Abstract

   Current research efforts in polymer and nanotechnology applications are
   primarily focused on cargo delivery to enhance the therapeutic index,
   with limited attention being paid to self‐molecularly targeted
   nanoparticles, which may also exhibit significant therapeutic
   potential. Long‐term and anomalous lipid accumulation in the liver is a
   highly relevant factor contributing to liver diseases. However, the
   development of the reliable medications and their pharmacological
   mechanisms remain insufficient. Herein, a polyguanide nanoinhibitors
   (PGNI) depot is constructed by copolymerizing biguanide derivatives in
   different proportions onto prepolymers. The nanoinhibitors for their
   ability to ameliorate lipid accumulation in vitro and in vivo is
   screened, and subsequently demonstrated that covalently polymeric
   guanidine chains exhibit superior efficacy in ameliorating hepatic
   lipid accumulation via heterogeneous mechanisms compared to
   small‐molecule guanidine. It is found that PGNIs stabilize guanidine
   metabolism in the liver, preferably for biosafety. More importantly,
   PGNI is ingested and localized in hepatocyte lysosomes and is locked to
   interact with vesicular adenosine triphosphatase (V‐ATPase) on
   lysosomes, leading to the inhibition of V‐ATPase and lysosomal
   acidification, thereby activating the AMPK pathway, reducing fatty acid
   synthesis, and enhancing lipolysis and fatty acid oxidation. These
   results imply that polymer‐formed nanoparticles can serve as targeted
   inhibitors, offering a novel approach for therapeutic applications.

   Keywords: AMPK, liver lipid accumulation, lysosome, polyguanide
   nanoinhibitor (PGNIs), V‐ATPase
     __________________________________________________________________

   In this work, a polyguanide nanoinhibitors (PGNI) depot is constructed
   by copolymerizing biguanide derivatives in different proportions onto
   prepolymers. The nanoinhibitors for their ability to ameliorate lipid
   accumulation in vitro and in vivo is screened, and subsequently
   demonstrated that covalently polymeric guanidine chains exhibit
   superior efficacy in ameliorating hepatic lipid accumulation via
   heterogeneous mechanisms compared to small‐molecule guanidine. More
   importantly, PGNI is ingested and localized in hepatocyte lysosomes and
   is locked to interact with vesicular adenosine triphosphatase
   (V‐ATPase) on lysosomes, leading to the inhibition of V‐ATPase and
   lysosomal acidification, thereby activating the AMPK pathway, reducing
   fatty acid synthesis, and enhancing lipolysis and fatty acid oxidation.
   These results imply that polymer‐formed nanoparticles can serve as
   targeted inhibitors, offering a novel approach for therapeutic
   applications.

   graphic file with name ADVS-12-2408906-g007.jpg

1. Introduction

   Nonalcoholic fatty liver disease (NAFLD), the most common chronic liver
   disease, affects one‐quarter of the global population worldwide and is
   associated with metabolic syndrome and obesity, whose incidence is
   increasing.^[ [52]^1 , [53]^2 , [54]^3 ^] Reducing hepatic lipid
   deposition is the key to inhibition of NAFLD progression.^[ [55]^4 ^]
   However, the use of medication to ameliorate of hepatic lipid
   deposition is limit.^[ [56]^5 , [57]^6 , [58]^7 ^] The reversal of
   excessive hepatic lipid deposition remains a challenge in drug design.

   At doses far greater than those tolerable in actual use (300 mg kg^−1
   in mice, converted to an adult dose of 2.7 g kg^−1), small‐molecule
   guanidine compounds, with metformin as a typical representative, have
   demonstrated the potential to improve intracellular lipid deposition to
   varying degrees in hepatocytes and animal experiments,^[ [59]^8 ,
   [60]^9 , [61]^10 , [62]^11 ^] however, the clinical data on their
   pharmacodynamics are still highly controversial.^[ [63]^12 , [64]^13 ^]
   Notably, the pharmacological effects of this class of drugs depend on
   the guanidine group.^[ [65]^14 ^] However, hydrophilic guanidineresults
   in low drug bioavailability. For example, many studies have shown that
   metformin uptake and elimination occur at similar rates in the liver
   via bidirectional transport.^[ [66]^15 , [67]^16 ^] This factor may
   limit the effective concentration of small‐molecule guanidine compounds
   in the liver and explain the intolerable dose of metformin for NAFLD
   amelioration and the clinical dispute mentioned above.

