Glutamine metabolism in the proliferation of GS-expression pituitary tumor cells

in Endocrine Connections

Correspondence should be addressed to H Yang: huiyangxinqiao@163.com

Objective

Many cancer cells cannot survive without exogenous glutamine (Gln); however, cancer cells expressing glutamine synthetase (GS) do not have this restriction. Previous metabolomics studies have indicated that glutamine metabolism is altered during pituitary tumorigenesis. However, the main role of Gln in pituitary adenoma (PA) pathophysiology remains unknown. The aim of this study was to evaluate the expression of GS and the main role of Gln in human PAs.

Methods

We used cell proliferation assay and flow cytometry to assess the effect of Gln depletion on three different pituitary cell lines and human primary PA cells. We then investigated the expression level of Gln synthetase (GS) in 24 human PA samples. At last, we used LC-MS/MS to identify the differences in metabolites of PA cells after the blockage of both endogenous and exogenous Gln.

Results

PA cell lines showed different sensitivities to Gln starvation, and the sensitivity is correlated with GS expression level. GS expressed in 21 out of the 24 human PA samples. Furthermore, a positive p53 and ki-67 index was correlated with a higher GS expression level (P < 0.05). Removal of both endogenous and exogenous Gln from GS-expressing PA cells resulted in blockage of nucleotide metabolism and cell cycle arrest.

Conclusions

Our data indicate that GS is needed for PA cells to undergo proliferation during Gln deprivation, and most human PA cells express GS and might have a negative response to exogenous Gln depletion. Moreover, Gln is mainly responsible for nucleotide metabolism in the proliferation of GS-expressing pituitary tumor cells.

Abstract

Objective

Many cancer cells cannot survive without exogenous glutamine (Gln); however, cancer cells expressing glutamine synthetase (GS) do not have this restriction. Previous metabolomics studies have indicated that glutamine metabolism is altered during pituitary tumorigenesis. However, the main role of Gln in pituitary adenoma (PA) pathophysiology remains unknown. The aim of this study was to evaluate the expression of GS and the main role of Gln in human PAs.

Methods

We used cell proliferation assay and flow cytometry to assess the effect of Gln depletion on three different pituitary cell lines and human primary PA cells. We then investigated the expression level of Gln synthetase (GS) in 24 human PA samples. At last, we used LC-MS/MS to identify the differences in metabolites of PA cells after the blockage of both endogenous and exogenous Gln.

Results

PA cell lines showed different sensitivities to Gln starvation, and the sensitivity is correlated with GS expression level. GS expressed in 21 out of the 24 human PA samples. Furthermore, a positive p53 and ki-67 index was correlated with a higher GS expression level (P < 0.05). Removal of both endogenous and exogenous Gln from GS-expressing PA cells resulted in blockage of nucleotide metabolism and cell cycle arrest.

Conclusions

Our data indicate that GS is needed for PA cells to undergo proliferation during Gln deprivation, and most human PA cells express GS and might have a negative response to exogenous Gln depletion. Moreover, Gln is mainly responsible for nucleotide metabolism in the proliferation of GS-expressing pituitary tumor cells.

Background

Pituitary adenoma (PA) is the second most common brain tumor. Except for prolactinoma, pituitary tumors associated with acromegaly (growth hormone-secreting) and Cushing’s disease (adrenocorticotropic hormone-secreting) as well as clinical nonfunctioning pituitary adenomas (NFPAs) are generally treated with transsphenoidal surgery as the first-line therapy. However, since some PAs invade the cavernous sinus or surrounding bone in the sellar region, it is sometimes difficult to achieve total surgical resection (1). Moreover, the currently available pharmacotherapies are not satisfactory. Thus, there is a need to explore novel approaches for treating PA.

Metabolic abnormalities, including the Warburg effect and glutamine (Gln) addiction, have been considered as a hallmark of cancer cells (2). The Warburg effect refers to the observation that cancer cells tend to favor anaerobic glycolysis even when oxygen is abundant. Despite PA being a benign tumor, we previously reported that lactate dehydrogenase A (LDHA), a key glycolysis enzyme, was involved in promoting its progression. Therefore, LDHA might be a promising therapeutic target for treating PA (3).

Gln is the most abundant amino acid in the human body and is a non-essential amino acid in the physiological state as it is endogenously synthesized. However, most cancer cells die when they lack exogenous Gln, a phenomenon termed as ‘Gln addiction’ (4). Gln constitutes a major metabolic fuel for some types of tumor cells and is a precursor for various biological macromolecules, such as amino acids, proteins, lipid, and nucleotides. Therefore, it is not surprising that Gln depletion induces severe metabolic stress that may lead cells to undergo apoptosis (5, 6). Therapeutic agents targeting Gln transporters or glutaminase have been shown to be effective (7, 8, 9, 10).

However, not all cancer types rely on exogenous Gln to survive, such as those that express Gln synthetase (GS). The gene GLUL codes for GS, which is an enzyme that catalyzes Gln synthesis from glutamate (Glu) and NH4+. The presence or absence of GS expression is an important survival factor after exogenous Gln depletion. Therefore, cancers such as luminal breast cancer (11), MET-induced liver cancer (12), glioma (13), and low invasive ovarian cancer (14) are insensitive to Gln deprivation owing to the high GS expression.

