Introduction
Tuberous sclerosis complex (TSC) is a genetic disorder characterized by the presence of benign hamartomas in any organ system, with highly variable, unpredictable, and potentially devastating neurological outcomes [
1,
2]. TSC is the second most common identified neurocutaneous disorder with an estimated incidence of 1:6000, affecting more than 1 million individuals worldwide [
3,
4]. The majority of patients have central nervous system involvement that manifests as structural brain abnormalities, epilepsy, and cognitive, behavioral, and psychiatric deficits including autism spectrum disorder (ASD) [
5,
6]. All types of seizures are seen, often in combination. Half of individuals have normal intelligence, but almost none are free from neuropsychiatric problems [
7]. Importantly, ASD is diagnosed in approximately 40–60 % of patients with TSC, and TSC accounts for 3–4 % of ASD [
8‐
10].
TSC is caused by mutations in either of two tumor suppressor genes,
TSC1 or
TSC2, encoding hamartin and tuberin, respectively [
11]. TSC is inherited in an autosomal dominant pattern and haploinsufficiency of
TSC causes neurological phenotypes including learning disability and social deficits [
10,
11].
De novo mutations account for approximately 80 % of TSC cases.
TSC2 mutations are four times as common as
TSC1 mutations among
de novo cases, whereas the prevalence of
TSC1 and
TSC2 mutations is approximately equal among familial TSC cases [
12].
TSC1 and
TSC2 mutations lead to essentially identical phenotypic manifestations, although there have been some suggestions that the
TSC2 phenotype is typically more severe [
13,
14]. The TSC1 and TSC2 proteins act as a heterodimer to suppress mammalian target of rapamycin (mTOR), a serine/threonine protein kinase that regulates cell growth and division [
5,
15]. Loss of either
TSC1 or
TSC2, followed by a “second hit” of the remaining functional allele thereby preventing formation of the heterodimeric complex, causes loss of regulatory control over mTOR and leads to overactive cell growth and proliferation [
1,
16]. Thus, at the cellular level, loss of
TSC1 or
TSC2 results in upregulation of the mTOR pathway [
10,
17]. The molecular understanding of the TSC pathophysiology has opened up possibilities for molecular targeted treatments of the neuropsychiatric phenotype in TSC using mTOR inhibitors such as rapamycin [
16]. Notably, rapamycin treatments have been shown to successfully reverse the deficits in behavior and synaptic plasticity in rodent models of TSC [
10,
18‐
20].
Recently, the mTOR inhibitors everolimus and sirolimus have been shown to exhibit efficacy for the treatment of several manifestations of TSC such as subependymal giant cell astrocytomas (SEGA), seizures, renal angiomyolipomas, lymphangioleiomyomatosis, and facial angiofibroma lesions in patient with TSC [
21‐
24]. Moreover, human and animal studies suggest that mTOR inhibitors improve deficits of sociability, learning and neurodevelopment in TSC mouse models and patients with TSC [
18,
25,
26]. On the basis of these findings, although some clinical trials have been completed or initiated to test whether everolimus treatment might improve neurocognition, features of autism, and the neuropsychological deficits in children with TSC (clinicaltrials.gov study ID: NCT01289912, NCT01730209), in the present study we present a case of a family with a novel
TSC2 mutation in which the behavioral phenotypes of a 3-year-old boy with TSC accompanied by severe autism could be dramatically improved by everolimus treatment.
Discussion
In the present study, we describe a case of a Korean family in which three of the members have been diagnosed with TSC. Among the affected members, only a boy (III-2 in Fig.
2a) showed diverse behavioral and cognitive deficits including autistic phenotypes. Whole exome analysis revealed a novel deletion mutant in the
TSC2 gene (
TSC2 c.700–701 del).
We found that the expression of the newly identified mutant TSC2 protein enhanced the activation of the mTOR signaling pathway in HEK293T cells, suggesting that the novel mutation represents a loss-of-function mutation as are other
TSC mutations associated with TSC [
28]. In our biochemical analyses, we found that the protein expression level of the deletion mutant was significantly decreased (Fig.
3a,
b). The reduced TSC2 c.700–701 del mutant expression level might be caused by nonsense-mediated mRNA decay as a quality-control mechanism [
35‐
37]. In addition, it has been reported that the TSC1-TSC2 interaction is important for stability of both TSC1 and TSC2 [
38,
39]. In this patient, it is hypothesized that the TSC1 protein would be destabilized since the mutation in TSC2 is located within the essential region for interacting with TSC1 (amino acids 1–900) [
31,
40] (Fig.
