2.1. Clinical Features and Biochemical Measurements of the Patients

All patients with APOC2-related HTG were recruited from the literature using a search conducted on available databases including HGMD and PubMed using the given terms (please see below). Genotype and clinical features of patients were recorded. In addition, data of heterozygotes were also analyzed. In silico analyses were applied to predict the pathogenicity of APOC2 variants. Genetic and clinical information of a cohort of Iranian patients were also investigated.

The number of homozygous and heterozygous individuals was extracted from the published studies to analyze the clinical characteristics including age, sex, hepatosplenomegaly, lipemia retinalis, eruptive xanthomas, pancreatitis, and abdominal pain. Profiles of TG and TC patents were also investigated. APOC2 variants and the genotypes of patients were determined as well. Frequencies of different variants were checked in the studied populations. The subgroup analysis was done for different groups of HTG patients as well as domain-specific analysis of the variants.

2.2. Molecular Investigations

Fifty HTG patients referring to Rajaie Cardiovascular Center, Tehran, Iran, were also studied. Inclusion criteria were TG above 250 mg/dL. These patients were selected based on serum TG levels according to the blood test that was taken. The study was approved by the Ethics Committee of Iran University of Medical Sciences.

After taking an informed consent form, genomic DNA was extracted from peripheral blood leucocytes by a standard procedure. To assess quantity of DNA, a spectrophotometer (NanoDrop ND2000c, Thermo Scientific) was applied. All exons and introns, including the intron/exon boundaries of APOC2, were amplified using the forward and reverse primers (available upon request). Briefly, polymerase chain reaction (PCR) was executed in a final volume of 50 μL, consisting of 10 pmol forward and reverse primers, 150 ng template DNA, 0.2 units/μL Taq DNA polymerase, 1.5 mmol/L MgCl2, and 0.4 mmol/L of each dNTP in the reaction. The following PCR program was implemented: 5 minutes at 94°C initial denaturation and 30 cycles of 30 seconds at 94°C denaturation, 30 seconds at 63°C, and 30 seconds at 72°C extension, followed by 10 minutes at 72°C final extension. The PCR products were directly sequenced on an ABI PRISM™ 3500 (PE Applied Biosystems) sequencing analyzer using a BigDye termination method, and the reactions were analyzed on the GeneMapper software package.

2.3. Literature Review of Published APOC2 Mutation Cases

A detailed comprehensive search was done in PubMed, Springer, John Wiley, and Elsevier inclusive until Jan 2023 using the keywords APOC2[ti/ab] variant AND/OR Apolipoprotein C-II[title/abstract] to find all the reported patients with APOC2 pathogenic variants.

Demographic and clinical presentations, lipid profile (TG, TC), family history, phenotypic manifestations (HSM: hepatosplenomegaly, AS: asymptomatic, LR: lipemia retinalis, EX: xanthomas, Pan: pancreatitis, and AP: abdominal pain), and genotype/zygosity of individuals were extracted. Homozygous and heterozygous individuals were more investigated for being symptomatic and HTG forms. The variants were named using standard nomenclature and categorized into premature chain termination (PCT) including nonsense and frameshift variants, missense, gross deletions, and intronic and regulatory variants based on location and functional effect of the variants.

2.4. Bioinformatic Analyses

All APOC2 likely/pathogenic variants according to the ClinVar were analyzed using available software tools to predict their pathogenicity. We used MutationTaster (http://www.mutationtaster.org/), CADD (https://cadd.gs.washington.edu/home), FATHMM (http://fathmm.biocompute.org.uk/), and PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/). The variants were classified using VarSome (https://varsome.com/). According to the guidelines provided by the American College of Medical Genetics and Genomics, sequence variants are categorized into interpretative categories based on expert opinion and empirical data. The interpretive categories are then used along with a systematic algorithm to determine the significance of the sequence variants as benign, likely benign, uncertain significance (VUS), likely pathogenic, and pathogenic [13].

A multiple amino acid sequence alignment of APOCII was performed using UniProt protein family members (UniProtKB/Swiss-Prot P02655). Pathogenicity analysis of APOC2 variants was also done through aggregation of homologous human protein domains using MetaDome (https://stuart.radboudumc.nl/metadome/). MetaDome is a tool that makes use of the data from gnomAD and ClinVar to analyze mutation tolerance at each position in a human protein. It increases the analysis of your chosen gene by doing a parallel analysis of all homologous domains across the entire human genome. This helps to enhance the interpretation of the gene in the context of its full domain [14].

