This study was approved by the Ethics Committee of Shandong Provincial Hospital affiliated with Shandong University. The study protocol was in line with the Declaration of Helsinki (as revised in Brazil 2013). The consent obtained from all the participants was both informed and written.

Two patients diagnosed with severe hypertriglyceridemia from our hospital were involved in the study. All patient data were collected at the first visit. All family members received specific physical and laboratory examinations in Shandong Provincial Hospital. Peripheral blood specimens were collected from each member for genetic analysis.

We followed these two patients from their first presence and closely tracked their clinical and biochemical information.

Using a QIAamp DNA Mini Kit (Qiagen, Hilden, Germany), genomic DNA was isolated from peripheral blood leukocytes. Next, DNA from peripheral blood was used for WES. We performed genomic DNA fragmentation, paired-end adaptor ligation, amplification, and purification and then captured human exons using a SeqCap EZ Med Exome Enrichment Kit (Roche NimbleGen, Madison, WI, USA). By postcapture amplification and purification, a DNA library was generated and then sequenced with the Illumina HiSeq sequencing platform. To obtain the coverage and mean read depth of target regions, sequence data alignment to the human genome reference (hg19) and variant calling were performed using NextGene V2.3.4 software. The average coverage of the exome was >100×, which allowed a deep examination of the target region to accurately match >99% of the target exons. Mutations with low coverage in the target area were screened and filtered to ensure the accuracy of data analysis.

In addition, the frequency of normal populations [data from Genome Aggregation Database (GnomAD), Exome Aggregation Consortium (ExAC), Trans-Omics for Precision Medicine (TOPMED), Human Gene Mutation Database (HGMD), Clinvar and Online Mendelian Inheritance in Man (OMIM) databases] was obtained by NextGene V2.3.4 and our in-house scripts. A variant was identified as a mutation when it was not found in 500 Chinese controls, in dbSNP (http://www.ncbi.nlm.nih.gov/snp/) and the exome variant server (http://evs.gs.washington.edu/EVS/), or when the allele frequency was <0.001 in the database. According to Standards and Guidelines for the Interpretation of Sequence Variants published by the American College of Medical Genetics (ACMG) in 2015, pathogenic variants were determined with the Human Genome Variation Society (HGVS) nomenclature.

Candidate variants were detected by WES. Pathogenic or suspected pathogenic variants were verified by Sanger sequencing. Universally tagged sequencing primers were designed by using Primer3 version 1.1.4 (http://www.sourceforge.net) and GeneDistiller 2014 (http://www.genedistiller.org/). Polymerase chain reaction (PCR) was performed in a 50-μl system including 5 μl 10 × PCR buffer, 4 μl dNTPs, 4 μl genomic DNA, 1 μl forward and reverse primers, and 0.3 μl Taq Hot Start (Takara Bio, Otsu, Japan). The PCR conditions were as previously described (10). Amplicons were sequenced using an ABI 3730 system (Applied Biosystems, Foster City, CA, USA) and were analyzed by the autoassembler software Chromas 2.6 and visual inspection.

Preheparin blood samples were taken after fasting overnight. Then, heparin was injected intravenously (60 IU/kg), and the postheparin blood was collected from the contralateral arm 10 min later. The subjects stopped taking insulin and other medications in the morning to eliminate the interference caused by medications. The blood sample was centrifuged at 3,000 rpm for 10 min at 4°C to obtain plasma. A Lipoprotein Lipase Activity Assay Kit (Fluorometric, Biovision, Milpitas, CA, USA) was used to measure LPL activity. Mouse (C57BL/6) postheparin plasma was used as a positive control.

Statistical analysis was performed using the SPSS 24.0 software package (SPSS Inc., Chicago, IL, USA). The Kolmogorov–Smirnov test was used to determine the distribution of continuous variables. Continuous variables with a normal distribution are given as the mean ± standard deviation (SD) and were compared by the Mann–Whitney U test. The results were considered statistically significant when the p-value was <0.05.

Proband 1, female, was hospitalized with acute necrotizing pancreatitis and underwent abdominal surgery at the age of 24. At 25 years old, she was found to have high triglyceride (TG) levels (approximately 17 mmol/L). However, TG levels increased further after she was administered some lipid-lowering drugs (including fenofibrate, rosuvastatin, Chinese medicine, etc.). In recent years, TG levels fluctuated at 13.63–33.94 mmol/L, and TC levels fluctuated at 4.92–10.51 mmol/L after drug withdrawal ( Figures 1A, B ). At the age of 49, the proband presented symptoms of hyperglycemia with fasting plasma glucose at 15.22 mmol/L and developed vomiting and diarrhea after oral administration of Chinese medicine decoction. Abdominal computerized tomography (CT) scan ( Figure 1C ) showed that the pancreas was irregular in shape; no obvious abnormality was observed in other parts. After giving patient 1 the low-fat diet, insulin, and liver protection drugs, the last follow-up in June 2020 showed that the TG, TC, and fasting plasma glucose were 15.37, 9.43, and 6.01 mmol/L, respectively ( Table 1 ). The family history of the patient showed that the patient’s parents and her sister were deceased and could not be tested. However, it was recalled that her sister had also suffered from recurrent pancreatitis ( Figure 2A ).

