Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 May 6.
Published in final edited form as: Eur J Endocrinol. 2024 Jun 5;190(6):467–478. doi: 10.1093/ejendo/lvae055

Cabergoline Targets Multiple Pathways to Inhibit PRL Secretion and Increases Stromal Fibrosis

Dongyun Zhang 1, Willy Hugo 1, Marvin Bergsneider 2, Marilene B Wang 3, Won Kim 2, Karam Han 4, Harry V Vinters 4, Anthony P Heaney 1,2
PMCID: PMC12054100  NIHMSID: NIHMS2075112  PMID: 38781434

Abstract

Objective:

Unravel potential mechanism(s) of the on- and off-target actions of dopamine agonist therapy in both human prolactinoma tumor and neighboring stromal and immune cells.

Design and methods:

Five surgically resected prolactinomas from 3 cabergoline (CBG)-treated and 2 treatment naive patients were analyzed by single cell RNA sequencing (scRNA-seq) to compare the cellular composition and transcriptional landscape.

Results:

Six major cell populations that included tumor (88.2%), immune (5.6%), stromal (4.9%), progenitor cells (0.6%), proliferating cells (0.4%), and erythrocytes (0.2%) were observed. Tumor cells from CBG-treated patients expressed lower levels of genes that regulated hormone secretion, such as SCG2, VGF, TIMP1, NNAT, and CALD1, consistent with the inhibitory effects of CBG on hormone processing and secretion. Interestingly, we also observed an increased number of CD8+ T cells in the CBG-treated tissues. These cytotoxic CD8+ T cells expressed killing granule components, such as perforin and the granzymes GZMB, GNLY and KLRD1 as well as the inflammatory cytokine CCL5. Immune cell activation of these CD8+ T cells was further analyzed in a compartment-specific manner, and increased CD25 (IL2R) expression was noted in the CD8+ T cells from CBG-treated samples. Additionally, and confirming prior reports, we noted a higher stromal cell population in CBG-treated samples.

Conclusions:

Our scRNAseq studies revealed key differences in the transcriptomic features of CBG-treated and untreated PRLomas in both tumor and microenvironment cellular constituents, and for the first time describe previously unknown activation of CD8+ T cells following CBG-treatment which may play a role in the tumoricidal actions of CBG.

Significance Statement:

Prolactinomas (PRLomas) are comparatively common and make up ~50% of all pituitary adenomas. Dopamine agonist (DA) therapy is an extremely effective first line medical treatment but some patients require long-term DA therapy to maintain normal PRL levels and others are either refractory to, or intolerant of DAs. For the first time, we have used single cell RNA sequencing to compare the cellular composition and transcriptional landscape in surgically resected prolactinomas from 3 patients who had received and were responsive to cabergoline treatment with 2 treatment naïve patients. In new exciting findings, we describe a potential role of cabergoline to modulate immune cell activation.

Keywords: Pituitary tumor, Prolactinomas, Cabergoline, scRNAseq, Tumor Microenvironment

Introduction

Prolactinomas (PRLomas) account for ~40–50% of all pituitary adenomas and frequently cause hypogonadism and subfertility (1). The majority of PRLomas are microadenomas (<10 mm), cause PRL levels typically between 100 to 200 μg/dL and most commonly occur in pre-menopausal women. In PRL-secreting macroadenomas (>10 mm), serum PRL levels are typically >200 μg/dL, and they cause mass effect symptoms including headache, visual disturbances and varying degrees of hypopituitarism (2).

Medical treatment of PRLomas with the dopamine agonists (DAs), bromocriptine (BRC) and cabergoline (CBG), can normalize serum PRL and cause a > 50% tumor volume reduction in 80% of cases. However, ~30% of PRLomas are either biochemically and/or radiologically largely unresponsive to DAs despite escalating DA doses (2). Additionally, some patients although responsive, are intolerant of DAs due to side effects, including headache, nausea, postural hypotension, and psychiatric disturbances (3). In recent years, concern has also arisen regarding the actions of high dose DAs to cause cardiac valve thickening (4). Surgical tumor debulking can be offered to DA-refractory and/or -intolerant patients but complete resection of these often invasive tumors is difficult to achieve (5). Although radiation therapy can be offered and is extremely good at controlling tumor growth, normalization of serum PRL levels is slow (6).

To better understand the actions and targets of DA treatment in PRLomas, we used single cell RNA sequencing (scRNA-seq) to compare cellular composition and cell-type specific transcriptional changes in 5 surgically-resected PRLomas, 3 derived from patients who had received and were responsive to DAs, and 2 derived from DA-treatment naïve patients. In this first scRNA-seq study of PRLomas, we describe the striking cellular and transcriptomic differences in the tumor cells and tumor microenvironment that reflect both the on- and off-target effects of CBG treatment.

Results:

Cabergoline exhibits broad actions on PRLoma hormone secretion

Using a 10x Genomics platform (v3) and Cell Ranger pipeline for alignment and mapping, we studied 5 surgically resected histopathologically confirmed pituitary lactotroph tumors. Two patients had not received any medical treatment, and 3 had been treated with and responded to cabergoline (CBG) for periods between 2–14 months (Table 1). Using Seurat v5 for read pre-filtering, normalization, integration and unbiased clustering (7), we identified 36,256 cells. Based on canonical cell type gene marker expression, these cells were then categorized into 6 major cell populations (Fig. 1A). These included tumor (31,988 cells, 88.2%), immune (2,045 cells, 5.6%), stromal (1,767 cells, 4.9%), progenitor cells (235 cells, 0.6%), proliferating cells (140 cells, 0.4%), and erythrocytes (81 cells, 0.2%, Table 2). Using the “mast” method (FC. threshold = 1.5), we identified a total of 3,949 differentially expressed genes (DEGs) in these 6 cell populations (Supplementary Table 1, p<10−4). The most highly expressed genes associated with each cell population are depicted in the Fig. 1B heatmap. The top DEGs in the tumor cell population were involved in hormone activity (GO:0005179, such as PRL, NTS, and CHGB), peptidase regulatory activity (GO:0061135, such as PCSK1N, and SPINT2), and calcium signaling (GO:0005544, such as PCLO and SYT1, Supplementary Table 2, p<0.05, and Fig. 1B).

Table-1.

Patient Characteristics and Sample Information.

Sample ID Gender Age Tumor Size (mm) & CV invasion (Y/N) Pituitary Function Testing Pathology Ki-67 Labelling Index Serum PRL before treatment CBG Dose mg/week Serum PRL (ng/mL) on Therapy % Tumor Shrinkage on CBG Total Time on CBG (months) Indication for TNTS
PRLoma1 F 22 4×8 (Y) Hypogonadism PRL+, patchy GH 1–3% 1,585 2 10 50% 14 Severe CBG-induced memory loss & depression
PRLoma2 M 32 8×9×7 (N) Hypogonadism PRL+ 1–2% 140 2 36 Too short duration 2 Responsive to CBG but patient declined continuing therapy
PRLoma3 M 28 43 × 36 × 28 (Y) Hypogonadism PRL+ <1% > 2,000 2 131 30% 6 Responsive to CBG but severe OC compression
PRLoma4 F 32 6 × 7 × 6 (N) Hypogonadism PRL+ 2–4% 104 NA NA NA NA Declined medical therapy as first line
PRLoma5 F 42 8 × 4 (N) Hypogonadism PRL+ 3–5% 149 NA NA NA NA Declined medical therapy as first line, conception desired
Open in a new tab

Abbreviations:

PRLoma: prolactinoma. CBG: cabergoline. Prolactin normal range: Female 3–23ng/ml; Male 3–19 ng/ml. TNTS: trans nasal trans sphenoidal surgery. NA: Not applicable

Figure 1. The cellular composition of CBG-treated (n=3) and untreated (n=2) PRLoma tissues. (A).

