p300-mediated histone H3K18 lactylation promotes mitochondrial ROS accumulation via mitophagy inhibition to potentiate dopamine agonists efficacy in prolactinomas

TL;DR

This study shows that boosting p300 activity improves the effectiveness of dopamine agonists in treating prolactinomas by increasing mitochondrial stress.

Contribution

The paper identifies a new p300–H3K18la–NDUFS7/WASH1 pathway that enhances dopamine agonist efficacy in prolactinomas.

Findings

p300 levels are inversely correlated with dopamine agonist resistance in prolactinomas.

p300-dependent H3K18la promotes mitochondrial ROS accumulation by inhibiting mitophagy.

Activating p300 synergizes with dopamine agonists to inhibit tumor growth in cell and animal models.

Abstract

Prolactinomas are the most common functional pituitary adenomas, and dopamine agonists (DAs) are the first-line therapy; however, approximately 10–30% of patients develop resistance, highlighting the need for effective sensitization strategies. In clinical specimens, we observed reduced p300 expression in tumors with poor DA responsiveness, and p300 levels were inversely associated with DA dosage. In cellular and xenograft models, DAs decreased p300 by suppressing the cAMP/PKA/CREB pathway. We therefore tested whether upregulating or activating p300 could enhance DA efficacy and investigated the underlying mechanism using immunohistochemistry, immunofluorescence, Western blot, genetic manipulations, RNA sequencing, CUT&Tag, ChIP–qPCR, Seahorse metabolic assays, flow cytometry, co-immunoprecipitation, and GST pull-down assays. Augmenting p300 markedly potentiated DA-induced antitumor effects in vitro and in vivo, a process accompanied by the elevated histone H3K18 lactylation (H3K18la). Mechanistically, p300-dependent H3K18la promoted transcriptional upregulation of Ndufs7 and Washc1. NDUFS7 induction was associated with increased mitochondrial ROS, whereas WASH1 bound the ubiquitin-associated domain of p62, impairing recognition and clearance of damaged mitochondria, suppressing mitophagy, and thereby sustaining mitochondrial ROS accumulation and apoptosis. Moreover, YF-2, a p300 HAT-domain activator, synergized with DAs to inhibit tumor growth in MMQ and AtT-20 cells. Together, these data identify a p300–H3K18la–NDUFS7/WASH1 axis that links mitophagy inhibition to mitochondrial ROS accumulation and provide a mechanistic rationale for targeting p300 as an adjuvant approach to improve DAs efficacy in prolactinomas.

Image 1

Figures

A). Experiments showed that p300 expression in rat tumor tissues was significantly reduced after treatment with dopamine agonists (BRC or cabergoline, CAB) ([Fig. 1](#fig1)H and I). We further validated this finding in vitro using pituitary adenoma cell lines (MMQ, AtT-20) and patient-derived primary prolactinoma cells (PA). Although AtT-20 cells differ from MMQ cells in subtype and hormone secretion profile, they were included in the study due to the expression of a certain level of dopamine D2 receptor (DRD2) on their cell membranes (Supplementary Fig. 1B). Results demonstrated that nuclear

A–M, Supplementary Fig. 3A–G).Fig. 2Upregulation of p300 synergizes with DA to promote cell apoptosis by increasing mitochondrial ROS in pituitary tumor cells. (A-B) Cell Counting Kit-8 (CCK-8) assay was used to detect cell viability in MMQ and AtT-20 cells transfected with empty vector or p300 overexpression plasmid (OE-p300) treated with BRC (10 μM) for different durations (0, 3, 6, 9, 12, 15, 18, 21, 24, 30, 36, 42, 48 h) (n = 3). (C-D) CCK-8 assay was used to detect cell viability in MMQ and AtT-20 cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) or CAB (25 μM) for

A–C). To determine whether cell apoptosis is induced by increased ROS, we co-treated MMQ and AtT-20 cells with upregulated or activated p300 with BRC and two ROS scavengers (N-acetylcysteine, NAC; reduced glutathione, GSH). Results showed that ROS scavengers reversed the pro-apoptotic effect induced by increased ROS ([Fig. 2](#fig2)J–L, Supplementary Fig. 4D–F) and attenuated the reduction in cell viability (Supplementary Fig. 4G–J).

WASH1 inhibits mitophagy by binding to the ubiquitin-associated (UBA) domain of p62 in pituitary tumor cells. (A) MMQ and AtT-20 cells in the vector group, OE-p300 group, OE-p300 with Washc1 knockdown group 1 (sh1+OE), and OE-p300 with Washc1 knockdown group 2 (sh2+OE) were treated with BRC (10 μM) for 48 h, followed by WB analysis to detect the expression of LC3BI/II, p62, translocase of outer mitochondrial membrane 20 (TOM20), translocase of inner mitochondrial membrane 23 (TIM23), and cytochrome *c* oxidase subunit IV (COX IV) (n = 3). (B) WASH1-Flag protein was overexpressed in AtT-20 cell

C-D, [Fig. 3](#fig3)K–M, Supplementary Fig. 7G–I). CoIP experiments further verified the interaction between p300 and H3K18la, and its HAT domain may mediate histone H3K18la in MMQ and AtT-20 cells ([Fig. 3](#fig3)N–P, Supplementary Fig. 6E). Finally, to exclude the influence of other p300-mediated acylation modifications, we used Galloflavin (Gal), a specific LDHA inhibitor [[[36]](#bib36), [[37]](#bib37), [[38]](#bib38)], to inhibit lactate production and further detected the effects of p300 upregulation combined with BRC on mitochondrial ROS and apoptosis. Results showed that inhibition of

A–D). In primary cells, activation of p300 combined with BRC also significantly promoted histone lactylation modification ([Fig. 3](#fig3)J, Supplementary Fig. 7E–F).

D–E). Flow cytometry results showed that Ndufs7 knockdown partially reversed the increase in mitochondrial ROS, intracellular hydrogen peroxide levels, and pro-apoptotic effect induced by p300 upregulation or activation combined with BRC ([Fig. 4](#fig4)J–L, Supplementary Fig. 10B–F). NDUFS7 is a key subunit of mitochondrial respiratory chain Complex I and serves as a core component maintaining Complex I assembly and electron transfer efficiency [[39](#bib39),[40](#bib40)]. Ndufs7 knockdown significantly inhibited the increase in Complex I electron leak mediated by p300 upregulation, suggestin

H, red arrows indicate autophagic mitochondria). Given that mitophagy can significantly affect intracellular ROS levels by regulating mitochondrial quality, we intersected the intersection genes in [Fig. 4](#fig4)F with the autophagy-related gene set and identified the Washc1 gene (WASP and SCAR homolog complex subunit 1, [Fig. 4](#fig4)P). Compared with the Vec + BRC group, the transcriptional level of Washc1 in the OE + BRC group was increased to 2.79 times that of the Vec + BRC group ([Fig. 4](#fig4)Q). Validation by qPCR and WB confirmed that compared with the Vec + BRC group, the OE + BRC

A). Molecular docking results suggested that WASH1 may bind to the ubiquitin-associated (UBA) domain of p62 ([Fig. 5](#fig5)C). Accordingly, we constructed GST-tagged UBA protein (GST-UBA), His-tagged wild-type WASH1 protein (His-WASH1-WT), and His-tagged mutant WASH1 protein (His-WASH1-Mut; with point mutations at arginine 25 and glutamic acid 28) in vitro. GST pull-down assay verified that the wild-type WASH1 protein could bind to the UBA domain of p62, while the mutant WASH1 protein failed to interact with the UBA protein (Supplementary Fig. 12B). Given that p62 acts as a ubiquitin adaptor

G–H). To verify whether YF-2 exerts a synergistic anti-tumor effect with DA by activating the HAT domain of p300, we detected mitochondrial ROS levels and cell apoptosis levels in MMQ and AtT-20 cells by flow cytometry. Results showed that compared with the single-drug group, YF-2 combined with BRC more significantly increased mitochondrial ROS levels and promoted cell apoptosis. Overexpression of mutant p300 (OE-Mut) reversed the further increase in mitochondrial ROS levels and cell apoptosis levels induced by overexpression of wild-type p300 (OE-WT) after combined treatment with BRC and YF-2

Full text

Introduction

1

Prolactinomas are the most common subtype of functional pituitary adenomas (PAs), accounting for approximately 50% of all PAs [1]. Currently, dopamine receptor agonists (DAs) serve as the first-line clinical therapy for prolactinoma. However, 10%–30% of patients develop resistance to DAs [2]. Drug-resistant prolactinomas are prone to recurrence following surgery or combined radiotherapy, which complicates clinical management [3].

