Genetic and biochemical analysis of severe hypertriglyceridemia complicated with acute pancreatitis or with low post-heparin lipoprotein lipase mass

Abstract

Severe hypertriglyceridemia is a pathological condition caused by genetic factors alone or in combination with environmental factors, sometimes leading to acute pancreatitis (AP). In this study, exome sequencing and biochemical analyses were performed in 4 patients with hypertriglyceridemia complicated by obesity or diabetes with a history of AP or decreased post-heparin LPL mass. In a patient with a history of AP, SNP rs199953320 resulting in LMF1 nonsense mutation and APOE rs7412 causing apolipoprotein E2 were both found in heterozygous form. Three patients were homozygous for APOA5 rs2075291, and one was heterozygous. ELISA and Western blot analysis of the serum revealed the existence of apolipoprotein A-V in the lipoprotein-free fraction regardless of the presence or absence of rs2075291; furthermore, the molecular weight of apolipoprotein A-V was different depending on the class of lipoprotein or lipoprotein-free fraction. Lipidomics analysis showed increased serum levels of sphingomyelin and many classes of glycerophospholipid; however, when individual patients were compared, the degree of increase in each class of phospholipid among cases did not coincide with the increases seen in total cholesterol and triglycerides. Moreover, phosphatidylcholine, lysophosphatidylinositol, and sphingomyelin levels tended to be higher in patients who experienced AP than those who did not, suggesting that these phospholipids may contribute to the onset of AP. In summary, this study revealed a new disease-causing gene mutation in LMF1, confirmed an association between overlapping of multiple gene mutations and severe hypertriglyceridemia, and suggested that some classes of phospholipid may be involved in the pathogenesis of AP.

HYPERTRIGLYCERIDEMIA is a metabolic disorder characterized by elevated plasma triglyceride levels. Epidemiological data have shown that hypertriglyceridemia is associated with atherosclerotic cardiovascular diseases (ASCVD) [1]. Severe hypertriglyceridemia, with triglycerides around 1,000 mg/dL or higher, has also been associated with acute pancreatitis (AP) [2, 3].

Recent advances in researching the causes of primary hyperchylomicronemia revealed the 5 candidate genes or proteins associated with the disease [4]. These include lipoprotein lipase (LPL), apolipoprotein C-II (apoCII), apolipoprotein A-V (apoAV), glycosylphosphatidylinositol-anchored high-density lipoprotein-binding protein 1 (GPIHBP1), and lipase maturation factor 1 (LMF1). LPL plays a critical role in triglyceride hydrolysis, and apoCII and apoAV activate LPL function. LMF1 is involved in the normal processing of LPL in the parenchymal cells (myocytes and adipocytes), and GPIHBP1 is critical for the transport of LPL across the endothelial cells (ECs) and for anchoring LPL on the EC luminal surface. In addition to the mutations causing primary hyperchylomicronemia in these five genes, genome-wide association studies (GWASs) have also shown that several SNPs in other genes are associated with triglyceride levels in the general populations.

In the present study, we performed exome-sequencing analyses in 4 severely hyper-triglyceridemic patients who have recurrent episodes of AP or showed reduced post-heparin LPL mass. We also conducted biochemical and lipidemic analyses of the serum samples to gain insight into the mechanism of mutant apoAV protein for causing hypertriglyceridemia, and to elucidate the changes in the phospholipid classes, some of which possess biological properties and have been found to be altered in human inflammation-associated diseases [5-12].

Materials and Methods

Subjects of this study

The present study was conducted according to the Declaration of Helsinki. The study protocol was approved by the Institutional Research Ethics committee of Teikyo University (Approval number: 20-007). The subjects of this study were patients who underwent outpatient or inpatient care at Teikyo University hospital between February 2021 and January 2022, with severe hypertriglyceridemia complicated with recurrent acute pancreatitis episode or with low post-heparin LPL mass. Four patients were selected by the screening and included in this study. The aim of the study was explained to all participants, and written informed consent was obtained from each participant prior to participation. Informed consent was also obtained from normal controls.

The detailed clinical course of each patient is provided in the supplemental file (Supplementary Text). Briefly stating, Cases 1, 2 and 4 showed low post-heparin LPL mass, and Cases 3 and 4 had a history of recurrent AP. Cases 1, 2 and 3 are obese and Cases 2, 3 and 4 are diabetic. The lipid and apolipoprotein profiles of patients at the time of sample collection are summarized in Table 1, together with the post-heparin LPL mass value, the presence or absence of past history of AP, the presence or absence of diabetes mellitus, BMI, and diabetes-related clinical data.

Table 1

Clinical profiles of the 4 analyzed cases of severe hypertriglyceridemia

Case	TC (mg/dL)	TG (mg/dL)	HDL-C (mg/dL)	ApoAI (mg/dL) (126–165)	ApoAII (mg/dL) (24.6–33.3)	ApoB (mg/dL) (66–101)	ApoCII (mg/dL) (1.5–3.8)	ApoCIII (mg/dL) (5.4–9)	ApoE (mg/dL) (2.8–4.6)	ApoAV (ng/mL) (110–397)	PH-LPL (ng/mL) (164–284)
1	405	3,051	28	112	26.3	75	18.9	42.8	29.2	389.2	73
2	1,440	12,020	25	93	21.7	188	78.8	235.8	117.5	254	112
3	249	621	34	110	31.1	158	10.1	20.5	10.2	783.1	190
4	324	1,589	24	75	18.5	128	22.4	24.9	13.8	138.9	108
Case	Past history of AP	Diabetes Mellitus	Insulin Therapy	BMI	HbA1c (%)	FBS (mg/dL)	CPR (ng/mL)
1	Absent	Absent	No	30.1	5.3	85	NM
2	Absent	Present	Yes	35.2	12.4	329	1.6
3	Present	Present	Yes	29.3	9.8	254	2.2
4	Present	Present	Yes	24.1	9.4	168	3.1

Lipid and apolipoprotein levels at the time of sample collection are presented. Levels of post-heparin LPL mass (PH-LPL), the presence or absence of a past history of AP, the presence or absence of diabetes mellitus, BMI, and diabetes-related clinical data are also presented. AP, acute pancreatitis; TC, total cholesterol; TG, triglycerides; HDL-C, HDL cholesterol; apo, apolipoprotein; BMI, body mass index; FBS, fasting blood glucose; CPR, C-peptide immunoreactivity; NM, not measured. Reference levels of apolipoproteins and PH-LPL are shown in parentheses under the units.

Sample collection

Blood samples were obtained after an overnight fast. Leucocytes were subjected to genomic DNA extraction, and serum samples were used for biochemical and lipidomics analyses. Serum samples were also subjected to sequential density gradient ultracentrifugation [13], and the resulting lipoprotein fractions and lipoprotein-free fractions were collected and used for biochemical analyses, as were serum samples from normolipidemic 3 controls.

