Isolation and identification of bone marrow-derived mesenchymal stem cells from mouse bone marrow

MSCs were isolated from the femurs and tibias of 6-week-old C57 mice.19 After 4–6 passages, MSCs demonstrated a fibroblast-like or spindle-like shape in culture (Fig. 1a). Then, we identified the surface markers of MSCs with or without ammonia treatment via flow cytometry analysis,20 including the positive markers of MSCs (CD44, Sca-1, CD29, CD105) and the negative markers of MSCs (CD31, CD86, CD11b, CD34, CD45) (Fig. 1b). To evaluate the multidifferentiation capacity of MSCs, we used a differentiation assays in vitro by differentiating these cells into different lineages, including osteoblasts, adipocytes and chondrocytes. Then, we verified osteogenic induction with positive alkaline phosphatase staining, adipogenic induction with positive oil red O staining and chondrogenic induction with positive toluidine blue O staining (Fig. 1c). To confirm that ammonia could not affect the characteristics of mesenchymal stem cells, we also detected MSC-related genes, including Bglap, Runx2, Alpl, Acan and Pparg.16 After treatment with or without ammonia, we found no significant difference between the two groups (Fig. 1d), demonstrating that ammonia could not change the characteristics of mesenchymal stem cells.

Fig. 1

Isolation and identification of mesenchymal stem cells from mouse bone marrow. a Fibroblast-like or spindle-shaped morphology of mesenchymal stem cells appears at passages 4–6. b Passage 4–6 cells were harvested and stained with positive surface markers of mouse BM-MSCs: CD29, CD44, CD105, and Sca-1 and negative markers: CD31, CD34, CD45, CD86, and CD11b with their corresponding isotype control via FACS analysis. c Adipogenesis of MSCs was observed with oil red O staining, osteoblastogenesis was assayed with in situ alkaline phosphatase staining, and chondrocytic cells were identified with toluidine blue staining. d RT-PCR results showing the relative expression of bglap, runx2, alpl, acan and pparg with or without NH4Cl treatment

Ammonia promotes bone marrow-derived mesenchymal stem cell proliferation

To assess the effect of ammonia on MSC proliferation, we added NH4Cl in culture with different concentrations of ammonia (0 µmol·L−1, 40 µmol·L−1, 80 µmol·L−1, 160 µmol·L−1, 320 µmol·L−1, 640 µmol·L−1, 1.25 mmol·L−1, 2.5 mmol·L−1 and 5 mmol·L−1), which included the physiological ammonia concentration and an ammonia concentration as high as that in tumors.21 Then, we observed the proliferation of mesenchymal stem cells within a week after NH4Cl treatment via RTCA (Real Time Cellular Analysis). The results revealed that the NH4Cl treatment group, especially the 5 mmol·L−1 group, showed a higher cell index than the other groups. After a 96 h NH4Cl treatment, there was a downward trend in all growth curves (Fig. 2a). Moreover, there were more MSCs in the NH4Cl treatment group than in the non-NH4Cl treatment group under the microscope (Fig. 2b), especially in the 5 mmol·L−1 ammonia treatment group. When the ammonia concentration exceeded 5 mmol·L−1, the effect of ammonia-induced proliferation of MSCs was reduced. These data demonstrated that 5 mmol·L−1 NH4Cl is the optimal concentration for the proliferation of MSCs. The cell number count of MSCs showed the same result (Fig. 2c). We measured cell proliferation by a 7-day CCK-8 assay, which showed that the ammonia treatment group increased cell proliferation compared to the control group; moreover, the most significant effect on proliferation was on the 2nd day (Fig. 2d). Then, we detected the cell cycle distribution of MSCs with or without 48 h ammonia treatment via PI flow cytometry analysis.22 The results demonstrated that the percentage of S-phase cells increased in the NH4Cl treatment group compared with the control group (Fig. 2e, f). We also assessed cell proliferation by BrdU incorporation with or without 48 h of treatment with NH4Cl23 and then determined the percentage of BrdU-positive cells by flow cytometry analysis and immunofluorescence analysis. The results showed that there was a greater percentage of BrdU-positive cells in the NH4Cl treatment group than in the control group both in the flow cytometry analysis (Fig. 2g, h) and immunofluorescence analysis (Fig. 2i, j). These results indicated that ammonia at a certain concentration could promote mesenchymal stem cell proliferation.

