Liver-bone organoids reveal senescence-driven interorgan crosstalk

Abstract

The liver-bone axis plays a critical role in age-related diseases. However, current models inadequately capture its complex inter-organ communication. Here, we established novel, physiologically relevant senescent liver and bone organoid models using engineered bionic hydrogels combined with doxorubicin (DOX)-induced senescence. These models successfully recapitulated hallmark aging characteristics: bone organoids exhibited reduced mineralization accompanied by elevated senescence markers, and liver organoids demonstrated DNA damage along with structural deterioration. Notably, aged mouse serum effectively induced senescence in both organoid types, confirming the existence of systemic aging regulators. The platform demonstrated robust bidirectional crosstalk, with senescent liver organoid-conditioned medium potently driving degradation in bone organoids and senescent bone organoid-conditioned medium aggravating dysfunction in liver organoids. Mechanistically, we identified 27-hydroxycholesterol (27-OHC) as a novel hepatocyte-derived factor mediating liver-to-bone communication. 27-OHC not only induced bone organoids senescence but also synergized with DOX treatment to exacerbate bone loss, a finding corroborated by in vivo mouse studies that validated the relevance of our platform in the context of pathological damage. This study pioneers the first organoid-based platform that elucidates multi-organ aging mechanisms, uncovering 27-OHC as a pivotal regulator of liver-bone axis dysfunction and proposing novel treatment strategies for age-related systemic disorders.

Keywords: 27-OHC; Bidirectional crosstalk; Bone organoids; Liver organoids; Senescence.

Graphical abstract

Fig. 1

The schematic diagram illustrates the construction of senescent liver and bone organoids, and the screening of 27-hydroxycholesterol as a key mediator of senescence propagation through this senescent liver-bone organoid model.

Fig. 2

Senescent bone organoids construction by replication. IHC staining of SA-β-gal in bone tissues (A) and primary BMSCs (B) of 4-month-old and 14-month-old mice. Scale bars, 200 μm and 100 μm. (C) Diagram of senescent bone organoids construction via 3D bioprinting. (D) Proliferation of P3 and P23 BMSCs. (E) IF staining of SA-β-gal in P3 and P23 BMSCs using a fluorescent probe (red). Scale bars, 100 μm. (F) IF staining of p21 (green), SA-β-gal (red), γ-H2A.X (purple) and DAPI (blue) in bone organoids constructed by P3 and P23 BMSCs. Scale bars, 200 μm. (G) Mineralization of bone organoids assessed by micro-CT scanning. (H–J) Changes and quantitative analysis of BMD (H), BV/TV (I) and Tb.Th (J). Data are presented as the mean ± SD. Data are considered statistically significant at ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

Fig. 3

Senescent bone organoids construction by inducer. (A) IHC staining of SA-β-gal in primary BMSCs treated with 0.5 μM DOX for 48 h. Scale bars, 50 μm. (B) Diagram of DOX-induced senescent bone organoids construction. (C–F) Represented images of 3D mineralization of bone organoids treated with 0.5 μM DOX for 3 and 6days (C) and quantification of BMD (D), BV/TV (E) and Tb.Th (F). (G–H) IF staining of p21 (green), SA-β-gal (red), γ-H2A.X (purple) and DAPI (blue) in bone organoids treated with 0.5 μM DOX (G), D-gal and H₂O₂ (H). Scale bars, 200 μm. (I–L) Represented images of 3D mineralization of bone organoids treated with D-gal and H₂O₂ for 6days (I) and quantification of BMD (J), BV/TV (K) and Tb.Th (L). Data are presented as the mean ± SD. Data are considered statistically significant at ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

Fig. 4

Construction of senescent liver organoids. (A) IF staining of SA-β-gal in liver tissues of 4-month-old and 14-month-old mice. Scale bars, 200 μm. (B) Diagram of senescent liver organoids construction from aged mice. (C) IF staining of SA-β-gal in liver organoids constructed from the duct of 4-month-old and 14-month-old mice. Scale bars, 100 μm. (D) SA-β-gal detection in AML12 cells treated with 0.5 μM DOX for 24 h. Scale bars, 50 μm. (E) The protein level of p53, p21 and p16. (F) Relative mRNA levels of p21. (G) Diagram of DOX-induced senescent liver organoids construction. (H) IF staining of p21 (green), SA-β-gal (red), γ-H2A.X (purple) and DAPI (blue) in liver organoids treated with 0.5 μM DOX for 3 and 6 days. Scale bars, 100 μm. (I) Lipid accumulation of liver organoids. Scale bars, 100 μm. Data are presented as the mean ± SD. Data are considered statistically significant at ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

Fig. 5

Senescence propagation between liver and bone organoids. (A–B) IF staining of senescence markers in liver organoids (SA-β-gal) (A) and bone organoids (p21, SA-β-gal and γ-H2A.X) (B) treated with 10 % serum (excluded FBS) from 4-month-old and 14-month-old mice. Scale bars, 100 μm and 200 μm. (C–F) Represented images of 3D mineralization (C) and quantification of BMD (D), BV/TV (E) and Tb.Th (F). (G) Diagram of the senescent conditioned medium exchange between liver and bone organoids. (H) IF staining of SA-β-gal in liver organoids treated with conditioned medium from senescent bone organoids. Scale bars, 100 μm. (I) IF staining of p21 (green), SA-β-gal (red), γ-H2A.X (purple) and DAPI (blue) in bone organoids treated with conditioned medium from senescent liver organoids. Scale bars, 200 μm. (J–M) Represented images of 3D mineralization (J) and quantification of BMD (K), BV/TV (L) and Tb.N (M). Data are presented as the mean ± SD. Data are considered statistically significant at ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

Fig. 6

Comparative Mass Spectrometry proteomics analysis. (A) Principal component analysis (PCA) of liver proteomics of aged mice. (B) Volcano plot of protein abundance from liver proteomic analysis. (C–D) GO enrichment analysis of biological processes (BP) (C) and KEGG pathways (D) in DEPs. (E) GSEA analysis of DEPs in aged mice enriched steroid hormone. (F) Volcano plot of protein abundance from aged-mice serum proteomic analysis. (G) GO enrichment analysis of BP in DEPs. (H) Venn diagram of common altered factors from proteomic analysis between aged-mice serum and conditioned medium from senescent liver organoids. (I) Heatmap of the common altered proteins both in serum and conditioned medium.

Fig. 7

27-hydroxycholesterol drives senescence propagation to bone organoids. (A) IF staining of senescence markers (p21, SA-β-gal and γ-H2A.X) in bone organoids treated with 2.5 μM 27- OHC for 6 days. Scale bars, 200 μm. (B–E) Represented images of 3D mineralization of bone organoids treated with DOX and 27-OHC (B) and quantitative analysis of BMD (C), BV/TV (D) and Tb.Th (E). (F) H&E, Masson's Trichrome and ARS staining of bone organoids treated with DOX and 27-OHC. Scale bars, 200 μm. 10-week-old male mice were administered the 27-OHC via intraperitoneal injection daily for 20 days after HDT PT3-NRAS and AKT along with transposase (SB100) (n = 5 per group). (G–K) Represented images of 3D scanning (G) and quantification of BMD (H), BV/TV (I), Tb.N (J), and Tb.Sp (K). Data are presented as the mean ± SD. Data are considered statistically significant at ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

figs1