Exome sequence analysis

Genomic DNA was extracted from whole blood using NucleoSpin^®Blood Kit (Macherey-Nagel Inc. PA, USA). After confirming sample quality, the DNA was sent to Novogene Co. Ltd. (Beijing, China) for exome sequencing. Briefly, the genomic DNA was randomly sheared into short fragments ranging in size from 180 to 280 bp, and those fragments were used to construct a library for multiplexed paired-end sequencing using the SureSelect^XT Reagent Kit (Agilent Technologies, CA, USA). That library was then hybridized to biotinylated cRNA oligonucleotide baits from the SureSelect Human All Exome V6 Kit (Agilent Technologies) for target enrichment. Targeted sequences were purified by magnetic beads with streptavidin, amplified, and sequenced on an Illumina platform using a NovaSeq 6000 system.

After the quality control of paired sequencing reads, Burrows-Wheeler Aligner (BWA, ver.0.7.17) [14] was utilized to map the paired-end clean reads to the human reference genome [15]. The original mapping result was obtained in BAM format, with SAMtools (ver.1.8) used for sorting the BAM files, and Picard (ver. 2.18.9) [16] utilized to mark duplicate reads. GATK (ver. 4.0) [17] was used for detecting and filtering SNPs/InDels. Following genomic variant detection, variants were annotated with the ANNOVAR tool [18] in multiple aspects, including protein coding changes, genomic regions affected by the variants, allele frequency, deleteriousness prediction, etc.

Afterwards, exonic mutations of the genes whose association with plasma triglyceride levels were proposed in the GWAS catalogue [19] (Table 2) were extracted, and mutations other than synonymous single nucleotide variants (SNVs) were chosen, followed by selection using SIFT score less than 0.05 or “not determined.” Among the selected mutations, we further selected those having a Polyphen2 score greater than 0.908 or “not determined” and a frequency less than 0.2 or “not known” within the “gnomAD (The Genome Aggregation Database) [20] exome AF.”

Table 2

List of genes reported to be associated with TG levels in GWAS catalogue and analyzed in the present study

A1CF	APOH	FRMD5	MLXIPL
ABCA1	ARL15	GALNT2	NCAN
ADH4	BUD13	GCK	NPC1
AFF1	CAPN3	GCKR	PCK1
ALDH1A2	CD300LG	GPIHBP1	PCSK6
ALDH2	CD33	HGFAC	PCSK7
ANGPTL3	CETP	INHBC	PDXDC1
ANGPTL4	CILP2	INPP5A	PGS1
ANGPTL8	CMIP	INSR	RSPO3
APOA4	COBLL1	JMJD1C	SIK3
APOA5	COL18A1	LIPC	SLC22A1
APOB	CTCFL	LMF1	SLC39A8
APOC1	DLEU1	LPA	SLCO1B1
APOC2	DOCK6	LRP1	TM6SF2
APOC3	DOCK7	LRPAP1	TRIB1
APOC4	FADS1	MAP1A	ZNF101
APOE	FADS2	MAU2	ZPR1

Confirmation of the mutations with direct sequencing

In order to confirm the mutations obtained in exome sequence analysis, PCR amplification of gene regions of interest was performed with the genomic DNA purified as described above, and the PCR products were directly sequenced using Sanger sequencing methods (Fasmac Co., Ltd., Kanagawa, Japan). The list of primers used for amplification and sequencing is shown in Supplementary Table 1.

Analysis of apoAV with ELISA and Western blot

ApoAV levels in the serum, along with lipoprotein fractions and lipoprotein-free fractions, were measured using an Enzyme-Linked Immuno Sorbent Assay (ELISA) kit (Immuno-Biological Lab. Co. Ltd., Gunma, Japan). To explore the distribution of apoAV protein among lipoprotein and lipoprotein-free fractions, apoAV protein mass per mL serum in each fraction was calculated utilizing the volume of each fraction collected. For Western blotting, aliquots of serum and lipoprotein-fractionated samples were subjected to 10% SDS-PAGE, and the proteins were transferred to a PVDF membrane. ApoAV was detected using a rabbit polyclonal antibody to human apoAV (Proteintech Group Inc, IL, USA) and peroxidase-labelled goat anti-rabbit IgG antibody (GE Healthcare Life Sciences, MA, USA), followed by visualization with the ECL reagent (GE Healthcare) and the LAS-4000mini imager (GE Healthcare).

Lipidomics analysis of serum samples

To elucidate the changes in the serum phospholipid component in these patients, the serum samples were subjected to lipidomics analysis for glycerophospholipid and sphingophospholipid, including lysophosphatidylcholine (LPC), lysophosphatidylserine (LPS), lysophosphatidylethanolamine (LPE), lysophosphatidylinositol (LPI), lysophosphatidylglycerol (LPG), phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylinositol (PI), phosphatidylglycerol (PG), and sphingomyelin (SM), using two independent LC-MS/MS methods with an LC8060 system consisting of a quantum ultra-triple quadrupole mass spectrometer (Shimadzu, Japan). We have described these methods in detail in previous papers [21, 22]. After measurement, the same classes of phospholipids were summed up to calculate total serum levels of glycerophospholipids. We utilized the values obtained in our previous experiment as control values [10] and managed those control values by presenting data separately for subjects without diabetes (n = 28) and with diabetes (n = 97).

Results

Genetic mutations found in the genes associated with triglyceride levels in GWAS

The gene mutations found in whole-exome sequencing and selected by the criteria described in the Materials and Methods section are listed in Table 3. These included mutations in apolipoproteins crucial for lipid metabolism as well as mutations in genes known to cause familial hyperchylomicronemia. For comparison, Table 4 shows the general population frequencies of these SNPs, obtained by accessing data from databases of whole genome analysis for global, Japanese, Asian, and European populations (databases: gnomAD [20], 14KJPN [23], The PAGE Study [24], 1000 Genomes [25]).

