Department of Biological Sciences (BIO)

Molecular Biology Master theses submitted in 2020

Main content

Lars Erik Eriksen Bohlin

Lars Erik Eriksen Bohlin, Sushma Grellscheid, Rune Male

Connections between MVP, p21 and Stress Granules in Senescent wild type and G3BP mutant U2OS Cells

Senescence can be studied as aging on a multicellular organism or on the cellular senescent level. Cells subjected to chemotherapeutic substances like etoposide experience severe DNAdamage that can induce cells into apoptosis or the alternative stress-induced non-dividing state of senescence.

Vaults with the Major Vault Protein (MVP) as primary constituent has been linked to apoptotic resistance and upregulation in senescence across diverse cell types. Here we attempted to test if MVP and minor vault proteins (TEP1 and VPARP) were upregulated in senescence for the U2OS cell type. P21 is proven to be reliable senescent marker as it guides cells towards the senescent phenotype both dependently and independent of p53. The thesis also investigates the role of the Stress Granule-nucleating proteins (G3BP) and how their presence and absence affect the proteomics of U2OS cells and the protein levels of MVP, TEP1, VPARP and p21. The project involved culturing Wild Type U2OS cells, and the knockdown line U2OS ΔΔG3BP1/2, induce senescence and verify with Senescence Associated
b-Galactosidase assay (SABG). Further to probe for MVP, p21 and G3BP with Western Blotting and qPCR and visualize G3BP1 and MVP in indirect immunofluorescence.
The SABG assay and p21 upregulation was positive across cultures and cells showed the enlarged, protruding and irregular morphology indicative of the senescent cellular phenotype. In the studied cell lines, MVP, TEP1, VPARP and p21 were upregulated in Senescent U2OS cells in the proteomic dataset. The total of the evidence including cell culturing of WT and U2OS ΔΔG3BP1/2 points towards the fact that MVP and therefore probably vaults are upregulated and possibly involved in the apoptotic resistance and steering towards senescence previously found across diverse cell types. TEP1 and VPARP were also found to be upregulated but have free cytosolic roles, outside of vaults, and are not considered as indicative of vault levels. The cell lines with one or both G3BP genes knocked out showed in proteomics fever upregulated proteins of strong signals. This might indicate that when stress granules, that protect mRNA are not allowed to form, RNAs are damaged by cellular stresses. Simultaneously MVP had its strongest upregulation in U2OS ΔΔG3BP1/2. Stress Granules and MVP showed up in immunofluorescent tagging in U2OS WT.

Ivan Abbedissen

Ivan Abbedissen, Karianne Fjeld, Anny Gravdal og Anders Molven

Characterization of the CEL-MODY mouse – A new disease model for chronic pancreatitis

Carboxyl ester lipase (CEL) is a digestive enzyme produced by the acinar cells of the pancreas. Two single-base deletion variants of the CEL gene cause a frameshift in the variable number of tandem repeats region located in the last exon. These CEL variants lead to a syndrome of pancreatic exocrine dysfunction and diabetes, known as Maturity-Onset Diabetes of the Young (CEL-MODY).
To get new insight into the disease mechanisms of CEL-MODY, our research group has developed a novel CEL-MODY knock-in mouse model. The overall aim of this project was to characterize this model. More specifically, we wanted to study the effect of the CEL-MODY protein at an organ level, with regard to pancreatic exocrine and endocrine dysfunction.
The CEL-MODY mice showed signs of pathological changes in the pancreas at 3 months of age, while severe, irreversible changes were observed at 6 months of age for both males and females. We observed exocrine atrophy, fibrosis and fatty infiltration, which are well known features of chronic pancreatitis (CP). Immunostaining indicated loss of Cel expression in atrophic exocrine tissue. We also detected signs of Cel protein aggregation, suggesting that CEL-MODY cause disease through the misfolding-dependent pathway of CP. No signs of diabetes were observed at 6 months of age, indicating that the islets of Langerhans remain functional in our mouse model, despite severe exocrine damage.
In summary, we found that the CEL-MODY mice developed spontaneous CP, making this the first mutant lipase mouse model for CP. However, the mice did not present any signs of diabetes development at 6 months of age.

