Abstract

   graphic file with name ao3c09983_0006.jpg

   Hyperlipidemia (HLP) is a prevalent systemic metabolic disorder
   characterized by disrupted lipid metabolism. Statin drugs have long
   been the primary choice for managing lipid levels, but intolerance
   issues have prompted the search for alternative treatments. Matrine, a
   compound derived from the traditional Chinese medicine Kushen, exhibits
   anti-inflammatory and lipid-lowering properties. Nevertheless, the
   mechanism by which matrine modulates lipid metabolism remains poorly
   understood. Here, we investigated the molecular mechanisms underlying
   matrine’s regulation of lipid metabolism. Employing quantitative
   proteomics, we discovered that matrine increases the expression of LDL
   receptor (LDLR) in HepG2 and A549 cells, with subsequent experiments
   validating its role in enhancing LDL uptake. Notably, in hyperlipidemic
   hamsters, matrine effectively lowered lipid levels without affecting
   body weight, which highlights LDLR as a critical target for matrine’s
   impact on HLP. Moreover, matrine’s potential inhibitory effects on
   tumor cell LDL uptake hint at broader applications in cancer research.
   Additionally, thermal proteome profiling analysis identified lipid
   metabolism-related proteins that may interact with matrine. Together,
   our study reveals matrine’s capacity to upregulate LDLR expression and
   highlights its potential in treating HLP. These findings offer insights
   into matrine’s mechanism of action and open new avenues for drug
   research and lipid metabolism regulation.

Introduction

   Hyperlipidemia (HLP) is a prevalent systemic metabolic disorder
   characterized by dysregulated lipid metabolism, resulting in elevated
   levels of total cholesterol (TC), triglyceride (TG), and low-density
   lipoprotein cholesterol (LDL-C) in circulation.^[42]1,[43]2 This
   perturbation in lipid metabolism contributes to the progression of
   atherosclerosis and can lead to cardiovascular diseases (CVDs).^[44]3
   As a consequence, HLP is recognized as a significant risk factor for
   CVDs.^[45]4 Furthermore, lipid accumulation can trigger common
   complications such as thrombus-induced myocardial infarction (MI) and
   stroke,^[46]3,[47]5,[48]6 underscoring the urgency for early
   intervention in HLP.

   Across the past few decades, diverse lipid-lowering medications have
   been developed, with statins being a representative example. Statins
   function by inhibiting intracellular cholesterol synthesis, leading to
   an elevation in the expression of low-density lipoprotein receptor
   (LDLR) and aiding in the maintenance of plasma cholesterol levels via
   LDL-C endocytosis. However, a subset of patients experience intolerance
   to statin treatments.^[49]7 Therefore, the exploration of nonstatin
   lipid-lowering alternatives assumes significance.^[50]8,[51]9

   Natural products represent a significant reservoir of potential drug
   candidates, owing to their distinctive strengths, such as chemical
   diversity, biological activity, and so on. Consequently, the discovery
   and development of novel natural products have become an important area
   of research in anti-HLP treatments.^[52]10,[53]11 Kushen, derived from
   the roots of Sophora flavescens, a well-known traditional Chinese
   medicine belonging to the leguminous family, is indigenous to China,
   India, Japan, Korea, and Russia.^[54]12 Historically, the dried roots
   of Kushen have been utilized for addressing various health issues,
   including tumors and inflammation.^[55]13 Kushen contains diverse
   alkaloid compounds including matrine (molecular formula:
   C[15]H[24]N[2]O, molecular weight: 248.36 g/mol), oxymatrine, and
   sophorine.^[56]13,[57]14 Recent studies have highlighted matrine’s
   anti-inflammatory, antioxidant, antimicrobial, and analgesic
   properties.^[58]13,[59]15 In addition, emerging evidence suggests that
   matrine effectively reduces TC and TG levels in high diet feed (HFD)
   mice,^[60]16−[61]19 indicating its potential in HLP treatment. However,
   the mechanism by which matrine modulates lipid metabolism remains
   poorly understood, and elucidating this mechanism holds significant
   potential for advancing its clinical use in HLP treatment.

   This study delves into the molecular mechanism underlying the
   regulation of lipid metabolism disorder by matrine. Through the
   application of quantitative proteomics, we assessed the dynamic changes
   in protein expression within HepG2 cells following matrine treatment.
   Interestingly, we observed a noteworthy increase in LDLR levels.
   Subsequent validation experiments substantiated that matrine bolsters
   LDL uptake by modulating LDLR transcription and augmenting protein
   stability. Furthermore, in vivo experiments showcased matrine’s
   effectiveness in mitigating elevated TC, TG, and LDL-C levels in
   hyperlipidemic hamsters that had been induced by a high-fat diet.
   Remarkably, this effect was achieved without influencing body weight or
   food consumption. Additional investigation unveiled heightened gene
   transcription and protein expression of LDLR in the livers of
   matrine-treated hamsters with HLP induced by a high-fat diet.
   Consequently, LDLR emerges as a pivotal target for matrine’s impact on
   the treatment of HLP.

   Similar to the aforementioned phenomenon, we observed that matrine also
   upregulates the expression of LDLR in A549 cells. A recent study
   highlighted the potential detrimental effects of excessive low-density
   lipoprotein (LDL) uptake on breast cancer cells.^[62]20 Therefore, this
   finding speculates that matrine might exert inhibitory effects on tumor
   cells by intervening in tumor cell LDL uptake. Furthermore, to identify
   potential interacting proteins of matrine, we conducted thermal
   proteome profiling (TPP) analysis, revealing four lipid
   metabolism-related proteins that could potentially serve as matrine
   target proteins.

