Melanie Modder

Globally, noncommunicable diseases such as obesity remain on the rise despite rapid advancements in biomedical sciences and healthcare. At the moment at least 5% of worldwide mortality can be attributed to high body fat mass due to associated comorbidities like metabolic dysfunction-associated steatotic liver disease (MASLD), type 2 diabetes (T2D), and atherosclerotic cardiovascular disease (asCVD), and which collectively strain healthcare systems.

Chapter 1 serves as a general introduction describing how disbalance between energy intake and expenditure can lead to fat mass gain and eventually to the development of obesity, MASLD, T2D, and asCVD. A key approach to combating obesity is through lifestyle improvements, such as adopting healthier eating patterns and increasing physical activity. Unfortunately, such measures are not always effective and can be difficult to maintain in the long term. One significant challenge is the lasting changes in the body induced by obesity, including alterations in the hormonal systems that regulate hunger and satiety. However, recent advancements in pharmacological treatments, including the development of long-acting glucagon-like peptide-1 (GLP-1) receptor agonists and even dual or triple agonists targeting additional receptors involved in food intake regulation, have made it possible to address these issues, making reversal of obesity a more realistic prospect for a growing number of individuals. Furthermore, new therapeutic compounds are reaching the market that effectively manage the (cardio)metabolic diseases associated with obesity. Nonetheless, both existing and novel therapeutic strategies can still be further developed to reach greater therapeutic potential, specifically by utilizing recent technical and pharmaceutical advances that enable more precise and effective targeting of proteins and pathways involved in lipid metabolism and hormone signalling. Chapter 1 of this thesis describes three promising strategies in this context: (1) targeting the right tissue, (2) refinement of downstream signalling pathways, and (3) implementing multi-targeted approaches. In this thesis, I have explored new therapeutic strategies that combat obesity related diseases by making use of these approaches.

One way to target multiple proteins is via microRNAs (miRNAs), which are a sub-class of non-coding single-stranded RNAs that inhibit translation of mRNAs by binding to complementary sequences. Usually, a 6-8 nucleotide miRNA seed sequence is sufficient to induce inhibition, even if the complementarity is not perfect. As a result, miRNAs can target a multitude of different genes, often inducing shifts in entire metabolic pathways. In this thesis, we investigated whether miR-33a, which plays an important role lipid metabolism, might be a promising multi-targeting approach to lower hyperlipidaemia and asCVD. In Chapter 2, we showed that both strands of miR-33a (i.e. miR-33a-3p and 5p) indirectly increase low-density lipoprotein receptor (LDLR) protein abundance in hepatocytes at post-transcriptional level. miR-33a-3p and 5p blunt LDLR degradation in hepatocytes by inhibiting expression of proprotein convertase subtilisin/kexin type 9 (PCSK9) and inducible degrader of low-density lipoprotein receptor (IDOL) through non canonical binding. In line with these effects, treatment with a liver-targeted miR-33a-3p mimic in diet-induced obese (DIO) mice increased hepatic LDLR levels and lowered circulating cholesterol. Furthermore, our investigation revealed that miR-33a-3p, but not miR-33a-5p, exerts a direct inhibitory effect on angiopoietin-like 3 (ANGPTL3) expression. Thereby, antisense-mediated depletion of miR-33a-3p in heterozygous Ldlr knockout (Ldlr+/-) mice resulted in increased serum ANGPTL3 and PCSK9 protein levels and concomitant elevated cholesterol levels. In APOE*3-Leiden.CETP mice, a model for human-like lipoprotein metabolism, treatment with the liver-targeted miR-33a-3p mimic attenuated postprandial triglyceride (TG) and cholesterol levels due to increased TG derived fatty acid uptake by white adipose tissue and subsequent hepatic uptake of lipoprotein remnants, accompanied by reduced plasma ANGPTL3 levels. Long-term miR 33a-3p mimic treatment enhanced stability of atherosclerotic plaques in APOE*3-Leiden.CETP mice, and miR-33a-3p overexpressing homozygous Ldlr-/- mice displayed strongly reduced atherosclerotic lesion formation. This study demonstrates how miR-33a functions as critical component of SREBP-2-regulated maintenance of cholesterol homeostasis, by preventing degradation of newly synthesized LDLR. Additionally, we underscore miR-33a-3p mimic treatment as potential alternative therapeutic strategy for managing LDL-cholesterol in hypercholesterolemia.

