Intestinal villus length and crypt base columnar cells (CBCs) were reduced in aging.
To explore the morphology of intestinal epithelium in three ages mice, we used hematoxylin-eosin staining to count the villi height and crypt depths. Villi height was no difference between young and middle groups, but decreased in old group compared with young and middle groups. Crypt depth was increased in middle group compared with young group, but decreased in old group compared with middle group (Fig. 1a). To count the major kinds of epithelium, we used Alcian blue and Periodic Acid-Schiff staining, MMP-7 immunohistochemistry staining and Olfm4 immunofluorescence to mark goblet cells, Paneth cells and ISCs respectively. Goblet cells number in crypt-villi axis were no difference among three groups (Fig. 1b). Paneth cells number was increased in middle group compared with young group, and was no difference between middle group and old group (Fig. 1c). Olfm4+ crypt base columnar cells (CBCs) were increased in the middle group compared with young group and then decreased in old group (Fig. 1d). Crypt depth and number of Olfm4+CBCs showed similar tendency in aging. These results indicated that the number of ISCs and proliferation might be increased in middle age and decreased in old.
Lgr5 + intestinal stem cells (ISCs) were reduced in aging mice and intestinal organoids.
To verify the ISCs number in different ages, we used Lgr5-EGFP knock-in transgenic mice to mark the Lgr5+ISCs. The number of Lgr5+ISCs in each crypt increased in middle group and then decrease in old group (Fig. 2a), which is consistent with the number of Olfm4+CBCs (Fig. 1d). Next, we isolated the crypts from Lgr5-EGFP mice in different ages and cultured organoids for 7 days to explore whether the ISCs from different ages mice were distinguish in culture condition. In the view of single organoids, the buddings number and projected area were decreased both in middle group and old group compared with young group and were no difference between middle group and old group (Fig. 2b). In the view of holes, the Lgr5-GFP area tended to decrease in old group but had no statistical difference (P=0.067 among three groups) (Fig. 2c). In the view of sections, the ratio of Lgr5+ISCs number to DAPI number decreased in old group compared with young group (Fig. 2d). These results above suggested that the number of Lgr5+ISCs decreased in old mice both in vivo and in vitro.
Intestinal cellular proliferation was downregulation in aging mice and intestinal organoids.
To further verify the proliferation in aging, we used PCNA and Ki67 to mark the proliferating cells. The results showed that proliferating cells in crypts increased in middle group compared with young group, and then decreased in old group compared with middle group. The young group and old group were no difference (Fig. 3a and 3b). To explore the movement of epithelium, we intraperitoneally injected BrdU (5-bromo-2’-deoxyuridine) into mice and collected the samples after 72 hours. After BrdU immunofluorescence staining, we calculated the distance from the middle point of BrdU+ cells to the bottom of villi. The result showed that the distance decreased gradually from young group to old group (Fig. 3c). To verify the proliferation in organoids, we used organoids section for Ki67 immunofluorescence. The ratio of Ki67+ cells decreased in old group compared with young group (Fig. 4a). These results suggested that the proliferation decreased in old groups. In middle group, the proliferation increased only in tissue.
Proliferation of Lgr5 + ISCs was restrained in aging mice.
According to the results above, we found that the number of Lgr5+ ISCs and its related proliferation function changed in aging. Next, we intended to explore which status in Lgr5+ ISCs caused such results, proliferation, apoptosis or cells senescence. We used Lgr5-EGFP mice to mark the Lgr5+ ISCs. Meanwhile, we used Ki67 immunofluorescence to mark the proliferating Lgr5+ ISCs and TUNEL to mark the apoptosis Lgr5+ ISCs. The proportion of proliferating Lgr5+ ISCs decreased in middle and old group compared with young group. The results of middle group and old group were no difference (Fig. 4a and 4d). Among all the three group, none of Lgr5+ ISCs were TUNEL positive. Although some TUNEL+ fragments were observed on the crypt basement in old group, these fragments did not overlap with nuclear or Lgr5-GFP signals (Fig. 4b).
