Methods and systems for detecting familial hypercholesterolemia and related disorders

Abstract

In an aspect, the present disclosure provides a method for detecting a presence or an absence of familial hypercholesterolemia (FH) in a subject, comprising: (a) determining a genotype or a gene expression of PDZK1 in the subject, and (b) detecting the presence or the absence of the FH in the subject based at least in part on the determined genotype or gene expression of PDZK1 in the subject.

Description

Docket No.: 60115-706.601METHODS AND SYSTEMS FOR DETECTING FAMILIAL HYPERCHOLESTEROLEMIA AND RELATED DISORDERSCROSS-REFERENCE

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 692,279, filed September 9, 2024, each of which are incorporated by reference herein in its entirety.BACKGROUND

[0002] Familial hypercholesterolemia (FH) is an inherited lipid disorder characterized by extremely high levels of low-density lipoprotein cholesterol in the blood with concomitantly low levels of high-density cholesterol.SUMMARY

[0003] The present disclosure provides methods and systems for detecting familial hypercholesterolemia (FH) and related disorders.

[0004] Familial hypercholesterolemia (FH) is an inherited lipid disorder characterized by extremely high levels of low-density lipoprotein cholesterol in the blood with concomitantly low levels of high-density cholesterol. The disorder has an observed population frequency of 1 :300 and is transmitted as a dominant trait, although recessive FH has also been reported with an earlier onset and severely more aggressive symptomology and progression. In all cases, undiagnosed and untreated FH may lead to life-threatening complications, most prominently for coronary heart disease (50% chance before the age of 50 for men and 30% risk for women by the age of 60.

[0005] FH remains underdiagnosed and undertreated, with some estimates placing the disease burden at 14-34 million people worldwide. Despite the magnitude of impact and the subsequent burden of cardiovascular disease to individuals, families, and national health systems, strategies geared toward early diagnosis and management of the disorder are lacking, with genetic studies often applied retrospectively, after adult-onset clinical disease has impacted cardiovascular health.

[0006] The present disclosure provides the discovery of a gene that regulates LDL cholesterol offers significant new applications. For example, the development and validation of a new clinical genetic diagnostic panel for familial hypercholesterolemia for the detection of both rare and common SNV and CNV events. This new tool may increase the diagnostic yield for FH cases continue to improve the guidance of treatment. For this purpose, it was designed a new diagnostic test that includes the analysis of both single nucleotide variants and structural variants for each of APOB, LDLR, PCSK9 and PDZK1, performed in a single assay.Docket No.: 60115-706.601

[0007] In an aspect, the present disclosure provides a method for detecting a presence or an absence of familial hypercholesterolemia (FH) in a subject, comprising: (a) determining a genotype or a gene expression of PDZK1 in the subject, and (b) detecting the presence or the absence of the FH in the subject based at least in part on the determined genotype or gene expression of PDZK1 in the subject.

[0008] In some embodiments, (a) comprises determining the genotype of PDZK1 in the subject. In some embodiments, the genotype comprises a single nucleotide variant (SNV). In some embodiments, the genotype comprises a copy number variant (CNV). In some embodiments, the genotype comprises a whole gene duplication.

[0009] In some embodiments, (a) comprises determining the gene expression of PDZK1 in the subject. In some embodiments, determining the gene expression of PDZK1 in the subject comprises determining a ribonucleic acid (RNA) expression of PDZK1 in the subject. In some embodiments, determining the gene expression of PDZK1 in the subject comprises determining a protein expression of PDZK1 in the subject.

[0010] In some embodiments, the method further comprises administering a treatment to the subject, responsive to detecting the presence of the FH in the subject. In some embodiments, the treatment comprises a statin, a cholesterol absorption inhibitor, a bile acid sequestrant, a PCSK9 inhibitor, a lipoprotein apheresis, a bempedoic acid, lomitapide, a liver transplant, or evinacumab. Treatments for FH may be described by, for example, Lambert et al., “Current Treatment of Familial Hypercholesterolaemia”, Eur. Cardiol., 2014 Dec; 9(2): 76-81, which is incorporated by reference herein in its entirety.

[0011] Another aspect of the present disclosure provides a non-transitory computer readable medium comprising machine executable code that, upon execution by one or more computer processors, implements any of the methods above or elsewhere herein.

[0012] Another aspect of the present disclosure provides a system comprising one or more computer processors and computer memory coupled thereto. The computer memory comprises machine executable code that, upon execution by the one or more computer processors, implements any of the methods above or elsewhere herein.

[0013] Additional aspects and advantages of the present disclosure will become readily apparent to those skilled in this art from the following detailed description, wherein only illustrative embodiments of the present disclosure are shown and described. As will be realized, the present disclosure is capable of other and different embodiments, and its several details are capable of modifications in various obvious respects, all without departing from the disclosure.Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.Docket No.: 60115-706.601

[0014] Another aspect of the present disclosure provides a non-transitory computer readable medium comprising machine executable code that, upon execution by one or more computer processors, implements any of the methods above or elsewhere herein.

[0015] Another aspect of the present disclosure provides a system comprising one or more computer processors and computer memory coupled thereto. The computer memory comprises machine executable code that, upon execution by the one or more computer processors, implements any of the methods above or elsewhere herein.

[0016] Additional aspects and advantages of the present disclosure will become readily apparent to those skilled in this art from the following detailed description, wherein only illustrative embodiments of the present disclosure are shown and described. As will be realized, the present disclosure is capable of other and different embodiments, and its several details are capable of modifications in various obvious respects, all without departing from the disclosure.Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.INCORPORATION BY REFERENCE

[0017] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. To the extent publications and patents or patent applications incorporated by reference contradict the disclosure contained in the specification, the specification is intended to supersede and / or take precedence over any such contradictory material.BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The features of the disclosure are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present disclosure will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the disclosure are utilized, and the accompanying drawings (also “Figure” and “FIG.” herein), of which:

[0019] FIG. 1 illustrates a schematic representation of the duplication (chrl : 145670985- 145707368) detected in two unrelated patients projected on the UCSC Genome Browser (genome.ucsc.edu / ).

[0020] FIG. 2 shows a phenome-wide view for variant rs2976384 showing phenome-wide significant association with traits from the UK Biobank including hypertension (A), essential hypertension (B) and hyperglyceridemia (C).Docket No.: 60115-706.601

[0021] FIG. 3 shows a phenome-wide view (PheWAS) for variants rsl553701870 and rsl649823291 showing phenome-wide significant association with traits from the FinMetSeq study including saturated fatty acids (A), LDL cholesterol (B) and APOB levels (C).

[0022] FIG. 4A shows a demographic pyramid that displays the age and sex distribution of participants.

[0023] FIG. 4B shows a distribution of BMI categories across the cohort.

[0024] FIG. 4C shows a distribution of lipids including total cholesterol, LDL, HDL, and triglycerides.

[0025] FIG. 4D shows a distribution of health conditions including corneal arcus, coronary heart disease, diabetes and tendon xanthomatosis.

[0026] FIG. 5 shows an ancestry analysis distribution chart.

