Sulev Koks

Additional file 2. Table S1: A Linkage Disequilibrium Score Regression (LDSC) calculated genetic correlation (rg) results with standard errors, p-values and FDR corrected p-values; Table S2: Data fields from the UK Biobank used to ascertain first occurrences of this studies respective psychiatric and neurodegenerative disorder outcomes; Table S3: IVW-MR results for all metabolite-neuropsychiatric outcome pairs calculated in this study; Table S4: Sensitivity test results for all metabolite-outcome pairs with IVW-MR significant at p_FDR < 0.05 in our primary Mendelian randomisation analysis; Table S5: Reverse MR estimates for the effects of neuropsychiatric outcomes on metabolites passing sensitivity analyses; Table S6: MR-Egger intercept tests for metabolites with causal effects on psychiatric/neurodegenerative risk and passing sensitivity analyses; Table S7: Cochran's Q heterogeneity tests for metabolites with causal effects on psychiatric/neurodegenerative risk and passing sensitivity analyses; Table S8: Leave-one-out outlier tests for metabolites with causal effects on psychiatric/neurodegenerative risk and passing sensitivity analyses; Table S9: Metabolic pathways for polygenic metabolites with causal effects on psychiatric/neurodegenerative disorders and passing sensitivity analyses; Table S10: Metabolic pathways for single instrument metabolites with causal effects on psychiatric/neurodegenerative disorders and passing sensitivity analyses; Table S11 - LD matrix for influential outlying instruments in the FADS gene cluster implicated in LOO analysis; Table S12: Re-calculated Wald ratio estimates for metabolites with single influential instruments, calculated using the influential instrument for that metabolite only; Table S13: Re-calculated IVW-MR estimates for polygenic metabolites following mt-COJO conditional analysis of the outcome phenotype; Table S14: LDSC calculated genetic correlation matrix for polygenic metabolites, calculated as a filtering step in MR-BMA ; Table S15: MR-BMA results, with the marginal inclusion probability for each metabolite following multivariable analysis in each respective outcome; Table S16: Polygenic risk score results testing the association between MR-BMA prioritised metabolites and their respective psychiatric/neurodegenerative outcome identified in MR analysis; Table S17: Colocalisation results testing for the presence of a shared causal variants between metabolites and their respective psychiatric/neurodegenerative outcomes in regions of single influential variants identified in LOO MR analysis; Table S18: Table of linoleic and arachidonic containing metabolites and their direction of effect in the Wald ratio tests; Table S19: Genes in the regions with suggestive colocalisation between a metabolite and neuropsychiatric outcome; Table S20: Colocalisation results testing for the presence of a shared causal variants between metabolites, psychiatric/neurodegenerative outcomes and gene expression eQTLs in regions where evidence of colocalisation was identified between a metabolite and neuropsychiatric outcome; Table S21: Results of SMR follow-up analysis for SPRYD4 gene expression, Alzheimer's disease and the ratio of histidine-to-glutamine; Table S22: MR results of the effect of glutamine on Alzheimer's disease; Table S23: Leave-one-out analysis for the effect of glutamine on Alzheimer's disease; Table S24: Colocalisation results between glutamine and Alzheimer's and between glutamine and SPRYD4 expression in the region of the influential variant identified in LOO analysis; Table S25: Follow-up SMR results for SPRYD4 expression and glutamine.

Psoriasis is a common inflammatory skin disease with heterogeneous presentation. Up to 30% of individuals have severe disease with a greater surface area of skin involvement, co-morbidity burden and impact on quality of life. Prognostic biomarkers of psoriasis severity could improve allocation of clinical resources and enable earlier intervention to prevent disease progression, and a genetic biomarker would be cost-effective, stable over time, and unaffected by treatment or comorbidity.
Psoriasis severity was studied in four European population-based biobanks (Estonian Biobank, HUNT, FinnGen, UK Biobank) and classified based on level of clinical intervention received, with criteria for severe disease including hospitalisation due to psoriasis, use of systemic immunomodulating therapy or phototherapy. Common genetic variants, polygenic risk scores and traditional epidemiological risk factors were tested for association with severe psoriasis in each of the constituent biobanks and combined through meta-analysis. The distribution of psoriasis polygenic risk was also evaluated in a cohort of 4151 participants in the UK-based severe psoriasis registry, BSTOP, and a cohort of 1461 participants from Novartis clinical trials of secukinumab for psoriasis.
In the population-based datasets, 9738 of 44,904 individuals with psoriasis (21.7%) were classified as having severe disease. Genetic variants within the major histocompatibility complex (MHC) and the TNIP1 and IL12B psoriasis susceptibility loci were associated with severe disease at genome-wide significance (P < 5.0 × 10
). Furthermore, a strong positive correlation was observed between psoriasis susceptibility and severity effect sizes across all psoriasis susceptibility loci. An individual's genetic liability to psoriasis as measured with a polygenic risk score (PRS) strongly associated with disease severity, with a magnitude of effect comparable to established severity risk factors such as obesity and smoking. The top 5% of psoriasis cases by genetic liability to psoriasis were 1.23-to-2.00 times as likely than the average psoriasis case to have severe disease. Psoriasis cases in our external validation datasets (BSTOP registry and Novartis clinical trials) were enriched for a PRS that exceeded the 95th percentile established among UK Biobank psoriasis cases by 3.06-fold and 2.32-fold respectively.
The psoriasis susceptibility PRS demonstrates utility and may be more effective than established epidemiological factors, as a stratification tool to identify those individuals that are at greatest risk of severe disease and may benefit most from early intervention.

