INTRODUCTION
During prenatal development, the fetus resides in a relatively low-glucose environment compared to the mother. In rodents, fetal blood glucose is about 3.7 mM, compared to about 5 mM in the maternal circulation
1, a gradient conserved across mammals
2,3. This gradient is not pathological, but represents a physiological adaptation essential for facilitated glucose transfer via the GLUT transporters across the placenta
4. Furthermore, during intrauterine development, the fetus has to adapt to fluctuations in maternal nutrient supply
1; for instance, a 24 h maternal fast in rats can reduce fetal glucose to 2 mM or lower
1. Remarkably, despite this intrinsically low and variable glucose milieu, the fetus sustains rapid growth, which requires high anabolic activity, particularly in organs such as the liver, the site central to developmental protein synthesis and hematopoiesis
5,6. This prompts a fundamental question: how does the fetus, especially its liver, maintain anabolic metabolism under chronic low glucose conditions?
Metabolic adaptation to glucose availability is orchestrated by two central regulators: mTOR Complex 1 (mTORC1), which promotes anabolism, and AMP-activated protein kinase (AMPK), which stimulates catabolism
7. It was shown that glucose levels can directly control the activities of mTORC1 and AMPK
8-13. In ample glucose, mTORC1 is active and promotes anabolism
14-16. Conversely, during periods of glucose and/or energy deficiency, AMPK is activated, which promotes catabolism and inhibits anabolism
17-19. This inhibition occurs through a glucose-sensing pathway, in which vacuolar H
+-ATPase (v-ATPase), a lysosomal proton pump, plays a pivotal role
20,21. Low glucose leads to a decrease in the glycolytic intermediate fructose-1,6-bisphosphate (FBP), leaving aldolase, an enzyme associated with v-ATPase which binds and cleaves FBP to dihydroxyacetone phosphate (DHAP) and glyceraldehyde-3-phosphate (G3P), in an unoccupied state
22,23. This unoccupied aldolase then inhibits neighboring transient receptor potential vanilloid (TRPV) cation channels on the endoplasmic reticulum (ER), reducing local Ca
2+ levels at ER–lysosome contact sites
12,20,23. The subsequent interaction between TRPV and v-ATPase reconfigures the aldolase–v-ATPase complex, resulting in inhibition of the v-ATPase
20. This inhibition permits the scaffold protein AXIN, in complex with LKB1, to translocate to the lysosomal surface by utilizing v-ATPase and its associated Ragulator (comprised of 5 LAMTOR subunits, LAMTOR1–5) as docking sites
21,24. AXIN binding promotes the conformational changes of Ragulator, disabling its ability to release GTP from RAGC (one component of the small GTPase RAGs (RAGA to RAGD))
21,25, and hence triggering the switch of RAG heterodimers from an “on” state (RAGA
GTP–RAGC
GDP) to an “off” state (RAGA
GDP–RAGC
GTP)
26. Consequently, mTORC1 is dissociated from the Ragulator–RAGs complex
27, and hence also detached from its allosteric activator, the lysosomal pool of Rheb GTPase
28,29. Concurrently, upon translocation onto the lysosomal surface, the AXIN-carried LKB1, an upstream kinase, phosphorylates and activates the lysosomal pool of AMPK
21. Activated AMPK can further inhibit mTORC1 by phosphorylating its component Raptor
30,31, or the upstream inhibitor TSC2 to inhibit Rheb
10. It is noteworthy that the dissociation of mTORC1 from the lysosome is autonomous in low glucose, as it can still be inhibited in the absence of AMPK
32,33, albeit at a slightly slower rate
22,34-36. In addition to glucose, the components of the lysosomal glucose-sensing pathway can also be utilized by amino acids and growth factors, which also play critical roles in maintaining the activity of mTORC1. For example, high leucine and arginine can activate RAGs via both the v-ATPase–Ragulator complex
26,37-41 and the GATOR1/2 complex
42-44. Amino acids could also inhibit TSC2 or elevate intracellular Ca
2+ that activates Rheb
45-47, or inhibit the ubiquitination of mTOR that allosterically activates mTORC1
48. In parallel, growth factors activate Rheb through the PI3K-AKT pathway that inhibits TSC2
49-51.
It was shown that mTORC1 activity is indispensable for normal fetal development
52,53. However, since the fetal liver is constantly exposed to low glucose levels, it remains unclear how mTORC1 is regulated to maintain anabolic output to support growth.
RESULTS
mTORC1 is active in fetal hepatocytes in low glucose
We subjected BNL-Cl2 cells, a cell line derived from mouse fetal liver, to treatment with different glucose concentrations. In low glucose (< 5 mM), we found that BNL-Cl2 cells did not show inhibition of mTORC1 activity despite significant activation of AMPK (Fig. 1a), as evidenced by the phosphorylation of S6K and 4e-BP1, and the phosphorylation of AMPKα and ACC, respectively. As a control, AML12 cells (derived from adult mouse liver), HEK293T cells, and MEFs all exhibited rapid inhibition of mTORC1 when cultured in low glucose (Fig. 1a). We also tested the effects of low glucose over different time durations, showing that prolonged glucose starvation still did not inhibit mTORC1 in the BNL-CL2 cells (Fig. 1b). Moreover, in primary mouse hepatocytes derived from the fetal liver (PFH), we found that mTORC1 also remained active in low glucose (Fig. 1c, d). These results indicate that low glucose cannot inhibit mTORC1 in cells derived from the fetal liver. The mTORC1 activity of BNL-CL2 was also not inhibited by treatment with the AMPK agonist — MK8722
54, and was only slightly increased with treatment of the AMPK inhibitor — BAY-3827
55 (Supplementary Fig. S1a). At the organismal level, we fasted pregnant mice (fetuses of E18.5), along with unpregnant female adult mice, for 24 h. This resulted in a significant decrease of plasma glucose (from 3.5 to 2.1 mM in the fetus; from 9.3 to 4.4 mM in the adults, Supplementary Fig. S1b), and activation of hepatic AMPK in both fetus and adults (Fig. 1e; note that the level of AMPK activation in fetal liver was much higher than that in adult liver). Consistently, we observed that the nuclear translocation of TFEB, which is regulated by AMPK independently of mTORC1
56, is promoted in fetal liver cells during starvation (Supplementary Fig. S1c). We also observed active mTORC1 in the fetal liver, and inhibition of mTORC1 in the adult liver during starvation (Fig. 1e). However, for an unknown reason, we can still observe a decrease of TFEB-S122 phosphorylation in fetal livers (Fig. 1a–e), which is known to be catalyzed by mTORC1
57. In human liver samples taken from adult patients, we found a negative correlation between AMPK activation and mTORC1 activation (Fig. 1f–h). In comparison, we did not observe such a correlation in human fetal livers, where higher AMPK activation did not correspond to lower mTORC1 activation (Fig. 1i–k).
Overall, these data indicate that low glucose is unable to inhibit mTORC1 in fetal hepatocytes or in the fetal liver, even when AMPK is allosterically activated by increased levels of AMP. However, the resistance of mTORC1 to inhibition by low glucose occurs only in the fetal liver tissue, but not in the other tissues. For example, in the fetal heart, lung, and muscle tissues, mTORC1 was significantly inhibited after 24 h of maternal fasting (Fig. 1l–n). Of note, the resistance to low glucose-induced inhibition of mTORC1 was maintained during the perinatal state but disappeared from postnatal day 1 (Fig. 2a, b).
Inability to inhibit the v-ATPase–Ragulator–RAG complex underlies maintenance of mTORC1 activity in fetal hepatocytes
We next investigated the mechanisms through which the low glucose-induced inhibition of mTORC1 is blocked in the fetal liver. We examined the phosphorylation of TSC2 and Raptor, substrates of AMPK, which mediate the inhibition of mTORC1 by AMPK. As shown in Fig. 1e and Supplementary Fig. S1a, we found intact phosphorylation of TSC2 (at S1387)
10 and Raptor (S792)
30 in the fasted fetal liver tissue and BNL-CL2 after glucose starvation or MK8722 treatment. We also investigated whether a lack of glucose could dissociate mTORC1 from the lysosomal membrane as a consequence of detaching from RAG and Ragulator
25,26. We immunoprecipitated RagA, and observed that mTORC1 remained associated with RAG in fetal hepatocytes in low glucose (Supplementary Fig. S1d). In contrast, mTORC1 was not associated with RAG in adult hepatocytes (Supplementary Fig. S1d). Immunofluorescent staining further showed that mTORC1 remained co-localized with the lysosomal marker LAMP2 in fetal hepatocytes in low glucose, but not in adult hepatocytes (Fig. 2c–e; Supplementary Fig. S1e–h). We also determined the activity of v-ATPase, assessed by the fluorescent intensity of Lysosensor dye that positively correlates with the lysosomal acidity
58. We observed that glucose starvation did not affect the lysosomal acidity in fetal hepatocytes (Supplementary Fig. S1i–k), indicating that v-ATPase remains active. In contrast, low glucose effectively inhibited v-ATPase in adult hepatocytes (Supplementary Fig. S1l–n). As a control, forced inhibition of v-ATPase in fetal hepatocytes by concanamycin A (conA) effectively triggered the lysosomal translocation of AXIN (Supplementary Fig. S2a, b), disrupted the lysosomal localization of mTORC1 (Fig. 2c–e), and led to its inhibition (Fig. 2f). The results above indicate that the mTORC1 activity in the fetal liver is maintained due to the unaffected v-ATPase-Ragulator-RAG axis in low glucose.
We also examined the effects of amino acids and growth factors on mTORC1 in fetal liver cells, and found that mTORC1 is readily inhibited in these cells upon the withdrawal of amino acids or growth factors (serum), as evidenced by decreased p-S6K (Fig. 2g, h) and lysosomal localization of mTORC1 (Supplementary Fig. S2c–f). Therefore, mTORC1 in fetal hepatocytes is resistant specifically to glucose withdrawal, but not to the removal of amino acids or growth factors.