[0027] FIG. 6 shows ancestry ideograms for various samples and various variants associated with various ancestries.

[0028] FIG. 7 shows a plot illustrating the clustering of 40 pathogenic or likely pathogenic variants located within the exon and intron regions of the LDLR gene.

[0029] FIG. 8A shows a distribution of allele types in our cohort and restricting the analysis to only pathogenic or likely pathogenic variants.

[0030] FIG. 8B shows a distribution of variants found in 3 FH genes as well as in FH phenocopy genes.

[0031] FIG. 9A shows an LDL PRS distribution for controls and cases (carriers and noncarriers).

[0032] FIG. 9B shows a percentage of samples above the 95th percentile of a Normal distribution.

[0033] FIG. 9C shows a cumulative incidence of LDL for the age of the patient.

[0034] FIG. 10 shows demographic / clinical characteristics of 5 carriers identified in FH cohort and UK Biobank.DETAILED DESCRIPTION

[0035] While various embodiments of the invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, changes, and substitutions may occur to those skilled in the art without departing from the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed.Docket No.: 60115-706.601Terms and Definitions

[0036] As used herein, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. Any reference to “or” herein is intended to encompass “and / or” unless otherwise stated.

[0037] As used herein, the phrases “at least one,” “one or more,” and “and / or” are open-ended expressions that are both conjunctive and disjunctive in operation. For example, each of the expressions “at least one of A, B and C,” “at least one of A, B, or C,” “one or more of A, B, and C”, “one or more of A, B, or C” and “A, B, and / or C” means A alone, B alone, C alone, A and B together, A and C together, B and C together, or A, B and C together. As used herein, the phrase “at most three” may mean less than one, one, two, or three.

[0038] Reference throughout this specification to “some embodiments,” “further embodiments,” or “a particular embodiment,” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, the appearances of the phrase “in some embodiments,” or “in further embodiments,” or “in a particular embodiment” in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0039] The terms "subject," "individual," and "patient" may be used interchangeably and refer to humans, as well as non-human mammals (e.g., non-human primates, canines, equines, felines, porcines, bovines, ungulates, lagomorphs, rodents, and the like). In various embodiments, the subject may be a human (e.g., adult male, adult female, adolescent male, adolescent female, male child, female child) under the care of a physician or other health worker in a hospital, as an outpatient, or other clinical context. In certain embodiments, the subject may not be under the care or prescription of a physician or other health worker. In some embodiments, the subject may be under the care of a dental professional.

[0040] As used herein, “treatment” or “treating” refers to an approach for obtaining beneficial or desired results with respect to a disease, disorder, or medical condition including, but not limited to, a therapeutic benefit and / or a prophylactic benefit. In certain embodiments, treatment or treating involves administering a therapeutic to a subject. A therapeutic benefit may include the eradication or amelioration of the underlying disorder being treated. Also, a therapeutic benefit may be achieved with the eradication or amelioration of one or more of the physiological symptoms associated with the underlying disorder, such as observing an improvement in the subject, notwithstanding that the subject may still be afflicted with the underlying disorder.

[0041] Familial hypercholesterolemia (FH) is an inherited lipid disorder characterized by extremely high levels of low-density lipoprotein cholesterol in the blood with concomitantly lowDocket No.: 60115-706.601 levels of high-density cholesterol. The disorder has an observed population frequency of 1 :300 and is transmitted as a dominant trait, although recessive FH has also been reported with an earlier onset and severely more aggressive symptomology and progression. In all cases, undiagnosed and untreated FH may lead to life-threatening complications, most prominently for coronary heart disease (50% chance before the age of 50 for men and 30% risk for women by the age of 60.Methods for Detection of Disease

[0042] In an aspect, the present disclosure provides a method for detecting a presence or an absence of familial hypercholesterolemia (FH) in a subject. The detection may comprise determining a genotype or a gene expression of PDZK1 in the subject, and detecting the presence or the absence of the FH in the subject based at least in part on the determined genotype or gene expression of PDZK1 in the subject.

[0043] In some embodiments, the detection may comprise determining the genotype of PDZK1 in the subject. In some embodiments, the genotype may comprise a single nucleotide variant (SNV). In some embodiments, the genotype may comprise a copy number variant (CNV). In some embodiments, the genotype may comprise a whole gene duplication.

[0044] In some embodiments, the detection may comprise determining the gene expression of PDZK1 in the subject. In some embodiments, determining the gene expression of PDZK1 in the subject may comprise determining a ribonucleic acid (RNA) expression of PDZK1 in the subject. In some embodiments, determining the gene expression of PDZK1 in the subject may comprise determining a protein expression of PDZK1 in the subject.

[0045] In some embodiments, the method may further comprise administering a treatment to the subject, responsive to detecting the presence of the FH in the subject. In some embodiments, the treatment may comprise a statin, a cholesterol absorption inhibitor, a bile acid sequestrant, a PCSK9 inhibitor, a lipoprotein apheresis, a bempedoic acid, lomitapide, a liver transplant, or evinacumab. Treatments for FH may be described by, for example, Lambert et al., “Current Treatment of Familial Hypercholesterolaemia”, Eur. Cardiol., 2014 Dec; 9(2): 76-81, which is incorporated by reference herein in its entirety.

[0046] In some embodiments, rare pathogenic point mutations and structural variants in all FH genes, together with variants in APOE, CREB3L3, and PI NL contribute to a molecular FH diagnosis in 67% of families, including gene-disruptive CNVs which arose in a native American background. Second, ancestry-adjusted polygenic risk score analysis identified a significant liability for coronary artery disease, hypertension, LDL, HDL, and Type 2 Diabetes. The polygenic signal for LDL was present in patients with rare, pathogenic FH mutations and wasDocket No.: 60115-706.601 more prominent in individuals bereft of a molecular FH diagnosis. Finally, we report both a whole-gene duplication and common, non-coding variants in a locus, PDZK1, which contribute to the genetic burden of FH, a finding we replicated in the UK Biobank (UKB). Together, the analyses illustrate the value of genetic studies in non-European populations and reinforce the notion that individual risk to disease can arise from both rare, large effect alleles (i.e. alone or in combination across genes) and common variants that increase the mutational burden of a biological system. Genetic variants in four genes (LDLR, APOB, PCSK9, and APOK) were shown to contribute drive autosomal dominant FH. Moreover, rare cases of autosomal recessive FH have been observed, with mutations in LDLRAP1 pinpointed as an underlying genetic cause.

[0047] Here, the Mexican FH registry (www.fhmexico.org.mx) was utilized to study 302 samples from 169 families. In the registry, each patient was assigned a unique code in order to preserve anonymity. Follow-up data on each patient was also collected once a year, or in accordance with their normal clinical follow-up. Specifically, after screening all participants for mutations across the three FH-causing loci, we constructed a network composed of 102 proteins composed of either biological components of LDL biosynthesis and trafficking, or represent first-order protein-protein interactors of such components, the analyses have discovered two unrelated individuals with a whole gene duplication of PDZK1, a protein that may traffic the HDL receptor but which, surprisingly, has not been associated with LDL pathobiology in any species, including rodent models.