   ngly, we previously discovered that polymerized guanidine has more
   effective weight control and reduced liver lipid deposition in vivo
   than small‐molecule guanidine compounds.^[ [68]^17 ^] We speculate the
   unique long‐chain structure, molecular weight, and intermolecular
   forces of polymers cause the guanidine groups to exhibit entirely
   different quantitative and qualitative changes in terms of structural
   characteristics and biological effects. First, in terms of structure,
   each molecule of polymeric guanidine contains more guanidine groups.
   These additional guanidine groups may render it more efficacious
   compared to small molecular guanidine. Second, with the increase of the
   molecular weight of the polymeric compound, the molecular chain becomes
   longer and more complex, and the intermolecular forces will also
   change. The intermolecular forces between polymerized guanidine
   ameliorate small‐molecule guanidine compounds excessive water
   solubility to a certain extent while forming nanoparticles in the
   aqueous phase. Leveraging the natural targeting advantage of
   nanoparticles to the liver, polymeric guanidine has a longer retention
   time in the liver, enhancing the bioavailability of guanidine in the
   liver. However, these polymeric compounds also have complex
   characteristics. The molecular weight (Mw), architecture, assembly
   structure, and subcellular location affect the pharmacodynamics of
   polymeric drugs.^[ [69]^18 ^] To solve this problem, in this study, a
   methacrylic biguanide derivative
   N‐(2‐(3‐carbamimidoylguanidino)ethyl)methacrylamide (GuC) was designed
   and synthesized, and controllable copolymerization was performed on
   three universal medically degradable polymers (polyethylene glycol
   (PEG), polylactic‐co‐glycolic acid (PLGA), and polylactic acid (PLA))
   by reversible addition‐fragmentation chain transfer (RAFT) using a
   modified prepolymer in different proportions to fabricate 15
   guanidine‐derived copolymers, followed by their dynamic self‐assembly
   to spherical nanoinhibitors (including PEG‐ b ‐P(Gu)1‐5, PLGA‐ b
   ‐P(Gu)1‐5, PLA‐ b ‐P(Gu)1‐5).

   To identify the druggable interaction site, we focused on the vesicular
   adenosine triphosphatase (V‐ATPase) protein in lysosomes for three
   reasons: First, most of the macromolecules are mainly ingested into
   cells through endocytosis,^[ [70]^19 , [71]^20 , [72]^21 ^] similarly,
   PGNIs is localized to lysosomes; and the pathway related to
   phagocytosis in the liver is significantly activated after
   administration. Thus, we suspected that PGNIs mainly interacts with
   lysosomal related proteins for pharmacological effects, which is differ
   from the fact that the small molecule guanidine mainly enters
   hepatocytes through organic cation transporters 1 or 3 (OCT1, OCT3).^[
   [73]^22 , [74]^23 , [75]^24 ^] Second, recent studies have found an
   intimate relationship between V‐ATPase and guanidine groups.^[ [76]^25
   ^] Specifically, low doses of metformin target the lysosomal membrane
   protein presenilin enhancer 2 (PEN2), allowing PEN2 to gain affinity
   for the V‐ATPase subunit ATP6AP1 and form a complex. This process leads
   to the inhibition of V‐ATPase, which in turn changes the conformation
   of V‐ATPase and Ragulator (LAMTOR) and then initiates the recruitment
   of the axis protein (AXIN) and liver kinase B1 (LKB1) to the outer
   lysosomal leaflets, where they bind to V‐ATPase/LAMTOR. Finally,
   AMP‐activated protein kinase (AMPK) is activated.^[ [77]^25 ^]
   Activated AMPK inhibits de novo fatty acid synthesis, and promotes
   lipolysis and fatty acid oxidation, ultimately reducing cellular lipid
   accumulation. Third, in our previous study, carriers containing
   guanidine groups showed evidence of a lysosomal escape effect to a
   certain degree.^[ [78]^17 ^] Notably, this effect is often closely
   related to the consumed H^+ and the increased pH of lysosome. V‐ATPase
   is the molecular motor protein that controls the acidic environment of
   lysosomes. Inhibition of V‐ATPase will cause a decrease in the ability
   of the lysosomal pump to transport H^+, leading to an increase in pH.

   In this study, we fabricated a series of PGNI depots (Figure [79] 1A)
   by RAFT polymerization. By controlling the quantity of guanidine, it
   was copolymerized onto PEG/PLGA/PLA‐modified 4‐cyano‐4‐(phenyl
   thioformyl sulfur) valeric acid (PEG‐CPADN/PLGA‐CPADN/PLA‐CPADN), which
   self‐assembled into nanoparticles. As depicted in Figure [80]1B, upon
   administration, the PGNIs naturally accumulate in the liver and are
   taken up by hepatocytes. After entering the lysosome, the PGNIs
   interact with and target the V[0] region of V‐ATPase in the inner
   leaflets of the lysosome. This interaction leads to an inhibition of
   V‐ATPase activity, resulting in the formation of
   V‐ATPase‐Ragulator‐AXIN/LKB1‐AMPK complex and activation of the AMPK
   pathway. Furthermore, activated AMPK inhibits fatty acid de novo
   synthesis, and promotes lipolysis and β‐oxidation, ultimately
   ameliorating lipid accumulation in the liver (Figure [81]1C).

Figure 1.

   Figure 1
   [82]Open in a new tab

   Diagram of the synthesis of polyguanidine nanomedicine for ameliorating
   liver accumulation. A) The synthesis and self‐assembly process of PEG‐
   b ‐P(Gu), PLGA‐ b ‐P(Gu), and PLA‐ b ‐P(Gu). B) Polyguanidine
   nanomedicine ameliorates hepatic lipid accumulation in mice and its
   lysosomal uptake in hepatocytes. C) PEG‐ b ‐P(Gu) accumulates at the
   lysosome and further triggers AMPK activation. PEG‐ b ‐P(Gu) binds to
   the inner leaflet of the V‐ATPase and inhibits its activity, followed
   by AXIN and LKB1 being recruited to the V‐ATPase and Regulator (LAMTOR)
   for inducing AMPK phosphorylation and subsequently promoting lipolysis
   and reducing the fatty acid de novo synthesis.