Odd-Erik Bakkebø

Odde-Erik Bakkebø, Fabian Rentzsch og James Gahan

Transcription factor NvRfx4 as a potential regulator of neurogenesis in Nematostella vectensis

Nervous systems have a plethora of neurons for different purposes which develop from neural stem and progenitor cells through asymmetric and symmetric divisions. The understanding of neural progenitors and nervous system development is mainly confined to Bilateria. Here, I use a cnidarian model organism, Nematostella vectensis, to gain insight into neurogenesis in a non-bilaterian animal. Using colorimetric and double fluorescent in situ hybridization I found that the evolutionarily conserved transcription factor NvRfx4 is expressed in some NvSoxB(2)-expressing neural progenitor cells and otherwise has a scattered ectodermal and dense pharyngeal expression. I further show that all NvFoxQ2d-expressing neural cells co-express NvRfx4 and that NvPaxC is not co-expressed with NvRfx4 but is often found expressed in a neighbouring cell. In addition, NvRfx4 is expressed in uncharacterized endomesodermal and pharyngeal cells. The expression data support a scenario in which NvRfx4 has a function in the development of NvFoxQ2d-expressing sensory cells and in other, currently uncharacterized cells. As a tool to understand the function of NvRfx4, genome editing via CRISPR/Cas9 was used to generate an NvRfx4 mutant. From an evolutionary perspective, the data support the hypothesis that a role for Rfx4 genes in neural development arose early in animal evolution.

Ingvild Emma Losnegard Koløen

Ingvild Emma Losnegard Koløen, Janne Molnes, Bente Berg Johanson og Pål Njølstad

Personalized medicine in diabetes: Unravelling the disease causality of HNF1B gene coding variants in
two large Norwegian diabetes registries

Maturity-onset diabetes of the young (MODY) is a subclass of monogenic diabetes, caused by
heterozygous mutations in one of the 14 identified genes. The disease is characterized by an
autosomal dominant inheritance pattern, progressive β-cell dysfunction and early onset
diabetes, usually before 25 years of age. Pathogenic variants identified in the hepatocyte nuclear
factor-1 beta (HNF1B) gene causes various phenotypes, including diabetes, abnormalities in
kidneys, liver, pancreas and genital tract malformations. In the present study, nine HNF1B
variants found in the Norwegian Childhood Diabetes Registry (NCDR) and the Norwegian
MODY Registry (NMR) were subjected to functional investigations in vitro.
To characterize HNF1B variant pathogenicity, the transactivation potential, subcellular
localization, protein expression, mRNA expression and DNA-binding ability were investigated
in transiently transfected MIN6 cells. The transactivation assay was additionally performed in
HEK293 and INS-1 cells. The nonsense variants R177* and Q182* revealed significantly
reduced transactivation ability, abolished DNA-binding, impaired nuclear targeting and
severely reduced protein expression levels. For the missense variants G287V, N289K and
R295C, both transactivation and DNA-binding abilities were found to be compromised,
especially pronounced for R295C in transactivation. The remaining variants P60R, Q131H,
N327K and S362F demonstrated varying degrees of dysfunction in certain assays but
performed similar to wild type (WT) HNF-1B in the transactivation assay.
Taken together, our data demonstrates that the truncated R177* and Q182* variants and the
missense variant R295C, are indeed pathogenic loss-of-function variants, consistent with
clinical data on variant carriers. The functional data gathered for the N289K variant, in addition
to newly retrieved clinical information, leads to a re-classification from variant of unknown
significance (VUS) to pathogenic. The G287V variant is of high interest, but further
investigation is needed. The variants P60R and N327K are supportive towards a likely benign
classification, whereas the Q131H and S362F variants remain VUSs. We believe that for some
of the variants in the current study, detailed family history and clinical phenotypes, including
examination of renal abnormalities, will be necessary in order to make a final conclusion. A
precise diagnosis of the variant carriers is important for optimized treatment, proper prognosis
as well as follow-up of family-members.

Eunice Kabanyana Mchaina

Eunice Kabanyana Mchaina, Ingvild Aukrust, Per M Knappskog og Bjørn Ivar Haukanes

The functional effects of two pathogenic HUWE1 variants reported in patients with intellectual disabilities

