The Correlation Between Tuberous Sclerosis Complex Genotype and Renal Angiomyolipoma Phenotype

Tuberous sclerosis complex (TSC) is a rare multisystem autosomal dominant genetic disease that occurs between 1 in 6,000 and 1 in 10,000 live births. Additionally, renal angiomyolipoma is the most common form of renal disease in patients affected by TSC. Although a genetic mutation analysis of TSC is not rare, the correlation between the TSC gene mutation and renal angiomyolipoma phenotype is poorly understood. This study aims to analyze the mutation sites in 261 types of selected TSC patients. The results reveal that: (1) female patients develop more renal angiomyolipoma than male patients [p = 0.008, OR = 2.474, 95%CI (1.258–4.864)]; (2). The missense mutation of TSC1 led to a higher risk of renal angiomyolipoma [p < 0.01, OR = 15, 95%CI (2.859–78.691)], and in contrast, showed a reduced risk in patients with frameshift mutation [p = 0.03, OR = 0.252, 95%CI (0.07–0.912)]; (3). Patients with TSC2 mutations in the transcription activation domain 1 coding genes, had increased renal angiomyolipoma [p = 0.019, OR = 3.519, 95%CI (1.226–10.101)]. Therefore, our genotype-phenotype correlation study might shed light on the early monitoring and evaluation of renal angiomyolipoma in TSC patients.


INTRODUCTION
Tuberous sclerosis complex (TSC), also known as Bourneville's disease, is a rare autosomal dominant neurocutaneous syndrome. The estimated incidence rate of TSC is between of 1/6000 and 1/10,000 in live births and 1 in 20,000 in the population (Northrup et al., 2013). TSC is a genetic disorder that is characterized by the growth of benign hamartomas in various organs and is caused by TSC1 or TSC2 gene mutations (Ni et al., 2019). The TSC1 gene is located on chromosome 9q34 and consists of 23 exons that encode the hamartin (Slegtenhorst et al., 1997) and the TSC2 gene is located on chromosome 16p13 and consists of 42 exons that encode the tuberin (European Chromosome 16 Tuberous Sclerosis Consortium,, 1993). Interestingly, hamartin and tuberin form the TSC protein complex with TBC1 domain family member 7 (TBC1D7) (Dibble et al., 2012). Functionally, TSC1 is necessary to maintain the stability, activity, and the correct intracellular localization of the TSC1-TSC2 complex (Rosner et al., 2008). TSC2 contains a highly conserved GTPase activating protein that hydrolyzes the small GTPase Ras-homolog enriched in brain (RHEB), which inhibits the activation of mammalian rapamycin target protein 1 (mTOR1) which is vital in regulating cell growth and metabolism (Li et al., 2004). Thus, the TSC1/2 gene mutations inactivate the TSC protein complex and aberrations of mTORC1 results in the occurrence of TSC.
Severe renal manifestations occur with a high frequency in patients carrying TSC mutations. The estimated incidence rates of renal manifestations, including renal angiomyolipoma (RAML), renal cysts, and renal cell carcinoma, in patients carrying TSC mutations range from 48 to 80%. Among these, RAML, which is a benign tumor, is the most common type, with an incidence rate of 34-80% (Rakowski et al., 2006). Of note, compared with sporadic RAML, TSC-RAML had an early onset, with bilateral and multiple lesions and faster growth (Rabenou and Charles, 2015). Although most of the RAML patients had no obvious clinical symptoms, a great proportion of patients experience abdominal pain, low blood volume shock, renal insufficiency, or end-stage renal disease which can even lead to mortality (Swärd et al., 2020;Vabret et al., 2020). Overall, RAML is considered to be the major cause of death in TSC patients (Amin et al., 2017).
Accumulating studies have suggested a correlation between the severity of clinical symptoms and TSC1, and TSC2 gene mutations in patients with TSC (Lewis et al., 2004;Au et al., 2007;Bai et al., 2017). Nevertheless, the correlation between the types/sites of gene mutations and symptoms are fragmentary and therefore should be elucidated further (van Eeghen et al., 2012). The correlation between the genotype of the TSC gene and the RAML phenotype would benefit early detection and intervention as well as the prognosis and quality of life of patients with RAML.