2. Results and Discussion

2.1. Macromolecule Design and Synthesis

   As shown in Figure [83]S1 (Supporting Information), PEG‐CPADN (compound
   2), PLGA‐CPADN (compound 5), and PLA‐CPADN (compound 7) were
   synthesized. The ‐CONH‐ bonds of the compound structures were confirmed
   by infrared spectroscopy (Figures [84]S2–S4, Supporting Information).
   Subsequent RAFT copolymerizations at different molar ratios (Figure
   [85]S5, Supporting Information) of
   N‐(2‐(3‐carbamimidoylguanidino)ethyl)methacrylamide (GuC, compound 3)
   monomers afforded PEG‐ b ‐P(Gu)1‐5 (compound 4), PLGA‐ b ‐P(Gu)1‐5
   (compound 6), and PLA‐ b ‐P(Gu)1‐5 (compound 8) using PEG‐CPADN,
   PLGA‐CPADN, PLA‐CPADN, and azodiisobutyronitrile (AIBN) as
   macroinitiators. Further, ^1H nuclear magnetic resonance (^1H‐NMR,
   Figures [86]S6–S8, Supporting Information) and gel permeation
   chromatography (GPC, Figure [87]2B; Figure [88]S9, Supporting
   Information) revealed an increase in the proportion of guanidine
   monomers and the average molecular weight (Mw) of the polymers. The
   theoretical number of guanidine monomers was calculated based on the
   molecular weight (Figure [89]2A). Owing to the positively charged
   nature of guanidine, its zeta potential, which was also used to
   characterize the degree of guanidine polymerization (Figure [90]S10,
   Supporting Information), increased from 15 to 25 mV as the number of
   guanidine monomers increased.

Figure 2.

   Figure 2
   [91]Open in a new tab

   Basic characterization and lipid deposition lowering effect of
   polyguanidine nanomedicine in vitro. A) Average molecular weights (Mw)
   of a library of 15 polyguanidine nanomedicines synthesized by different
   feed ratios and the theoretical values of the guanidine chains
   calculated by Mw. B) Chemical formulas of PEG‐ b ‐P(Gu), PLGA‐ b
   ‐P(Gu), and PLA‐ b ‐P(Gu) and their 3D self‐assembly schematic
   diagrams. C) Molecular dynamics simulation of polyguanidine
   conformation change in 100 ns and D) TEM image of the polyguanidine
   nanomedicine library. Scale bar: 200 nm. E) Cell viability of HepG2
   cells treated with metformin, PEG‐ b ‐P(Gu)1‐5, PLGA‐ b ‐P(Gu)1‐5, and
   PLA‐ b ‐P(Gu)1‐5 in different concentrations. F) Triglyceride levels in
   HepG2 cells, including negative control, positive control (palmitic
   acid loading), metformin, PEG‐ b ‐P(Gu)1‐5, PLGA‐ b ‐P(Gu)1‐5, and PLA‐
   b ‐P(Gu)1‐5. G) Bodipy and Oil Red O staining and lipid content rate of
   HepG2 cells treated with PEG‐ b ‐P(Gu)1‐5, PLGA‐ b ‐P(Gu)1‐5, and PLA‐
   b ‐P(Gu)1‐5. Scale bar: 5 µm; ^* P < 0.05; ^** P < 0.01; bars represent
   the mean ± SD (n = 3). The P‐values are calculated using one‐way ANOVA
   with Bonferroni correction.

