Introduction
Lymphatic malformations (LMs) are congenital lesions caused by defects in the development of the lymphatic system and mainly observed in neonates or young children, with an incidence of 1 in 4000 to 1 in 2000 [
1]. Lymphatic malformations (LM) are characterized by the overgrowth of lymphatic vessels during pre- and postnatal development [
2]. LMs can cause adjacent structures compromise leading to airway obstruction even dyspnea, cosmetic deformity, swallowing impairment, infection or naturally diffuse, especially in head and neck region. Conservative observation and surgery are main treatments and the latter including surgical excision (e.g., partial or total excision), sclerotherapy, radiofrequency ablation, laser therapy. With the development of medical genetics of LMs, new therapies increasingly emerged such as oral medications (i.e., sildenafil, propranolol, and sirolimus), and vascularized lymph node transfer [
3,
4]. A systematic review recruited 20 trials including 71 patients with oral sirolimus, and found that the sirolimus was effective for LMs [
4]. However, patients with extensively infiltrating LMs, namely, intractable LMs (iLM), experienced high risks of recurrence and progression, routine treatment regimens are less effective for iLMs [
5].
Recently, increasing evidence shows that the PI3K/AKT/mTOR pathway is involved in the pathogenesis of isolated LMs and syndromic disorders in which LM is a component feature [
6‐
9]. For example, proteus syndrome patients who also have LMs carried a somatic mutation in
AKT1, which encodes RAC-alpha serine-threonine protein kinase and plays a role in lymphangiogenesis [
9,
10]. Moreover, somatic mutations that activate phosphatidylinositol 4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) have been found in approximately 79% of LMs [
8,
11]. Thus around 20% remain unexplained. Somatic
PIK3CA mutation is not only identified in isolated LM, but also in CLOVES syndrome or Klippel–Trenaunay–Weber syndrome [
9]. High activity of the PI3K-AKT- mTOR pathway was demonstrated by hyperphosphorylation of AKT-Ser473 in all LM-derived lymphatic endothelial cells (LECs) as compared to normal LECs while LM-derived fibroblasts did not possess such mutations [
12]. Several
PIK3CA somatic mutations have been shown to be oncogenic as well as function as the main pathogenic mechanisms of LMs and vascular malformations by promoting the hyperproliferation of endothelial cells [
13]. The International Society for the Study of Vascular Anomalies (ISSVA) has also identified a
PIK3CA mutation as a specific pathogenic cause for LMs. However, the molecular mechanism of LMs without
PIK3CA mutations is still unclear, and other genetic alterations have not been found in the disease. Thus, the discovery of new genetic events in LMs is crucial to identify the molecular mechanism of the pathogenesis and further develop novel targeted therapies. In this study, a novel candidate mutation in
PIK3CD was identified as an LM-associated mutation by whole-exome sequencing (WES) and was validated by in vitro functional studies.
Discussion
PI3Ks are a family of lipid kinases with critical roles in cell biology, including cell proliferation, differentiation, migration and survival [
15‐
17]. There are three categories of PI3Ks (Class IA, Class IB; Class II; and Class III). Class IA PI3Ks comprise a p110 catalytic subunit and a p85 regulatory subunit. The p110α, p110β and p110δ catalytic isoforms are encoded by the
PIK3CA,
PIK3CB and
PIK3CD genes, respectively. p110α is frequently involved in human cancers, including endometrial, breast, ovarian, colorectal and other various tumours, by affecting cell proliferation, migration and survival [
18‐
22]. Most of the LMs were caused by somatic mutations in the
PIK3CA gene, which could lead to the hyperproliferation of lymphatic endothelial cells [
7,
8,
13].
Among three Class IA PI3K catalytic isoforms, p110α and p110β are ubiquitously expressed, whereas p110δ is principally enriched in leukocytes and regulates immune functions [
23]. However, some non-leucocytes such as neurons [
24], ECs (endothelial cells) [
25] and lung fibroblasts [
26] also express p110δ, albeit at lower levels than in leucocytes. In addition, p110δ is generally overexpressed to induce cancer cell growth and invasion by activating the AKT-mTOR pathway in hepatocellular carcinoma, glioma, glioblastoma, neuroblastoma, colorectal cancer and breast cancer [
27‐
30]. It is well known that LMs present some similar characteristics as tumours, such as uncontrolled cell proliferation and extension into surrounding tissues. Lymphatic vascular endothelial cells in LMs usually exhibit abnormal proliferation due to mTOR activation [
31]. However, it was unclear whether genetic changes in
PIK3CD play a role in the pathogenesis of LMs. All these signs indicated the endothelial cells overproliferated and aggregated, leading to LMs. In the present study, in vitro functional studies demonstrated that exogenous overexpression of wild-type and most significantly mutant
PIK3CD increased the proliferation rate of HUVECs. In addition, phosphorylated protein levels of AKT, mTOR and S6 were significantly increased in cells with exogenous overexpression of the
PIK3CD mutant, suggesting that the elevated expression of mutant
PIK3CD in vascular endothelial cells may promote the overgrowth of endothelial cells and further affect lymphatic vessel development.