Literature Search
An extensive and comprehensive literature search was conducted with regard to the relationship between TSC genotypes and RAML phenotypes using PubMed and Google. The search terms included "(RAML OR renal angiomyolipoma OR renal hamartoma) AND (TSC OR tuberous sclerosis complex OR Bourneville's disease) AND (genotype OR mutation), ""(angiomyolipoma OR renal vascular smooth muscle lipoma OR renal tumors) AND (TSC OR tuberous sclerosis complex OR Bourneville's disease) AND (genotype OR mutation), ""(TSC OR tuberous sclerosis complex OR Bourneville's disease) AND (RAML OR renal angiomyolipoma OR renal hamartoma), " and "(TSC OR tuberous sclerosis complex OR Bourneville's disease) AND (angiomyolipoma OR renal vascular smooth muscle lipoma OR renal tumor), " (TSC OR tuberous sclerosis complex OR Bourneville's disease) AND (genotype OR phenotype). All retrieved papers were read to evaluate whether the reported patients met the inclusion criteria of the present study.

Inclusion Criteria
Patients (1) who met the 2012 NIH Tuberous Sclerosis Complex Diagnostic Criteria Updated (Northrup et al., 2013); (2) who underwent abdominal imaging examination (ultrasound, computed tomography or magnetic resonance imaging) for a definitive diagnosis of RAML; and (3) should undergo TSC gene testing to determine the pathogenic germline mutations.

Exclusion Criteria
Patients (1) who lacked abdominal imaging examination or did not undergo abdominal imaging examination for confirming the diagnosis of RAML; (2) with incorrect gene mutation information or if the base of the DNA mutation failed to comply the corresponding base in the reference sequence; (3) with two or more TSC gene variants detected, and pathogenic mutations could not be successfully determined; and (4) with only a specific mutation type or mutation site.
Based on the strict inclusion and exclusion criteria, 261 patients were enrolled in our study, of which 54 patients had the TSC1 gene mutation and 207 patients had the TSC2 gene mutation. The original data was collected from papers that were published from 1998 to April 2020. Patients were mainly recruited from six countries and regions including China, United States, Japan, Taiwan, Russia and Greece.

Mutation Analysis
The TSC gene mutation information was verified and analyzed for each patient using 1 Mutalyzer with the transcript reference number TSC1 (NM_000368.4) or TSC2 (NM_000548.3). The LOVD database was used to explore the exon number of 2 TSC1 and 3 TSC2.
The TSC2 gene contains 41 coding exons and 1 non-coding guiding exon (Satake et al., 1998). In previous studies, the number of exons were in disunity. In our study, we described all the TSC2 mutation sites according to the LOVD Databases with an encoding exon. Tuberin contains seven functional domains, that are arranged from the N-terminal to the C-terminal, and are as follows: overlapping amino acids between the TAD2 and CaBD domains, patients with partial amino acid changes that are located in the overlapping regions could not be located, and so two domains were included in one group during the analysis. Additionally, tuberin also has a hamartin interaction domain (HID). According to the location of encoding exon (E) and effect of the TSC2 gene mutations, the affected amino acids were divided into three subsets, HID-TID (E1-E22), GAP domain (E34-E41), and mutations between the two regions (E23-E33) (van Eeghen et al., 2012).
Moreover, the mutation types were divided into a protein truncation mutation (PT) and a protein non-truncation (NT) mutation based on differential regulations of the gene mutation on the protein structure. PT comprises frameshift mutations, splice site mutations, nonsense mutations, and large fragment mutations. NT contains missense mutations and in-frame mutations (van Eeghen et al., 2012). Large fragment mutations were excluded from the study when the correlation between the functional domain gene mutation and RAML was present.

Statistical Analysis
SPSS version 25.0 (IBM Corporation, Chicago, IL, United States) and GraphPad Prism version 8.0.2 were used for data analysis and illustration. A Chi-square test was used for categorical variables. When the expected value on contingency tables was below 5, then the results of the total sample size with ≥40 were corrected for continuity. When any expected value in the contingency table was below one or the total sample number was less than 40, then Fisher's exact test was used. Binary logistic regression was used to evaluate the risk of developing RAML, and p-values of <0.05 were considered to be statistically significant.

Patient Characteristics
A total of 261 patients were analyzed in this study. Briefly, 54 patients had TSC1 mutations, and 22 of 54 (40.74%) had RAML.

