Skip directly to content

Coming dissertations at MedFak

  • Exploring the Role of the PDZ Domain in a Supramodule Author: Louise Laursen Link: Publication date: 2021-04-29 10:34

    The postsynaptic density (PSD) is a large, dense and membraneless compartment of proteins associated below the postsynaptic membrane bilayer, and which constantly undergoes morphological alteration in response to synaptic activity. Formation of PSD is associated with liquid-liquid phase separation of scaffold proteins in complex with other PSD proteins. PSD-95, one of the most abundant scaffold proteins, contains five domains: PDZ1, PDZ2, PDZ3, Src homology 3 (SH3), and guanylate kinase-like (GK) domain. The domains are functionally divided in two supramodules: PDZ1-PDZ2 and PDZ3-SH3-GK (PSG). Multi-domain proteins are characterized through their isolated domains in most studies and represented by “beads on a string” model, which means that the function of a single domain is independent of the context. In this thesis, the properties of PDZ3 and PSG are compared to elucidate how and when PSD-95 can been characterized by the simple “beads on a string” model. Kinetic characterization of CRIPT binding to PDZ3 showed a two-state mechanism, but a more complex mechanism involving two conformational states upon binding to PSG. The results were consistent with recent structural findings of conformational changes in PSD-95, altogether showing that conformational transitions in supertertiary structures can shape the ligand-binding energy landscape and modulate protein-protein interactions. Next the allosteric networks in a PDZ:ligand complex were experimentally mapped, both in isolation and in the context of a supramodular structure. Data showed that allosteric networks in a PDZ3 domain has high dependency on the supertertiary structure. Furthermore, equilibrium and kinetic folding experiments were applied to demonstrate that the PDZ3 domain folds faster and independently from the SH3-GK tandem, which folds as one cooperative unit. However, concurrent folding of the PDZ3 domain slows down folding of SH3-GK by non-native interactions, resulting in an off-pathway folding intermediate. Finally, the interactome of PSG in PSD was mapped. PDZ3 and PSG show high specificity for peptides with type I PBM. Interestingly, two proteins called SynGap and AGRB1 only bind with high affinity to PSG and forms concentration dependent liquid droplets. The results show how context in terms of supertertiary structure alter affinity and function, and suggest a model for how PSD anchor to the postsynaptic membrane. Altogether, the findings in the thesis show that binding energy landscape, interactome, allosteric network, folding mechanism and phase separation are dependent on the context, which suggest that we need to be careful in interpretation of data obtained from isolated domains in multi-domain proteins.

  • The role of HMG-coenzyme A reductase (HMGCR) and statin medication in the Central Nervous System : Cognitive Functions, Metabolism, Feeding and Sleep Behaviour Author: Ahmed Alsehli Link: Publication date: 2021-04-29 10:25

