Identification of a novel LPL nonsense variant and further insights into the complex etiology and expression of hypertriglyceridemia-induced acute pancreatitis

Background Hypertriglyceridemia (HTG) is a leading cause of acute pancreatitis. HTG can be caused by either primary (genetic) or secondary etiological factors, and there is increasing appreciation of the interplay between the two kinds of factors in causing severe HTG. Objectives The main aim of this study was to identify the genetic basis of hypertriglyceridemia-induced acute pancreatitis (HTG-AP) in a Chinese family with three affected members (the proband, his mother and older sister). Methods The entire coding and flanking sequences of LPL, APOC2, APOA5, GPIHBP1 and LMF1 genes were analyzed by Sanger sequencing. The newly identified LPL nonsense variant was subjected to functional analysis by means of transfection into HEK-293 T cells followed by Western blot and activity assays. Previously reported pathogenic LPL nonsense variants were collated and compared with respect to genotype and phenotype relationship. Results We identified a novel nonsense variant, p.Gln118* (c.351C > T), in the LPL gene, which co-segregated with HTG-AP in the Chinese family. We provided in vitro evidence that this variant resulted in a complete functional loss of the affected LPL allele. We highlighted a role of alcohol abuse in modifying the clinical expression of the disease in the proband. Additionally, our survey of 12 previously reported pathogenic LPL nonsense variants (in 20 carriers) revealed that neither serum triglyceride levels nor occurrence of HTG-AP was distinguishable among the three carrier groups, namely, simple homozygotes, compound heterozygotes and simple heterozygotes. Conclusions Our findings, taken together, generated new insights into the complex etiology and expression of HTG-AP.


Introduction
Acute pancreatitis (AP) is an acute inflammatory disease that is characterized by local pancreatic inflammation and consequently systemic inflammatory response [1,2]. Gallstones, alcohol abuse and massive hypertriglyceridemia (HTG) are generally thought to be three leading etiologies of AP worldwide [3]. However, unlike in Western countries, HTG, rather than alcohol abuse, is the second leading cause of AP in China [4]. Hypertriglyceridemiainduced acute pancreatitis (HTG-AP) is defined by serum triglyceride (TG) level exceeding 11.3 mmol/L (1000 mg/ dL) or between 5.6 to 11.3 mmol/L (500~1000 mg/dL) together with lipemic serum [5,6]. As compared to other etiologies, HTG-AP is usually more severe and has higher recurrence rate [7,8].
According to the etiology, HTG can be divided into primary and secondary HTG. Secondary HTG is usually caused by metabolic syndrome, diabetes, alcohol consumption, obesity, chronic renal failure, etc. [9] Primary HTG is caused by genes defects related with TG metabolism, including lipoprotein lipase (LPL), apolipoprotein C-II (APOC2), apolipoprotein A-V (APOA5), glycosylphosphatidylinositolanchored high density lipoprotein-binding protein 1 (GPIHBP1) and lipase maturation factor 1 (LMF1). LPL is the key enzyme that catabolizes TG in non-hepatic tissues [10]. APOC2 and APOA5 act as essential LPL activators [11,12]. LMF1 is involved in the folding and expression of LPL [13]. GPIHBP1 mediates the transmembrane transport and binding of LPL [14].
However, in most cases, the cause of HTG is complex [15]. Severe HTG was recently shown to be primarily polygenic [16], and there is increasing appreciation of the interplay between primary and secondary etiological factors in causing severe HTG [17,18]. In this study, we reported a novel LPL nonsense variant in one typical Chinese family with HTG-AP history and discussed insights into the complex etiology of HTG-AP gleaned from the so far reported pathogenic LPL nonsense variants.

Ethical statement
This study was approved by the Ethics Committee of Jinling Hospital. Informed consent was obtained from all participants.

Family description
The male proband had been suffered from recurrent severe HTG-AP since 26 years old, respectively in 2003, 2007, 2014 and 2017. He has had hypertension for 7 years and abused alcohol for more than 5 years (250-350 g/d). His body mass index (BMI) was normal (22.7 kg/m 2 ). His mother and older sister also respectively had one-and two-times onset of HTG-AP.
Sequencing of the LPL, APOC2, APOA5, GPIHBP1 and LMF1 genes Genomic DNA was extracted from blood by the Gentra Puregene Blood kit (Qiagen, Dusseldorf, Germany) according to the manufacturer's instructions. All exons and exon/intron boundaries of the LPL, APOA5, APOC2, LMF1 and GPIHBP1 genes were analyzed by sanger sequencing [18].

Population allele frequency reference and variant nomenclature
Population allele frequencies of variants found in this study were evaluated using the Genome Aggregation Database (gnomAD) genome dataset [19] via VarSome [20]. Variant nomenclature was in accordance with Human Genome Variation Society (HGVS) recommendations [21]. NM_000237.3 was used as the LPL mRNA reference sequence.