Figure 1.

Open in a new tab

Uniform manifold approximation and projection (UMAP) was used to depict the 6 distinct cell types (left panel) and their percentages (right panel) in 5 individual PRLomas. (B) Heatmap illustrated the highly expressed genes in each of 6 cell types in PRLomas. (C) Tumor cell population was subset and the differentially expressed genes (DEGs) were analyzed in the CBG-treated and untreated samples. (D & E) Violin plots depicted the DEGs that were downregulated (D) and upregulated (E) by CBG treatment. p<10−4.

Table 2.

The composition of cell populations.

Tumor Immune Stroma Progenitor Proliferating RBCs Total
PRL-1-CBG 2,485 556 380 23 17 8 3,469
PRL-2-CBG 6,678 280 896 39 18 69 7,980
PRL-3-CBG 3,472 121 389 6 - 1 3,989
PRL-4-woCBG 12,077 405 65 39 95 1 12,682
PRL-5-woCBG 7,276 683 37 128 10 2 8,136
Total 31,988 2,045 1,767 235 140 81 36,256
Tumor (%) Immune (%) Stroma (%) Progenitor (%) Proliferating (%) RBCs (%)
PRL-1-CBG 71.6 16.0 11.0 0.7 0.5 0.2
PRL-2-CBG 83.7 3.5 11.2 0.5 0.2 0.9
PRL-3-CBG 87.0 3.0 9.8 0.2 0.0 0.0
PRL-4-woCBG 95.2 3.2 0.5 0.3 0.7 0.0
PRL-5-woCBG 89.4 8.4 0.5 1.6 0.1 0.0
Total 88.2 5.6 4.9 0.6 0.4 0.2
Open in a new tab

We used the subset function of Seurat to further analyze the tumor cell population (Supplementary Table 3, p<10−4, and Fig. 1C). We observed that many of the more highly expressed genes in untreated prolactinomas [(–)CBG] were involved in the regulation of hormone secretion (such as DLK1, SCG2, VGF, TIMP1, NNAT, and CALD1, Supplementary Table 3, p<10−4, Fig. 1D). Underscoring broader actions of CBG to inhibit PRL release, we also noted downregulation of the granin proteins, secretogranin II (SCG5) and VII (VGF, Supplementary Table 3, p<10−4, Fig. 1D), which regulate dense core secretory vesicle processing, transport, storage, and release of peptide hormones (8). Interestingly, we noted that tumor cells from the (+)CBG samples exhibited lower expression of major histocompatibility complex class I (MHC-I) components, such as HLA-A and B2M (Fig. 1D). This suggested that CBG may affect immune response in PRLomas and is supported by our associated finding below that CBG also increases cytotoxic CD8+ T cells. Other genes which were more highly expressed in the (+)CBG tumor cells were related to cell viability and metabolism, such as ARHGAP5, BTG2, BCAT1, DBP, SAMD5, WRN, BTF3, and MEIS2 (p<10−4, Fig. 1E).

CBG treatment results in increased CD8+ T cells in PRLomas

We next analyzed and compared the immune cell populations comprising 2,045 cells, with 957 cells from the (+)CBG and 1,088 from the (−)CBG samples (Fig. 2A). We delineated 8 immune cell clusters (Supplementary Table 4, p<10−4, and Fig. 2B) that included: early activated CD4+ T cells expressing IL7R, and IL32 (514 cells, 42.5%); CD8+ T cells expressing CCL5, NKG7, and GZMA (106 cells, 8.8%); naive and memory B cells (BANK1, CD79B, and TNFRSF13C, 53 cells, 4.4%) (9, 10); antibody secreting B cells (IGHG1, and IGHG4, 41 cells, 3.4%) (11); as well as myelocytes such as macrophage-1 expressing CD14, CLEC12A, FCN1, CSTA, VCAN, and LYZ (178 cells, 14.8%); macrophage-2 expressing APOE, C1QA, SELENOP, MRC1, FCLR2, and MS4A4A (210 cells, 17.4%); dendritic cells expressing NDRG2, RNASE6, HLA-DQA1/B1, HLA-DPA1/B1, and HLA-DRA/B1 (76 cells, 6.3%) (12, 13); and neutrophils expressing IFITM2, NAMPT, and S100A8/A9 (31 cells, 2.6%) (14). A cluster of uncharacterized cells mainly comprised of doublets was also observed but not included in further analysis. Among the 8 clusters of immune cells, the most significant finding was that the proportion of CD8+ T cells was strikingly higher in (+)CBG (13.38 ± 1.96%) compared to (−) CBG samples (4.99 ± 1.69%, p<0.01, Fig. 2B and supplementary Table-5). These cytotoxic CD8+ T cells expressed the killing granule components perforin (PRF1) and the granzymes GZMB, GNLY, KLRD1 as well as the inflammatory cytokine CCL5 (Fig. 2C). Further analysis of the upstream regulators of this pathway indicated activation of IL2, and involvement of the inflammatory transcription factors NFκB and STATs in CBG treated tumor cells (Supplementary Fig. 1A). The marked increase in CD8+ T cells in the (+)CBG samples was confirmed by multiplex immunofluorescent staining of CD8 (FITC/green), CD45 (CY5/yellow) and CD31 (TRITC/red) which was used to detect vascular structures [(+)CBG: 0.71±0.17% vs. (−)CBG: 0.28±0.16%, p<0.05, Fig. 2D]. We then used Nanostring GeoMX Digital Spatial Profiler (DSP) to evaluate 30 compartmentalized regions of interest (ROIs) selected from CD8+ T cells in both treated and untreated samples (example in Fig. 2E) to identify factors or pathways potentially involved in the CBG-associated CD8+ T cell activation. Twenty proteins, including checkpoint molecules and metabolic mediators of immune function, were assessed and in particular, increased CD25 (IL2R) expression was noted in the CD8+ T cells in the (+)CBG samples (p<0.05, Fig. 2E, far-right of figure circled in red).

Figure 2. Transcriptomic features of immune cell population in CBG–treated and untreated PRLomas.

Figure 2.

