Structural and proteomic repercussions of growth hormone receptor deficiency on the pituitary gland: Lessons from a translational pig model

Growth hormone receptor deficiency (GHRD) results in low serum insulin‐like growth factor 1 (IGF1) and high, but non‐functional serum growth hormone (GH) levels in human Laron syndrome (LS) patients and animal models. This study investigated the quantitative histomorphological and molecular alterations associated with GHRD. Pituitary glands from 6 months old growth hormone receptor deficient (GHR‐KO) and control pigs were analyzed using a quantitative histomorphological approach in paraffin (9 GHR‐KO [5 males, 4 females] vs. 11 controls [5 males, 6 females]), ultrathin sections tissue sections (3 male GHR‐KO vs. 3 male controls) and label‐free proteomics (4 GHR‐KO vs. 4 control pigs [2 per sex]). GHR‐KO pigs displayed reduced body weights (60% reduction in comparison to controls; p < .0001) and decreased pituitary volumes (54% reduction in comparison to controls; p < .0001). The volume proportion of the adenohypophysis did not differ in GHR‐KO and control pituitaries (65% vs. 71%; p = .0506) and GHR‐KO adenohypophyses displayed a reduced absolute volume but an unaltered volume density of somatotrophs in comparison to controls (21% vs. 18%; p = .3164). In GHR‐KO pigs, somatotroph cells displayed a significantly reduced volume density of granules (23.5%) as compared to controls (67.7%; p < .0001). Holistic proteome analysis of adenohypophysis samples identified 4660 proteins, of which 592 were differentially abundant between the GHR‐KO and control groups. In GHR‐KO samples, the abundance of somatotropin precursor was decreased, whereas increased abundances of proteins involved in protein production, transport and endoplasmic reticulum (ER) stress were revealed. Increased protein production and secretion as well as significantly reduced proportion of GH‐storing granules in somatotroph cells of the adenohypophysis without an increase in volume density of somatotroph cells in the adenohypophysis could explain elevated serum GH levels in GHR‐KO pigs.


| INTRODUCTION
Growth hormone receptor deficiency (GHRD) is a rare autosomal recessive hereditary disorder caused by loss-of-function mutations of the GHR gene. 1 Affected patients (human Laron syndrome [LS]) 2 as well as murine (the Ghr-KO mouse) 3 and porcine animal models (the GHR-KO pig) 4 display constantly elevated but non-functional levels of growth hormone (GH).The lack of GHR-mediated hepatic insulin-like growth factor 1 (IGF1) production leads to low circulating levels of IGF1 and a disturbed feedback regulation of pituitary GH release, causing constantly high GH levels in serum.The phenotypic consequences include inter alia postnatal growth retardation 2,4,5 and an increased accumulation of body fat [5][6][7] due the lack of lipolytic GH action. 8Another consequence of GHRD is increased insulin sensitivity, [5][6][7] potentially associated with the lack of GH antagonism on insulin. 9ile a correlation between low serum GH levels and reduced pituitary size has been observed in GH deficiency, 10 an enlargement of pituitary glands was not verified in LS patients using magnetic resonance imaging (MRI). 11In contrast, when corrected for body weight, Ghr-KO mouse pituitaries were relatively enlarged and revealed an increased amount and proportion of GH producing somatotroph cells in the adenohypophysis. 12Aim of the present study was to investigate the effects of a disturbed GH-IGF1 axis on structural and molecular characteristics of the pituitary gland in a large animal model for GHRDthe GHR-KO pig.We employed mass spectrometry-based label-free proteomics to reveal molecular changes and associated biological pathways in GHR-KO adenohypophysis.To complement molecular findings, further quantitative morphological analyses of the cellular composition of the adenohypophysis and the ultrastructure of somatotroph cells were performed.Specifically, the absolute volumes and volume proportions of the pituitary compartments, of the somatotroph cells in the adenohypophysis, and of the GH secretory granules in the somatotroph cells were determined, using unbiased quantitative-stereological analysis techniques. 13| MATERIALS AND METHODS

| Generation of pituitary gland tissue samples
All animal procedures were approved by the responsible animal welfare authority (Regierung von Oberbayern; permission 55. 21-54-2532-70-12)   and performed according to the German Animal Welfare Act and Directive 2010/63/EU for the protection of animals used for scientific purposes.