Although numerous studies have explored the potential mechanisms of prolactinoma resistance from various perspectives, most have focused on reduced expression of the dopamine D2 receptor (DRD2) on tumor cell membranes [2,[4], [5], [6], [7], [8]]. In our previous work, we found that the expression of the acetyltransferase p300 was reduced in tumor tissues from drug-resistant patients. Specifically, p300 downregulates DRD2 transcription by decreasing histone H3K18 and H3K27 acetylation, ultimately leading to secondary resistance; upregulating p300 can partially restore DRD2 expression and thereby sensitize cells to DAs [6]. However, additional sensitization mechanisms may be involved in this process. Recent studies have shown that p300-mediated histone lactylation plays a crucial role in regulating gene transcription, tumor cell proliferation, differentiation, apoptosis, and drug resistance, providing a new avenue for exploring the multifaceted role of p300 in DA sensitization [[9], [10], [11], [12]].

A previous study indicated that the regulation of reactive oxygen species (ROS) levels is closely linked to prolactinoma resistance [7]. In the present study, we found that upregulating or activating p300 in combination with DAs more potently promotes mitochondrial ROS (mtROS) production, thereby inducing tumor cell apoptosis. Mitochondria are a primary source of intracellular ROS, and studies have confirmed that mitophagy plays a key role in maintaining homeostatic intracellular ROS levels [[13], [14], [15], [16], [17]]. Damaged mitochondria with high ROS levels can be cleared through mitophagy to preserve intracellular homeostasis. Among these processes, p62-mediated mitophagy is central: following ubiquitination of damaged mitochondria, p62 specifically recognizes ubiquitinated mitochondria via its ubiquitin-associated domain (UBA), targets them to autophagosomes, and ultimately facilitates autolysosome formation for degradation [[18], [19], [20]]. This process eliminates mitochondria-derived ROS and sustains physiological intracellular ROS levels; if disrupted, damaged mitochondria accumulate persistently, exacerbating ROS release and ultimately inducing pathological consequences such as tumor cell apoptosis.

In summary, this study elucidates the key mechanism by which p300-mediated histone H3K18 lactylation (H3K18la) modulates mitochondrial ROS levels through mitophagy. It also confirms for the first time that the p300 activator YF-2 can exert a synergistic antitumor effect with DAs by activating the p300 histone acetyltransferase (HAT) domain, providing a basis for future clinical translation.

Methods

2

Clinical samples collection

2.1

Tumor specimens from 32 bromocriptine-naive and 86 bromocriptine-resistant prolactinoma patients (surgical cases) were collected at Tongji Hospital (Wuhan) from January 2021 to January 2025. This study was approved by Tongji Hospital Ethics Committee (TJ-IRB20220325), and all patients provided written informed consent. The 86 medicated patients were divided into a relatively sensitive group (43 cases) and a relatively insensitive group (43 cases) based on the reduction in maximum tumor diameter before and after bromocriptine treatment. Clinical drug resistance was defined as failure of serum prolactin (PRL) to return to normal or recover gonadal function, plus <30% reduction in tumor diameter or <50% in volume, after 3–6 months of continuous tolerable-dose medication [21,22]. See Supplementary Methods for details.

Cell lines and primary cell culture

2.2

Rat prolactinoma cell line MMQ, growth hormone adenoma cell line GH3, and mouse corticotropin adenoma cell line AtT-20 were obtained from the American Type Culture Collection (ATCC). Mouse non-functional adenoma cell line TtT/GF and 293T cells were from Procell. All cell lines tested negative for mycoplasma using a mycoplasma detection kit (Servicebio) according to the manufacturer's instructions. MMQ cells were cultured in RPMI-1640 medium (Servicebio) supplemented with 10% fetal bovine serum (FBS, Gibco); GH3, AtT-20, TtT/GF, and 293T cells were cultured in DMEM medium (Servicebio) with 10% FBS, all cell lines maintained at 37 °C with 5% CO_2_ and subcultured regularly. Primary prolactinoma cells were isolated from human tumor tissues: tissues were rinsed with PBS (Servicebio) and minced into <1 mm^3^ fragments. After centrifugation, 4 mL of 0.25% collagenase (Servicebio) was added, and the mixture was incubated at 37 °C with 80 rpm shaking for 2 h. Following another centrifugation, red blood cell lysis buffer (Servicebio) was added for 1 min. Cells were washed 2–3 times with serum-free medium, resuspended in complete medium, filtered through a 40 μm strainer (Servicebio), and aliquoted for culture. See Supplementary Methods for details.

Drug treatment

2.3

Drugs used included bromocriptine, cabergoline, forskolin, CTB, estradiol, sodium lactate, N-acetylcysteine (NAC), reduced glutathione (GSH), TTK21, and YF-2, all from MedChemExpress (MCE). Powders were dissolved in DMSO (Servicebio) or PBS to prepare stock solutions according to the manufacturer's instructions, then diluted to working solutions with a DMSO ratio >1000:1. See Supplementary Methods for details.

Plasmid construction and transfection

2.4

Plasmids included wild-type p300 overexpression plasmid, HAT domain-mutated p300 overexpression plasmid, p300/Ndufs7/Washc1 knockdown/knockout/overexpression plasmids, and packaging plasmids psPAX2 and pMD2.G, all constructed by Tsingke Biological Technology. Transfection was performed as previously described [23] by using Lipomaster 3000 reagent (Vazyme) according to the manufacturer's instructions, and transfected cells were screened with puromycin (MCE). See Supplementary Methods for details.

Immunofluorescence (IF), immunohistochemistry (IHC), and tyramide signal amplification (TSA) staining

2.5

IF and IHC staining were performed as previously described [6]. For TSA staining (used when primary antibodies are from the same species), a TSA Plus kit (Servicebio) was used according to the manufacturer's instructions. See Supplementary Methods for details.

Western blot (WB)

2.6

WB was performed as previously described [6].See Supplementary Methods for details.

Co-immunoprecipitation (Co-IP)

2.7

Co-IP was performed as previously described [6]. See Supplementary Methods for details.

Endosome isolation

2.8

Endosome isolation used the Minute™ Kit (cat. no. ED-028, Invent Biotechnologies). Columns/tubes were pre-cooled on ice, with all centrifugation at 4 °C. Cultured cells were processed with Buffer A, ice-incubated, and column-passed. Samples were centrifuged sequentially to remove impurities, mixed with Buffer B, incubated, and centrifuged to collect endosomes. Endosomes were resuspended in dedicated lysis buffer, protein quantified by BCA assay, and used for WB. See Supplementary Methods for details.

Chromatin immunoprecipitation (ChIP)

2.9

ChIP was performed as previously described [6]. See Supplementary Methods for details.

Real-time quantitative Polymerase chain reaction (qPCR)

2.10

qPCR was performed as previously described [6]. See Supplementary Methods for details.

Cell Viability Assay

2.11

Cell Viability Assay was performed as previously described [6]. See Supplementary Methods for details.

Flow cytometry

2.12

Mitochondrial reactive oxygen species (ROS) was investigated by using MitoSOX™ Green Mitochondrial Superoxide Indicator (Thermo Fisher) according to the manufacturer's instructions. Cell apoptosis was investigated by using an Annexin V-FITC/PI apoptosis assay kit (Vazyme) according to the manufacturer's instructions. Both detections used a flow cytometer (Beckman Coulter), and data were analyzed with FlowJo 10.8. See Supplementary Methods for details.

Biomolecule detection

2.13

Hydrogen peroxide was investigated by using a hydrogen peroxide detection kit (Beyotime) according to the manufacturer's instructions. Intracellular cAMP and supernatant PRL, ACTH were investigated by using enzyme-linked immunosorbent assay kits (Elabscience) according to the manufacturer's instructions. PKA activity was investigated by using a PKA kinase activity detection kit (Abcam) according to the manufacturer's instructions. Acetyl-CoA, lactic acid, and mitochondrial complex I activity were investigated by using detection kits (Elabscience) according to the manufacturer's instructions. OD values were measured, and concentrations/activities were calculated via standard curves. See Supplementary Methods for details.

Seahorse cell energy metabolism assay

2.14

Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured by using Seahorse kits (Agilent Technologies) according to the manufacturer's instructions. Data were analyzed with Seahorse Wave Controller 2.6. See Supplementary Methods for details.

Detection of mitochondrial complex I electron leak

2.15

Mitochondrial Complex I electron leak in cell samples was detected via high-resolution respirometry-fluorometry, based on/optimized from Salazar-Ramírez et al. and Makrecka-Kuka et al. [24,25] Log-phase cells were washed with pre-cooled PBS, permeabilized with digitonin on ice (validated by trypan blue assay). Cell suspension was added to an Oroboros Oxygraph-2k chamber containing MiR05 buffer, Amplex Red, HRP, and SOD post equilibration and stabilization for baseline recording. Complex I LEAK respiration (no ADP) was induced with specific substrates; OXPHOS was initiated with saturating ADP, ETS capacity determined by incremental uncoupler titration, and ROX measured via mitochondrial ETC inhibitors. Oxygen flux and fluorescence were recorded simultaneously; H_2_O_2_ flux was calibrated with standard additions. Electron leak was calculated as the ratio of corrected net H_2_O_2_ flux to mitochondrial net oxygen flux, normalized per 10^6^ cells. Technical replicates were included, with data processed using GraphPad Prism. See Supplementary Methods for details.

