Results
Effect of high glucose on RhoA and Drp1 activity
We previously demonstrated that phosphorylation of Drp1 was increased at Ser637 in the presence of high glucose and was decreased upon formoterol treatment. To determine whether this increase in phosphorylation also led to increased Drp1 activity and to determine the involvement of RhoA, a GTP-pulldown assay was used to determine Drp1 and RhoA activity. RPTC exposed to high glucose had increased GTP-bound RhoA (Fig 1A) and Drp1 (Fig 1B) compared to 0mM glucose and 17mM mannitol controls. Formoterol treatment reduced both RhoA and Drp1 activity to control levels.
To determine the involvement of the RhoA/ROCK1/Drp1 signaling pathway, CCG-1423, a pharmacological inhibitor of RhoA was used. In RPTC exposed to high glucose, RhoA activity increased in the presence of glucose. Treatment with either formoterol or CCG-1423 prevented the increase in RhoA activity (Fig 2A). In addition, treatment with CCG-1423 prevented glucose-induced activity of Drp1 (Fig 2B). Seahorse XF96 analysis was used to determine the effect of RhoA inhibition on FCCP-OCR, a marker of MB and maximal electron transport chain activity. Glucose exposure did not change basal respiration in any treatment group (Fig 2D). However, after FCCP injection, maximal respiration decreased in RPTC grown in the presence of high glucose compared to controls, and RhoA inhibition restored maximal respiration (Fig 2E).
Effect of ROCK1 inhibition on Drp1 activity and mitochondrial function
To determine the involvement of ROCK1, a downstream kinase of RhoA, we treated RPTC with the ROCK1 inhibitor Y-27632 and measured Drp1 activity. RPTC treated with either formoterol or Y-27632 restored Drp1 activity to control levels (Fig 2C). Similarly to what was observed after RhoA inhibition, inhibition of ROCK1 had no effect on basal-OCR under any conditions (Fig 2F) and restored maximal mitochondrial respiration to the same level as controls (Fig 2G). The specific Drp1 inhibitor Mdivi-1 was used to determine whether the effect of glucose on mitochondrial respiration and MB was a result of altered mitochondrial fission. Glucose did not change basal respiration, nor did treatment with formoterol or Mdivi-1 (Fig 3A). Formoterol increased FCCP-OCR in 0mM glucose and mannitol controls compared to vehicle, whereas Mdivi-1 did not (Fig 3B). However, RPTC grown in high glucose, both formoterol and Mdivi-1 treatment restored FCCP-OCR.
Effect of formoterol activation of the β2-adrenoceptor on RhoA signaling
To determine if modulation of the RhoA/ROCK1/Drp1 signaling pathway was mediated through the β2-adrenoceptor, carvedilol, a β2-adrenoceptor antagonist with a high affinity for the β2-adrenoceptor (KD=-9.40±0.08) was used to block the effect of formoterol on the receptor. While high glucose increased RhoA and Drp1 activity and formoterol restored both RhoA and Drp1 activity, carvedilol blocked this effect (Fig 4A and B). We previously discovered that formoterol activates the Gβγ subunit of the β2-adrenoceptor to induce MB (Cameron et al., 2017). Therefore, to determine whether formoterol works through Gβγ signaling to regulate Drp1. RPTC were grown in the presence or absence of glucose with the Gβγ inhibitor gallein. Gallein blocked the ability of formoterol to restore both RhoA and Drp1 activity (Fig 5A and B). Because RhoA is active when bound to GTP and this process is regulated by GDP-GTP exchange via RhoGEFs. RhoA was immunoprecipitated to determine potential interactions with RhoGEF proteins that may regulate RhoA activity. Among these RhoGEFs, we found that the interaction between RhoA and p114RhoGEF was increased in the presence of glucose (Fig 6). Treatment with formoterol reduced the interaction between RhoA and p114RhoGEF to control levels. Interestingly, co-treatment with gallein blocked the ability of formoterol to prevent this interaction.
Effect of high glucose and formoterol on Mfn1 signaling
We previously discovered that in addition to alteration of mitochondrial fission via Drp1, expression of the mitochondrial fusion protein Mfn1 was decreased in models of DKD. To investigate the signaling mechanism responsible for altered Mfn1 expression, RPTC were treated with the Raf inhibitor PLX4032. Since Mfn1 is also a GTPase and its activity can be detected by measuring GTP-bound protein, the same GTP pulldown assay to measure Mfn1 activity was used. In RPTC treated with vehicle in the presence of high glucose, Mfn1 activity was decreased compared to 0mM glucose and mannitol controls (Fig 7A). Cells treated with either formoterol or PLX4032 restored Mfn1 activity to control levels. Since Raf phosphorylates MEK1/2, the MEK1/2 inhibitor GSK11202 was used to determine the role of MEK1/2 and ERK1/2 in Mfn1 activation. Treatment with GSK11202 blocked ERK1/2 phosphorylation in RPTC grown in 0mM glucose, 17mM mannitol and 17mM glucose, while treatment with either vehicle or formoterol had no effect (Fig 7B). In addition, both GSK11202 and formoterol treatment prevented the decrease in Mfn1 activity compared to 0mM glucose and 17mM mannitol controls (Fig 7C).
To determine whether the Raf/MEK1/2/ERK1/2/Mfn1 pathway is also regulated by formoterol activation of Gβγ, gallein was used to block Gβγ signaling and subsequently measured Mfn1 activity. Formoterol, gallein or co-treatment of formoterol+gallein had no statistical effect on GTP-bound Mfn1 in either 0mM glucose or 17mM mannitol control groups (Fig 8). In the presence of glucose, GTP-Mfn1 was significantly reduced compared to controls. While formoterol prevented the decrease in Mfn1 activity, co-treatment with gallein blocked this effect. Importantly, phosphorylation of Akt after formoterol or gallein treatment was unchanged in RPTC in all groups (Fig 9).