Fig. 2

Ammonia promotes mesenchymal stem cell proliferation. a RTCA (real-time cellular analysis) of different concentrations of NH4Cl (0, 40 µmol·L−1, 80 µmol·L−1, 160 µmol·L−1, 320 µmol·L−1, 640 µmol·L−1, 1.25 mmol·L−1, 2.50 mmol·L−1, 5.00 mmol·L−1, and 10 mmol·L−1). b The growth of MSCs with or without NH4Cl treatment under a microscope after 48 h. c Cell counts after treatment with different concentrations of NH4Cl (0, 40 µmol·L−1, 80 µmol·L−1, 160 µmol·L−1, 320 µmol·L−1, 640 µmol·L−1, 1.25 mmol·L−1, 2.50 mmol·L−1, 5.00 mmol·L−1, and 10 mmol·L−1) after 48 h. d CCK-8 analysis with different concentrations of NH4Cl treatment (0, 40 µmol·L−1, 80 µmol·L−1, 160 µmol·L−1, 320 µmol·L−1, 640 µmol·L−1, 1.25 mmol·L−1, 2.50 mmol·L−1, 5.00 mmol·L−1, and 10 mmol·L−1) lasting 1 week (right). e PI flow cytometry analysis. f The statistical results of the percentage of S phase cells. f BrdU flow cytometry analysis. g The statistical results of the percentage of BrdU+ cells. h Immunofluorescence analysis of BrdU. j The statistical results of BrdU-positive cells/HPF. Values are the mean ± SEM of an experiment performed in triplicate (a two-tailed, paired Student’s t test). *P < 0.05 versus the controls. #P < 0.05 versus the 5 mmol·L−1 NH4Cl treatment group

Glutamine promotes bone marrow-derived mesenchymal stem cell proliferation

To investigate the mechanisms of the proliferation of MSCs by ammonia, as a nitrogen supplement of amino acids, we considered that ammonia might be converted with glutamate into glutamine via GS (Fig. 3a). Previously, glutamine was shown to play a prominent role in cell growth and proliferation.24 Additionally, more glutamine production was observed in the groups with NH4Cl treatment than in the control group (Fig. 3b), and after treatment with MSO, a specific inhibitor of glutamine synthetase,18 we found that combined treatment of NH4Cl and MSO resulted in decreased production of glutamine compared to single treatment of NH4Cl (Fig. 3b). We verified the effect of glutamine on MSC proliferation with 5 mmol·L−1 glutamine treatment within a week in vitro via RTCA (Real Time Cellular Analysis). The growth curve of MSCs showed a higher cell index in the glutamine-treated group than in the control group (Fig. 3c). Additionally, there were more MSCs in the glutamine treatment group than in the control group under the microscope (Fig. 3d), and the results of the cell counting analysis demonstrated that the number of glutamine-treated cells was obviously higher than that in the DMEM group (Fig. 3e). The results revealed increased cell proliferation in the glutamine treatment group compared to the control group by 7-day CCK-8 analysis, and the most significant effect on proliferation was observed on the 2nd day (Fig. 3f). Furthermore, the percentage of S phase cells increased in the glutamine treatment group compared to the control group, as shown by PI flow cytometry analysis (Fig. 3g, h), and the percentage of BrdU-positive cells was detected via both immunofluorescence analysis and flow cytometry analysis. The results demonstrated that an increase in BrdU-positive cells was observed in the glutamine treatment group (Fig. 3i–l). These results not only indicated that glutamine could promote MSC proliferation but also indicated that the effects of ammonia on MSC proliferation might be due to the conversion of ammonia to glutamine.

Fig. 3

Glutamine promotes mesenchymal stem cell proliferation. a Metabolic pathway of ammonia in GS-expressing MSCs. b RTCA (real-time cellular analysis) with or without glutamine treatment. c Determination of glutamine after NH4Cl treatment (left) and with or without both NH4Cl treatment and MSO treatment (right). d The growth of MSCs with or without glutamine treatment under a microscope. e The results of cell count analysis with or without glutamine treatment. f CCK-8 analysis with or without glutamine treatment lasting 1 week. g PI flow analysis. h The statistical results of the percentage of S phase cells. i BrdU flow analysis. j The statistical results of the percentage of BrdU+ cells. k Immunofluorescence analysis of BrdU. l The statistical results of BrdU+ cells/HPF. Values are the mean ± SEM of an experiment performed in triplicate (one-way analysis of variance). *P < 0.05 versus the controls. #P < 0.05 versus the 5 mmol·L−1 NH4Cl treatment group

Inhibition of glutamine synthetase reduces the ammonia-induced proliferation of bone marrow-derived mesenchymal stem cells