The glutamine- or glutamate-related pathways have also been reported to be altered in pituitary gland tumorigenesis using mass spectrometry or NMR spectrometry. Rahmi et al. compared blood samples from the bilateral inferior petrosal sinus in corticotropinoma patients and healthy controls. They found that Glu metabolism was the most affected metabolic pathway (15); however, they did not show how the Glu metabolism pathway was affected. Similarly, Oklu et al. also observed that Glu and Gln were up-regulated in prolactinoma tissues compared with gonadotropin-secreting PAs (16). These data indicate that Glu/Gln metabolism is altered in pituitary tumorigenesis and that glutamine/glutamate metabolism processes may differ for the different clinical types of PA. However, the specific role of Gln in the pathophysiology of PA is still unknown. Hence, in this study, we aimed to determine the main role of Gln in human PAs as well as GS expression.

Materials and methods

Cell culture and transfection

Rat GH3 (ATCC, CCL-82.1) and MMQ (ATCC, CRL-10609) cells as well as mouse AtT-20 (ATCC, CCL-89) PA cells were all cultured in Ham’s F-12K medium (GIBCO, 21127-022, with 2 mM Gln) containing 2.5% fetal bovine serum (GIBCO, 10100-147) and 15% horse serum (GIBCO, 16050-122) in a humidified atmosphere of 5% CO2 at 37°C. For Gln-free cell culture and certain experiments, custom-made Gln-free Ham’s F-12K medium (GIBCO, with 0 mM Gln) was used to dialyze fetal bovine serum and horse serum. Thereafter, 2.5% dialyzed fetal bovine serum and 15% horse serum were added to Gln-free Ham’s F-12K medium.

Cells were transfected with the negative control or GS siRNA (Gema, Shanghai, China) using Lipofectamine 2000 (Invitrogen Life Technologies, 11668-027) according to the manufacturer’s instructions.

The GS siRNA sequence was ‘GCATCAAGCAGATGTACAT’.

Cell proliferation assay

We used the WST-8 Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Mashiki-machi, Kumamoto, Japan) to measure cell proliferation based on the manufacturer’s instructions. The doubling time (DT) was obtained using the exponential growth equation with GraphPad Prism 6.0. The percentage of growth inhibition was calculated according to the following equation: (1 − (DT + Gln)/(DT − Gln)) × 100.

Colony-forming assay

GH3 cells were plated at 200 cells/well in 6-well plates in complete F-12K, in the absence or presence of Gln or 1 U/mL L-asparaginase (Sigma, A3809) and 1 mM L-methionine sulfoximine (Sigma, M5379) as indicated. After 14 days, cells were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. Visible colonies were manually counted.

Patients

We obtained 24 human pituitary tumor samples (five somatotroph adenomas, five prolactinomas, five corticotropinomas, and nine nonfunctional adenomas) in patients who underwent transsphenoidal surgery at the multidisciplinary center for pituitary adenomas in the Chongqing of the Xinqiao Hospital. Individual tumors were diagnosed according to clinical and endocrine assessments with additional information obtained by pathological evaluation. Tumor invasion was determined based on a preoperative radiological investigation using Knosp’s classification (14) combined with intraoperative findings. Based on the Knosp’s classification system, noninvasive adenomas were grade 0–2 and invasive adenomas were grades 3–4. Clinical data regarding sex, age, Knosp grade, ki67 index, p53, and hormonal type are summarized in Supplementary Table 1 (see section on supplementary materials given at the end of this article).

Immunohistochemistry

Tumor tissue was fixed in 4% paraformaldehyde for 24 h and embedded in paraffin. Embedded tumor tissues were sectioned at 5 µm for subsequent immunohistochemistry (IHC) staining. IHC staining was performed according to standard protocols using Anti-GS (1:50) antibodies.

The immunoreactivity (IR) of GS was evaluated as previously reported (17, 18). We used a Leica DM IRB microscope with a square grid inserted into the eyepiece to examine an area of 781.250 µm2 (200 high power nonoverlapping fields with widths of 0.0625 × 0.0625 mm). The staining intensity was assessed using a semiquantitative three-point scale within the scope of the IR and was defined as absent (−, 1), weak (+, 2), moderate (++, 3), or strong (+++, 4). These scores represented the predominant staining intensity in each section and were calculated as an average of the selected samples.

Primary culture of human PA cells

We obtained six primary human PA cells from harvested human PA specimens from patients undergoing transsphenoidal surgery at the Xinqiao Hospital in Chongqing, China (Supplementary Table 2). Fresh tissues were washed with 1× PBS and minced into small pieces. Tissue fragments were then digested with type I collagenase for 2 hours at 37°C. An equal volume of MEM (GIBCO, 11095-080) supplemented with 10% fetal bovine serum was added, and the cell suspension was filtered through a 200-µm filter. After two washes with PBS, the cells were resuspended in complete MEM and cultured in a humidified atmosphere of 5% CO2 at 37°C.

Western blot analysis

We used antibodies against the following proteins: GS (Abcam, ab176562), β-tubulin (Abcam, ab75991), Cyclin D1 (Cell Signaling Technology, 2978), Cyclin A2 (Beyotime, Shanghai, China, AF2524), and Cyclin E1 (Beyotime, AF2491).

Cell extracts equal to 40 µg protein were subjected to SDS-PAGE and then transferred onto polyvinylidene difluoride membranes. The membranes were blocked for 2 h at room temperature with 5% skimmed milk in Tris-buffered saline containing 0.05% Tween 20, incubated with rabbit antibodies against rat GS (1:1000), β-tubulin (1:10,000), Cyclin D1 (1:1000), Cyclin A2 (1:1000), and Cyclin E1 (1:1000).

Flow cytometry

For cell cycle analysis, cells were collected at the indicated time points for the relevant treatments and fixed in 70% ethanol at 4°C overnight. Cells were resuspended and stained in 1 mL PBS containing 50 μg/mL PI (BD Biosciences Pharmingen, 556463) and then analyzed using flow cytometry (FACScould; BD Biosciences Pharmingen). For cell apoptosis analysis, cells were collected at the indicated time points for the relevant treatments. Apoptosis was assessed using a FITC-Annexin V Apoptosis Detection Kit (BD Biosciences Pharmingen, 556547) based on the manufacturer’s instructions, and cells were detected using flow cytometry.