[0048] In some embodiments, comprehensive genetic analysis of this cohort was performed by coupling whole exome and short- and long- read whole genome sequencing with state-of-the art ancestry and ancestry-adjusted polygenic score algorithms, the contribution of both point mutations and structural variants were determined in all genes associated with dyslipidemias in humans. Moreover, it was examined how the genetic architecture of FH might be influenced by the concomitant contribution of rare and common alleles. Finally, it was determined whether some of the missing genetic burden of FH might be contributed either by previously unknown alleles (in FH genes) unique to this population or in previously unknown disease-associated loci.

[0049] Further PheWas analysis of the locus in publicly-available databases unearthed the involvement of common variants at this locus in a variety of co-morbidities relevant to FH, such as hypertension, hyperglyceridemia, and, crucially, abundance of APOB protein, thus offering insight into pathomechanism. Notably, none of these observations would have been obvious without the discovery of high liability PDZK1 alleles since none of the observed signals, although robust, could reach genome-wide significance.

[0050] This genetic architecture is reminiscent of the mechanism of disease of another important component of the LDL cholesterol pathway, PCSK9, wherein whole gene duplications causeDocket No.: 60115-706.601 phenotypes indistinguishable from PDZK1 whilst deletions cause the reverse effect (low LDL, high HDL and likely protection from cascade adverse effects associated with excessive cholesterol deposition). Based on the findings, developed was a sequencing panel that extends the diagnostic strategy for screening candidate FH patients. Various treatments may be administered to female patients for hyperlipidemia and related co-morbidities.EXAMPLES

[0051] The following examples are provided to further illustrate some embodiments of the present disclosure, but are not intended to limit the scope of the disclosure; it will be understood by their exemplary nature that other procedures, methodologies, or techniques may alternatively be used.Example 1: Genomic Architecture AnalysisRecruitment and clinical environment

[0052] The FH Mexican Registry design and rationale has been described. This research protocol was approved by the Institutional Research and Ethics Committees (Ref. 2571). Patients identified with baseline LDL-C levels >190 mg / dL (adults) or >160 mg / dL (children) are evaluated in accordance with the Dutch Lipid Clinic Network criteria 19. Only those cases that have a definite or probable diagnosis are included. Decision to participate in the study and informed consent form (ICF) are obtained from each subject. Within the web-based registry information is captured from subjects with heterozygous / homozygous familial hypercholesterolemia (FH) clinical diagnosis.Sequencing Twist kit (WES and DP)

[0053] In the laboratory facilities of Galatea bio, we sequenced 302 samples from 169 families using whole exome sequencing (WES), whole genome sequencing, and a custom low-pass genome sequencing approach. Combined, we were able to interrogate these samples for both common and rare genetic variants associated with FH. WES sequencing library was prepared using the Twist Exome 2.0 / Twist Library Preparation EF Kit 2.0 with Twist SNP Density Panel spiked in. The hybridized fragments were then sequenced as 2 x 150 bp reads on an Illumina NovaSeq 6000 instrument (Illumina, San Diego, CA, USA).SNV analysis

[0054] Initially, we focused the analysis on three genes that may cause FH: LDLR, APOB and PCSK9. Variants were annotated using Variant Effect Predictor (VEP) and filtered for population-based gnomAD allele frequency less than 1% (gnomad.broadinstitute.org / ). ClinVarDocket No.: 60115-706.601 database was used to classify the variants in terms of pathogenicity (ncbi.nlm.nih.gov / clinvar). Next, we examined non-classical genes that may cause FH (APOE, LIPA, ABCG5, ABCG ).

[0055] To discover potentially new genes for FH, we searched for rare variants in 47 genes in the core cholesterol pathway (KEGG pathway: hsa048979) and added interactors (55 genes) acquired from the Intact database.CNV calling

[0056] We performed CNV Calling using two algorithms, designed for WES data, Codex and Exom eDepth and using the recommended default settings of each program. Briefly, to generate the normalized coverage metrics for each exon, we supplied as inputs the bed files of the FH genes and the bam files processed in batches according to each WES run. CODEX normalizes the depth of coverage by applying a Poisson latent factor model that removes bias due to GC content, exon capture and amplification efficiency, and latent systemic artifacts. Five Poisson latent factors were included in the normalization model for this dataset. For each sample, ExomeDepth builds a reference set that includes samples with similar read count distribution and then performs CNV calling. We prioritized CNVs that were detected by both programs. CNVs were called for all FH genes, as well as for new candidate genes involved in or interacting with key effectors of the cholesterol pathway.CNV Validation using Taqman assay

[0057] To validate the CNV, we designed custom TaqMan probes and primers targeting the genomic region of PDZK1. We then conducted TaqMan assays in the laboratory to assess the presence of the CNV.PheWas studies

[0058] We utilized the pheWeb database to retrieve pheWas results for variants in PDZK1. This database contains the results of pheWas studies conducted in large datasets, including the UKBiobank, the Michigan Genomics Initiative and the FinMetSeq study.Results

[0059] Following CNV analysis, we identified a whole-gene duplication of PDZK1 in two unrelated FH patients (Figure 1). PDZK1, which encodes a scaffolding protein from the PDZ domain-containing family, was included in the list of candidate genes for FH because it interacts with and stabilizes the scavenger receptor class B type I (SR-BI), which in turn is involved in high-density lipoprotein (HDL) metabolism. Specifically, SR-BI facilitates the uptake of HDL cholesterol by the liver, a critical process for maintaining cholesterol homeostasis. Both patients were male and, aside from a confirmed diagnosis of FH, we were also able to access detailed clinical records for one of them. At the age of 26, this individual presented with a totalDocket No.: 60115-706.601 cholesterol level of 456 mg / dL, LDL cholesterol of 374 mg / dL, and HDL cholesterol of 44.8 mg / dL.

[0060] Table 1 presents the CNV metrics generated by the two programs, Codex and Exomedepth. In particular, Ratio Metrics such as LRatio for CODEX highlight the deviation of observed coverage from expected levels, while the Reads Ratio for Exomedepth quantifies the proportion of observed reads compared to expected reads. Coverage Metrics from Codex, include both raw and normalized coverage, while Exomedepth provides the expected and observed read counts to assess CNV presence and size.

[0061] Critically, this duplication was found only in 5 / 200,000 genomes from the All of Us Project (allofus.nih.gov / ), thus showing a significant enrichment (p< 1 O'4) in the FH cohort.

[0062] Table 1 : CNV metrics generated by the two programs, Codex and Exomedepth, indicating the presence of a duplication of the entire PDZK1 gene in two patients.

[0063] FIG. 1 shows a schematic representation of the duplication (chrl : 145670985- 145707368) detected in two unrelated patients projected on the UCSC Genome Browser (https: / / genome.ucsc.edu / ).

[0064] Next, testing was performed to determine any association of LDL with metabolismrelevant phenotypes in the general population. For this purpose, PheWas analyses were performed across several large cohort datasets.