To confirm our hypothesis that ammonia could be converted into glutamine to promote MSC proliferation, we used MSO, a specific inhibitor of GS, to determine whether the proliferative effect of ammonia on MSCs could be reduced after inhibition of GS activity. To verify the effect of MSO on MSC proliferation induced by ammonia, we performed RTCA (Real Time Cellular Analysis) within a week. The results demonstrated that after MSO treatment, the combined treatment of 5 mmol·L−1 NH4Cl and 5 mmol·L−1 MSO decreased the proliferative effect compared to the single treatment of 5 mmol·L−1 NH4Cl (Fig. 4a). More MSCs were observed in the NH4Cl treatment group than in the combined treatment group under the microscope (Fig. 4b), and the cell number was decreased in the NH4Cl combined with 5 mmol·L−1 MSO treatment group compared to the NH4Cl treatment group via cell counting analysis (Fig. 4c). The results of the 7-week CCK-8 assays showed that the NH4Cl combined with MSO group had decreased cell proliferation compared to the single NH4Cl treatment group, and significant effects were observed on the 2nd day (Fig. 4d). Additionally, the percentage of S phase cells decreased when MSO was added, as shown by PI flow cytometry analysis (Fig. 4e, f). Subsequently, we detected the positive incorporation of BrdU following NH4Cl and MSO treatment in MSCs. Both the flow cytometry analysis and immunofluorescence analysis showed increased BrdU-positive cells in the single NH4Cl treatment group, while they decreased in the combined NH4Cl and MSO treatment group (Fig. 4g–j). These results further indicated that ammonia promotes MSC proliferation by conversion to glutamine via glutamine synthetase activity; furthermore, proliferation could be reduced by inhibiting glutamine synthetase.

Fig. 4

Inhibition of glutamine synthetase reduces the ammonia-induced proliferation of mesenchymal stem cells. a RTCA (real-time cellular analysis) with or without both NH4Cl treatment or MSO treatment. b The growth of MSCs with or without both NH4Cl treatment and MSO treatment under a microscope. c The cell count results with or without both NH4Cl treatment and MSO treatment. d CCK-8 analysis with or without NH4Cl treatment or MSO treatment after 48 h (left) and lasting 1 week (right). e PI flow analysis. f The statistical results of the percentage of S phase cells. g BrdU flow analysis. h The statistical results of the percentage of BrdU+ cells. i Immunofluorescence analysis of BrdU. j The statistical results of BrdU+ cells/HPF. Values are the mean ± SEM of an experiment performed in triplicate (one-way analysis of variance). *P < 0.05 versus the controls. #P < 0.05 versus the 5 mmol·L−1 NH4Cl treatment group

The different effects of ammonia on GS-expressing cells and cells with no expression of GS in bone marrow

To determine the fate of ammonia in bone marrow, we further assessed the mechanism by which ammonia could promote MSC proliferation caused by the high expression of GS (Fig. 5b). CD45+ cells, such as neutrophils, DCs, monocytes and macrophages, without GS expression showed another fate: ammonia inhibited the growth of these cells as a toxic product. We sorted CD45+ cells from mouse tibias and femurs (Fig. 5c), and the expression of GS was reduced compared with that in MSCs (Fig. 5d, e). Furthermore, we sorted neutrophils, DCs, monocytes and macrophages via specific surface markers (Fig. 5f). The results showed that these cells notably reduced the expression of GS compared with MSCs (Fig. 5g, h). In summary, ammonia could promote proliferation in cells with GS expression, such as MSCs, but inhibited cells without GS expression, such as CD45+ cells in bone marrow (Fig. 5a).

Fig. 5

The different effects of ammonia on GS-expressing cells and cells with no expression of GS in bone marrow. a Schematic of the fates of ammonia in the GS-expressing cells or non-GS-expressing cells in bone marrow. b The expression of GS in MSCs with or without both NH4Cl treatment and MSO treatment via western blots. c CD45+ cells from mouse bone marrow sorted by flow cytometry. d The expression of GS analysis between MSCs and CD45+ cells via western blots, with HeLa cells as a positive control. e The expression of GS in MSCs and CD45+ cells via immunofluorescence, with HeLa cells as a positive control. f CD45+ cells, including neutrophils (CD45+CD11b+Ly6G+), DCs (CD45+CD11b+CD11c+), monocytes (CD45+CD11b+Ly6C+) and macrophages (CD45+CD11b+F4/80+), were sorted by flow cytometry. g The expression of GS in MSCs, neutrophils, DCs, monocytes and macrophages via western blots, with HeLa cells as a positive control. h The expression of GS in MSCs, neutrophils, DCs, monocytes and macrophages via immunofluorescence. HeLa cells were used as a positive control

mTOR activation is involved in ammonia-induced mesenchymal stem cell proliferation