Table 3

Summary of genetic variants identified in 4 cases

Gene	SNP	Mutation	SIFT	Polyphen2	Form of inheritance
					Case 1	Case 2	Case 3	Case 4
ABCA1	rs374190304	c.1913G>A, p.R638Q	0.018, D	0.998, D	—	—	—	Hetero
APOA5	rs2075291	c.553G>T, p.G185C	0.021, D	0.999, D	Homo	Hetero	Homo	Homo
APOB	rs13306194	c.1594C>T, p.R532W	0.0, D	1.0, D	—	—	—	Hetero
APOE	rs7412	c.526C>T, p.R176C	0.012, D	1.0, D	—	—	—	Hetero
APOH	rs1801690	c.1004G>C, p.W335S	0.022, D	0.999, D	Hetero	—	—	—
GCKR	rs146175795	c.307G>A, p.V103M	0.004, D	1.0, D	—	Hetero	—	—
GPIHBP1	rs11538389	c.41G>T, p.C14F	*	*	Hetero	—	Hetero	Hetero
JMJD1C	rs139722368	c.3473_3478del, p.Gly1158_Leu1159del	*	*	—	Hetero	Hetero	—
LMF1	rs199953320	c.697C>T, p.R233X	*	*	—	—	—	Hetero
LRP1	rs79435985	c.12161A>T, p.Y4054F	0.016, D	0.999, D	—	Hetero	—	—
MAP1A	rs148851436	c.4134GGT [1], p.Val1380del	*	*	Hetero	—	—	Hetero
SLC22A1	rs2282143	c.1022C>T, p.P341L	0.001, D	0.999, D	Hetero	—	—	—
SLCO1B1	rs4149056	c.521T>C, p.V174A	0.001, D	1.0, D	Homo	—	—	—

The mutations of genes selected using the criteria described in the “Materials and Methods” are listed here. On the left side of the panel, Gene name, rs (reference SNP ID) number, changes in nucleotide and amino acid residue, SIFT and Polyphen2 scores and prediction, and form of inheritance in each case are shown. D (in SIFT), Deleterious; D (in Polyphen2), Damaging; *, not determined; —, not found; Homo, homozygote; Hetero, heterozygote.

Table 4

Frequencies of genetic variants shown in Table 3 Data from databases of whole genome analyses are indicated.

Gene	SNP	Variant frequencies
		Population	Global			Japanese	Asian		European
		Project Name	gnomAD_ Exomes_AF	gnomAD_ Genomes	The PAGE Study	14KJPN	gnomAD_ Exomes	The PAGE Study	1000 Genomes
		Sample size	136,984– 251,280	128,718– 140,266	78,416– 78,698	28,248– 28,258	31,910– 49,010	8,310– 8,318	1,006
ABCA1	rs374190304		0.0019	0.000071	0.00011	0.00004	0.00941	0.0000	0.0000
APOA5	rs2075291		0.0064	ND	0.01016	0.07528	ND	0.0713	0.001
APOB	rs13306194		0.0113	0.00441	0.01788	0.12145	0.05132	0.1373	0.002
APOE	rs7412		0.0615	0.079856	ND	0.04315	0.0525	ND	0.0626
APOH	rs1801690		0.0479	0.039596	0.0286	0.10949	0.05453	0.0961	0.0567
GCKR	rs146175795		0.0027	0.001512	0.00651	0.01709	0.00322	0.015	0.0000
GPIHBP1	rs11538389		ND	0.104067	0.15402	0.32644	ND	0.3248	0.0932
JMJD1C	rs139722368		0.0533	0.032516	0.06265	0.19396	0.10984	0.2049	0.0179
LMF1	rs199953320		5.06E-05	0.000007	ND	0.00035	0.00015	ND	ND
LRP1	rs79435985		0.002	0.000663	0.00283	0.02750	0.00994	0.0226	0.0000
MAP1A	rs148851436		0.0104	0.00678	0.00936	0.07141	0.02271	0.0674	0.005
SLC22A1	rs2282143		0.0376	ND	0.06242	0.16431	ND	0.1511	0.0089
SLCO1B1	rs4149056		0.1329	0.121061	ND	0.14424	0.07884	ND	0.161

ND, data not shown.

Regarding APOA5, Cases 1, 3, and 4 were homozygous, and Case 2 was heterozygous, for reference SNP ID (rs)2075291 (c.G553T). Case 4 was heterozygous for LMF1 rs199953320 (c.C697T), which causes early termination of the LMF1 protein, and was also heterozygous for SNP rs7412 (c.C526T) in the APOE gene, which confers the apoE2 isoform [26], and for rs13306194 (c.C1594T) in APOB. GPIHBP1 rs11538389 (c.T41G) was found in Cases 1, 3 and 4 in the heterozygote form. These genetic mutations were confirmed by direct sequencing analysis (Supplementary Fig. 1).

Several SNPs in other genes, primarily in the heterozygous form, were also identified (Table 3). These genes include ABCA1, APOH, GCKR, JMJDIC, LRP1, MAP1A, SLC22A1 and SLCO1B1. No mutations or polymorphisms in APOC2 or LPL were found in any of the four patients.

Analysis of apoAV protein with ELISA and Western blot

The concentration of apoAV protein was measured by ELISA, using whole and lipoprotein-fractionated serum samples, and Western blotting analysis was also performed on those same samples. ELISA analysis showed that apoAV was detected not only in the lipoprotein fractions but also in the lipoprotein-free fraction in our four patients carrying APOA5 rs2075291, consistent with prior research [27]; however, the existence of apoAV in the lipoprotein-free fraction was also noted in normal subjects who did not carry SNP rs2075291 (Table 5).

Table 5

Distribution of apoAV protein in lipoprotein and lipoprotein-free fractions

		CM	VLDL	HDL	LPP-F	Serum level (ng/mL)
Case 1	mass (ng/mL serum)	53.1	27.4	46.9	50.3	389.2
	Distribution among LPPs	29.9%	15.4%	26.4%	28.3%
Case 2	mass (ng/mL serum)	77.2		8.8	28.2	254.0
	Distribution among LPPs	67.6%		7.7%	24.7%
Case 3	mass (ng/mL serum)	179.9	106.5	84.9	97.8	783.1
	Distribution among LPPs	38.3%	22.7%	18.1%	20.8%
Case 4	mass (ng/mL serum)	13.5		0.4	22.8	138.9
	Distribution among LPPs	36.8%		1.0%	62.2%
Cont 1	mass (ng/mL serum)			96.8	23.4	197.9
	Distribution among LPPs			80.5%	19.5%
Cont 2	mass (ng/mL serum)			71.7	16.9	77.9
	Distribution among LPPs			80.9%	19.1%
Cont 3 (before meals)	mass (ng/mL serum)	5.85		175.87	56.61	415.7
	Distribution among LPPs	2.5%		73.8%	23.8%
Cont 3 (after meals)	mass (ng/mL serum)		2.19	212.91	46.25	513.7
	Distribution among LPPs		0.8%	81.5%	17.7%

After measuring the apoAV concentration in each lipoprotein and lipoprotein-free fraction with ELISA, apoAV protein mass in each fraction per mL serum was calculated utilizing the volume of each fraction collected. From these values, the distribution of apoAV protein among lipoprotein and lipoprotein-free fractions was calculated and expressed as percentages. ApoAV levels measured with whole serum were also indicated. LPP, lipoprotein; LPP-F, lipoprotein-free fraction.

The results of Western blot analysis are shown in Fig. 1. The analysis of whole serum samples revealed that the apoAV protein was detected at molecular weight positions around 40 kDa and 50 kDa in all four patients and in the control subjects. When each lipoprotein fraction and lipoprotein-free fraction was analyzed, a 50 kDa band was detected in the chylomicron and lipoprotein-free fractions in all patients, and a 40 kDa band was detected in VLDL and HDL. The same findings were noted in the control subjects: a band of 50 kDa was detected in the chylomicron and lipoprotein-free fractions, and a band of 40 kDa in HDL. Western blot analysis performed without the first apoAV antibody but with the secondary antibody showed no band in all lipoproteins and lipoprotein-free fractions (data not shown).