[0065] In the UKBiobank the intron variant c.793+572T>C (rsl40340656) was detected in the PDZK1 locus to be associated with higher risk of hypertension (Beta=0.13, SE=0.033, p=6.7xl0'5); essential hypertension (Beta=0.13, SE=0.033, p=7.3*10'5); and hyperglyceridemia (Beta=3.5, SE=0.99, p=4.8x!0'4) (Figure 2).Docket No.: 60115-706.601

[0066] FIG. 2 shows an example of a phenome-wide view for variant rs2976384 showing phenome-wide significant association with traits from the UK Biobank including hypertension (A), essential hypertension (B) and hyperglyceridemia (C).

[0067] Furthermore, in the FinMetSeq cohort, a different intronic variant, c.211-69A>G (rsl553701870) was associated with significantly lower levels of saturated fatty acids (Beta = - 1.7, SE = 0.45, p = 1.4 x 10’4) and LDL cholesterol (Beta = -1.4, SE = 0.38, p = 3.3 x IO’4). Finally, a third intronic variant, c.416-195A>G (rsl649823291) was associated with reduced levels of ApoB (Beta = -1.1, SE = 0.33, p = 5.6 x 10'4) within the same cohort as shown in FIG. 2.

[0068] FIG. 3 shows a phenome-wide view (PheWAS) for variants rsl553701870 and rsl649823291 showing phenome-wide significant association with traits from the FinMetSeq study including saturated fatty acids (A), LDL cholesterol (B) and APOB levels (C).Example 2: Ancestry Analysis

[0069] Whole exome sequencing (WES) libraries were prepared with the Twist library Preparation EF Kit 2.0 with Twist Exome 2.0 / Twist SNP Diversity Panel in a 4 / 1 ratio (Twist Biosciences, San Francisco, CA, USA) according to manufacturer’s protocol. Briefly, genomic DNA was fragmented enzymatically, ligated to a universal adapter and amplified with Twist Unique Dual Index Primers. Amplified samples were then hybridized overnight with the Twist 2.0 and Diversity panels. The hybridized fragments were sequenced as 2 x 150 bp reads on an Illumina NovaSeq 6000 instrument (Illumina, San Diego, CA, USA). The exome (283,942 regions) and the SNP diversity panel (1,380,542 SNPs) were sequenced at an approximate mean depth of 50X and 10X coverage, respectively. In addition, a subset of samples carrying putative copy number variants or pathogenic missense variants were sequenced by both Illumina shortread WGS and Nanopore sequencing. Nanopore sequencing was done as follows: Sequencing libraries were prepared from 1 ug of DNA using the Oxford Nanopore Technologies (ONT) Ligation Sequencing kit V14 (SQK-LSK114). The samples were loaded on the PromethlON Flow Cells R10 (M Version) and were wash and reloaded after approximately 24 hours. The samples were sequenced for a total of 72 hours on a PromethlON 24 sequencer. Mapping, alignment, and variant calling of the long reads was performed with Sentieon DNAscope software (Sentieon, Inc., San Jose, CA, USA). Characteristics of the cohort are shown in FIGs.4A-4DVariant calling and Quality Control

[0070] We generated VCF files with the Illumina DRAGENTM Secondary Analysis (Version 4.2.4), which we filtered using default parameters, including a quality score < 10.41 for SNVs, aDocket No.: 60115-706.601 quality score < 7.83 for indels, a read depth < 1, and the removal of genotype calls inconsistent with the expected chromosome ploidy. To confirm survey-based family architecture, we used Plink vl.9 to perform a sex check and to calculate relatedness in all pairs of samples in the cohort. Finally, we imputed variants using the GLIMPSE2 algorithm with a reference panel from 4,091 individuals from the 1000 Genomes Project and the Human Genome Diversity Project, as implemented by Gencove Inc. (New York, NY, USA).Ancestry and PRS

[0071] We used, G-Nomix to calculate local and global ancestry proportions, the ancestry model was trained to detect genomic loci with the following ancestries: Native American, European, African, Southeast Asian, Central and South Asian, East Asian, Levantine and Middle Eastern, and Oceanian. We also used the Galatea Bio PRS, which includes an ancestry-adjusted PRS algorithm, to estimate the genetic risk of high LDL-C levels for all the samples in the cohort. The algorithm performs ancestry adjustment by normalizing the score with a mean and standard deviation computed using a neighborhood of genetically-close individuals from a proprietary database of diverse samples covering populations across the world. As controls, we used an - internal database of individuals from Mexico with no evidence of FH diagnosis. To reduce potential biases originated by using a population-level control, for each sample in the FH cohort, we select the closest individual (in the PCA space) of the population-level control cohort in order to generate a subset of control individuals with an ancestry composition and genetic structure similar as possible as the FH cohort. PRS for LDL-C levels were calculated using the variants and weights of the Galatea Bio LDL PRS model.Mutation analysis

[0072] SNV annotation

[0073] Initially, we focused the analysis on genes that may cause FH: LDLR, APOB, PCSK9 and LDBLAPP We annotated variants using Variant Effect Predictor (VEP) and filtered for population-based gnomAD allele frequency less than 1% (https: / / gnomad.broadinstitute.org / ). In addition to the classical genes, FH cases with pathogenic variants in APOE have been reported, suggesting that this locus should also be included in genetic screening for FH. Next, we examined non-classical genes implicated either with FH or with conditions mimicking FH; we searched for rare and likely pathogenic (per modified ACMG criteria) variants in these genes.

[0074] We also performed Al-based prioritization and scoring of candidate disease genes and diagnostic conditions with the Fabric Enterprise GEM algorithm (Fabric Genomics, Inc., Oakland, CA USA), which uses - variant files and patient’s HPO terms as input and combines variant prioritization algorithms and clinical / genomic database annotations using Bayesian scoring. - The algorithm adjusts its parameters based on proband-specific variant data and usesDocket No.: 60115-706.601 naive Bayes to integrate factors such as ancestry, gene location, inheritance patterns, and proband sex to refine variant scoring. Furthermore, the variants were classified according to the ACMG / AMP guidelines with the aid of Fabric ACE automated variant classification engine, adding manual curation for literature-based criteria to obtain a final classification.

[0075] CNV calling and validation

[0076] To call Copy Number Variation (CNVs) we used the WES read depth data. We ran CNV calling using two algorithms designed for WES data: Codex and ExomeDepth with the recommended default settings of each program. Briefly, to generate the normalized coverage metrics for each exon, we supplied as inputs the BED files of the target gene set and the BAM files processed in batches according to each WES run. CODEX normalizes the depth of coverage by applying a Poisson latent factor model that removes bias due to factors such as GC content, exon capture and amplification efficiency, and latent systemic artifacts. Five Poisson latent factors were included in the normalization model for this dataset. For each sample, ExomeDepth built a reference set that included samples with similar read count distribution and then performed CNV calling. We prioritized CNVs detected by both programs.

[0077] We confirmed candidate CNVs (deletions and duplications) involving multiple exons with custom-made TaqMan assays (ThermoFisher Scientific, Waltham, MA, USA). Single-exon CNVs were validated orthogonally against CNV calls derived from WGS data from the suspected cases. For WGS data, we employed the Illumina DRAGEN Secondary Analysis v4.2 CNV / SV caller with default parameters, which integrates sequencing depth and split-read analyses. The DRAGEN pipeline has shown superior sensitivity and specificity compared to other CNV callers used in WGS data.