Correlation Between TSC1 Mutations and RAML
In total, 54 patients with TSC1 mutations were included and analyzed to obtain the results in this study. The types of mutations in TSC1 included nonsense, frameshift, large fragment, and missense mutations. Discrete types and rates of mutations in RAML patients with the TSC1 gene mutation were observed when compared to non-RAML patients (Figures 2A,B). There were 41 PT mutations among the TSC1 gene mutations, and 11 (26.83%) of these patients had RAML. In addition, 13 people had NT mutations, and 11 (84.62%) of these patients had RAML ( Figure 2C). Therefore, PT mutations had a lower incidence of RAML than NT mutations (p < 0.01, OR = 0.067, 95%CI 0.13-0.35). The frameshift mutations were lower in patients with RAML (p = 0.03, OR = 0.252, 95%CI 0.07-0.912), and the rate of missense mutations was higher in patients with RAML (p < 0.01, OR = 15, 95%CI 2.859-78.691) compared with all other mutation types, except the missense mutation ( Figure 2D). There was no difference between other mutation types in the groups (Table 1). However, due to the small sample size of PT and NT TSC1 mutations included in this study, a bias might occur in our findings.
In the RAML patients, there were 41 (42.71%) gene mutations in the HID-TID domain and 21 (21.88%) gene mutations in exons encoding the GAP domain. There are 34 (35.42%) mutations in the middle exons of 23-33. In patients without RAML, 47 mutations (47%) were observed in the HID-TID domain, while it was 26 (26%) mutations in the GAP domain and 27 (27%) mutations in the middle coding region. Of note, TSC2 gene mutations favored accumulation in the HID-TID coding region. Nonetheless, the location of gene mutations in the exons that encode distinct proteins showed no correlation with the occurrence of RAML disease ( Figure 3D).
Interestingly, there were 41 gene mutations in exons that encoded the seven functional proteins in the TSC2 mutation patients with RAML, which included three in CCD1, one in CCD2, 15 in TAD1, eight in GAPD, and 14 in TAD2/CaBD. On the other hand, 36 mutations were found in the same loci in patients without RAML. Briefly, there are three in CCD1, four in CCD2, five in TAD1, 11 in GAPD, and 13 in TAD2/CaBD ( Figure 3E and Table 2).
Intrigued by the results in Table 2, we inferred that the mutations that occurred in the gene encoding TAD1 led to a higher risk of RAML than any other functional regions of tuberin. Our statistical results show that the TAD1 mutation in TSC2 patients has a higher risk of RAML [p = 0.019, OR = 3.519 (95%CI 1.226-10.101)]. There were eight nonsense mutations, one splice site mutation, three frameshift mutations, and three missense mutations in the exons encoding TAD1 in RAML patients. On the contrary, one missense mutation, two nonsense mutations, one splice site mutation, and one frameshift mutation were observed in the same loci in patients without RAML. However, there was no statistically significant difference with regard to the distribution of mutations between the two groups.