Mitochondrial transmission electron microscopy

2.16

Cells were fixed with electron microscopy fixative (Servicebio), post-fixed with osmium tetroxide, dehydrated, embedded, sectioned (60–80 nm), stained, and imaged with a Hitachi HT7800 microscope. See Supplementary Methods for details.

In vitro GST pull-down assay

2.17

Protein–protein interaction between p62 UBA and WASH1 was examined using the Pierce™ GST Protein Interaction Pull-Down Kit (Thermo Scientific, Cat. No. 21516) according to the manufacturer's instructions. Recombinant GST-p62 UBA and His-WASH1-WT/His-WASH1-Mut proteins were constructed and purified by Zoonbio Biotechnology. Eluates and inputs were analyzed by SDS–PAGE and Western blotting (anti-His for pulled-down His-WASH1; anti-GST for bait loading). See Supplementary Methods for details.

Mt-Keima adenoviral transduction and ratiometric quantification of mitophagy

2.18

Mitophagy was quantified using mt-Keima adenovirus (HanBio, Cat. No. HBAD001724OE) following the manufacturer-recommended protocol. After infection and indicated treatments, mt-Keima signals were measured on an Infinite® M Plex microplate reader (Tecan, Cat. No. 30190085) and expressed as the ratiometric index RFU(Ex550/Em620)/RFU(Ex440/Em620). See Supplementary Methods for details.

Generation of autophagy LC3 HiBiT reporter stable cells and bulk autophagic flux measurement

2.19

Stable autophagy reporter cells were generated using the Autophagy LC3 HiBiT Reporter Vector and Detection System (Promega, Cat. No. GA2550). Bulk autophagic flux was determined using the Nano-Glo® HiBiT Lytic Detection System (Promega, Cat. No. N3030) with or without bafilomycin A1 (MedChemExpress, Cat. No. HY-100558; final 50 nM, 4 h), and luminescence was recorded on an Infinite® M Plex microplate reader (Tecan, Cat. No. 30190085). Autophagic flux was calculated as ΔRLU = RLU(+BafA1) − RLU(−BafA1). See Supplementary Methods for details.

Ratiometric measurement of lysosomal pH

2.20

Lysosomal acidification was assessed using LysoSensor™ Yellow/Blue DND-160 (Thermo Fisher Scientific, Cat. No. L7545) according to the manufacturer's instructions. Fluorescence was measured on an Infinite® M Plex microplate reader (Tecan, Cat. No. 30190085) and expressed as RFU (Ex380/Em535)/RFU (Ex340/Em535) after subtraction of background from unstained wells. See Supplementary Methods for details.

Animal models

2.21

Models included subcutaneous xenografts (6-week-old male BALB/c nude mice, axillary cell inoculation, intraperitoneal treatment for 2 weeks) and orthotopic prolactinomas (4-week-old female Fischer 344 rats, ovariectomy, estradiol pumps from DURECT, same treatment). This experiment was performed as previously described [23]. Approved by Tongji Hospital Ethics Committee (TJH-202206015). See Supplementary Methods for details.

Bulk RNA sequencing/CUT&tag/IP-MS

2.22

Bulk RNA sequencing: RNA was extracted by using TRIzol (TIANGEN) according to the manufacturer's instructions, and qualified RNA was sequenced on Illumina Xplus. CUT&Tag: samples were treated with beads (Thermo Fisher) according to the manufacturer's instructions, then sequenced on Illumina NovaSeq Xplus. IP-MS: proteins were digested by using trypsin (Promega) according to the manufacturer's instructions, then detected via UHPLC-MS (Thermo Fisher). See Supplementary Methods for details.

Organ toxicity assessment

2.23

Nude mice bearing MMQ cell-derived subcutaneous xenografts were randomly divided into the control (i.p. PBS) and BRC + YF-2 (i.p. 10 mg/kg/d + 20 mg/kg/d) groups, with daily administration for 2 weeks. After treatment, serum was collected via centrifugation (3000 rpm, 15 min, 2-8 °C) and stored at −80 °C (avoid repeated freeze-thaw). Mice were sacrificed to isolate 10 organs (brain, heart, liver, spleen, lung, kidney, large intestine, small intestine, testis, ovary), which were fixed and subjected to H&E staining using Servicebio reagents (cat. no. G1101, G1128, G1076) and equipment. Serum ALT, AST, UREA, and CRE were detected with Leadman kits (cat. no. S03030, S03040, S03036, S03076) on a Chemray 800 analyzer. Pathological and biochemical results were interpreted against mouse normal reference ranges to assess organ toxicity. See Supplementary Methods for details.

Data analysis

2.24

Data were analyzed using GraphPad Prism 8.0.2. All data were tested for normality using the Shapiro-Wilk test (n ≤ 50) or Kolmogorov-Smirnov test (n > 50). Data conforming to normal distribution were expressed as mean ± standard deviation (SD). Pearson correlation analysis was used for correlation between two variables; independent sample t-test was used for comparison between two groups; one-way analysis of variance (ANOVA) was used for comparison among multiple groups. A p-value <0.05 was considered statistically significant, denoted as: ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05.

Results

3

DA downregulates p300 expression by inhibiting the cAMP/PKA/CREB pathway in pituitary tumor cells

3.1

To initially explore the association between p300 and bromocriptine (BRC)-mediated acquired resistance in prolactinomas, we collected tumor tissue samples from 86 patients with BRC-resistant prolactinomas. Results showed a moderate positive correlation between the percentage reduction in serum prolactin (PRL) levels and the percentage reduction in maximum tumor diameter before and after BRC treatment (Fig. 1A), indicating consistency between tumor volume changes and hormonal improvement. Based on the median percentage reduction in maximum tumor diameter, patients were divided into a relatively sensitive group (43 cases) and a relatively insensitive group (43 cases). Immunohistochemical (IHC) staining for p300 in tumor tissues revealed that p300 expression was significantly lower in the relatively insensitive group than in the relatively sensitive group (Fig. 1B and D), suggesting that low p300 expression may be associated with BRC-induced acquired resistance. Additionally, we collected tumor tissues from 32 prolactinoma patients who underwent direct surgical resection without prior BRC treatment. IHC results showed that p300 expression was significantly lower in tumor tissues from BRC-treated patients than in those from untreated patients (Fig. 1C and D), indicating that BRC exposure may be a key inducer of p300 downregulation. Meanwhile, a moderate negative correlation was observed between the total BRC dosage and p300 expression in patient tumor tissues (Fig. 1E), further confirming that BRC-mediated p300 downregulation is dose-dependent.Fig. 1DA downregulates p300 expression by inhibiting the cAMP/PKA/CREB pathway in pituitary tumor cells. (A) Correlation analysis was performed between the percentage reduction in tumor maximum diameter and the percentage decrease in serum prolactin (PRL) levels before and after bromocriptine (BRC) treatment in 86 patients with drug-resistant prolactinomas. Patients were classified into a relatively sensitive group (n = 43) and a relatively insensitive group (n = 43). (B) Immunohistochemistry (IHC) staining was used to detect p300 expression in the above two groups of patients. (C) IHC staining was performed to detect p300 expression in the BRC-untreated group (n = 32) and BRC-treated group (n = 86). (D) Representative IHC staining images of p300 in the BRC-untreated group, relatively sensitive group, and relatively insensitive group. (E) Correlation analysis was conducted between the total BRC dosage administered and p300 expression in 86 treated patients. (F) Flowchart showing the establishment of an orthotopic pituitary prolactinoma model in rats. (G) Representative images of successfully established orthotopic pituitary prolactinoma in rats. (H–I) After successful establishment of orthotopic pituitary prolactinoma in rats, the rats were intraperitoneally (i.p.) injected with PBS, BRC (10 mg/kg/d), or cabergoline (CAB, 15 mg/kg/d) for 2 weeks. Immunofluorescence (IF) staining was used to detect p300 expression levels in tumor tissues (n = 4). (J-K) Western blot (WB) analysis was performed to detect nuclear p300 expression in MMQ and PA cells treated with BRC (0, 5, 10 μM) or CAB (0, 12.5, 25 μM) for 48 h (n = 3). (L-M) WB analysis was used to detect nuclear p300 expression in MMQ and PA cells treated with BRC (10 μM) or CAB (25 μM) for different durations (0, 24, 48 h) (n = 3). MMQ and PA cells were treated with BRC (10 μM), CAB (25 μM), forskolin (FSK, 50 μM), BRC (10 μM) + FSK (50 μM), or CAB (25 μM) + FSK (50 μM) for 48 h. (N–O) Intracellular cyclic adenosine monophosphate (cAMP) concentrations in MMQ and PA cells were detected using a cAMP ELISA kit (n = 5). (P-Q) WB analysis was performed to detect the expression of total protein kinase A catalytic subunit (PKA-C), total cAMP response element-binding protein (t-CREB), phosphorylated CREB (p-CREB), CREB-binding protein (CBP), and p300 in MMQ and PA cells (n = 3). (R) Schematic diagram of the cAMP/PKA/CREB/p300 signaling pathway mediated by DA acting on dopamine D2 receptors (DRD2). Pearson correlation analysis was used for correlation between two variables. Independent sample t-test was used for comparison between two groups. One-way analysis of variance (ANOVA) was used for comparison among multiple groups. Data are presented as the mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.Fig. 1

Next, we established a rat orthotopic prolactinoma model (Fig. 1F and G, Supplementary Fig. 1A). Experiments showed that p300 expression in rat tumor tissues was significantly reduced after treatment with dopamine agonists (BRC or cabergoline, CAB) (Fig. 1H and I). We further validated this finding in vitro using pituitary adenoma cell lines (MMQ, AtT-20) and patient-derived primary prolactinoma cells (PA). Although AtT-20 cells differ from MMQ cells in subtype and hormone secretion profile, they were included in the study due to the expression of a certain level of dopamine D2 receptor (DRD2) on their cell membranes (Supplementary Fig. 1B). Results demonstrated that nuclear p300 expression was significantly downregulated in all three cell types after treatment with 10 μM BRC or 25 μM CAB for 24 h (Fig. 1J–M, Supplementary Fig. 1C–D).