To date, LMs are often treated with rapamycin or rapamycin analogues such as everolimus to cure the lesions and improve quality of life [
32]. Accepted paper from our laboratory also demonstrated that rapamycin could effectively reduce volume of LMs especially for Macrocystic LMs [
33]. Newly paper showed that a combination of VEGFC inhibition with rapamycin is much more potent inducing even LM regression in mice, although this is a contraindication for VEGF inhibition in children [
34].
However, given almost 80% LM patients carried
PI3KCA mutations, mutation-specific inhibitors or combination of inhibitors have become a promising choice for the treatment of LMs. Similar to developmental tumours, LMs carrying a single mutation might be more sensitive to targeted therapies than tumours carrying multiple mutations [
35]. Studies have shown that p110α-specific inhibitors could normalize aberrant PI3K signalling, thereby reducing or eliminating PIK3CA-driven vascular malformations. The p110α-specific inhibitor BYL719 was also successfully applied for the treatment of patients with PIK3CA-related overgrowth syndrome, which gives hope to patients with LMs [
36]. In the future, we will enrol a larger group of patients with LMs to detect the mutation frequency of
PIK3CD and to elucidate the mechanism of its pathogenicity. As this is a promising gene for novel targeted therapies, we will also evaluate the effects of
PIK3CD mutant-specific inhibitors on the reversal of cellular dysfunction.
Materials and methods
Patients and sample collection
All 6 patients were admitted to and diagnosed by clinicians in the Department of Otolaryngology at Beijing Children's Hospital Affiliated to Capital Medical University. The clinical characteristics were collected from their medical records (Table
1). Guardians of all the participants signed informed consent forms (ICFs) designed in accordance with the Declaration of Helsinki. The tissue specimens of LMs were obtained under the human subject protocol approved by the Human Ethics Committee of Beijing Children's Hospital Affiliated to Capital Medical University (ID: 2019-k-66, approved on February 2019).
Whole-exome sequencing (WES)
Peripheral blood and tissue specimens of LMs from all 6 children were sent to Running Gene Inc. (Beijing, China) for WES (Additional file
1: Figure S2). Average depth of coverage was × 142 in blood and × 166 in tissue (Additional file
1: Table S1). DNA samples were isolated from the peripheral blood and lymphatic tissue specimens with a DNA Isolation Kit (Bioteke, AU1802 and AU18016). The DNA concentrations were measured with a Qubit dsDNA HS Assay Kit (Invitrogen, Q32851) on a Qubit fluorometer (Invitrogen, Q33216). High-quality DNA samples were fragmented into 200–300 bp by a Covaris Acoustic System (Covaris, Massachusetts, USA), and the resulting DNA fragments were processed with a KAPA Library Preparation Kit (Kapa Biosystems, KR0453) to construct a DNA library. The libraries were estimated with a Qubit dsDNA HS Assay kit (Invitrogen, Q32851), after which hybridization of pooled libraries to the capture probes was conducted with an Agilent SureSelectXT2 Target Enrichment System (Agilent, Santa Clara, USA). Probe-captured DNA fragments were then enriched by post-capture PCR. The final products were sequenced on an Illumina HiSeq X10 platform (Illumina, San Diego, USA) as 150 bp paired-end reads.
Raw data from the HiSeq X10 platform were processed for quality control and then aligned against the human reference genome (GRCh37/hg19) using the Burrows-Wheeler Alignment tool (
http://bio-bwa.sourceforge.net/). Duplicate reads were identified using GATK software (
www.broadinstitute.org/gatk), and single-nucleotide polymorphisms and insertions and deletions were examined. Low-quality variants were filtered out based on quality by depth (< 2.0), mapping quality (< 40.0), Fisher strand (> 60.0), mapping quality rank sum test (< -12.5) and read position rank sum test (< -8.0). All the called variants were annotated by ANNOVAR (annovar.openbioinformatics.org/en/latest/) based on public databases (1000 Genomes Project, ExAC, gnomAD, ESP6500, CCDS, RefSeq, Ensembl, etc.). The potential impacts of candidate single-nucleotide variants were predicted by the MutationTaster, SIFT, Provean and Polyphen-2 programs.