Consistent with the constitutive activity of mTORC1, maternal fasting did not decrease the concentrations of leucine, arginine, and glutamine, all of which could independently sustain mTORC1 in an active state when growth factors and glucose were present
59,60 (Supplementary Fig. S2g–j; see also the contents of other amino acids in Supplementary Fig. S3a–q). Similarly, maternal fasting did not inhibit the PI3K-AKT pathway in the fetal liver (Fig. 2i), although it did lower the serum levels of insulin, IGF-1, and IGF-2 (Supplementary Fig. S4a–c). Consistently, when insulin levels in the culture medium were reduced from 200 pg/mL to 100 pg/mL, reflecting the changes in serum insulin in the fetus before and after 24 h of fasting, the activity of mTORC1, along with AKT phosphorylation, remained unchanged in cultured fetal hepatocytes (Supplementary Fig. S4d, e). In adult hepatocytes, insulin levels observed in adult mice, both pre-fasting and post-fasting, did not alter the activation of AKT, but they further enhanced the high-glucose-induced mTORC1 activation during the pre-fasting stage (Supplementary Fig. S4f, g). We also investigated the potential roles of glucagon, which induced calcium influx in liver cells
61 and may help compensate for the decrease in TRPV-released calcium during fasting. We observed that glucagon levels were slightly higher in fetal mice (29.0–31.6 pg/mL at embryonic days 16.5 and 18.5) compared to those seen at postnatal day 2 and 8 weeks (21.0–22.7 pg/mL) (Supplementary Fig. S4h). However, supplementation of glucagon at the concentration found in fetal mice did not restore mTORC1 activation in adult liver cells cultured in low glucose (Supplementary Fig. S4i). Therefore, the changes in insulin and glucagon during fasting do not influence mTORC1 activation in the liver.
Low glucose-triggered lysosomal AMPK pathway is blocked in the fetal liver
We next investigated why the v-ATPase-Ragulator-RAG axis in the fetal liver is not inhibited by low glucose. In adult liver or hepatocytes, low glucose renders the v-ATPase-associated aldolase unoccupied with FBP, triggering inhibition of v-ATPase and consequently the inhibition of the Ragulator–RAG complex
20,23. We thus determined the levels of FBP and DHAP in fetal hepatocytes in low glucose, and observed a strong decrease in FBP and DHAP levels in both fetal liver cell lines and primary fetal hepatocytes in low glucose (Supplementary Fig. S5a–l). Such a decrease could also be observed in the fetal liver after maternal fasting (Supplementary Fig. S5m–o), similarly to the adult mouse livers (Supplementary Fig. S5p–r); the levels of other glycolytic intermediates are shown in Supplementary Figs. S6a–n, S7a–n, and S8a–n. Therefore, the maintenance of fetal liver mTORC1 activity in low glucose is not due to sustained intracellular glucose, FBP, or DHAP.
We also investigated whether aldolase and TRPV are affected by low glucose. We found that in fetal hepatocytes, all three isozymes of aldolase (ALDOA to ALDOC) could bind to TRPV (Fig. 3a; represented by TRPV4). TRPV4 had the highest mRNA levels in both BNL-CL2 (according to transcriptomics) and fetal liver tissues (according to qPCR) compared to other TRPVs (Supplementary Fig. S9a, b), and it could be detected in the immunoblotting in fetal hepatocytes and livers (Supplementary Fig. S9c). In addition, low glucose/FBP enhanced the aldolase–TRPV4 interaction (Fig. 3a), as in HEK293T cells
20. Surprisingly, the TRPV activity, evidenced by the fluorescent GCaMP6s-TRPV4 indicator
20, remained unaffected in low glucose levels in fetal hepatocytes (Fig. 3b, c), whereas the TRPV activity decreased in adult liver or 293T cells under the same low glucose condition (Fig. 3d–f). When TRPV was pharmacologically inhibited by the antagonist BCTC, as evidenced by the decreased GCaMP6s-TRPV4 fluorescence (Fig. 3b, c), mTORC1 activity was decreased significantly in fetal hepatocytes (Fig. 3g), and mTORC1 was dissociated from the lysosomes, regardless of glucose availability (Fig. 3h, i). We also found that BCTC could inhibit the activity of v-ATPase, likely as a consequence of the inhibition of TRPV (Supplementary Fig. S9d, e), and effectively triggered the lysosomal translocation of AXIN in fetal hepatocytes (Supplementary Fig. S9f, g). In addition, the TRPV agonist GSK101 did not change the already high mTORC1 activity in fetal hepatocytes; in contrast, GSK101 sufficiently overrode the low-glucose-induced inhibition of TRPV to restore mTORC1 activity in adult hepatocytes (Fig. 3j, k). These results suggest that TRPVs are not inhibited by low glucose in fetal liver cells, such that the low glucose signal cannot be transmitted to inhibit v-ATPase or to trigger the lysosomal translocation of AXIN, events required for mTORC1 dissociation from the lysosome
21,23. We also investigated why AMPK can still be activated in low glucose in fetal hepatocytes while the lysosomal pathway is apparently not operating. To that end, we determined the levels of ATP, ADP, and AMP. We found that glucose starvation significantly increased the ratios of AMP:ATP and ADP:ATP in fetal hepatocytes (Supplementary Fig. S9h, i). This indicates that the fetal AMPK is allosterically activated through the canonical, AMP-dependent mechanism
17.
Acetylation of TRPV during intrauterine development confers insensitivity to low glucose
We next investigated how TRPV is resistant to inhibition by FBP-unoccupied aldolase under low glucose conditions. We found that the protein level of TRPV in fetal livers was lower than that in adult livers (Supplementary Fig. S10a, b). However, as BCTC could inhibit fetal liver TRPV (Fig. 3b, c, g–i; Supplementary Fig. S9d–g), we concluded that the relatively low protein levels of TRPV in fetal livers, compared to those in adult livers, cannot explain the resistance of fetal TRPV to inhibition by low glucose. We therefore explored whether alteration of the post-translational modifications of TRPV may contribute to the constitutive activation of this cation channel. Forty-one acetylated lysine residues were identified by protein mass spectrometry. We then individually mutated the 41 lysine residues to arginine (K to R mutation) to mimic the deacetylated state, and found that the TRPV4-K608R mutant, when expressed in fetal hepatocytes, rendered mTORC1 sensitive to low glucose levels (Fig. 4a). Importantly, the levels of acetylated TRPV4-K608 were significantly higher in fetal liver/hepatocytes than in adult liver/hepatocytes, as confirmed by quantitative mass spectrometry analysis (Supplementary Fig. S10c–m). The K608 residue is also highly conserved among various species and among different members of the TRPV family (Supplementary Fig. S11a). In addition, we found that v-ATPase inhibition (Supplementary Fig. S11b–d), AXIN translocation to the lysosome (Supplementary Fig. S11e–h), and mTORC1 dissociation from the lysosome (Fig. 4b–e), could all be observed in fetal hepatocytes expressing TRPV4-K608R in low glucose conditions, as seen in adult cells, although the mutation of K608 did not affect the basal activity of TRPV4 (Supplementary Fig. S11i, j) or change its affinity with the FBP-unoccupied aldolase (Supplementary Fig. S11k). As it has been reported that acetylation of lysine residues is related to ubiquitination and degradation of proteins
62-64, we then determined the ubiquitination of TRPV4. We found that TRPV4-K608R did not alter the ubiquitination of TRPV4, although the mutation slightly decreased the half-life of TRPV4 (Supplementary Fig. S12a, b). Of note, we found that the TRPV4-K608Q mutation, which is expected to mimic the acetylated state, failed to replicate this effect, as the ectopic expression of TRPV4-K608Q in AML12 cells did not block the low glucose-induced mTORC1 inhibition (Supplementary Fig. S12c). This finding may align with the previous studies indicating that not all K to Q mutants can effectively mimic the state of acetylation
65. We therefore conclude that acetylation of K608 renders TRPV4 insensitive to the low glucose-/FBP-unoccupied aldolase-induced inhibition.
P300 mediates the acetylation of TRPV during intrauterine development
We next studied how the high TRPV acetylation observed during intrauterine development is maintained. We found that the levels of acetyl-CoA, a substrate for acetyltransferases
66,67, were much lower in fetal liver compared to adult liver, suggesting that the elevated acetylation of TRPV4-K608 in the fetal liver is not due to the concentration of acetyl-CoA (Supplementary Fig. S12d). We then screened the acetyltransferase(s) and deacetylase(s) that are required for the acetylation and deacetylation of TRPV4-K608, respectively. As shown in Fig. 4f and 4g, we discovered that ectopic expressing the deacetylases SIRT2 or SIRT7 or knocking down acetyltransferase P300 in BNL-CL2 cells allowed mTORC1 to be inhibited in low glucose. In contrast, knockdown of 15 other acetyltransferases or ectopic expression of 16 other deacetylases did not result in such effects (Supplementary Figs. S13a–o, S14a–c). Moreover, ectopic expression of SIRT2 or SIRT7 in BNL-CL2 cells significantly reduced the acetylation level of TRPV4, similar to that after the knockdown of P300 (Fig. 4h; Supplementary Figs. S15a–h, S16a, b). Importantly, the expression levels of P300 in the liver decreased dramatically after birth (Fig. 4i). In contrast, the expression levels of SIRT2 and SIRT7 remained largely unchanged before and after birth (Fig. 4i). In addition, we investigated a possible direct acetylation of TRPV4 by P300 using a cell-free assay with bacterially expressed and purified TRPV4 protein and P300 protein purified from HEK293T cells. We found that the acetylation of TRPV4-K608 was detected only in the presence of P300 (Fig. 4j). Therefore, we conclude that P300 plays a crucial role in maintaining the acetylation of TRPV during intrauterine development.