The present study examined pituitary glands (PG) from nine GHR-KO pigs (5 males and 4 females; age = 6 months) and 11 age-matched non-transgenic control littermates (5 males and 6 females) which were obtained in a previous study. 4Pituitary glands were extracted as described previously 14 and weighed to the nearest milligram.The total volume of the PG (V (PG) ) was determined from its weight (m (PG) ) and its specific weight (γ (PG) = 0.89 g/cm 3 ) as V (PG) = m (PG) /γ (PG) . 13,15dy weight and brain weight of the animals were also assessed in the same previous study. 4r holistic proteome analysis, a sample of approximately 1 mm 3 was excised from the adenohypophysis of four GHR-KO and four control animals (2 males and 2 females each) and immediately frozen on dry ice.4][15][16] After fixation of the tissue in neutrally buffered 4% formaldehyde solution for 24 h at room temperature, the pituitary gland was exhaustively sliced into approximately 1 mm thick, parallel slabs.
For electron microscopy, each three tissue specimen of approximately 1 mm 3 were excised from the adenohypophysis (pars distalis) of three male GHR-KO and three male WT control animals, using a systematic random sampling approach. 13These tissue samples were postfixed in 6.25% glutaraldehyde in Sorensen's phosphate buffer (Serva, USA) for 6 h and subsequently embedded in epoxy resin (Epon, aka glycidyl ether 100; methyl nadic anhydride; 2,4,6-tris [dimethylaminomethyl]phenol; 2-dodecyl succinic acid anhydride [Serva, USA]) in random orientation. 13Subsequently, ultrathin tissue sections of approximately 70 nm were prepared and mounted on 1% formvar coated 200 mesh nickel grids (Veco, Neatherlands).Next, sections were treated with sodium metaperiodate (Merck, Germany) for 10 min and washed with deionized distilled H 2 O (ddH 2 O) for another 5 min.Consecutively, sections were incubated with 1% bovine serum albumin (BSA: Sigma Aldrich, USA) in ddH 2 O solution for 5 min, 0.1 M phosphate-buffered saline (PBS) glycine buffer for 10 min and phototonic bandgap (PBG) buffer containing 0.2 g gelatin (Merck, Germany) and 0.5 g BSA three times for 2 min.Blocking was performed with normal donkey serum (Aurion, Netherlands) diluted 1:10 in PBS for 30 min.For unambiguous identification of somatotroph cells with GH-containing hormone granules and differentiation from other pituitary cell types, immuno-gold electron microscopy was used, as described previously. 17For this, ultrathin sections were incubated with goat anti-GH antibody (R&D Systems, USA, 1:200 in PBS) at 4 C over night in a humid container.After washing with PBG six times for 2 min, sections were incubated with a 10-nm immuno-gold conjugated donkey anti-goat IgG antibody (Science Service GmbH, Germany, 1:50 in PBS) for 1 h at room temperature.Labeled sections were then washed with PBG, 0.1 M glycine-PBS and ultimately with ddH 2 O. Finally, the slides were contrasted with 2% uranyl acetate (Science Service GmbH, Germany) and 3% lead citrate (Ultrostain II Leica, Germany) according to standard protocols.The sections were examined using and EM10 transmission electron microscope (Zeiss, Germany) at different magnifications.
For qualitative and quantitative histomorphological analyses, the remaining formalin-fixed pituitary gland tissue slabs were embedded in paraffin, while maintaining the orientation of their section surfaces.