Germline mutations involved in either PI3K/AKT/mTOR or Ras pathways
Low-quality variants were filtered out based on quality by depth (< 8.0). The remaining variants were filtered against 1000 Genomes Project_EAS, ExAC and gnomAD, with a minor allele frequency (MAF) < 1% for autosomal and X-linked recessive mutations and an MAF < 0.01% for autosomal and X-linked dominant mutations. Based on the Human Gene Mutation Database, nonsense, frameshift, and splicing mutations annotated as disease mutations were retained. Only candidate genes associated with both PI3K/AKT/mTOR and Ras pathways were included. No definite pathogenic germline variant was identified.
Somatic mutations involved in either PI3K/AKT/mTOR or Ras pathways
Germline mutations appearing in the peripheral blood were filtered out. The remaining mutations were selected based on quality by depth (< 8.0) and against 1000 Genomes Project_EAS, ExAC and gnomAD, with an MAF < 0.01%. Only exonic and splicing variants were included. Synonymous variants and variants with low number of alteration (alt < 4) were excluded as well. Finally, only candidate genes associated with the both PI3K/AKT/mTOR and Ras pathways were included.
Digital polymerase chain reaction (PCR)
Digital PCR were conducted to verify the remaining
IRS1, MTOR,
TSC1, TSC2,
PIK3CA and
PIK3CD variants. DNA samples were mixed with 2X ddPCR Supermix for probes (Bio-Rad Laboratories, Inc., USA), probes, primers and ddH2O (Table
3). The mixture and droplet generation oil (Bio-Rad Laboratories, Inc., USA) were separately loaded on the DG8 cartridge. Then, targeted droplets were generated by a QX200 Droplet Digital PCR system (Bio-Rad Laboratories, Inc., USA) and transferred to 96-well plates. After PCR in a Bio-Rad thermal cycler T100, the digital PCR data were read and briefly analysed on the QX200 Droplet Digital PCR system.
Table 3
Sequence of primers and probes used for digital polymerase chain reaction (ddPCR)
PIK3CA-F | GCTCAAAGCAATTTCTACACGA |
PIK3CA-R | CTTACCTGTGACTCCATAGAAAATC |
PIK3CA-P-G | 6-FAM- TGAAATCACTGAGCAGGA- BHQ-X |
PIK3CA-P-A | HEX- TGAAATCACTAAGCAGG- BHQ-X |
PIK3CD-F | TCCGAGATGCACGTGCC |
PIK3CD-R | CCTTCATGTGGTGGGTGCT |
PIK3CD-P–T | 6-FAM -TTCGGCCTCATCCT- BHQ-X |
PIK3CD-P–C | HEX -TTCGGCCCCATCC- BHQ-X |
Cell culture and infection
Human umbilical vein endothelial cells (HUVECs) were donated by Beijing Belife Bio-Medical Technology LTD and cultured in endothelial cell medium (ECM) (cat no. 1001; ScienCell, San Diego, California, USA) supplemented with 5% foetal bovine serum (FBS) (cat no. 0025; ScienCell, San Diego, California, USA), 1% Endothelial Cell Growth Supplement (ECGS) (cat no. 1052; ScienCell, San Diego, California, USA) and 1% penicillin/streptomycin solution (cat no. 0503; ScienCell, San Diego, California, USA). Cells were cultured at 37 °C in a humidified atmosphere containing 5% CO2.
cDNAs coding wild-type
PIK3CD (GenBank accession NM_005026.4) or
PIK3CD mutations were synthesized for adenovirus packaging (Vigene Bioscience, Shandong, China). The vector used was a bi-cistronic construct with EGFP. HUVECs were seeded in each well of 6-well plate (5 × 10
3 cells/well). Approximately 18–24 h later, the medium was replaced with fresh medium containing different viruses (final concentration, 2.5 × 10
6 pfu/mL). At 48 h after initial virus treatment, the infection efficiency was evaluated using GFP fluorescence imaging (Additional file
1: Figure S3).
Cell viability assay
Infected cells (5 × 103/well) in 100 μL of culture medium were seeded into 96-well plates and incubated for 18–24 h, after which the medium was replaced with a virus suspension (final concentration 2.5 × 106 pfu/mL) in fresh medium. Afterward, the cell viability assay was performed by adding 10 µL of reagent from Cell Counting Kit-8 (CCK8) (Meilunbio, cat no. MA0218-L, Dalian, China) into each well and incubating the plates for another 2 h; then, absorbance at 450 nm was detected with a microplate reader (Molecular Devices, Silicon Valley, CA, USA). The cell survival rate was calculated as follows: (OD value of wild-type PIK3CD or mutant PIK3CD group/OD value of the control group) × 100%.