Plasma lipid profile analysis
Blood samples were taken from the proband after fasting for 12 h. Serum TG, TC, HDL, LDL levels were measured enzymatically on an automatic analyzer (Hitachi High-Tech, 7600-120, Japan).

Post-heparin LPL mass analysis
Post-heparin blood samples were collected into Na-EDTA tubes 10 min after intravenous heparin injection (60 IU/kg body weight) and fasting for 12 h. Postheparin plasma LPL mass was detected by immunoassay using the Human LPL Elisa kit (TSZ Biological Trade, USA).

LPL activity analysis
LPL activity was in principle measured through detecting free fatty acid (FFA) concentration [22]. The reaction substrate, termed buffer A, was composed of 1 ml TGrich serum (TG concentration, > 3000 mg/dL) from Gpihbp1-deficient mice (Gpihbp1 −/− ) [23], 0.18 g 10% fatty acid-poor bovine serum albumin (BSA) (Miles, West Haven, CT), 0.031 mg heparin, 0.012 g NaCl and 0.3 mmol Tris-HCl Buffer (pH 8.5), in a final volume of 5 mL. 5 μL buffer A were mixed with 5 μL serum from wild-type rats and 5 μL test sample, and incubated at 37°C for 60 min. [Note that serum from either Gpihbp1deficient mice or wild-type rats was pre-incubated for 10 min at 62.5°C in order to inactivate any residual endogenous lipase activity.] FFA concentration was determined in triplicate on a spectrophotometer (Thermo Multiskan GO) using the Wako kit, NEFA-HR (2).
In the case of human serum test sample, the FFA concentration represented the total post-heparin lipase activities that comprised LPL and hepatic lipase (HL) activities. To correct for the contribution from HL, 1 M NaCl was added and incubated for 60 min, so that the LPL activity can be completely inhibited [24]. LPL activity was then calculated by the difference between total post-heparin lipase activity and HL activity. All assays were performed in triplicate.

Plasmid construction and transfection
Human wild-type and c.352C > T mutant LPL coding sequences were synthesized and cloned into pcDNA3.1 (Vigene Biosciences), respectively. Sequence accuracy of the inserts was confirmed by Sanger sequencing.

Collation of previously published pathogenic LPL nonsense variants
Key words including "lipoprotein lipase", "mutation", "nonsense" and "termination" were used for searching previously reported pathogenic LPL nonsense variants in PubMed.

Clinical findings and treatment of the proband
At his latest bout of AP in 2017, the proband was transferred into our severe acute pancreatitis therapy center in Jinling Hospital. His TG level was 71.3 mmol/L   (Fig. 1) and his plasma was milky (Fig. 2a), fulfilling the definition of extreme HTG [25]. Physical examination revealed epigastric tenderness without rebound tenderness or Murphy's sign. Laboratory examination revealed elevations in amylase level (446 U/L), white blood cell count (19.86 × 10 9 /L), inflammation biomarkers CRP (263.4 mg/L), IL-6 (121.6 ng/L) and PCT (2.93 μg/L), and renal function biomarkers CRE (289 μmol/L) and BUN (13.3 mmol/L). Abdominal computed tomography showed evidence of AP (Fig. 2b).
Moderate-severe AP was diagnosed in accordance with the 2012 revision of Atlanta classification [5].
The proband was treated with enzyme inhibitors, antiinfection agents and fasting. His TG level decreased rapidly to 17.8 mmol/L (1576.19 mg/dL) 2 days later, and was determined to be 5.5 mmol/L (487.05 mg/dL) when discharged on May 2nd, 2017. During his hospitalization, the patient well tolerated enteral nutrition and adhered to a low-fat diet plus the lipid-lowering drug fenofibrate (0.3 g/d).
In most of the follow-up period, the patient has taken 0.3 g/d fenofibrate, low-fat diet, dry out and exercise (1 h /day), keeping the TG level within the mild to moderate range (defined as 2~9.9 mmol/L in accordance with Dron et al. [16]). However, once he took 0.1 g/d fenofibrate, an obvious increase in TG level was observed (Fig. 1).

Genetic findings
Sequencing of the LPL, APOA5, APOC2, LMF1 and GPIHBP1 genes in the proband detected four gene variants, as one LPL nonsense variant (Fig. 3) and three LMF1 synonymous variants (Supplemental. Fig. S1). All three LMF1 synonymous variants are common in the general populations (Table 1) and therefore were excluded from further consideration. The LPL nonsense variant, p.Gln118* (c.351C > T), is absent from the gno-mAD database ( Table 1). The LPL p.Gln118*was also detected in the proband's mother and sister (Fig. 3a), but not in others 256 unrelated HTG-AP patients. Additionally, the LPL p.Gln118* nonsense variant has not previously been reported. In vitro analysis of the LPL p.Gln118* variant LPL wild-type and p.Gln118* mutant expression plasmids were transiently transfected into HEK-293 T cells, respectively. LPL protein expression was analyzed by Western blot using transfected cell proteins and LPL activity was analyzed using transfected cell media. The LPL p.Gln118* mutant resulted in no detectable LPL activity and as shown in Fig. 4, the LPL p.Q118X resulted in undetectable LPL mass and activity, as compared to the positive results of LPL wild-type.