Fig. 1

Western blot analysis of apoAV protein in serum and lipoprotein-fractionated samples

Western blotting analyses were performed utilizing whole serum (A) or lipoprotein- fractionated samples (B). A: 1 μL of whole serum was subjected to 10% SDS-PAGE, protein was transferred to a PVDF membrane, and apoAV was detected by a monoclonal antibody to human apoAV. B: Lipoprotein and lipoprotein-free fractions, containing 2 μg protein, were subjected to Western blotting analysis in the same way as described in (A). Only fractions in which apoAV protein was detected with ELISA were utilized for the analysis. CM, chylomicron; VLDL, very low density lipoprotein; HDL, high density lipoprotein; LPP-F, lipoprotein-free fraction.

Lipidomics analysis

Fig. 2 illustrates the lipidomics analysis. Briefly, whole serum samples were analyzed, and the values obtained for the same classes of phospholipid and lyso-phospholipid were summed up. Among the monitored phospholipids, serum levels of PC, PS, PE, LPE, PG, PI, and SM deviated notably upwards from the reference range of both healthy and diabetic populations. Interestingly, as shown in Table 1, Case 1 had the highest values for both serum TC and triglycerides among the four patients, followed sequentially by Cases 2, 3, and 4. This sequence did not coincide with the sequence for any class of phospholipid, indicating that changes in the level of any phospholipid species will differ from one individual to another, even under similar conditions of hypertriglyceridemia. In addition, we noted marked elevation of serum LPC in Case 4 and elevation of LPS and LPG in Cases 2 and 3. Cases 3 and 4, both of whom had a history of pancreatitis, showed higher serum LPI, PC, and SM than the controls, Case 1, or Case 2. Serum PG, LPG, and LPI levels were higher in Case 3 than in the other patients, and Case 4 showed higher PE and LPE levels than any of the other cases or controls.

Fig. 2

Dynamic modulation of serum phospholipids in 4 patients with hypertriglyceridemia

Serum levels of phospholipids were determined using LC-MS/MS methods as described in the “Materials and Methods.” As a control, the distributions of values for subjects without diabetes (n = 28) as well as those with diabetes (n = 87), as described in Ref. 10, are shown as violin plots. (A) lysophosphatidylcholine; (B) phosphatidylcholine; (C) lysophosphatidylserine; (D) phosphatidylserine; (E) phosphatidylethanolamine; (F) lysophosphatidylethanolamine; (G) phosphatidylglycerol; (H) lysophosphatidylglycerol; (I) phosphatidylinositol; (J) lysophosphatidylinositol; (K) sphingomyelin.

Discussion

Hyperchylomicronemia is a group of diseases that may be caused by genetic factors alone or by the combination of comorbid disease with genetic factors. In the present study, we investigated genetic factors using exome sequencing analysis in four hyperchylomicronemic patients who showed low LPL protein levels after intravenous heparin infusion and/or had a history of AP. Lipidomics analysis and apoAV protein analysis were also performed.

Our study found that Case 4 was heterozygous for LMF1 rs199953320, a novel mutation not previously reported in relation to hyperchylomicronemia but noted at extremely low frequency levels in genome projects. LMF1 rs199953320 is a mutation that causes termination of the protein between transmembrane domain (TM)3 and TM4 [28]. This termination results in a truncated LMF1 protein having only 232 amino acids and lacking the C-terminal portion that has been shown to play a crucial role in proper LPL protein maturation [28]. Previously, researchers have reported that the variants LMF1 (Y439X) [29] and LMF1 (W464X) [30], disease-causing mutations that terminate the protein at a later point in the LMF1 protein than for the rs199953320 mutation, caused loss of potency in LPL secretions from cells in vitro [30]. Those findings suggest that the rs199953320 mutation would surely interfere with normal processing of the LPL protein, similar to the effects of LMF1 (Y439X) and LMF1 (W464X). Notably, the patient who was homozygous for the LMF1 (Y439X) mutation was reported to be resistant to treatment, whereas the patient who was heterozygous for the LMF1 (W464X) mutation was reported to be relatively responsive to treatment. In addition to LMF1 rs199953320, Case 4 was also heterozygous for apoE2, which lacks lipoprotein receptor binding capacity and is associated with type III dyslipidemia [31] and postprandial hyperlipidemia [32], and was homozygous for the APOA5 rs2075291 (role of rs2075291 is described below). Together with these multiple genetic factors, the environmental factor of diabetes might have made the control of dyslipidemia difficult in Case 4, and this patient also harbored a heterozygous APOB rs13306194 mutation, an SNP in the common sequence of apoB48 and apoB100. ApoB48 is a structural protein of chylomicron, and it is possible that this mutation may affect triglyceride hydrolysis or on the uptake of chylomicron remnants in the liver. However, triglyceride levels in patients with familial hypercholesterolemia and APOB rs13306194 have been reported as no higher than in patients who do not harbor this SNP [33]. Future research might clarify the significance of this mutation.

Regarding GPIHBP1 rs11538389, which encodes an amino acid residue in the part of the leader sequence and may decrease the LPL protein level by half [34], was found in the heterozygote form in three patients. However, the frequency of rs11538389 is not so low, especially in Japanese and Asian (Table 4), suggesting that this SNP may contribute less to hyperchylomicronemia than other SNPs such as LMF1 rs199953320 and APOA5 rs2075291.

It should be noted that APOA5 rs2075291, which has been reported to be associated with high triglyceride levels or hyperchylomicronemia [35-42], was observed as homozygous in three patients and heterozygous in one. The frequency of this mutation is very low in European populations (around 0.1%), but higher among Japanese and Asians (around 7%), suggesting that many Japanese and Asian patients with severe hypertriglyceridemia may have this SNP.

In addition to these genetic mutations, Cases 1, 2, and 3 were obese, and Cases 2, 3, and 4 had poorly controlled diabetes mellitus, as shown in Table 1 and Supplementary Text (detailed clinical course). These metabolic factors are known to contribute to hypertriglyceridemia, and their contribution to hypertriglyceridemia has also been clarified in previous reports in relation to APOA5 rs2075291 [40]. Therefore, these metabolic factors probably contributed to the marked hyperchylomicronemia in all four of our patients.

Exome sequencing analysis also detected SNPs of several other genes associated with triglyceride levels in GWAS. These genes are reportedly associated with other diseases or conditions, but the literature has not yet reported associations between these genes and hyperchylomicronemia.

