Fig 1.
Histological assessment of NASH and liver injury.
Histological sections were scored per established criteria (see Methods) in 19 of 20 HFWD mice and 20 of 20 HFWD-Ca mice. A. Individual components of NASH-related liver injury. B. NAS, calculated as the sum of the individual scores for the three component parameters. Inflammation and overall NAS were significantly lower in the HFWD-Ca group. (Mann-Whitney U, Steatosis, p = 0.1; Inflammation, p = 0.014; ballooning degeneration, p = 0.25; Overall NAS, p = 0.023). C-H: Representative histological features of NASH-related liver injury. Steatosis (C and D) consisted of macrovesicular cytoplasmic fat deposition distributed multifocally or diffusely. Multifocal lesions were most commonly in an azonal (random) or pericentral (zone 3) pattern. Inflammation (E and F) consisted of variably sized foci of infiltrating macrophages and neutrophils, most commonly in midzonal or pericentral (zone 3) areas (asterisk). Ballooning degeneration (G and H) consisted of markedly swollen hepatocytes with cleared cytoplasm and non-displaced nuclei (arrows). Hematoxylin and eosin staining. Scale bars = 200μm [C,D] and 100μm [E-H].
Fig 2.
Histological scoring of fibrosis and necrosis.
A. Fibrosis as indicated by Sirius red-staining for collagen was significantly lower in the HFWD-Ca group (asterisk, Mann-Whitney U, p = 0.0007). Necrosis assessed in H&E stained sections was also lower in the HFWD-Ca group, although this difference was not significant. B-E. Representative histological features. Fibrosis (B and C) varied from increased pericellular (“chicken wire” pattern) to pericentral (lobular) deposition in Sirius red stained sections. Necrosis (D) consisted of areas of devitalization (asterisk) that most commonly occurred within large regenerative hyperplastic nodules (RH) in HFWD mice. (E) depicts liver from a HFWD-Ca mouse with steatosis but no necrosis for comparison. Hematoxylin and eosin staining. Scale bar = 100μm.
Fig 3.
Regenerative hyperplasia (RH).
Representative RH lesions are shown grossly (A) and histologically (B and C). RH nodules were distinguished from neoplastic masses by preservation of hepatic structures (ie. portal triads, central veins, 1–2 cell thickness of hepatic cords) within the area of proliferation. D. Frequency of RH nodules. Six of 20 HFWD mice had liver masses histologically classified as RH compared to 1 of 20 HFWD-Ca mice (Chi-square, 2-tailed, p = 0.0375). (Scale bars = 1mm [B] and 100μm [C]).
Table 1.
Serum biomarkers in diet groups, excluding animals with liver massesa,b.
Table 2.
Serum cytokines in diet groups, excluding animals with liver massesa,b.
Fig 4.
Increased gut microbial diversity in HFWD-Ca group and differential segregation of gut microbial communities.
A. Increased microbial diversity in cecal (top) and fecal (bottom) microbial communities of HFWD-Ca group at study endpoint (18 months). B. PCoA depicting differential microbial segregation based on Illumina sequencing of the V4 region of 16S rRNA gene in cecal (left) and fecal (right) samples. Segregation between HFWD and HFWD-Ca groups was significant based on θYC distances (AMOVA, p<0.0001).
Fig 5.
Microbial community composition within dietary groups.
Cecal bacterial community compositions of individual samples. Each row represents the cecal bacterial community of one sample based on relative abundances of OTUs (97% sequence similarity) identified from V4 region sequences of 16S rRNA genes. Each column represents one OTU with darker shades of purple indicating higher relative abundances. OTUs were organized by taxonomic classification and only the OTUs that comprised at least 2% of one sample were included.
Fig 6.
Relative abundance of selected taxa.
Significant differences in relative abundance of major OTUs driving separation between HFWD-Ca and HFWD groups for cecal (A) and fecal (B) samples. Asterisks represent significance determined by Mann-Whitney U (p values above individual panels).
Table 3.
Significant OTUs with ten highest LDAa scores per diet group.
Fig 7.
Hepatobiliary bile acid profiles.
A. Total bile acids. Total bile acids and total conjugated bile acids were lower (but did not reach significance) in livers of mice from the HFWD-Ca group. B. Muricholic acid (MCA) and derivatives. TbMCA was significantly reduced in livers of HFWD-Ca mice (asterisk, p<0.00001, unpaired t test, Holm-Sidak correction for multiple comparisons). C. Low-prevalence MCA derivatives were also lower in the HFWD-Ca group, although this did not reach significance. D. Cholic acid (CA), chenodeoxycholic acid (CDCA) and derivatives. No differences between groups were seen (note y axis values indicating lower overall prevalence of CA/CDCA derivatives in mice in comparison to MCA). E. Low-prevalence CA derivatives. No differences between groups were seen. Data shown for all panels is from randomly selected mice without tumors (n = 7 for HFWD, n = 5 for HFWD-Ca).
Fig 8.
Proposed downstream site of calcium protection from liver injury.
In the current study, calcium supplementation did not affect weight gain or hepatic steatosis, but did decrease the incidence of NASH-related histological changes, including fibrosis and regenerative hyperplasia associated with “end-stage” NASH. Thus, calcium supplementation could prevent the adverse downstream effects of obesity in situations where weight loss cannot be achieved or maintained. Although the mechanism is not known, potential factors could include gut chelation of bile acids or altered bile acid pools that may be related to gut microbial shifts. Calcium supplementation in this study was associated with decreased hepatic tauro-β- muricholic acid, which reduces FXR antagonism in the liver, permitting FXR-mediated inhibition of total bile acid synthesis. The decrease in hepatic total bile acid synthesis may decrease available substrate for downstream intestinal bacterial conversion of primary bile acids to cytotoxic secondary bile acids (LCA, DCA). Since a percentage of these bile acids are re-absorbed in the colon and transported to liver, the decrease in their synthesis could limit hepatic exposure to potentially cytotoxic BA molecular species.