[0078] Breakpoint analysis

[0079] We generated WGS data for three samples with candidate CNVs: a carrier of a whole gene duplication in PCSK9 and two carriers of an exonl3-14 LDLR deletion. By using the sequencing reads of the WGS we identified the breakpoints of each CNV event using the Manta algorithm. Next, we extracted the region surrounding the breakpoints (+-1000 bp on each side) and searched for repetitive elements using RepeatMasker. The repetitive elements at the left and right breakpoints were then compared for similarity with BLASTn.

[0080] Analysis for genes participating in cholesterol pathway

[0081] In addition to all dyslipidemia genes, we also studied rare variants (SNVs and CNVs) in all genes that encode components of the cholesterol pathway. In particular, we searched for likely deleterious variants in 47 genes defined to participate in cholesterol biosynthesis and transport (KEGG pathway: hsa048979) along with their interactors (55 genes) extracted from the Intact database (Table 2.) Ancestry analysis results are shown in FIG. 5, a bar chart whereDocket No.: 60115-706.601 each bar corresponds to an individual and the height of each color within the bar indicates the proportion of the individual's genome that can be attributed to the corresponding ancestral population. The "unassigned" fraction of the genome represents regions that could not be confidently assigned to any specific ancestry.

[0082] Table 2: Ancestry analysis for samples carrying ultra-rare or unreported variants in gnomAD.

[0083] CNV calling in UK Biobank

[0084] To replicate the variants with respect to their involvement in hypercholesterolemia, we examined the cholesterol profiles of UK Biobank (UKB) participants in the UK. Copy number variant (CNV) detection was performed using PennCNV version 1.0.5, with log R ratio and B- allele frequency (BAF) data provided by the UK Biobank, setting missing values to NA. Across the cohort, CNVs on chromosome 1 were analyzed separately for each of the 106 genotyping batches, each containing approximately 4,700 samples. Batch-specific population frequency of the B allele (PFB) files were generated by calculating the mean BAF for each batch. Probes with missing PFB values were excluded in a batch-specific manner. The hidden Markov model (HMM) file for the Affymetrix genome-wide 6.0 array, included in the PennCNV-Affy package, was used directly for analysis.Results

[0085] Cohort demographics, phenotypes, and genetic structure

[0086] We studied 170 families from the Mexican FH National Registry. Of these, 111 were composed of a single affected individual with an available DNA sample, while in 59 families weDocket No.: 60115-706.601 were able to phenotype and recruit multiple members across several generations; we focused on individuals with a “definite” or “probable” FH diagnosis as defined by the Dutch Lipid Clinic Network Score (DLCNS), which includes LDL-C levels >190mg / dl in adults and 160mg / dl in children. A summary of cohort characteristics is shown in Fig. 1 and Table 1, where we recorded standard anthropometric data; lipid profiles (including secondary dyslipidemia phenotypes such as xanthomas or corneal arcus); and other relevant phenotypic data such as the incidence of stroke, coronary heart disease, and Type 2 Diabetes.

[0087] Seventy-seven individuals received targeted genetic testing for / . / J / . / ?, APOB and PCSK9, the data from which have been reported but we pursued a cohort-wide, comprehensive analysis of all LDL-relevant genes and signatures by performing joint Illumina short-read WES and low-pass genotype-by-sequencing (GBS) of -1.4M Diversity Panel SNPs. QC showed that >98% of the WES-targeted regions were covered at a minimum of 20x, whilst the diversity SNP set achieved 2-10x coverage for -80% of sites. We excluded five samples from further analyses: three were proven to be duplicates and two were of low quality, leaving us with data for 300 individuals from 167 families.

[0088] We first assessed all survey-based familial relationships. Identity-by-descent analysis confirmed kinships for 278 / 300 individuals. Eight individuals did not map to any family and were thus reassigned as singletons, while the remaining 14 individuals were found to be 1stor 2nd degree relatives of existing multi-member families and were reassigned as such.

[0089] Next, we determined eight-label ancestries for each individual in the study (Fig. 5 / Table 2.). For the genetic analyses from this geographic region, 6 individuals showed nearly 100% Amerindian (n=5) or European (n=l) ancestry, while the -remained of the cohort were admixed between these two ancestral groups. We did not determine any significant representation from other ancestries, with the exception of ~10 individuals whose genome contained 10% of more segments of East Asian, African or Middle Eastern origin. This pattern has been observed in similar admixed samples from Mexico, derived from the three-way admixture process that took place after the arrival of Europeans to the continent.Analysis of FH genes

[0090] As a first step toward mapping the genetic architecture of FH in this cohort, we interrogated all rare variants in the three “classical” genes dominant FH genes: LDLR, APOB and PCSK9, as well as LDLRAP1 and APOE that have been reported recently to contribute deleterious variants in autosomal recessive and dominant FH.SNV discovery

[0091] Consistent with The in FH patients from multiple ancestral backgrounds, rare SNVs in LDLR contributed the bulk of pathogenic mutations in the cohort. Using the modified ACMGDocket No.: 60115-706.601Criteria (see Methods), the Al-based Fabric GEM algorithm that prioritizes variants based on HPO terms, and prior functional studies, we identified pathogenic and likely pathogenic variants in 141 individuals from 63 families (Table 3). Amongst the 59 families with multiple affected individuals, 59% carried pathogenic variants: twenty-seven were missense, while eight are predicted to induce truncations that trigger nonsense-mediated decay. Amongst singletons, the diagnostic yield was lower, 27 in 108 individuals. This is not surprising, given that elevated LDL cholesterol is common among older individuals (and prominent in Mexico, where hypercholesterolemia affects between 11% and 36% of young men and women). Variants annotated as VUS or benign are listed in Table S2.

[0092] Three variants were recurrent: it was determined that p.Cys352Tyr in four families, p.Glul 13GlyfsTerl7 in two families and p.Tyr400_Phe402del in two families, with no evidence of direct relatedness, intimating more distant common ancestors. In addition, and consistent with the ancestry data, 39 / 65 pathogenic LDLR SNVs (27 missense, 12 splice / nonsense), 13 have been reported only in patients from Mexico, Italy or the Iberian Peninsula, while five SNVs are (Table 3). Ancestry ideograms were shown for samples in FIG. 6, showing LDLR: p.Glul 13GlyfsTerl7, LDLR:p.Met298IlefsTer3, APOB:exonl deletion, and PCSK9:p.Arg218Lys. These ideograms illustrate the local ancestry composition at gene loci where pathogenic variants were identified.Docket No.: 60115-706.601

[0093] Table 3: Pathogenic Variants Identified in FH Genes (LDLR, APOB, PCSK9, andLDRAP1) Within the FH Cohort, Along with Population-Based Frequencies from gnomAD,Functional studies and reported ethnicities of carriers in other studies. Source studies for this data can be found in supplementary references

[0094] To probe this observation further, we derived the local ancestry around mutations that were ultra-rare or unreported in gnomAD and asked whether these variants might have arisen in an Amerindian (AMR) background, it was determined that this to be true in numerous instances (Table 3). For example, it was determined that ADAR: p.Glul 13GlyfsTerl7 and p.Met298IlefsTer3 primarily in AMR / AMR chromosomes (73% and 86%, respectively). In contrast, ADAR:p.Leu346AlafsTerl2 was mostly EUR (86%). Similarly, thePCSK9.^ .Arg218Lys and LDLRAP 1 :c.459+2T>G alleles mapped exclusively in an AMR background. These findings highlight population-specific contributions to these rare variantsDocket No.: 60115-706.601(Table 3; Fig. 6) and also suggest that some alleles might have either arisen independently or predate the settlement of the Americas.

