mTOR pathway in human cardiac hypertrophy caused by LEOPARD syndrome: a different role compared with animal models?

LEOPARD Syndrome (LS) is a very rare autosomal dominant disease, and named for its major symptoms of Lentigines, Electrocardiography conduction abnormalities, Ocular hypertelorism, Pulmonic stenosis, Abnormal genitalia, Retardation of growth, and sensorineural Deafness [1, 2]. Since it was first reported by Zeisler and Becker in 1936, only a few hundreds of patients have been reported worldwide. PTPN11 mutation is the most common cause of LS and accounts for approximately 90% of the patients with LS. PTPN11 gene encodes protein tyrosine phosphatase Shp2, which mediates several cellular signaling pathways (Fig. 1). Loss-of-function mutation of Shp2 protein in LS leads to the pleiotropic manifestations aforementioned. So far, at least 9 mutations involving 7 amino acid sites have been reported. These mutations may affect different functions of the Shp2 protein. More interestingly, gain-of-function mutations of PTPN11 also lead to a systemic presentation that overlaps quite much with syndrome caused by loss-of-function mutations. Therefore, understanding the abnormalities of signaling pathways will be vital in the treatment of LS.

Aside from the characteristic features, patients with LS are also commonly complicated by hypertrophic cardiomyopathy (HCM). HCM is an early-onset complication which may affect the survival of the patients with LS [3, 4], thus the prevention of cardiac hypertrophy is vital in the management of LS. In mouse LS model induced by PTPN11 p.Y279C mutation, the activity of mammalian target of rapamycin (mTOR) activity was increased. Rapamycin inhibition of mTOR activity can reverse cardiac hypertrophy in PTPN11-mutated LS mice [5]. This result was also observed in animal studies on p.Q510E mutation. Thus, enhanced activity of mTOR was thought to be responsible for systemic manifestations, including HCM [6, 7].

In a later study, exploratory treatment using rapamycin analog improved heart function in an LS infant with severe cardiac hypertrophy [8]. The cardiac hypertrophy, however, was not reversed and resulted in heart transplant. Moreover, patients with gain-of-function mutations of PTPN11 and attenuated mTOR pathway may also have cardiac hypertrophy, which cannot be explained by the animal studies. This raised the concern that the abnormalities observed in animal models might not be the primary cause of cardiac hypertrophy in human. Evidence from human myocardial samples is warranted to understand the poor treatment effect of rapamycin.

The first LS patient (LS1) was a 20-year-old male. He presented with multiple lentigines in the face, body, and limbs (Fig. 2a). Echocardiography showed biventricular hypertrophy and double outflow tract obstruction (Fig. 2c). Electrocardiogram showed ST-T segment changes and left ventricular hypervoltage without conduction block. The second LS patient (LS2) was a 10-year-old male. He had typical facial features of LS syndrome including multiple lentigines, hypertelorism, flat nasal bridge, and low-set ears (Fig. 2b). Echocardiography showed asymmetric left ventricular hypertrophy and left outflow tract obstruction (Fig. 2d).

Four male patients with HCM who underwent myectomy were included as positive controls. They were a similar age to the two LS patients: 34, 15, 20, and 13 years old. All the patients with HCM had no concomitant cardiac disease such as atrial fibrillation, coronary artery disease, or intrinsic valvular dysfunction. All patients with LS or HCM were using beta-blockers for symptom relief before septal myectomy.

The four patients with HCM were known to have myosin binding protein C, cardiac (MYBPC3) or myosin heavy chain 7 (MYH7) mutations before they were included (Table 1). Whole-exome sequencing found a heterozygous c.A836G (p.Y279C) PTPN11 mutation in LS1 and a heterozygous c.C1403T (p.T468 M) PTPN11 mutation in LS2. Subsequent Sanger sequencing validated both mutations. Further, both mutations had previously been shown as causative for LS by multiple studies [9]. No pathological mutations in other LS-related genes (namely, RAF1 and B-Raf proto-oncogene, serine/threonine kinase [BRAF]) were found in the patients.