Finally, we explored the potential molecular mechanism by which DAs downregulate p300. DAs are known to reduce PRL transcription primarily by inhibiting the cAMP/PKA/CREB pathway, thereby alleviating endocrine symptoms [26,27]; p300 often forms a complex with CREB-binding protein (CBP) and is recruited to the promoter regions of nuclear genes by phosphorylated CREB (p-CREB) to regulate transcription [28,29]. This suggests that DA may downregulate p300 expression by inhibiting the cAMP/PKA/CREB signaling pathway. To verify this, we treated MMQ, AtT-20 cells, and primary cells with DAs and detected the expression of proteins related to the cAMP/PKA/CREB pathway by Western blot (WB). Results showed that DA treatment significantly reduced cytoplasmic cAMP levels, the activity and expression of the PKA catalytic subunit (PKA-C), and the expression of total CREB (t-CREB). Meanwhile, nuclear expression of p-CREB, CBP, and p300 was also significantly decreased. Co-treatment with forskolin (FSK), a cAMP agonist, significantly reversed the DA-induced downregulation of PKA/CREB/p300 pathway-related proteins (Fig. 1N–Q, Supplementary Fig. 1F–J). We simultaneously detected cytoplasmic and nuclear p300 expression in MMQ and AtT-20 cells treated with DA, and speculated that DA may reduce p300 nuclear recruitment by inhibiting the downstream PKA/CREB/p300 pathway, thereby downregulating nuclear p300 expression (Supplementary Fig. 1E). Co-immunoprecipitation (Co-IP) experiments further confirmed that p-CREB, CBP, and p300 can form a complex (Supplementary Fig. 1K). A schematic diagram of this mechanism is shown in Fig. 1R.

Upregulation of p300 synergizes with DA to promote cell apoptosis by increasing mitochondrial ROS in pituitary tumor cells

3.2

Previous studies have confirmed that p300 overexpression (OE-p300) combined with DA treatment exerts a more significant anti-tumor effect [6]. In this section, we first replicated this phenotypic experiment and used N-(4-chloro-3-trifluoromethylphenyl)-2-ethoxybenzamide (CTB), an activator of the p300 histone acetyltransferase (HAT) domain, to verify through in vitro and in vivo experiments that upregulation or activation of p300 combined with DA treatment significantly reduced cell viability, increased cell apoptosis, inhibited tumor proliferation, and reduced tumor volume in MMQ cells, AtT-20 cells, and primary cells (Fig. 2A–D, Supplementary Fig. 2A–M, Supplementary Fig. 3A–G).Fig. 2. Upregulation of p300 synergizes with DA to promote cell apoptosis by increasing mitochondrial ROS in pituitary tumor cells. (A-B) Cell Counting Kit-8 (CCK-8) assay was used to detect cell viability in MMQ and AtT-20 cells transfected with empty vector or p300 overexpression plasmid (OE-p300) treated with BRC (10 μM) for different durations (0, 3, 6, 9, 12, 15, 18, 21, 24, 30, 36, 42, 48 h) (n = 3). (C-D) CCK-8 assay was used to detect cell viability in MMQ and AtT-20 cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) or CAB (25 μM) for 48 h (n = 5). (E) AtT-20 cells (transfected with empty vector or OE-p300) were treated with BRC (10 μM) for 24 h, followed by bulk RNA sequencing and KEGG pathway enrichment analysis of differentially upregulated genes (n = 3). (F–I) Flow cytometry was used to detect mitochondrial ROS levels in MMQ and AtT-20 cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) or CAB (25 μM) for 48 h. Flow cytometry was used to detect mitochondrial ROS levels in PA cells treated with BRC (10 μM), CAB (25 μM), N-(4-chloro-3-triffuoromethyl-phenyl)-2- ethoxy-benzamide (CTB, 50 μM), BRC (10 μM) + CTB (50 μM), or CAB (25 μM) + CTB (50 μM) for 48 h (n = 3). (J-L) Flow cytometry was used to detect cell apoptosis levels in MMQ and AtT-20 cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) alone or combined with N-acetylcysteine (NAC, 5 mM) or glutathione (GSH, 2 mM) for 48 h (n = 3). (M) Schematic diagram of molecular structures of wild-type p300 and histone acetyltransferase (HAT) domain-mutated p300. (N–S) Flow cytometry was used to detect mitochondrial ROS levels and apoptosis levels in MMQ and AtT-20 cells (transfected with empty vector, wild-type p300 overexpression plasmid [OE-WT-p300], or HAT domain-mutated p300 overexpression plasmid [OE-Mut-p300]) treated with BRC (10 μM) for 48 h (n = 3). One-way ANOVA was used for comparison among multiple groups. Data are presented as the mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.Fig. 2

To explore the potential mechanism underlying this synergistic effect, we treated AtT-20 cells with empty vector (Vec) or p300 overexpression (OE) with BRC for 24 h and performed bulk RNA sequencing (RNA-seq). KEGG pathway enrichment analysis of differentially upregulated genes between the two groups (top 20 pathways) showed significant enrichment of oxidative phosphorylation, ROS, and apoptosis-related signaling pathways (Fig. 2E). Excessive ROS accumulation in tumor cells is known to induce programmed cell death, thereby exerting anti-tumor effects, suggesting that ROS may play a key role in mediating cell apoptosis. To confirm this, we treated MMQ cells, AtT-20 cells, and primary cells with upregulated or activated p300 with DA and detected mitochondrial ROS levels by flow cytometry. Results showed that DA significantly increased mitochondrial ROS levels (consistent with previous studies) [7,30], and p300 upregulation or activation further elevated ROS levels (Fig. 2F–I, Supplementary Fig. 4A–C). To determine whether cell apoptosis is induced by increased ROS, we co-treated MMQ and AtT-20 cells with upregulated or activated p300 with BRC and two ROS scavengers (N-acetylcysteine, NAC; reduced glutathione, GSH). Results showed that ROS scavengers reversed the pro-apoptotic effect induced by increased ROS (Fig. 2J–L, Supplementary Fig. 4D–F) and attenuated the reduction in cell viability (Supplementary Fig. 4G–J).

Finally, we established stable MMQ and AtT-20 cell lines overexpressing HAT domain-mutated p300 (OE-Mut, Fig. 2M) to explore the role of the p300 HAT domain in the above processes. We treated p300 wild-type overexpression (OE-WT) and OE-Mut cells with BRC and detected mitochondrial ROS and cell apoptosis levels by flow cytometry. Results showed that p300 HAT domain mutation eliminated the significant differences in mitochondrial ROS and cell apoptosis levels between the OE-WT and BRC-only groups (Fig. 2N–S), indicating that the p300 HAT domain plays a key role in regulating mitochondrial ROS levels and cell apoptosis.

Upregulation of p300 synergizes with DA to promote histone H3K18 lactylation in pituitary tumor cells