[0095] We also noted a non-random distribution of mutations across LDLR (Fig. 7). Six of 27 missense variants map to exon 4, followed by further clustering in exons 6, 8, 9 and 14. Exon 4 encodes the ligand binding domain and has been found to harbor the highest number of pathogenic mutations in FH patients, even after adjusting for size of exons (Fig. 7).

[0096] APOB variants were more challenging to interpret; the locus harbors numerous rare and ultra-rare variants of unknown functional effect, with scant evidence to enable inclusion or rejection from a role in FH. The APOB was a rare cause of FH in Mexico, it was determined that a single SNV (p.Tyr3560Cys) that segregated in the three affected members of one family. This allele has been implicated in FH in patients from Portugal and Australia, and was absent from the internal control cohort (n=l,500 chromosomes) All remaining candidate variants bore either inconclusive genetic evidence and / or lacked functional insights. Notably, we did not find any instances of the R3500W variant, the most frequent APOB mutation in Northern European populations. Similarly, the less frequent but recurrent Northern European alleles 3500Q and R3531C were likewise absent.

[0097] Finally, it was determined that deleterious SNVs in each o APOE and LDLRAPL For APOE, we detected a three-amino acid deletion (p.Leul67del), shown previously to cause adult Dominant Hypercholesterolemia and FH. In LDLRAP1 , the only recessive locus in this series, it was determined that a single individual bearing a homozygous splice variant (c.459+2T>G) reported previously in patients from Mexico, unrelated to the individual in the study.

[0098] CNV analysis

[0099] The structural variation accounts for a considerable percentage of FH cases. For example, copy number variations (CNVs) in LDLR have been shown in as much 10% of FH cases. Here, using a combination of exome depth coverage and orthogonal validations by Taqman PCR assays, it was determined that two CNVs in LDLR (Table 3): a deletion of exons 13-14 that segregated with the disease in all three patients from family 107 and a deletion of exons 1-15 in the sole affected individual of family 82.

[0100] In contrast to LDLR, CNVs have not been reported for APOB. In the cohort, however, it was determined that a deletion of exon 1 cL APOB in four unrelated individuals (families 165, 196, 33 and 84). The variant is likely pathogenic since it deletes the start Methionine and induces a frameshift mutation and premature termination. Notably, the patient from family 165 was also heterozygous for the pathogenic LDLR variant p.Lys393Glu, raising the possibility that the two variants may act synergistically to contribute to FH. We failed to find the 4 / YV> deletion in 741 internal, ancestry-adjusted normolipidemic controls as well as in the gnomAD and theDocket No.: 60115-706.601UKB databases. Finally, we identified an individual with a duplication of PCSK9 described previously in two unrelated Canadian FH patients.

[0101] All observed CNVs were confirmed by custom Taqman RT-PCR assays, which were also used for segregation analysis. In addition, for the families bearing the LDLR exon 13-14 CNV and the PCSK9 duplication, we had sufficient DNA to perform Illumina short-read wholegenome sequencing and Oxford Nanopore long-read sequencing. In addition to confirming these lesions with both technologies, we also mapped all four breakpoints. For LDLR, both the right and left breakpoints mapped within 4 / z / SP sequences in introns 12 and 14 respectively that shared 86% sequence similarity and were oriented in the same direction, supporting s Alu-Alu recombination mechanism, it was determined that the same genomic organization in PCSK9'. breakpoint analysis identified 4 / z / SN sequences at the breakpoint junctions. These elements also exhibited 85% similarity and maintained the same orientation.

[0102] Phenocopy rate

[0103] Amongst the 59 families with multiple members available to this study, we identified 24 patients who did not inherit the family mutation but were diagnosed with FH because of their lipid profile and family history. In Family 7, for example, we identified four patients who carry the pathogenic LDLR p.C352Y variant, along with one first-degree relative who tested negative for the mutation yet had been considered affected by virtue of family history, high LDL cholesterol (189) and low HDL cholesterol (29). Considering the entire cohort, but restricting the analysis to families with 2+ affected individuals, it was determined that a maximal phenocopy rate of 12.5% (24 / 192 individuals), likely driven by the fact that FH is clinically indistinguishable from sporadic hyperlipidemia, a common disease.

[0104] Overlap with primary dyslipidemias and lipodystrophies

[0105] Variants in ABCG5, ABCG8, axA LPA have been causally associated with sitosterolemia, a condition that shares symptoms with FH, including elevated cholesterol levels, xanthomas, and increased cardiovascular risk while variants m A CA l can cause hypoalphalipoproteinemia, characterized by low plasma HDL levels. However, the two conditions differ in terms of pathophysiology: sitosterolemia is characterized by the accumulation of dietary plant sterols in the blood and tissues due to defective sterol transporters, whereas FH is marked by elevated levels of LDL cholesterol due to impaired LDL clearance. Furthermore, sitosterolemia is found as an autosomal recessive trait.

[0106] None of the patients carried homozygous or compound heterozygous variants nABCAl, A CG5, or ABCG8. In one patient it was determined that two variants in ABCG5 (p.R105L and p. A98G); the first variant is absent from all databases, while the second has been reported once in an FH patient in the ClinVar database, is conserved across species, but is predicted to beDocket No.: 60115-706.601 tolerated, thereby receiving a preliminary designation as a VUS. Similarly, it was determined that two VUSs in LIPA.

[0107] However, we did discover two pathogenic variants, one in PLIN1 and one in CREB3L3, each reported previously in dominant familial partial lipodystrophy (FPL) and hypertriglyceridemia. Common signs with FH include high levels of triglycerides and dyslipidemia in patients with FPL while patients with hypertriglyceridemia have high levels of total cholesterol, triglycerides and low levels of HDL. Although we cannot exclude the possibility that these are patients misdiagnosed with FH, we consider that possibility unlikely; typical FPL patients exhibit a striking loss of adipose tissue in their limbs and hips, which we did not observe. Instead, we favor the possibility that these cases represent phenotypic expansion PLIN 1 and CREB3L3.

[0108] Diagnostic yield

[0109] Focusing on families with multiple affected members (and as such more likely to capture a true FH diagnosis), we identified pathogenic variants in 125 / 192 patients across 40 of 59 families, resulting in a diagnostic yield of 65%. Unsurprisingly, the diagnostic yield in singleton cases was lower at 32% (35 / 108 cases), further underscoring the importance of family history as a criterion for FH diagnosis. FIGs. 8A-8B illustrates the distribution of allele types in the cohort.