    Millions of people are currently on statin medications (HMGCR inhibitors) to prevent cardiovascular diseases. Despite considerable central nervous system expression, little is known about HMGCR function in the brain. In Paper I, we used Drosophila and rodent models and found that inhibiting Hmgcr expression in the insulin-producing cells of the Drosophila hypothalamus equivalent, known as the pars intercerebralis (PI), throughout development, significantly reduces the expression of Insulin–like peptides 2 and 3 (ILP2 and ILP3), severely decreasing insulin signalling. This reduction causes decreased body size, hyperglycemia, increased lipid storage, and hyperphagia. We also discovered that Farnesyl pyrophosphate synthase (Fpps), an enzyme downstream of Hmgcr in the mevalonate pathway, is required for ILP2 expression in the PI. In rodents, acute inhibition of hypothalamic Hmgcr stimulates food intake as well. Furthermore, in rats, we found two regions within the hypothalamus that had significantly increased neural activity, the paraventricular nucleus and arcuate nucleus, which are known to regulate food intake. In Paper II, we explored the effects of statins on cognition and performed an observational study on a population-based sample from the UK Biobank. Cognitive performance in terms of reaction time, working memory and fluid intelligence was analysed at baseline and two follow-ups. Subjects were classified depending on age (up to 65 and over 65 years). The effect of statin use differed between the two age groups, with a beneficial effect on reaction time in older persons and fluid intelligence in both age groups, and a negative effect on working memory in younger subjects. In Paper III, we examined association of single nucleotide polymorphisms within the HMGCR gene, rs17238484 and rs12916, with self-reported insomnia symptoms. We found that statin users are associated with a higher risk for self-reported insomnia. The HMGCR genetic variants were also associated with self-reported insomnia, but in different manner. Carriers the rs12916-T risk allele had a protective effect from insomnia symptoms. No associations were found for either statin takers or carriers of these HGCMR risk alleles and late evening chronotype. The increased risk of insomnia noted with statins is partially explained by a mechanism that might be independent of HMGCR inhibition. In Paper IV, we discovered a novel role for Hmgcr in sleep regulation in Drosophila, where lacking of pan-neuronal Hmgcr expression causes sleep-promoting effects. We also found that loss of Hmgcr expression specifically in the PI insulin-producing cells, recapitulates the effect of pan-neuronal Hmgcr inhibition. Conversely, inhibiting Hmgcr in only six PI DH44 expressing neurons has the opposite effect on sleep, increasing sleep latency and decreasing sleep duration. This bi-functional property of Hmgcr in the fly brain underlies its importance in sleep regulation. Furthermore, loss of Hmgcr showed no effect on circadian rhythm, suggesting that Hmgcr regulates sleep by pathways distinct from the circadian clock.

  • Development of a robust and rapid microfluidics-based antibiotic susceptibility test : From prototype to clinical implementation Author: Christer Malmberg Link: Publication date: 2021-04-29 10:15

    Even though bacteria normally are rapidly cleared from the blood by the immune system, a blood stream infection may arise, in turn possibly leading to sepsis and septic shock. Sepsis is a life-threatening condition, where the inflammatory response to infection targets tissues and organs. Mortality in sepsis is very high, at 25-30%, but can be reduced by early and appropriate antibiotic treatment. The time until appropriate antibiotic therapy is started impacts mortality and morbidity to a large extent, from 1-7% increase in mortality per hour of delayed treatment in sepsis and septic shock. However, with increasing antimicrobial resistance globally, the likelihood of successful treatment is continuously reduced. Antimicrobial resistance emphasizes the need for diagnostics to guide therapy, but traditional antimicrobial susceptibility testing is often inadequate in time-critical disease such as sepsis. Subsequently, there is an urgent need for new, more rapid and accurate antibiotic susceptibility tests. One challenge to improving the speed of phenotypic susceptibility testing is the need to rapidly, accurately and non-invasively capture data from a very large collection of cells. Furthermore, development of new diagnostic methods for patients in critical condition, such as sepsis, demands high reliability and accuracy of the method – a false test result can lead to suboptimal therapy, and in turn increased morbidity and ultimately death. This thesis presents a series of prototype rapid antibiotic susceptibility testing (AST) systems using a low-magnification, wide-field optical setup with high cell-mass resolution for simultaneously quantifying bacterial growth rates of tens of thousands of bacterial cell clusters growing in antibiotic gradients generated using microfluidics. Performance data from spiked reference blood samples show that the analytical performance of the system is good, with mean essential agreement of 83.2% against reference methods. The average time to result for the reference dataset was 180 min. For clinical samples, the method was demonstrated to have high categorical agreement with disc diffusion (94.9%), as tested at Uppsala University Hospital. Furthermore, the time from patient sampling until test result availability (turnaround-time) was reduced by 40%. The thesis concludes with a discussion of the recent experimental work and a summary concerning the potential applications of this technology. In summary, rapid diagnostics capable of shorter turnaround times could enable earlier de-escalation of broad-spectrum empirical therapy, as well as enable rapid adjustments to treatments when coverage is lacking. This is likely to reduce mortality as well as healthcare costs associated with increasing resistance.