Liquid chromatograph mass spectrometer (LC-MS) analysis for metabolites

Cells were seeded in 10-cm dishes. After 24 h growth, the medium was substituted with fresh medium in the presence (2 mM) or absence of Gln combined with MSO. After the stipulated experimental times, cells were washed with ice-cold PBS, and metabolites were extracted using 1 mL cold methanol acetonitrile-water solution (v/v, 2:2:1). Ultrasonic crushing was performed at low temperature, and the protein was incubated at −20°C for 1 h for precipitation at 13,000 g. The protein was then centrifuged at 4°C for 15 min, and the supernatant was removed, freeze-dried, and stored at −80°C for later use. Next, we added 100 µL acetonitrile for mass spectrometry analysis. The samples were re-dissolved in an aqueous solution (acetonitrile:water = 1:1, v/v), vortexed, centrifuged at 14,000 g at 4°C for 15 min, and then the liquid supernatant was removed for analysis. We subsequently separated the obtained samples using Agilent 1260 HPLC system. Agilent 6460 QqQ mass spectrometer (Agilent Technologies) was used and mass spectrometry analysis was performed as previously described (19).

Statistical analysis

The data were expressed as means ± s.e.m. The correlations between the GS levels and PA clinical characteristics were determined using the chi-square test. Additionally, we used the two-tailed Student’s t-test to determine the statistical significance of between-group differences. Data analyses were conducted using SPSS for Windows version 13.0 (SPSS Inc). P values less than 0.05 were considered statistically significant.

Results

PA cell lines showed different sensitivities to Gln starvation

To explore the response of PA cell lines to Gln starvation, we used Gln lacking F-12K medium, and the serum was dialyzed to remove Gln. Compared with the normal control, Gln withdrawal showed no significant effect on proliferation of GH3 cells; however, it inhibited the proliferation of MMQ and AtT20 cells at 64% and 20%, respectively (Fig. 1A and B). Flow cytometric apoptosis assay revealed that Gln withdrawal induced apoptosis in MMQ cells but had no significant effect on GH3 and AtT20 cells (Fig. 1C).

Figure 1
Figure 1

PA cell lines showed different sensitivity to Gln starvation. (A) GH3, MMQ, and AtT20 cell proliferation with/without Gln were tested by the CCK-8 assay (n = 5, Mean ± s.e.m.). (B) Growth inhibition caused by Gln starvation. (C) Apoptosis of cell lines incubated for 48 h with/without Gln. (n = 3, Mean ± s.e.m.). (D) Cells were incubated for 48 h in medium with/without L-asparaginase (ASNase). OD values were tested by CCK-8 assay (n = 5, Mean ± s.e.m.). Gln (+): containing 2 mM Gln; Gln (−): containing 0 mM Gln; ns: no significant. ***P < 0.001.

Citation: Endocrine Connections 9, 3; 10.1530/EC-19-0515

We also used L-asparaginase (ASNase), which can simultaneously decompose both asparagine (Asn) and Gln, and found that the proportion of cells that it inhibited was similar to the proportion of cells inhibited in a Gln-lacking medium (Fig. 1D). These results indicated that the response of PA cells to exogenous Gln starvation is heterogeneous.

Sensitivity to Gln starvation is correlated with GS expression level

GS localizes in the cytoplasm and catalyzes Gln synthesis endogenously. To determine the mechanism underlying the heterogeneity of the response of PA cells to exogenous Gln starvation, we first assessed the expression of GS in GH3, MMQ, and AtT20 cell lines. As expected, GS was significantly higher expressed in GH3 cells compared to AtT20 or MMQ cells (P < 0.001). Furthermore, Gln starvation significantly increased GS expression in GH3 (P < 0.001) and AtT20 (P < 0.05) cells but showed no significant effect on MMQ cells (Fig. 2A). Then we tested primary cells derived from three human NFPA patients (case 1–3) and three human growth hormone (GH) secreting PA patients (case 4–6). Five out of the six patients’ primary cells also significantly increased GS expression levels under Gln starvation (Fig. 2B).

Figure 2
Figure 2

Sensitivity to Gln starvation is correlated with GS expression level. (A) GH3, MMQ, and AtT20 Cells were incubated for 48 h with/without Gln and GS protein expression levels were examined by Western blot. (B) Primary PA cells derived from 3NFPAs (case 1–3) and 3 GH-secreting PA patients (case 4–6) were incubated for 48 h with/without Gln and GS protein expression levels were examined by Western blot. (C) GH3 cells were transfected with control or GS siRNA for 24 h before the medium was changed to with/without Gln, followed by cell proliferation assay. (D) GH3 cells were incubated for 48 h in medium with different concentrations of ASNase alone or in combination with L-methionine sulfoximine (MSO). (E) Growth curve of GH3, MMQ, and AtT20 cells as indicated. (F&G) Cells were incubated for 14 days with/without Gln in medium supplemented with 1 mM ASNase or/and 1 mM MSO as indicated. Colonies obtained in representative wells are shown (the best result of three repetitions). ns: no significant; Gln (+): containing 2 mM Gln; Gln (−): containing 0 mM Gln; *P < 0.05; ***P < 0.001.