2.2. Molecular Dynamics Simulation and Characterization of PGNIs

   To explore the dynamic behavior of polyguanidines in the aqueous phase,
   molecular dynamics (MD) was used to simulate the formation processes of
   PGNIs. The formulae and self‐assembled conformations are shown in
   Figure [92]2B. Furthermore, cluster analysis was performed on the
   simulated trajectories ranging from 80 to 100 ns using the GROMACS
   cluster module. As shown in Figure [93]2C, the polyguanidines
   interacted with each other and formed spherical nanoclusters. To
   further investigate the driving forces during the interaction and
   aggregation processes of the system, the interactions between the
   polyguanidines were analyzed. As shown in Figure [94]S11A (Supporting
   Information), the N atoms on the polyguanidines formed a large number
   of hydrogen bonds with other surrounding polyguanidines. The number of
   hydrogen bonds (Figure [95]S11B, Supporting Information),
   root‐mean‐square deviation (RMSD; Figure [96]S11C, Supporting
   Information), and solvent‐accessible surface area (SASA; Figure
   [97]S11D, Supporting Information) tended to stabilize after 60 ns,
   indicating stable aggregation of the polyguanidines. Next, the
   interaction energies between the polyguanidines in the system were
   calculated. Figure [98]S11E (Supporting Information) shows the total
   interaction energy in the system. With polyguanidine aggregates in the
   system, the interaction energy between the molecules also increased.
   The electrostatic interaction between polyguanidines is significantly
   stronger than van der Waals interactions (the Coulomb potential
   represents electrostatic interactions, with hydrogen bonding as the
   main type, whereas the Lennard‐Jones potential represents van der Waals
   interactions), which suggests that the stronger electrostatic
   interaction energy between the two molecules drives their self‐assembly
   in aqueous solvents to form stable nanoclusters (Figure [99]S11F,
   Supporting Information). Finally, transmission electron microscopy
   (TEM; Figure [100]2D) revealed that each of PEG‐ b ‐P(Gu)1‐5, PLGA‐ b
   ‐P(Gu)1‐5, and PLA‐ b ‐P(Gu)1‐5 self‐assembled into nanoparticles. The
   sizes of PEG‐ b ‐P(Gu) and PLA‐ b ‐P(Gu) were approximately in the
   range of 150–200 nm, whereas those of PLGA‐ b ‐P(Gu) were 30–50 nm.
   These differences in particle size may be attributed to the ester bonds
   of lactic acid and glycolic acid in PLGA molecular chains, which endow
   these nanoparticles with stronger hydrophobic interactions and
   self‐assembly capabilities during nanoparticle formation. This leads to
   a tendency to form tighter particle structures and consequently smaller
   sizes. In addition, we used Dynamic Light Scattering (DLS) and PDI
   characterizations that evaluate the particle size distribution and
   aggregation behavior of these nanoparticles in aqueous solutions, to
   assess their stability. Based on these results (Figure [101]S12,
   Supporting Information), we observe that the particle sizes of the
   nanoparticles remain relatively unchanged within the 7‐day period,
   exhibiting preferable colloidal stability.

2.3. Small Molecule Guanidines and PGNIs Ameliorate Lipid Accumulation In
Vitro

   In a cytotoxic test using the CCK8 assay, we found that the viability
   of HepG2 and MIHA cells treated with PGNIs was also facilitated, and
   kupffer cells is maintained at normal levels (Figure [102]2E; Figures
   [103]S13A and [104]S14, Supporting Information). To evaluate the effect
   of PGNIs on lipid deposition, the levels of triglycerides (TG)
   (Figure [105]2F; Figure [106]S13B, Supporting Information) and
   Bodipy‐labeled lipid droplets (Figure [107]2G) were detected in HepG2
   cells, and Oil Red O staining was performed (Figure [108]2G). All
   results indicated that both small‐molecule guanidine, represented by
   metformin, and large‐molecule guanidine PGNIs ameliorated liver lipid
   deposition in vitro, with PGNIs exhibiting a slightly better effect
   than small‐molecule guanidine.

2.4. PGNIs Ameliorate Liver Lipid Accumulation In Vivo

   Based on the above results, for subsequent experiments, we selected
   those numbered three for each type of PGNI. Because these PGNIs
   demonstrated a relatively prominent and consistent performance in
   ameliorating hepatic triglyceride levels. While PGNIs numbered 4 and 5
   also showed some improvement in lipid deposition, this enhancement did
   not follow a pronounced dose‐response trend. Consequently, the
   selection of PGNI number 3 aligns more closely with the principles of
   economy and high efficiency. The capacity of small‐molecule guanidine
   and PGNIs to ameliorate lipid deposition was first investigated in the
   liver. As a proof of concept, HFD‐fed mice were intravenously (i.v.)
   injected with metformin and different PGNIs twice a week for 16 weeks
   (Figure [109]3A). As Figure [110]3B,C show, the administration of PGNIs
   flattened the steep body and liver weight gains. This anti‐obesity
   effect was not caused by reduced food intake (Figure [111]3D); it was
   exclusively accounted for by the inhibition of lipid deposition in the
   body, liver, and serum by PGNIs, especially PEG‐ b ‐P(Gu)
   (Figure [112]3E–H; Figure [113]S15, Supporting Information). On the
   contrary, at this dose, metformin did not have any effect. Moreover,
   PGNIs relieved insulin resistance (Figure [114]3I,J). To exclude the
   liver lipid‐lowering effects caused by the charge, other compositions,
   and sizes of the nanoparticles, a classic cationic dendritic polymer
   (P‐G3)^[ [115]^26 ^] and PEG were selected for administration to
   HFD‐fed mice. As depicted in Figure [116]S16A–D (Supporting
   Information), PEG did not have any effect on body and liver weights,
   while suppression was observed in mice that were administered P‐G3;
   however, it was not as significant as in the mice treated with PGNIs.
   In terms of improving lipid deposition in the liver and serum, there
   was no significant difference between the P‐G3 and PEG‐administered
   groups compared to the HFD group (Figure [117]S16E–H, Supporting
   Information). Based on these results, we speculated that the guanidine
   group in PGNIs might induce this pharmacological activity.

Figure 3.