Open in a new tab

(A) The subset function of Seurat was used to analyze the immune cell subpopulation. (B) Eight immune cell clusters were identified, including CD4+ T cells, CD8+ T cells, naïve B cells and antibody secreting B cells for lymphocytes; and macrophages, dendritic cells, and neutrophils for myeloid cells. (C) Cytotoxic CD8+ T cells were highly expressed in CBG-treated PRLomas, and expressed tumoricidal granule components, including perforin (PRF1) and the granzymes GZMB, GNLY, and KLRD1 as well as the inflammatory cytokine CCL5. (D) Multiplex immunostaining was used to analyze CD8+ T cells in scRNA-matching samples and one treatment naïve samples (n = 6 total). The results confirmed high CD8+ T cell content in CBG-treated tissues. (E) GeoMx digital spatial profiling was employed in scRNA-matching samples and one treatment naïve samples (n = 6 total), and demonstrated high expression of CD25 (IL2R) of CD8+ T cells in CBG-treated tissue (p<0.05).

Transcriptomic profile of the stromal population in CBG-treated PRLomas

We also noted that the proportion of stromal cells was significantly higher in (+)CBG (10.67 ± 0.76%) compared to (−)CBG samples (0.5 ± 0%, p<0.001, Table 2, Fig. 3A). The stromal cells in the (+)CBG samples comprised two main subgroups (Supplementary Table-6, Fig 3B). Firstly, endothelial cells were characterized by high expression of PECAM1, VWF, ENG, VEGFC, CDH5, PLVAP, EMCN and CLEC14A (Fig. 3C, top). Using ENRICHR Gene Ontology analysis, we demonstrated that these endothelial cell DEGs were involved in cadherin-dependent cell-cell/substrate junction and focal adhesion, transmembrane receptor- (IGFR, and TGFBR) associated signal transduction, antigen presentation and cytokine signaling, which collectively regulate cell migration, angiogenesis, and immune activation (Fig. 3D). The second stromal subgroup comprised fibroblasts characterized by expression of TAGLN, PDGFRB, SOD3, HIGD1B, DCN, COL1A2, NOTCH3 and AGT (Fig. 3C, bottom). The fibroblasts were subdivided into 3 subgroups, namely F1, F2, and F3 (Fig. 3E). Using a Venn diagram approach, we compared the common and distinct DEGs amongst these 3 fibroblast subgroups (Fig. 3F). The mural cell/pericyte markers PDGFRB (15) and ITGB1 (16) as well as secretory ECM components LGALS1 and SPRAC were commonly expressed in all the subtypes, suggesting that some mesenchymal adventitial features were shared among these fibroblasts (Fig. 3F). The F1 subtype expressed higher levels of canonical resident fibroblast markers, such as collagens (COL1A2, COL3A1, COL4A3, and COL6A3), ECM genes (MGP, DCN, NTS, BGN and FBLN1) and inflammatory factors (C1S, C1R and STEAP4, Fig. 3G, left) (15, 17). The F2 fibroblasts expressed high levels of inflammatory transcription factors, such as AP-1 (FOS, FOSB, JUN, JUNB/D) and CEBPD, and kinase regulators (PHLDA1 and CAMK2N1), indicative of their activated status potentially triggered by neighboring inflammatory cues (Fig. 3G, middle) (18, 19). Finally, the F3 subtype expressed archetypal myofibroblast markers including the myosin genes TPM1, TPM2, MYL6/9/K, and MYH11 which regulate intermediate contractile force; along with matrix components DSTN, FLNA, TAGLN, CAV1, MCAM, and ACTA2, reflective of their state as active myofibroblasts (Fig. 3G, right) (20).

Figure 3. Striking intratumor stromal cell heterogeneity between CBG-treated and untreated PRLoma tissues.

Figure 3

Open in a new tab

(A) Increased numbers of stromal cells were noted in CBG-treated versus untreated PRLomas (p<0.005). (B) The subtypes of the stromal cells were analyzed (upper panel), and their percentages were shown in histogram (lower panel). (C) The markers of endothelial cells and fibroblasts were depicted by feature plots. (D) ENRICHR was used to analyze GO molecular function, cellular component, and biological process of endothelial cell population. (E) The fibroblast cell populations were further sub-grouped into F1, F2, and F3. (F) The common features of fibroblasts were analysis by Venn diagraph. (G) According to distinguished canonical transcript markers, the fibroblast sub-groups were identified as resident fibroblasts (F1), inflammatory fibroblasts (F2), and myofibroblasts (F3).

Trajectory analysis of the phenotypic transition between progenitor and stromal cells.

To try and further understand the origins of the increased stromal cells in the (+)CBG samples, we conducted a pseudotime trajectory using the Monocle2 algorithm (21) (Fig. 4A). Using ENRICHR from the ChIP-x Enrichment Analysis (ChEA) 2022 database, the transcription factor SOX2 was predicted as the master regulator of the progenitor population (Supplementary Fig. 1B, and Supplementary Table 7). The specificity of SOX2 and stromal markers TAGLN (fibroblasts) and VWF (endothelial) was confirmed as illustrated in Supplementary Fig. 1C. SOX2+ progenitor cells were identified as the origin of the pseudotime trajectory, whereby TAGLN+ fibroblasts and VWF+ endothelial cells branched off from progenitors into two distinctive groups (Fig. 4B). Pseudotime ordering denoted 5 states with State-4 being composed of progenitors, State-1 mainly of endothelial cells and States 2, 3, 5 composed of a mixture of fibroblasts (Fig. 4C). A mixed group of resident fibroblast-like F1 cells and myofibroblast-like F3 cells emerged at an early state/stage of the pseudotime trajectory (State-5), followed by a heterogenous group composed of a small number of F1, F2 and endothelial cells (State-3), while the late group (State-2) mainly contained F2 inflammatory fibroblasts (Fig. 4D). Analysis of the effect of CBG treatment on this trajectory track indicated that whereas the (−)CBG cells predominantly remained in the progenitor state with limited differentiation toward fibroblasts and endothelial cells, the (+)CBG cells were distributed along a pseudotime track toward stromal cell differentiation (Fig. 4E), consistent with the higher proportion of stromal cells observed in the (+)CBG group (Fig. 3A). This CBG-driven expansion of stromal cells may provide new insights into CBG-mediated fibrotic effects (22, 23).

Figure 4. Trajectory analysis of the phenotypic transition between progenitor and stromal cells.

Figure 4.

Open in a new tab

(A) The subset function of Seurat was used to analyze the progenitor and stromal cell populations. (B) Taking the progenitor cells as the root of a pseudotime trajectory, stromal fibroblasts and endothelial cells appeared to branch into two distinctive groups. (C & D) Cell state analysis revealed progenitors in State-4, endothelial cells in State-1, whereas State-2, 3, 5 contained a mixture of fibroblasts. (E) The (+)CBG samples were distributed along a pseudotime track directed toward stromal cells, while the (–)CBG samples remained in the progenitor state without differentiation, suggesting that CBG treatment may promote progenitor cell differentiation into stromal cells.