3.3

Mitochondria are the main source of intracellular ROS, and excessive ROS accumulation may damage mitochondrial function and structure [31,32]. To verify this, we treated Vec and OE groups of MMQ and AtT-20 cells with BRC for 48 h and observed mitochondrial morphological changes by transmission electron microscopy. Results showed that mitochondria in the BRC-only group maintained basic structural integrity, with clear inner and outer mitochondrial membranes and mitochondrial cristae. In contrast, mitochondria in the p300 upregulation combined with BRC group exhibited typical vacuolar degeneration and impaired structural integrity (Fig. 3A, Supplementary Fig. 5A, indicated by red arrows). To explore the impact of mitochondrial structural damage on cellular metabolism, we detected the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) using a Seahorse XF Analyzer. Results showed that compared with the single-drug group, the p300 upregulation combined with BRC group exhibited significantly reduced basal respiration rate, ATP production rate, and maximum respiration rate, indicating decreased cellular oxidative phosphorylation levels. Meanwhile, the basal glycolysis rate was significantly increased, suggesting that mitochondrial dysfunction induces a metabolic shift toward glycolysis (Fig. 3B and C, Supplementary Fig. 5B–M).Fig. 3. Upregulation of p300 synergizes with DA to promote histone H3K18 lactylation in pituitary tumor cells. (A) Transmission electron microscopy (TEM) was used to observe mitochondrial morphological changes in MMQ cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) for 48 h, the structures indicated by the red arrows are the mitochondria. (n = 3). (B) MMQ cells (transfected with empty vector or OE-p300) were treated with BRC (10 μM) for 48 h, followed by detection of extracellular acidification rate (ECAR) using a Seahorse cell energy metabolism assay. (C) Glycolytic baseline and glycolytic capacity were calculated based on ECAR (n = 3). (D-G) MMQ and AtT-20 cells (transfected with empty vector or OE-p300) were treated with BRC (10 μM) for different durations (0, 3, 6, 12, 18, 24, 48, 72 h). (D, F) Intracellular lactic acid and acetyl-CoA in MMQ and AtT-20 cells were detected using a lactic acid detection kit and acetyl-CoA detection kit, and the OE-p300 group at each time point was normalized using the Vec + BRC group ("nc" in the figure) from the corresponding time point. (E, G) The ratios of lactic acid to acetyl-CoA contents in MMQ and AtT-20 cells were calculated separately (n = 3). (H–I) WB analysis was used to detect pan lactyl-lysine and pan acetyl-lysine protein expression in MMQ (H) and AtT-20 (I) cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) for 24 h or 48 h (n = 3). (J) WB analysis was used to detect pan lactyl-lysine and pan acetyl-lysine protein expression in PA cells treated with BRC (10 μM) or BRC (10 μM) + CTB (50 μM) for 24 h or 48 h (n = 3). (K-L) WB analysis was used to detect H3K18 lactylation (H3K18la) expression in MMQ and AtT-20 cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) or CAB (25 μM) for 48 h (n = 3). (M) WB analysis was used to detect H3K18la expression in PA cells treated with BRC (10 μM), CAB (25 μM), CTB (50 μM), BRC (10 μM) + CTB (50 μM), or CAB (25 μM) + CTB (50 μM) for 48 h (n = 3). (N) Schematic diagram of molecular structures of wild-type p300 and HAT domain-mutated p300. (O) WB analysis was used to detect H3K18la expression in MMQ and AtT-20 cells (transfected with empty vector, OE-WT-p300, or OE-Mut-p300) treated with BRC (10 μM) for 48 h (n = 3). (P) Co-IP experiments were performed using anti-p300 antibodies in MMQ and AtT-20 cells (transfected with OE-WT-p300 or OE-Mut-p300) treated with BRC (10 μM) for 48 h, followed by WB analysis to detect H3K18la expression (n = 3). One-way ANOVA was used for comparison among multiple groups. Data are presented as the mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.Fig. 3

Acetyl-CoA (Ac-CoA) and lactate are key metabolites of oxidative phosphorylation and glycolysis, respectively, and both can serve as substrates for p300-mediated protein acylation. As reported in the literature, different substrate concentrations can affect the type of p300-mediated protein modification [[33], [34], [35]]. Therefore, we detected changes in intracellular lactate and Ac-CoA levels in MMQ and AtT-20 cells. Results showed that after BRC treatment, lactate levels in p300-upregulated MMQ and AtT-20 cells continued to increase with prolonged BRC exposure (12 h to 48 h) compared with the Vec group. In contrast, Ac-CoA levels slightly increased at 12 h of BRC treatment (possibly related to enhanced oxidative phosphorylation induced by mild mitochondrial oxidative stress) and significantly decreased at 48 h in both cell types (possibly related to impaired mitochondrial structure and function). Meanwhile, the ratio of lactate to Ac-CoA gradually increased with prolonged BRC treatment, indicating a gradual increase in the proportion of glycolytic metabolism and further lactate accumulation (Fig. 3D–G). Given that significantly enhanced glycolytic levels suggest a potential metabolic shift, we detected the expression levels of lactate dehydrogenase A (LDHA) and lactate dehydrogenase B (LDHB) by WB and found that under BRC treatment, p300-upregulated cells showed significantly increased LDHA levels and significantly decreased LDHB levels, further verifying the occurrence of metabolic reprogramming (Supplementary Fig. 6A). Additionally, we detected the expression levels of lactate shuttle-related enzymes (monocarboxylate transporter 1/4, MCT1/4) by WB, excluding the possibility that BRC or p300 upregulation increases exogenous lactate uptake (Supplementary Fig. 6B).

Intracellular accumulated lactate can serve as a direct substrate for p300-mediated protein lactylation. Therefore, we detected the expression of pan-lysine acetylation/lactylation proteins by WB. Results showed that after 48 h of BRC treatment, significant protein acetylation and lactylation modifications near the molecular weight of histones (15 kDa) were observed in both cell types. Compared with the Vec group, p300 upregulation significantly promoted histone lactylation modification in both cell types (Fig. 3H and I, Supplementary Fig. 7A–D). In primary cells, activation of p300 combined with BRC also significantly promoted histone lactylation modification (Fig. 3J, Supplementary Fig. 7E–F).

Next, we screened common lysine sites of histone lactylation modification by WB and confirmed that lactylation modification of lysine 18 on histone H3 (H3K18la) was significantly upregulated after p300 upregulation combined with DA treatment (Supplementary Fig. 6C-D, Fig. 3K–M, Supplementary Fig. 7G–I). CoIP experiments further verified the interaction between p300 and H3K18la, and its HAT domain may mediate histone H3K18la in MMQ and AtT-20 cells (Fig. 3N–P, Supplementary Fig. 6E). Finally, to exclude the influence of other p300-mediated acylation modifications, we used Galloflavin (Gal), a specific LDHA inhibitor [[36], [37], [38]], to inhibit lactate production and further detected the effects of p300 upregulation combined with BRC on mitochondrial ROS and apoptosis. Results showed that inhibition of lactate production significantly reduced mitochondrial ROS levels and apoptosis rate, indirectly confirming that p300-mediated lactylation modification plays a more critical role (Supplementary Fig. 8A–C).

Upregulation of p300 synergizes with DA to elevate mitochondrial ROS levels via H3K18la-mediated transcription of Ndufs7 and Washc1 in pituitary tumor cells