Sections of approximately 4 μm thickness were stained with hematoxylin and eosin (HE), and subjected to immunohistological detection of growth hormone (GH) using a specific polyclonal goat anti-rat GH antibody (#AF1566 R&D Systems, USA; 1:3000) and biotinylated rabbit anti-goat IgG solution (DAako, USA; 1:100), diaminobenzidine as chromogen and hemalum as nuclear counterstain, as previously described in detail. 132 | Quantitative morphological analyses 2.2.1 | Determination of the volume of the adenohypophysis (AH) and the total volume of somatotrophs (SOM) in the AH Volume densities (V V(X/Y) ) were estimated based on area densities using the principle of Delesse.18,19 According to this principle, the volume density of a target structure V (X) in a reference compartment V (Y) is equal to the quotient of their respective cumulative section areas ( P A (X) / P A (Y) ).The so-called point counting method allows a simple determination of the area density (A A(X/Y) ) by calculating the quotient of the total number of points hitting section profiles of the target structure ( P Pt (X) ) and the total number of points hitting the reference compartment ( P Pt (Y) ) in the same sections.20,21 The points can be represented by crosses to establish uniform rules regarding the "hitting" or "not hitting" of a point (Table S1).
The volume density of the adenohypophysis (AH) in the pituitary gland (PG) (V V(AH/PG) ) and the other pituitary compartments was determined as the fraction of the section areas of the AH (i.e., target structure) and the total section profile area of PG tissue (i.e., reference compartment) in HE stained paraffin sections of the PG (11 control and 9 GHR-KO) at 12.5Â microscopic magnification based on the point density of the AH in the PG (Pt Pt(AH/PG) ) as described above and previously. 13On the average, 750 ± 77 for the adenohypophysis, 665 ± 73 for the pars intermedia and 2260 ± 151 points were counted per case for the neurohypophysis (mean ± SE).
The volume density of somatotroph cells in the AH (V V(SOM/AH) ) was analogously determined in immunohistological sections (11 control and 9 GHR-KO) in which somatotroph cells were identified by detection of GH, at 200Â microscopic magnification, as the quotient of the total number of points hitting GH-positive cell section profiles ( P Pt (SOM) ) and total number of points hitting AH-tissue section area ( P Pt (AH) ) per case.On the average, 3326 ± 74 points were counted per case.The total volume of the AH (V (AH, PG) ) and the other compartments was then calculated as the product of V (PG) and V V(AH/PG) and the total volume of somatotrophs in the adenohypophysis (V (SOM, AH) ), respectively, as the product of V V(SOM/AH) and V (AH, PG) (Table S1).

| Determination of the volume of growth hormone granules in somatotroph cells by unbiased quantitative stereology
To estimate the volume density of GH containing granules (GH-GRAN) in somatotroph cells of the adenohypophysis (V V(GH-GRAN/SOM) ), ultrathin sections of two systematic uniform random (SUR) sampled locations of the adenohypophysis derived from three male control and three male GHR-KO pigs were analyzed by electron microscopy.In these sections, 29.8 ± 2.0 SUR sampled fields of view per case were photographed at 7000 x magnification.Within these areas, per case 51.0 ± 18.0 somatotroph cell section profiles were identified, based on their typical cellular morphological characteristics and detection of immuno-gold labeled GH-containing secretory granules.
V V(GH-GRAN/SOM) was determined, using the point counting method like described above and previously. 13Per case, 2211.3 ± 287.0 points on somatotroph cells containing GH-containing granules per case were counted.The total volume of GH-granules in the somatotroph cells (V (GH-GRAN,SOM) ) was calculated by multiplication of V V(GH-GRAN/SOM) and the total volume of somatotroph cells in the adenohypophysis (V (SOM, AH) ) of the respective animal.