[0110] Dissection of the remaining missing variation[OHl] We turned to the of genetic causality in the 35% of patients who lacked a molecular diagnosis, which even under the assumption of a 12.5% phenocopy rate, remains substantial. We explored two possibilities: a) that the remainder of the genetic burden for FH is driven by common variants; and b) that some of the mutation-negative patients might harbor rare, high liability alleles in hitherto unknown genes.Polygenic Risk Score Analysis

[0112] To answer the first question, we computed the polygenic risk score for high LDL across the entire cohort (irrespective of their genetic burden in the FH+genes) and compared that to the PRS distribution of a control cohort.

[0113] Our LDL PRS model predicted significantly different risk scores between the FH cohort and controls (Figs. 9A-9C). Among the patients with a definitive molecular diagnosis, 10.6% mapped above the 95thcentile of LDL PRS distribution, compared to ancestry-adjusted controls (3.0%; Figs. 9A-9B). The enrichment was even more pronounced among the FH patients without a molecular diagnosis, 19.7% of whom scored above the 95th centile of risk. By using an Exact Fisher’s Exact Tail with Holm-Bonferroni correction, we determined the difference in the 95thcentile tails -to be significant when compared to either controls (p-value < 0.001) orDocket No.: 60115-706.601 non-carriers of pathogenic variants (p-value =0.03). Furthermore, the differences between the average scores between groups were also significant when applying a Welch’s t-test: when comparing cases vs controls (p-value < 0.001) and carriers vs non-carriers (p-value < 0.001). By applying a linear regression from ancestry-adjusted scores to cumulative LDL years, we observed that individuals at the tail of the distribution carried the same risk as individuals with high liability heterozygous pathogenic alleles, suggesting that some patients do not necessarily carry rare, high liability alleles but distributed additive risk from common variants (Fig. 9C). Finally, we measured the risk scores using over 4000 models available on the PGSCatalog (www.pgscatalog.org) and after applying a Benhamin-Hochberg correction, a total of 17 models showed a significant p-value from which five are LDL models, and 14 are cholesterol-related. Variants in genes encoding cholesterol pathway components

[0114] Finally, it was determined whether some of the remaining affected FH individuals might carry rare, high liability variants in hitherto unknown FH genes. To test this idea, we focused on the 47 genes that may to encode components of cholesterol biosynthesis and transport and an additional 55 genes that encode lst-order interactors (total 102 genes).

[0115] We first filtered for loci that fulfill the following three criteria: a) they carry non- synonymous, predicted deleterious alleles with a population frequency below the carrier frequency of FH (1 :200 ) in public databases such as UKB and gnomAD; b) such alleles ought to be absent from the internal, normolipidemic, ancestry-adjusted controls; and c) they either show evidence of segregation within the mutation-negative families or show recurrent alleles in at least two undiagnosed, unrelated individuals.

[0116] Two unrelated (pi-hat=O) individuals in the cohort carry a whole-gene duplication of PDZK1. PDZK1 encodes a scaffolding protein that interacts with membrane proteins, including the high-density lipoprotein (HDL) scavenger receptor class B type I (SR-BI). Typing the control set, we were encouraged to see the variant absent from 741 normolipidemic individuals.

[0117] Given these findings, we turned to the UKB, where we analyzed array data from 498,200 individuals. We identified three carriers of the PDZK1 duplication; in each case, their LDL and total cholesterol levels mapped to the top quartile of the entire cohort for LDL and total cholesterol, with two individual recording LDL-C >190 and the third individual recording elevated LDL-C (163).

[0118] PheWas studies

[0119] To investigate further the possible involvement of PDZK1 variation in FH, we mined the PheWeb database for relevant phenotypes that include a clinical diagnosis of dyslipidemia hypertension, or cardiovascular disease, as well as direct lipid measurements, all phenotyped in large cohorts, such as the UK Biobank, the Michigan Genomics Initiative, and FinMetSeq.Docket No.: 60115-706.601

[0120] We found two intronic variants and one intergenic variant 41kb upstream of PDZK1 to be associated with hypertension and lipid levels. Specifically, the intronic variant c.793+572T>C (rsl40340656) showed association with an elevated risk for hypertension (Beta = 0.13, SE = 0.033, p = 6.7 x 10'5), essential hypertension (Beta = 0.13, SE = 0.033, p = 7.3 x 10'5), and hypertriglyceridemia (Beta = 3.5, SE = 0.99, p = 4.8 x 10'4) in the UK Biobank. The second intronic variant, c.211-69A>G (rsl553701870), exhibits a protective effect in FinMetSeq, with significantly lower levels of saturated fatty acids (Beta = -1.7, SE = 0.45, p = 1.4 x IO-4) and LDL cholesterol (Beta = -1.4, SE = 0.38, p = 3.3 x 10'4). Finally, the intergenic variant rsl649823291 was associated with reduced levels of APOB abundance (Beta = -1.1, SE = 0.33, p = 5.6 x 10'4), which was protective of cardiovascular disease and mortality secondary to heart attack and stroke.

[0121] A 300-strong cohort from the Mexican FH registry was used. In some cases, across populations, we attribute >80% of FH cases with a definitive molecular diagnosis to dominant mutations in LDLR. The most common mutation was an insertion in exon 4(p.Glul 13GlyfsTerl7), reported previously in Mexican FH patients. Likewise, another three variants (p.Cys352Tyr, p.Leu346AlafsTerl2, p.Cys289Arg) have only been reported in the Mexican population. More broadly, almost a third of the LDLR mutations in the cohort have only been seen in Spanish, Portuguese, and Italian populations, but not in Northern Europeans (despite the supernumerary studies in the latter). In one family, we also found the p.Cys681Ter allele, that may to account for 60% of FH cases in Lebanon, but also reported in other Latin American cohorts. The enrichment of alleles from Mediterranean Europe and the Middle East reflects the distinct genetic background of these geographic regions and their contribution to the genetic architecture of Latin America. Of note, six variants (p.Met298IlefsTer3, p.Lys393Glu, p.His407Gln, p.Ser648Phe, c,1060+2del, deletion of exonl-15) are and might represent variants arisen in a Native American background. Discovery of such alleles remains critical, both for improving diagnostic capabilities in admixed individuals and for informing the biology of the disorder.

[0122] APOB and PCSK9 mutations were rare in the cohort. Of note, the most frequent FH- causing mutation in “Europeans”, p.Arg3500Gln, was absent, further underscoring the fact that “Europeans” represent a sometimes simplistic label with limited utility in interpreting genetic variation in admixed Latin American populations. Moreover, among the rare variants it was determined that in APOB, only two offer sufficiently compelling evidence for pathogenicity. The first is a p.Tyr3560Cys variant, reported previously in FH patients of Portuguese and Dutch ancestry. The second is a novel, heterozygous deletion of exon 1; we do not know whether the discovery of this allele reflects a variant arisen in a Native AmericanDocket No.: 60115-706.601 background, or that single-exon CNVs are often disregarded or not detected in exome-based diagnostic tests.