TRPV acetylation is crucial for normal intrauterine development
We next investigated the physiological significance of TRPV acetylation during fetal development in utero, by generating transgenic mice with liver-specific expression of TRPV4-K608R (induced by
Alb-
Cre; validated in Supplementary Fig. S17a–c). Compared to wild-type mice, the levels of TRPV4 in the liver of mice containing liver-specific knock-in of TRPV4-K608R (LKI) remained unchanged (Supplementary Fig. S17d, e), and TRPV4-K608 acetylation was almost undetectable (Supplementary Fig. S18a–m). We then examined the phenotypes of the fetuses in these mice. As shown in Supplementary Fig. S19a, approximately 18% (2/11) of embryos with liver-specific expression of TRPV4-K608R (TRPV4-K608R-LKI) died during intrauterine development. Among the surviving embryos, we observed a significant decrease in mTORC1 activity in the liver (Fig. 5a–d), while the activation of AMPK was unaffected (Supplementary Fig. S19b–d), which is consistent with the notion that AMPK is activated through a canonical mechanism, which is independent of TRPV. In line with the reduced anabolic activity as a consequence of mTORC1 inhibition, the livers of TRPV4-K608R-LKI embryos displayed lower weights, as well as decreased levels of albumin, hepatocyte growth factor (HGF), and IGF-1 compared to their wild-type littermates (Fig. 5e–i; Supplementary Fig. S19e–n). In addition, the protein content in the livers was reduced; mass spectrometry revealed that out of 9,219 detected and quantified protein species, 1,498 showed significantly decreased levels (
P < 0.05), and 308 decreased by more than 2-fold (Supplementary Fig. S19o). Simultaneously, the overall protein content in the whole fetus (with the liver removed) was also diminished (Supplementary Fig. S19p, q). We also observed a lower proliferation rate in the livers of TRPV4-K608R-LKI embryos, as determined by Ki67 staining (Supplementary Fig. S20a, b), which may correlate with the reduced liver weight noted in these mice. Of note, the levels of hemoglobin, erythropoietin and blood glucose remained unchanged in the TRPV4-K608R-LKI embryos (Supplementary Fig. S20c–j). In addition, we quantified the number of surviving offspring from these mice, and found that none of the offspring expressing K608R-TRPV4 survived to the weaning stage (Details shown in Supplementary Table. S1). Consistently, we found that rapamycin administration in wild-type embryos, which inhibits mTORC1 in the livers, recapitulates the postnatal death observed in the TRPV4-K608R-LKI embryos (Supplementary Fig. S21a–f). Note that although we attempted to block the fetal mTORC1 signaling by knocking out
Raptor in the liver as well, it turned out to be unsuccessful. The protein levels of Raptor remained mostly unaffected in the fetal liver (Supplementary Fig. S21g–i), which is in stark contrast to those in adult livers (significantly decreased in the liver of
RaptorF/F mice crossed with the
Alb-Cre mice)
68. This observation may also explain why liver-specific knockout of
Raptor can survive normally and remain fertile, although smaller sizes of the adult livers were observed
68. Together, we found that the TRPV4 K608 acetylation, through maintaining the mTORC1 activity under low glucose condition in the fetal liver, plays a critical role in embryo development.
DISCUSSION
During pregnancy, most of the glucose provided by the uterine artery is consumed by the uterus and placenta, and only about a fraction of one-third can reach the fetus
69,70. Therefore, the glucose concentrations in the fetus are lower than maternal levels, displaying a transplacental gradient of glucose concentration during embryonic development
71. In addition, it was reported that in pregnant women with insulinoma, who constantly experience higher rates of insulin secretion and subsequently suffer from hyperinsulinemia, there are no significant defects in the survival and development of the fetus
72,73. This requires a mechanism that sustains mTORC1 activity, as it is essential for maintaining the anabolic activities in the fetal livers amid maternal glucose fluctuations.
During embryogenesis, the liver is formed from hepatoblasts, which are derived from definitive endoderm at around 3–4 weeks post-conception in humans, or at E8.5–E9.0 in mice
74. In mice, the differentiation of hepatoblasts into hepatocytes and cholangiocytes starts at E13.5, while in humans, mature hepatobiliary cells can be observed at around 7 weeks post-conception
75. Our work also identified mTORC1 as a critical mediator for the development of the liver, in addition to its general roles in cell proliferation in both extraembryonic and embryonic compartments
52. As we observed lowered fetal glucose levels from E10.5 to P1 in pregnant mice without dietary restrictions (Fig. 2a), and the TRPV4-K608R-LKI fetal liver, in which mTORC1 is inhibited by low glucose, displays developmental abnormalities, it is evident that a constant activity mTORC1 plays a crucial role during the developmental stage of the fetal liver, which is continually exposed to low glucose levels. In support of such a notion is that reduced activity of mTORC1 seen in the fetus of TRPV4-K608R-LKI mice, leads to decreased protein levels throughout the entire fetal mouse body. Thus, sustained mTORC1 activity by blocking the inhibition in the intrinsically low glucose conditions ensures that the liver generates sufficient proteins and other molecules to provide to the entire fetus, helping to explain why only the liver possesses the unique ability to maintain mTORC1 activity, while other tissues cannot.
Our current study has revealed the mechanism that prevents the glucose-sensing pathway from operating in the fetal liver for the sustainability of anabolic activities. We have found that acetylation of TRPV4 on K608 residue renders TRPV4 unaffected by the FBP-unoccupied aldolase, which highlights TRPV as a checkpoint for the glucose sensing-triggered AMPK activation and mTORC1 inhibition. Models of glucose-sensing pathways in adult and fetal livers are shown in Fig. 5j.
The glucose sensing pathway begins to operate in the liver right after birth, highlighting the importance of regulating mTORC1 activity for maintaining the balance between anabolism and catabolism during postnatal development. P300 regulates the acetylation level of TRPV4 K608 in fetal liver, and the expression of P300 decreased significantly after birth, while it has been reported that P300 can acetylate Raptor, leading to high activity of mTORC1
76. Previous studies have shown that sustained RagA activity through introducing RagA-Q66L mutant (RagA
GTP) allows mTORC1 to maintain constitutive activity. While
RagAGTP/GTP mice can develop normally in the uterus, they are unable to survive for more than 24 h after birth
32. Therefore, the transition of mTORC1 from the constitutive activity in the fetus to dynamically regulation after birth is crucial for the normal development of the newborns.
MATERIALS AND METHODS
Studies in animals
Wild-type C57BL/6J (#000664),
RaptorF/F (#013191, generated as in previous study)
68, and
Alb-
Cre mice (#003574) were obtained from the Jackson Laboratory.
Raptor-LKO mice were generated by crossing
RaptorF/F mice with
Alb-
Cre mice. Genotyping primers for
Raptor and
Alb-
Cre were supplied in Supplementary Data S2. For identification of
RaptorF/F mice, the wild-type allele was present, the PCR reaction generated a 140-bp product, while the presence of the
3’-LoxP element resulted in a 174-bp amplicon. The PCR reaction generated an amplicon of approximately 570 bp, indicating the presence of
Alb-
Cre. To ensure the reliability of the PCR system, a fragment of approximately 150 bp was used as a positive control. The TRPV4 wild-type (N7-100246) and TRPV4-K608R (N7-100247) knock-in mice were customized by Shanghai Model Organisms Center, Inc. by introducing the CAG-LSL-
TRPV4-WT (or K608R)-HA-WPRE-pA cassette into the recipient C57BL/6J mice at the
Rosa26 locus, following the protocol as described previously
77 (see details in the Supplementary Data S1). Briefly, to create these mice,
Cas9 mRNA and gRNA were first produced by
in vitro transcription. Then, a donor vector containing a 3.3-kb 5’ homologous arm, the knock-in sequence, and a 3.3-kb 3’ homologous arm was constructed by the in-fusion cloning strategy. Next, the
Cas9 mRNA, gRNA, and the donor vector were mixed, and micro-injected into fertilized eggs of C57BL/6J mice, and the F0 mice with successful knockin, were identified by PCR and sequencing, and were crossed with wild-type C57BL/6J mice to generate the F1 mice. The homozygous
TRPV4F/F and
TRPV4-K608RF/F mice resulting from this process were viable, fertile, and normal with no apparent phenotypic abnormalities, and were crossed with
Alb-
Cre mice to generate the
TRPV4-LKI and
TRPV4-
K608R-LKI mice. Primers for genotyping
TRPV4-K608R and
TRPV4-WT were supplied in the Supplementary Data S2. The PCR reaction generated an amplicon of 967 bp when the wild-type allele site was present, and of 580 bp when the
TRPV4-K608R was present. The PCR reaction generated an amplicon of 967 bp when the wild-type allele site was present and of 737 bp when the
TRPV4-WT was present.
Alb-
Cre was genotyped as in the
Raptor-LKO mice.
The genotyping process was carried out as follows: initially, a 1.5-mm length of the mouse’s tail was cut using a surgical blade (to determine the genotype of fetal mice, a limb or entire tail of the mice was cut) and placed in a DNA-free, 1.5-mL EP tube. 100 μL of 25 mM NaOH solution was then added to the 1.5-mL EP tube, followed by centrifugation at 3,000× g for 1 min to ensure that the tail tissues were fully immersed in the NaOH solution, and then heated in a metal bath at 100 °C for 1 h. After heating, the tube was centrifuged (3,000× g, 1 min) to ensure that all liquids or tissues were pelleted at the bottom of the tube, and the mixture was then neutralized by the addition of 100 μL of 40 mM Tris-HCl, pH 8.0 solution. The PCR reaction system consists of 10 μL of 2× Taq PCR Master Mix, 1 μL of mouse tail DNA sample, 0.5 μL (10 µM) of forward primer, 0.5 μL (10 µM) of reverse primer, and 8 μL of ddH2O. The programs were set as follows:
a) For Alb-Cre:
- Initial denaturation at 95 °C for 3 min
- 35 cycles of denaturation at 95 °C for 30 s, annealing at 58 °C for 40 s, extension at 72 °C for 30 s
- Final extension at 72 °C for 5 min
- PCR products were analyzed using a 2% DNA agarose gel.
b) For RaptorF/F:
- Initial denaturation at 95 °C for 5 min
- 10 cycles of denaturation at 95 °C for 30 s, annealing starting at 65 °C with a decrease of 0.5 °C per cycle, extension at 72 °C for 30 s
- Followed by 28 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s, extension at 72 °C for 30 s
- Final extension at 72 °C for 5 min
- PCR products were identified using a 3% DNA agarose gel.
c) For TRPV4F/F and TRPV4-K608RF/F:
- Initial denaturation at 95 °C for 5 min
- 35 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s, extension at 72 °C for 30 s
- Final extension at 72 °C for 5 min
- PCR products were identified using a 2% DNA agarose gel.