| Statistical analysis
Quantitative morphological data was statistically evaluated using a general linear models procedure (PROC GLM; SAS 8.2), taking the fixed effects of Group (GHR-KO, WT), Sex (male, female), and the interaction Group*Sex into account.Since none of the parameters was significantly affected by Sex and most were not affected by the interaction of Group*Sex (Table S2), least squares means (LSMs) and standard errors (SEs) of LSMs were calculated for the two groups and compared using student's t-tests.Correction for multiple testing was performed with the Bonferroni method using the ADJUST = BON statement.The ultrastructural measurements (GH granular) in three male control and three male GHR-KO pigs were compared using student t-tests using GraphPad Prism (GraphPad version 5.04).MaxQuant (version 1.6.7.0) 22 using its built-in search engine Andromeda 23 and the NCBI RefSeq Sus scrofa database (downloaded in April 2020) alongside the MaxQuant contaminants fasta file.Protein intensities were normalized with the MaxLFQ label-free normalization approach. 24All statistical and bioinformatic analyses were done using Perseus (version 1.6.7.0) 25 or the R framework. 26oteins detected in at least 70% of all replicates in at least one condition were kept for quantitative analysis.To handle missing values, data imputation from a normal distribution was performed using the Perseus default parameters (width -0.3 and down-shift -

1.8). A volcano plot was generated with Perseus using two-tailed
Student's t-test statistics and permutation-based FDR cutoff of 0.05, together with an s0-parameter of 0.1 to additionally consider fold changes. 27When appropriate, comparisons between multiple conditions were performed using a two-way analysis of variance (ANOVA).
Resulting p-values were corrected for multiple hypothesis testing using the Benjamini-Hochberg (BH) procedure.Proteins with a significant two-way ANOVA interaction effect (adjusted p-value <.05) were followed by a Tukey's HSD post hoc test.For unsupervised cluster analysis, principal component analysis (PCA) (prcomp() function with default parameters in R) and hierarchical clustering (ComplexHeatmap R package 28 ) was applied.Overrepresentation analysis, based on significantly changed proteins was performed using WebGestalt 29 and the functional category "GO Biological Process nonRedundant".False discovery rate was controlled using BH procedure.All data visualization was performed in R. 26 3 | RESULTS 3.1 | GHR-KO pigs display significantly reduced body weights, pituitary-, adenohypophyseal-, and total somatotroph-volumes GHR-KO pigs investigated in this study showed a significant, 60% reduction in body weight (Figure 1A), as compared to controls (38.5 ± 3.6 kg vs. 97.5 ± 3.3 kg; p < .0001).Correspondingly, the absolute weights and volumes of the pituitary glands (V (PG) ) of GHR-KO pigs were also significantly reduced by 54% (126.6 ± 8.4 mg/115.1 ± 7.6 mm 3 in GHR-KO vs. 273.8± 7.6 mg / 248.9 ± 6.9 mm 3 in controls; p < .0001)(Figure 1B).When corrected for body weight, the relative weight of GHR-KO pituitary glands was only 1.15-times larger than in control pigs (p = .0367)(Figure 1C).The absolute brain weight was only reduced by approximately 15% in GHR-KO pigs (83.4 ± 2.6 g in GHR-KO vs. 95.6 ± 2.3 g in controls; p = .0017)(Table 1).Accordingly, relative brain weight of GHR-KO animals was more than doubled (by factor 2.2 of control animal's values; p < .0001;see Table 1).
The reduction in pituitary gland weight and volume was not in proportion to brain weight reduction (see Table 1).
While the volume proportion of the adenohypophysis (i.e., pars distalis) in the pituitary gland (V V(AH/PG) ) was not significantly different in GHR-KO and control pigs (0.65 ± 0.02 in GHR-KO vs. 0.7 ± 0.02 in controls; p = .0506;Figure 1D), the absolute volume of the adenohypophysis (V (AH, PG) ) in GHR-KO pigs was also significantly lower than in controls (75.4 ± 6.8 mm 3 in GHR-KO vs. 176.1 ± 6.2 mm 3 in controls; p < .0001; Figure 1E).While the reduction of volumes of pituitary compartments appeared in correspondence to the reduction in body weight, the neurohypophysis (i.e., pars nervosa) in the pituitary gland was less affected by growth retardation, as the volume proportion of the neurohypophysis in the pituitary gland (V V(NH/PG) ) appeared increased (0.27 ± 0.02 in GHR-KO vs. 0.22 ± 0.01 in controls; p = .039)as well as the relative volume (rel. V (NH, PG) ; 0.81 ± 0.05 mm 3 per kg BW in GHR-KO vs. 0.57 ± 0.05 mm 3 per kg BW in controls; p = .0026)when corrected for body weight (see Table 1).