We explored the mechanism of MSC proliferation induced by ammonia. mTOR plays a fundamental role in supporting cell proliferation.12 As we mentioned previously, glutamine can activate the Akt/mTOR/S6K pathway by increasing phosphorylation of the mTOR pathway (Fig. 6a).25 Then, we investigated whether ammonia could activate the Akt/mTOR/S6K pathway by detecting phosphorylation of mTOR (Ser2448) and, upstream and downstream of mTOR, phosphorylation of Akt (Ser473) and S6K (Thr308) via western blot analysis.26 Our results showed enhanced phosphorylation of mTOR, Akt and S6K in the 5 mmol·L−1 NH4Cl treatment group, as well as the 5 mmol·L−1 glutamine treatment group, compared to the control group, and the 10 mmol·L−1 NH4Cl treatment group showed decreased phosphorylation of mTOR, Akt and S6K compared to the 5 mmol·L−1 NH4Cl treatment group, which is consistent with the proliferative effect of ammonia mentioned above. The expression of total mTOR, total Akt and total S6K showed no significant differences among the groups (Fig. 6b). We next assessed whether combined treatment with NH4Cl and MSO reduced the phosphorylation of mTOR (Ser2448), phosphorylation of Akt (Ser473) and phosphorylation of S6K (Thr308) compared to those of the NH4Cl treatment group. In addition, the expression of total mTOR, total Akt and total S6K showed no significant differences among the groups (Fig. 6c). These results revealed that ammonia could activate the Akt/mTOR/S6K pathway to promote MSC proliferation, and the activation of the Akt/mTOR/S6K pathway could be reduced by inhibiting the key enzyme for ammonia conversion to glutamine, glutamine synthetase, to decrease the effect of ammonia-induced proliferation on MSCs.

Fig. 6

mTOR activation and cell cycle regulation are involved in ammonia-induced mesenchymal stem cell proliferation. a Schematic of the metabolic pathway of ammonia in GS-expressing MSCs. b Western blot analysis of the AKT/mTOR/S6K pathway with or without NH4Cl or glutamine treatment. c Western blot analysis of the AKT/mTOR/S6K pathway with or without both NH4Cl treatment or MSO treatment. d Western blot analyses of cyclin D/CDK4 and cyclin E/CDK2 with or without NH4Cl or glutamine treatment. e Western blot analysis of cyclin D/CDK4 and cyclin E/CDK2 with or without both NH4Cl treatment or MSO treatment

Cell cycle regulatory proteins and DNA synthesis modulated by mTOR activation

We further investigated how ammonia promotes MSC proliferation regulated by the Akt/mTOR/S6K pathway. As mentioned above, we found an obvious increase in the percentage of S-phase cells in the NH4Cl treatment group (Fig. 2e) as well as in the glutamine treatment group (Fig. 3g), and this increase in S phase could be reduced by inhibition of glutamine synthetase (Fig. 4e). We further detected mTOR pathway-related cell cycle proteins and cyclin-dependent kinases, including cyclin D, CDK4, cyclin E and CDK2, which could induce cells from G1 phase to S phase to support MSC proliferation (Fig. 5a). The western blot results showed elevated expression of cyclin D, CDK4, cyclin E and CDK2 in the 5 mmol·L−1 NH4Cl treatment group (Fig. 6d). Elevated levels were also observed in the glutamine treatment group (Fig. 6d), and as expected, the expression of cell cycle proteins and cyclin-dependent kinases was reduced by inhibiting glutamine synthetase with combined treatment of NH4Cl and MSO compared to single treatment with NH4Cl (Fig. 6e). These results further revealed that activation of the Akt/mTOR/S6K pathway could regulate the distribution of the cell cycle to promote MSCs from G1 phase to S phase via activation of the cell cycle proteins and cyclin-dependent kinases related to proliferation.27 MSCs in S phase showed an increase in DNA synthesis for proliferation.