Differential expression of serotonin receptors in CBG-treated or treatment naive patients

Off target actions of CBG via the serotonin receptors has been implicated in fibroblast mitogenesis and increased cardiac valve thickening (24, 25). Of the 17 HTR sub-types, we detected HTR1F, HTR2B and HTR4 in our scRNAseq dataset (Fig. 5A). HTR4 was mainly expressed in the tumor cell population, whereas HTR1F and HTR2B expression predominated in stromal cells (Fig. 5A). Our analysis further revealed that HTR1F was mainly expressed in the myofibroblast-like F3 cells, whereas low HTR2B expression was seen in the resident fibroblasts-like F1 cells (Fig. 5B). Expression of HTR1F and HTR2B in stromal cells was confirmed by immunofluorescent staining where HTR2B was distributed throughout the normal pituitary (Fig. 5C, yellow, top panel), partially overlapping with pericyte-derived pleiotrophin (PTN, Fig. 5C, orange, 2nd panel, solid arrow). In contrast to the diffuse distribution of HTR2B observed in the normal pituitary, HTR2B was expressed in clusters within the prolactinomas and more highly expressed in (+)CBG samples (Fig. 5D, yellow, top panel) compared to the (−)CBG samples (Fig. 5E, yellow, top panel). Similar to the normal pituitary, HTR2B co-localized with PTN+ pericytes in the PRLoma samples (Figs. 5D & E, solid arrow), indicating a close association with the tumor vasculature. Whereas little if any HTR1F expression was seen in either the normal pituitary (Fig. 5F, red, top panel) or the (−)CBG samples (Fig. 5H, red, top panel), HTR1F was clearly visible around PTN+ luminal structures in the (+)CBG samples (Fig. 5G, open arrow), but distinct from HTR2B+ vessels in the (+)CBG samples (Fig. 5D, solid arrow).

Figure 5. Evaluation of HTR2B and HTR1F expression using multiplex immunostaining.

Figure 5.

Open in a new tab

(A) Dotplot depiction of serotonin receptor expression in 6 cell populations from CBG-treated and untreated PRLomas. (B) HTR1F was highly expressed in the myofibroblast-like F3 cells, and HTR2B in the resident fibroblasts-like F1 cells. (C-H) Multiplex immunostaining was used to assess expression of HTR2B and HTR1F in normal pituitary of autopsy sample (C & F), CBG-treated (D & G) and untreated (E & H) PRLoma samples. PTN was used as a reference for vessel structures. Solid arrows indicated colocalization of HTR2B with PTN, and open arrows indicated colocalization of HTR1F with PTN.

Discussion

Pituitary lactotrophs are under the inhibitory control of hypothalamic-derived dopamine which acts via the G-protein coupled dopamine receptor family comprising 5 subtypes D1-D5 (26). Our scRNAseq revealed that DRD2, and DRD4, which associate with an inhibitory Gαi subunit to inhibit AC thereby blocking cAMP formation, were detected (Supplementary Fig. 1D). DRD2 predominated in the tumor population as expected (Supplementary Fig. 1D) (26).

Our scRNAseq analysis of these prolactinomas provides new important insights into the mechanisms of CBG to regulate prolactin production and secretion. Firstly, in addition to inhibiting PRL transcription, CBG-induced downregulation of the secretogranins, (SCG5 and VGF) underscored broader mechanisms by which cabergoline inhibits pituitary hormone secretion (Fig. 1D). Secondly, we noted that several genes involved in cell viability including Rho GTPase activating protein 5 (ARHGAP5), BTG anti-proliferation factor 2 (BTG2), and branched chain amino acid transaminase 1 (BCAT1, p<10−4) were increased in the tumor cells of (+)CBG samples. ARHGAP5 for example is a negative regulator of small RHO GTPases (27). ARHGAP5 knockout animals are 30% reduced in size, have impaired insulin/IGF-1 signaling and exhibit defects in brain and lung development (28) (29).

A further major feature that we observed in the tumor cell population was that the MHC-I components HLA-A and B2M were reduced in the (+)CBG samples (Fig. 1D). The major histocompatibility complex (MHC)/human leukocyte antigen (HLA) is a highly conserved protein complex which governs adaptative immunity. MHC-I molecules are ubiquitously expressed in all nucleated cells that display antigens from pathogens or transformed tumor cells to CD8+ cytotoxic T cells. In contrast, MHC-II which is restricted to antigen-presenting cells, such as B lymphocytes, macrophages and dendritic cells, activates CD4+ helper T cells under homeostatic conditions (30). Successful recognition of the cognate antigen presented on the tumor cell surface by the T cell receptor (TCR) and its coreceptor CD8 leads to caspase-dependent tumor cell apoptosis and destruction by release of perforin, granzymes, and granulysin. Concordant with a reduction in HLA-A/B2M in the (+)CBG tumor cells, we observed increased numbers of CD8+ T cells in the (+)CBG samples (Fig. 2).

Prolactin itself can promote T cell maturation (31) and modulate CD4+ T cell response in a dose-dependent manner (32). Thus the actions of CBG on the CD8+ T cell population could be multifactorial, in part through i) normalizing the hyperprolactinemia associated immune response; ii) directly activating dopaminergic receptors on immune cells; and/or iii) exhibiting a tumoricidal effect by facilitating tumor antigen presentation and T cell activation. We also demonstrated increased CD25 (IL2R) protein expression in the CD8+ T cells, suggesting IL2 may play a role in CBG-induced CD8+ T cell activation (Fig, 2E). IL-2 is produced by CD4+ T helper type 1 (Th1) cells and acts to enhance the activation of effector cells, such as cytotoxic NK cells, T cells and monocytes, against a wide variety of tumors (33). Our findings raise the intriguing possibility that activation of CD8+ T cells may play a previously unknown role in the tumoricidal actions of CBG, perhaps mediated by IL-2.

Tumor infiltrating immune cells (TILCs) derive from the innate (monocytes, macrophages, dendric cells, and natural killer cells) and adaptive immune systems (T and B cells) (34). Macrophages and T lymphocytes, particularly CD8+ and CD4+ T cells are the most abundant TILC populations reported in pituitary tumors (35). The immune microenvironment of a lactotroph tumor is particularly interesting given prolactin can function not only as a bioactive peptide hormone, but also a circulating cytokine which modulates the threshold for B cell activation. Hyperprolactinemia has been shown to play a role in autoimmune disorders such as rheumatoid arthritis, autoimmune thyroid disease and multiple sclerosis (36) and patients with PRLomas exhibit a higher prevalence of autoimmune diseases (mainly thyroid disease, 28%) compared to patients with non-functioning pituitary adenomas (16%) and the general population (5–8%) (37, 38, 39). In line with this, our scRNAseq analysis demonstrates the presence of naive and memory B cells (53 cells, 4.4%) and antibody secreting B cells (41 cells, 3.4%, Fig. 2B). Furthermore, 2 populations of macrophages were identified with similar proportions in the CBG-treated and untreated samples (Fig. 2B & Supplementary Table 5). The macrophage-1 subtype (178 cells, 14.8%) represented proinflammatory FCN1+ M1 Φ (Supplementary Table 4), and expressed high level of Ficolin 1 (FCN1), a soluble pattern-recognition receptor of innate immunity, which activates host immune defenses against invading pathogens (40, 41). The macrophage-2 subtype (210 cells, 17.4%) expressed anti-inflammatory markers such as apolipoprotein E (APOE), mannose receptor C-type 1 (MRC1), folate receptor beta (FOLR2), membrane spanning 4-domains A4A (MS4A4A), and DAB adaptor protein 2 (DAB2), representing SELENOP+ M2 Φ (Supplementary Table 4) (42, 43, 44).