   In summary, our quantitative proteomics study has revealed matrine’s
   capacity to upregulate LDLR expression in both HepG2 and A549 cells.
   Furthermore, we validated that matrine-induced LDLR upregulation
   significantly contributes to rectifying disordered lipid metabolism in
   vivo. Our findings thus present a novel avenue for exploring matrine’s
   HLP treatment mechanism and offer a fresh perspective on investigating
   drug molecular mechanisms.

Methods

Cell Lines and TMT Agent

   The A549 cell line (catalog number: SCSP-503) and HepG2 cell line
   (catalog number: SCSP-510) were obtained from the National Collection
   of Authenticated Cell Culture. TMT 10plex was purchased from Thermo
   Fisher (catalog number: 90111).

Animal Model

   We purchased six-week-old healthy male Syrian golden hamsters from
   Beijing Vital River Laboratory and housed them in the SPF-level
   Laboratory Animal Room of Shanghai Institute of Materia Medica, Chinese
   Academy of Sciences. All animal experiments and protocols used in this
   study were approved by the Animal Ethics Committee of the Shanghai
   Institute of Materia Medica. This study was conducted under the
   guidelines and ethics of the Association for Assessment and
   Accreditation of Laboratory Animal Care International (AAALAC).

   We housed hamsters in a light-controlled room at a temperature of 22
   °C. The hamsters could access water for free, and they were fed a
   high-fat diet food (0.5% cholesterol, 23% fat, Research Diets). Before
   treatment, hamsters were orderly and evenly assigned into the HFD-fed
   vehicle, HFD-fed matrine-treated, and HFD-fed berberine (BBR)-treated
   groups according to their serum lipid levels. All of the hamsters were
   fed a HFD for 4 weeks and orally treated with matrine (50 or 100 mg/kg)
   or BBR (100 mg/kg) daily for the last 3 weeks. Next, hamsters were
   euthanized, and their body weights were recorded. The livers of
   hamsters were immediately removed, cut into small pieces, and saved at
   −80 °C.

Cell Culture

   HepG2 and A549 cells were cultured in Dulbecco’s modified Eagle’s
   medium (DMEM, high glucose) supplemented with 10% FBS (164210-50) and
   1% P/S ([63]PB180120) in 10 cm dishes (Corning) at 37 °C, 5% CO[2]. For
   proteomics analysis, HepG2 cells were divided into experimental groups
   and control groups and treated with or without matrine. A549 cells were
   cultured in DMEM free of lysine/arginine supplemented with 10% dialyzed
   FBS and either “light” or “heavy” amino acids with 1% penicillin and
   streptomycin in a 5% CO[2] atmosphere at 37 °C. To ensure that the
   efficiency of labeling was more than 98%, the cells were cultured for
   at least 7 generations. A549 cells were treated with matrine or DMSO
   for 48 h to perform proteomics analysis. For TPP analysis, when A549
   cells grew to approximately 80% confluence in the dishes, we divided
   the cells into experimental groups and control groups and treated them
   with 1 mM matrine or DMSO for 1.5 h, respectively.

Proteomics Sample Analysis

   The treated HepG2 cells and A549 cells were washed with precooled PBS
   (pepstatin, leupeptin, PMSF) three times. Then, we added RIPA buffer to
   lyse the cells on ice for 30 min and sonicated them (15 W, 3 min).
   After centrifugation (16,000g at 4 °C for 10 min), we used the BCA
   assay to determine the protein concentration. For HepG2 cells, 100 μg
   of proteins were transferred into new tubes. For A549 cells, we equally
   mixed the proteins from “heavy” and “light” cells and then transferred
   100 μg of proteins into new tubes. Next, TCA was used to precipitate
   the proteins at a final concentration of 10%, and precooled acetone was
   used to wash the proteins twice. Then, 25 mM Tris buffer (pH 7.5) was
   used to resuspend the proteins. We added trypsin to the proteins at a
   ratio of 1:50 overnight to obtain peptide samples. After digestion, the
   samples were reduced with dithiothreitol (5 mM, 56 °C, 30 min) and
   alkylated with iodoacetamide (11 mM, room temperature, 30 min).

Thermal Proteome Profiling

   The A549 cells were collected after being treated with matrine or DMSO.
   The collected cells were washed three times and resuspended with
   precooled PBS. Then, the two groups of cells were equally divided into
   eight fractions (each of the fractions contained 96 μL of samples), and
   each fraction was transferred into PCR tubes. The tubes were heated in
   parallel at different temperatures ranging from 37 to 67 °C at a
   gradient of 5 °C for 3 min, followed by cooling at 25 °C for 3 min.
   Subsequently, 10% NP-40 was added to each of the tubes at a final
   concentration of 4%, and the tubes were incubated on ice for 20 min.
   After that, the tubes were put into liquid nitrogen to freeze thaw
   three times. Then, the tubes were subjected to ultracentrifugation at
   160,000g for 10 min at 4 °C. The supernatant was obtained and moved
   into fresh tubes. We detected its protein concentration through the BCA
   assay. Next, we removed 40 μL of the supernatant, reduced it with
   dithiothreitol (5 mM, 56 °C, 30 min), and alkylated it with
   iodoacetamide (11 mM, room temperature, 30 min). Finally, the proteins
   were subjected to short SDS gel electrophoresis.

TMT Labeling

   After short SDS gel electrophoresis, we cut the entire lanes of the gel
   into slices and added ACN to dehydrate them later. After dehydration,
   the ACN was removed. Trypsin was added to digest the samples in 5 mM
   TEAB buffer (pH 8.0) at 37 °C overnight. Then, we added formic acid
   until the pH reached 2 to stop the digestion. Next, we added a mixture
   containing 1% formic acid and 50% ACN to elute and desalt the peptides
   with a C18-tip after digestion. The desalted peptides were dissolved in
   20 μL of 50 mM HEPES at pH 8.5. Subsequently, the TMT (50 μg) solution,
   which was dissolved in ACN, was added to the peptides and incubated at
   25 °C for 1 h. Hydroxylamine was added to stop the reaction at a final
   concentration of 3% after incubation. Finally, we used a mixture
   containing 10% FA and 10% ACN to acidize the samples and evaporated
   them to dryness.

