Abstract
Cardiovascular diseases (CVDs) are the major cause of death globally. Elevated lipoprotein(a) [Lp(a)] levels are an independent risk factor for multiple CVDs. Lp(a) is a lipoprotein composed of apolipoprotein(a) [apo(a)] bound to a low-density lipoprotein (LDL) molecule. Apo(a) comprises two plasminogen kringle domains (KIV and KV) and an inactive protease domain. One of the KIV domains, KIV-2, occurs as a copy number variation (CNV) that varies among individuals. Lp(a) levels are inversely correlated with the KIV-2 CNV that determines apo(a) isoform size. Studies showed that Lp(a) levels are mostly genetically regulated, and most of the variants that alter Lp(a) levels reside within the LPA gene, which codes for apo(a). Several variants have been associated with Lp(a) levels, with most of these being associated with lower levels. The few variants that were associated with high levels have been mostly linked to small apo(a) isoform size. This study aimed to identify variants associated with the high Lp(a) phenotype independent of the apo(a) isoform size. Consequently, functional studies were done to determine their role in altering Lp(a) levels, to identify potential novel therapeutic targets for Lp(a)-lowering. As the KIV-2 complex repeat region of LPA is still very challenging to sequence, this study also aimed to develop a method to sequence the KIV-2 repeat region using long-read sequencing technology.
Illumina sequencing of the LPA gene was performed on a discovery set comprising 48 very high Lp(a) samples (>60 mg/dL) from the Otago LPA European cohort. From this, 39 variants were detected. The rs3124784 SNP in the protease domain was selected for further analysis and was associated with Lp(a) levels and apo(a) isoform size in two European cohorts, the Otago LPA (n=589) and the GCKD cohorts (n=4930). The minor allele of rs3124784 was significantly associated with high Lp(a) levels in both cohorts, but the minor allele was also significantly associated with small apo(a) isoform size. The effect size was ~4mg/dL after adjusting for the apo(a) isoform sizes, and the SNP accounted for only 0.5% of the variance in Lp(a) levels. Therefore, this SNP was not further analysed functionally.
To combat the inability of the short-read Illumina sequencing to locate and determine the frequency of the identified KIV-2 variants, Oxford Nanopore long-read sequencing, and adaptive sampling were used to enrich for and sequence the KIV-2 repeat region of LPA in six high Lp(a) samples from the Otago LPA cohort. Assembly of the long reads generated contigs that covered some but not the entire repeat region in all six samples. Some of the variants detected via Illumina sequencing in those samples were located to a specific repeat, and their frequency was determined in some of the sequenced samples. However, the read length and the depth need to be further optimised.
Twelve other candidate genes were Illumina sequenced in 54 high Lp(a) samples (>50mg/dL) from the Otago LPA European cohort. Some genes code for receptors that interact with Lp(a), and others are nearby LPA. Sequencing detected 253 variants. Seven variants in the SLC22A1-3 genes were genotyped and associated with Lp(a) levels and apo(a) isoform sizes in both study cohorts. The minor alleles of all SLC22A1-3 variants near to LPA were significantly associated with high Lp(a) levels. However, only the two SNPs (rs1810126 and rs3088442) in SLC22A3 were not associated with small apo(a) isoform size. The two highly linked SNPs in SLC22A3 accounted for 3% of the variance in Lp(a) and the allelic effect size was ~9 mg/dL. Also, the two SLC22A3 SNPs were associated with an increased LPA expression in liver tissues on GTEx.
Lastly, this study further investigated the possible role of the SLC22A3 transporter in Lp(a) uptake since SLC22A3 transports serotonin, which increases Lp(a) surface binding. Results showed that HepG2 liver cells transiently overexpressing SLC22A3 had reduced Lp(a) uptake. However, it is hypothesised that a more pronounced effect will be observed in stably transfected HepG2 liver cell lines.
In conclusion, this study provided a new insight into the association between the rs1810126 and rs3088442 SNPs in the SLC22A3 gene and Lp(a). The two SNPs were associated with high Lp(a) levels independent of the apo(a) isoform size, and they accounted for the greatest variation in Lp(a) levels in this study. The two SLC22A3 SNPs were also associated with an increased LPA expression in liver tissues on GTEx. Preliminary results showed that transiently overexpressing the SLC22A3 transporter in HepG2 cells reduced Lp(a) uptake, but the effect could be improved in stably transfected cell lines. Establishing the role of the SLC22A3 transporter in Lp(a) uptake will determine if SLC22A3 can be a potential therapeutic target to lower Lp(a). Moreover, this study proposed a novel approach to sequence the complex KIV-2 repeat region in LPA. Improving the proposed approach will enable us to unravel the secrets of the KIV-2 repeat region and several other repeat regions which have been challenging to sequence and analyse.