid gland, heart) only rarely regenerate to a full extent [8,35,36].

In addition to the transgenically targeted lactotropes, somatotropes also decline in the PRLCre/iDTR model. Because less than 1% of GH⁺ cells show transgenic promoter activity in the PRLCre mouse [17]—also encompassing the GH⁺ cells that have expressed PRL at some time point during their development—, it is unlikely that the 20% somatotrope reduction is due to a direct impact of DT through DTR expressed in these cells. The acute decrease in somatotropes may be explained by their transdifferentiation to complement the loss in the affected lactotrope population, which, however, does not succeed because the newborn lactotropes are killed as long as DT is present [12,15,23 –25]. Alternatively, the dying lactotropes may liberate factors inducing death of neighboring cells, most of which are somatotropes. Finally, we cannot exclude that loss of trophic signals from the disappearing lactotropes may play a role [37]. However, this paracrine impact on the somatotropes would then occur at a very fast pace. Later, during the regeneration period (4–6 weeks), somatotropes are restored and further kept stable despite the putative transdifferentiation to lactotropes. Self-duplication may be one underlying mechanism, since we observed proliferating somatotropes in the PRLCre/iDTR, but not the control pituitary. Transdifferentiation from the surviving and/or newly formed lactotropes may represent another possibility. Thus, somatotropes recur here in a different manner than after their more vigorous ablation in the GHCre/iDTR model, where increased proliferation and transdifferentiation are not observed [12].

In conclusion, the mature pituitary can restore more than one cell type after injury, and restorative mechanisms seem to depend on the cell type that is affected. Further studies deciphering molecular pathways (including hypothalamic influence) underlying pituitary cell regeneration, may guide to potential clinical applications. As suggested by this and our previous study [12], different therapeutic approaches may be needed for different cell types.

Footnotes

Acknowledgments

This work was supported by the Fund for Scientific Research-Flanders (Belgium) (FWO-Vlaanderen) and the Research Fund (Onderzoeksfonds) of the KU Leuven. Q.F. received a Selective Bilateral Agreement (SBA) Scholarship (KU Leuven).

Author Disclosure Statement

No competing financial interests exist.

References

Vankelecom

. 2010. Pituitary stem/progenitor cells: embryonic players in the adult gland? Eur J Neurosci, 32:2063–2081.

Chen

, Hersmus

, Van Duppen

, Caesens

, Denef

, Vankelecom

. 2005. The adult pituitary contains a cell population displaying stem/progenitor cell and early embryonic characteristics. Endocrinology, 146:3985–3998.

Chen

, Gremeaux

, Fu

, Liekens

, Van Laere

, Vankelecom

. 2009. Pituitary progenitor cells tracked down by side population dissection. Stem Cells, 27:1182–1195.

Fauquier

, Rizzoti

, Dattani

, Lovell-Badge

, Robinson

. 2008. SOX2-expressing progenitor cells generate all of the major cell types in the adult mouse pituitary gland. Proc Natl Acad Sci U S A, 105:2907–2912.

Garcia-Lavandeira

, Quereda

, Flores

, Saez

, Diaz-Rodriguez

, Japon

, Ryan

, Blasco

, Dieguez

, Malumbres

, Alvarez

. 2009. A GRFa2/Prop1/stem (GPS) cell niche in the pituitary. PLoS One, 4:e4815.

Gleiberman

, Michurina

, Encinas

, Roig

, Krasnov

, Balordi

, Fishell

, Rosenfeld

, Enikolopov

. 2008. Genetic approaches identify adult pituitary stem cells. Proc Natl Acad Sci U S A, 105:6332–6337.

Pevny

, Rao

. 2003. The stem-cell menagerie. Trends Neurosci, 26:351–359.

Beltrami

, Barlucchi

, Torella

, Baker

, Limana

, Chimenti

, Kasahara

, Rota

, Musso

et al. 2003. Adult cardiac stem cells are multipotent and support myocardial regeneration. Cell, 114:763–776.

Shin

, Lee

, Guo

, Kim

, Lim

, Qu

, Mysorekar

, Beachy

. 2011. Hedgehog/Wnt feedback supports regenerative proliferation of epithelial stem cells in bladder. Nature, 472:110–114.

10.

Wong

, Reiter

. 2011. Wounding mobilizes hair follicle stem cells to form tumors. Proc Natl Acad Sci U S A, 108:4093–4098.

11.

Vankelecom

. 2007. Stem cells in the postnatal pituitary? Neuroendocrinology, 85:110–130.

12.

, Gremeaux

, Luque

, Liekens

, Chen

, Buch

, Waisman

, Kineman

, Vankelecom

. 2012. The adult pituitary shows stem/progenitor cell activation in response to injury and is capable of regeneration. Endocrinology, 153:3224–3235.

13.

Landolt

. 1973. Regeneration of human pituitary. J Neurosurg, 39:35–41.

14.

Saeger

, Warnecke

. 1980. Ultrastructural examination of the regeneration of the rat adenohypophysis after partial hypophysectomy. Virchows Arch A Pathol Anat Histopathol, 387:279–288.

15.