3.4

To further clarify the association between target genes regulated by p300-mediated H3K18la and mitochondrial ROS, we divided AtT-20 cells into four groups: Vec, Vec + L-Na, Vec + BRC, and OE + BRC, performed bulk RNA sequencing, and screened the set of genes upregulated by p300-mediated histone lactylation modification (Fig. 4A–D). Subsequently, we used CUT&Tag technology in the Vec + BRC and OE + BRC groups of AtT-20 cells to screen the set of genes regulated by H3K18la modification (Fig. 4E and F). Intersection of the above two sets of differentially upregulated genes with the genes screened in Fig. 4C yielded the set of genes potentially upregulated by p300-mediated H3K18la (Fig. 4F). Further intersection of this gene set with the ROS-related gene set identified the Ndufs7 gene (NADH dehydrogenase (ubiquinone) iron-sulfur protein 7, Fig. 4G). Compared with the Vec + BRC group, the transcriptional level of Ndufs7 in the OE + BRC group was increased to 2.14 times that of the Vec + BRC group (Fig. 4H). Validation by quantitative real-time PCR (qPCR) and WB confirmed that compared with the Vec + BRC group, the OE + BRC group significantly upregulated the transcriptional level of Ndufs7 and its protein level in mitochondria (Supplementary Fig. 9A–B), which was consistent with the results of in vivo tumor xenograft experiments (Supplementary Fig. 9C). Chromatin immunoprecipitation-qPCR (ChIP-qPCR) further confirmed that H3K18la can bind to the promoter region of the Ndufs7 gene. Moreover, reducing lactate content inhibits the enrichment of H3K18la at the Ndufs7 promoter region and suppresses its transcription (Fig. 4I, Supplementary Fig. 9A, Supplementary Fig. 10A).Fig. 4. Upregulation of p300 synergizes with DA to elevate mitochondrial ROS levels via H3K18la-mediated transcription of Ndufs7 and Washc1 in pituitary tumor cells. (A-D) AtT-20 cells (transfected with empty vector or OE-p300) were treated with sodium lactate (5 mM) or BRC (10 μM) for 24 h, followed by bulk RNA sequencing. (A) Volcano plot showing the distribution of differentially expressed genes in the vector + sodium lactate group (Vec + L-Na) compared with the vector group (Vec). (B) Volcano plot showing the distribution of differentially expressed genes in the OE-p300+BRC group compared with the Vec + BRC group. (C) Venn diagram of the intersection of upregulated differentially expressed genes from groups A and B. (D) KEGG pathway enrichment of the 2273 genes obtained from the intersection in C (n = 3). (E) Peak plot showing H3K18la enrichment in genome-wide promoter regions of cells from the OE-p300+BRC group and Vec + BRC group in AtT-20 cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) for 24 h. (F) Venn diagram of the intersection between the 2273 genes obtained from the intersection in C and the upregulated differentially expressed genes enriched by CUT&Tag using anti-H3K18la antibody in AtT-20 cells (transfected with OE-p300) treated with BRC (10 μM) for 24 h. (G) Venn diagram of the intersection between the 36 genes obtained from the intersection in F and the ROS-related gene set (KEGG: mmu05208). (H) Peak plot showing H3K18la enrichment in the NADH dehydrogenase [ubiquinone] Fe–S protein 7 (Ndufs7) promoter region in the Vec + BRC group and OE-p300+BRC group in AtT-20 cells. (I) Chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR) of H3K18la enrichment in the Ndufs7 promoter region in the Vec + BRC group, OE-p300+BRC group, and OE-p300+BRC + Gal group in AtT-20 cells (n = 3). AtT-20 cells in the vector group were treated with BRC (10 μM) or BRC (10 μM) + CTB (50 μM) for 48 h. AtT-20 cells in the OE-p300 group, OE-p300 with Ndufs7 knockdown group 1 (sh1+OE), and OE-p300 with Ndufs7 knockdown group 2 (sh2+OE) were treated with BRC (10 μM) for 48 h, followed by flow cytometry to detect mitochondrial ROS levels (J), hydrogen peroxide detection kit to measure intracellular hydrogen peroxide levels (K), and flow cytometry to detect cell apoptosis levels (L) (n = 3). (M) After AtT-20 cells in the vector group, OE-p300 group, OE-p300 with Ndufs7 knockdown group 1 (sh1+OE), and OE-p300 with Ndufs7 knockdown group 2 (sh2+OE) were treated with BRC (10 μM) for 36 h, electron leak was detected via high-resolution respirometry combined with fluorometry (n = 3). (N) AtT-20 cells (transfected with empty vector or OE-p300) were treated with BRC (10 μM) for 24 h, followed by bulk RNA sequencing and KEGG pathway enrichment analysis of differentially downregulated genes (n = 3). (O) TEM was used to observe mitochondrial morphological changes in AtT-20 cells (transfected with empty vector or OE-p300) treated with BRC (10 μM) for 24 h, the process indicated by the red arrows is mitophagy. (n = 3). (P) Venn diagram of the intersection between the 36 genes obtained from the intersection in F and the autophagy-related gene set (GO:0010506). (Q) Peak plot showing H3K18la enrichment in the WASH complex subunit 1 (Washc1) promoter region in the vector group and OE-p300+BRC group in AtT-20 cells. (R) ChIP-qPCR of H3K18la enrichment in the Washc1 promoter region in the Vec + BRC group, OE-p300+BRC group, and OE-p300+BRC + Gal group in AtT-20 cells (n = 3). AtT-20 cells in the vector group were treated with BRC (10 μM) or BRC (10 μM) + CTB (50 μM) for 48 h. AtT-20 cells in the OE-p300 group, OE-p300 with Washc1 knockdown group 1 (sh1+OE), and OE-p300 with Washc1 knockdown group 2 (sh2+OE) were treated with BRC (10 μM) for 48 h, followed by flow cytometry to detect mitochondrial ROS levels (S), hydrogen peroxide detection kit to measure intracellular hydrogen peroxide levels (T), and flow cytometry to detect cell apoptosis levels (U) (n = 3). (V) Schematic diagram showing that NDUFS7 increases mitochondrial ROS and WASH1 inhibits mitophagy, thereby inducing cell apoptosis. One-way ANOVA was used for comparison among multiple groups. Data are presented as the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.Fig. 4

To clarify the impact of NDUFS7 on mitochondrial ROS and cell apoptosis, we established Ndufs7 knockdown stable cell lines in p300-upregulated MMQ and AtT-20 cells (Supplementary Fig. 9D–E). Flow cytometry results showed that Ndufs7 knockdown partially reversed the increase in mitochondrial ROS, intracellular hydrogen peroxide levels, and pro-apoptotic effect induced by p300 upregulation or activation combined with BRC (Fig. 4J–L, Supplementary Fig. 10B–F). NDUFS7 is a key subunit of mitochondrial respiratory chain Complex I and serves as a core component maintaining Complex I assembly and electron transfer efficiency [39,40]. Ndufs7 knockdown significantly inhibited the increase in Complex I electron leak mediated by p300 upregulation, suggesting that increased mitochondrial ROS production may be associated with enhanced Complex I electron leak induced by NDUFS7 upregulation (Fig. 4M, Supplementary Fig. 10G).

Notably, Ndufs7 knockdown only partially reversed the increase in ROS induced by p300 upregulation or activation combined with BRC, suggesting the existence of other mechanisms mediating increased cellular ROS. Therefore, we performed KEGG pathway enrichment analysis of differentially downregulated genes between the Vec + BRC and OE + BRC groups and found significant enrichment of autophagy and mitophagy-related pathways (Fig. 4N). Meanwhile, electron microscopy results showed that mitophagy levels were lower in the OE + BRC group than in the Vec + BRC group (Fig. 4O, Supplementary Fig. 10H, red arrows indicate autophagic mitochondria). Given that mitophagy can significantly affect intracellular ROS levels by regulating mitochondrial quality, we intersected the intersection genes in Fig. 4F with the autophagy-related gene set and identified the Washc1 gene (WASP and SCAR homolog complex subunit 1, Fig. 4P). Compared with the Vec + BRC group, the transcriptional level of Washc1 in the OE + BRC group was increased to 2.79 times that of the Vec + BRC group (Fig. 4Q). Validation by qPCR and WB confirmed that compared with the Vec + BRC group, the OE + BRC group significantly upregulated Washc1 expression (Supplementary Fig. 9F–G), which was consistent with the results of in vivo tumor xenograft experiments (Supplementary Fig. 9H). ChIP-qPCR also confirmed that H3K18la can bind to the promoter region of the Washc1 gene. Moreover, reducing lactate content inhibits the enrichment of H3K18la at the Washc1 promoter region and suppresses its transcription (Fig. 4R and Supplementary Fig. 9F, Supplementary Fig. 10I).

We further established Washc1 knockdown stable cell lines in p300-upregulated MMQ and AtT-20 cells (Supplementary Fig. 9I–J). Flow cytometry detection of mitochondrial ROS and cell apoptosis levels showed that Washc1 knockdown partially reversed the increase in ROS levels and pro-apoptotic effect induced by p300 upregulation or activation combined with BRC (Fig. 4S–U, Supplementary Fig. 10J–N).

Finally, to confirm that Ndufs7 and Washc1 are the main drivers, we performed dual knockdown of Ndufs7 and Washc1 in p300-overexpressing MMQ and AtT-20 cells. Results showed that simultaneous knockdown of Ndufs7 and Washc1 more significantly reversed the increase in ROS levels and pro-apoptotic effect mediated by p300 upregulation (Supplementary Fig. 11A–B).

These experiments indicate that p300-mediated H3K18la can promote the transcription of Ndufs7 and Washc1, thereby upregulating mitochondrial ROS levels in pituitary tumor cells and promoting tumor cell apoptosis. A schematic diagram of this mechanism is shown in Fig. 4V.

WASH1 inhibits mitophagy by binding to the ubiquitin-associated domain of p62 in pituitary tumor cells

3.5

To explore the impact of WASH1 protein on mitophagy, we detected the expression levels of autophagy-related proteins (LC3B, p62) and mitochondrial membrane structure-related proteins (TOM20, TIM23, COX IV) in MMQ and AtT-20 cells by WB. Results showed that under BRC treatment, the LC3B-II/LC3B–I ratio was significantly increased in p300-upregulated cells, suggesting enhanced autophagy initiation or reduced lysosomal degradation. Significant accumulation of p62, LC3B-II, and mitochondrial membrane proteins suggested potential inhibition of autophagic processes downstream of p62. Washc1 knockdown restored autophagic activity, indicating that WASH1 may inhibit mitophagy (Fig. 5A).Fig. 5WASH1 inhibits mitophagy by binding to the ubiquitin-associated (UBA) domain of p62 in pituitary tumor cells. (A) MMQ and AtT-20 cells in the vector group, OE-p300 group, OE-p300 with Washc1 knockdown group 1 (sh1+OE), and OE-p300 with Washc1 knockdown group 2 (sh2+OE) were treated with BRC (10 μM) for 48 h, followed by WB analysis to detect the expression of LC3BI/II, p62, translocase of outer mitochondrial membrane 20 (TOM20), translocase of inner mitochondrial membrane 23 (TIM23), and cytochrome c oxidase subunit IV (COX IV) (n = 3). (B) WASH1-Flag protein was overexpressed in AtT-20 cells with Washc1 knockout, and immunoprecipitation-mass spectrometry (IP-MS) was used to analyze potential interacting proteins of WASH1. (C) Mouse WASH1 protein and the UBA domain of p62 were subjected to molecular docking using the HDOCK server (http://hdock.phys.hust.edu.cn/). Appropriate prediction models were selected, and docking results were analyzed using PyMOL (Version 3.1). (D) In MMQ and AtT-20 cells with p300 overexpression and Washc1 knockout, wild-type Washc1 (OE-WT-Flag) or mutant Washc1 (OE-Mut-Flag) was then overexpressed. Co-IP experiments were conducted using anti-Flag antibodies, and subsequent WB analysis was used to detect the expression of WASH1-Flag and p62 (n = 3). (E) MMQ and AtT-20 cells with p300 overexpression and Washc1 knockout in the vector group, OE-WT-Flag group, or OE-Mut-Flag group were treated with BRC (10 μM) for 48 h, followed by WB analysis to detect the expression of LC3BI/II, p62, TOM20, TIM23, and COX IV (n = 3). (F) MMQ and AtT-20 cells with p300 overexpression and Washc1 knockout in the vector group, OE-WT-Flag group, or OE-Mut-Flag group were treated with BRC (10 μM) for 24 h, followed by mt-Keima ratiometric analysis to quantify mitophagy levels (n = 4). (G) MMQ and AtT-20 cells in the vector group, OE-WT-Flag group, or OE-Mut-Flag group were treated with BRC (10 μM) for 24 h, and confocal fluorescence microscopy was used to observe the co-localization of mitochondria (TOM20) and lysosomes (lysosome-associated membrane protein 1, LAMP1) (n = 3). (H–I) MMQ and AtT-20 cells with p300 overexpression and Washc1 knockout in the vector group, OE-WT-Flag group, or OE-Mut-Flag group were treated with BRC (10 μM) for 48 h, followed by flow cytometry to detect mitochondrial ROS and cell apoptosis levels (n = 3). One-way ANOVA was used for comparison among multiple groups. Data are presented as the mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.Fig. 5