   Figure 3
   [118]Open in a new tab

   Amelioration of HFD‐induced lipid deposition in mice by polyguanidine
   nanomedicine administration. Small molecule guanidines (e.g.,
   metformin) do not show this lipid lowering effect. A) Schematic of
   hepatic lipid deposition reduction in C57BL/6J mice and the treatment
   schedule when using PEG‐ b ‐P(Gu), PLGA‐ b ‐P(Gu), or PLA‐ b ‐P(Gu). B)
   Body weight, C) liver weight, and D) feed intake of C57BL/6 mice
   treated with different medicines. Bars represent the mean ± SD (n = 6).
   E) Images of the whole‐body and liver of mice at the end of treatment.
   Scale bar: 1 cm. F) Hepatic total triglyceride and G) free fatty acids
   levels in mice after 16 weeks of treatment with metformin, PEG‐ b
   ‐P(Gu), PLGA‐ b ‐P(Gu), or PLA‐ b ‐P(Gu). Bars represent the mean ± SD
   (n = 6). H) H&E and Oil Red O staining images of liver cells collected
   from mice administrated with metformin, PEG‐ b ‐P(Gu), PLGA‐ b ‐P(Gu),
   or PLA‐ b ‐P(Gu). I) Results of glucose tolerance tests (GTTs) and(J)
   insulin tolerance tests (ITTs) conducted on HFD‐fed mice before and
   after the administration of metformin and polyguanidine nanomedicine
   (n = 6). ^*** P < 0.001; ^** P < 0.01; ^* P < 0.05. The P‐values are
   calculated using one‐way ANOVA.

2.5. Biosafety of PGNIs

   In addition to studying the ameliorating effect of PGNIs on lipid
   deposition, their safety was assessed in animal toxicology studies
   (Figure [119]S17, Supporting Information). No abnormal levels of
   aminotransferase (ALT), aspartate aminotransferase (AST), urea nitrogen
   (BUN), or creatinine (S‐Cr) were observed in mice that were
   administered PEG‐ b ‐P(Gu), PLGA‐ b ‐P(Gu), or PLA‐ b ‐P(Gu).
   Hematoxylin and eosin (H&E) staining of major organ sections also
   indicated no observable influence on other organs.

2.6. Biodistribution In Vitro and In Vivo

   To explore the reasons for the inconsistent lipid‐lowering effects of
   guanidines with different molecular weights or prepolymer types in
   vitro and in vivo, we investigated the dynamic state of the PGNIs in
   vivo. First, PGNIs labeled with indocyanine green (ICG/PEG‐ b ‐P(Gu),
   ICG/PLGA‐ b ‐P(Gu), and ICG/PLA‐ b ‐P(Gu)) were produced (Figure
   [120]4A). The following experiment showed that the three different
   aggregation types of PGNIs exhibited different uptake abilities in
   liver cells (Figure [121]4B). Four groups of animals were given a
   single dose of ICG: ICG/PEG‐ b ‐P(Gu), ICG/PLGA‐ b ‐P(Gu), or ICG/PLA‐
   b ‐P(Gu). As shown in Figure [122]4C, the three types of PGNIs showed
   significant liver accumulation at 0.5–4 h. 8 h after injection,
   fluorescence in the liver gradually decreased. In contrast, the ICG
   group exhibited rapid and complete metabolism, because fluorescence was
   observed in the kidneys after 4 h. In addition, the ex vivo
   fluorescence image and quantitative diagram (Figure [123]4D) of the
   excised organs further confirmed that ICG/PEG‐ b ‐P(Gu) showed a higher
   fluorescence signal in the liver than in the other groups. Next,
   immunofluorescence staining was performed on the liver at 4 h after
   dose administration to further explore the dynamics of PGNIs within the
   liver. The PEG‐ b ‐P(Gu)/ICG group showed a higher localization in
   hepatocytes (marked as HNF4‐α) compared to other groups
   (Figure [124]4F), which might explain why PEG‐ b ‐P(Gu) exhibits better
   lipid‐lowering effects in vivo. Hence, PEG‐ b ‐P(Gu)3 was used for
   further experiments. As shown in Figure [125]4E, three endocytosis
   inhibitors, amiloride, methyl‐β‐cyclodextrin (MβCD), and
   chlorpromazine, were employed to verify the macropinocytosis,
   caveolae‐mediated endocytosis, and clathrin‐mediated endocytosis
   efficiency, respectively.^[ [126]^27 ^] The endocytosis efficiency was
   significantly decreased in cells treated with chlorpromazine,
   suggesting that PEG‐ b ‐P(Gu) entered the cells via macropinocytosis
   and caveolae‐ and macropinocytosis‐related pathways, as all inhibitors
   slowed down cellular uptake. An analysis of the drug concentration‐time
   curve suggested that the half‐life of PEG‐ b ‐P(Gu) was ≈4 h longer
   than that of metformin (0.5 h) (Figure [127]4G). At 4 h, PEG‐ b ‐(Gu)
   was mainly localized in the lysosomes within the liver cells
   (Figure [128]4H) rather than in other organelles (Figure [129]S18,
   Supporting Information). Finally, fluorescence resonance energy
   transfer (FRET) was used to investigate the action of nanoparticles
   within cells. This process occurs when the emission spectrum of the
   donor overlaps with the absorption spectrum of the acceptor and the
   donor and acceptor are in close proximity.^[ [130]^28 ^] As shown in
   Figure [131]4I, the ratio of the FRET fluorescence pair did not undergo
   any significant change during the time in which the nanoparticles
   resided within the lysosomes, suggesting that the nanoparticles
   produced the effect as intact particles.