Citation: Endocrine Connections 9, 3; 10.1530/EC-19-0515

Next, we used GS-siRNA to perform a knockdown of GS expression in GH3 cells. Compared with controls, GS knockdown cells were more sensitive to exogenous Gln starvation, but no significant effect was observed in the presence of exogenous Gln (Fig. 2C). L-methionine sulfoximine (MSO), an irreversible inhibitor of GS, also sensitized GH3 cells to Gln starvation. MSO alone also had no effect on cell proliferation in the presence of exogenous Gln. A combination of ASNase and MSO significantly suppressed the proliferation of all PA cell lines (Fig. 2E). The colony formation assay also yielded similar results (Fig. 2F and G). Furthermore, Asn depletion alone had no effect on GH3 cell proliferation, and Asn supplementation could not rescue GH3 cell proliferation under Gln starvation (Fig. 2D). These data indicate that GS expression determines the sensitivity of PA cells to exogenous Gln depletion.

GS expression in human PA tissues

Since GH3, MMQ, and AtT20 cells secrete GH, prolactin (PRL), and adrenocorticotropic hormone (ACTH), respectively, we sought to determine whether GS was differentially expressed among the various types of human hormone-secreting PAs. Thus, we analyzed GS expression of 24 human PA tissues using IHC. In total, four hormone types of PAs were included: five GH-secreting PAs, five ACTH-secreting PAs, five PRL-secreting PAs, and nine NFPAs. We divided the observed expression into four grades from low to high (score 1–4) (Fig. 3A). Results from IHC staining indicated that GS was expressed in the vast majority (21/24) of patients and was only absent in three patients (Fig. 3B). We further analyzed the relationship between GS expression level and patients’ age, sex, tumor size, hormone type, cavernous sinus invasion, and ki-67 index (Table 1). The results showed that a positive ki-67 index (ki-67 >3%) and positive p53 both correlated with a higher GS expression (P < 0.05) (Fig. 3C and D). Comparison among the four hormone types showed that GH-secreting PAs had the highest GS expression level, consistent with that observed in the cell lines (Fig. 3E).

Figure 3
Figure 3

GS expression in human PA tissues. (A) The expression of GS protein in human PA samples was examined by IHC staining. The staining intensity was defined as absent (1), weak (2), moderate (3), or strong (4), original magnification (×400). Scale bar represents 50 μm. (B) The staining degrees of 24 PA samples are shown in a pie chart. (C, D, and E) Scatter diagram showed the correlation between immunoreactivity scores and ki-67 index, p53 and hormone subtypes. *P < 0.05.

Citation: Endocrine Connections 9, 3; 10.1530/EC-19-0515

Table 1

Demographic and clinical characteristics of patients’ samples used for IHC. ACTH, adrenocorticotropic hormone; GH, growth hormone; NF, nonfunctioning; PRL, prolactin.

CharacteristicsNo. (%) of patientsP
Sex0.094
 Male12 (50.0)
 Female12 (50.0)
Age
 Median 44 years (19–64 years)0.563
 <4011 (45.8)
 ≥4013 (54.2)
Hormone type0.087
 NF9 (37.5)
 ACTH5 (20.8)
 PRL5 (20.8)
 GH5 (20.8)
Cavernous sinus Invasion0.185
 Yes7 (29.2)
 No17 (70.8)
Tumor Size0.196
 Microadenoma4 (16.7)
 Macroadenoma20 (83.3)
Ki-67<0.05
 ≥3%8 (33.3)
 <3%16 (66.7)
P53<0.05
 Positive3 (12.5)
 Negative21 (87.5)

Gln deprivation induced cell cycle arrest but not cell death in GS-expressing PA cell

Previous IHC results have indicated that most human PAs expressed GS. Among the three cell lines, GH3 cells represented the most significant GS expression (Fig. 2A), therefore, we used GH3 cells to further explore the main function of Gln in PA cells. GH3 cells were still in the adherent state after 3 days of both endogenous and exogenous Gln deprivation (Fig. 4A). In addition, GH3 cell proliferation was rescued after Gln addition to the medium after 3 days of exogenous and endogenous Gln deprivation (Fig. 4A and B). Furthermore, the CCK-8 assay revealed that exogenous and endogenous Gln deprivation did not influence cell survival in five of six human primary PA cells (Supplementary Fig. 1). These phenomena indicated that exogenous and endogenous Gln deprivation only inhibited the cell cycle but did not induce cell death. To confirm this hypothesis, we analyzed the cell cycle distributions and apoptosis using Annexin V flow cytometry. We observed no significant difference in the percentage of apoptotic cells (Fig. 4C). However, the cell cycle distribution showed that Gln deprivation arrested cell cycle progression at the S phase (Fig. 4D). EDU-labeled cells were also remarkably reduced in the Gln deprivation group (Fig. 4E). Cyclins are a family of proteins containing the conserved cyclin box, which mediate binding to cyclin-dependent kinases and are involved in cell cycle control. Cyclin D controls the G1 phase whereas cyclin A and cyclin E drives G1/S phase transition (20). A combination of ASNase and MSO significantly decreased cyclin A, cyclin D1, and cyclin E levels after 48 h (Fig. 4F).

Figure 4
Figure 4

Gln deprivation induced cell cycle arrest but not cell death. Representative microscopic fields (A) and CCK-8 (B) assay of cells incubated for 3 days with 1 mM ASNase alone or in combination with MSO, then both replaced by Gln (+) medium. Scale bar represents 100 μm. Apoptosis (C) and cell cycle distributions (D) for GH3 cells incubated in medium containing 1 mM ASNase alone or in combination with MSO were analyzed by Annexin V flow cytometry. (E) GH3 cells labeled with EDU. Representative confocal microscopy images. Scale bar represents 100 μm. (F) For protein expression analysis, we collected GH3 cells grown in media containing 1 mM ASNase alone or in combination with MSO at appropriate time points and subsequently performed Western blotting using appropriate antibodies. ns: no significant. *P < 0.05; **P < 0.01; ***P < 0.001.