Parkinson’s disease (PD) is a complex neurodegenerative disease that involves many interlinking pathways and genetic elements that remain to be fully understood and characterised. Non-coding genetic elements have long been overlooked, however recent advancements in the field have highlighted their importance with an area of interest being transposable elements. SINE-VNTR-Alu (SVA) elements are the youngest and smallest subset of retrotransposons that are only found within hominid species. SVAs have been shown to have strong regulatory impacts within our genome and can affect progression of neurodegenerative disease such as PD. Previous studies identified an SVA, polymorphic for its presence/absence, that was associated with changes in gene expression at the MAPT locus. This particular SVA is located within a long non-coding RNA (lncRNA) and is known as SVA_67. Here, we evaluated the SVA67-lncRNA effects on gene expression within the MAPT locus, a region associated with several neurodegenerative diseases in the SH-SY5Y cell line. The expression of SVA67-lncRNA in the SH-SY5Y cell line was associated with differential expression of several genes at the MAPT locus including MAPT, KANSL1, ARL17A/B, LRRC37A/2, and NSF. This study provides the first analysis of this SVA67-lncRNA and potential evidence for its involvement in complex diseases, such as PD.

Genome-wide association study of Parkinson's disease (PD) identified common variants associated with lysosomal mechanism, including TMEM175, SCARB2, and CTSB. We investigated the association between common and rare variants across populations using cohorts from the Global Parkinson's Genetics Program (GP2) (33,733 cases and 18,703 controls from ten ancestries). In the European cohort, we confirmed significant associations with PD risk for all known genetic risk variants across the three genes and TMEM175 p. Met393Thr as an independent genome-wide significant signal. Additionally, a novel independent signal, SCARB2 rs11547135, was detected. The burden analysis linked PD to SCARB2 in African American, Ashkenazi Jewish and East Asian cohorts. Single variants-based tests identified rare missense variants in SCARB2 in several populations. Our study reinforces the association of lysosomal genetic variants with PD risk, revealing genetic heterogeneity across populations.

Risk prediction models play a crucial role in advancing healthcare by enabling early detection and supporting personalized medicine. Nonetheless, polygenic risk scores (PRS) for Parkinson’s disease (PD) have not been extensively studied across diverse populations, contributing to health disparities. In this study, we constructed 105 PRS using individual-level data from seven ancestries and compared two different models. Model 1 was based on the cumulative effect of 90 known European PD risk variants, weighted by summary statistics from four independent ancestries (European, East Asian, Latino/Admixed American, and African/Admixed). Model 2 leveraged multi-ancestry summary statistics using a p-value thresholding approach to improve prediction across diverse populations. Our findings provide a comprehensive assessment of PRS performance across ancestries and highlight the limitations of a “one-size-fits-all” approach to genetic risk prediction. We observed variability in predictive performance between models, underscoring the need for larger sample sizes and ancestry-specific approaches to enhance accuracy. These results establish a foundation for future research aimed at improving generalizability in genetic risk prediction for PD.

Hereditary diseases, or genetic diseases, are generally considered to be caused only by single-nucleotide mutations. Typically, the assumption is that a single mutation is solely responsible for the clinical syndrome. This perspective is common, as geneticists search for one specific nucleotide mutation or variant that can account for the clinical phenotype. However, this approach has led to an oversimplified understanding of the genetic pathology or causes of these diseases, resulting in a diagnostic gap for many hereditary conditions. Consequently, we encounter patients with apparent hereditary issues yet lack a meaningful genetic explanation. This has led to reduced diagnostic yield and is one of the main challenges for genetic counsellors. Moreover, GWAS studies have shown limited success in fully elucidating the inheritance of most of the complex traits studied, typically reaching the explained heritability of not more than 30%. These limitations have given rise to the term “missing heritability”, which signifies our inability to clarify genetic inheritance using mainstream methods. This chapter presents a potential solution to the missing heritability problem. It demonstrates that much of the heritability remains unaccounted for due to its complexity, arising from transposable or repetitive elements in the human genome. The genetic basis for heritable conditions is intricate and necessitates a sophisticated approach to reveal the parts of missing heritability.