We also noted a striking increase in stromal cells in (+)CBG samples (Fig. 3A). Effects of high dose ergot derived DAs to cause fibrosis of the lung (45, 46), retroperitoneum (47, 48) and cardiac valves (49) are well described (50, 51). However, the association with CBG doses used in PRLoma treatment and cardiac valve abnormalities is not so clear (52). Our trajectory analysis suggested that CBG treatment may promote progenitor cells to differentiate into fibroblasts as well as endothelial cells (Fig. 4). Interestingly, whereas HTR4 was the predominant serotonin receptor expressed in tumor cells, HTR1F and HTR2B predominated in the stromal cell population (Fig. 5). HTR1F is widely distributed in the human brain and brain vasculature (53) and has been a therapeutic target for migraine (54, 55). Furthermore, high HTR1F mRNA expression has been reported in the avian pituitary and the HTR1F agonist LY344864 inhibited VIP-stimulated chicken in vitro PRL secretion (56). Similarly, we demonstrate immunocytochemical HTR1F expression close to vasculature in the (+)CBG samples (Fig. 5). Whether HTR4 expression in tumor cells or HTR1F in the stroma represent therapeutic targets in lactotroph tumors may warrant further investigation.

In conclusion, our scRNAseq studies demonstrate for the first-time a potential role of CBG to modulate immune cell activation, especially CD8+ T cells. Additionally, our studies shed further light on the fibrosis enhancing off-target actions of this drug class which could potentially limit response to DA therapy and in parallel identify several serotonin receptor sub-types as potential therapeutic targets in prolactinomas.

Materials and Methods:

Patient Information

Aliquots of freshly surgically resected tumor tissue were collected from 3 CBG-treated, and 2 CBG-treatment naive surgically resected PRLomas by our neurosurgical collaborators (MB, WK). Patient demographics are included in Table 1.

Ethics

The study was performed in accordance with our approved UCLA Institutional Review Board protocol (IRB#20–002235) and written informed consent was obtained from all patients. The procedures of this study were performed in compliance with the Declaration of Helsinki.

Single-cell RNA-Sequencing

Single cell suspensions of PRLomas were obtained by mechanical (MACS dissociator, Miltenyi Biotec Inc., Germany) and enzymatic digestion (Tumor Dissocation Kit, human, Cat# 130-095-929). Libraries were generated on a 10x Genomics Chromium Controller following the manufacturer’s protocol for the v3 reagent kit (10x Genomics). In brief, cell suspensions were loaded onto a Chromium Single Cell A Chip, aiming for 10,000 cells per channel for generation of single-cell gel bead-in-emulsions (GEMs), following which reverse transcription was performed. Post-GEM reverse transcription products were cleaned using DynaBeads MyOne silane beads (Thermo Fisher Scientific, Waltham, MA). cDNA was amplified, cleaned and quantified, then enzymatically fragmented and size selected prior to library construction. Libraries were quantified by KAPA quantitative PCR for Illumina adapters (Roche, Pleasanton, CA), library size determined by Agilent TapeStation D1000 tapes and then sequenced on NextSeq 500 and Novaseq 6000 sequencers (Illumina, San Diego, CA).). Fastq files are deposited in the NCBI BioSample database (http://www.ncbi.nlm.nih.gov/biosample/) under accession number SAMN39182476 (PRLoma1), SAMN39182477 (PRLoma2), SAMN39182478 (PRLoma3), SAMN39182480 (PRLoma4), and SAMN39182481 (PRLoma5).

Bioinformatic analyses of scRNAseq data

Demultiplexed fastq files generated at UCLA Technology Center for Genomics & Bioinformatics (TCGB) were analyzed with the 10x Genomics Cell Ranger 2.1.1. The pipeline aligned the reads to the University of California Santa Cruz (UCSC) human reference (GRCh38) transcriptome using the RNAseq alignment program STAR. Data were imported and analyzed using Seurat package within Rstudio (57). For quality assurance, cells were selected for downstream analysis using the following conservative cut-offs: 1) Cell barcodes associated with the most UMIs were employed by estimating the number of cells captured as 5% of the input beads and retained this number of cell barcodes for downstream analysis; 2) Only cells with >200 and <7,500 unique genes detected, and UMI >2,000 and <50,000 were analyzed; and 3) Only cells with <20% of their counts mapping to MT genes were included; and 4) Only genes detected in >5 cells were included. Cumulatively, from all 5 PRLoma tissue samples, we obtained 36,256 cells for the subsequent analysis. SCTransform was used for normalization and integration to reduce batch effect (7). Principal component analysis (PCA) and uniform manifold approximation and projection (UMAP) were then performed. Cell clusters were annotated using the “FindNeighbors” function, and the “mast” method (FC. threshold = 1.5) used to identify differentially expressed genes (DEGs) using the “FindAllMarkers” function in Seurat within Rstudio (57). Following subsetting of the inividdual cell populations, standard Seurat workflow was performed, including RunPCA, FindNeighbours, FindClusters, RunUMAP, to characterize the subgroups of each population (7). No customized codes were used. We employed monocle2 to study pseudotime trajectories of cells (58). The UMI matrix was used as the input, and the various genes that had been detected by Seurat were used to generate a building trace.

Multiplex IHC Staining

Multiplex IHC staining was performed at the UCLA translational pathology core laboratory (TPCL) using the Opal Multiplex IHC Kit (Akoya Bioscences). A total of 7 samples were analyzed, including 5 scRNAseq-matched samples, one treatment naïve PRLoma (PRLoma6), and one post-mortem autopsy-derived normal pituitary tissue. Primary antibodies specific to CD8 (Cat#M7103, Clone/Lot#C8/144B, 1:100 dilution), CD45 (Cat#M0701, Clone/Lot#2B11+PD7/26, 1:200 dilution) and CD31 (Cat#M0823, Clone/Lot#JC70A, 1:100 dilution) were purchased from DAKO. HTR1F (Cat#HPA005555) and HTR2B (Cat# HPA012867) were purchased from Sigma. PTN (Cat#sc-74443) was purchased from Santa Cruz. For the CD8, CD45, and CD31 staining panel, the fluorophores used were Opal 520 FITC for CD8, Opal 570 TRITC for CD31, and Opal 690 CY5 for CD45. The images were then scanned at 20x magnification (Vectra Polaris, Akoya). The number of CD8+ T cells were manually counted for each patient in three different high-power fields, and statistical analysis was performed using Prism with a nonparametric Mann Whitney test. For the HTR2B, PTN, and HTR1F staining panel, the fluorophores used were Opal 570 TRITC for HTR2B, Opal 620 CY3 for PTN, Opal 690 CY5.5 for HTR1F. The original scan files were uploaded to Figshare and access is provided upon request.