Citation: Endocrine Connections 9, 3; 10.1530/EC-19-0515

Gln deprivation affected nucleotide metabolism in GS-expressing PA cell

To further determine the key metabolic role of Gln in GS-expressing cells, we used LC-MS/MS to identify the differences in metabolites. The most changed metabolites are shown in the heat map (Fig. 5A). We observed that the levels of tricarboxylic acid (TCA) cycle intermediates derived from Gln anaplerosis, such as α-KG and its downstream metabolites succinate and fumarate, remain unchanged under both endogenous and exogenous Gln deprivation (Fig. 5B). However, as observed for dCTP, dCMP, CTP, ATP, and guanosine diphosphate (GDP), we found a significant decrease in the intracellular level of inosine monophosphate (IMP), the key metabolite of the nucleotide de novo synthesize pathway, under Gln deprivation (Fig. 5C), indicating a blockage of the nucleotide de novo synthesis pathway. Conversely, we observed a significant increase in the intracellular levels of inosine, guanosine, cytidine, and uridine (Supplementary Fig. 2), indicating a blockage of the nucleotide salvage pathway. Pathway enrichment of the changed metabolites also indicated significant changes in the purine and pyrimidine metabolic pathways (Supplementary Fig. 3).

Figure 5
Figure 5

Metabolomics analysis of GH3 cells cultured in a medium containing Gln (N, with 2 mM Gln) vs in a medium lacking Gln (0 mM) but containing 1 mM MSO (W). (A) Heat map of the most changed metabolites between N and W. (B) The main metabolites involved in Gln-related anaplerosis in the two groups. (C) Key metabolites involved in nucleotide metabolism in the two groups.

Citation: Endocrine Connections 9, 3; 10.1530/EC-19-0515

Our previous results have confirmed that the S phase of cell cycle was blocked when glutamine was depleted (Fig. 4D and F). DNA replicates among S phase, and this process need a lot of nucleotides. We checked the levels of the deoxynucleotides in our data, which were the direct synthetic material of DNA. As expected, we found that all the detected deoxynucleotides were decreased, including dCTP, dCMP, deoxyguanosine, and deoxycytidine. We then tried to use deoxynucleotides to rescue cell proliferation. The results showed that dCTP and dUTP could partly rescue GH3 cell proliferation in W condition. Besides, L-glutamine is a coding amino acid in protein synthesis. However, we found that cycloheximide (CHX), an eukaryote protein synthesis inhibitor (4), did not affect the rescue effect of dCTP and dUTP (Supplementary Fig. 4A), suggesting that protein synthesis blockage is not the main reason for the suppression of cell proliferation. Furthermore, we noticed that after Gln deprivation for 72 h, the DNA concentration of GH3 cells significantly increased after readdition of glutamine (Supplementary Fig. 4B). Together, the results indicate that glutamine is mainly used to support nucleotides synthesis in the proliferation of GS-expressing pituitary tumor cells.

Discussion

The renewed attention to the metabolic characteristics of cancer cells and the recognition that changes in cell metabolism are a critical hallmark of cancer that have stimulated the exploration of targeting the metabolic specificity of cancer cells as a new therapeutic strategy. The important role of Gln has been confirmed in many cancers; however, little attention has been paid to the potential role of Gln metabolism in the progression of PA. To the best of our knowledge, this is the first study to demonstrate the effects of Gln depletion on PA cells. We show that Gln depletion may lead to increased GS expression in most PA cells, and the sensitivity to Gln depletion depends on the presence or absence of GS expression. We also observed GS expression for most human PA samples and that the positive p53 and ki-67 index was correlated with higher GS expression. The proliferation of GS-expressing cell lines was completely suppressed after application of a combination of ASNase (a glutamine scavenger) and MSO (an irreversible GS inhibitor) due to blocking of nucleotide metabolism and cell cycle arrest.

Gln depletion induced apoptosis and suppressed cell growth in MMQ cells but had no significant effect on GH3 and AtT20 cells. These data indicate heterogeneity in the endogenous glutamine synthesis among different tumor types. The differences in the GS expression levels, especially the degree of compensatory enhancement after Gln depletion, explained this heterogeneity well. GS expression level has been shown to be negatively correlated with Gln depletion sensitivity in many cancer types (11, 12, 14). Pharmacological or siRNA-mediated GS inhibition significantly suppressed proliferation of glutamine-deprived cells but had no effect on cells grown under normal culture conditions. Taken together, these data indicate that GS is needed for PA cells to undergo proliferation during Gln deprivation.

To determine the GS expression levels in human PAs, we analyzed 24 human PA samples using IHC. The results indicated a correlation between a high ki-67 index and a high GS expression level. High GS expression has been reported to correlate with high aggressiveness and worse survival in glioblastoma (13, 21). We also noted that the GH subtype had the highest GS expression level among the four main subtypes of PA, which was consistent with the result that GH3, a GH-secreting cell line, had the highest GS expression level among the three cell lines. However, different to PRL-secreting cell line MMQ, human prolactinomas did not show a very low GS expression level. One possible explanation is that most prolactinomas were treated using drugs instead of surgery, and so the data of prolactinomas included in the present study did not represent the commonly observed prolactinomas. More importantly, we observed that irrespective of expression level, most PAs expressed GS, indicating that the majority of PAs might be insensitive to Gln depletion. Further, we tested and confirmed this phenomenon in five out of six primary PA cells. Despite cancer cells themselves, stromal-epithelial cells or astrocytes may express high levels of GS and synthesize Gln to support cancer cells regardless of whether they express GS (13, 22). Therefore, it may be necessary to include a GS inhibitor. However, MSO, which is the best-characterized GS inhibitor, has been largely avoided as a clinical cancer therapeutic agent due to CNS toxicity (23). Further examination of strategies to limit CNS toxicity by MSO and GS inhibition are needed, such as combination therapy, to lower the necessary dose or tumor-specific drug delivery. One such strategy involving the combination of MSO and L-asparaginase showed antitumor activity in a preclinical hepatocellular carcinoma xenografts model (24).