Mouse fasting treatments
Protocols for all mouse experiments were approved by the Institutional Animal Care and the Use Committee of Xiamen University (XMULAC20210014). Unless stated otherwise, mice were housed with free access to water and a standard diet (65% carbohydrates, 11% fat, 24% protein) under specific pathogen-free conditions. The lighting was on from 8:00 to 20:00, with the temperature maintained at 21–24 °C and humidity at 40–70%. Unless stated otherwise, male and female mice were randomly selected in this study at an approximately 1:1 ratio. Mice used for collection of the liver tissues or isolation of primary hepatocytes were also selected randomly, without distinction of sex. Mice were individually caged for one week before each treatment. For fasting experiments, the mice were placed in a new cage at 15:00 p.m., and the experiment was conducted the next day at 15:00 p.m.
Rapamycin treatment of mice
The E11.5 C57BL/6 dams were randomly divided into three groups. The dams received intraperitoneal injections of either corn oil or 5 mg/kg rapamycin (MedChemExpress, cat. HY-10219 dissolved in anhydrous ethanol with a stock solution of 5 mg/mL and freshly diluted to a working concentration of 0.5 mg/mL by diluting the stock solution with 9 volumes of corn oil), or 10 mg/kg rapamycin once a day from E12.5 to E17.5. Rapamycin was injected slowly due to the high viscosity of corn oil to prevent bubble formation. In addition, extra caution should be taken when administering drugs as the gestational age increases to avoid puncturing the amniotic sac with the needle when the abdominal bulge of pregnant mice becomes more pronounced. At 18.5 days of pregnancy, the pregnant mice were anesthetized with isoflurane, and the fetal livers, hearts, lungs, and kidneys were excised and weighed or quickly lysed for the determination of the levels of p-AMPKα, AMPKα, p-S6K, S6K, Raptor, and β-Actin by immunoblotting.
Primary hepatocytes
Mouse adult primary hepatocytes were isolated with a modified two-step perfusion method, as described previously
77,78. Briefly, mice were anaesthetized before the isolation, followed by the insertion of a 0.72 × 19 mm intravenous catheter into the post-cava. After cutting off the portal vein, mice were perfused with 50 mL of liver perfusion medium (A buffer; prepared by mixing 49 mL of Kreb buffer with 1 mL of 50 mM EGTA buffer) at 5 mL/min, followed by 50 mL of liver digest buffer (B buffer; prepared by dissolving 25 mg of Type IV collagenase (Sigma, cat. C5138) in a solution made by mixing 49 mL of Kreb buffer with 1 mL of 250 mM CaCl
2 (Sango Biotech, cat. A610050-0500)) at a rate of 2.5 mL/min. The Kreb buffer (pH 7.4) was prepared by dissolving 7 g of NaCl (Sango Biotech, cat. A610476), 2 g of NaHCO
3 (Sango Biotech, cat. A500873-0500), 1.1915 g of HEPES (Gibco, cat. 15630130), and 1 g Glucose (Sigma, cat. G7021-1kg), in a solution containing 10 mL of Solution C (180 mM KCl (Sango Biotech, cat. A100395), 120 mM MgSO
4·7H
2O (Sigma, cat. M1880-500G), 120 mM KH
2PO
4 (Sango Biotech, cat. A100781-0100)) and 990 mL of deionized water, followed by filtration with 0.22 μm pore size filters (Sartorius, cat. 16541K). The digested liver was then briefly rinsed with PBS and dissected by gently tearing apart the Glisson’s capsule with two sterilized, needle-pointed tweezers on a 6-cm dish containing 3 mL of PBS. The dispersed cells were mixed with 10 mL of cold William’s medium E plus 10% fetal bovine serum (FBS) and filtered by passing through a 100-μm cell strainer (BD Falcon). Cells were then centrifuged at 50×
g at 4 °C for 2 min, followed by washing twice with 10 mL of ice-cold William’s medium E plus 10% FBS (Gibco, cat. 30044333). Cells were then immediately plated (at 60–70% confluence) in collagen-coated 6-well plates in William’s medium E plus 10% FBS, 100 IU penicillin, and 100 mg/mL streptomycin. After 4 h of attachment, the medium was replaced with fresh Dulbecco’s modified Eagle’s medium with 10% BSA for another 48 h before further use. In this manuscript, we did not add insulin during the glucose starvation treatment of PAH, except for Fig. 2h.
Mouse FPHs were isolated following the procedures described previously
79, with minor modifications:
a) Preparation of HEPES buffer:
HEPES buffer was prepared by adding 8 g NaCl, 0.2 g KCl, 0.1 g Na2HPO4·12H2O, and 10 mL of 1 M HEPES solution to 750 mL of autoclaved distilled water, followed by adjusting the pH to 7.65 and the volume to 1 L. The solution was filtered with a 0.22-μm filter and stored at 4 °C.
b) Preparation of basal culture media and complete culture media:
Basal culture media were prepared by adding 50 mL of FBS, 5 mL of L-glutamine (200 mM stock solution), and 5 mL of the antibiotic (streptomycin 10,000 μg/mL, penicillin 10,000 IU/mL) solution to William’s E medium to a final volume of 500 mL.
Complete culture media were prepared by adding HGF (final concentration 25 ng/mL), EGF (25 ng/mL), Insulin (5 μg/mL), Hydrocortisone (0.5 μM), and Dexamethasone (0.1 μM) to the basal culture media.
c) Preparation of Liver Dissociation Solution:
We dissolved the lyophilized liberase (5 mg of collagenase I) in 2 mL of sterile water to get a 13 U/mL enzyme solution.
d) Isolation of fetal livers from anesthetized pregnant mice:
Pregnant mice were anesthetized, and the fetal liver tissues were excised using sterilized tweezers and scissors. Approximately 300 mg of liver tissue was immediately transferred into ice-cold HEPES buffer (4 °C) in a biosafety cabinet.
e) Preparation of fetal liver tissue:
The connective tissue and the gastrointestinal tract attached to the fetal liver tissue were removed using sterile forceps. The liver was then transferred to the clean, cool HEPES buffer and was rinsed twice by the cool HEPES buffer, followed by being sliced into small pieces by sterilized blades, and then incubated in 10 mL of liberase-HEPES-calcium buffer for 5 min at 37 °C. The homogenates were then aspirated using a 10-mL glass pipette once per min. Note that any pipetting (e.g., using the 1-mL pipette) should be avoided, as it will severely decrease cell viability. After 5 min of incubation, tissues were allowed another 5 min of incubation at 37 °C, followed by mixing with 25 mL of basal media to terminate the digestion.
f) Cell isolation and culture:
The homogenates were then filtered using a 70-µm filter (Corning, cat. 352350). After transferring the filtrate to a sterile, 50-mL centrifuge tube, cells in the filtrate were collected by centrifuging at 1,000× g for 2 min at 4 °C. Cells were then resuspended with 25 mL of basal medium, followed by a wash with the complete culture medium, and then resuspended in complete culture medium for culturing, followed by transfer to insulin-free medium (basal culture media) upon reaching full confluence.
Note that in this study, the isolated fetal liver cells were not further purified by magnetic bead sorting
79 as we did not observe any significant difference in the context of this study. The primary fetal hepatocytes were cultured for 2–3 days before being subjected to further immunofluorescence, immunoblotting, and metabolite analysis.
Clinical specimens
Fetal liver tissues (n = 12) were collected from women who voluntarily terminated the pregnancy at the Obstetrics and Gynecology Hospital of Fudan University. Adult liver tissues (n = 10) were the adjacent tissue of benign lesions from surgery at Fudan University Shanghai Cancer Center. The procedures involving adult participants (adult liver tissues) were approved by the Ethics Committee of Fudan University Shanghai Cancer Center (050432-4-2108*). The procedures involving pregnant participants (fetal liver tissues) were reviewed and approved by the Ethics Committee of Obstetrics and Gynecology Hospital of Fudan University (2023-66). Written informed consents were obtained from all participants. All tissues were collected in sterile tubes and quickly stored at −80 °C at the surgical site.
Tissue lysates were prepared as follows: appropriate amounts of tissue were cut (30–50 mg), and 10 μL/mg of protein lysis solution, pre-mixed with phosphatase inhibitors, was added to a 1.5-mL EP tube containing the tissues. A handheld homogenizer was used to break the tissue, and an ultrasound was applied to further lyse the cells. Centrifuged at 14,000× g for 10 min, and the supernatant was collected and mixed with an equal volume of 2× SDS solution. The mixture was well mixed, and the proteins were denatured at 100 °C for 15 min. The levels of p-AMPKα, AMPKα, p-S6K, S6K, and β-actin proteins were detected in the clinical specimens by Western blot. Based on the levels of p-AMPKα, the samples were categorized into two groups: AMPK-relatively inactive and AMPK-relatively active.