Therefore, we overexpressed WASH1-Flag protein in p300-upregulated AtT-20 cells and analyzed potential interacting proteins of WASH1 by co-immunoprecipitation-mass spectrometry (CoIP-MS). Results showed that p62 can specifically bind to WASH1 protein (Fig. 5B, Supplementary Fig. 12A). Molecular docking results suggested that WASH1 may bind to the ubiquitin-associated (UBA) domain of p62 (Fig. 5C). Accordingly, we constructed GST-tagged UBA protein (GST-UBA), His-tagged wild-type WASH1 protein (His-WASH1-WT), and His-tagged mutant WASH1 protein (His-WASH1-Mut; with point mutations at arginine 25 and glutamic acid 28) in vitro. GST pull-down assay verified that the wild-type WASH1 protein could bind to the UBA domain of p62, while the mutant WASH1 protein failed to interact with the UBA protein (Supplementary Fig. 12B). Given that p62 acts as a ubiquitin adaptor to mediate mitophagy [[17], [18], [19]], we constructed rat and mouse Washc1 mutants (point mutations at arginine 25 and glutamic acid 28) to explore whether WASH1 affects mitophagy by binding to the UBA domain of p62. After Washc1 knockout in p300-upregulated MMQ and AtT-20 cells, we overexpressed wild-type Washc1-Flag (OE-WT-Flag) and mutant Washc1-Flag (OE-Mut-Flag), respectively (Supplementary Fig. 12C–D). WASH1 is mainly localized in early endosomes and can perform biological functions such as endosomal sorting and vesicular transport by binding to the Arp2/3 complex. To exclude the possibility that the mutant may interfere with the normal physiological functions of the WASH protein complex, we verified the structural and functional stability of the WASH1 mutant. Results showed no significant functional difference between the WASH1 mutant and the wild-type (Supplementary Fig. 12E–G). Meanwhile, we also ruled out the effects of wild-type and mutant WASH1 on cellular lysosomal function and bulk autophagic flux (Supplementary Fig. 12H–I). Subsequently, Co-IP experiments confirmed that the mutant WASH1 protein could not bind to p62 (Fig. 5D). Meanwhile, Western blot (WB) detection showed that compared with the OE-WT-Flag group, the OE-Mut-Flag group reversed the inhibition of mitophagy flux and the significant accumulation of autophagy-related proteins and mitochondrial membrane-related proteins mediated by WASH1 (Fig. 5E and F). These results indicate that in pituitary tumor cells, WASH1 can block p62's recognition of ubiquitinated damaged mitochondria by binding to the UBA domain of p62, thereby inhibiting mitophagy. Subsequent colocalization experiments between lysosomes and mitochondria showed that the colocalization level was restored and increased in the OE-Mut-Flag group, which indirectly confirmed WASH1-mediated mitophagy impairment (Fig. 5G).

Finally, we further verified the phenotype and found that under BRC treatment, the OE-Mut-Flag group reversed the increase in ROS levels and enhanced cell apoptosis mediated by the OE-WT-Flag group, further indicating that the constructed Washc1 mutation site is a key site regulating mitochondrial ROS and cell apoptosis levels indirectly by binding to p62 (Fig. 5H and I).

p300 activator YF-2 combined with DA exerts a synergistic anti-pituitary adenoma effect

3.6

Given that p300 upregulation or activation can sensitize DA efficacy in pituitary tumor cells by increasing mitochondrial ROS levels, p300 is expected to be a potential drug target for clinically resistant prolactinomas.

We selected three reported p300 activators (CTB, TTK21, YF-2) [[41], [42], [43]], all of which enhance acetyltransferase activity by targeting the active pocket of p300 or regulating its conformation. We first evaluated the effects of different concentrations of the three activators on MMQ and AtT-20 cells in vitro. Based on the screening criterion of the smallest half-maximal inhibitory concentration (IC50), YF-2 was ultimately selected for subsequent experiments (Fig. 6A, Supplementary Fig. 13A–B).Fig. 6p300 activator YF-2 combined with DA exerts a synergistic anti-pituitary adenoma effect. (A) CCK-8 assay was used to detect cell viability in MMQ and AtT-20 cells treated with YF-2 at different concentrations (0, 1.25, 2.5, 5, 10, 20 μM) for 48 h, after which dose-response curves were fitted based on cell viability and half-maximal inhibitory concentration (IC50) was calculated (n = 4). (B–C) Synergy indices for MMQ cells treated with combinations of YF-2 (0, 2.5, 5, 10 μM) and either BRC (0, 5, 10, 20 μM) or CAB (0, 12.5, 25, 50 μM) for 48 h were calculated using the ZIP synergy scoring model via SynergyFinder Version 3.0 (https://synergyfinder.fimm.fi), with a synergy index >10 indicating synergy and a white dashed box denoting the concentration range with the highest potential for maximum synergy (n = 4). (D-G) MMQ and AtT-20 cells were treated with BRC (10 μM), CAB (25 μM), YF-2 (5 μM), BRC (10 μM) + YF-2 (5 μM), or CAB (25 μM) + YF-2 (5 μM) for 48 h. (D, F) CCK-8 assay was used to detect cell viability in MMQ and AtT-20 cells (n = 5). (E) PRL concentration in the supernatant of MMQ cells was detected using a PRL ELISA kit (n = 4). (G) ACTH concentration in the supernatant of AtT-20 cells was detected using an ACTH ELISA kit (n = 4). (H–I) Nude mice were subcutaneously implanted with MMQ and AtT-20 cells, followed by i.p. injection of PBS, BRC (10 mg/kg/d), YF-2 (20 mg/kg/d), or BRC (10 mg/kg/d) + YF-2 (20 mg/kg/d) for 2 weeks. Representative images of subcutaneous xenograft tumors (left), average volume of excised tumors (middle), and average weight of excised tumors (right) (n = 6). (J-K) Quantification of IF staining for Ki-67 expression in tumor tissue sections (n = 6). (L-O) MMQ and AtT-20 cells in the vector group were treated with BRC (10 μM), YF-2 (5 μM) or BRC (10 μM) + YF-2 (5 μM) for 48 h. MMQ and AtT-20 cells transfected with wild-type p300 overexpression plasmid (OE-WT), or HAT domain-mutated p300 overexpression plasmid (OE-Mut) treated with BRC (10 μM) + YF-2 (5 μM) for 48 h, followed by flow cytometry to detect mitochondrial ROS and cell apoptosis levels (n = 3). One-way ANOVA was used for comparison among multiple groups. Data are presented as the mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.Fig. 6

Next, we detected the effects of different concentrations of YF-2 combined with different concentrations of DA on MMQ and AtT-20 cells. Synergy index analysis showed that 5 μM YF-2 combined with 10 μM BRC or 25 μM CAB exerted a significant synergistic anti-proliferative effect on tumor cells, and the PRL level in cell supernatants also decreased accordingly (Fig. 6B–G, Supplementary Fig. 13C–D). The same conclusion was obtained in combined drug experiments with primary cells (Supplementary Fig. 13E–F).