Figure 4.

   Figure 4
   [132]Open in a new tab

   Biodistribution and pharmacokinetics profile of polyguanidine
   nanomedicine in vitro and in vivo. A) Schematic of self‐assembly of
   polyguanidine nanomedicine with the corresponding fluorescent labels.
   B) FCM analysis and quantification of polyguanidine nanomedicine
   cellular uptake in HepG2 cells (n = 3). C) Luminescence images D)
   quantification of C57BL/6J mice following treatment with PEG‐ b
   ‐P(Gu)/ICG, PLGA‐ b ‐P(Gu)/ICG, or PLA‐ b ‐P(Gu)/ICG in vitro and in
   vivo (n = 3) E) Inhibition of endocytosis by different inhibitors,
   including amiloride, mβCD, chlorpromazine, and 4 °C for detecting
   cellular uptake pathways (n = 3). F) Immunofluorescence image of mouse
   liver treated with PEG‐ b ‐P(Gu)/ICG (blue: nucleus; green: hepatic
   parenchymal cell; red: PEG‐ b ‐P(Gu); scale bar: 50 µm). G) Drug
   concentration‐time curve of PEG‐ b ‐P(Gu) and metformin in mouse liver.
   Bars represent the mean ± SD (n = 6). H) CLSM and Mander's overlap of
   PEG‐ b ‐P(Gu) cellular uptake and accumulation in the lysosome. Blue:
   nucleus; green: lysosome; red: PEG‐ b ‐P(Gu); yellow: merged. Scale
   bars: 50 µm. I) FRET ratio of particles prepared with Dio and Dil after
   incubation in HepG2 cells. Bars represent the mean ± SD (n = 3). ^***
   P < 0.001; ^** P < 0.01; ^* P < 0.05. The P‐values are calculated using
   one‐way ANOVA.

2.7. PEG‐b‐P(Gu) Inhibits Lysosomal Activity and Raises Lysosomal pH

   In addition to the dynamic state in vivo, the mechanisms underlying the
   pharmacological effects of polyguanidine must be investigated. The
   transcriptome differences in liver tissues were analyzed using RNA
   sequencing. The volcano plot (Figure [133]5A) revealed that, compared
   with the HFD‐fed group (control), 646 genes were upregulated, and 172
   genes were downregulated after the administration of PEG‐ b ‐P(Gu).
   KEGG analysis indicated that phagocytosis‐related pathways were
   significantly activated (Figure [134]5B), and this activation was not
   attributed to inflammation (Figure [135]S19, Supporting Information).
   Among these phagocytosis‐related proteins, the increase in the
   transcriptome (Figure [136]5C) and protein expression (Figure [137]S20,
   Supporting Information) of ATP6V0D2, a “d” subunit of V‐ATPase, has
   attracted our attention. V‐ATPase primarily controls the acidic
   environment of lysosomes and is closely associated with their
   activity.^[ [138]^29 ^] In its V[0] domain, V‐ATPase is mainly divided
   into multiple subunits such as “a”, “c”, and “d”. The “a” subunit is
   the core component of V‐ATPase and directly participates in the
   conduction of H^+. The “c” subunit is a major component of the proton
   channel, while the “d” subunit is mainly used to maintain the stability
   of the proton channel,^[ [139]^30 , [140]^31 ^] furthermore, study have
   also shown evidence that ATP6V0D2 can promote the fusion of
   autophagosomes and lysosomes.^[ [141]^32 ^] Therefore, despite the
   lysosomal pathway not being prominent in the KEGG analysis
   (Figure [142]5C), as part of the phagocytic pathway, it and V‐ATPase
   may still play an important role. Thus, we further examined lysosomal
   activity in HepG2 cells after treatment with PEG‐ b ‐P(Gu) and observed
   a significant downregulation (Figure [143]5D–F) and pH increase (Figure
   [144]S21, Supporting Information), but the pH remained within the
   optimum range of lysosome (4.5–5.0). An increase in lysosomal pH is
   generally detrimental to lipid degradation and autophagy,^[ [145]^33 ^]
   current research even developed nanoparticles that directly act on
   lysosomal acidification.^[ [146]^34 ^] However, the effect of V‐ATPase
   inhibition on lipid deposition is not solely dependent on lysosomal pH
   modulation. Rather, it may be attributed to V‐ATPase's intricate roles
   in cellular signaling and lipid metabolism regulation. But what is even
   more important is that, since ATP6V0D2 does not participate in proton
   transport within the V‐ATPase, it is worthwhile to further investigate
   exactly how PEG‐ b ‐P(Gu) leads to an increase in the lysosomal pH
   value, which subsequently upregulates ATP6V0D2, there may be a feedback
   regulation mechanism involved.

Figure 5.