The volume proportion of somatotroph cells in the adenohypophysis (V V(SOM/AH) ) was also not significantly different in GHR-KO and control pigs (0.21 ± 0.02 in GHR-KO vs. 0.18 ± 0.02 in controls; p = .3164; Figure 1F and 2A), while the total volume of somatotroph cells in the adenohypophysis (V (SOM, AH) ) of GHR-KO pigs was correspondingly significantly lower as compared to controls (15.9 ± 3.2 mm 3 in GHR-KO vs. 32.3 ± 2.9 mm 3 in controls; p = .0016;Figure 1G).
None of the investigated parameters revealed an effect for sex (see Table S2).The volume density (H), as well as the absolute volume of growth hormone-containing secretory granules in somatotroph cells (I), however, is significantly reduced in GHR-KO pigs (3 male GHR-KO vs. 3 male control pigs; compare to Figure 2B).The box plots show medians, 25th and 75th percentiles (box), and extremes (whiskers).
pituitary gland showed an effect for the interaction of group*sex (V V(pars intermedia/PG) p = .0321)(see Table S2).

| Somatotroph cells in GHR-KO pigs display significantly reduced relative and absolute volumes of secretory granules
Ultrastructural quantitative-morphological analysis of somatotroph cells revealed a significant reduction of the relative volume of secretory granules in GHR-KO pigs (Figures 1H and 2B).The volume density of secretory granules in somatotroph cells in the adenohypophysis (V V(GH-GRAN/SOM) ) in GHR-KO pigs (23.5 ± 2.2%) was reduced by 65% (p < .0001),as compared to control pigs (67.7 ± 2.2%).Correspondingly, the total volume of GH-containing secretory granules in somatotroph cells of GHR-KO pigs (V (GH-GRAN, SOM) ) was also significantly reduced (3.0 ± 0.4 mm 3 in GHR-KO vs. 28.0 ± 4.2 mm 3 in controls; p = .0039)(Figure 1I).Occasionally, mild dilation of the ER was observed in somatotroph cell section profiles of GHR-KO pigs.However, this observation was not further T A B L E 1 Growth parameters and quantitative stereological data of control and GHR-KO pigs.Note: Body weight, brain weight and pituitary data was assessed from 11 control (5 males; 6 females) and nine GHR-KO pigs (5 males; 4 females) in a previous study (Hinrichs et al., 2018) and were statistically evaluated using a general linear models procedure (PROC GLM; SAS 8.2), taking the fixed effects of Group (GHR-KO, WT), Sex (male, female), and the interaction Group*Sex into account.Since none of the parameters were significantly affected by Sex and most were not affected by the interaction of Group*Sex (Table S2), least squares means (LSMs) and standard errors (SEs) of LSMs were calculated for the two groups and compared using student t-tests.Correction for multiple testing was performed with the Bonferroni method using the ADJUST = BON statement.The ultrastructural measurements (GH granula) in three male control and three male GHR-KO pigs were compared using student t-tests using GraphPad Prism (GraphPad Version 5.04).Abbreviations: AH, adenohypophysis; BW, body weight; GH-GRAN, GH-containing secretory granules; NH, neurohypophysis; PG, pituitary gland; Rel., relative; SOM, somatotroph cells.
quantitatively analyzed on a morphological level due to occurrence in only a minor proportion of the examined somatotroph cell cross section profiles.
Overrepresentation analysis using the significantly changed proteins revealed 36 gene sets significantly enriched in GHR-KO pig's adenohypophysis and 45 gene sets enriched in control pig's adenohypophysis (FDR <0.05) (Table S5).Inter alia, gene sets such as response to endoplasmic reticulum stress, cargo loading into vesicle, protein exit from endoplasmic reticulum, response to topologically incorrect protein, protein folding, and cytoplasmic translation were overrepresented among the proteins increased in abundance in GHR-KO pig's adenohypophysis (Figure 4), whereas gene sets such as platelet degranulation, protein activation cascade, homotypic cell-cell adhesion, regulation of response to wounding, and humoral immune response were overrepresented among the proteins increased in abundance in control pig's adenohypophysis (Figure 3D).In addition, two-way ANOVA revealed five proteins that were differently abundant with an effect for sex and the abundance of only four proteins were significantly influenced by the interaction of group x sex (Table S6).The aim of this study was to investigate the histomorphological and molecular alterations of the adenohypophysis in the pathophysiological condition of GHRD.