The variant apoAV (G185C) protein has been shown to retard LPL activity [43]. Sharma et al. proposed that replacement with cysteine residue led to the formation of aberrant hetero-disulfide bonds, which forced the variant apoAV (G185C) to detach from the lipoprotein and appear in lipoprotein-free fractions, resulting in the modulation of triglyceride metabolism [27]. However, as shown in Table 5 and Fig. 1, apoAV protein was detected in the lipoprotein-free fraction even in normal subjects possessing wild type apoAV. Further research is needed to explore the precise mechanism of apoAV (G185C) for causing hyperchylomicronemia. The present study also clarified that apoAV on chylomicron and in lipid-free fractions was detected as a protein approximately 50 kDa in size, while on VLDL and HDL particles it was detected as a 40 kDa protein, indicating that the modification of the protein depends on the fraction in which apoAV is present. This finding might provide a clue for elucidating the mechanism by which APOA5 rs2075291 causes hypertriglyceridemia.

Another new finding in this study is that even in normal control subjects, chylomicrons are present in the fasted state, and this presence can be detected by the presence of apoAV protein, although the cholesterol and triglycerides levels were too low to measure in chylomicron fraction.

A previous study in mice showed that apoAV is more distributed in triglycerides-rich lipoproteins after oral lipid loading or with increasing plasma triglycerides levels [44]. We also tried to investigate this phenomenon in Case 4 by analyzing a new sample obtained at a different time point, but could not clearly show the redistribution of apoAV, probably due to the persistent extremely high triglycerides level in Case 4 (data not shown). However, as shown in Table 5, the distribution of apoAV among lipoprotein classes is quite different between cases and normal controls, indicating that apoAV would be more distributed in triglycerides-rich lipoproteins in hypertriglyceridemic cases.

Severe hypertriglyceridemia is a risk for AP; fatty acid toxicity and impaired blood flow at the capillary level have been proposed as mechanisms by which pancreatitis develops in hypertriglyceridemic patients [45], but these have not been fully elucidated. Furthermore, not all severe hypertriglyceridemia patients develop the complication of AP, and there may be some differences between those who develop AP and those do not. In addition, modulations of phospholipids in severe hypertriglyceridemia have not yet been elucidated in detail. In the present study, although the number of patients was limited, we analyzed phospholipid levels in each patient, noted interesting elevations in many phospholipid classes, and detected differences in the levels of some classes of phospholipid (i.e., PC, LPI, SM) between those who experienced AP and those who did not.

As for LPC and PC, some studies have reported an association between LPC and AP; LPC activated nuclear factor-kappaB and activator protein 1 in pancreatic cells, leading to apoptosis [46, 47]. Enhancement or inhibition of phospholipase A-2, an enzyme that catalyzes the synthesis of LPC from PC, reportedly induced or suppressed, respectively, AP in animal models [48-50]. Furthermore, LPC not only has proinflammatory properties [51] but is also further converted by autotaxin to LPA, a molecule with proinflammatory functions [52]. In our study, the LPC level was higher in Case 4, and the precursor PC was particularly elevated in Cases 3 and 4, leading us to speculate that these molecules may be responsible for AP.

Regarding LPI/PI axis, although serum PI levels were elevated in all four patients, serum LPI levels were higher than controls only in Cases 3 and 4, both of whom had a history of AP. Recently, we have demonstrated that LPI can induce inflammatory responses in macrophages [53] and another group has reported that LPI can activate endothelial cells [54], suggesting that higher LPI levels might contribute to susceptibility to pancreatitis. In addition, LPI could induce GLP-1 [55, 56], which is also suspected of being associated with pancreatitis [57].

Along with LPI and PC, SM showed remarkable increases, especially in Cases 3 and 4. SM itself mainly forms cell membranes and is considered to have no biological activity. However, its products, ceramide and sphingosine 1-phosphate, have been associated with inflammation and thrombosis [58-61], and alterations in sphingolipid metabolism have been reported in AP [62]. Further research in more patients is absolutely needed, and the associations between these phospholipids and AP should be investigated.

For a better understanding of this study, the known or newly identified causes of severe hypertriglyceridemia and hypertriglyceridemia-related pancreatitis are shown schematically in Fig. 3. In summary, we found a new pathogenic gene mutation of hyperchylomicronemia in LMF1 and confirmed that multiple gene mutations together with environmental factors can cause hyperchylomicronemia. Although the precise molecular mechanism of variant apoAV (G185C) for this condition needs to be further elucidated, it is anticipated that this variant will be highly prevalent among Japanese patients with hyperchylomicronemia. Some phospholipids such as LPC, PC, LPI, and SM might be responsible for AP in hyperchylomicronemia. Future large-scale case studies as well as basic research will be needed to prove this hypothesis.

Fig. 3

Causes of severe hypertriglyceridemia and hypertriglyceridemia-related pancreatitis

Amino acids substitutions found in this study are indicated in parentheses. Novel findings obtained in this study are indicated by red letters. Dotted arrows indicate possible associations. PC, phosphatidylcholine; LPC, lysophosphatidylcholine; LPI, lysophosphatidylinositol; SM, sphingomyelin.

Abbreviations

ASCVD, atherosclerotic cardiovascular diseases; AP, acute pancreatitis; LPL, lipoprotein lipase; ApoCII, apolipoprotein C-II; ApoAV, apolipoprotein A-V; GPIHBP1, glycosylphosphatidylinositol-anchored high-density lipoprotein-binding protein 1; LMF1, lipase maturation factor 1; EC, endothelial cell; GWAS, genome-wide association study; SNP, single-nucleotide polymorphism; rs, reference SNP ID; RLP, remnant-like particle; VLDL, very low density lipoprotein; BMI, body mass index; ELISA, Enzyme-Linked Immuno Sorbent Assay; LC-MS/MS, liquid chromatography-mass spectrometry; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PI, phosphatidylinositol; PS, phosphatidylserine; LPC, lysophosphatidylcholine; LPE, lysophosphatidylethanolamine; LPG, lysophosphatidylglycerol; LPI, lysophosphatidylinositol; LPS, lysophosphatidylserine; SM, sphingomyelin; SD, standard deviation; TM, transmembrane domain

Acknowledgement

We are deeply grateful to Ms. Kazue Murata for her assistance in the experiments.

Disclosure of Conflicts of Interests

The authors have declared that no conflict of interest exists. Kazuhisa Tsukamoto is a member of Endocrine Journal’s Editorial Board.

Author Contributions

TSu, KU and KT designed the research; AI, TU, HT, DM, RS and TSe collected samples; TSu, MK, YS, KU, NE, TI and KT performed the research; TSu, MK, KU-E, KM, TI and KT analyzed data; TSu, MK, KU-E, KU, KM, TI and KT wrote the manuscript. All authors read and approved the final version of the manuscript for submission.

Funding

This work was supported by scholarship donation from Kowa Company, Ltd. and Taisho Pharmaceutical.