[0123] Extending the analysis to non-classical FH genes, we identified several pathogenic variants, starting with a p.Leul67del variant in 4 / YVi. This variant, reported previously in FH patients of Italian and Spanish descent, disrupts the leucine zipper near the 4th a-helical structure of APOE, a region that mediates its interaction with LDLR. The mutation increases the binding affinity of APOE to LDLR, preventing receptor transport back to the hepatocyte surface to clear cholesterol. Carriers of the p.Leul67del mutation exhibit a better response to lipid lowering therapy than LDLR carriers further motivating the addition of APOE in FH genetic panels. Based at least in part on these carriers, pathogenic mutations were detected and verified in PLIN1 and CREB3L3, each detected previously in familial partial lipodystrophy (FPL). However, given that the patients do not exhibit the adipose deposits that hallmark FPL, we once again suggest that a more “lipodystrophy” centered approach to analyzing panels, exomes and genomes for FH might serve patients and their families better.

[0124] In multiplex families, the diagnostic yield approaches 70%, at the upper bound of reported work and significantly higher than cross-sectional population-based studies, where some studies have reported a mutational yield in the single digits. In the cohort, even in singletons, the diagnostic yield exceeded 30%. We attribute this diagnostic yield to the stringent application of the Dutch diagnostic criteria for FH; to the expansion of genes interrogated to include both loci that cause other clinical forms of dyslipidemia and the remaining components of the cholesterol pathway; and the accurate detection of CNVs, which contribute 5-10% to the mutational burden but are often undetected in targeted sequencing clinical tests. We also note that application of appropriate PRS analysis has revealed two points that merit attention. First, we observed that, even in the presence of a rare, high liability allele in a FH gene, a significant fraction of the patients maps to the 95thor 99thcentile of PRS-derived risk, suggesting that common alleles likely exert an effect on rare, in principle penetrant mutations. Longitudinal studies on disease onset, progression and cardiovascular outcomes, as well as response to a variety of medications, will be necessary to understand the effect of such variation on the phenotype. In addition, we also found that the PRS signal was significantly more pronounced in bona fide FH patients bereft of high liability alleles, arguing that individuals will manifest FH because of their common allele burden and not rare, peri-Mendelian variants. In that context, we support the development of clinical PRS tools and suggest that, for some disorders such as FH, they might have diagnostic and clinical utility.

[0125] Finally, we draw attention to the fact that most clinical genetic tests for FH (and other disorders) focus on single nucleotide variants and small indels, even though CNVs can accountDocket No.: 60115-706.601 for as much as 10% of the genetic burden, the findings further underscore the importance of structural variants in FH. The continued development of robust CNV analysis and validation tools, both through improved exome methods and by transitioning to whole-genome sequencing will be important to support molecular diagnosis and new gene discovery for FH and other disorders.

[0126] The detection of clinically-relevant duplications can be challenging, yet carry significant ramifications. For example, one of the patients was a carrier of a whole-gene duplication of PCSK9. This finding has treatment implications, as statin-induced upregulation of PCSK9 has been suggested to be less effective due to the presence of an additional PCSK9 copy. Second, duplication analysis was a catalyst for the discovery of a candidate FH locus via the identification, and subsequent molecular confirmation of a duplication in PDZK1 in two unrelated patients. We recapitulated this finding in the UKB, where it was determined that the same lesion in three individuals, all of whom showed elevated lipid profiles. The absence of any phenotypic signal for loss-of-function events at this locus suggests that dosage sensitivity is the most likely contributory mechanism to FH. Consistent with this notion, we identified two intergenic and one intragenic SNP that show association with hypertension and lipid levels. We are particularly intrigued by the fact that one of the SNPs (rsl553701870) likely exerts a protective effect. Based on population-wide PheWAS of UKB participants with loss-of function PDZK1 mutations and a deeper analysis of -600 individuals with gene-disruptive deletions, we suggest that loss-of function at this locus is not likely to impart obvious pathology in humans. Indeed, heterozygous knockout mice are asymptomatic, while homozygous null PDZK1 mice likewise exhibit normal liver and renal physiology, with albeit increased HDL abundance.Together, the human and murine data suggest that suppression or ablation of PDZK1 might be of therapeutic benefit and offer the development of a class of drugs orthogonal to the current set of offerings such as statins and CGS' inhibitors. We are intrigued by the recent association of PDZK1 with gout, a phenotype strongly associated with dyslipidemia, most notably in a metaanalysis of 514,000 individuals from the Korean National Health Insurance Service-Health Screening Cohort. We speculate that the renal, not the liver role(s) of PDZK1 might be most relevant to its therapeutic potential and merit further investigation. More broadly, this study highlights the persistent and urgent need to study well-phenotyped cohorts across the globe and to combine iterative learnings that can inform and ultimately improve patient management and public health policy.

[0127] While preferred embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. It is not intended that the invention be limited by the specific examplesDocket No.: 60115-706.601 provided within the specification. While the invention has been described with reference to the aforementioned specification, the descriptions and illustrations of the embodiments herein are not meant to be construed in a limiting sense. Numerous variations, changes, and substitutions will now occur to those skilled in the art without departing from the invention. Furthermore, it shall be understood that all aspects of the invention are not limited to the specific depictions, configurations or relative proportions set forth herein which depend upon a variety of conditions and variables. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention. It is therefore contemplated that the invention shall also cover any such alternatives, modifications, variations or equivalents. It is intended that the following claims define the scope of the invention and that methods and structures within the scope of these claims and their equivalents be covered thereby.

Claims

Docket No.: 60115-706.601CLAIMSWHAT IS CLAIMED IS:

1. A method for detecting a presence or an absence of familial hypercholesterolemia (FH) in a subject, comprising:(a) determining a genotype or a gene expression of PDZK1 in the subject, and(b) detecting the presence or the absence of the FH in the subject based at least in part on the determined genotype or gene expression of PDZK1 in the subject.

2. The method of claim 1, wherein (a) comprises determining the genotype of PDZK1 in the subject.

3. The method of claim 2, wherein the genotype comprises a single nucleotide variant (SNV).

4. The method of claim 2, wherein the genotype comprises a copy number variant (CNV).

5. The method of claim 2, wherein the genotype comprises a whole gene duplication.

6. The method of claim 1, wherein (a) comprises determining the gene expression ofPDZK1 in the subject.

7. The method of claim 6, wherein determining the gene expression of PDZK1 in the subject comprises determining a ribonucleic acid (RNA) expression of PDZK1 in the subject.

8. The method of claim 6, wherein determining the gene expression of PDZK1 in the subject comprises determining a protein expression oiPDZKl in the subject.

9. The method of any one of claims 1-8, further comprising administering a treatment to the subject, responsive to detecting the presence of the FH in the subject.

10. The method of claim 9, wherein the treatment comprises a statin, a cholesterol absorption inhibitor, a bile acid sequestrant, a PCSK9 inhibitor, a lipoprotein apheresis, a bempedoic acid, lomitapide, a liver transplant, or evinacumab.

Images