3a,
c). Furthermore, the loss of the C-terminal GTPase activating protein (GAP) domain of TSC2 would also be predicted to disrupt the action of the TSC1-TSC2 complex on the GTPase Ras homolog expressed in brain (RHEB), thus activating RHEB–GTP-dependent stimulation of the mammalian target of rapamycin complex 1 (mTORC1) [
41]. Therefore, one of the downstream mTORC1 targets, S6K T389 phosphorylation, would be elevated (Fig.
3a,
d). Additionally, we examined the effect of everolimus treatment on elevated S6K T389 phosphorylation induced by
TSC2 c.700–701 del mutation and found that everolimus could reduce the activation of the mTOR signaling pathway (Fig.
4a,
b).
The efficacy of mTOR inhibitors for treating TSC-associated phenotypes has been demonstrated in multiple animal models. For example, Ehninger and colleagues showed that rapamycin treatment reversed the deficits in learning and in hippocampal synaptic plasticity in
Tsc2
+/− mice [
18]. Rapamycin or everolimus (RAD001, 40-O-(2-hydroxyethyl)-rapamycin) treatment rescued lethality, brain enlargement, and hyperactivity in
Tsc1 conditional knockout mice [
18,
19]. Notably, deleting
Tsc1 in mouse cerebellar Purkinje cells resulted in autistic-like behaviors, which can also be reversed by rapamycin treatment [
20]. In addition, deficits in social interaction in
Tsc1
+/− and
Tsc2
+/− mice were also reversed by rapamycin treatment [
26]. These animal studies support the hypothesis that mTOR activation is responsible for the ASD-associated phenotypes in TSC [
16,
26,
42] and thereby mTOR inhibition might be an effective treatment strategy for ASD symptoms in TSC patients [
10,
25].
In our study, everolimus treatment resulted in a rapid and marked reduction of behavioral deficits and improved cognition, attention, social interaction, and language development in the patient. As indicated by changes in the ATEC subscale scores, the classification of severe autism with scores in the 90th percentile was modified to that of semi-independent autism with scores in the 30th percentile level. The total score dropped by 83 points, which is considered to be a remarkable improvement for an autism treatment. In addition to the improvement of autistic features, everolimus showed marked effectiveness in mediating the intractable seizures of TSC, such that the seizure frequency and the median seizure duration decreased by over 90 % in the patient. However, we cannot insist that everolimus would have therapeutic effect specifically on ASD or epilepsy in patients with TSC induced by TSC2 c.700–701 del mutation because grandmother and father carrying TSC2 c.700–701 del mutation has normal intelligence without epilepsy or neuropsychiatric symptoms and we could not suggest an evidence of genotype (TSC2 c.700–701 del mutation)-phenotype correlation in TSC patient. Therefore, we consider that individuals affected by TSC2 c.700–701 del mutation could exhibit a high variability in clinical findings and further studies are needed to identify a key mechanism underlying the therapeutic effect of everolimus on TSC symptoms with ASD.
Nevertheless, early recognition of ASD in patients with TSC and proper management with everolimus might give a life-enhancing effect on the long-term outcome of the disorder. Consistent with our findings, Ishii and colleagues recently reported that everolimus treatment improved the irritability, stereotypic behavior, inappropriate speech, and social behavior in a 27 year-old female patient with TSC [
43]. In addition, Wheless showed that everolimus reduced the seizure frequency in a 13-year-old girl with TSC-associated epilepsy after 1.5 years of treatment [
44]. Kruger and colleagues also demonstrated the antiepileptic effect of everolimus in the majority of TSC patients [
45]. Focal cortical malformations such as cortical tuber are highly associated with epileptogenesis and epilepsy related to cortical tuber is often refractory to antiepileptics [
46]. mTOR inhibitor treatment also reduces the size of cortical tubers in TSC patients [
47].
Together with our results, these studies strongly support that the inhibition of mTOR signaling represents a valid treatment strategy for the neurological and psychiatric manifestations associated with TSC. As additional clinical trials to test the efficacy of everolimus for treating the psychiatric symptoms associated with TSC have been completed or launched [
10,
17], it will likely soon be determined whether the use of everolimus or other mTOR inhibitors will be approved to treat the features of ASD associated with TSC in the clinic.
In summary, we identified a novel small deletion mutation in TSC2 associated with severe TSC in a Korean family that enhances the activation of mTOR signaling in vitro. Moreover, everolimus treatment showed not only reduction in SEGA size, but improved behavioral deficits including autistic phenotypes and seizures in the patient.