DISCUSSION
Currently, there are fragmentary studies that focus on the correlation between TSC gene mutations and phenotypes. Although the genotype-phenotype correlation in patients with TSC-RAML has been conducted in many studies (Li et al., 2018;Ni et al., 2019), the sample size of these remained too small. Thus, a further analysis with a larger sample size is required. Our research included a larger study cohort by extensively searching the studies published in the last two decades and patients from multiple countries were included to make our study more convincible.
Several studies have indicated that TSC1 mutations are associated with milder clinical outcomes in patients with TSC (Jones et al., 1999;Dabora et al., 2001;Sancak et al., 2005). In addition, the incidence of RAML is higher, and the disease is more severe in patients with TSC2 mutations than in those with TSC1 mutations (Rakowski et al., 2006;Au et al., 2007). However, another study revealed that only milder developmental delay and/or intellectual disability was observed in patients with TSC1 mutations than in those with TSC2 (Niida et al., 2013). We herein revealed that the prevalence of the TSC2 gene mutation with RAML is higher than that of TSC1, without reaching a statistical difference. Moreover, female patients had a higher risk of RAML than male patients [p = 0.008, OR = 2.474, 95%CI (1.258-4.864)], which is in agreement with that elucidated in the previous studies (Rakowski et al., 2006;Au et al., 2007).
Regardless of the existence of RAML, mutations in the TSC1 gene involved the trend of N-accumulation. Patients with RAML bear more frequent variants in the TID domain of the TSC1 protein versus patients without RAML, even though the difference was not statistically significant. TSC1 maintains stability, activity, and transportation of the TSC1-TSC2 complex (Rosner et al., 2008). Amino acids that are close to the N-terminal region of TSC1 are dominant in TSC1 functions. The alterations in the amino acid sequence of this region disrupt the formation of large TSC1 integration and promote proteasome-mediated TSC1 degradation, leading to an mTOR signaling cascade. Indeed, rapid degradation of the mutant TSC1 isoforms in vivo removed the suppressive effects of TSC1-TSC2 -variants on mTOR signaling (Mozaffari et al., 2009), promoting the progression of TSC. Among the mutations in TSC1, missense mutation was shown to be higher in patients with RAML [p < 0.05,OR = 15,]. Meanwhile, due to fewer amino acid substitutions, missense mutations ameliorated clinical symptoms in patients with TSC (van Eeghen et al., 2012. How the TSC1 missense mutation contributed to the phenotype remains unclear (Nellist et al., 2009). Of interest, some missense mutations not only change at just an amino acid, but also affect the splice site. For example, the c.737G > A (p. Arg246Lys) transition is predicted to be the most likely pathogenic splice-site mutation. It was predicted to destroy the splice donor site at the 3 end of exon 8 (Nellist et al., 2009). Specific missense mutations might lead to increased incidence of RAML in our study. However, due to the small sample size of the TSC1 mutation included in this study, a bias might occur in our findings.
Research shows that tuberin plays a novel role in redox process which may contribute to the development of kidney tumors in patients with a tuberous sclerosis complex (Habib and Abboud, 2016). PT mutation was the most dominant category of TSC2 mutations and was not related to the presence of RAMLthis is in agreement with previous findings (Lewis et al., 2004). Uniform TSC2 gene mutation sites are mainly located in the proximal N-terminal HID-TID functional region, in both RAML and non-RAML patients. The N-terminal region of tuberin contributed to the interaction of hamartin and tuberin (Hodges et al., 2001). Mutations that occur near the N-terminal region that potentially disturb the HID-TID functional region might hamper the formation of the TSC complex, leading to TSC progression. In addition, the GAP region of tuberin contains RapGAP homology that negatively regulates Rheb. RAML and non-RAML patients also had similar mutation rates for the genes encoding GAP. Furthermore, TAD1 is a transcriptional activation domain that is encoded by exons 29-30 (amino acids 1163-1259). Patients with mutations in the TAD1 of tuberin had a higher risk of developing RAML than patients with mutations in other regions. [p = 0.019, OR = 3.519, 95%CI (1.226-10.101)], albeit comparable mutation types in the TAD1 region in patients with and without RAML. The mutation in the C-terminal gene regulates the function of Rheb. The identification of TADs in TSC2 warrants an indepth study with regard to the transcription-related function of TSC2. Indeed, TSC2 can modulate the transcriptional activity of the steroid receptor superfamily via binding to the estrogen receptor alpha (ERα) by TADs (Henry et al., 1998). Anzick reported the direct correlation of tumors, including breast and ovarian cancer, with the overexpression of AIB1, a member of the steroid receptor coregulators (Anzick et al., 1997). It might be hypothesized that TSC2 binding to Erα can also regulate the cell cycle and differentiation, leading to the formation of multiple system hamartomas, including RAML. TSC2 also negatively regulates the transcription level of EREG through promoter sub-regions of EREG (Pradhan et al., 2014). The upregulation of EREG has been found in several cancers and TSC patients. Mutations of TADs may disrupt the transcription factor function of TSC2 and may cause the upregulation of EREG and the occurrence of RAML. Nonetheless, these studies mainly focused on the TADs of the C-terminal region. Thus, our understanding on TAD1 function is limited and warrants further investigation.
The present study found that the missense mutation of TSC1 and the mutation of TSC2 encoding TAD1 showed a correlation with the high risk of RAML. The frameshift mutation of TSC1 also lowered the risk of RAML. This helps in early detection and intervention of TSC patients with RAML. However, PT mutations in the proximal terminal region of the gene scan leads to a base change of the distal region. Due to insufficient gender and age information of the included subjects, this study failed to correct the confounding factors, resulting in a bias. A larger study cohort with comprehensive basic information is therefore warranted to obtain more accurate results.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

AUTHOR CONTRIBUTIONS
QG and HX designed the study. NZ, XW, and HX wrote the manuscript. ZT, XQ, ZG, and DH analyzed the data. All authors aided in data interpretation, reviewed, and approved the final manuscript.

FUNDING
This work was supported by the Fundamental Research Funds for the Central Universities (20ykpy109) to HX.
The funder had no role in the design of the study, collection, analysis, and interpretation of data or in writing the manuscript.