HUWE1 (HECT, UBA, and WWE domain-containing E3 ubiquitin-protein ligase 1) is an E3
ligase with a catalytic HECT domain in the C-terminus and is a participant in the ubiquitination
cascade. Covalent ubiquitin linkage targets substrates to various cellular pathways, like
degradation by the proteasome. The substrates structure and functions can also be modified by
ubiquitination to enable them to participate in different cellular mechanisms. HUWE1 is located
on Xp11.22, and pathogenetic variants in this gene are associated with X-linked intellectual
disability. HUWE1 is a large protein of 486 kDa with six functional domains. Most of the
pathogenic variants identified so far are located in the HECT domain. The other domains are
associated with, e.g., protein-protein interactions and non-covalently binding of ubiquitin. The
N-terminal DUF908/DUF913 domains have unknown structures and functions.
This study has focused on the functional properties of three HUWE1 (NM_031407.6) variants;
c.329G>A p.(Arg110Gln) (located in DUF908), c.12225C>G p.(Asn4075Lys) and
c.13022G>C p.(Cys4341Ser) (both located in the HECT domain), of which the first two have
been identified in patients with intellectual disabilities. This study has shown that Arg110Gln
had lower protein expression in transfected HEK293 cells compared to wild type (WT) DUF
(although not significant). In HEK293 cells the DUF domains (both WT and Arg110Gln) were
shown to be ubiquitinated and also stabilized by ubiquitin. In transfected HeLa cells,
Arg110Gln was localized both to the cytoplasm and nucleus in contrast to WT DUF that was
only found in the cytoplasm. In contrast to the HECT domain, there was no expression of the
DUF domains in E. coli cells, even though the mRNA encoding the DUF domains were present
in the cells and different conditions like varying IPTG and ethanol concentrations were tested.
Functional studies of the variant Asn4075Lys showed that it has reduced self-ubiquitination
activity compared to WT HECT, and the variant was also less degraded by the proteasome.
Cys4341Ser showed significantly lower protein expression and self-ubiquitination activity than
WT HECT, although it was degraded by the proteasome to some extent (although less than
WT). These findings show that studying the functional properties of HUWE1 variants can
provide a better understanding of their effects on function and their implications in patients
with intellectual disabilities, and indicates that variants in the DUF908/DUF913 domains versus
the HECT domain may have different functional properties. This could explain the different
phenotypes found in the patients carrying these variants, however further studies are needed to

Eirik Nordahl-Pedersen

Eirik Nordahl-Pedersen, Johan Fernø og Martha E Haugstøl

Fat-depot specific differences in adipose tissue inflammation in the setting of obesity and insulin resistance

The prevalence of obesity has increased substantially in the recent decades, along with associated co-morbidities such as insulin resistance, type 2 diabetes mellitus (T2DM) and cardiovascular diseases (CVDs). Moreover, obesity is associated with a state of chronic low-grade inflammation which is thought to be one of the main driving forces for the development of insulin resistance and T2DM. However, the causal relationship is still controversial. The inflammation in adipose tissue has been characterized by the secretion of pro-inflammatory cytokines and infiltration of pro-inflammatory macrophages. In particular, an increased tendency of accumulating fat in the visceral adipose tissue have been associated with adipose tissue inflammation and an increased risk of associated co-morbidities.
In this project, we aimed to quantify the level of macrophage infiltration into subcutaneous (SAT) and visceral adipose tissue (VAT) of severely obese subjects undergoing bariatric surgery. Furthermore, we phenotypically characterized pro-inflammatory M1-like macrophages (CD11c+CD206+) and anti-inflammatory M2-like macrophages (CD11c-CD206+) by the expression of known pro-inflammatory receptors and investigated whether the level of pro- and anti-inflammatory macrophages associated with insulin resistance and other clinical parameters relevant for metabolic syndrome. We also analyzed the expression level of pro- and anti-inflammatory genes and performed functional analysis on isolated mature adipocytes from SAT and VAT, focusing on differences between the two depots.
We found that macrophages are more abundant in SAT than in VAT. Moreover, insulin resistance (HOMA-IR) was associated with a higher ratio of the anti-inflammatory M2/M1-like macrophages and a higher expression of the anti-inflammatory receptors CD163 and TREM2 in SAT. We also found that systemic inflammation was associated with a higher ratio of pro-inflammatory M1/M2-like macrophages and a pro-inflammatory gene expression pattern in VAT. Moreover, the chemotactic receptor, CCR2, was found to be highly expressed by M1-like macrophages and almost absent from M2-like macrophages. Further, we found that lipolysis was significantly higher in VAT compared to SAT of obese subjects, and the lipolytic response correlated negatively with insulin sensitivity in VAT.