Methods
Participants
The study participants consisted of a three-generation family whose members were diagnosed with TSC. TSC showed autosomal dominant inheritance throughout the family history; the grandmother was the first affected family member. Six individuals including the grandmother, father, mother, second and third sons, and daughter participated in this study.
Diagnostic evaluation
Diagnosis of TSC
Brain MRI, EEG, complete ophthalmologic evaluation including dilated fundoscopy, detailed dental examination, careful skin examination with a Wood’s lamp, ultrasonography and ECG of the heart, HRCT of the chest, blood pressure, abdominal MRI, and GFR tests were performed. In addition to genetic diagnosis (described below), clinical diagnosis was made according to the “Tuberous Sclerosis Complex Diagnostic Criteria Update: Recommendations of the 2012 International Tuberous Sclerosis Complex Consensus Conference” [
48].
Measures and procedure
A baseline psychomotor developmental evaluation including the DDST-II, WPPSI-III, CARS, SMS, and speech and language evaluations was performed prior to everolimus treatment. The ADOS and ADI-R were conducted by an experienced examiner. For ADOS, module 1 for 31 months and older children with pre-verbal/single words was applied. The ATEC was also implemented for the severity evaluation of autistic features prior to and following completion of the intervention.
Intervention
Everolimus treatment was initiated, at 5 mg/m2/day, administered once daily in the morning and rounded to the nearest 2.5 mg/dose. A serum everolimus level was obtained at every 2 weeks, and the treatment dose was adjusted to obtain a target range between 5 and 15 ng/ml. No further adjustments were made after the stable target range was obtained, unless a severe side effect or considerable weight gain was noticed. Safety and efficacy of treatment was observed for 1 year.
Whole exome sequencing and variant calling
Whole blood was obtained from the family members (the grandmother, father, mother, elder brother, and little sister of the patient) after informed consent for the protocol (KNUH 2013-07-011-004) guided by the KNUH IRB was obtained. Genomic DNA was extracted from the blood from each subject and about 3 μg genomic DNA from each family member was subjected to library preparation and exome capture following the Agilent SureSelect Human All Exon v4 Illumina Paired-End Sequencing Library Prep Protocol (Agilent Technologies, Santa Clara, CA, USA). The prepared sequencing library was used to perform sequencing on an Illumina HiSeq-2000 system (San Diego, CA, USA) as a 100 bp paired-end run. In each sample, about 51–65 million reads were generated. The raw sequencing reads were checked and trimmed using the sickle program (version 1.33) to ensure the quality of the raw reads. Processed reads were mapped to the reference human genome sequence GRCh37 using the Burrow-Wheeler Aligner (BWA, version 0.7.10) [
49]. To reduce the potential bias problems caused by the sequencing processes, the mapped duplicated reads were marked using Picard (version 1.118). Insertion and realignment (INDEL) realignment and base quality recalibration were performed using GATK (version 3.2.2) [
50]. The average of the mean target coverage was 65 and 85 % of the target exome was covered to 20×. Using the alignments, both small nucleotide variants (SNVs) and small INDELs were called by the GATK HaplotypeCaller and called variants were filtered using the GATK variant quality score recalibration process. Finally, for each family member, 84,356–86,901 variants were obtained. The variants were further filtered to exclude those included in dbSNP142, 1000 Genomes Project (Oct 2014), NHLBI-ESP project with 6500 exomes, or ExAC 65,000 exomes at the level of 5 % minor allele frequency (MAF) by ANNOVAR [
51]. In addition, variants were filtered out if they existed in in-house genome and exome databases.
Sanger sequencing
To validate the small deletion variant in the TSC2 gene, Sanger sequencing was performed. The target site of the variant and the flanking sequences of the DNA template from each family member were amplified with specific primers (forward primer 5′-ACAGTGACAGGGACGTCAGGTG-3′ and reverse primer 5′-ACAACCATTCATGGGAGACAGGA-3′) and the amplified products were directly sequenced on an ABI PRISM 3730 automated sequencer (Applied Biosystems, Foster City, CA, USA). The results were compared with the reference human genome sequence, GRCh37, to confirm the deletion variant.