3.1. Patient Demographics

3.1.2. Symptomatic Heterozygous Individuals

Ten of 39 heterozygous individuals were symptomatic. The mean TG and TC were 434.2 and 227.85 mg/dL, respectively (Table 1). The distribution of HTG frequencies across different groups is depicted in Figure 1. Approximately 44% of the individuals studied had TG levels that were within normal range followed by 27% of moderate HTG. However, severe HTG was observed in 2% of these subjects. Pancreatitis was also detected in 2 of the cases.

3.2. Clinical Features in Different Types of Variants

A total of 23 patients had PCT variants, with mean TG and TC values of 4506.98 and 395.45, respectively. Of 9 HSM cases, 8 had PCT variants. In addition, PCT variants were detected in 7 out of 12 cases of pancreatitis and 7 out of 11 cases of lipemia retinalis.

The mean TG and TC levels among patients with missense variants were 2225.34 and 250.96, respectively. Missense variants were found in 4 patients with pancreatitis and 3 patients with AP. Meanwhile, the mean TG and TC levels for splice and regulatory variants were 4580.33 and 194.33, respectively. The mean age of these patients was 15.98 years as depicted in Figure 2.

A total of 30 variants in the APOC2 gene were found, comprising 8 missense, 6 nonsense, 6 indel, 4 splice, and 2 regulatory variants, along with 3 gross deletions and 1 gross duplication. In total, 12 variants were categorized as premature chain terminations (including nonsense and indel mutations) (Table 3). A detailed representation of the positions of the variants in different regions of the protein is displayed in Table 4 and Figure 2. Some protein regions had specific variant; O-glycosylated had nonsenses and lipid-binding region showed only missenses. The c.1A>G (p.Met1Val) and c.274C>T (p.Gln92Ter) were the most common variants, each found in 2 unrelated families.

Five variants c.122A>C (p.Lys41Thr), c.56-30G>A, c.178G>A (p.Glu60Lys), c.229A>C (p.Lys77Gln), and c.206A>T (p.Glu69Val) found in heterozygous individuals were classified as likely benign/benign/VUS. Other variants which were found in homozygous patients were likely pathogenic/pathogenic variants according to ACMG 2015 (Table 3).

3.3. Distribution of the Variants

In European subjects, 15 variants were detected, with c.10C>T and c.177C>A each being identified in two unrelated families. Of these variants, 7 were PTC. A total of 7 variants were detected in Asian patients, of which 3 were PTC. Similarly, 3 variants were detected in American patients, all of which were PTC. In the same way, only 1 variant was detected in each of the Oceanic and African populations (Table 2).

3.4. Molecular Findings and In Silico Analysis

Pathogenicity analysis of APOC2 variants through aggregation of homologous human protein domains was done using MetaDome (Figure 3). As shown in Figure 3, 14 variants were located in alpha helices. The majority of variants (12 variants) were in exon 3.

4.1. Clinical Features of APOC2 Homozygotes and Heterozygotes

The most common features of patients with two variants in APOC2 gene are as follows: pancreatitis (observed in 55% of these patients), lipemia retinalis (28.95%), abdominal pain (23.68%), hepatosplenomegaly (23.68%), and xanthomas (18.42%). These findings imply that when an APOC2 variant is present, it is likely to correlate with clinical manifestations in accordance with our observations.