Under physiological conditions (functional GHR), GH is secreted in a pulsatile manner from secretory granules storing GH (reviewed in 31 ).
In contrast, baseline serum GH levels are constantly elevated in GHRD and show higher peaks 32 due to the lack of IGF1 mediated negative feedback regulation.In association with high serum GH levels, a study in Ghr-KO mice pituitaries indicated an increased proportion of GH producing somatotroph cells. 12Other adenohypophyseal cell populations (lactotroph, corticotroph, tyrotroph and gonadotroph) remained with unaltered distribution and proportional reduction.Nevertheless, no pituitary hypertrophy was detected in human LS patients despite high levels of GH. 11 Porcine pituitaries show close similarities to human pituitaries regarding histological and stereological composition. 33In humans, somatotroph cells make up about 50% of hormone-producing cells in the adenohypophysis, 34 which appears in line with observations in pigs 33 and also in mice, 35 with a dependency on age and genetic background.We recently confirmed that GHR-KO pigs resemble the hallmark for GHRD, which are elevated GH and decreased IGF1 levels. 4For an objective quantitative assessment of the effects of GHRD on relevant morphological characteristics of the porcine pituitary gland and somatotroph cells, established unbiased quantitative-stereological analyses were [13][14][15][16] applied, demonstrating that the volume proportions of the adenohypophysis in the pituitary gland, and of somatotroph cells in the adenohypophysis were not significantly altered in GHR-KO pigs.
Interestingly, the volume density of GH-containing secretory granules in somatotroph cells of GHR-KO pigs was significantly reduced.This is in line with the 3.5-fold decreased abundance of somatotropin precursor (GH1) revealed by proteomic analysis.Together with their increased systemic GH-levels, 4 the reduced proportion of secretory granules in somatotroph cells of GHR-KO pigs could be interpreted as a consequence of a permanently increased release of GH due to a lack of feedback inhibition.An increased protein production is also indicated in proteomic data of GHR-KO pig adenohypophysis as several 40S (RPSA, RPS26, RPS29) and 60S ribosomal subunits (RPLP2, RPL15, RPL36) were increased in abundance, which are inter alia associated with the enriched gene sets "translational initiation", "cytoplasmic translation" and "protein localization to endoplasmic reticulum".Following an increased protein translation, several proteins located in the ribosome, such as disulfide isomerases (inter alia ERP44, PDIA3, PDIA4, PDIA6) and chaperones (inter alia UGGT1, HSP90B1), associated with the enriched gene set "protein folding" showed an increase in abundance.For the transport from the ER to the Golgi complex, proteins are packed in COPIIcoated vesicles.Their components are formed by protein transport proteins whose abundance was significantly increased in GHR-KO adenohypophysis (SEC13, SEC23A, SEC24A, SEC24D, SEC31A), which were related to the enriched gene set "cargo loading vesicle" and "protein exit from ER".The increased demand for peptide hormones requires an adaptation by expanding ER membranes and can increase the amount of unfolded proteins in cells, which can cause ER stress. 36Accordingly, the most prominent proteins increased in GHR-KO adenohypophysis were related to "response to ER stress", followed by "response to topologically incorrect protein".Those pathways contained prominent proteins involved in ER-associated degradation (ERAD) of misfolded proteins, such as ERLEC1, HSP90B1, DNAJB11, UBQLN1, and chaperones (MANF, HYOU1, HSPA5).
In conclusion, proteomic analysis together with quantitative histomorphological analysis findings suggest increased production and release of GH, while the volume proportion of the adenohypophysis (i.e., pars distalis) in the pituitary gland and the volume proportion of somatotroph cells in the adenohypophysis remain unaffected.This might explain the observation of unaffected pituitary size in human LS patients.