Supplementary Table 1

Sequence of primers used for PCR amplification and direct sequencing

Gene	rs	Forward Primer (5' -> 3')	Reverse Primer (5' -> 3')
APOA5	rs2075291	ACCAGGCTCTCGGCGTAT	ACTGAAGCCCTACACGATGG
APOE	rs7412	GCACGGCTGTCCAAGGAGCTGCAGGC	GGCGCTCGCGGATGGCGCTGAG
LMF1	rs199953320	CACGCTCACCTCATAGTGGA	CTTCGTGGATGGTTCGTCTT
APOB	rs13306194	TAGTACCTTCCAAATCCTTGT	AGAACTCAAGTCTTCAATCCT
GPIHBP1	rs11538389	CAGAGTCAGGGACACAGCAG	TCCCCTCCTTCTTCCTAAGC

Supplementary Fig. 1

Direct sequencing analysis of the SNP found in exome sequencing analysis

Polymorphic bases are indicated with red circles and black arrows. (A) APOA5 rs2075291; (B) APOE rs7412; (C) LMF1 rs199953320; (D) APOB rs13306194; (E) GPIHBP1 rs11538389

Supplementary Text: Clinical Course of Patients

Case 1: 37-year-old female, body mass index (BMI) 30.1 kg/m²

The patient had been obese since childhood. She was told she had hypertriglyceridemia at every health check-up after reaching adulthood but did not follow up with a physician. She is a social drinker. In 2020, her general physician reported triglycerides at 4,092 mg/dL and prescribed pemafibrate, which reduced triglycerides to around 400 mg/dL. However, the patient hoped to undergo infertility treatment and came to our hospital to pursue alternative therapy without medication.

We found no signs of eruptive xanthoma, and there was no history of AP. Lipoproteins, analyzed when the patient’s triglycerides were at 3,051 mg/dL and remnant-like particle (RLP) cholesterol was at 114 mg/dL, showed increases in chylomicron and very low density lipoprotein (VLDL). Strict fat restriction of 20 g/day brought the patient’s triglycerides down into the approximate range of 472 to 995 mg/dL without medication. Post-heparin LPL mass was 73 ng/mL (reference level: 164–284).

Case 2: 19-year-old female, BMI 35.2

The patient had suffered from epilepsy and autism since childhood. Medical history showed no episodes of AP. In 2020, the patient visited a dermatology department for an ingrown toenail. Eruptive xanthoma was noted in the extremities and abdomen, and blood tests in April 2021 showed triglycerides at 13,504 mg/dL and uncontrolled diabetes with HbA1c at 12.4%. Parenteral infusion therapy was initiated under fasting conditions, together with insulin therapy, but even after 5 days’ fasting, triglycerides declined only to 5,369 mg/dL. When the patient was switched to an oral diet, bezafibrate therapy was initiated. Fat consumption was initially limited to 5 g/day and gradually increased to 20 g/day. Triglycerides decreased to 475 mg/dL one month after the initiation of bezafibrate. Lipoprotein analysis showed increased chylomicron and VLDL, and post-heparin LPL mass was 73 ng/mL.

Case 3. 47-year-old female, BMI 29.3

The patient had experienced episodes of abdominal pain since childhood. In her 30s, when living in the United States, she was hospitalized for AP. There she was diagnosed with diabetes and started on insulin therapy. After returning to Japan in 2016, she was hospitalized twice for AP. A letter introducing her medical history since 2019 showed triglycerides ranging from 319 to 1,998 mg/dL under treatment with pemafibrate, ethyl icosapentate, and pitavastatin. She was referred to our hospital for further evaluation in April 2021, and dietary fat was restricted to approximately 20 to 30 g/day in addition to treatment with the prescription drugs noted above. Under this treatment regimen, triglycerides ranged from 483 to 689 mg/dL. Glycemic control has been poor since 2019, with HbA1c values ranging from 7.9% to 12.1% even with multiple insulin injections. Lipoprotein analysis showed increases in chylomicron and VLDL, and post-heparin LPL mass was 190 ng/mL. She stated that her parents were not consanguineous but that both were hypertriglyceridemic and that her father had experienced AP.

Case 4: 58-year-old male, BMI 24.1

Findings of dyslipidemia were first mentioned to the patient when he was about 40 years of age, but no treatment was initiated until his first AP attack in 2011, after which dyslipidemia therapy was initiated. In 2013 he was referred to our hospital for AP, where he was also diagnosed with diabetes mellitus (HbA1c 7.5%) and hypertriglyceridemia (TG 1,589 mg/dL under therapy with bezafibrate and niceritrol), and treatment with glimepiride was initiated. The patient stopped regular alcohol intake from 2013 but continued to experience AP events one to four times a year. Even under medication with bezafibrate, tocopherol nicotinate, ethyl icosapentetate, and atorvastatin, together with dietary fat restrictions, his peak triglyceride value reached 5,034 mg/dL. In 2019, when the patient’s HbA1c was at 11.0%, insulin therapy was initiated for glycemic control. One week of fasting for treatment of AP in 2021 provided only a reduction to 494 mg/dL in triglyceride levels. The post-heparin LPL mass was 108 ng/mL. The lipoprotein analysis revealed an increase in chylomicron and VLDL. No eruptive xanthoma was noted.