LC–MS/MS Analysis

   For TPP analysis, the peptide samples were dissolved in 2 μL of 0.1%
   FA, and 2 μg of the peptides were loaded on an EASY-nLC 1200 UHPLC
   system connected to a Q Exactive HF-X mass spectrometer. Then the
   peptides were injected into a reversed-phase C18 column (20 cm length ×
   75 μm ID, 1.9 μm particle size, Dr. Maisch GmbH, Germany) in a 60 min
   gradient from 10 to 90% solvent B (A, 0.1% formic acid; B, 80%
   acetonitrile in 0.1% formic acid) at 225 nL/min. The eluted peptides
   were analyzed by the mass spectrometer, which was operated in a
   data-dependent acquisition mode. During peptide elution, the mass
   spectrometer method was operated in positive ion mode. Full mass scans
   were acquired with a mass resolution of 60,000 when the automatic gain
   control (AGC) number was 3 × 10^6 and the maximum injection time was 50
   ms. The 20 most intensive ions were subjected to fragmentation, with 20
   s of dynamical exclusion. The MS2 fragmentation spectra can be acquired
   at a resolution of 45,000 when the number of AGC for MS2 is 5 × 10^4
   and the maximum injection time is 65 ms. The fragmentation analysis
   used HCD with an NCE of 28. The respective data-dependent settings were
   set with parameters: exclude isotopes as “on”; dynamic exclusion of 60
   s using Xcalibur software (available from Thermo Fisher Scientific).

Western Blotting

   A549 cells, HepG2 cells, and liver samples were added to RIPA buffer
   for lysis on ice for 30 min and sonicated at 15 W for 3 min. After
   centrifugation (16,000g at 4 °C for 10 min), we used the BCA assay to
   determine the protein concentration. Next, proteins were separated by
   SDS-PAGE electrophoresis and transferred to a PVDF membrane. Then, the
   membranes were incubated with antibodies against LDLR (Proteintech,
   Cat# 10785-1-AP, China; 1:2000), PCSK9 (Abcam, Cat# ab181142, UK;
   1:2000), and ACTIN (Proteintech, Cat# 66009-1-1g, China; 1:5000). The
   bands were visualized with an ECL kit and were recorded on a Tanon 4600
   (Tanon).

Analysis of Gene Transcription by RT-qPCR

   We extracted RNA from cells or liver tissue using TransZol Up regent
   (TRAN, Cat# ET111, China) and reverse-transcribed RNA to cDNA (Vazyme,
   Cat# R333-01, China). Real-time PCR was performed on a Steponeplus
   real-time PCR system with SYBR Green PCR Master mix (Vazyme, Cat#
   Q712-02, China) and gene-specific primers (LDLR, forward,
   5′-CTGAAATCGCCGTGTTACTG-3′, reverse, 5′-GCCAATCCCTTGTGACATCT-3′; PCSK9,
   forward, 5′-CCAAGCCTCTTCTTACTTCACC-3′, reverse,
   5′-GCATCGTTCTGCCATCACT-3′; and Antin, forward,
   5′-AGAGCTACGAGCTGCCTGAC-3′, reverse, 5′-AGCACTGTGTTGGCGTACAG-3′). The
   relative changes in gene expression were normalized against actin mRNA
   expression.

Dil-LDL Uptake Analysis

   We cultured HepG2 cells in 24-well plates with DMEM supplemented with
   1% P/S for 24 h. Then, the cells were incubated in DMEM with 2% LPDS
   and 20 mg/mL DiI-LDL for 4 h at 37 °C in the dark. After incubation, we
   used precooled PBS containing 0.4% albumin to wash the cells twice and
   wash them three times with precooled PBS. Then, we added 400 μL
   isopropanol into each well and incubated them on a constant shaker at
   real room temperature for 20 min. Then, 200 μL isopropanol was taken
   from each well for analysis with a SpectraMax M2e Microplate Reader
   (Molecular Devices, 520–570 nm).^[64]21

Serum Lipid Analysis

   We collected the blood from hamsters after 16 h of fasting. The serum
   TC, TG, and LDL-C were enzymatically measured with a kit (Wako,
   Maccura, Shino, Japan) in an automatic biochemical analyzer (7020,
   Hitachi, Japan).

Data Analysis

   After LC–MS/MS, we used MaxQuant (v1.6.15.0) to retrieve the mass data
   from the UniProt Human protein database (20376 entries,
   [65]https://www.uniprot.org). We used TMT 10plex labeling to carry out
   quantitative analysis. The following settings were used: the maximum
   missed cleavage sites were 2, carbamidomethylation of cysteine was set
   as the fixed modification, and the oxidation of methionine and
   N-terminal acetylation were specified as variable modifications. We set
   the lowest temperature point at 1 and calculated the comparison between
   the relative abundances of the TMT reporter ions with the lowest
   temperature to determine melting points. Then, we used the TPP R script
   to normalize the data, and the following equation was used to fit the
   melting curve of proteins. When half of the protein is denatured at
   this temperature, the melting point of the protein is f(T) = 0.5.
   graphic file with name ao3c09983_m001.jpg

Bioinformatic Analysis

   For the cellar pathway analysis, GO analysis, KEGG pathway analysis,
   and GSEA were performed for a hypergeometric test in the
   clusterProfiler package in R. The ClueGO app was used to find
   overrepresented KEGG pathway and Reactome pathway and a network of
   connected those terms was created. The proteins identified as having
   matrine-dependent thermostability were put into the STRING database
   (v11, [66]http://www.string-db.org/) to build the protein–protein
   association network and visualized in Cytoscape (v.3.8.2). Finally,
   StringAPP embedded in Cytoscape (v1.7.0) was used to enrich the key
   network of protein–protein interactions (PPIs) by function.^[67]22

