Pathogenic Role of mTORC1 and mTORC2 in Pulmonary Hypertension

Visual Abstract

The phosphoinositide 3-kinase (PI3K)/ AKT/mammalian target of rapamycin (mTOR) pathway, one of the critical signaling cascades involved in cell proliferation (16), can be activated by various growth factors and mitogenic cytokines (17)(18)(19). We and other investigators have shown that the PI3K/Akt1/ mTOR signaling pathway plays an important role in the regulation of PASMC proliferation and the development of PH (16,20). Activation of the PI3K/AKT/mTOR pathway in PASMCs is through various stimuli such as platelet-derived growth factor (PDGF), endothelin-1 (21,22), stress, and hypoxia (23). protective effect on experimental PH in PTEN transgenic mice (16).
The downstream signaling protein, mTOR, in the PI3K/AKT/mTOR pathway is a serine/threonine kinase that belongs to the PI3K-related kinase family (26). We previously reported that conditional and inducible KO of mTOR in smooth muscle cells almost completely inhibited the development of PH in mice (16). These data provide compelling evidence that the PI3K/AKT1/ mTOR signaling pathway in PASMC plays an important role in the development of PH. Specifically targeting signaling proteins and kinases in the PI3K/AKT1/ mTOR cascade may help develop novel therapeutic approaches for idiopathic and associated PAH, as well as PH associated with lung diseases and hypoxia. mTOR is a downstream signaling protein and a serine/threonine kinase of AKT1. mTOR is also a major kinase present in 2 functionally distinct complexes: the mTOR complex 1 (mTORC1) and the mTOR complex 2 (mTORC2) (27). mTORC1 is composed of mTOR, Raptor (regulatory associated protein of mammalian target of rapamycin), Pras40, GbL, and DEPTOR, and is inhibited by rapamycin and KU 0063794; mTORC2 is composed of mTOR, Rictor (rapamycin insensitive companion of mammalian target of rapamycin), GbL, Sin1, PRR5/Protor-1, and DEPTOR (26), and is inhibited by KU 0063794 (28). The individual protein complexes of mTORC have different upstream and downstream regulators (26). mTORC1 is a master growth regulator that promotes cell proliferation in response to growth factors, extracellular nutrients, and amino acids; mTORC2 promotes cell survival by activating AKT, regulates cytoskeletal dynamics by activating protein kinase C alpha, and controls ion transport and cell growth via serum/glucocorticoid-inducible kinase 1 phosphorylation. Global deletion of mTOR would disrupt the function of both mTORC1 and mTORC2.
A relatively new topic of research in the field of pulmonary vascular disease is to understand the individual or differential roles of mTORC1 and mTORC2 in PASMC proliferation and the development of PAH/PH. The mTORC1 and mTORC2 also differ in their sensitivity to rapamycin; that is, short-term treatment with rapamycin inhibits mTORC1, but long-term treatment with rapamycin can inhibit both mTORC1 and mTORC2 (29).
Significant research is being conducted in understanding the role of mTOR, as a common component in both mTORC1 and mTORC2, in the development of hypoxia-induced PH by promoting PASMC proliferation (30,31). The aim of the present study was to examine whether mTORC1 and mTORC2 potentially play a differential role in the development of PH. We generated the following: 1) smooth muscle (SM)-specific Raptor KO mice (Raptor SMÀ/À ) to inhibit mTORC1 function in PASMCs; and 2) SM-specific Rictor KO mice (Rictor SMÀ/À ) to inhibit mTORC2 function in PASMCs. We then conducted a series of experiments in wild-type (WT) Raptor SMÀ/À and Rictor SMÀ/À mice using combined techniques of in vitro cell and molecular biology, and in vivo hemodynamic measurement in intact mice, to define whether mTORC1 and mTORC2 are differentially involved in the development of PH and whether inhibition of mTORC1 and mTORC2 exerts the same therapeutic effect on experimental PH.