As shown in Fig. 3, PI3K was hyperphosphorylated in LS but its downstream effector PDK1 was approximately the same. Akt³⁰⁸ was markedly hypophosphorylated in the HCM group, yet only slightly hypophosphorylated in patients with LS compared with normal controls. This led to subsequent mild hypophosphorylation of mTOR in LS and HCM. Consequently, targets of the mTOR complex (mTORC)-1, S6K and 4E-BP1, were slightly changed. However, S6 was significantly hypophosphorylated in both the LS and HCM groups. Although initially enhanced, the mTORC1 pathway gradually became attenuated during subsequent signal transduction. Akt⁴⁷³, a target site of mTORC2, was hypophosphorylated in the HCM group and slightly hyperactivated in the LS group (Fig. 4). Our results also showed hyperphosphorylated ERK1/2 in the HCM and LS groups (Fig. 5). Moreover, their regulatory molecules, MEK1/2, were also significantly hyperphosphorylated.

The present study is the first and the only study on the human sample with PTPN11-mutated Leopard syndrome. Our findings were different from the results from animal models [5]. mTOR/Akt/S6K pathway was not enhanced in human myocardial samples from LS patients, in contrast to the animal models. Moreover, MEK/Erk was significantly enhanced in the present study but showed attenuation after stimulation in the animal model.

LS is also called Noonan Syndrome with Multiple Lentigines because it overlaps with Noonan Syndrome in multiple phenotypes. Indeed, it also shares the main causative gene (PTPN11) with Noonan Syndrome, although the underlying molecular mechanisms are entirely different. PTPN11 mutations in LS are loss-of-function mutations, while those in Noonan Syndrome are gain-of-function mutations [9‐11]. Interestingly, both types of PTPN11 mutations can lead to HCM.

The two patients with LS in our study manifested typical features of LS. T468 M and Y279C of PTPN11 are the most common mutations in LS [12]. Previous studies have shown that enhanced activity of mTORC1 is a key factor of HCM in LS. Rapamycin can reverse HCM in a mouse model of LS. Moreover, although a later attempt of rapamycin treatment in an infant patient did not change the final outcome, the symptoms were greatly alleviated. Thus, the efficacy of rapamycin treatment in humans remains to be validated. In previous studies, hyperphosphorylated Akt⁴⁷³ and S6K³⁸⁹ were regarded as evidence of elevated mTORC1 activity [5, 6]. According to some studies, however, Akt⁴⁷³ is a substrate of mTORC2, and Akt³⁰⁸ is the true catalytic site of the mTORC1 pathway [13]. Our study showed that Akt³⁰⁸ was hypophosphorylated in hypertrophic myocardial tissue, which might change the initial hyperactive effect of PTPN11 mutation. Downstream signaling molecules, including mTOR and S6K, were slightly hypoactivated due to attenuated Akt³⁰⁸ activity. Additionally, our data showed marked hypophosphorylation of S6 in both sarcomere-HCM and LS. This finding implicates multiple branching downstream factors in the regulation of PTPN11/PI3K/Akt/mTORC1. Activated S6K also has a negative feedback effect on insulin receptor substrate 1, which is an upstream regulator of PI3K. Altogether, hypertrophic pathophysiology may have a greater impact on this signaling pathway than mutation.

Rapamycin was reported to be associated with suppressing various types of cardiac hypertrophy, regardless of primary hyperactivity of mTORC1 [14, 15]. In contrast, RAF1 and BRAF mutations can also lead to LS [1, 16, 17]. These two proteins are components of the RAS/RAF/ERK pathway, which is also regulated by PTPN11. Theoretically, they do not have a significant impact on mTORC1 activity. These findings suggest that PTPN11 mutation may induce LS beyond the mTORC1 pathway, while primary hyperactivation of mTORC1 is not a prerequisite for the hypertrophy-reversing effect of rapamycin. In this observational study, differences between LS1 and LS2 also showed that the functional status of different PTPN11 mutations might be distinct, which raises the concern that determining the dose of rapamycin would be hard in clinical practice.