Lastly, we used a canonical cell senescent marker-β-galactosidase staining on isolated crypts and villi, 24-hours cultured organoids and 7-days cultured organoids to detect the cell senescence. The result showed that all of the villi and the top of the crypts were blue, however, the bottom of the crypts were unstaining in all three groups, which suggested the majority of intestinal epithelium were already senescent at three months while ISCs were not. In addition, organoids from different groups were almost staining blue, while some newborn buddings were unmarked (Fig. 4c and 4d).
All the results showed that the proliferation Lgr5+ISCs function degenerated obviously.
Parp3 gene was upregulated in intestinal Lgr5+ stem cells of aging mice.
Since we found that the proliferation of Lgr5+ ISCs degenerating in aging impacted the renovation of intestinal epithelium, to understand the molecular mechanism of Lgr5+ ISCs in different age, we isolated the crypts, separated them into single cell and sorted the Lgr5high cells by FACS for bulk RNA sequence (Fig. 5a). The flow chart showed that the percentage of Lgr5high cells increased in middle group and decreased in old group (Fig. 5B), which was consistent with former results (Fig. 2a). The whole heat map showed that 9 samples were divided into three clusters according to gene expression, which were consistent of the age groups (Fig. 5c). The majority of genes expression were similar between middle and old group, with only 55 genes up-regulated and 85 genes down-regulated in old group. And 211 genes upregulated and 93 genes downregulated in middle group compared with young group. Furthermore, we used the pathway enrichment to analysis the signal transduction among the three groups. For example, there were statistical differences in immune-related signaling pathways between the young and middle group such as antigen processing and presentation pathway. (Fig. 5d) Between middle group and old group, pathways such as DNA methylation and Meiotic Recombination had statistically significant difference (Fig. 5e). We noticed that the Parp3 had significantly statistical difference in the young group compared with both middle group and old group (Fig. 5f). Parp3 was also into the KEGG pathway “cell growth death”. We hypothesized that this gene might relate to the alternation of Lgr5+ ISCs proliferation in aging.
Parp3 was overexpressed in intestinal mucosa and intestinal organoids of aging mice.
To further verify the Parp3 expression in different ages, we used RT-PCR, western-blotting and Parp3 immunofluorescence to determine the gene expression and protein expression of Parp3. Parp3 gene expression increased in old group compared with young group and showed a tendency of increase in middle group compare with young group but had no statistical difference (Fig. 6a). Parp3 protein expression increased in middle group and old group compared with young group. And the protein expression was no difference between middle group and old group (Fig. 6b). The results of Parp3 immunofluorescence on tissue and organoids also showed Parp3 protein level increased in middle and old group (Fig. 6c and 6d). These results indicated that Parp3 gene expression and protein level increased in aging.
Inhibition of Parp3 depressed intestinal organoids growth in aging.
Since we found that the Parp3 increased from the middle age and keeping high level in the old, next, we explored the effect of Parp3 on epithelium. We cultured the organoids from 12 months Lgr5-EGFP mice. We used two kinds of Parp3 non-selective inhibitors, AZD-2461 and ME0328, with a concentration of 2µmol/L and 10µmol/L respectively. The inhibitors were added into the culture medium in day 4, and the culture medium was changed every day until day 7. We added the equally volume of DMSO in culture fluids as control groups. We traced the organoids in these groups in day 4 and compared the organoids shape every day after adding the inhibitors. The organoids in both of AZD-2461 groups and ME0328 group grew less volume and buddings compared with control group (Fig. 7a-c). To observe the expression of Parp3, we used Parp3 immunofluorescence to detect the Parp3 protein in organoids after 7-days cultured. The results showed that Parp3 had a lower lever in the two groups adding inhibitors compared with control group (Fig. 7d). All the results showed that inhibition of Parp3 depressed the middle age organoids growth.