RESULTS
We first conducted in vivo experiments using WT and various KO mice to examine whether SM-specific KO of mTOR (mTOR SMÀ/À ), Raptor (Raptor SMÀ/À ), and Rictor (Rictor SMÀ/À ) exerted protective effects on experimental PH. In vitro Western blot experiments were then conducted by using PA isolated from WT and KO mice to examine whether functional disruption of mTORC1 in mTOR SMÀ/À and Raptor SMÀ/À mice or mTORC2 in mTOR SMÀ/À and Rictor SMÀ/À mice affects protein expression of platelet-derived growth factor receptor (PDGFR) a and PDGFRb in PASMCs.  ; red), and mammalian target of rapamycin (mTOR; green) in the cross-section of small pulmonary artery (PA) in lung tissues from wild-type (WT) (mTOR-Oil) and mTOR SMÀ/À (mTOR-Tam) mice (a). Summarized data (mean AE SE; n ¼ 5 in each group) for DAPI, SMA, and mTOR fluorescence intensity are shown in panels b. It is noted that the mTOR (green) expression is almost abolished in the SMA-positive PA wall in mTOR-Tam mice but preserved in the mTOR-Oil mice. Student's t-test (DAPI and SMA level) and Welch's t-test (mTOR level), **p < 0.01 and ***p < 0.001 versus mTOR-Oil. (C) Representative record of right ventricular pressure (RVP) in WT and mTOR SMÀ/À mice exposed to normoxia (room air, 21% oxygen) and hypoxia (10%  To induce the KO of mTOR, we treated the Cre þ m-TOR F/F mice with tamoxifen 5 times consecutively and waited for 1 to 2 weeks prior to exposing the mice to hypoxia (for 3 weeks) (Figure 1Ab). The control mice were treated with vehicle (Oil). Immunohistochemical staining of the PA in lung tissues showed that mTOR was expressed in all cell types, including PASMCs, in Cre þ mTOR F/F mice treated with the vehicle (Oil), whereas the mTOR expression was hardly detected in PASMCs or in PA in mTOR SMÀ/À mice after treatment with tamoxifen ( Figure 1B).
Deletion of mTOR would disrupt the function of both mTORC1 and mTORC2 (32). In this study, we repeated the experiments ( Figure 1C) showing that SMspecific deletion of mTOR significantly inhibited the development of HPH. Chronic hypoxia-mediated increases in right ventricular systolic pressure (RVSP) (Figures 1Ca and 1Cb) in Cre þ /mTOR F/F mice treated with vehicle Oil (WT) was significantly attenuated compared with Cre þ /mTOR F/F mice treated with tamoxifen (mTOR SMÀ/À ). The reduced RVSP was associated with the down-regulation of mTOR protein expression level in PASMCs (Figure 1Bb).
In addition to the change in RVSP, the hypoxiainduced pulmonary vascular wall thickening, determined by the PA wall thickness, was also significantly inhibited in mTOR SMÀ/À mice compared with WT mice ( Figure 1D). Hypoxia also resulted in polycythemia, indicated by the increased number of red blood cells, hemoglobin concentration (grams per deciliter) and hematocrit percentage ( Figure 1E). It is noted that the hypoxia-induced polycythemia effect on red blood cell count and hemoglobin was also significantly inhibited in mTOR SMÀ/À mice compared with WT mice. These data, consistent with our previous study (16), indicate that SM-specific deletion of mTOR, a serine/threonine kinase that is pivotal for the function of both mTORC1 and mTORC2, significantly inhibits the development of pulmonary arterial remodeling and HPH. The next set of experiments was designed to investigate whether mTORC1 and mTORC2 complexes are differentially involved in HPH. In contrast to mTOR SMÀ/À and Raptor SMÀ/À mice, the Rictor SMÀ/À mice exhibited a slight, but not sta- Rictor SMÀ/À mice. It was noted that the nonsignificant difference in RVSP and the Fulton index between hypoxic WT mice and hypoxic Rictor SMÀ/À mice was somehow related to a slight, but not statistically sig-

nificant, increase in basal level of RVSP and RVH in
Rictor SMÀ/À mice (Figures 3Cb and 3Cc). These results indicate the following: 1) that SM-specific deletion of showing the peak value of RVSP (b) (Kruskal-Wallis test, p < 0.001) and the Fulton index Raptor F/F ) and Raptor SMÀ/À (Tam-Cre þ /Raptor F/F ) mice exposed to normoxia (room air, 21% oxygen) and hypoxia (10% oxygen for 3 weeks). Dunn test, ***p < 0.001, **p < 0.01 versus Normoxia-WT; #p < 0.05 versus Hypoxia WT. The numbers of experiments (n) for each group are indicated in each bar. Abbreviations as in Figure 1.    Figure 4A) and then examined whether intraperitoneal injection of imatinib, a tyrosine kinase inhibitor that inhibits PDGFRs with high affinity, had a reversal effect on established PH in Rictor SMÀ/À mice. As shown in Figure 4B, Rictor SMÀ/À mice exhibited a significantly higher RVSP (Figures 4Ba and   4Bb) and Fulton index (Figure 4Bc    Abbreviations as in Figure 1. The expression level of PDGFRa and PDGFRb were greater in PA isolated from Rictor SMÀ/À mice than in PA isolated from WT mice (Figures 6Aa, b and 6Ba, b).
The KO of Rictor would lead to inhibition of mTORC2.
Indeed, pAKT (at S473 but not at T308), a major downstream signaling protein of mTORC2, was significantly decreased in the PA isolated from Rictor SMÀ/À mice compared with the PA isolated from WT mice ( Figures 6A and 6B). The total AKT protein expression level, however, did not differ in isolated PA between WT and Rictor SMÀ/À mice (Figures 6Aa   and 6Ba).   (Figures 7Ec and 7Ed). The low dose of imatinib (20 mg/kg) had no effect on RVSP (Figures 7Ea and 7Eb) but slightly (with statistical significance) inhibited hypoxia/Sugen -mediated increases in the Fulton index ( Figure 7Ec) and the ratio of right ventricular weight to body weight (Figure 7Ed). However, combination of the low doses of rapamycin and imatinib resulted in w60% inhibition of the hypoxia/Sugenmediated increase in RVSP (Figures 7Ea and 7Eb).
These data suggest that blockade of PDGFRs using  ( Figures 8Bb and 8Bc), whereas rapamycin did not significantly decrease pAKT at S473 ( Figure 7D) and actually decreased the nuclear FOXO3a (Figure 8Bc).
Taken together, these data indicate that inhibition of mTORC2 increases the activity of FOXO3a in PASMCs, which may be the mechanism in which PDGFR is negatively regulated by mTORC2.    (29). These findings suggest cooperative mechanisms between the signals from mTORC1 and mTORC2, and it becomes essential to understand the differential role played by mTORC1 and mTORC2 in stimulating cell proliferation and survival.
Many studies indicate that mTORC1 and mTORC2 function differently in regulating gene expression, cell proliferation, and growth (44). KO of mTOR would disrupt the function of both mTORC1 and mTORC2 (60). The goal of the present study was to examine the potential divergent or differential role of In consistent with our previously published data using mTOR À/À mice (16) Tang et al.