References

• 1   Nordestgaard  BG (2016) Triglyceride-rich lipoproteins and atherosclerotic cardiovascular disease: new insights from epidemiology, genetics, and biology. Circ Res 118: 547–563.
• 2   Fortson  MR,  Freedman  SN,  Webster PD  3rd (1995) Clinical assessment of hyperlipidemic pancreatitis. Am J Gastroenterol 90: 2134–2139.
• 3   Forsmark  CE,  Vege  SS,  Wilcox  CM (2016) Acute Pancreatitis. N Engl J Med 375: 1972–1981.
• 4   Okazaki  H,  Gotoda  T,  Ogura  M,  Ishibashi  S,  Inagaki  K, et al. (2021) Current diagnosis and management of primary chylomicronemia. J Atheroscler Thromb 28: 883–904.
• 5   Kurano  M,  Suzuki  A,  Inoue  A,  Tokuhara  Y,  Kano  K, et al. (2015) Possible involvement of minor lysophospholipids in the increase in plasma lysophosphatidic acid in acute coronary syndrome. Arterioscler Thromb Vasc Biol 35: 463–470.
• 6   Kurano  M,  Dohi  T,  Nojiri  T,  Kobayashi  T,  Hirowatari  Y, et al. (2015) Blood levels of serotonin are specifically correlated with plasma lysophosphatidylserine among the glycero-lysophospholipids. BBA Clin 4: 92–98.
• 7   Kurano  M,  Kano  K,  Dohi  T,  Matsumoto  H,  Igarashi  K, et al. (2017) Different origins of lysophospholipid mediators between coronary and peripheral arteries in acute coronary syndrome. J Lipid Res 58: 433–442.
• 8   Uranbileg  B,  Kurano  M,  Kano  K,  Sakai  E,  Arita  J, et al. (2022) Sphingosine 1-phosphate lyase facilitates cancer progression through converting sphingolipids to glycerophospholipids. Clin Transl Med 12: e1056.
• 9   Emoto  S,  Kurano  M,  Kano  K,  Matsusaki  K,  Yamashita  H, et al. (2017) Analysis of glycero-lysophospholipids in gastric cancerous ascites. J Lipid Res 58: 763–771.
• 10   Kurano  M,  Okamoto  K,  Jubishi  D,  Hashimoto  H,  Sakai  E, et al. (2022) Dynamic modulations of sphingolipids and glycerophospholipids in COVID-19. Clin Transl Med 12: e1069.
• 11   Kurano  M,  Jubishi  D,  Okamoto  K,  Hashimoto  H,  Sakai  E, et al. (2022) Dynamic modulations of urinary sphingolipid and glycerophospholipid levels in COVID-19 and correlations with COVID-19-associated kidney injuries. J Biomed Sci 29: 94.
• 12   Sugimoto  N,  Morita  Y,  Sakai  E,  Yatomi  Y,  Kurano  M (2023) Modulations of urinary lipid mediators in acute bladder cystitis. Prostaglandins Other Lipid Mediat 164: 106690.
• 13   Havel  RJ,  Eder  HA,  Bragdon  JH (1955) The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest 34: 1345–1353.
• 14   Li  H,  Durbin  R (2009) Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25: 1754–1760.
• 15  National Library of Medicine. Genome assembly GRCh38.p13. https://www.ncbi.nlm.nih.gov/assembly/GCF_000001405.39/ accessed on August 25, 2023.
• 16  Picard A set of tools for working with high-throughput sequencing data. http://sourceforge.net/projects/picard/ accessed on August 25, 2023.
• 17   DePristo  MA,  Banks  E,  Poplin  R,  Garimella  KV,  Maguire  JR, et al. (2011) A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat Genet 43: 491–498.
• 18   Wang  K,  Li  M,  Hakonarson  H (2010) ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 38: e164.
• 19  GWAS Catalog. The NHGRI-EBI Catalog of human genome-wide association studies. https://www.ebi.ac.uk/gwas/docs/file-downloads accessed on August 25, 2023.
• 20   Karczewski  KJ,  Francioli  LC,  Tiao  G,  Cummings  BB,  Alföldi  J, et al. (2020) The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581: 434–443.
• 21   Kurano  M,  Sakai  E,  Yatomi  Y (2022) Understanding modulations of lipid mediators in cancer using a murine model of carcinomatous peritonitis. Cancer Med 11: 3491–3507.
• 22   Morita  Y,  Kurano  M,  Sakai  E,  Nishikawa  M,  Sawabe  M, et al. (2019) Evaluation of lysophospholipid measurement in cerebrospinal fluid samples using liquid chromatography-tandem mass spectrometry. Lipids 54: 487–500.
• 23  jMorp. https://jmorp.megabank.tohoku.ac.jp/ accessed on August 25, 2023.
• 24   Wojcik  GL,  Graff  M,  Nishimura  KK,  Tao  R,  Haessler  J, et al. (2019) Genetic analyses of diverse populations improves discovery for complex traits. Nature 570: 514–518.
• 25   Auton  A,  Brooks  LD,  Durbin  RM,  Garrison  EP,  Kang  HM, et al. (2015) A global reference for human genetic variation. Nature 526: 68–74.
• 26   Tsukamoto  K,  Watanabe  T,  Matsushima  T,  Kinoshita  M,  Kato  H, et al. (1993) Determination by PCR-RFLP of apo E genotype in a Japanese population. J Lab Clin Med 121: 598–602.
• 27   Sharma  V,  Witkowski  A,  Witkowska  HE,  Dykstra  A,  Simonsen  JB, et al. (2014) Aberrant hetero-disulfide bond formation by the hypertriglyceridemia-associated p.Gly185Cys APOA5 variant (rs2075291). Arterioscler Thromb Vasc Biol 34: 2254–2260.
• 28   Doolittle  MH,  Ehrhardt  N,  Péterfy  M (2010) Lipase maturation factor 1: structure and role in lipase folding and assembly. Curr Opin Lipidol 21: 198–203.
• 29   Péterfy  M,  Ben-Zeev  O,  Mao  HZ,  Weissglas-Volkov  D,  Aouizerat  BE, et al. (2007) Mutations in LMF1 cause combined lipase deficiency and severe hypertriglyceridemia. Nat Genet 39: 1483–1487.
• 30   Serveaux Dancer  M,  Di Filippo  M,  Marmontel  O,  Valéro  R,  Piombo Rivarola  MDC, et al. (2018) New rare genetic variants of LMF1 gene identified in severe hypertriglyceridemia. J Clin Lipidol 12: 1244–1252.
• 31   Mahley  RW,  Weisgraber  KH,  Huang  Y (2009) Apolipoprotein E: structure determines function, from atherosclerosis to Alzheimer’s disease to AIDS. J Lipid Res 50: S183–S188.
• 32   Weintraub  MS,  Eisenberg  S,  Breslow  JL (1987) Dietary fat clearance in normal subjects is regulated by genetic variation in apolipoprotein E. J Clin Invest 80: 1571–1577.
• 33   Lye  SH,  Chahil  JK,  Bagali  P,  Alex  L,  Vadivelu  J, et al. (2013) Genetic polymorphisms in LDLR, APOB, PCSK9 and other lipid related genes associated with familial hypercholesterolemia in Malaysia. PLoS One 8: e60729.