Buch

, Heppner

, Tertilt

, Heinen

, Kremer

, Wunderlich

, Jung

, Waisman

. 2005. A Cre-inducible diphtheria toxin receptor mediates cell lineage ablation after toxin administration. Nat Methods, 2:419–426.

16.

Luque

, Lin

, Córdoba-Chacón

, Subbaiah

, Buch

, Waisman

, Vankelecom

, Kineman

. 2011. Metabolic impact of adult-onset, isolated, growth hormone deficiency (AOiGHD) due to destruction of pituitary somatotropes. PLoS One, 6:e15767.

17.

Castrique

, Fernandez-Fuente

, Le Tissier

, Herman

, Levy

. 2010. Use of a prolactin-Cre/ROSA-YFP transgenic mouse provides no evidence for lactotroph transdifferentiation after weaning, or increase in lactotroph/somatotroph proportion in lactation. J Endocrinol, 205:49–60.

18.

Luque

, Amargo

, Ishii

, Lobe

, Franks

, Kiyokawa

, Kineman

. 2007. Reporter expression, induced by a growth hormone promoter-driven Cre recombinase (rGHp-Cre) transgene, questions the developmental relationship between somatotropes and lactotropes in the adult mouse pituitary gland. Endocrinology, 148:1946–1953.

19.

Stahl

, Kendall

, Brinkmeier

, Greco

, Watkins-Chow

, Campos-Barros

, Lloyd

, Camper

. 1999. Thyroid hormone is essential for pituitary somatotropes and lactotropes. Endocrinology, 140:1884–1892.

20.

Kobayashi

, Yamamoto

, Kikuyama

, Machida

, Kobayashi

. 2009. Impaired development of somatotropes, lactotropes and thyrotropes in growth-retarded (grt) mice. J Toxicol Pathol, 22:187–194.

21.

Otto

, tom Dieck

, Bauer

. 1996. Dipeptide uptake by adenohypophysial folliculostellate cells. Am J Physiol, 271:C210–C217.

22.

Gremeaux

, Fu

, Chen

, Vankelecom

. 2012. Activated phenotype of the pituitary stem/progenitor cell compartment during the early-postnatal maturation phase of the gland. Stem Cells Dev, 21:801–813.

23.

Luquet

, Perez

, Hnasko

, Palmiter

. 2005. NPY/AgRP neurons are essential for feeding in adult mice but can be ablated in neonates. Science, 310:683–685.

24.

Bennett

, van Rijn

, Jung

, Inaba

, Steinman

, Kapsenberg

, Clausen

. 2005. Inducible ablation of mouse Langerhans cells diminishes but fails to abrogate contact hypersensitivity. J Cell Biol, 169:569–576.

25.

Thorel

, Nepote

, Avril

, Kohno

, Desgraz

, Chera

, Herrera

. 2010. Conversion of adult pancreatic alpha-cells to beta-cells after extreme beta-cell loss. Nature, 464:1149–1154.

26.

Baltus

, Kowalski

, Zhai

, Tutter

, Quinn

, Wall

, Kadam

. 2009. Acetylation of Sox2 induces its nuclear export in embryonic stem cells. Stem Cells, 27:2175–2184.

27.

Kineman

, Faught

, Frawley

. 1992. The ontogenic and functional relationships between growth hormone- and prolactin-releasing cells during the development of the bovine pituitary. J Endocrinol, 134:91–96.

28.

Frawley

, Boockfor

. 1991. Mammosomatotropes: presence and functions in normal and neoplastic pituitary tissue. Endocr Rev, 12:337–355.

29.

Wynick

, Small

, Bacon

, Holmes

, Norman

, Ormandy

, Kilic

, Kerr

, Ghatei

et al. 1998. Galanin regulates prolactin release and lactotroph proliferation. Proc Natl Acad Sci U S A, 95:12671–12676.

30.

Nolan

, Levy

. 2009. The trophic effects of oestrogen on male rat anterior pituitary lactotrophs. J Neuroendocrinol, 21:457–464.

31.

, D'Hoker

, Stangé

, Bonné

, De Leu

, Xiao

, Van de Casteele

, Mellitzer

, Ling

et al. 2008. Beta cells can be generated from endogenous progenitors in injured adult mouse pancreas. Cell, 132:197–207.

32.

Rando

. 2006. Stem cells, ageing and the quest for immortality. Nature, 441:1080–1086.

33.

Conboy

, Conboy

, Smythe

, Rando

. 2003. Notch-mediated restoration of regenerative potential to aged muscle. Science, 302:1575–1577.

34.

Kwan

, White

, Segil

. 2009. Development and regeneration of the inner ear. Ann N Y Acad Sci, 1170:28–33.

35.

Baddour

, Sousounis

, Tsonis

. 2012. Organ repair and regeneration: an overview. Birth Defects Res C Embryo Today, 96:1–29.

36.

Ozaki

, Matsubara

, Seo

, Okamoto

, Nagashima

, Sasaki

, Hayase

, Murata

, Liao

et al. 2012. Thyroid regeneration: characterization of clear cells after partial thyroidectomy. Endocrinology, 153:2514–2525.

37.

Denef

. 2008. Paracrinicity: the story of 30 years of cellular pituitary crosstalk. J Neuroendocrinol, 20:1–70.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.85 MB

0.81 MB

0.09 MB

1.01 MB