Cell lines
MEF cells were established as described previously
80. BNL-CL2, HEK293T, and MEF cells were maintained in Dulbecco’s modified Eagle’s medium (Gibco, cat. 11960044) supplemented with 10% FBS (Gibco, cat. 10099158), 100 IU penicillin, 100 µg/mL streptomycin at 37 °C in a humidified incubator containing 5% CO
2. The AML12 cells were cultured in DMEM/F12 (Gibco, cat. 11320033) supplemented with 10% FBS, 100 IU penicillin, 100 µg/mL of streptomycin, 4 ng/mL of dexamethasone, and ITS liquid media supplement (Sigma, cat. I3146). All cell lines were verified to be free of mycoplasma contamination. Cells were plated in 6-well plates one day before the experiments. On the next day, cells were rinsed once with PBS and then incubated with amino-acid-free EBSS and/or glucose-free DMEM supplemented with 10% FBS and relevant concentrations of drugs (Rapamycin, BCTC, Concanamycin A, etc.) for the indicated time. 1 L Amino acid-free EBSS buffer (pH 7.4): 0.2 g CaCl
2 (Sango Biotech, cat. A610050-0500), 0.2 g MgSO
4.7H
2O (Sigma, cat. M1880-500G), 0.4 g KCl (Sango Biotech, cat. A100395), 2.2 g NaHCO
3 (Sango Biotech, cat. A500873-0500), 6.8 g NaCl (Sango Biotech, cat. A610476), 0.14 g NaH
2PO
4.H
2O (Sango Biotech, cat. A100823) in 1,000 mL of deionized water. The dissolved EBSS liquid was filtered and sterilized using a filter with a pore size of 0.22 µm. Taking a 6-well plate as an example, to study the effect of amino acids on mTORC1 and AMPK in fetal hepatocytes, we first washed the cells once with PBS, added 2 mL of EBSS solution to each well, and then added the corresponding reagents in sequence: +G+A (25 mM glucose (Sigma, cat. G7021-1Kg) + 1× MEM Amino Acids (Gibco, cat. 11130-051) + 1× MEM Non-essential Amino Acid solution (Sigma, cat. M7145)), –G–A (No glucose, no amino acid), +G–A (25 mM glucose, no amino acid), –G+A (No glucose, + 1× MEM Amino Acid (Gibco, cat. 11130-051) + 1× MEM Non-essential Amino Acid solution). Taking the 6-well plate as an example, if only the effect of glucose starvation on AMPK and mTORC1 in fetal liver was studied, the cells were washed once with PBS, and then 2 mL of glucose-free DMEM medium containing 10% FBS was added to each well, followed by the corresponding reagents (Different concentrations of glucose). The glucose storage solution was 2.5 M (Sigma, cat. G7021-1Kg), with working concentrations of 0 mM, 1 mM, 3 mM, 5 mM, 10 mM, and 25 mM. Taking a 6-well plate as an example, to study the effects of growth factors on AMPK and mTORC1 in fetal hepatocytes, the cells were first treated overnight with high-glucose DMEM without serum, and then 2 mL of glucose-free DMEM medium without serum was added to each well. Finally, corresponding reagents were added: +G+I (5 µg/mL insulin (Procell, cat. PB180432) + 25 mM glucose), +G–I (no insulin + 25 mM glucose), –G+I (no glucose + 5 µg/mL insulin), –G–I (no insulin, no glucose). Then, the cells were lysed, and relevant proteins were detected by Western blot.
Plasmids
The restriction enzyme cleavage sites of pBOBI are BamHI and XhoI. The WT human TRPV4 sequence was based on CCDS9134.1. Point mutations of TRPV4 were performed by PCR-based site-directed mutagenesis using PrimeSTAR HS polymerase (Takara, cat. R010A). The construction method for the deacetylation point mutation involves mutating the lysine at the corresponding site to arginine. The full sequence of TRPV4 WT, TRPV4 K608R, and GCaMP6s was shown in Supplementary Data S1. Purified PCR products and pBOBI plasmid (digested with BamHI and XhoI) were incubated with Exonuclease III and the relevant buffer (NEW ENGLAND Biolabs, cat. M0206L) on ice for 1 h. Then, we added 1 μL of 0.5 M EDTA to terminate the enzyme digestion reaction and heated the reaction at 65 °C for 10 min to inactivate the enzyme. After cooling, the reaction products were added to competent cells of the STABLE 3 Escherichia coli on ice for 30 min. Thermal activation was performed at 42 °C for 90 s, followed by a 20-min incubation on ice. The strains were subsequently plated on LB solid culture medium containing ampicillin resistance. After incubating at 37 °C for 16 h, the solid culture medium was removed. We selected monoclonal colonies and placed them in 25 mL of ampicillin-resistant liquid LB medium. The liquid LB culture medium was incubated with colonies on a 37 °C shaker for 20 h. Finally, we lysed the bacteria and extracted the plasmid. All mutation proteins were stably expressed in BNL CL.2 with pBOBI for lentivirus packaging. The methods for constructing stable gene expression cell lines were as follows: (1) First, 300 μL of OPTI-MEM (Gibco, cat. 31985070), 1.5 μg of plasmid DNA, and 1.5 μg of Lenti-vector (pMOL: VSV-G: REV = 5:3:2) were thoroughly mixed. (2) Subsequently, 7 μL of Lip2000 Transfection Reagent (Biosharp, cat. BL623B) was added, and the mixture was gently mixed and incubated at room temperature for 30 min. (3) A 10-cm dish of 293T cells at 100% confluency was digested and resuspended in 900 μL of DMEM. The plasmid mixture and cell suspension were then combined and seeded into 6-well plates. Each well contained 700 μL of nutrient-rich DMEM (Dulbecco’s Modified Eagle’s Medium; Gibco, cat. 11960044) supplemented with 10% FBS (Gibco, cat. 10099158), 1× MEM Non-Essential Amino Acids (Sigma, cat. M7145-100mL), 1× GlutaMAX™ (Gibco, cat. 35050061), 1 mM sodium pyruvate (Sigma, cat. S8636-100ML), 100 IU/mL penicillin, and 100 μg/mL streptomycin, along with 160 μL of 293T cell suspension and 300 μL of plasmid mixture. (4) The cells were incubated at 37 °C, and the medium was replaced after 10 h with 2 mL of fresh nutrient-rich DMEM per well. (5) After 48 h, the culture medium was collected, centrifuged, and the supernatant was harvested and stored as lentiviral stock for subsequent cell infection. (6) A 10 mg/mL stock solution of polybrene (Hexadimethrine Bromide, Sigma, cat. H9268) was added to the lentiviral solution at a final dilution of 1:1,000, thoroughly mixed, and centrifuged. (7) BNL-CL2 cells were resuspended and seeded into 6-well plates at a density of 1.67 × 105 cells per well. After incubation for 30 min, the culture medium was replaced with 500 μL of DMEM (Gibco, cat. 11960044) supplemented with 10% FBS (Gibco, cat. 10099158), 100 IU/mL penicillin, 100 μg/mL streptomycin, and 2 mL of lentiviral solution. (8) The plates were centrifuged at 30 °C and 2,500 rpm for 30 min using a horizontal centrifuge. (9) Following centrifugation, the cells were incubated at 37 °C for 48 h. Plasmid expression was confirmed by immunoblotting.
Protein expression in cell lines
This study involved the overexpression of target proteins, including TRPV4-WT, TRPV4-K608R, aldolase A, aldolase B, aldolase C, TRPV4 WT-GCaMP6s, and TRPV4 K608R-GCaMP6s in BNL-CL2 cells, as well as TRPV4 WT-GCaMP6s in AML12 and 293T cells. In Fig. 3a, BNL-CL2 cells overexpressed TRPV4 WT-HA and aldolase A/B/C-Flag. The experimental procedures were as follows: A total of 1.5 μg of pBOBI Homo sapiens TRPV4-WT-HA (N-terminal tag), 1.5 μg of pBOBI Homo sapiens aldolase A/B/C-Flag (N-terminal tag), 1.5 μg of lentiviral packaging plasmids (pMDL:VSVG:REV = 5:3:2), and 300 μL of OPTI-MEM were mixed in a 1.5-mL sterile EP tube by vortexing and briefly centrifuged (3,000× g for 1 min). Subsequently, 7 μL of Lipofectamine 2000 was added, and the mixture was vortexed, centrifuged briefly, and incubated at room temperature for 30 min. 293T cells were cultured to 80–90% confluence in a 10-cm dish. The culture medium was removed, the cells were rinsed with 10 mL of PBS, and 1 mL of trypsin was added. The cells were incubated at 37 °C for approximately 1 min until complete digestion occurred. Digestion was terminated by adding 1 mL of complete medium (high-glucose DMEM supplemented with 10% FBS and penicillin-streptomycin). The cell suspension was transferred to a sterile tube and centrifuged at 1,000× g for 3 min. The cell pellet was resuspended in 1 mL of nutrient-rich DMEM (Gibco, cat. 11960044) supplemented with 10% FBS (Gibco, cat. 10099158), 1× MEM Non-Essential Amino Acids (Sigma, cat. M7145-100mL), 1× GlutaMAX™, 1 mM sodium pyruvate, 100 IU/mL penicillin, and 100 μg/mL streptomycin. The plasmid mixture was then combined with the 293T cell suspension and seeded into 6-well plates. Each well contained 700 μL of nutrient-rich DMEM, 160 μL of 293T cell suspension, and 300 μL of plasmid mixture. The plates were incubated at 37 °C. After 10 h, the medium was replaced with 2 mL of fresh nutrient-rich DMEM per well. After 48 h, the culture supernatant was collected, centrifuged, and the clarified supernatant was harvested as lentiviral stock for subsequent infection of BNL-CL2 cells. Polybrene (10 mg/mL; Hexadimethrine Bromide, Sigma, cat. H9268) was added to the lentiviral solution at a final dilution of 1:1,000, gently mixed, and centrifuged. BNL-CL2 cells were digested, resuspended, and seeded into 6-well plates at a density of 1.67 × 105 cells per well. After 30 min, the medium was replaced with 500 μL of DMEM supplemented with 10% FBS, 100 IU/mL penicillin, 100 μg/mL streptomycin, and 2 mL of lentiviral solution. The plates were centrifuged at 30 °C and 2,500 rpm for 30 min and subsequently incubated at 37 °C for 48 h until 70–80% confluence was reached. The infected cells were transferred to 6-cm dishes and cultured to 70–80% confluence. Half of the cells were lysed for protein extraction, and target protein expression was confirmed by immunoblotting using mouse anti-Flag (1:1,000) and mouse anti-HA (1:1,000) antibodies. Detection of TRPV4 WT-HA and aldolase A/B/C-Flag bands confirmed successful overexpression in BNL-CL2 cells. The remaining infected cells were used for subsequent experiments shown in Fig. 3a. A similar strategy was applied to other target genes, with differences in plasmid constructs and recipient cell types (BNL-CL2, 293T, or AML12).