GeoMx digital spatial profiling

Formalin fixed paraffin embedded (FFPE) slides from 6 individual PRLoma samples, including our 5 scRNAseq-matched samples and one treatment naïve PRLoma (PRLoma6), were analyzed with GeoMx Digital Spatial Profiling (DSP) Proteomics Analysis. FFPE sections were stained for DNA, PanCK, CD45 and CD8 as morphology markers, and together with GeoMx Human Protein Core for NGS panel, Immune Activation Status Panel and IO Drug Target Panel from Nanostring Technologies (Seattle, WA). Thirty Region of Interest (ROIs) were selected and collected by the DSP. Libraries were made and sequenced with Miseq PE 2×75bp. Reads were aligned by GeoMX software. Data quality control analysis was performed with GeoMx DSP according to manufacturer’s instructions. Sample read counts were normalized with Geometric-mean using H3 and S6 as housekeeping controls. Differential expression was determined by sample comparison using an unpaired t-test. The original scan files were uploaded to Figshare and access is provided upon request.

Statistical analysis:

The “FindAllMarkers” function in the Seurat package was used to identify DEGs based on wilcox testing with Bonferroni correction. Significant DEGs were selected from genes with adjusted p values (p_value_adj) <10−4.

Supplementary Material

Supple-Table-3

Supplementary Table 3. List of DEGs observed in the tumor cell population from CBG-treated and untreated samples.

Supple-Table-1

Supplementary Table 1. List of DEGs observed in the 6 cell type populations.

Supple-Table-2

Supplementary Table 2. GO molecular function analysis of the tumor cell population top DEGs by ENRICHR.

Supple-Table-5

Supplementary Table 5. Percentage of various immune cell subpopulations in (–)CBG and (+)CBG samples.

Supple-Table-4

Supplementary Table 4. List of DEGs observed in the immune cell subpopulations.

Supple-Table-6

Supplementary Table 6. List of DEGs of the stromal cell subpopulations.

Supple-Table-7

Supplementary Table 7. Enrichment analysis of the master regulators of the progenitor cells using ENRICHR.

Supple-Figure

Supplementary Figure 1. (A) Comparative analysis of upstream regulators of the immune cell population in CBG-treated and untreated PRLomas implicated activation of IL2, and involvement of the inflammatory transcription factors NFκB and STATs in response to CBG treatment. (B) The master regulators of the progenitors were predicted using ENRICHR from ChEA library, and the bar chart shows the top 10 enriched terms along with their corresponding p-values. An asterisk (*) next to a p-value indicates the term also has a significant adjusted p-value (<0.05). (C) Dotplot depicted the specificity of SOX-2 and stromal markers TAGLN (fibroblasts) and VWF (endothelial). (D) Two dopamine receptors (DRD2, and DRD4) were detected by scRNAseq and DRD2 was the main subtype in the tumor population.

Acknowledgements:

We thank Dr. Yunfeng Li at the UCLA Translational Pathology Core Laboratory for her technical support for multiplex immunostaining. Also, we appreciate the helps from Dr. Xinmin Li, Dr. Chao Niu, and Dr. Yu-chyuan Su at the UCLA Technology Center for Genomics & Bioinformatics (TCGB), and Dr. Jerid Robinson from NanoString Technologies for technical support and the analysis of GeoMx digital spatial profiling. We thank Dr. Lu Sun and Dr. Lizhong Ding for their valuable assistance with scRNAseq analysis. We thank Dr. Shino Magaki and Christopher K. Williams for providing post-mortem autopsy-derived normal pituitary tissues.

Funding:

This work was supported by the National Cancer Institute NIH/NCI R01CA251930 (APH), NIH/NCI R21CA264838 (APH), and the Warley Trust (APH).

Footnotes

Declaration of interest: The authors have no conflict of interest to declare.

References:

  • 1.Chanson P, Maiter D. The epidemiology, diagnosis and treatment of Prolactinomas: The old and the new. Best practice & research Clinical endocrinology & metabolism. 2019;33(2):101290. [DOI] [PubMed] [Google Scholar]
  • 2.ML J-R. From resistant to aggressive and malignant prolactinomas: A translational approach. J Endocr Disord. 2014;1:1012. [Google Scholar]
  • 3.Webster J, Piscitelli G, Polli A, Ferrari CI, Ismail I, Scanlon MF. A Comparison of Cabergoline and Bromocriptine in the Treatment of Hyperprolactinemic Amenorrhea. New England Journal of Medicine. 1994;331(14):904–9. [DOI] [PubMed] [Google Scholar]
  • 4.Auriemma RS, Pirchio R, De Alcubierre D, Pivonello R, Colao A. Dopamine Agonists: From the 1970s to Today. Neuroendocrinology. 2019;109(1):34–41. [DOI] [PubMed] [Google Scholar]
  • 5.Vermeulen E, D’Haens J, Stadnik T, Unuane D, Barbe K, Van Velthoven V, et al. Predictors of dopamine agonist resistance in prolactinoma patients. BMC Endocrine Disorders. 2020;20(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Mehta AE, Reyes FI, Faiman C. Primary radiotherapy of prolactinomas: Eight- to 15-year follow-up. The American Journal of Medicine. 1987;83(1):49–58. [DOI] [PubMed] [Google Scholar]
  • 7.Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM III, et al. Comprehensive Integration of Single-Cell Data. Cell. 2019;177(7):1888–902.e21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Kim T, Gondré-Lewis M, Arnaoutova I, Loh Y. Dense-core secretory granule biogenesis. Physiology. 2006;21:124–33. [DOI] [PubMed] [Google Scholar]
  • 9.Le Berre L, Chesneau M, Danger R, Dubois F, Chaussabel D, Garand M, et al. Connection of BANK1, Tolerance, Regulatory B cells, and Apoptosis: Perspectives of a Reductionist Investigation. Frontiers in Immunology. 2021;12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Smulski CR, Eibel H. BAFF and BAFF-Receptor in B Cell Selection and Survival. Frontiers in Immunology. 2018;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Sanz I, Wei C, Jenks SA, Cashman KS, Tipton C, Woodruff MC, et al. Challenges and Opportunities for Consistent Classification of Human B Cell and Plasma Cell Populations. Frontiers in Immunology. 2019;10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Choi S-C, Kim KD, Kim J-T, Kim J-W, Yoon D-Y, Choe Y-K, et al. Expression and regulation of NDRG2 (N-myc downstream regulated gene 2) during the differentiation of dendritic cells. FEBS Letters. 2003;553(3):413–8. [DOI] [PubMed] [Google Scholar]
  • 13.Choi SC, Kim KD, Kim JT, Kim JW, Lee HG, Kim JM, et al. Expression of human NDRG2 by myeloid dendritic cells inhibits down-regulation of activated leukocyte cell adhesion molecule (ALCAM) and contributes to maintenance of T cell stimulatory activity. J Leukoc Biol. 2008;83(1):89–98. [DOI] [PubMed] [Google Scholar]
  • 14.Shaath H, Vishnubalaji R, Elkord E, Alajez NM. Single-Cell Transcriptome Analysis Highlights a Role for Neutrophils and Inflammatory Macrophages in the Pathogenesis of Severe COVID-19. Cells. 2020;9(11). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Xie T, Wang Y, Deng N, Huang G, Taghavifar F, Geng Y, et al. Single-Cell Deconvolution of Fibroblast Heterogeneity in Mouse Pulmonary Fibrosis. Cell Rep. 2018;22(13):3625–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Scully KM, Skowronska-Krawczyk D, Krawczyk M, Merkurjev D, Taylor H, Livolsi A, et al. Epithelial cell integrin β1 is required for developmental angiogenesis in the pituitary gland. Proc Natl Acad Sci U S A. 2016;113(47):13408–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Zhang D, Hugo W, Bergsneider M, Wang MB, Kim W, Vinters HV, et al. Single Cell RNA Sequencing in Silent Corticotroph Tumors Confirms Impaired POMC Processing and Provides New Insights into Their Invasive Behavior. European Journal of Endocrinology. 2022:EJE-21–1183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Florin L, Hummerich L, Dittrich BT, Kokocinski F, Wrobel G, Gack S, et al. Identification of novel AP-1 target genes in fibroblasts regulated during cutaneous wound healing. Oncogene. 2004;23(42):7005–17. [DOI] [PubMed] [Google Scholar]
  • 19.Davidson S, Coles M, Thomas T, Kollias G, Ludewig B, Turley S, et al. Fibroblasts as immune regulators in infection, inflammation and cancer. Nature Reviews Immunology. 2021;21(11):704–17. [DOI] [PubMed] [Google Scholar]
  • 20.Peyser R, MacDonnell S, Gao Y, Cheng L, Kim Y, Kaplan T, et al. Defining the Activated Fibroblast Population in Lung Fibrosis Using Single-Cell Sequencing. American Journal of Respiratory Cell and Molecular Biology. 2019;61(1):74–85. [DOI] [PubMed] [Google Scholar]
  • 21.Qiu X, Mao Q, Tang Y, Wang L, Chawla R, Pliner HA, et al. Reversed graph embedding resolves complex single-cell trajectories. Nature Methods. 2017;14(10):979–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Yamamoto M, Uesugi T. Dopamine agonists and valvular heart disease in patients with Parkinson’s disease: evidence and mystery. Journal of Neurology. 2007;254(5):74–8. [Google Scholar]
  • 23.Choi J, Horner K. Dopamine Agonists. . In: StatPearls [Internet] Treasure Island (FL): StatPearls Publishing. 2022 [Google Scholar]
  • 24.Oana F, Onozuka H, Tsuchioka A, Suzuki T, Tanaka N, Kaidoh K, et al. Function and expression differences between ergot and non-ergot dopamine D2 agonists on heart valve interstitial cells. J Heart Valve Dis. 2014;23(2):246–52. [PubMed] [Google Scholar]
  • 25.Zanettini R, Antonini A, Gatto G, Gentile R, Tesei S, Pezzoli G. Valvular Heart Disease and the Use of Dopamine Agonists for Parkinson’s Disease. New England Journal of Medicine. 2007;356(1):39–46. [DOI] [PubMed] [Google Scholar]
  • 26.Oh MC, Aghi MK. Dopamine agonist-resistant prolactinomas. J Neurosurg. 2011;114(5):1369–79. [DOI] [PubMed] [Google Scholar]
  • 27.Héraud C, Pinault M, Lagrée V, Moreau V. p190RhoGAPs, the ARHGAP35- and ARHGAP5-Encoded Proteins, in Health and Disease. Cells. 2019;8(4):351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Sordella R, Classon M, Hu K-Q, Matheson SF, Brouns MR, Fine B, et al. Modulation of CREB Activity by the Rho GTPase Regulates Cell and Organism Size during Mouse Embryonic Development. Developmental Cell. 2002;2(5):553–65. [DOI] [PubMed] [Google Scholar]
  • 29.Birnbaum MJ. RhoGAP: The Next Big Thing for Small Mice? Developmental Cell. 2002;2(5):521–3. [DOI] [PubMed] [Google Scholar]
  • 30.Mak TW, Saunders ME. 10 - MHC: The Major Histocompatibility Complex. In: Mak TW, Saunders ME, editors. The Immune Response. Burlington: Academic Press; 2006. p. 247–77. [Google Scholar]
  • 31.Carreño PC, Sacedón R, Jiménez E, Vicente A, Zapata AG. Prolactin affects both survival and differentiation of T-cell progenitors. Journal of neuroimmunology. 2005;160(1–2):135–45. [DOI] [PubMed] [Google Scholar]
  • 32.Tomio A, Schust DJ, Kawana K, Yasugi T, Kawana Y, Mahalingaiah S, et al. Prolactin can modulate CD4+ T-cell response through receptor-mediated alterations in the expression of T-bet. Immunol Cell Biol. 2008;86(7):616–21. [DOI] [PubMed] [Google Scholar]
  • 33.Nie D, Fang Q, Li B, Cheng J, Li C, Gui S, et al. Research advances on the immune research and prospect of immunotherapy in pituitary adenomas. World Journal of Surgical Oncology. 2021;19(1):162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Wang Z, Guo X, Gao L, Deng K, Lian W, Bao X, et al. The immune profile of pituitary adenomas and a novel immune classification for predicting immunotherapy responsiveness. The Journal of Clinical Endocrinology & Metabolism. 2020;105(9):e3207–e23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Marques P, Silva AL, López-Presa D, Faria C, Bugalho MJ. The microenvironment of pituitary adenomas: biological, clinical and therapeutical implications. Pituitary. 2022;25(3):363–82. [DOI] [PubMed] [Google Scholar]
  • 36.Borba VV, Zandman-Goddard G, Shoenfeld Y. Prolactin and Autoimmunity. Front Immunol. 2018;9:73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Larouche V, Correa JA, Cassidy P, Beauregard C, Garfield N, Rivera J. Prevalence of autoimmune disease in patients with prolactinomas and non-functioning pituitary adenomas. Pituitary. 2016;19(2):202–9. [DOI] [PubMed] [Google Scholar]
  • 38.Pilli T, Cardinale S, Dalmiglio C, Secchi C, Fralassi N, Cevenini G, et al. Autoimmune thyroid diseases are more common in patients with prolactinomas: a retrospective case–control study in an Italian cohort. Journal of endocrinological investigation. 2019;42(6):693–8. [DOI] [PubMed] [Google Scholar]
  • 39.Elenkova A, Racheva P, Kirilov G, Zacharieva S. Clinical course of autoimmune thyroid diseases in women with prolactinomas: Results from a prospective study in a single tertiary centre. Endocrinología, Diabetes y Nutrición. 2022. [DOI] [PubMed] [Google Scholar]
  • 40.Zhang F, Mears JR, Shakib L, Beynor JI, Shanaj S, Korsunsky I, et al. IFN- γ and TNF- α drive a CXCL10 + CCL2 + macrophage phenotype expanded in severe COVID-19 and other diseases with tissue inflammation. bioRxiv. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Zhang L, Li Z, Skrzypczynska KM, Fang Q, Zhang W, O’Brien SA, et al. Single-Cell Analyses Inform Mechanisms of Myeloid-Targeted Therapies in Colon Cancer. Cell. 2020;181(2):442–59.e29. [DOI] [PubMed] [Google Scholar]
  • 42.Abdelaziz MH, Abdelwahab SF, Wan J, Cai W, Huixuan W, Jianjun C, et al. Alternatively activated macrophages; a double-edged sword in allergic asthma. Journal of Translational Medicine. 2020;18(1):58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Puig-Kröger A, Sierra-Filardi E, Domínguez-Soto A, Samaniego R, Corcuera MT, Gómez-Aguado F, et al. Folate Receptor β Is Expressed by Tumor-Associated Macrophages and Constitutes a Marker for M2 Anti-inflammatory/Regulatory Macrophages. Cancer Research. 2009;69(24):9395–403. [DOI] [PubMed] [Google Scholar]
  • 44.Sanyal R, Polyak MJ, Zuccolo J, Puri M, Deng L, Roberts L, et al. MS4A4A: a novel cell surface marker for M2 macrophages and plasma cells. Immunol Cell Biol. 2017;95(7):611–9. [DOI] [PubMed] [Google Scholar]
  • 45.Belmonte Y, de Fàbregues O, Marti M, Domingo C. Pleuropulmonary Toxicity of Another Anti-Parkinson’s Drug: Cabergoline. Open Respir Med J. 2009;3:90–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Eda S, Orii K, Hori S, Gomi E, Hashimoto T. Slowly progressive lung fibrosis (dominant in right upper lobe) after encapsulated pleurisy induced by cabergoline in a patient with parkinsonism. B40 CASE REPORTS: PLEURAL DISEASES: DIAGNOSTIC DILEMMAS. 183: American Thoracic Society; 2011. p. A2946-A. [Google Scholar]
  • 47.Alberti C Drug-induced retroperitoneal fibrosis: short aetiopathogenetic note, from the past times of ergot-derivatives large use to currently applied bio-pharmacology. G Chir. 2015;36(4):187–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Jarzynska A, Khan HH, Chandran S, editors. A case of retroperitoneal fibrosis on low dose cabergoline. Endocrine Abstracts; 2019: Bioscientifica. [Google Scholar]
  • 49.Vroonen L, Lancellotti P, Garcia MT, Dulgheru R, Rubio-Almanza M, Maiga I, et al. Prospective, long-term study of the effect of cabergoline on valvular status in patients with prolactinoma and idiopathic hyperprolactinemia. Endocrine. 2017;55(1):239–45. [DOI] [PubMed] [Google Scholar]
  • 50.Newman-Tancredi A, Cussac D, Quentric Y, Touzard M, Verrièle L, Carpentier N, et al. Differential actions of antiparkinson agents at multiple classes of monoaminergic receptor. III. Agonist and antagonist properties at serotonin, 5-HT(1) and 5-HT(2), receptor subtypes. Journal of Pharmacology and Experimental Therapeutics. 2002;303:815–22. [DOI] [PubMed] [Google Scholar]
  • 51.Elangbam CS. Drug-induced Valvulopathy: An Update. Toxicologic Pathology. 2010;38(6):837–48. [DOI] [PubMed] [Google Scholar]
  • 52.Lafeber M, Stades AM, Valk GD, Cramer MJ, Teding van Berkhout F, Zelissen PM. Absence of major fibrotic adverse events in hyperprolactinemic patients treated with cabergoline. Eur J Endocrinol. 2010;162(4):667–75. [DOI] [PubMed] [Google Scholar]
  • 53.Bhalla P, Sharma HS, Wurch T, Pauwels PJ, Saxena PR. Molecular cloning and expression of the porcine trigeminal ganglion cDNA encoding a 5-ht(1F) receptor. Eur J Pharmacol. 2002;436(1–2):23–33. [DOI] [PubMed] [Google Scholar]
  • 54.Targeted Vila-Pueyo M. 5-HT(1F) Therapies for Migraine. Neurotherapeutics. 2018;15(2):291–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Bouchelet I, Case B, Olivier A, Hamel E. No contractile effect for 5-HT1D and 5-HT1F receptor agonists in human and bovine cerebral arteries: similarity with human coronary artery. Br J Pharmacol. 2000;129(3):501–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Sun C, Qiu Y, Ren Q, Zhang X, Cao B, Zou Y, et al. Molecular Cloning and Functional Characterization of Three 5-HT Receptor Genes (HTR1B, HTR1E, and HTR1F) in Chickens. Genes (Basel). 2021;12(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Luecken MD, Theis FJ. Current best practices in single-cell RNA-seq analysis: a tutorial. Molecular Systems Biology. 2019;15(6):e8746. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Trapnell C, Cacchiarelli D, Grimsby J, Pokharel P, Li S, Morse M, et al. The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nature Biotechnology. 2014;32(4):381–6. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supple-Table-3