In addition to GH from somatotroph cells, the anterior pituitary secrets glycoproteins derived from gonadotropic (follicle-stimulating and luteinizing hormone -FSH and LH) and thyrotrophic cells (thyroid-stimulating hormone [TSH]) formed as dimers from a shared alpha chain and a hormone-specific beta subunit. 37Since the glycoprotein hormones alpha chain precursor (CGA) showed an increased abundance in GHR-KO adenohypophysis samples and disturbances in at least thyroid hormone levels are reported from mice, 38 future studies should aim to investigate glycoproteins in GHRD.
In contrast to the majority of other tissues and organs, development and growth of the brain appear to be largely independent of GH action, resulting in an increased brain weight to body weightratio in GHR-KO pigs 4 and Ghr-KO mice. 39In contrast, giant transgenic mice overexpressing GH have a decreased relative brain weight. 40A study by Asa et al. 12 inter alia assessed the weight of pituitary glands in Ghr-KO mice and observed that the reduction of pituitary weights was less than the reduction of overall body weight.In the present study, we showed that the pituitary compartments are differently affected by GHRD.While absolute volumes of all pituitary compartments were reduced, only the volumes of the adenohypophysis and pars intermedia were reduced in proportion to body weight, while the neurohypophysis was less affected.In contrast to the adenohypophysis and pars intermedia, which are derived from the ectoderm, the neurohypophysis is derived from the base of the diencephalon as a part of the neuroectoderm. 41erefore, the same (unclassified) mechanisms leading to the GH-independent growth of the brain, might possibly also affect the neurohypophysis.

| CONCLUSION
This study investigated histomorphological, ultrastructural and molecular alterations of the pituitary gland in a translational pig model for GHRD.
Quantitative stereological analyses identified functionally relevant morphological alterations of somatotroph cells in the pituitary gland supporting the interpretation of alterations of the pituitary proteome profile and increased serum GH levels.The latter are neither associated with a relative enlargement of the adenohypophysis in GHR-KO pituitaries nor a significant increase in the volume density of somatotroph cells in the adenohypophysis.In fact, high serum GH levels appear to represent a consequence of increased protein production and secretion, leading to a depletion of GH-containing granules in somatotroph cells.

Frozen tissue samples from 2 . 5 |
the adenohypophysis derived from four GHR-KO and four control animals (2 males and 2 females each) were transferred into precooled tubes and cryopulverized using a CP02 Automated Dry Pulverizer (Covaris, USA) according to the manufacturer's instructions.Next, $10 mg of tissue powder was lysed in 8 M urea/50 mM ammonium bicarbonate by ultrasonication (18 cycles of 10 s pulse and 20 s pause) using a Sonopuls HD3200 (Bandelin, Germany).Protein content of lysates was quantified using a Pierce 660 nm Protein Assay (Thermo Fisher Scientific, USA).From each biological replicate, 20 μg of protein were further processed.Before digestion, disulfide bonds were reduced (45 mM dithiothreitol/20 mM tris[2-carboxyethyl] phosphine, 30 min, 56 C) and cysteine residues were alkylated (100 mM iodoacetamide, 30 min, room temperature), followed by quenching of a residual iodoacetamide with dithiothreitol (90 mM, 15 min, room temperature).Proteins were digested with modified porcine trypsin (Promega, USA) for 16 h at 37 C (1:50 enzyme to protein ratio).Digested peptides were dried before analysis using a vacuum centrifuge.Nano-liquid chromatography (LC)-tandem mass spectrometry (MS/MS) analysis and bioinformatics LC-MS/MS was performed on an UltiMate 3000 nano LC system online coupled to a Q-Exactive HF-X mass spectrometer via a nanoelectrospray ion source (Thermo Fisher Scientific).Briefly, 1.5 μg of peptides were transferred to a PepMap 100 C18 trap column (100 μm Â 2 cm, 5 μM particles, Thermo Fisher Scientific) at a flow rate of 5 μL/min mobile phase A (0.1% formic acid and 1% acetonitrile in water).Separation was performed on an analytical column (PepMap RSLC C18, 75 μm Â 50 cm, 2 μm