Cell lysates and immunoprecipitation
Cells were rinsed once with PBS and lysed in ice-cold lysis buffer (40 mM HEPES, pH 7.4, 2 mM EDTA or 5 mM MgCl2, 10 mM pyrophosphate, 10 mM glycerophosphate, and 0.3% CHAPS or 1% Triton X-100) containing one tablet of EDTA-free protease inhibitors (Roche) per 25 mL. The soluble fractions of the lysates were collected by centrifugation at 13,000 rpm for 10 min. For immunoprecipitation, primary antibodies and protein A/G beads were added to the lysates and incubated with rotation overnight at 4 °C. The immunoprecipitated complexes were washed three times with fresh lysis buffer. Proteins were denatured by adding an appropriate volume of 2× SDS sample buffer and boiling for 15 min, resolved by 8–15% SDS–PAGE, and subsequently subjected to immunoblotting analysis. For Flag affinity purification, the Flag M2 affinity gel was washed three times with lysis buffer prior to use. A total of 10 μL of beads (50% slurry) was added to pre-cleared cell lysates and incubated with rotation for 12 h at 4 °C. The beads were then washed three times with lysis buffer. Bound proteins were eluted by addition of 2× SDS sample buffer and boiling for 15 min.
Screening the acetyltransferase is sensitive to low glucose in BNL-CL2
The
Kat2a,
Kat2b,
Hat1,
Atf2,
Cbp,
Ep300 (
P300),
Kat5,
Kat6a,
Kat6b,
Kat7,
Ncoa-1,
Ncoa-2,
Ncoa-3,
Taf1 (
TAFII250),
Gtf3c1, and
Clock shRNAs were designed according to the instructions of the website (
http://biodev.extra.cea.fr/DSIR/DSIR.html). The above shRNA sequences were ligated into the lentivirus-based vector pLVX-shRNA2-zsGreen-PGK-puro (Addgene #197991). The full-length cDNA of human
HDACs (
HDAC1-HDAC11) and
SIRTs (
SIRT1-SIRT7) fragments were ligated into the pBOBI vector for lentivirus packaging, as described previously
81. The shRNA and the ectopic expression of
HDACs and
SIRTs lentivirus were prepared, and BNL-CL2 cells were prepared according to the “Protein expression” procedure in “Cell lines” section in Materials and Methods. All such stable expression cells were subjected to starvation for 4 h, after which the p-S6K were determined via immunoblotting. The rapamycin (100 nM, administered for 1 h) was utilized as a control to inhibit S6K phosphorylation. The phosphorylation of AMPKα was also detected in these cells. Meanwhile, 20 10-cm dishes of AML12, sh
Ep300, OE-SIRT2, and OE-SIRT7 stable expression BNL-CL2 cells were subjected to lysis in 500 μL of lysis buffer, respectively. Following sonication, all samples were subjected to centrifugation at 12,000 rpm, after which an appropriate volume of 5× SDS sample buffer was added, subsequently followed by boiling at 100 °C for 15 min. Then, samples were subjected to SDS–PAGE and staining with R250, and the gel was cut to include the TRPV4 band and sent for K608 acetylation identification by quantitative mass spectrometry analysis.
Recombinant TRPV4 and in vitro acetylation assay
The human
TRPV4 cDNA was cloned into the vector pGEX-4T-1. The
GST and
GST-TRPV4 were transformed into the
E. coli strain BL21 (DE3) (Cat. EC0114, ThermoFisher Scientific), followed by culturing in LB medium in a shaker at 200 rpm at 37 °C. Protein expression was induced with 0.5 mM IPTG at an OD
600 of 0.6–0.8, followed by incubation for an additional 12 h at 26 °C with shaking at 160 rpm. Cells were harvested by centrifugation and resuspended in ice-cold GST binding buffer (PBS supplemented with 10 mM β-mercaptoethanol and 1% Triton X-100). Cell suspensions were sonicated on ice. The lysates were centrifuged at 2,000×
g for 30 min at 4 °C. The supernatant containing GST-tagged proteins was incubated with Glutathione Sepharose 4 Fast Flow resin at 4 °C overnight. The resin was washed with 100 bead volumes of ice-cold PBS, and GST-tagged proteins were eluted with GST elution buffer (50 mM Tris-HCl, 10 mM reduced glutathione, pH 8.0) at 4 °C. To minimize degradation of full-length TRPV4, a relatively large volume of GST binding buffer was used during cell lysis (e.g., 300 mL buffer for 3,600 mL bacterial culture at OD
600 ≈ 1.0). Sonication was performed at low power (< 25% maximal output) using a VCX 750 sonicator (Sonics) equipped with a 6-mm stepped microtip, with 100 cycles of 3 s pulses separated by 3 s intervals for 50 mL of lysate. Purified proteins were concentrated to approximately 3 mg/mL by ultrafiltration (Millipore, UFC905096) at 4 °C and further subjected to size-exclusion chromatography using a Superdex 200 column (Cytiva) equilibrated with PBS, as described previously
82,83.
To express P300 proteins, the human
P300 cDNA was cloned into the vector pcDNA3.3-N-HA. HEK293T cells cultured in fifteen 10-cm dishes (20 µg plasmid per dish) were transfected using PEI. After 48 h of transfection, the P300-HA tag proteins were purified. The acetylation reaction was performed as described previously
84. Briefly, purified GST-TRPV4 (10 μg) or immunopurified TAG synthetic enzymes (expressed in HEK293T cells) were incubated with HA-P300 (about 1 μg) or GST (about 1 μg) in 40 μL of reaction buffer containing 20 mM Tris-HCl, pH 8.0, 20% glycerol, 100 mM KCl, 1 mM dithiothreitol (DTT), and 0.2 mM EDTA with 100 μM acetyl-CoA. After a 2-h incubation at 37 °C, the reaction was stopped by addition of 10 μL of 5× SDS sample buffer. Samples were then subjected to SDS–PAGE and staining with R250, and the gel was cut to include the GST-TRPV4 band and sent for K608 acetylation identification by quantitative mass spectrometry analysis. Fetal liver tissues fromTRPV4-WT
+-Alb-cre
+ and K608R
+-Alb-cre
+ mice were crushed in lysis buffer (2 μL of lysis buffer per 3 mg of tissue). After sonication, samples were centrifuged at 15,000 rpm, and the relevant 5× SDS sample buffer was added and boiled at 100 °C for 10 min. Then, samples were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and staining with R250, and the TRPV4 band was excised and submitted for K608 acetylation identification via quantitative mass spectrometry analysis.
Qualitative mass spectrometry analyses
BNL-CL2 cells stably expressing HA-tagged TRPV4 (80% confluence; fifty 10-cm dishes per experiment) were lysed in Triton lysis buffer containing 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% (v/v) Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, and protease inhibitor cocktail. Cells were harvested at room temperature, collected by centrifugation at 1,000× g for 2 min, and resuspended in approximately 40 mL of lysis buffer. Cell lysates were sonicated on ice and centrifuged at 10,000× g for 10 min at 4 °C to remove debris. The clarified supernatants were incubated with anti-HA antibody (200 μL) and protein A/G beads (200 μL) at 4 °C overnight with rotation. Beads were collected by centrifugation at 3,000× g for 1 min at 4 °C and washed three times with 30 mL of lysis buffer. Bound proteins were eluted with 2× SDS sample buffer (bromophenol blue–free) and subjected to SDS–PAGE. Gels were stained with 0.25% Coomassie Brilliant Blue R-250 and destained in methanol/acetic acid solution. Gel fragments corresponding to TRPV4 (approximately 85–125 kDa) were excised, subjected to in-gel chymotrypsin digestion, and dried. Peptides were dissolved in 0.1% formic acid and analyzed using a nano-Elute system coupled to a timsTOF Pro mass spectrometer (Bruker) equipped with a CaptiveSpray source. Peptides were separated on a homemade C18 column (35 cm × 75 μm, 1.9 μm particle size, 100 Å pore size) using a linear gradient of 3–35% acetonitrile in 0.1% formic acid at a flow rate of 0.3 μL/min for 60 min. Mass spectrometry data were acquired in PASEF mode and analyzed using PEAKS Studio Xpro software. Data were searched against the human UniProt Reference Proteome database with precursor and fragment mass tolerances of 20 ppm and 0.05 Da, respectively. Semi-specific digestion was allowed, with a maximum of two missed cleavages per peptide.
Quantitative mass spectrometry
Stable overexpression of TRPV4 cell lines was constructed in BNL-CL2 and AML12, and TRPV4 protein was enriched from lysed cells. The specific steps for protein enrichment and preparation of gel samples were elaborated in the “Mass spectrometry analyses” section in Materials and Methods. Samples were subjected to in-gel digestion, dried, and then analyzed using an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific) equipped with an EASY-IC ion source. Peptides were eluted for 120 min with linear gradients of 3–35% acetonitrile in 0.1% formic acid at a flow rate of 300 nL/min. Raw data acquired in data-dependent acquisition mode were processed using Proteome Discoverer 2.5 against the UniProt Mus musculus database to determine peptide m/z, charge state (z), and retention time. Although TRPV4 was of human origin in the overexpression system, the high sequence conservation between human and mouse TRPV4, including complete identity in the region surrounding K608, permitted database matching using the mouse reference proteome. Two TRPV4 peptides were selected as reference peptides for acetylation analysis:
TGTYSIMIQK (+42.01): m/z 592.299 (z = 2), retention time (RT) 55.01 min
GVPNPIDLLESTLYESSVVPGPK: m/z 1206.1390 (z = 2), RT 128.19 min
These peptides were subsequently used as reference transitions for parallel reaction monitoring (PRM) analysis of mouse liver tissues. The 8-week-old mice, D1 newborn, E18.5 fetal liver tissues were collected, homogenized in lysis buffer (2 μL lysis buffer / 3 mg tissues), and followed by sonication and centrifugation at 4 °C for 15 min. Then, the supernatant was extracted and mixed with an equal volume of 2× SDS sample buffer (bromophenol blue free), followed by subjecting it to SDS–PAGE. After staining with 0.25% (m/v) Coomassie Brilliant Blue (Sigma, cat. C.I.42660B0149) dissolved in 45% (v/v) methanol and 5% (v/v) acetic acid in water for 30 min, the gels were decolorized (in 45% (v/v) methanol and 5% (v/v) acetic acid in water) overnight (during this period, it was necessary to replace the decolorizing solution with a fresh one), and the excised gel segments were subjected to in-gel chymotrypsin digestion and then dried. We selected the gel segment between 85 kDa and 125 kDa where TRPV4 was located. Further PRM parameters including an automatic gain control (AGC) of 1 × 10
5, a maximum injection time of 1,000 ms, and a precursor isolation window width of m/z 1. Skyline-daily 21.2.1.424 was used to analyze the PRM data
85. The acetylated amino acid sites of TRPV4 identification in liver tissues were compared and matched with the cell line TRPV4 acetylation peptides.