To further develop a miR-33a-3p mimic for use in the clinic, it is crucial to know which tissues are responsible for the beneficial effects, so that these tissues can be specifically targeted, as mentioned in Chapter 1. Interestingly, while miR-33a-3p overexpression in Ldlr-/- attenuates atherosclerosis development largely by reducing plasma lipids, our preliminary data also hinted towards a role in immune cells. In Chapter 3, we therefore set out to explore whether miR-33a-3p in immune cells contributes to the anti-atherogenic effects observed in Ldlr-/- mice. To this end, we studied how hematopoietic miR-33a-3p overexpression affects hyperlipidaemia and atherosclerotic plaque development, by transplanting Ldlr-/- mice with bone marrow (BM) from either miR-33a-3p transgenic (Tg) mice or wild-type (WT) mice. Hematopoietic miR-33a-3p overexpression, somewhat unexpectedly, resulted in larger atherosclerotic lesions that were more frequently classified as severe. Plasma lipid levels were not affected in mice transplanted with miR-33a-3p Tg BM, nor were cellular cholesterol efflux and polarization of BM-derived macrophages. Rather, hematopoietic miR-33a-3p overexpression induced a significant shift in circulating leukocytes, with a reduction in the fraction of B-cells and increased fractions of both pro and anti-inflammatory monocytes. In conclusion, hematopoietic miR-33a-3p overexpression in Ldlr-/- mice aggravates atherosclerosis development, possibly as a result of increased circulating monocytes. We thus anticipate that miR-33a-3p overexpression in non-hematopoietic cells is dominant in the anti-atherogenic effect of whole body miR-33a 3p overexpression in Ldlr-/- mice.

The disparity between the effects of systemic miR-33a-3p overexpression in Chapter 2 versus hematopoietic overexpression in Chapter 3 highlights the importance of targeting the right tissue for therapeutic purposes. In line with that consideration, in Chapter 4, we investigated the therapeutic potential of silencing Angptl4 specifically in the liver to attenuate hyperlipidaemia and atherosclerosis development in comparison to liver-specific Angptl3 silencing. While pharmacological inhibition of ANGPTL3 is used in the clinic as lipid-lowering strategy, ANGPTL4 inhibition was previously not considered a viable therapeutic strategy due to adverse effects observed after global inhibition. In this study, APOE*3-Leiden.CETP mice, a well-established humanized model for lipoprotein metabolism, were treated with liver-targeted antisense oligonucleotides against Angptl3, Angptl4, both, or a scrambled oligonucleotide as control. Liver-targeted Angptl4 silencing reduced plasma TGs and total cholesterol, explained by higher very low-density lipoprotein (VLDL)-derived fatty acid uptake by brown adipose tissue (BAT) and lower VLDL production by the liver. Accordingly, Angptl4 silencing reduced atherosclerotic lesion size and improved lesion stability. Hepatic Angptl3 silencing similarly attenuated hyperlipidaemia and atherosclerosis development. While Angptl3 and Angptl4 silencing lowered plasma TGs specifically in the refed and fasted state, respectively, combined Angptl3/4 silencing lowered plasma TGs independent of the nutritional state. In cynomolgus monkeys, anti-ANGPTL4 ASO treatment was well tolerated without adverse effects.

Together, these data support liver-targeted Angptl4 silencing as a promising way to combat hyperlipidaemia and reduce atherosclerosis development, warranting further clinical development of liver-targeted anti-ANGPTL4 ASOs. Excitingly, following the favourable safety profile observed in a phase-I-clinical trial, anti-ANGPTL4 ASO Lipisense® is currently being evaluated in phase-II clinical testing.