Results

Quantitative Proteomics Analysis of HepG2 Cells upon Matrine Treatment

   Matrine is a natural alkaloid with pharmacological activity in the
   treatment of HLP ([68]Figure [69]1
   A). In order to delve into the mechanism through which matrine
   regulates disrupted lipid metabolism, we examined its impact on the
   HepG2 proteome by treating cells with or without matrine. The IC[50] of
   matrine on HepG2 cells was determined to be 5.066 mM following a 48 h
   treatment, exhibiting minimal impact on HepG2 cell proliferation when
   the dose was below 1.25 mM. Following treatment, we extracted proteins
   from both groups, which were subsequently subjected to trypsin
   digestion and prepared for LC–MS/MS analysis. Data analysis was
   performed using MaxQuant software ([70]Figure [71]1A).

Figure 1.

   [72]Figure 1
   [73]Open in a new tab

   Matrine upregulated proteins enriched in lipid metabolism pathways in
   HepG2 cells. (A) Chemical structure of matrine and workflow of
   quantitative proteome analysis. HepG2 cells were treated with or
   without matrine. Subsequently, proteins were extracted from both groups
   and subjected to trypsin digestion. The samples were then analyzed
   using LC–MS/MS. (B) Volcano plot illustrating the dynamic regulation of
   proteins by matrine. Orange: representative lipid proteins among
   upregulated proteins; blue: representative lipid proteins among
   downregulated proteins. (C) KEGG pathway and GO enrichment analysis of
   upregulated proteins. The top seven enrichment clusters encompassing
   the KEGG pathway, biological process, cellular component, and molecular
   function. (D) Gene set enrichment analysis of upregulated proteins.

   The quantitative proteome analysis through LC–MS/MS yielded the
   quantification of a total of 7582 proteins ([74]Figure [75]1
   B, Supporting Information [76]Table S1). Out of these, 854 proteins
   displayed distinct expression patterns when compared with the control
   group (|log[2] (Fold change) | > 1, P value <0.05). Notably, 600
   proteins exhibited upregulation upon matrine treatment (log[2] (Fold
   change) > 1, P value <0.05), whereas 253 proteins showed downregulation
   (log[2] (Fold change) < −1, P value <0.05). Importantly, a significant
   portion of these dynamically altered proteins showed correlations with
   lipid metabolism, exemplified by LDLR (log[2] (Fold change) = 1.379).

Pathways and Biological Process in HepG2 Cells Affected by Matrine Treatment

   To elucidate the potential cellular pathways influenced by the
   upregulated proteins induced by matrine, we performed enrichment
   analysis using the KEGG database ([77]Figure [78]1
   C). Interestingly, the upregulated proteins exhibited enrichment in
   cholesterol metabolism (adjusted p = 3.08 × 10^–6 -), lysosome
   (adjusted p = 9.01 × 10^–5), and thermogenesis (adjusted p = 1.75 ×
   10^–3). These enrichment results suggest a robust impact of matrine on
   lipid metabolism, particularly cholesterol metabolism.

   To further explore the influence of matrine on cellular process, we
   performed enrichment analysis using the Gene Ontology (GO) database
   ([79]Figure [80]1
   C). The top seven enriched biological process clusters primarily
   revolved around material transport, encompassing lipid transport
   (adjusted p = 1.3 × 10^–8) and lipid localization (adjusted p = 1.1 ×
   10^–8). This result underscores matrine’s influence on lipid transport,
   aligning with the KEGG database results. Additionally, the top seven
   enriched cellular component clusters revealed that the upregulated
   proteins mainly constituted the cell membrane components, such as the
   lysosomal membrane (adjusted p = 4.1 × 10^–21). Similarly, the enriched
   clusters of molecular function highlighted that the upregulated
   proteins predominantly engaged in transmembrane transporter activity,
   such as anion transmembrane transporter activity (adjusted p = 1.2 ×
   10^–10) and active transmembrane transporter activity (adjusted p = 1.0
   × 10^–9). The molecular function enrichment implies that matrine
   enhances cellular metabolism. Together, these results suggested that
   matrine-induced upregulated proteins primarily reside within the cell
   membrane, exhibit transmembrane transporter functions, and are related
   to lipid transport.

   Next, we performed Gene Set Enrichment Analysis (GSEA) with the GO
   database ([81]Figure [82]1
   D). The upregulated proteins significantly enriched processes such as
   lipid transport (adjusted p = 2 × 10^–2), cholesterol transport
   (adjusted p = 2 × 10^–2), cholesterol homeostasis (adjusted p = 2 ×
   10^–2), regulation of lipid transport (adjusted p = 2 × 10^–2), lipid
   homeostasis (adjusted p = 2 × 10^–2), cholesterol metabolic process
   (adjusted p = 2 × 10^–2), steroid metabolic process, and lipid
   metabolic process (adjusted p = 2 × 10^–2). This outcome further
   substantiates matrine’s role in enhancing lipid metabolism, cholesterol
   metabolism, and sterol metabolism. Together with the above results,
   matrine may execute its disordered lipid metabolism regulatory function
   by bolstering cellular lipid metabolism and lipid transport.

   To further elucidate the mechanism under lipid metabolism regulation of
   matrine, we conducted PPI analysis among the upregulated proteins using
   the STRING database and employed StringAPP, which facilitates the
   integration of STRING networks into Cytoscape^[83]22 ([84]Figure [85]2
   ). In the interaction network, darker connections signify closer
   associations. The upregulated proteins demonstrated enrichment in
   cholesterol metabolism, vitamin digestion, and vitamin absorption. The
   transport of cholesterol and the metabolism of vitamins are pertinent
   to lipid metabolism.^[86]23,[87]24 In addition, proteins presented as
   diamonds are linked to lipid metabolism. This observation offers
   further evidence that matrine indeed stimulates lipid metabolism.