Measurement of ATP, ADP, and AMP
ATP, ADP, and AMP from cells or liver tissues were analyzed by capillary electrophoresis–mass spectrometry (CE–MS). For each measurement, cells were collected from one 10-cm dish, or 30–100 mg of liver tissue was harvested using a freeze-clamp method. Subsequent procedures were performed as previously described
22,86-89.
Following anesthesia, mice were placed supine and the abdomen was sterilized with 75% ethanol. An inverted T-shaped incision (approximately 1 cm transverse and 0.5 cm longitudinal) was made beneath the sternum. The liver was gently exteriorized without compression and immediately freeze-clamped using liquid nitrogen-precooled forceps. Approximately 50–100 mg of liver tissue was collected and stored in liquid nitrogen. Frozen tissues were homogenized in 1 mL methanol containing internal standard 1 (IS1), followed by addition of 1 mL chloroform and 400 μL water. After vortexing and centrifugation (15,000× g, 15 min, 4 °C), 450 μL of the aqueous phase was collected and filtered through a 5-kDa cutoff filter (12,000× g, 3 h, 4 °C). For cellular adenylate measurement, cells cultured in 10-cm dishes were rinsed with 5% mannitol and immediately quenched in liquid nitrogen. Cells were extracted with 1 mL methanol containing IS1 and scraped. Quality control samples were generated by pooling aliquots of each extract. Filtered extracts were vacuum-dried at 4 °C and reconstituted in 100 μL water containing internal standard 2 (IS2). Samples (20 μL) were subjected to CE–MS analysis using an Agilent 7100 CE system coupled to an Agilent 6545 Q-TOF mass spectrometer. Separation was performed in anion mode at −30 kV for 40 min following capillary preconditioning and equilibration. Sheath liquid was delivered at 10 μL/min to the MS via a flow splitter. Mass spectrometry was conducted in negative ionization mode with reference mass correction. Spectra were acquired over m/z 50–1,100. AMP, ADP, and ATP were quantified in full-scan mode at m/z 346.0558, 426.0221, and 505.9885, respectively. In-source fragmentation-derived overlaps were corrected based on retention time. Total ADP and ATP levels were calculated by summing the corresponding resolved peaks. Retention time alignment was performed using isotope-labeled standards and internal standards (IS1 and IS2).
Measurement of intracellular amino acids
Normal-fed and 24-h–fasted pregnant E18.5 C57BL/6 female mice were anesthetized with isoflurane, and approximately 50 mg of fetal liver tissue was collected for intracellular amino acid analysis by HPLC–MS (SCIEX QTRAP 5500) as previously described
77,90.
a) The fetal livers were ground and lysed in 1 mL of ice-cold methanol. The lysates were then mixed with 1 mL of chloroform and 400 µL of water (containing 4 µg/mL [U-13C]-glutamine as an IS), followed by vortexing for 20 s.
b) After centrifugating samples at 15,000× g for another 15 min at 4 °C, 600 µL of the aqueous phase was collected and lyophilized in a vacuum concentrator at 4 °C for 8 h, and then we dissolved them in 30 µL of 50% (v/v, in water) acetonitrile, followed by loading 20 µL of solution into an injection vial (Cat# 5182-0714, Agilent; with an insert (Cat# HM-1270, Zhejiang Hamag Technology)) equipped with a snap cap (Cat# HM-2076, Zhejiang Hamag Technology).
c) A total of 2 µL of each sample was loaded onto a HILIC column (ZIC-pHILIC, 5 μm, 2.1 × 100 mm, PN: 1.50462.0001, Millipore). The mobile phase A was composed of 15 mM ammonium acetate containing 3 mL/L ammonium hydroxide (> 28%, v/v) in the LC–MS-grade water, and mobile phase B was composed of LC–MS-grade 90% (v/v) acetonitrile in LC–MS-grade water. The mobile phase was operated at a flow rate of 0.2 mL/min.
d) The amino acids were separated with the following HPLC gradient elution program: 95% B for 2 min, decreased to 45% B over 13 min, held at 45% B for 3 min, and then returned to 95% B over 4 min. The mass spectrometer was run on a Turbo V ion source in negative mode with a spray voltage of −4,500 V, source temperature at 550 °C, gas no. 1 at 50 psi, gas no. 2 at 55 psi, and curtain gas at 40 psi.
e) Metabolites were measured using the multiple reaction monitoring (MRM) mode, and declustering potentials and collision energies were optimized using analytical standards. The data were collected using Analyst 1.7.1 software (SCIEX), and the relative amounts of metabolites were analyzed using MultiQuant 3.0.3 software (SCIEX). The protein wet weight of each sample was determined by Bradford assay after dissolving the naturally dried protein sediment with 0.2 M KOH at room temperature.
Measurement of glycolytic intermediates
Changes of glycolytic intermediates in the hepatic cell lines, primary hepatocytes, and fetal livers after glucose starvation were measured as follows:
a) Fetal liver tissues were prepared in the following way: the normal and 24-h-fasted pregnant E18.5 C57BL/6 female mice were anesthetized with isoflurane, and the 50 mg fetal livers were frozen in liquid nitrogen. The sample preparation method has been described in the “Measurement of intracellular amino acids” section in Materials and Methods.
b) The adult and fetal primary hepatocytes were isolated and cultured as in the “Primary hepatocytes” section in Materials and Methods. And the primary hepatocytes, cell lines BNL-CL2 and AML12, were treated with glucose starvation for 2, 4, 8, and 24 h, and then the cells in a 10-cm dish were rinsed with 20 mL of PBS and instantly frozen in liquid nitrogen. The 10 cm cell dish was removed from liquid nitrogen and placed on ice for approximately 1 min. Subsequently, 1 mL of ice-cold methanol was added, and the cells were collected using a cell scraper. The lysates were then mixed with 1 mL of chloroform and 400 µL of water (containing 4 µg/mL [U-13C]-glutamine as an IS), followed by 20 s of vortexing. The subsequent sample preparation steps were consistent with the tissue sample preparation.
c) Glycolytic intermediates were detected by HPLC-MS in the “Measurement of intracellular amino acids” section in Materials and Methods.
Measurement of Acetyl-CoA
Fetal liver tissues and adult liver tissues were prepared as follows: the normal pregnant E18.5 C57BL/6 female mice and 10-week-old C57BL/6 female mice were anesthetized with isoflurane, and the 50 samples of fetal and adult livers were cryopreserved in liquid nitrogen. The sample preparation method has been described in the “Measurement of intracellular amino acids” section in Materials and Methods.
Acetyl-CoA was detected by HPLC-MS in the “Measurement of intracellular amino acids” section in Materials and Methods.
RNA-seq and qRT-PCR
The BNL-CL2 cells were cultured in high-glucose DMEM medium supplemented with 10% FBS, 100 IU penicillin, and 100 mg/mL streptomycin. The mRNAs were isolated using 1 ml of TRIzol™ LS Reagent (Thermo Fisher Scientific, Catalog. 10296010), and the samples were delivered on dry ice to the Shanghai Ouyi Biomedical Technology Co., Ltd. for RNA-seq library construction. The Fastp
91 (v0.23.1) (
https://github.com/OpenGene/fastp) was used to trim the Illumina next-generation sequencing adapters, and low-quality and short-length reads were filtered from the RNA-seq data. The RNA-seq reads were processed using the ENCODE RNA-seq pipeline (
https://github.com/ENCODE-DCC/rna-seq-pipeline). Briefly, cleaned FASTQ reads were aligned to the mm9 reference genome, and unique genome-wide signal coverage tracks were generated using samtools (1.9.0)
92, bedGraphToBigWig, and bedSort UCSC tools (
https://hgdownload.soe.ucsc.edu). Then, the Fragments Per Kilobase of exon model per Million mapped fragments (FPKM) value was calculated using the TopHat-Cufflinks pipeline based on the count number. The DESeq2 R package (v1.36.0)
93 was used to identify differentially expressed genes with a
P value < 0.05 and log
2Fold-change > 1 as cut-offs.
Total RNA of 50 mg of E18.5d fetal liver was extracted with 900 µL of TRIzol™ LS Reagent and subjected to three rounds of freeze/thaw cycles before being homogenised. The homogenate was mixed with 200 µL of chloroform and then left at room temperature for 5 min, followed by centrifugation at 4 °C, 12,000×
g for another 15 min. The upper aqueous layer (approximately 450 µL) was carefully transferred to a new RNase-free tube. To precipitate the RNA, 450 µL of isopropanol was added, and then it was centrifuged at 12,000×
g for 30 min at 4 °C. The resulting pellet was washed twice with 75% ethanol and once with 100% ethanol before being dissolved in 20 µL of DEPC-treated water. The RNA concentration was measured by a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). Approximately 1 µg of RNA was diluted with DEPC-treated water to a final volume of 10 µL, heated to 65 °C for 5 min, and then immediately placed on ice. Reverse transcription was performed using the RT Reagent Kit and gDNA Eraser (Takara, Japan), followed by incubation at 37 °C for 15 min, and then at 98 °C for 5 min in a thermocycler. The primers for
TRPV1-6 described in reference
20 and the primers for internal reference
GAPDH described as the reference
94 were used for qPCR assay. The reverse-transcribed cDNA was quantified using the Maxima SYBR Green/ROX qPCR Master Mix on a LightCycler 480 II System (Roche) with the following program: pre-denaturation at 95 °C for 10 min; denaturation at 95 °C for 10 s, followed by annealing and extension at 65 °C for 30 s for a total of 45 cycles. The parameters for annealing and extending were optimized based on amplification curves, melting curves, and bands from agarose gel electrophoresis, adjusting the annealing temperature for each primer pair accordingly.