Supplementary Table 3. List of DEGs observed in the tumor cell population from CBG-treated and untreated samples.

NIHMS2075112-supplement-Supple-Table-3.xlsx (10.6KB, xlsx)
Supple-Table-1

Supplementary Table 1. List of DEGs observed in the 6 cell type populations.

NIHMS2075112-supplement-Supple-Table-1.xlsx (40.5KB, xlsx)
Supple-Table-2

Supplementary Table 2. GO molecular function analysis of the tumor cell population top DEGs by ENRICHR.

NIHMS2075112-supplement-Supple-Table-2.xlsx (151KB, xlsx)
Supple-Table-5

Supplementary Table 5. Percentage of various immune cell subpopulations in (–)CBG and (+)CBG samples.

NIHMS2075112-supplement-Supple-Table-5.xlsx (18.5KB, xlsx)
Supple-Table-4

Supplementary Table 4. List of DEGs observed in the immune cell subpopulations.

NIHMS2075112-supplement-Supple-Table-4.xlsx (274KB, xlsx)
Supple-Table-6

Supplementary Table 6. List of DEGs of the stromal cell subpopulations.

NIHMS2075112-supplement-Supple-Table-6.xlsx (12.9KB, xlsx)
Supple-Table-7

Supplementary Table 7. Enrichment analysis of the master regulators of the progenitor cells using ENRICHR.

NIHMS2075112-supplement-Supple-Table-7.xlsx (321.7KB, xlsx)
Supple-Figure

Supplementary Figure 1. (A) Comparative analysis of upstream regulators of the immune cell population in CBG-treated and untreated PRLomas implicated activation of IL2, and involvement of the inflammatory transcription factors NFκB and STATs in response to CBG treatment. (B) The master regulators of the progenitors were predicted using ENRICHR from ChEA library, and the bar chart shows the top 10 enriched terms along with their corresponding p-values. An asterisk (*) next to a p-value indicates the term also has a significant adjusted p-value (<0.05). (C) Dotplot depicted the specificity of SOX-2 and stromal markers TAGLN (fibroblasts) and VWF (endothelial). (D) Two dopamine receptors (DRD2, and DRD4) were detected by scRNAseq and DRD2 was the main subtype in the tumor population.

NIHMS2075112-supplement-Supple-Figure.pdf (280.6KB, pdf)

RESOURCES