   Figure 5
   [147]Open in a new tab

   Inhibition of lysosomal V‐ATPase activity by PEG‐ b ‐P(Gu), which
   targets the V[0] region. A) Volcano plots of differentially expressed
   genes in HFD‐fed mouse liver and comparison with those of
   PEG‐administered mouse liver. FC, fold change. The upregulated and
   downregulated genes are shown in red and blue, respectively. B) KEGG
   pathway enrichment analysis and the corresponding scores. The
   enrichment score representing the enrichment degree among the
   differential pathways is shown in ‐log10. C) Heat map showing the
   expression levels of 23 phagosome relative genes in mouse liver after
   treatment with PEG‐ b ‐P(Gu). The control group consists of mice fed
   with HFD. D) CLSM image of the deacidification property of lysosomes in
   HepG2. The control group means palmitic acid loading of HepG2 cells;
   fluorescence intensities of left: lysosensor; right: PEG‐ b ‐P(Gu);
   scale bar: 10 µm. E) FCM analysis and F) quantification of the
   deacidification property of lysosomes in HepG2 after treatment with
   V‐ATPase inhibitor bafilomycin and PEG‐ b ‐P(Gu) (n = 3). The positive
   control group means palmitic acid loading of HepG2 cells. G) Assembly
   structure of PEG‐ b ‐P(Gu) after molecular dynamics simulation (red
   represents guanidine, gray represents the other groups). H) Molecular
   docking between PEG‐ b ‐P(Gu) and V‐ATPase (green stripes represent the
   guanidine groups and red dashed lines represent the interaction
   forces). I) Localization of PEG‐ b ‐P(Gu) and V[0] region in HepG2
   (red: PEG‐ b ‐P(Gu); blue: nucleus; green: V[0] region of V‐ATPase;
   yellow: merged; scale bar: 10 µm; and magnification: 5 µm). J) The
   vitality of V‐ATPase and K) proton transport activity of HepG2 after
   treatment with PEG‐ b ‐P(Gu) (n = 3). The control group means palmitic
   acid loading of HepG2 cells. ^*** P < 0.001; ^** P < 0.01; ^* P < 0.05;
   statistical analysis was performed using one‐way ANOVA.

2.8. PEG‐b‐P(Gu) Targets V[0] Region, Inhibits ATPV0A, and Triggers ATPV0D2
Upregulation

   Therefore, we docked the PEG‐ b ‐P(Gu) molecule with V‐ATPase to
   determine the approximate range of interaction sites. As shown in
   Figure [148]5G,H, guanidine groups are distributed throughout all the
   polymers and interact with the V[0] region of V‐ATPase (the V[0] region
   exists in the inner leaflets of the lysosome, while V[1] is located in
   the outer leaflets^[ [149]^29 ^]). Further fluorescence imaging
   (Figure [150]5I) also revealed an interaction between PEG‐ b ‐P(Gu) and
   the V[0] domain. Then, we used ATP6V0A1, ATP6V0C, and ATP6V0D2
   recombinant proteins to dock with PEG‐ b ‐P(Gu) through surface plasmon
   resonance (SPR). As anticipated by the molecular dynamics simulations,
   the SPR results (Figure [151]S22, Supporting Information) showed that
   although all three proteins could bind to PEG‐ b ‐P(Gu) to varying
   degrees, but in terms of dissociation constants (K[D]), ATP6V0A1 had a
   higher affinity than ATP6V0C, which in turn had a higher affinity than
   ATP6V0D2. This suggests that PEG‐ b ‐P(Gu) is likely to target the “a”
   subunit of V‐ATPase. In addition, the docking of PEG‐ b ‐P(Gu) with
   V‐ATPase also resulted in a decrease in V‐ATPase activity
   (Figure [152]5J) and H^+ pumping capability (Figure [153]5K), as the
   “a” subunit is the core component of V‐ATPase and directly participates
   in the conduction of H^+.^[ [154]^31 ^] Due to the observed weakening
   of the H^+ pumping ability of V‐ATPase, we speculate that PEG‐ b ‐P(Gu)
   may inhibit the H^+ conduction of the “a” subunit based on its
   function. After sensing this inhibition, the cell acts on the ATPV0D2
   subunit through feedback regulation to maintain the stability of the
   proton channel, which may be the reason why we observed high expression
   of ATPV0D2 at both the transcriptomic and protein levels.