Thus, our work in Chapter 4, along with others, has demonstrated that liver-targeted therapy represents a promising strategy to attenuate dyslipidaemia and atherosclerosis development while circumventing potential adverse effects associated with systemic treatment. Beyond the liver, the central nervous system plays a crucial role in maintaining metabolic health, and the crosstalk between peripheral metabolic organs and the central nervous system is essential for body weight control. As such, the brain represents an interesting target to combat obesity. However, targeting the brain with pharmacological agents can be challenging due to the existence of a protective blood-brain barrier, which prevents most drugs from entering the brain. Only at specific sites, known as circumventricular organs (CVOs), does the blood-brain barrier make place for a more permeable endothelium that facilitates exchange of substances between the peripheral bloodstream and cerebrospinal fluid. At the median eminence, one of the CVOs located at the base of the hypothalamus, specialized cells called β-tanycytes surround the portal capillaries and function as gatekeepers to regulate transfer of molecules from the circulation into the cerebrospinal fluid and vice versa. In Chapter 5, we investigated the role of the neuroplasticity gene doublecortin-like (DCL), highly expressed by β-tanycytes, in body weight control and whole-body energy metabolism. We demonstrated that DCL knockdown through a doxycycline-inducible shRNA expression system prevents body weight gain by reducing adiposity in mice. DCL-knockdown slightly increased whole-body energy expenditure possibly as a result of elevated circulating thyroid hormones. In WAT, triglyceride-derived fatty acid uptake was increased while the average adipocyte cell size was reduced. At histological level we observed clear signs of browning, and thus increased thermogenesis in WAT. We found no indications for stimulated thermogenesis in BAT. Altogether, we demonstrate an important, though subtle, role of tanycytic DCL in body weight control through regulation of energy expenditure, and specifically WAT browning. Further elucidation of mechanisms underlying the role of DCL in regulating brain peripheral crosstalk might identify new treatment targets for obesity.

Another approach to increase therapeutic efficacy explored in this thesis is via refinement of downstream signalling pathways, which we investigated in the context of biased glucagon-like peptide-1 (GLP-1) receptor agonism. While GLP-1 agonists are currently the most successful medications to treat obesity and provide significant therapeutic benefit for many individuals, several challenges remain in optimizing their use. A considerable proportion of patients continue to experience adverse effects associated with GLP-1 receptor agonists or have reduced long-term efficacy. These challenges form an obstacle for treatment adherence, thereby lowering real-world therapeutic efficacy. GLP-1 receptor agonists might be further optimized by creating compounds with favourable signalling bias, optimizing the intracellular routes activated after receptor binding. In Chapter 6, we compared the effects of two oppositely biased GLP-1 receptor agonists on lipid metabolism in hyperlipidaemic mice. The two agonists used, [D3,G40,K41.C16 diacid]exendin-4 (acyl ExD3) and [F1,G40,K41.C16 diacid]exendin-4 (acyl-ExF1) preferentially promote β-arrestin recruitment or G protein-induced signalling, respectively. The latter is more favourable in glycaemic control and induces a steeper reduction in body weight in DIO mice. To compare the effects of the two agonists on lipid metabolism, APOE*3-Leiden.CETP mice were treated with saline, acyl-ExD3, or acyl-ExF1, either via intraperitoneal injections or continuous infusion directly into the ventricular system of the brain, circumventing the need for transfer across the blood-brain barrier. Upon peripheral treatment, body weight gain was prevented and plasma glucose levels were reduced by G protein-biased acyl-ExF1, while circulating lipids were not affected by either acyl-ExF1 or β-arrestin-biased acyl ExD3. In contrast, central administration of either agonist strongly reduced plasma triglyceride and cholesterol levels, accompanied by increased VLDL-derived fatty acid uptake by WAT and BAT. To summarize, the oppositely biased GLP-1 receptor agonists acyl-ExD3 and acyl-ExF1 do not differentially affect lipid metabolism in APOE*3-Leiden.CETP mice, while effects on glucose homeostasis and prevention of body weight gain are more pronounced upon peripheral acyl-ExF1 treatment.

Finally, the results of the experiments described in this thesis were placed in context of current scientific literature in Chapter 7. The potency and challenges of the novel therapeutic approaches were discussed. With the research as conducted in this thesis, I hope to bring new insights into the potential ways by which tissue-targeting, multi-targeted approaches and refinement of intracellular signalling pathways can contribute to increased precision and efficacy of therapies combating obesity-related diseases. I am excited to find out what the future will hold with regards to optimizing treatment strategies for obesity.