• 34   Yamamoto  H,  Onishi  M,  Miyamoto  N,  Oki  R,  Ueda  H, et al. (2013) Novel combined GPIHBP1 mutations in a patient with hypertriglyceridemia associated with CAD. J Atheroscler Thromb 20: 777–784.
• 35   Kao  JT,  Wen  HC,  Chien  KL,  Hsu  HC,  Lin  SW (2003) A novel genetic variant in the apolipoprotein A5 gene is associated with hypertriglyceridemia. Hum Mol Genet 12: 2533–2539.
• 36   Matsunaga  A,  Arishima  H,  Niimura  H,  Zhang  B,  Uehara  Y, et al. (2007) Strong linkage disequilibrium and association of –1131T>C and c.553G>T polymorphisms of the apolipoprotein A5 gene with hypertriglyceridemia in a Japanese population. Circ J 71: 746–752.
• 37   Pullinger  CR,  Aouizerat  BE,  Movsesyan  I,  Durlach  V,  Sijbrands  EJ, et al. (2008) An apolipoprotein A-V gene SNP is associated with marked hypertriglyceridemia among Asian-American patients. J Lipid Res 49: 1846–1854.
• 38   Liu  ZK,  Hu  M,  Baum  L,  Thomas  GN,  Tomlinson  B (2010) Associations of polymorphisms in the apolipoprotein A1/C3/A4/A5 gene cluster with familial combined hyperlipidaemia in Hong Kong Chinese. Atherosclerosis 208: 427–432.
• 39   Hishida  A,  Morita  E,  Naito  M,  Okada  R,  Wakai  K, et al. (2012) Associations of apolipoprotein A5 (APOA5), glucokinase (GCK) and glucokinase regulatory protein (GCKR) polymorphisms and lifestyle factors with the risk of dyslipidemia and dysglycemia in Japanese—a cross-sectional data from the J-MICC Study. Endocr J 59: 589–599.
• 40   Lee  MJ,  Chien  KL,  Chen  MF,  Stephenson  DA,  Su  TC (2013) Overweight modulates APOE and APOA5 alleles on the risk of severe hypertriglyceridemia. Clin Chim Acta 416: 31–35.
• 41   Qian  X,  Li  Y,  Liu  X,  Li  L,  Yang  K, et al. (2018) The “T” allele of apolipoprotein A5 rs2075291 is significantly associated with higher total cholesterol and triglyceride and lower high-density lipoprotein cholesterol levels in Asians: a meta-analysis. Nutr Res 56: 11–22.
• 42   Pu  N,  Yang  Q,  Shi  XL,  Chen  WW,  Li  XY, et al. (2020) Gene-environment interaction between APOA5 c.553G>T and pregnancy in hypertriglyceridemia-induced acute pancreatitis. J Clin Lipidol 14: 498–506.
• 43   Huang  YJ,  Lin  YL,  Chiang  CI,  Yen  CT,  Lin  SW, et al. (2012) Functional importance of apolipoprotein A5 185G in the activation of lipoprotein lipase. Clin Chim Acta 413: 246–250.
• 44   Fruchart-Najib  J,  Bauge  E,  Niculescu  LS,  Pham  T,  Thomas  B, et al. (2004) Mechanism of triglyceride lowering in mice expressing human apolipoprotein A5. Biochem Biophys Res Commun 319: 397–404.
• 45   Tsuang  W,  Navaneethan  U,  Ruiz  L,  Palascak  JB,  Gelrud  A (2009) Hypertriglyceridemic pancreatitis: presentation and management. Am J Gastroenterol 104: 984–991.
• 46   Masamune  A,  Sakai  Y,  Yoshida  M,  Satoh  A,  Satoh  K, et al. (2001) Lysophosphatidylcholine activates transcription factor NF-kappaB and AP-1 in AR42J cells. Dig Dis Sci 46: 1871–1881.
• 47   Masamune  A,  Sakai  Y,  Satoh  A,  Fujita  M,  Yoshida  M, et al. (2001) Lysophosphatidylcholine induces apoptosis in AR42J cells. Pancreas 22: 75–83.
• 48   Zhang  KJ,  Zhang  DL,  Jiao  XL,  Dong  C (2013) Effect of phospholipase A2 silencing on acute experimental pancreatitis. Eur Rev Med Pharmacol Sci 17: 3279–3284.
• 49   Zhang  MS,  Zhang  KJ,  Zhang  J,  Jiao  XL,  Chen  D, et al. (2014) Phospholipases A-II (PLA2-II) induces acute pancreatitis through activation of the transcription factor NF-kappaB. Eur Rev Med Pharmacol Sci 18: 1163–1169.
• 50   Wu  Y,  Liao  GY,  Ke  HJ,  Liu  P (2020) Effects of Snake-Derived Phospholipase A2 Inhibitors on Acute Pancreatitis: In vitro and in vivo Characterization. Drug Des Devel Ther 14: 4765–4774.
• 51   Law  SH,  Chan  ML,  Marathe  GK,  Parveen  F,  Chen  CH, et al. (2019) An updated review of lysophosphatidylcholine metabolism in human diseases. Int J Mol Sci 20: 1149.
• 52   Yatomi  Y,  Kurano  M,  Ikeda  H,  Igarashi  K,  Kano  K, et al. (2018) Lysophospholipids in laboratory medicine. Proc Jpn Acad Ser B Phys Biol Sci 94: 373–389.
• 53   Kurano  M,  Kobayashi  T,  Sakai  E,  Tsukamoto  K,  Yatomi  Y (2021) Lysophosphatidylinositol, especially albumin-bound form, induces inflammatory cytokines in macrophages. FASEB J 35: e21673.
• 54   Li  X,  Wang  L,  Fang  P,  Sun  Y,  Jiang  X, et al. (2018) Lysophospholipids induce innate immune transdifferentiation of endothelial cells, resulting in prolonged endothelial activation. J Biol Chem 293: 11033–11045.
• 55   Harada  K,  Kitaguchi  T,  Kamiya  T,  Aung  KH,  Nakamura  K, et al. (2017) Lysophosphatidylinositol-induced activation of the cation channel TRPV2 triggers glucagon-like peptide-1 secretion in enteroendocrine L cells. J Biol Chem 292: 10855–10864.
• 56   Arifin  SA,  Paternoster  S,  Carlessi  R,  Casari  I,  Ekberg  JH, et al. (2018) Oleoyl-lysophosphatidylinositol enhances glucagon-like peptide-1 secretion from enteroendocrine L-cells through GPR119. Biochim Biophys Acta Mol Cell Biol Lipids 1863: 1132–1141.
• 57   Storgaard  H,  Cold  F,  Gluud  LL,  Vilsbøll  T,  Knop  FK (2017) Glucagon-like peptide-1 receptor agonists and risk of acute pancreatitis in patients with type 2 diabetes. Diabetes Obes Metab 19: 906–908.
• 58   Kurano  M,  Yatomi  Y (2018) Sphingosine 1-phosphate and atherosclerosis. J Atheroscler Thromb 25: 16–26.
• 59   Kurano  M,  Tsukamoto  K,  Hara  M,  Tsuneyama  K,  Nishikawa  T, et al. (2020) Modulation of sphingosine 1-phosphate by hepatobiliary cholesterol handling. FASEB J 34: 14655–14670.
• 60   Juchnicka  I,  Kuźmicki  M,  Szamatowicz  J (2021) Ceramides and Sphingosino-1-Phosphate in Obesity. Front Endocrinol (Lausanne) 12: 635995.
• 61   Cirillo  F,  Piccoli  M,  Ghiroldi  A,  Monasky  MM,  Rota  P, et al. (2021) The antithetic role of ceramide and sphingosine-1-phosphate in cardiac dysfunction. J Cell Physiol 236: 4857–4873.
• 62   Konończuk  T,  Łukaszuk  B,  Mikłosz  A,  Chabowski  A,  Żendzian-Piotrowska  M, et al. (2018) Cerulein-Induced Acute Pancreatitis Affects Sphingomyelin Signaling Pathway in Rats. Pancreas 47: 898–903.