DNA constructs
Full-length
TSC1 with a Myc-tag at the C-terminus, wild-type
TSC2, and the pathogenic
TSC2 mutation p.R611Q were kindly provided by Dr. Mark Nellist (Eramus Medical Centre, The Netherlands). pcDNA-Myc S6K1 was a gift from Jie Chen (Addgene plasmid #26610, Cambridge, MA, USA) [
52]. The novel
TSC2 c.700–701 deletion construct was derived from the wild-type
TSC2 construct by site-directed mutagenesis using a QuikChange II XL Site-Directed Mutagenesis Kit (Agilent Technologies, Santa Clara, CA). N-terminal HA-tagged wild-type
TSC2,
TSC2 p.R611Q, and
TSC2 c.700–701 deletion mutants were subcloned into the pcDNA3.1(+) vector using
BamHI and
XhoI sites. The complete open reading frame of the each construct was verified by sequencing. Construct DNA were prepared using the PureYield™ Plasmid Midiprep System (Promega, Madison, WI, USA).
Antibodies
Phospho-p70 S6 kinase Thr389 rabbit monoclonal (108D2 for Fig.
3 or 9205S for Fig.
4), Myc-tag mouse monoclonal (9B11), and HRP-conjugated anti-rabbit IgG (#7074) antibodies were purchased from Cell Signaling Technology (Danvers, MA, USA). HRP-conjugated goat anti-rat IgG (#AP183P), HRP-conjugated anti-mouse IgG (H+L) (#SA001-500), and anti-HA-Fluorescein, high affinity (#11 988 506 001) antibodies were obtained from Millipore (Darmstadt, Germany), GenDEPOT (Barker, TX, USA), and Roche (Basel, Switzerland), respectively.
Immunoblotting
A transfection-based immunoblot assay for the functional assessment of the TSC2 variants was performed as described previously [
30,
31]. HEK293T cells were plated on a 12-well plate and grown overnight in Dulbecco’s modified eagle’s medium (DMEM) (Hyclone, Logan, UT, USA) with 10 % fetal bovine serum (FBS). Confluent cells (70–90 %) were transfected with 0.4 μg
TSC2, 0.8 μg
TSC1, and 0.2–0.3 μg
S6K1 constructs using 4 μL Lipofectamine® 2000 (Thermo Fisher Scientific, Waltham, MA, USA) in Opti-MEM. The transfected medium was replaced with DMEM with 10 % FBS at 4 h after transfection. And then, everolimus (10 nM, LC Laboratories, Woburn, MA, USA) was added and incubated for another 20 h. The transfected cells were harvested, washed with cold phosphate-buffered saline, and lysed with 200 μL RIPA buffer (150 mM NaCl, 1.0 % Triton X-100, 0.5 % sodium deoxycholate, 0.1 % SDS, and 50 mM Tris-Cl, pH 8.0) containing a protease inhibitor cocktail and a phosphatase inhibitor cocktail (Roche). After full-speed centrifugation for 15 min at 4 °C, the supernatant fractions were quantified using a Thermo Scientific™ Pierce™ BCA Protein Assay kit. The protein samples (4.5 μg each) were electrophoresed on Bolt® 4–12 % Bis-Tris Plus Gels (Thermo Fisher Scientific) and transferred to PVDF membranes (Millipore) according to the manufacturer’s recommendations.
The blots were blocked for 10 min at room temperature with either 5 % Blotto, non-fat dry milk (sc-2325, Santa Cruz Biotechnology, Dallas, TX, USA) or 5 % bovine serum albumin in TBST (Tris-buffered saline plus 0.1 % Tween-20). Blots were incubated overnight at 4 °C with the following primary antibodies: 1:1000 dilutions of rabbit monoclonal anti-p-S6K (T389) or rat anti-HA antibody, or 1:10,000 dilutions of mouse monoclonal anti-myc tag antibody. After washing three times for 10 min in TBST, the blots were incubated for 2 h at room temperature with 1:5000 dilutions of secondary antibodies. After washing three times for 10 min in TBST, the blots were scanned using a ChemiDoc™ XRS System (Bio-Rad, Hercules, CA, USA) after applying detection reagents. To estimate the ratio of p-S6K (T389) to total S6K, the scans were analyzed using Image Lab™ Software (Bio-Rad).
Statistics
One-way analysis of variance (ANOVA) with appropriate post-hoc tests (Dunnett’s or Bonferroni’s test) was used for western blotting results. Each “n” indicates an independent set of experiment. Differences are considered significant at the level of p <0.05.
Acknowledgements
We thank the patient and family who participated in this study and Dr. Mark Nellist for sharing the plasmids for TSC1, TSC2, and TSC2-R611Q.