Gln involved in multiple metabolic pathways supports cell bioenergetics and biosynthesis. In several human cancer models, Gln has been shown to sustain cell growth through anaplerosis (25, 26, 27). Furthermore, Gln is also involved in many other pathways. In glioma cells, Gln fuels de novo purine biosynthesis rather than anaplerosis (13). However, Gln is necessary for sarcoma cells to support both optimal mitochondrial bioenergetics and nucleotide synthesis (28). Additionally, exogenous Gln is needed in oligodendroglioma cells for cell proliferation independent of anaplerosis or glutathione biosynthesis (29). We found that simultaneously blocking endogenous and exogenous Gln did not kill GH3 cells and that replenishing exogenous Gln could still restart their proliferation after Gln starving for 72 h. These phenomena indicated that Gln deprivation arrested the cell cycle, but did not induce cell death, which was confirmed by flow cytometry and EDU. Metabolomics analysis and replenishment results added to the evidence that the main role of Gln in the proliferation of GS-expressing cells is nucleotide metabolism rather than anaplerosis or protein translation.

However, the number of human PA samples in the present study is low and in vivo tests are lacking. Therefore, further clinical studies are needed to verify the present results and for further discovery.

Conclusions

In summary, the present study indicated that PA cells require GS to survive during Gln starvation. The majority of human PAs expressed GS, and positive p53 and ki-67 index correlated with higher GS expression. Blocking both endogenous and exogenous Gln in GS-expressing cells induced nucleotide metabolism disturbance and cell cycle arrest. Our results further the understanding of Gln metabolism in PA cells and may provide information for future clinical studies to develop novel therapeutic agents.

Supplementary materials

This is linked to the online version of the paper at https://doi.org/10.1530/EC-19-0515.

Declaration of interest

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

Funding

This work was supported by National Natural Science Foundation of China (No. 81801369) and Foundation of Xinqiao Hospital (No. 2016D417).

Ethics and consent to participate

This study was approved by the Ethics and Clinical Research Committee of Xinqiao Hospital (China). Registration number: AF/SC-08/1.0. All procedures performed in this study involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. We added this paragraph to ‘Ethics and consent to participate’ in the revised manuscript.

Consent for publication

Informed consent was obtained from all patients or their relatives.

Availability of data and materials

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Author contribution statement

J H and S L conceived and designed the research and drafted the manuscript. J H, Q C, X D and X Z performed the experiments, analysis, and interpretation of data. X T and J H evaluated the IHC score. H Y critically reviewed and edited the manuscript. All authors have read and approved the final manuscript.

References

  • 1

    MolitchME. Diagnosis and treatment of pituitary adenomas: a review. JAMA 2017 317 516524. (https://doi.org/10.1001/jama.2016.19699)

  • 2

    HanahanDWeinbergRA. Hallmarks of cancer: the next generation. Cell 2011 144 646674. (https://doi.org/10.1016/j.cell.2011.02.013)

  • 3

    AnJZhangYHeJZangZZhouZPeiXZhengXZhangWYangHLiS. Lactate dehydrogenase A promotes the invasion and proliferation of pituitary adenoma. Scientific Reports 2017 7 4734. (https://doi.org/10.1038/s41598-017-04366-5)

    • Search Google Scholar
    • Export Citation
  • 4

    YangLVennetiSNagrathD. Glutaminolysis: a hallmark of cancer metabolism. Annual Review of Biomedical Engineering 2017 19 163194. (https://doi.org/10.1146/annurev-bioeng-071516-044546)

    • Search Google Scholar
    • Export Citation
  • 5

    WillemsLJacqueNJacquelANeveuxNMacielTTLambertMSchmittAPoulainLGreenASUzunovMet al. Inhibiting glutamine uptake represents an attractive new strategy for treating acute myeloid leukemia. Blood 2013 122 35213532. (https://doi.org/10.1182/blood-2013-03-493163)

    • Search Google Scholar
    • Export Citation
  • 6

    AltmanBJStineZEDangCV. From Krebs to clinic: glutamine metabolism to cancer therapy. Nature Reviews: Cancer 2016 16 619634. (https://doi.org/10.1038/nrc.2016.71)

    • Search Google Scholar
    • Export Citation
  • 7

    WangQBeaumontKAOtteNJFontJBaileyCGvan GeldermalsenMSharpDMTiffenJCRyanRMJormakkaMet al. Targeting glutamine transport to suppress melanoma cell growth. International Journal of Cancer 2014 135 10601071. (https://doi.org/10.1002/ijc.28749)

    • Search Google Scholar
    • Export Citation
  • 8

    Osanai-SasakawaAHosomiKSumitomoYTakizawaTTomura-SurukiSImaizumiMKasaiNPohTWYamanoKYongWPet al. An anti-ASCT2 monoclonal antibody suppresses gastric cancer growth by inducing oxidative stress and antibody dependent cellular toxicity in preclinical models. American Journal of Cancer Research 2018 8 14991513.