The proportion of bone marrow-derived MSCs was elevated in ammonia-loaded mice

To investigate the growth conditions of bone marrow-derived MSCs with hyperammonemia in vivo, we established a chronic ammonia-loaded model. The ammonia level of the ammonia-loaded mice was significantly higher than that of the control group (Fig. 7a). To explore whether a high ammonia concentration in vivo could also promote MSC proliferation and whether proliferation could be inhibited by injection of MSO in vivo, we collected the tibias and femurs of the mice and digested them with collagenase, and then, the collected cells were analyzed by FACS. The percentage of ammonia-loaded mouse MSCs was elevated compared with that in the control group, and that in the ammonia-loaded mice with MSO injection was reduced (Fig. 7b, c). We further measured the proportion of CD45+ cells, which indicates the non-GS-expressing cells or MSCs from bone marrow. The ammonia-loaded mice showed an increase in MSCs and a decrease in CD45+ cells. (Fig. 7d–f). HE staining of the spinal column showed more fibroblasts and fewer hematopoietic cells in the ammonia-loaded mice than in the control mice (Fig. 7g). These results indicated that ammonia could not only promote MSC proliferation in vitro but also elevate the proportion of MSCs in vivo.

Fig. 7

The proportion of bone marrow-derived MSCs was elevated in ammonia-loaded mice. a The ammonia concentration (µmol·L−1) of the wild-type mouse group, NH4Cl-loaded mouse group and NH4Cl-loaded mouse groups with MSO injection. b The percentage of MSCs in CD45- cells from mouse tibias and femurs in each group. c The statistical results of the percentage of MSCs in BM. d The percentage of the expression of MSC surface markers (CD29, CD44, CD105, Sca-1) and the expression of CD45 in each group. e The statistical results of the percentage of MSC surface marker expression in each group. f The statistical results of the percentage of CD45 expression in each group. g HE staining of the spinal column in each group. Red arrowheads, BM-derived fibroblasts; blue arrowheads, hematopoietic cells. Values are the mean ± SEM of an experiment performed in triplicate (one-way analysis of variance). *P < 0.05 versus the WT group. #P < 0.05 versus the NH4Cl-loaded group

The proportion of bone marrow-derived MSCs was elevated in the tumor infiltration model and uremic model

To further investigate the proportion of MSCs in bone tissue in vivo with both normal and high ammonia concentrations, we established two high-level ammonia concentration models: a uremic model and a tumor infiltration model.28 First, we detected the peripheral blood ammonia concentration of the untreated uremic model and tumor infiltration model. The results revealed an expected difference in the ammonia concentration of the tumor infiltration group and uremic group compared to the untreated group (Fig. 8a). To assess the distribution of MSCs and HSCs in bone tissues in vivo, we separated mouse femurs and tibias and assessed them by flow cytometry analysis. The results demonstrated that the positive markers of MSCs, including CD29, CD44, CD105 and Sca-1, increased in the uremic group and tumor infiltration group compared to the untreated group (Fig. 8b–e), while the positive markers of HSCs and CD45 decreased in the uremic group and tumor infiltration group compared to the untreated group (Fig. 8f). On the basis of these phenomena, to further assess the presence of MSCs in bone tissue in situ, we used spinal section staining with HE (Fig. 8g). The results revealed that more fibroblasts and fewer hematopoietic cells appeared at a higher proportion in both the uremic model group and the tumor model group compared to the untreated group. Moreover, we found that in the bone metastasis tissue sections of human lung cancer and breast cancer, more fibroblasts appeared around cancer cells, the residual hematopoietic cells appeared distal to the cancer cells at the early stage, and fibroblasts were present instead of myeloid cells around cancer cells at the late stage (Fig. 8h). These results showed that some pathological conditions with hyperammonemia, such as uremia or tumors, could also elevate the proportion of MSCs in vivo.

Fig. 8

The proportion of bone marrow-derived MSCs was elevated in the tumor infiltration model and uremic model. a The ammonia concentration (µmol·L−1) of wild-type mice, uremia model mice and tumor infiltration mice. b The percentage of MSCs in CD45- cells from mouse tibias and femurs in each group. c The statistical results of the percentage of MSCs in BM. d The percentage of the expression of MSC surface markers (CD29, CD44, CD105, Sca-1) and the expression of CD45 in each group. e The statistical results of the percentage of MSC surface marker expression in each group. f The statistical results of CD45 expression among in group. g HE staining of the spinal column in each group. Red arrowheads, BM-derived fibroblasts; blue arrowheads, hematopoietic cells; black arrowheads, LL2 cells; black arrowheads with diamond, completed bone trabecula; black arrowheads with circle, broken bone trabecula. h HE staining of bone with infiltrating lung cancer or breast cancer cells from patients. Red arrowheads, BM-derived fibroblasts; blue arrowheads, hematopoietic cells; black arrowheads, cancer cells. Values are the mean ± SEM of an experiment performed in triplicate (one-way analysis of variance). *P < 0.05 versus the WT group