Figure 2.

   [88]Figure 2
   [89]Open in a new tab

   PPI analysis of upregulated proteins mediated by matrine in HepG2
   cells. The graph depicts PPIs, with each pattern representing a
   protein. Diamonds symbolize proteins associated with lipid metabolism,
   and darker areas indicate closer connections.

LDLR Is Upregulated in HepG2 Cells upon Matrine Treatment

   Subsequently, our investigation into matrine’s potential as a
   therapeutic intervention for HLP directed our attention toward the
   identification of specific proteins associated with lipid metabolism
   within the cohort of the upregulated proteins ([90]Figure [91]3
   A).

Figure 3.

   [92]Figure 3
   [93]Open in a new tab

   Matrine upregulated gene transcription and expression of LDLR in HepG2
   cells. (A) Candidate proteins related to lipid metabolism. (B) Western
   blot validation of matrine-induced dose-dependent upregulation of LDLR
   protein expression in HepG2 cells, while having little effect on PCSK9
   protein. (C) RT-qPCR validation of matrine-induced dose-dependent
   upregulation of LDLR gene transcription in HepG2 cells. (D) RT-qPCR
   validation indicating that matrine had little effect on PCSK9 gene
   transcription in HepG2 cells. (E) Dil-LDL uptake analysis revealing
   that matrine promoted the uptake of LDL by HepG2 cells.

   ATP-binding cassette transporter A1 (ABCA1) (log[2](Fold change) =
   2.247) emerges as a crucial member of the ABCA subfamily of ATP-binding
   cassette transporters, pivotal in facilitating the loading of apoA-1
   with phospholipids to form high-density lipoproteins (HDLs).^[94]25
   Notably, mutations in ABCA1 have been implicated in Tangier disease, an
   uncommon and severe form of HDL deficiency characterized by an
   impairment in cellular cholesterol efflux.^[95]26,[96]27 Thus, it is
   conceivable that matrine may elevate ABCA1 expression, thereby
   fostering HDL formation and potentially ameliorating disordered lipid
   metabolism.

   Carcinoembryonic antigen-related cell adhesion molecule 1 (CEACAM1)
   (log[2](Fold change) = 1.396), a transmembrane glycoprotein primarily
   found in the liver,^[97]28 has been associated with insulin resistance,
   fatty liver disease, and obesity when subjected to mutations. CEACAM1
   overexpression has been demonstrated to possess the ability to
   counteract hyperinsulinemia, notably reducing insulin resistance and
   hepatic lipid accumulation.^[98]29,[99]30 Hypothetically, matrine’s
   efficacy in treating HLP might be attributed to its upregulation of
   CEACAM1 expression.

   The involvement of NPC intracellular cholesterol transporter 1 (NPC1)
   (log[2](Fold change) = 1.188), a transmembrane glycoprotein pivotal in
   transferring LDL-C from late endosome or lysosome to the endoplasmic
   reticulum for esterification or the plasma membrane for efflux,^[100]31
   presents another avenue. Studies have shown that NPC1 upregulation
   enhances cholesterol transport and inhibits foam cell
   formation.^[101]31,[102]32 Hence, matrine-induced upregulation of NPC1
   might contribute to the regulation of lipid metabolism.

   Remarkably, LDLR (log[2](Fold change) = 1.378), a cell surface
   glycoprotein that plays a pivotal role in maintaining plasma
   cholesterol levels through endocytosis of LDL-C,^[103]33 could
   potentially benefit from matrine-induced upregulation, facilitating
   increased extracellular uptake of LDL-C and aligning with its role in
   treating HLP.

   Intriguingly, our investigation demonstrated matrine’s dose-dependent
   upregulation of LDLR ([104]Figure [105]3
   B). Moreover, we evaluated the influence of matrine on Pro-protein
   convertase subtilisin-like/kexin type 9 (PCSK9), a factor involved in
   mediating LDLR degradation.^[106]34 We employed BBR as a positive
   control, a compound known to downregulate PCSK9 and consequently
   enhance LDLR expression. Encouragingly, our results indicated the
   minimal influence of matrine on PCSK9. The upregulation of LDLR protein
   level typically results from either PCSK9-mediated mechanisms or the
   promotion of LDLR gene transcription. Therefore, to ascertain that
   matrine-mediated LDLR upregulation occurs through gene transcription
   promotion, we performed RT-qPCR analysis. Our findings aligned with
   protein expression data, revealing that matrine indeed upregulated LDLR
   mRNA levels in a dose-dependent manner. Notably, matrine exerted little
   influence on PCSK9 mRNA levels ([107]Figure [108]3C,D).

   Moreover, the primary function of LDLR is the uptake of LDL-C from
   extracellular environments.^[109]35 To evaluate matrine’s impact on the
   function of LDLR, we cultured HepG2 cells and treated cells with
   matrine under starving conditions and subsequently exposed cells to
   fluorescently labeled Dil-LDL. Our fluorescence-based assessment of
   cellular LDL uptake mirrored the protein expression outcomes,
   underscoring matrine’s capacity to promote LDL uptake in a
   dose-dependent manner ([110]Figure [111]3
   E).

Disordered Lipid Metabolism Is Reversed by Matrine in Hyperlipidemic Hamsters

   In order to validate the lipid-lowering efficacy of matrine in vivo, we
   established a hyperlipidemic hamster model by administering a HFD
   ([112]Figure [113]4
   A). The hamsters subjected to the HFD regimen were subsequently treated
   with either a vehicle (0.5% CMC-Na) or matrine (at doses of 50 or 100
   mg/kg) for a duration of 21 days. The positive control group received
   treatment with BBR at a dose of 100 mg/kg. Following the treatment
   period, we assessed the level of TC, TG, and LDL-C in the hamster
   blood.