particles, Thermo Fisher Scientific) at 250 nL/min flow rate with a 160-min gradient of 3%-25% of solvent B, followed by a 10-min raise to 40% and 5-min raise to 85%.Solvent A was 0.1% formic acid in water and solvent B was 0.1% FA in acetonitrile.For mass spectrometry data acquisition cycles of one full scan and 15 datadependent MS/MS scans were used.Raw files were processed with Solely for the absolute and relative volume of the pars intermedia (V (pars intermedia, PG) p = .0464;rel.V (pars intermedia, PG) p = .0192)and the volume proportion of the pars intermedia in the total volume of the adenohypophysis V (AH, PG) U R E 1 (A, C) GHR-KO pigs (9 GHR-KO [5 males; 4 females] vs. 11 control pigs [5 males; 6 females]) display significantly reduced body weights (A) and pituitary gland (PG) volumes (B), as compared to control pigs.The volume of GHR-KO pituitary glands appears just slightly increased when corrected for body weight (C).(D-I) Quantitative morphological analysis of the pituitary gland and somatotroph cells.The volume density of the adenohypophysis (AH) in the pituitary gland (D) and the volume density of the somatotroph cells (SOM) in the AH (F) do not significantly differ between control-and GHR-KO pigs.Corresponding to the reduced volume of the PG, the absolute volumes of the AH-and the somatotroph cells (E, G) are significantly reduced in GHR-KO pigs.

2. 7 )
, and 60S ribosomal protein L36 (RPL36; l2fc 2.2) were among the most significantly elevated proteins.The set of proteins most decreased in abundance in GHR-KO adenohypophysis contained among other proteins involved in retinol metabolism, such as retinoic F I G U R E 2 (A) Immunohistological detection of growth hormone (brown color) in somatotroph cells in the adenohypophysis of Control-and GHR-KO pigs.In GHR-KO pigs, somatotroph cell section profiles display a granular staining pattern with a slightly weaker staining intensity as compared to control pigs.Formalin-fixed, paraffinembedded tissue sections.Bars = 50 μm and 25 μm in insets.(B) Somatotroph cell ultrastructure.Somatotrophs (som) of GHR-KO pigs apparently display fewer hormone granule section profiles than controls.Transmission electron microscopy.Arrows indicate immuno-gold labeled GH-containing secretory granules.Bars = 2.5 μm and 500 nm in insets.

F I G U R E 3
Quantitative proteome analysis of adenohypophysis from GHR-KO and control pigs.(A) Label-free quantitation (LFQ) intensity values of adenohypophysis proteomes of each sample (GHR-KO: n = 2 female, n = 2 male; control, n = 2 female, n = 2 male) were grouped by unsupervised hierarchical clustering and visualized as a heatmap.The color code indicates z-score normalized abundance values.(B) Principal component analysis indicates genotype (GHR-KO vs. control) as the strongest contributor to the intersample variance of the adenohypophysis proteome.The shape of each spot corresponds to the sex, and the color to the genotype.(C) Volcano plot visualizes the protein abundance change between conditions (GHR-KO/control).Proteins significantly altered in abundance (FDR < 0.05) are colored in red and blue for increased and decreased abundance, respectively.Black curves represent the permutation-based false discovery rate (FDR) significance cutoff.(D) Overrepresentation analysis of adenohypophysis proteins less abundant in GHR-KO (left column) and more abundant in GHR-KO (right column).Size of each bubble indicates the corresponding number of differentially abundant proteins (referred to as genes mapped in the figure) and color the significance of enrichment.Enrichment score is magnitude of overrepresentation.Multiple testing correction was performed using BH method.

F I G U R E 4
Increased protein production and secretion in GHR-KO adenohypophyses.Bubble plot of differentially abundant proteins that are involved in the following processes: vesicle cargo loading, response to ER stress, protein localization to ER, protein folding and cytoplasmic translation.The color of the bubble corresponds to the log2-fold change of protein (red downregulation, blue upregulation) and the size of the bubble indicates the significance of the protein change (bottom).ER, endoplasmic reticulum.The top part of this figure was generated using BioRender.