Immunofluorescence assays
Immunofluorescence-based imaging of LAMP2 and mTOR
95 was performed as follows: (1) Cell preparation: 0.2 mL of culture medium was added to each well of a 6-well plate, and cells were seeded onto sterile glass coverslips placed in each well. (2) Cell growth: the cells were cultured until they reached 50–80% confluence prior to glucose starvation or drug treatment. (3) Post-treatment: after completing glucose starvation or drug treatment, the culture medium was aspirated. (4) Fixation: cells in each well were fixed with 2 mL of 4% paraformaldehyde for 15 min at room temperature. (5) Permeabilization: after three washes with 2 mL of PBS, cells were incubated with 2 mL of 0.05% Triton X-100 (Sigma, Cat. P9416-100ML) for 15 min at room temperature. (6) Blocking: the cells were incubated with PBS containing 5% bovine serum albumin to block non-specific binding. (7) Primary antibody incubation: primary antibodies were diluted in PBS containing 5% FBS. LAMP2 (Abcam, cat. ab13524) was diluted at 1:120 (120 µL diluent + 1 µL antibody), and mTOR (CST, cat. 2983t) was diluted at 1:80 (80 µL diluent + 1 µL antibody). The cells were incubated with the primary antibodies overnight at 4 °C, protected from light. (8) Secondary antibody incubation: after washing with PBS, the cells were incubated with Alexa Fluor 488- or 594-conjugated secondary antibodies (Invitrogen, Carlsbad, CA, USA) diluted at 1:100 (100 µL of diluent +1 µL of antibody) overnight at 4 °C in the dark. (9) Imaging: The tiled images were acquired by an inverted epifluorescence microscope (Zeiss LSM 780 or LSM 980) with a 63× 1.4 NA oil objective at regular intervals. The exposure time for each channel was kept constant for all slides on a given day.
Immunofluorescence-based imaging of LAMP2 and AXIN was performed similarly, with the following adjustments: (1) Cell growth: cells were cultured in 6-well plates until they reached 50–80% confluence prior to glucose starvation or drug treatment. (2) Primary antibody incubation: the primary antibodies were diluted as follows: LAMP2 (Abcam, Cat. ab13524) was used at a dilution of 1:120 (120 µL diluent + 1 µL antibody), and AXIN (Santa Cruz Biotechnology, Cat. sc-8567) was used at a dilution of 1:20–50 (20–50 µL diluent + 1 µL antibody). Colocalization analysis was quantified using ZEN 3.4 (blue) (Mander’s overlap coefficient) or ImageJ (Pearson’s overlap coefficient) as described in “Image analysis” section in Materials and Methods.
The Ki-67 staining on frozen tissue sections was performed as follows: (1) Sectioning and fixation: the frozen liver sections were prepared and fixed in 4% paraformaldehyde (PFA) at RT for 30 min. (2) Washing and permeabilization: sections were washed 3 times with PBS and then permeabilized with 0.3% Triton X-100 in PBS for 15 min at room temperature (RT). (3) Blocking: block sections were incubated in an appropriate blocking buffer for 1 h at RT (50 mL of block buffer: 5 mL of Triton X-100 (Sigma, cat. P9416-100ML) + 1g BSA in PBS). (4) Primary antibody incubation: The sections were washed once with PBS and then incubated with primary antibody diluted in antibody dilution buffer overnight at 4 °C (50 mL antibody dilution buffer: 5 mL of Triton X-100 (Sigma, cat. P9416-100ML) + 0.5 g BSA in PBS). The Ki-67 (Thermo Fisher Scientific, cat. MA5-14520) was applied at a dilution ratio of 1:200. (5) Secondary antibody incubation: sections were washed one to two times with PBS and then incubated with Alexa Fluor 594-conjugated secondary antibody (Invitrogen, Carlsbad, CA, USA) at a dilution of 1:100 for 1–2 h at RT. (6) Nuclear staining: the sections were washed three times with PBS, followed by nuclear staining with Hoechst (Sigma, cat. B2261-100MG) at a dilution of 1:6,000 for 10 min at RT, after which final washes were performed as needed. Tissue sections were digitally scanned by Zeiss AxioScan7. The positive ratio of Ki-67 was calculated by ImageJ.
Indirectly reflecting v-ATPase activity by detecting lysosomal pH
Firstly, 2 mL of culture medium was added to a 35-mm glass-bottom dish (Cellvis, Cat. D35-20-0-N) or a 4-chamber glass-bottom dish (Cellvis, Cat. D35C4-20-0-N), followed by cell seeding. Glucose starvation or drug treatment experiments were performed when the cells reached 70–80% confluence in the 35-mm glass-bottom dish. Next, the fluorescent dye solution was prepared. The required volume of culture medium was calculated based on 2 mL of dye-containing medium per 35-mm dish. LysoSensor™ Green DND-189 (Thermo Fisher Scientific, Cat. L7535; 1 mM stock solution) was added to the culture medium to achieve a final working concentration of 1 μM. The original culture medium was carefully removed and replaced with 2 mL of culture medium containing 1 μM LysoSensor™ Green DND-189. Cells were incubated for 30 min at 37 °C. Subsequently, the fluorescence intensity of the cells was detected using an inverted epifluorescence microscope (Zeiss LSM 780 or LSM 980). The excitation wavelength was 488 nm. We captured images using Zeiss LSM 780 or LSM980 with a 63× 1.4 NA oil objective at regular intervals. During imaging, we maintained live cells at 37 °C 5% CO
2 in a humidified incubation chamber (ZEISS, Incubator PM S1). We kept the laser intensity and voltage settings constant for slides from the same batch. Due to the variation in fluorescence intensity with prolonged incubation time, it was necessary to capture samples from the same batch as quickly as possible. The activity of v-ATPase, assessed by the fluorescent intensity of Lysosensor dye, positively correlated with the lysosomal acidity
21. Signal intensity was quantified using ZEN 3.4 (blue) as described in “Image analysis” section in Materials and Methods.
TRPV4 activity measurement
The BNL-CL2 cell line expressing the TRPV4-GCaMP6s fusion protein was imaged using an inverted epifluorescence microscope (Zeiss LSM 780 or LSM 980) with an excitation wavelength of 488 nm
20,96. The live cells were maintained at 37 °C with 5% CO
2 in a humidified incubation chamber (ZEISS, Incubator PM S1) during imaging. Images were acquired utilizing a Zeiss LSM 780 equipped with a 63× 1.4 NA oil immersion objective at consistent intervals. To mitigate cellular damage and fluorescence quenching induced by excessive laser exposure, the laser intensity and voltage must be minimized, and the imaging should be conducted expeditiously. Signal intensity was quantified utilizing ZEN 3.4 (blue) as detailed in the “Image analysis” section in Materials and Methods.
Image analysis
Quantification of fluorescence intensity, pixel location, and hepatocyte size was performed using the imaging software ZEN 3.4 (blue) (Mander’s overlap coefficient) and ImageJ (Pearson’s overlap coefficient). The analysis was performed on grayscale images (1,048 × 1,048 pixels) acquired at 63× magnification from cells immunostained with an anti-mTOR antibody. These images were utilized for measuring cellular mTOR and AXIN localization.
Statistical analysis
Statistical analyses were performed using Prism 9 and Prism 10 (GraphPad software). The normality of each group of data was assessed using the following tests, where applicable: Anderson-Darling test, Kolmogorov-Smirnov test, or Shapiro-Wilk test. The homogeneity of variance test uses the F-test. When the data conformed to both homogeneity of variance and normal distribution simultaneously, an unpaired two-tailed Student’s t-test was used to determine the significance between the two groups. If the variances between the groups were unequal, Welch’s correction was applied. For data that did not follow a normal distribution or homogeneity of variance, an unpaired two-tailed Mann-Whitney test was used to determine the significance between the groups. The adjusted mean and standard error of the mean (SEM) were calculated and recorded when the data met the above standards.
For comparisons between multiple groups with one fixed factor, multiple-group comparisons were analyzed using one-way ANOVA, which is similar to two-group comparisons. The Kruskal-Wallis test and a multiple comparisons test based on data characteristics were used for non-normally distributed data. When equal standard deviations and equal sample sizes were assumed, the Tukey test was used. We used the Sidak test when group sizes differed but standard deviations were equal. We used the Dunnett test to compare each treatment group to a single control group with equal standard deviations. For Welch-corrected unequal standard deviations, Dunnett's T3 multiple comparisons test was used. For multiple groups with two fixed factors, we used ordinary two-way ANOVA followed by Tukey or Sidak. In particularly, Gsisser-Greenhouse's correction was used before this analysis, and therefore there is no need to consider the homogeneity of SD. Differences were considered statistically significant when the P value was less than 0.05.
DATA AVAILABILITY
Supplementary information is available for this paper. Correspondence and requests for materials should be addressed to H.-F.H. (huanghefg@hotmail.com). All software and data used are freely available either online through various servers (see Supplementary Data S2). The MS proteomics data have been deposited to the ProteomeXchange Consortium (
http://proteomecentral.proteomexchange.org) through the iProX partner repository with the dataset identifier IPX0009924000. The transcriptome raw data of mice fetal liver cell line BNL-CL2 can be accessed from NCBI SRA BioProject ID: PRJNA1171108 (SRA records will be accessible with the following link after the indicated release date:
https://www.ncbi.nlm.nih.gov/sra/PRJNA1171108). The analysis was performed using standard protocols with previously described computational tools. No custom code was used in this study. Full immunoblots are provided as Supplementary Data S1 as Supplementary information.
The Author(s) 2026. Published by Higher Education Press. This is an Open Access article distributed under the terms of the CC BY license (https://creativecommons.org/licenses/by/4.0/).