    • Search Google Scholar
    • Export Citation
  • 9

    GrossMIDemoSDDennisonJBChenLChernov-RoganTGoyalBJanesJRLaidigGJLewisERLiJet al. Antitumor activity of the glutaminase inhibitor CB-839 in triple-negative breast cancer. Molecular Cancer Therapeutics 2014 13 890901. (https://doi.org/10.1158/1535-7163.MCT-13-0870)

    • Search Google Scholar
    • Export Citation
  • 10

    ZhuMFangJZhangJZhangZXieJYuYRuanJJChenZHouWYangGet al. Biomolecular interaction assays identified dual inhibitors of glutaminase and glutamate dehydrogenase that disrupt mitochondrial function and prevent growth of cancer cells. Analytical Chemistry 2017 89 16891696. (https://doi.org/10.1021/acs.analchem.6b03849)

    • Search Google Scholar
    • Export Citation
  • 11

    KungHNMarksJRChiJT. Glutamine synthetase is a genetic determinant of cell type-specific glutamine independence in breast epithelia. PLoS Genetics 2011 7 e1002229. (https://doi.org/10.1371/journal.pgen.1002229)

    • Search Google Scholar
    • Export Citation
  • 12

    YunevaMOFanTWAllenTDHigashiRMFerrarisDVTsukamotoTMatesJMAlonsoFJWangCSeoYet al. The metabolic profile of tumors depends on both the responsible genetic lesion and tissue type. Cell Metabolism 2012 15 157170. (https://doi.org/10.1016/j.cmet.2011.12.015)

    • Search Google Scholar
    • Export Citation
  • 13

    TarditoSOudinAAhmedSUFackFKeunenOZhengLMileticHSakariassenWeinstockAWagnerAet al. Glutamine synthetase activity fuels nucleotide biosynthesis and supports growth of glutamine-restricted glioblastoma. Nature Cell Biology 2015 17 15561568. (https://doi.org/10.1038/ncb3272)

    • Search Google Scholar
    • Export Citation
  • 14

    YangLMossTMangalaLSMariniJZhaoHWahligSArmaiz-PenaGJiangDAchrejaAWinJet al. Metabolic shifts toward glutamine regulate tumor growth, invasion and bioenergetics in ovarian cancer. Molecular Systems Biology 2014 10 728. (https://doi.org/10.1002/msb.20134892)

    • Search Google Scholar
    • Export Citation
  • 15

    OkluRDeipolyiARWickySErgulEDeikAAChenJWHirschJAWojtkiewiczGRClishCB. Identification of small compound biomarkers of pituitary adenoma: a bilateral inferior petrosal sinus sampling study. Journal of NeuroInterventional Surgery 2014 6 541546. (https://doi.org/10.1136/neurintsurg-2013-010821)

    • Search Google Scholar
    • Export Citation
  • 16

    IjareOBBaskinDSPichumaniK. Ex vivo (1)H NMR study of pituitary adenomas to differentiate various immunohistochemical subtypes. Scientific Reports 2019 9 3007. (https://doi.org/10.1038/s41598-019-38542-6)

    • Search Google Scholar
    • Export Citation
  • 17

    ZangZLiSZhangWChenXZhengDShuHGuoWZhaoBShenKWeiYet al. Expression patterns of TRPC1 in cortical lesions from patients with focal cortical dysplasia. Journal of Molecular Neuroscience 2015 57 265272. (https://doi.org/10.1007/s12031-015-0615-5)

    • Search Google Scholar
    • Export Citation
  • 18

    YueJLiWLiangCChenBChenXWangLZangZYuSLiuSLiSet al. Activation of LILRB2 signal pathway in temporal lobe epilepsy patients and in a pilocarpine induced epilepsy model. Experimental Neurology 2016 285 5160. (https://doi.org/10.1016/j.expneurol.2016.09.006)

    • Search Google Scholar
    • Export Citation
  • 19

    LinRMoYZhaHQuZXiePZhuZJXuYXiongYGuanKL. CLOCK acetylates ASS1 to drive circadian rhythm of ureagenesis. Molecular Cell 2017 68 198.e6209.e6. (https://doi.org/10.1016/j.molcel.2017.09.008)

    • Search Google Scholar
    • Export Citation
  • 20

    LimSKaldisP. Cdks, cyclins and CKIs: roles beyond cell cycle regulation. Development 2013 140 30793093. (https://doi.org/10.1242/dev.091744)

    • Search Google Scholar
    • Export Citation
  • 21

    RosatiAPolianiPLTodeschiniACominelliMMedicinaDCenzatoMSimonciniELMagriniSMBuglioneMGrisantiSet al. Glutamine synthetase expression as a valuable marker of epilepsy and longer survival in newly diagnosed glioblastoma multiforme. Neuro-Oncology 2013 15 618625. (https://doi.org/10.1093/neuonc/nos338)

    • Search Google Scholar
    • Export Citation
  • 22

    YangLAchrejaAYeungTLMangalaLSJiangDHanCBaddourJMariniJCNiJNakaharaRet al. Targeting stromal glutamine synthetase in tumors disrupts tumor microenvironment-regulated cancer cell growth. Cell Metabolism 2016 24 685700. (https://doi.org/10.1016/j.cmet.2016.10.011)

    • Search Google Scholar
    • Export Citation
  • 23

    BrusilowWSPetersTJ. Therapeutic effects of methionine sulfoximine in multiple diseases include and extend beyond inhibition of glutamine synthetase. Expert Opinion on Therapeutic Targets 2017 21 461469. (https://doi.org/10.1080/14728222.2017.1303484)

    • Search Google Scholar
    • Export Citation
  • 24

    ChiuMTarditoSPillozziSArcangeliAArmentoAUggeriJMissaleGBianchiMGBarilliADall’AstaVet al. Glutamine depletion by crisantaspase hinders the growth of human hepatocellular carcinoma xenografts. British Journal of Cancer 2014 111 11591167. (https://doi.org/10.1038/bjc.2014.425)