Discussion
In this study, we reported a novel heterozygous nonsense variant in the LPL gene, p.Gln118* (c.351C > T), in one typical Chinese family with HTG-AP history. Presumably, this variant should cause a complete functional loss of the affected LPL allele due to its significant truncation of the 475 amino acid protein. Indeed, the LPL mass and activity in the proband's plasma revealed a roughly 50% reduction as compared to normal controls. And in vitro, the results confirmed that the p.Gln118* mutant resulted in undetectable LPL protein and activity. Taken together, LPL p.Gln118* could be a novel and pathogenic LPL gene variant.
In this typical HTG-AP family, the proband, his mother and sister all had the LPL p.Gln118* nonsense variant, and separately had four, one and two times of HTG-AP onset. Moreover, HTG-AP was milder in the mother and older sister than the proband, as mild compared to moderate-severe. This variable clinical expression may be, at least partly, explained by the existence of an established secondary etiological factor, alcohol abuse, in the proband but not in the diseased mother and older sister. Although interplay between primary and secondary etiological factors in causing HTG-AP has been described in the literature [17,18], to our best knowledge, the present study is the first to demonstrate the possible effect of alcohol abuse in modifying the Fig. 4 Functional characterization of the LPL p.Gln118* variant. Proteins prepared from the HEK293T cells that were transiently transfected with the LPL wild-type (WT) expression vector, the LPL p.Gln118* mutant expression vector and empty vector (EV), respectively, were used for Western blot analysis. GAPDH, loading control clinical expression of a pathogenic genetic variant in the context of a HTG-AP family that exhibited a mode of monogenic inheritance. In this regard, it is pertinent to mention that, given the apparent effect of alcohol abuse in inducing or worsening HTG-AP, the proband has been required to quit alcohol since May 2017.
In this study, both in vivo and in vitro results showed that the LPL p.Gln118* nonsense variant could be pathogenic as resulting in complete functional loss of LPL mass and activity. Moreover, we briefly reviewed all the reported LPL nonsense variants together with the clinical features, as to evaluate the complex etiology of HTG or HTG-AP from a perspective of genotype and phenotype relationship. As shown in Table 2, the reported 12 LPL nonsense variants were detected in 20 subjects worldwide. As in detail, 8 were homozygotes, 2 were heterozygotes and 10 were compound heterozygotes. As expected, simple homozygotes showed no or barely detectable LPL mass and activity in the proband. The compound heterozygotes showed variable LPL mass and activity levels from zero to 50% of normal (depending upon the functional effect of the variant in trans). And, the heterozygotes showed LPL mass and activity levels that were around 50% of normal. We found correlation between mutation status and TG levels or occurrence of HTG-AP, as most informative adult patients had suffered recurrent AP with extreme high TG level, however, some patient is quite special. Taken p.Trp14* (c.42G > A) for example, the patient was had homozygote LPL variant, a 74years-old Japanese with a complete LPL deficiency, never developed AP whilst the compound heterozygote, a 61years-old Chinese whose LPL mass was~50% of normal, had three times of AP (Table 1). Above all, the zygosity and AP occurrence of all the nonsense patients reported before emphasized again the complex etiology and expression of HTG-AP. Findings from this comparative analysis provide new evidence suggesting that new layers of complexity, beyond known genetic risk factors, predispose to, or prevent, the development of HTG-AP. Nowadays, there are some new genetic drugs for these LPL deficient patients like alipogene tiparvovec, LCQ908 etc., beyond the dietary management and usual pharmacologic therapies, these new genetic treatments can be certainly promising and effective therapy basing on the patients' genetic background [44].

Conclusion
To sum up, in this report, a novel LPL p.Gln118* (c.351C > T) variant was detected in one typical Chinese family with HTG-AP history, and had been verified to be pathogenic as resulting complete loss of LPL function in vitro. In particular, we highlighted a role of alcohol abuse in modifying the clinical expression of the disease in the proband. Moreover, we briefed reviewed all reported LPL nonsense variants, together with the phenotype, which may give new insights into the complex etiology of HTG-AP.
Additional file 1: Figure S1. Sanger sequencing electropherograms showing the three heterozygous LMF1 synonymous variants detected in the proband.