2.9. PEG‐b‐P(Gu) Increases the Energy Expenditure by Activating AMPK
Phosphorylation

   Subsequently, we explored how the inhibition of V‐ATPase affects energy
   metabolism both in vivo and in vitro. An indirect calorimetric analysis
   of energy homoeostasis revealed that PEG‐ b ‐P(Gu) treatment increased
   energy expenditure without affecting locomotor activity or the
   respiratory exchange ratio in mice (Figure [155]6A–C). Furthermore, we
   measured the oxygen consumption rate in HepG2 cells and found that PEG‐
   b ‐P(Gu) significantly increased basal respiration, maximal
   respiration, and adenosine triphosphate (ATP) production compared to
   the control and metformin groups, suggesting that PEG‐ b ‐P(Gu)
   treatment increased energy metabolism and fatty acid oxidation, thereby
   contributing to reduced lipid accumulation in the liver.
   (Figure [156]6D; Figure [157]S23, Supporting Information). Based on
   these results, we focused on AMPK, as it is a key metabolic regulator
   that plays a pivotal role in maintaining cellular energy homeostasis.^[
   [158]^35 ^] Once activated (phosphorylation level increases), AMPK
   stimulates catabolic pathways that generate ATP while inhibiting
   anabolic pathways that consume ATP.^[ [159]^36 ^] In addition,
   activated AMPK inhibits the activity of acetyl‐CoA carboxylase (ACC) by
   increasing its phosphorylation.^[ [160]^10 ^] The suppression of the
   AMPK pathway can exacerbate fatty liver changes and lipid metabolic
   disorders, thereby playing a significant role in the development and
   progression of NAFLD.^[ [161]^37 ^] According to Figure [162]6E, the
   phosphorylation of AMPK or ACC increased in mouse livers following PEG‐
   b ‐P(Gu) treatment, suggesting that AMPK was activated, and ACC was
   inhibited. Consistently, the expression of de novo fatty acid synthesis
   markers in the liver was suppressed, including FASN and SCD1,^[
   [163]^38 ^] whereas lipolysis and fatty acid oxidation markers (ACSL,
   ATGL, HSL, and CPT1)^[ [164]^39 ^] were dramatically increased
   (Figure [165]6F). Notably, AMPK activation was also caused by the
   interaction of the inhibited V‐ATPase with LKB1/AXIN and AMPK in the
   mouse livers after PEG‐ b ‐P(Gu) treatment (Figure [166]6G,H).
   Similarly, after the treatment of V‐ATPase antagonists (concanamycin
   A),^[ [167]^40 ^] the interaction between these proteins ceased on
   HepG2 cells (Figure [168]6I). The AMPK inhibitors (compound C)^[
   [169]^41 ^] also inhibited the effects of PEG‐ b ‐P(Gu) on HepG2 cells
   (Figure [170]S24A–C, Supporting Information).

Figure 6.

   Figure 6
   [171]Open in a new tab

   PEG‐ b ‐P(Gu) increases the energy expenditure in vitro and in vivo by
   activating AMPK. A) Activity level, B) respiratory quotient, and C)
   energy consumption of C57BL/C mice after receiving 10 mg k^−1g PEG‐ b
   ‐P(Gu) i.v. twice a week since the beginning of HFD feeding. The mice
   were singly housed in metabolic cages for calorimetric analysis
   (n = 6). D) Basal respiration, maximal respiration, and ATP production
   were calculated through oxygen consumption rate (OCR) in HepG2 cells as
   measured using a Seahorse XF24 analyzer (n = 3). () Expression and
   phosphorylation levels of AMPK and ACC protein, and their quantitative
   data in mouse liver (n = 3). F) qPCR analysis of the gene expression of
   fatty acid synthesis, lipolysis, and β‐oxidation markers in mouse liver
   (n = 6). G) Schematic diagram showing that PEG‐ b ‐P(Gu) binds to
   V‐ATPase via interacting with ATPase V[0] region, thereby inhibiting
   the activity of V‐ATPase and then activating AMPK pathway. H)
   Interaction among LAMTOR1, AXIN, LKB1, and AMPK detected by CO‐IP in
   mouse liver and I) in HepG2 cells. ^*** P < 0.001; ^** P < 0.01; ^*
   P < 0.05; statistical analysis was performed using one‐way ANOVA.

3. Conclusion

   In the realm of innovative drug development, traditional drugs may face
   challenges amidst the rapid technological advancements and emergence of
   novel therapies. However, it is also essential to recognize that even
   compounds based on known mechanisms, through meticulous structural
   optimization and targeted design, may uncover novel biological effects
   or enhance efficacy, thereby finding new application spaces in
   therapeutic areas. In this study, through the synthesis and screening
   of a series of PGNIs, we identified that polyguanidine can greatly
   ameliorate lipid accumulation compared to small‐molecule guanidine. In
   addition to lipid lowering, the pharmacokinetics and interaction sites
   of guanidines are shifted because of nanoization resulting from
   polymerization. PGNIs are mainly located in hepatocytes and are taken
   up by lysosomes. Inside lysosomes, PGNI combined with the “a” subunit
   in V[0] residue of V‐ATPase inhibits its activity and feedback the
   increase of ATP6V0D2. Subsequently, V‐ATPase and Regulator (LAMTOR)
   allow LKB1 and AXIN to be recruited to activate AMPK. This process
   further decreases de novo fatty acid synthesis, promotes lipolysis and
   fatty acid oxidation, and ameliorates liver lipid accumulation.
   Although this mechanism similarities with metformin in certain aspects,
   we aspire to uncover its potential in reducing lipid accumulation
   through further investigation, aiming to provide more diverse and
   effective treatment options for NAFLD through a multi‐faceted approach.
   In summary, we discovered a novel pharmacological interaction mode
   between guanidine‐based compounds and potential target sites for fatty
   liver disease, which may be applicable to a wider range of
   high‐solubility and low‐permeability drugs, thereby yielding better
   efficacy in treating diseases.

Conflict of Interest

   The authors declare no conflict of interest.

Supporting information

   Supporting Information
   [172]ADVS-12-2408906-s001.docx^ (25.5MB, docx)

Acknowledgements