    • Search Google Scholar
    • Export Citation
  • 25

    WiseDRDeBerardinisRJMancusoASayedNZhangXYPfeifferHKNissimIDaikhinEYudkoffMMcMahonSBet al. Myc regulates a transcriptional program that stimulates mitochondrial glutaminolysis and leads to glutamine addiction. PNAS 2008 105 1878218787. (https://doi.org/10.1073/pnas.0810199105)

    • Search Google Scholar
    • Export Citation
  • 26

    WangQHardieRAHoyAJvan GeldermalsenMGaoDFazliLSadowskiMCBalabanSSchreuderMNagarajahRet al. Targeting ASCT2-mediated glutamine uptake blocks prostate cancer growth and tumour development. Journal of Pathology 2015 236 278289. (https://doi.org/10.1002/path.4518)

    • Search Google Scholar
    • Export Citation
  • 27

    van den HeuvelAPJingJWoosterRFBachmanKE. Analysis of glutamine dependency in non-small cell lung cancer: GLS1 splice variant GAC is essential for cancer cell growth. Cancer Biology and Therapy 2012 13 11851194. (https://doi.org/10.4161/cbt.21348)

    • Search Google Scholar
    • Export Citation
  • 28

    IssaqSHMendozaAFoxSDHelmanLJ. Glutamine synthetase is necessary for sarcoma adaptation to glutamine deprivation and tumor growth. Oncogenesis 2019 8 20. (https://doi.org/10.1038/s41389-019-0129-z)

    • Search Google Scholar
    • Export Citation
  • 29

    ChiuMTaurinoGBianchiMGOttavianiLAndreoliRCiociolaTLagrastaCAMTarditoSBussolatiO. Oligodendroglioma cells lack glutamine synthetase and are auxotrophic for glutamine, but do not depend on glutamine anaplerosis for growth. International Journal of Molecular Sciences 2018 19 E1099. (https://doi.org/10.3390/ijms19041099)

    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

 

     European Society of Endocrinology

     Society for Endocrinology

Sept 2018 onwards Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 281 281 118
PDF Downloads 77 77 18
  • View in gallery

    PA cell lines showed different sensitivity to Gln starvation. (A) GH3, MMQ, and AtT20 cell proliferation with/without Gln were tested by the CCK-8 assay (n = 5, Mean ± s.e.m.). (B) Growth inhibition caused by Gln starvation. (C) Apoptosis of cell lines incubated for 48 h with/without Gln. (n = 3, Mean ± s.e.m.). (D) Cells were incubated for 48 h in medium with/without L-asparaginase (ASNase). OD values were tested by CCK-8 assay (n = 5, Mean ± s.e.m.). Gln (+): containing 2 mM Gln; Gln (−): containing 0 mM Gln; ns: no significant. ***P < 0.001.

  • View in gallery

    Sensitivity to Gln starvation is correlated with GS expression level. (A) GH3, MMQ, and AtT20 Cells were incubated for 48 h with/without Gln and GS protein expression levels were examined by Western blot. (B) Primary PA cells derived from 3NFPAs (case 1–3) and 3 GH-secreting PA patients (case 4–6) were incubated for 48 h with/without Gln and GS protein expression levels were examined by Western blot. (C) GH3 cells were transfected with control or GS siRNA for 24 h before the medium was changed to with/without Gln, followed by cell proliferation assay. (D) GH3 cells were incubated for 48 h in medium with different concentrations of ASNase alone or in combination with L-methionine sulfoximine (MSO). (E) Growth curve of GH3, MMQ, and AtT20 cells as indicated. (F&G) Cells were incubated for 14 days with/without Gln in medium supplemented with 1 mM ASNase or/and 1 mM MSO as indicated. Colonies obtained in representative wells are shown (the best result of three repetitions). ns: no significant; Gln (+): containing 2 mM Gln; Gln (−): containing 0 mM Gln; *P < 0.05; ***P < 0.001.

  • View in gallery

    GS expression in human PA tissues. (A) The expression of GS protein in human PA samples was examined by IHC staining. The staining intensity was defined as absent (1), weak (2), moderate (3), or strong (4), original magnification (×400). Scale bar represents 50 μm. (B) The staining degrees of 24 PA samples are shown in a pie chart. (C, D, and E) Scatter diagram showed the correlation between immunoreactivity scores and ki-67 index, p53 and hormone subtypes. *P < 0.05.

  • View in gallery

    Gln deprivation induced cell cycle arrest but not cell death. Representative microscopic fields (A) and CCK-8 (B) assay of cells incubated for 3 days with 1 mM ASNase alone or in combination with MSO, then both replaced by Gln (+) medium. Scale bar represents 100 μm. Apoptosis (C) and cell cycle distributions (D) for GH3 cells incubated in medium containing 1 mM ASNase alone or in combination with MSO were analyzed by Annexin V flow cytometry. (E) GH3 cells labeled with EDU. Representative confocal microscopy images. Scale bar represents 100 μm. (F) For protein expression analysis, we collected GH3 cells grown in media containing 1 mM ASNase alone or in combination with MSO at appropriate time points and subsequently performed Western blotting using appropriate antibodies. ns: no significant. *P < 0.05; **P < 0.01; ***P < 0.001.

  • View in gallery

    Metabolomics analysis of GH3 cells cultured in a medium containing Gln (N, with 2 mM Gln) vs in a medium lacking Gln (0 mM) but containing 1 mM MSO (W). (A) Heat map of the most changed metabolites between N and W. (B) The main metabolites involved in Gln-related anaplerosis in the two groups. (C) Key metabolites involved in nucleotide metabolism in the two groups.