Carcinogenesis is based on acquiring the mutations that will trigger and maintain the malignant cellular growth. According to the widely accepted concept, these mutations are single-nucleotide variants, and environmental factors mainly generate them. In this chapter, we describe an alternative understanding based on the human genome’s internal capacity to generate structural variants within the “dark matter” of the human genome. The “dark matter” of the genome contains transposable elements (TEs) that can transpose and generate mutations in the genomes. These de novo variants affect the functioning of the genome and are the leading cause of the aberrant transcription that is the basis of many, if not most, of the complex diseases. In this chapter, I focus on the impact of TEs on cancer development.

PARK7 mRNA encodes DJ-1 protein, which functions as a protective agent against oxidative stress and cell damage within the brain cells. Mutations in the mRNA can lead to reduced production of DJ-1 and initiate brain diseases such as Parkinson’s disease. Transport of appropriate mRNA to damaged brain cells may provide a suitable treatment. Mesoporous silica nanoparticles (MSNPs), particularly pore-expanded and dye-labeled varieties, are regarded as potential carriers for large therapeutic agents such as mRNA. This study explored the influence of alterations in reaction conditions on the structural characteristics of MSNPs to produce nanoparticles with favorable characteristics for delivering large therapeutic agents to target sites. One-stage and two-stage procedures were compared for the introduction of 3-aminopropyltriethoxysilane (APTES) and APTES-dye adduct, in conjunction with two different surfactants, cetyltrimethylammonium bromide (CTAB) and cetyltrimethylammonium chloride (CTAC). Analysis of the MSNPs shows that the two-stage method using CTAB as a surfactant produced amine-functionalized, dye-labelled particles with smaller overall size and better uniformity than the one-stage approach. However, due to their small pore size (<10 nm), these particles were unable to encapsulate the PARK7 mRNA (926 nucleotides). The one-stage method via CTAC produced MSNPs with large (150 nm), broad pore distribution (10–20 nm), and high aggregation, limiting their suitability for brain-targeted gene delivery. In comparison, the two-stage method using CTAC yielded well-ordered MSNPs with an optimal size (80 nm) and pore diameters (15–20 nm), enabling effective encapsulation of the large PARK7 mRNA and offering strong potential for future brain gene therapy studies.

Age-related dysfunction in neuroendocrine signaling, which influences adipose tissue homeostasis, has been implicated in numerous diseases, including breast cancer. Caloric restriction has been shown to improve metabolic health and prolong lifespan, yet the molecular mechanisms underlying its long-term effects are not fully understood. In this study, we investigated the impact of long-term chronic (CCR) and intermittent caloric restriction (ICR) on the whole transcriptome of mammary fat pad tissue (MFP) in a breast cancer mouse model. Transgenic female Mouse Mammary Tumor Virus-Transforming Growth Factor-Alpha (MMTV-TGF-ɑ) C57BL/6 mice were randomized into ad libitum (AL), CCR, and ICR groups. Total RNA was isolated from the samples collected at weeks 10 (baseline), 49/50 (adult), and 81/82 (old), were then subjected to RNA sequencing. Differential gene expression analysis identified significant age-related transcriptomic shifts. Specifically, Malat1 expression levels, a long non-coding RNA associated with cancer progression, were elevated with aging, suggesting increased tumorigenic susceptibility in this model. Pathways linked to neuroendocrine signaling were downregulated with age, reflecting a potential decline in neuro-adipose cross-talk. Remarkably, ICR appeared to mitigate this age-related decline in neuroendocrine signaling by upregulating genes involved in neurotransmitter support and downregulating extracellular matrix organization and positive regulation of angiogenesis. In contrast, CCR did not effectively alter the whole transcriptome profile, particularly in long-term. Our findings reveal that ICR mitigates age-related transcriptional shifts in MFP tissue, providing a novel insight into dietary strategies for maintaining adipose tissue function with potential implications for cancer susceptibility.

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder presenting progressive weakness of the bulbar and extremity muscles, leading to a wide-ranging clinical phenotype. More than 30 genes have been associated to genetically inherited ALS yet, approximately 85%–90% of ALS cases are sporadic. Short tandem repeats expansions, have recently been found in clinically diagnosed ALS patients and are currently investigated as potential genetic biomarkers. In this paper we compare the investigation of pathological tandem repeat expansions on a group of ALS patients by comparing the standard short-read sequencing (SRS) technique with a long-read-sequencing (LRS) method which has recently become more accessible. Blood samples from 47 sporadic ALS cases were subjected to SRS by Illumina Whole Genome Sequencing. The genome-wide tandem repeat expansions were genotyped using GangSTR, while wANNOVAR was used for variant annotation. Uncertain cases were further explored using LRS. SRS identified pathological expansions in HTT , ATXN2 , and CACNA1A genes in one patient, which were not confirmed with LRS. The latter identified large tandem repeat expansions in the C9orf72 gene of one patient that were missed by SRS. Our findings suggest that LRS should be preferred to SRS for accurate identification of pathological tandem repeat expansions.