Figure 4.

   [114]Figure 4
   [115]Open in a new tab

   Matrine reversed disordered lipid metabolism in HFD hamsters. (A)
   Experimental procedure for the animal study. Hamsters were subjected to
   a high-fat diet (HFD) for 4 weeks and subsequently treated with matrine
   (50 or 100 mg/kg) or BBR (100 mg/kg) during the last 3 weeks. Serum
   lipids analysis was conducted on the last day. (B) Graph illustrating
   the cumulative food intake of hamsters. (C) Body weight of hamsters in
   the last day. (D) Alterations in TC levels in response to matrine
   administration in hamsters. (E) Alterations in TG levels in response to
   matrine administration in hamsters. (F) Alterations in LDL-C levels in
   response to matrine administration in hamsters. (G) Western blot
   validation of matrine-induced upregulation of LDLR protein expression
   in hamster livers. (H) RT-qPCR validation of matrine-induced
   upregulation of LDLR gene transcription in hamster livers.

   Our observations indicated that the cumulative food intake and the body
   weight of the hamster remained unaffected by the administration of
   matrine or BBR ([116]Figure [117]4
   B,C). However, there was a noteworthy reduction in the levels of TC,
   TG, and LDL-C after the matrine or BBR treatments ([118]Figure
   [119]4D–F). Interestingly, matrine exhibited a superior regulatory
   effect on serum TC, TG, and LDL-C levels in comparison to BBR,
   particularly at the higher dosage (100 mg/kg).

   To investigate whether matrine’s impact on LDLR levels in vivo mirrored
   its effect in cells, we assessed the hepatic LDLR expression in
   hamsters ([120]Figure [121]4
   G). As anticipated, both matrine and BBR treatment led to an
   upregulation in the expression of LDLR. Remarkably, the high-dosage
   matrine group (100 mg/kg) displayed a more pronounced increase in LDLR
   expression compared to the low-dosage group (50 mg/kg). Similarly, at
   the same dose, matrine (100 mg/kg) exhibited a more potent upregulation
   of LDLR compared to BBR (100 mg/kg). Furthermore, we analyzed the mRNA
   levels of LDLR ([122]Figure [123]4H). These results clearly
   demonstrated an upregulation of LDLR levels following matrine or BBR
   treatment. Interestingly, BBR exhibited elevated levels of LDLR mRNA
   expression; however, this did not correspond with protein expression
   levels under high-dose matrine treatment. Emerging studies have
   suggested that numerous processes influence the relationship between
   protein and mRNA abundance, including translational regulation,
   modulation of protein stability, protein synthesis delay, and
   others.^[124]36 Consequently, the induction of LDLR protein expression
   by BBR may also be subject to these processes, resulting in a lack of
   high consistency between its abundance and mRNA levels. Collectively,
   these results suggested that matrine can attenuate serum TC, TG, and
   LDL-C levels by stimulating gene transcription and protein expression
   of LDLR in vivo.

   Although studies have reported potential liver toxicity from
   administering matrine at doses of 118 and 154 mg/kg/day for 21 days in
   mice,^[125]13 our study did not observe matrine-induced liver toxicity,
   even at a high dose (100 mg/kg/day).

Influence of Matrine on the Proteome of A549 Cells

   A growing body of evidence has demonstrated that matrine possesses
   therapeutic potential for various cancers, including lung
   cancer.^[126]37,[127]38 Given that disordered lipid metabolism has been
   identified as a crucial factor in the progression of lung
   cancer,^[128]39 we hypothesize that matrine’s antitumor effects might
   stem from its regulation of lipid metabolism.

   To elucidate the mechanism underlying matrine’s efficacy against lung
   cancer, we conducted a comprehensive investigation of its impact on the
   proteome of A549 cells. In this strategy, we employed stable isotope
   labeling of amino acids in cell cultures (SILAC). After the treatment
   with matrine (“light” labeling) or DMSO (“heavy” labeling), we
   extracted proteins from both groups and combined them at a 1:1 ratio.
   In a biological replicate experiment, the SILAC labeling is reversed.
   These samples were then subjected to trypsin digestion and prepared for
   LC–MS/MS analysis, with data analysis executed using the MaxQuant
   software ([129]Figure [130]5
   A). To enhance the robustness of our findings, we performed
   label-swapping repetitions.

Figure 5.

   [131]Figure 5
   [132]Open in a new tab

   Matrine-induced upregulated proteins enriched in pathways associated
   with lipid metabolism in A549 cells. (A) Workflow of quantitative
   proteome analysis. A549 cells were cultured in a medium with either
   “light” or “heavy” amino acids and treated with or without matrine.
   After treatment, proteins from both groups were equally combined after
   extraction and subjected to trypsin digestion. The samples were
   subsequently analyzed by LC–MS/MS. (B) Graph illustrating quantified
   proteins in two replicates. Orange: dynamic lipid proteins; blue:
   stable lipid proteins. (C) KEGG and Reactome pathway analysis of
   upregulated proteins. Each node represents a term, and the colors
   represent the pathway group. The edges reflect the relationships
   between the terms based on the similarity of their associated genes.
   (D) GO enrichment analysis of upregulated proteins. The top ten
   enrichment clusters of biological processes are presented. (E) Gene Set
   Enrichment Analysis of upregulated proteins. (F) Western blot
   validation of matrine-induced dose-dependent upregulation of LDLR
   protein expression in A549 cells, with little effect on PCSK9 protein.