We further evaluated the in vivo anti-tumor efficacy of YF-2 using a nude mouse subcutaneous xenograft model constructed with MMQ and AtT-20 cells. Results showed that YF-2 significantly enhanced the anti-tumor effect of BRC, as evidenced by significantly reduced tumor volume, decreased tumor weight, inhibited tumor proliferation, and downregulated Ki-67 expression (Fig. 6H–K, Supplementary Fig. 13G–H). To verify whether YF-2 exerts a synergistic anti-tumor effect with DA by activating the HAT domain of p300, we detected mitochondrial ROS levels and cell apoptosis levels in MMQ and AtT-20 cells by flow cytometry. Results showed that compared with the single-drug group, YF-2 combined with BRC more significantly increased mitochondrial ROS levels and promoted cell apoptosis. Overexpression of mutant p300 (OE-Mut) reversed the further increase in mitochondrial ROS levels and cell apoptosis levels induced by overexpression of wild-type p300 (OE-WT) after combined treatment with BRC and YF-2 (Fig. 6L–O, Supplementary Fig. 14A–B). The same conclusion was obtained in combined drug experiments with primary cells (Supplementary Fig. 14C–D). Previously, we identified Ndufs7 and Washc1 as key target genes of p300-mediated lactylation. Therefore, we detected the expression of NDUFS7 and WASH1 in in vitro cell lines and in vivo xenograft tissues. Results showed that compared with the single-drug group, YF-2 combined with BRC significantly promoted the expression of NDUFS7 and WASH1 both in vitro and in vivo (Supplementary Fig. 14E–F). Finally, we evaluated the changes in body weight as well as the structure and function of major organs in mice during the combined treatment with BRC and YF-2, and no obvious drug toxicity was observed (Supplementary Fig. 15A–E).

In summary, in vitro and in vivo experiments showed that YF-2 combined with BRC synergistically inhibits the proliferation of MMQ and AtT-20 cells, thereby enhancing the therapeutic effect of BRC. The mechanism is related to activating the HAT domain of p300 and further increasing mitochondrial ROS levels. This combination strategy is expected to become a potential therapeutic approach for resistant prolactinomas.

Discussion

4

Prolactinomas are the most common functional pituitary adenomas, and DAs resistance remains a key clinical challenge for which effective combination therapies are currently lacking [[2], [3], [4],26,44]. Addressing this unmet need, the present study first confirmed that DAs treatment itself may drive resistance by downregulating p300—an effect that likely contributes to secondary resistance in certain patients. Subsequent in vitro and in vivo experiments demonstrated that p300 upregulation or activation, when combined with DAs, more potently induced mtROS production in tumor cells, ultimately triggering apoptosis.

At the mechanistic level, we found that p300 upregulation combined with DA promotes a metabolic shift toward glycolysis in pituitary tumor cells, leading to lactate accumulation that serves as a direct substrate for p300-mediated histone lactylation. Numerous studies have established that p300-mediated histone lactylation plays a critical role in tumorigenesis, immune escape, invasion, and drug resistance [[9], [10], [11], [12],28,29]; simultaneously, research in other tumor types has explored the reciprocal regulatory relationship between H3K18 lactylation and intracellular ROS levels [[45], [46], [47], [48]]. The current study elucidates the mechanism by which p300 histone acetyltransferase (HAT) domain-mediated H3K18 lactylation (H3K18la) modulates mtROS levels in pituitary tumor cells, and identifies Ndufs7 (NADH dehydrogenase (ubiquinone) iron-sulfur protein 7) as a key gene transcriptionally upregulated by the p300-H3K18la axis. NDUFS7 is a core structural subunit of mitochondrial respiratory chain Complex I(CI), and its normal expression is critical for Complex I assembly, electron transport, proton pump activity, and mitochondrial oxidative phosphorylation efficiency [39,40]. Previous studies have shown that NDUFS7 deficiency impairs the assembly and enzymatic activity of mitochondrial complex I (CI), disrupts electron transport chain (ETC) function, decreases the NADPH/GSH ratio, and aggravates oxidative stress, ultimately leading to ROS accumulation and cell death. Notably, SLC7A11-mediated cystine uptake can mitigate this damage by promoting GSH biosynthesis. In a rotenone-induced CI dysfunction model, restoration of NDUFS7 expression can cooperate with subunits such as NDUFV2 to improve CI assembly and electron transport efficiency, thereby reducing neuronal ROS production and lipid peroxidation and exerting a neuroprotective effect [49,50]. In our study, p300 upregulation combined with DA increased NDUFS7 levels in CI and was accompanied by elevated ROS and increased CI electron leakage. One possible explanation relates to the position of NDUFS7 within the Q module of CI, where it contributes to the ubiquinone-binding/reduction region; changes in the abundance or conformation of this region could influence Q-site electron-transfer kinetics and, under some conditions, favor ROS generation [51]. In addition, prior work suggests that respiratory supercomplex organization, particularly the association between CI and complex III, can affect CI-derived ROS output, raising the possibility that altered CI assembly stoichiometry or supercomplex stability might contribute to the increased ROS observed here [52]. The underlying mechanism in this study requires further investigation in future work.

Mitochondria are the main source of intracellular ROS. Under physiological conditions, cells maintain redox homeostasis via antioxidant systems and mitophagy-mediated clearance of damaged mitochondria [32,53]. This study found that upregulation of p300 combined with DA significantly inhibits mitophagy in pituitary tumor cells, suggesting that mitophagy may play a pivotal role in regulating mitochondrial ROS homeostasis. Based on this, we further identified the key transcriptionally upregulated gene Washc1 mediated by p300-H3K18la which encodes the core subunit of the WASH complex (WASP family verprolin-homologous protein complex). The WASH complex was initially shown to be mainly localized in endosomes, promoting F-actin polymerization by activating the Arp2/3 complex, thereby regulating endosomal sorting and vesicular transport processes [54]. Endosomes and autophagy also form a functional network through multi-level crosstalk [55].In our experimental system, WASH1 overexpression had no significant effect on WASH complex physiological functions or global autophagic flux, but exerted a specific inhibitory effect on mitophagic flux. Studies in recent years have clearly established negative regulatory role in autophagy of WASH complex. On one hand, WASH complex can directly bind to Beclin1, a core autophagic molecule, competitively blocking the interaction between AMBRA1 and Beclin1, thereby inhibiting AMBRA1-DDB1-CUL4A complex-mediated K63-linked ubiquitination of Beclin1, downregulating Vps34/PI3K kinase activity and downstream autophagosome formation. On the other hand, WASH complex can act as a molecular scaffold to recruit the E3 ubiquitin ligase RNF2, specifically mediating K48-linked ubiquitination and proteasomal degradation of AMBRA1, a positive regulator of autophagy, further impairing the assembly efficiency of the Beclin1-Vps34 complex [[56], [57], [58]]. This study further uncovers a novel regulatory mechanism of WASH1, the core functional subunit of the WASH complex: WASH1 can specifically bind to the UBA domain of the ubiquitin adaptor p62, preventing p62 from recognizing and binding to ubiquitinated damaged mitochondria. This leads to the failure of damaged mitochondria with high ROS load to be effectively cleared through the autophagic pathway, which in turn causes sustained intracellular ROS accumulation and ultimately induces tumor cell apoptosis. Together with this novel mechanism identified in the present study, these findings suggest that WASH family proteins regulate autophagy through multi-target and multi-pathway characteristics.

Finally, this study evaluated the potential application of p300 activator YF-2 in prolactinomas. Previous studies have shown that YF-2 has anti-tumor activity [43], but its application in pituitary adenomas has not been reported. In our models, YF-2 increased p300 HAT activity in MMQ and AtT-20 cells and enhanced the anti-tumor effects of DAs, supporting the idea that activating the p300 axis may be a useful adjuvant strategy. At the same time, these data are preliminary and do not support immediate clinical translation. Because p300 and CBP have overlapping functions and pharmacological activation can have broad transcriptional consequences, contributions from CBP or other off-target effects cannot be excluded. In addition, our in vivo assessment was limited and lacked detailed PK/PD and longer-term safety evaluation. Further work is needed to define selectivity, establish exposure–response relationships, and assess safety across additional models.

Taken together, this study elucidates that p300 upregulation or activation combined with DAs promotes mtROS accumulation in prolactinomas via H3K18la-mediated upregulation of Ndufs7 and Washc1, uncovers a novel mechanism whereby WASH1 binds p62's UBA domain to selectively inhibit mitophagy, validates the adjuvant effect of the p300 activator YF-2 in enhancing DA efficacy, and lays a preliminary foundation for novel combination therapies targeting DA-resistant prolactinomas.

Data

The raw data of CUT&Tag sequencing and transcriptome sequencing have been deposited in the Gene Expression Omnibus (GEO) database under the accession numbers GSE313568 and GSE313569, respectively. These data are publicly available upon database release.

CRediT authorship contribution statement

Sihan Li: Conceptualization, Investigation, Methodology, Writing – original draft. Qian Jiang: Data curation, Investigation, Methodology, Validation. Quanji Wang: Investigation, Methodology. Xingbo Li: Investigation, Methodology. Zihan Wang: Investigation, Methodology. Linpeng Xu: Investigation, Methodology. Shuyan Luo: Methodology. Yaorui Wang: Investigation. Huaqiu Zhang: Supervision, Validation, Writing – review & editing. Kai Shu: Supervision, Validation, Writing – review & editing. Ting Lei: Funding acquisition, Supervision, Validation, Writing – review & editing. Yimin Huang: Funding acquisition, Supervision, Validation, Writing – review & editing. Zhuowei Lei: Conceptualization, Funding acquisition, Supervision, Validation, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing ffnancial interests or personal relationships that could have appeared to inffuence the work reported in this paper.