Proband 2 came to our hospital for treatment because his blood sample showed severe hyperlipemia during the physical examination at 9 months old. On the second day of admission, his TG level was 20.35 mmol/L, and his TC level was 8.89 mmol/L. A low-fat diet and 4 U of continuous intravenous infusion of insulin were ordered to lower blood lipids. After 3 days, the TG level was 22.77 mmol/L, and the TC was 7.60 mmol/L. After 7 days, the TG level was 12.32 mmol/L, and the TC level was 4.55 mmol/L. After 10 days, the proband was discharged, his blood lipids improved significantly, with TG at 10.25 mmol/L and TC at 4.93 mmol/L. After being discharged from the hospital, he continued to consume a low-fat diet without any special discomfort. The last follow-up conducted in June 2020 showed that the serum TG was 3.35 mmol/L, the TC was 2.72 mmol/L, and the fasting plasma glucose was 5.60 mmol/L ( Table 1 ). The family history of the patient shows that the parents of patient 2 have thus far been healthy ( Figure 2B ).

To further identify disease-causing genes to facilitate diagnosis, we subsequently applied the WES technique for genetic analysis of the two pedigrees. According to HGMD, we found a novel compound variant of LPL (c.3G>C and c.835_836delCT, Figure 3A ) in proband 1, which may cause truncation mutations (p. M1)?, and frameshift mutations (p. L279Vfs*3) leading to premature amino acid stop codes for protein synthesis, respectively. The p. M1? variant has been reported in a related clinical case, and its region is the translation initiation codon of this protein (11). The p. L279Vfs*3 variant is absent in the GnomAD, TOPMED, and ExAC databases. Notably, the son and daughter of proband 1 carry the p. M1? variant derived from the maternal line, as verified by Sanger sequencing, but their lipid levels and other phenotypes were normal.

We also identified proband 2 as carrying a novel compound variant of LPL (c.188C>T and c.662T>C, Figure 3B ) that resulted in a change in the 63rd amino acid of the encoded protein from serine to phenylalanine (p. Ser63Phe) and a change in the 221st amino acid of the encoded protein from isoleucine to threonine (p. Ile221Thr), respectively. The p. Ser63Phe variant is absent in the GnomAD, TOPMED, and ExAC databases, while the p. Ile221Thr variant has been reported in three unrelated probands causing lipoprotein lipase deficiency (12). In addition, the parents of proband 2 each carried a variant. The Ser63Phe variant comes from the father, and the p. Ile221Thr variant comes from the mother, but they have thus far been healthy.

Accordingly, the postheparin LPL activity of patient 1 (0.46 ± 0.13 mU/ml) and patient 2 (0.75 ± 0.11 mU/ml) actually showed a decrease of 72.22 ± 9.46% (p<0.01) and 54.60 ± 9.03% (p<0.01), respectively, compared with the control (1.66 ± 0.09 mU/ml). The differences between the son and daughter of proband 1, the parents of proband 2, and the control seem to be negligible ( Figure 4 ). Therefore, two novel compound heterozygous variants of LPL may cause changes in LPL enzyme activity in the patient’s plasma, leading to defects in triglyceride metabolism and resulting in extremely elevated serum triglycerides.

To understand the effect of the identified LPL variants at the protein level, we transiently transfected the plasmids containing pcDNA3.1-LPL-WT and pcDNA3.1-LPL-MU into HEK293T/17 cells. HEK293T/17 cells transfected with the p. M1? mutation, p. L279Vfs*3 mutation, or p. M1? and p. L279Vfs*3 mutation showed virtually no protein in the cell lysate (p<0.01, Figure 5 ). Compared with cells transfected with the plasmids containing the pcDNA3.1-LPL-WT, there was no significant difference in the amount of protein reduction produced by the cells transfected with the p. S63F mutation or the two variants (p. S63F and p. I221T), but protein synthesis was reduced by 48.33 ± 1.78% in cells transfected with the p. I221T variant (p<0.01, Figure 5 ).

Next, we tested LPL activity in media of HEK 293T/17 cells. Wild-type LPL medium was used as a positive control. Except for the p. L279Vfs*3 mutation, LPL activity in the medium transfected with other plasmids containing pcDNA3.1-LPL-MU was significantly reduced compared with the LPL activity of the wild type. Specifically, cells transfected with the p. M1? or the two variants (p. M1? and p. L279Vfs*3) showed reductions in protein activity of approximately 55.93 ± 3.28% (p<0.01) or 59.84 ± 4.47% (p<0.01), while LPL activity was reduced by 62.76 ± 9.90% (p<0.01) or 63.25 ± 10.55% (p<0.01) in the media of HEK 293T/17 cells transfected with p. Ser63Phe or the two variants (p. Ser63Phe and Ser63Phe+Ile221Thr). The media of cells transfected with p. Ile221Thr demonstrated a maximum reduction in LPL activity of approximately 81.56 ± 4.35% (p<0.01, Figure 6 ).