Basal substrates of the RAS/RAF pathway were also examined. Unexpectedly, MEK and ERK were significantly hyperphosphorylated in both LS and sarcomere-HCM, which is another significant paradox. LS-related PTPN11 mutations reportedly diminish ERK activation, while in our study, hypertrophic myocardial samples presented enhanced ERK activation [12, 18]. The lowest molecules of this pathway showed similar changes in LS and sarcomere-HCM. Consequently, branching regulations are strong enough to reverse the primary catalytic alteration. Considering the same changes in a different genetic background, the role of the RAS/RAF pathway in hypertrophy cannot be ignored. A previous study reported that calcium signaling dysregulation might play a role in both sarcomere and non-sarcomere HCM. These intricate signaling pathways are widely dysregulated in myocardial hypertrophy and can further affect each other. As a result, effector molecules might not just behave as theories.

Although gene mutation is the primary cause of the LS, straight-forward analysis of the downstream signaling may be misleading. PTPN11 gene and downstream pathways widely involve many metabolic activities and are also regulated by these metabolic molecules. The homeostasis of the signaling network is a balance between gene function and environmental stimulation. Functional studies on model animals cannot reflect the real pathogenic mechanism in human beings. Using metabolomic methodology to unveil what is happening in the cells may help our understanding of the pathophysiology in LS [19].

In patients with LS, the mTOR signaling pathway shows similar activity to HCM and is attenuated compared with normal controls. Rapamycin treatment for LS should be cautious, especially for patients with LS who have developed cardiac hypertrophy.

Two patients were presumptively diagnosed with LS based on their clinical manifestation and family history. Subsequent gene sequencing confirmed the diagnosis. They underwent septal myectomy because of left ventricular outflow tract obstruction. Myocardial samples from four gender- and age-matched obstructive HCM patients with pathogenic sarcomere mutation were included as HCM controls. Myocardial specimens were resected during surgical operation, and immediately cut and preserved in liquid nitrogen. Myocardial samples from four healthy donors who had accidental deaths were included as healthy controls. Left ventricular muscle was used for protein analysis.

DNA was isolated from myocardial tissues of the presumptive patients with LS, and whole-exome sequencing was performed to identify causing mutation of the disease. Pathogenic mutations were subsequently validated by Sanger sequencing. The sarcomere mutations of HCM controls were detected by panel sequencing as described previously [20] and confirmed by Sanger sequencing.

Myocardial samples were lysed on ice in lysis buffer (CWBiotech, Beijing, China) containing protease inhibitor (Roche, Basel, Switzerland) and protein phosphatase inhibitor (Roche). Extracted proteins were preserved at − 80 °C for use after quantification and denaturation. Prepared protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Samples were then blotted onto polyvinylidene fluoride membranes with a bore diameter of 0.22 μm (Merck Millipore, Darmstadt, Germany). Membranes were blocked with 5% nonfat milk in Tris-buffered-saline with Tween-20 at room temperature. Next, the membranes were incubated overnight at 4 °C with primary antibodies. Subsequently, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. Specific bands were detected using the SuperSignal West Femto Maximum Sensitivity Substrate (Pierce, Rockford, IL, USA). Glyceraldehyde 3-phosphate dehydrogenase was used as the reference protein. The following protein and phosphorylation sites were examined with the corresponding primary antibodies: total phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K) p85, phospho-PI3K p85⁴⁵⁸, total phosphoinositide-dependent kinase 1 (PDK), phospho-PDK²⁴¹, total Akt, phospho-Akt³⁰⁸, phospho-Akt⁴⁷³, total tuberous sclerosis complex 2 (TSC2), phospho-TSC2¹⁴⁶², total mTOR, phospho-mTOR²⁴⁴⁸, total ribosomal protein S6 kinase (S6K), phospho-S6K³⁸⁹, total ribosomal protein S6 (S6), phospho-S6^235/236, total eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1), phospho-4E-BP1^37/46, total mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) kinase (MEK1/2), phospho-MEK1/2^217/221, total ERK1/2, phospho-ERK1/2^202/204, and glyceraldehyde 3-phosphate dehydrogenase. All antibodies were from Cell Signaling Technology, Danvers, MA, USA. The pathway examined is shown in Fig. 1 [13, 21].