   Through the quantitative proteome analysis, we successfully quantified
   a total of 2533 proteins ([133]Figure [134]5
   B, Supporting Information [135]Table S2). Among these, 76 proteins were
   upregulated (log[2] (Fold change) > 1, P value <0.05) while 7 proteins
   were downregulated (log[2] (Fold change) < −1, P value <0.05).
   Particularly noteworthy was the remarkable upregulation of LDLR among
   the elevated proteins in response to matrine treatment. Intriguingly, a
   separate study has proposed that increased LDL can foster tumor growth,
   yet enhanced and prolonged LDL uptake is associated with heightened
   caspase-3 cleavage, a detrimental process for breast cancer
   cells.^[136]20

   As a consequence, we postulate that matrine’s antilung cancer effects
   might stem from its ability to elevate LDLR expression and sustain LDL
   uptake, thereby potentially impeding lung cancer progression. This
   insight underscores the intricate interplay between matrine, lipid
   metabolism, and cancer modulation.

Pathways and Biological Process in A549 Cells Affected by Matrine Treatment

   To elucidate the potential pathway modulated by matrine in A549 cells,
   we performed KEGG and Reactome pathway enrichment analysis using
   upregulated proteins induced by matrine treatment ([137]Figure [138]5
   C). Our observations revealed significant enrichment of upregulated
   proteins within pathways related to cholesterol biosynthesis, plasma
   lipoprotein assembly, remodeling, and clearance, and fatty acyl-CoA
   biosynthesis. This result underscores the substantial impact of matrine
   on cell metabolism. Notably, the upregulation of the steroid
   biosynthesis (adjusted p = 1.299 × 10^–3) and cholesterol metabolism
   (adjusted p = 9.38 × 10^–2) pathways implies that matrine continues to
   influence lipid-related metabolic pathways within A549 cells.

   Moreover, we performed enrichment analysis with the GO database to
   probe the effects of matrine on the cellular process ([139]Figure
   [140]5
   D). The most prominent biological process clusters include the alcohol
   metabolic process (adjusted p = 2.128 × 10^–8), secondary alcohol
   metabolic process (adjusted p = 2.128 × 10^–8), and cholesterol
   metabolic process (adjusted p = 1.586 × 10^–7). The results of the
   biological process enrichment analysis parallel those obtained from the
   KEGG database, signifying the profound influence of matrine on lipid
   metabolism in A549 cells.

   Additionally, we performed GSEA utilizing the Reactome database
   ([141]Figure [142]5
   E). As expected, the upregulated proteins exhibited pronounced
   enrichment in fatty acid metabolism (adjusted p = 2.18 × 10^–2),
   metabolism of lipids (adjusted p = 2.18 × 10^–2), and LDL clearance
   (adjusted p = 2.99 × 10^–2). To delve further, we explored the
   interplay among differentially expressed proteins through the STRING
   database (Supporting Information [143]Figure S1). Consistent with the
   above results, the upregulated proteins exhibited strong correlations
   with lipid metabolism.

   Together, matrine unquestionably wields a potent influence on lipid
   metabolism pathways. Furthermore, we corroborated the enhancement of
   LDLR expression by matrine ([144]Figure [145]5
   F). Nevertheless, whether matrine exerts its antilung cancer function
   through LDLR upregulation, subsequently prolonging LDL uptake,
   necessitates further in-depth investigation.

Identification of Potential Protein Targets of Matrine through TPP

   In pursuit of potential protein targets of matrine, we performed TPP
   analysis. TPP is a method that integrates the Cellular Thermal Shift
   Assay (CETSA) with quantitative mass spectrometry. This approach is
   based on the concept of ligand-induced thermostabilization of target
   proteins, providing insights into the affinity between drugs and target
   proteins derived from cellular or tissue samples.^[146]40−[147]42

   Considering the practical research requirements, we conducted the TPP
   experiments only on A549 cells. Initially, A549 cells were treated with
   matrine or DMSO for 1.5 h. Subsequently, cells from both groups were
   harvested and divided into eight aliquots for gradual heating. After
   heating, the samples underwent lysis using 1% NP-40, and the
   supernatant containing soluble proteins was collected through
   centrifugation. Although pellet analysis is considered helpful in
   identifying more binding proteins in a small number of studies,^[148]43
   the most widely accepted TPP method currently involves analyzing the
   supernatant after heating. Therefore, we employed the well-established
   method. The soluble proteins were then labeled using Tandem Mass Tag
   (TMT) reagent and subjected to LC–MS/MS analysis (Supporting
   Information [149]Figure S2A). To ensure the robustness of our
   experiment, we conducted two replications, and a high degree of
   correlation was evident between these replications (Supporting
   Information [150]Figure S2B).

   Data analysis of the TPP revealed a total of 5398 identified proteins
   (Supporting Information [151]Table S3). Within these, 4322 proteins
   were identified in the matrine-treated group, while 4319 proteins were
   identified in the DMSO-treated group. Given the denaturation and
   insolubilization of proteins upon exposure to heat,^[152]44 a reduction
   in the abundance of soluble proteins was indeed observed with
   increasing temperature (Supporting Information [153]Figure S2C).

   Next, we selected 3729 proteins that were identified in both
   matrine-treated and DMSO-treated groups to generate melting curves.
   Unfortunately, 1206 proteins were unable to generate a melting curve
   due to the presence of some missing values. We subsequently compared
   the melting points of the remaining 2523 proteins. The results
   indicated an increase in the melting point for some proteins in the
   matrine-treated group compared to the same proteins in the DMSO-treated
   group (Supporting Information [154]Figure S2D). These proteins were
   ranked based on the magnitude of their ΔTm (ΔTm = |Tm(matrine) –
   Tm(DMSO)|) (Supporting Information [155]Figure S2E). Of note, 115
   proteins were shown as orange data points with ΔTm > 1 and a p value
   <0.05, signifying a significant thermal shift (Supporting Information
   [156]Figure S2F). To reinforce the reliability of the experiment,
   proteins were required to be identified in both biological replicates
   (Supporting Information [157]Figure S2G). Ultimately, we pinpointed 76
   candidate proteins.