Since pathogenic variants in the APOC2 gene are usually inherited in a recessive fashion, it is anticipated that heterozygous individuals will not exhibit any clinical symptoms. However, it was discovered in our study that approximately 25% of the heterozygous individuals were symptomatic. We have previously reported symptomatic patients with congenital adrenal hyperplasia due to heterozygous pathogenic variants in CYP21A2 gene [19, 20]. Possible reasons for this observation are as follows: (1) for the variants in regulatory regions, some individuals may harbor variants in distant regions that regulate gene expression, which have not been previously identified. This may occur even in a heterozygous state. (2) In large deletions and duplications, these variants could be undetectable due to technical limitations in some studies. (3) Digenic and triallelic inheritance is possible, whereby variants in other genes may collaborate with the identified variant in causing the disease. This is also known as digenic or triallelic inheritance. (4) As for gain-of-function variants, some variants may act in a dominant manner; however, functional studies are required to confirm this. (5) The possibility exists that environmental factors, in particular related to hyperlipidemia, may further exacerbate symptoms caused by genetic variants in heterozygous individuals, leading to enhanced lipid levels. On the contrary, it is possible that asymptomatic homozygotes, which comprise approximately 18% of biallelic cases, are the result of environmental and genetic modifying factors as well; these factors may contribute in compensating for APOC2 deficiency, leading to no clinical symptoms despite increased levels of TG. Takase et al. revealed a Japanese patient who was diagnosed with significantly reduced plasma apoC-II levels, yet they did not find a causative variant present in the APOC2 gene [21]. This suggests that unidentified factors, such as alternative genes or agents, might play a role in the development of APOC2-hypertriglyceridemia.

Pancreatitis has been previously reported in homozygous case reports with both missense and PTC variants [18, 22, 23], but to our knowledge, there has not been a study that defines how common pancreatitis is among homozygous patients. In our study, we found that more than half of the cases showed evidence of this condition. Different types of variants may lead to various forms of HTG with different severity in clinical features. PCT variants usually result in more severe HTG and clinical phenotypes; the mean age of patients with such variants is less than that of patients having missense ones (Table 4). PTC variants in APOC2 are more common rather than missense variants. Human apoC2 protein has three amphipathic helices, with the C-terminal helix responsible for LPL activation. The majority of the variants is distributed on alpha helices of the protein indicating the important roles of these regions. The exact location of binding site on LPL for the activating cofactor, apolipoprotein C2, and its associated mechanisms have been ambiguous. Kumari et al. demonstrated that APOC2’s C-terminal alpha helix interacts with regions of LPL near the catalytic pocket [24]. The findings by Kumari et al. indicate that these alpha helices have an important role in APOC2 and variations in this region potentially cause pathogenic effects, leading to dyslipidemia. To develop effective therapeutic modalities and design cell- and molecular-based therapies, attention to the roles and the structures of different domains of this protein could be helpful.

The APOC2 variants have been reported from many populations worldwide, although we could not see any variant in Iranian patients which is similar to Oman population [25]. APOC2 variants may be extremely rare in these populations although it is recommended to study more patients to find its frequencies in these countries. The frequency of AOPC2 variants is apparently more than other continents although it could be due to that more people have been studied from the continent. The type and frequencies of the variants can vary greatly among different cohorts. Moreover, significant genetic differences may result in different clinical presentation, especially regarding lipid profiles. In previous research, we have reported the impact of both common and population-specific variants on the incidence of other disorders [26–28], although the frequency and effects of different variants in APOC2 gene have not been published yet. Various APOC2 variants have been reported across different populations, which imply that certain positions in the locus may be considered a hotspot for alteration. For instance, c.274C>T has been found among patients in both Europe and the United States. Conversely, some variants have been documented solely within specific demographics, denoting the possibility of a founder effect; for example, c.177C>A was detected among two unrelated Italian patients.

4.2. Genetic Variant Distribution along the Protein

The human pro-apoliporpotein C-II protein contains 101 amino acids in which amino acids 1-22 are signal peptide. Apolipoprotein C-II has 3 amphipathic alpha helices spanning the approximate residues 16-38, 45-57, and 65-74 (Table 5) [29]. The majority of the variants were in these regions (Table 5 and Figure 3). The lipid-binding and LPL-binding domains of the protein are localized to a region in the N-terminal and C-terminal region, respectively [30, 31]. It has been known that some residues (63, 66, 69, and 70) are important for LPL activation [32].

Since we did not directly participate in patient selections, clinical evaluations, or technical experiments, there were limitations that we should keep in mind, such as the possibility of having missed patients with APOC2 variants, even during our search process. Future research involving a wider pool of patients or even a multicenter study is needed to provide a more comprehensive examination of the APOC2 variants, eliminating the limitations we encountered during our study. Nevertheless, our study is the first of its scale to study the connection between APOC2 variants and their respective phenotypes.