Quantitative real-time PCR (RT-qPCR)
Total RNA was isolated from HUVECs using an RNA isolation kit according to the manufacturer's protocol (cat no. 220010; Shanghai feijie biological, Inc, Shanghai, China). The RNA was subsequently reverse transcribed into cDNA using a KR106-02 reverse transcription system according to the manufacturer’s instruction (cat. no. KR106-02; TIANGEN, Beijing, China). To detect RNA expression, qPCR analyses were carried out in triplicate using SYBR Green PCR Master Mix (cat. no. FP205-02; TIANGEN, Beijing, China) and run on a Roche LightCycler 96 (Roche Diagnostics, Indianapolis, IN, USA). Relative expression was calculated by the 2 − ΔΔCt method with GAPDH as the endogenous control. The primer sequences for the specific targets were showed in Table
4.
Table 4
Sequence of primers used for quantitative real-time PCR (RT-qPCR)
GAPDH-F | 5ʹ- GGAGCGAGATCCCTCCAAAAT -3ʹ |
GAPDH-R | 5ʹ- GGCTGTTGT CATACTTCTCATGG -3ʹ |
mTOR-F | 5ʹ- ATGCTTGGAACCGGACCTG -3ʹ |
mTOR-R | 5ʹ-TCTTGACTCATCT CTCGGAGTT -3ʹ (reverse) |
AKT(Human)-F | 5′-CTACCCACACAGCAGTACGC-3' |
AKT(Human)-R | 5′-AAGTCGCTGGTGTTAAGCCG-3' |
S6-F | 5ʹ- AGGGTTATGTGGTCCGAATCA -3ʹ |
S6-R | 5ʹ- TTGGTCTGT AACAGGAATGCC -3ʹ |
Western blot
Total proteins were extracted from HUVECs using RIPA buffer (cat. no, XSY-WB-001; B-Belife, Beijing China) mixed with 1% protease inhibitor cocktail (cat. no. 04693116001; Roche Molecular Biochemicals, Mannheim, Germany). The protein concentration was determined using a BCA Protein assay kit (cat. no. CW0014S; CWBIO, Beijing, China). Protein extracts were then mixed with 5 × SDS loading buffer and boiled for 10 min. Then 20–50 μg per sample were separated via SDS-PAGE under reducing or non-reducing conditions on a 10% polyacrylamide gel and then electrotransferred onto PVDF membranes (Millipore, Bedford, MA, USA) in vertical buffer tanks. The membranes were blocked with 5% non-fat milk in TBST buffer (50 mM Tris–HCl (pH 7.4), 0.9% NaCl, and 0.1% Tween 20) before they were incubated with primary antibodies (Table
5) for 2–3 h at room temperature or overnight at 4 °C. After the addition of the HRP-conjugated secondary antibodies for 1 h, signals were detected with an electrogenerated chemiluminescence (ECL) detection reagent (cat. no. MA0186; Meilunbio, Dalian, China). Relative target protein expression levels were normalized to those of tubulin and visualized using ImageJ software.
Table 5
Antibodies for western blot
1 | Anti-p-AKT antibody | Abcam | ab8933 |
2 | Anti-AKT antibody | Abcam | ab8805 |
3 | Anti-p-mTOR antibody | Abcam | ab109268 |
4 | Anti-mTOR antibody | Abcam | ab32028 |
5 | Anti-p-S6 antibody | CST | 4858 |
6 | Anti-S6 antibody | CST | 2217 |
Wound healing assay
The migration ability of HUVECs were assessed by the wound healing assay. A sterile tip was used to create a wound in a cell monolayer, which was then washed 3 times to remove non-adherent cells, and fresh medium was added to the cultures. Images were captured at 0, 4, 8 and 12 h after scratching. Photoshop software (Adobe Photoshop CS6) was used to measure the area of the wound at each time point and calculate the wound healing rate as follows: migration area (%) = (A0 − An)/A0 × 100, where A0 represents the area of initial wound area, an represents the remaining area of wound at the metering time point.
Statistical analysis
All in vitro experiments performed in this study were repeated three times. Statistical analysis was performed using GraphPad Prism software, and all comparisons between groups were assessed using Student’s t test. All clinical data are indicated as mean ± standard deviation (Mean ± SD), with the significant statistical threshold of two-tailed p value < 0.05.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.