   Among these candidate proteins, we identified four linked to lipid
   metabolism (Supporting Information [158]Table S4 and [159]Figure S3).
   Notably, voltage-dependent anion-selective channel protein 1 (VDAC1) is
   a mitochondrial porin involved in apoptosis and influenced by
   lipids.^[160]45 Delta (24)-sterol reductase (DHCR24) is a critical
   enzyme in cholesterol biosynthesis.^[161]46,[162]47 Acyl-coenzyme A
   diphosphatase FITM2 (FITM2) is an endoplasmic membrane protein
   implicated in promoting lipid accumulated in cell culture.^[163]47 In
   addition, ABCD3 is an ATP-dependent transporter within the ATP-binding
   cassette (ABC), playing a pivotal role in fatty acids or fatty
   acyl-coenzyme A transport.^[164]48 Moreover, we performed enrichment
   analysis with the GO database to probe the effects of the other 72
   proteins (Supporting Information [165]Figure S4). Interestingly, the
   most prominent biological process clusters of these proteins were
   mainly involved in the cellular nitrogen compound metabolic process
   (adjusted p = 4.90 × 10^–3) and nitrogen compound metabolic process
   (adjusted p = 6.20 × 10^–3). These pathways are closely related to
   lipid metabolism, as lipid synthesis is influenced by the glutamate
   dehydrogenase pathway, and the main product of this pathway, glutamine,
   plays a crucial role in nitrogen metabolism.^[166]49,[167]50 This
   result suggests a potential association between matrine and lipid
   metabolism, with these four proteins serving as potential targets of
   matrine in cancer therapy.

Discussion

   This study delves into the molecular mechanism underlying Matrine’s
   potential for HLP treatment using quantitative proteomics. In HepG2
   cells, we observed that matrine upregulated both the mRNA and protein
   levels of LDLR, subsequently enhancing Dil-LDL uptake by the cells.
   Remarkably, this effect extended to the livers of hamsters, with
   matrine similarly upregulating LDLR expression. Additionally, we
   demonstrated that matrine effectively ameliorated disordered lipid
   metabolism in hamsters fed with HFD. Both in vitro and in vivo
   experiments firmly established that matrine’s HLP-treating prowess
   involves the upregulation of LDLR expression. Based on the results from
   our study and literature, we speculate that matrine may potentially
   impact LDLR mRNA expression through various mechanisms, such as by
   influencing the activity of transcription factors to regulate LDLR gene
   transcription, or by affecting the stability of LDLR mRNA to regulate
   its levels. However, our current data are not sufficient to derive the
   exact mechanism. To explore potential protein targets of matrine, we
   attempted TPP experiments and observed significant effects of matrine
   on several lipid metabolism proteins. Further analysis revealed that
   matrine can also influence the lipid biosynthetic process and nitrogen
   metabolic pathway. Based on these findings, we hypothesize that matrine
   not only alters LDLR expression levels through transcriptional
   regulation but also modulates lipid metabolism by affecting metabolic
   pathway proteins. In addition, since matrine primarily influences lipid
   metabolism and other metabolic pathways, metabolomics research will
   undoubtedly enhance our understanding of matrine’s metabolic regulation
   in future studies,^[168]51,[169]52 which will provide a more
   comprehensive insight into the molecular pathways modulated by matrine.

   Recognizing the connection between disrupted metabolism and cancer
   development, and taking into account the documented potential of
   matrine for antilung cancer activity, we also sought to determine
   whether matrine’s anticancer effect is mediated by its regulation of
   disordered lipid metabolism. Astonishingly, our investigation revealed
   that matrine-induced upregulated proteins in A549 cells were
   predominantly enriched in lipid metabolism-related pathways. More
   intriguingly, matrine also upregulated LDLR expression in A549 cells.
   Given that increased LDLR expression and extended LDL uptake can prove
   detrimental to cancer cells, our primary results have provided valuable
   insights into the anticancer properties of matrine.

   Currently, statins remain the primary choice for HLP treatment.^[170]7
   Statins, acting as inhibitors of HMG-CoA reductase, curb cholesterol
   biosynthesis, induce LDLR upregulation through feedback, and
   subsequently bolster LDL uptake. However, about 15% of patients fail to
   achieve target HLP treatment goals with statins alone.^[171]53,[172]54
   Therefore, alternative nonstatin medications have been explored. For
   example, PCSK9 inhibitors, including monoclonal antibodies and small
   molecules, have emerged as clinical options.^[173]55,[174]56 Given the
   differing mechanisms of action among matrine, PCSK9 inhibitors, and
   statins, their combined usage holds the potential to yield favorable
   therapeutic effects.

   TPP is currently one of the most powerful tools for investigating
   drug–protein interactions, yet it also has the potential to result in
   the loss of such interaction information. For instance, certain
   proteins may degrade at extremely high or low temperatures, or some may
   show insignificant changes in thermal stability following matrine
   binding. Moreover, mass spectrometry-based proteomics techniques may
   not comprehensively reflect all proteins in the sample, particularly
   those with lower abundance yet crucial in the system. In our study, TPP
   analysis identified 76 potential matrine-binding proteins, primarily
   concentrated in metabolic pathways. Despite matrine exhibiting potent
   pharmacological activity, its binding proteins largely remain
   unidentified, thereby rendering our findings valuable in providing data
   support for matrine target discovery. Further biological experiments
   are also necessary to confirm the existence of these targets.

Conclusions

   This study investigated the molecular mechanism underlying matrine’s
   potential for treating HLP in in vitro and in vivo experiments.
   Employing quantitative proteomics, we discovered that matrine increases
   the expression of LDLR in cells, with subsequent experiments validating
   its role in enhancing LDL uptake. It effectively lowers lipid levels in
   hyperlipidemic hamsters without affecting body weight, indicating LDLR
   as a critical target for matrine’s impact on HLP.

Acknowledgments