Mette Ovesen

Mette Ovesen, Øyvind Halskau, Aurora Martinez, Jarl Underhaug, Espen Bariås og Martin Jakubec

The Attenuating Effects of 14-3-3h in Parkinson’s Disease-Related a-Synuclein Aggregation

Characterized by the progressive loss of dopaminergic neurons, Parkinson’s Disease (PD) is one of the most common neurodegenerative diseases (NDs). A pathological hallmark of PD is the presence of neuronal inclusions termed Lewy Bodies (LBs). The main component of LBs has been identified as a-Synuclein (aS): a small synaptic protein with an intrinsically disordered structure in its native state. In parkinsonian brains, aS abnormally aggregates into fibrillar b-sheet rich structures that accumulate in LBs. Consequently, the misfolding and

aggregation of aS has been established as central in PD pathology.

In recent years, several studies in neuroprotection have focused on 14-3-3 proteins – a family of adaptor proteins with chaperone activities, consisting of seven human isoforms. With regards to PD, several 14-3-3 isoforms have been detected in LBs and shown to interact with PD related proteins like aS and Parkin. In particular, the 14-3-3h isoform has been found to coimmunoprecipitate with aS in parkinsonian brains. Previous aggregation studies have also demonstrated that this isoform interferes with aS aggregation by rerouting the fibrillation process. Moreover, studies in aging mice overexpressing aS demonstrated an upregulation of 14-3-3h, altogether indicating a role of 14-3-3h in aS aggregation and PD in general.

In this study, we investigated how 14-3-3h in different aS:14-3-3 ratios affect the kinetics of aS aggregation in vitro by Thioflavin T (ThT) monitored aggregation assays. Our findings show 14-3-3h to decrease aS fibril growth rate by more than 50% in ratios of 10:1 and 2:1. The fibrillation lag time, and thus onset, was not significantly affected by 14-3-3h. However, this could be attributed to an  insufficient number of replicates, and we concluded that 14-3-3h likely

attenuates aS fibrillation onset as well. We further sought to locate the 14-3-3h residues involved in aS binding by nuclear magnetic resonance (NMR) spectroscopy and to characterize the binding by surface plasmon resonance (SPR) experiments. Although time restrictions prevented binding site analysis, we found that 2H13C15N-labeling and TROSY-HSQC experiments considerably improves the 14-3-3h fingerprint spectrum, enabling future binding analysis. SPR experiments indicated no binding of 14-3-3h to monomeric aS, suggesting that the binding occurs at oligomeric stages of aS aggregation. In sum, we propose that 14-3-3h interacts with aS oligomers to attenuate aS fibril growth, and possibly fibrillation onset.

Malene Skuseth Slinning

Malene Skuseth Slinning og Aurelia Lewis

PARP-1: A novel nuclear polyphosphoinositide effector protein

Polyphosphoinositides (PPIns) are a family of seven signalling lipids that are important for
many cellular processes. While their functional roles in the cytoplasm have been extensively
studied, their roles in the nucleus are still poorly understood. To date, several members of the
PI3K pathway have been identified within the nucleus, where they have shown to be involved
in several nuclear functions such as DNA replication, DNA double strand break repair, and
ribosome biogenesis. Previously, our group identified the class I PI3K p110b and its lipid
product, PtdIns(3,4,5)P3 (PIP3) in the nucleoplasm and the nucleolus. To better understand the
nuclear functions of PIP3, a nuclear PIP3 interactome was mapped using nuclear extracts from
HeLa cells. Interestingly, many of the identified PIP3 binding proteins were annotated to the
nucleolus, a subnuclear structure which is known as the site of ribosome biogenesis.
Poly(ADP-ribose) polymerase-1 (PARP-1), an abundant nuclear protein involved in DNA
repair and enriched in the nucleolus, was identified as a potential PIP3 binding protein. This
was further confirmed by in vitro binding of PARP-1 to PPIns, including PIP3. Moreover,
PARP-1 was shown to co-localize with PIP3 in the nucleolus in HeLa cells. In the present study,
we aimed to determine the specific PIP3 interaction sites of PARP-1. We showed that PARP-1
binds to PPIns including PIP3 via two polybasic regions (PBRs) and through binding of one
reverse K/R motif, as deletion of either of the PBRs or mutations in the K/R motif reduced or
completely abolished all binding to PPIns. PARP-1 has been shown to be localized to the
nucleolus, however no nucleolar localization signal has been determined so far. Using the
nucleolar sequence detector (NoD) algorithm, we identified that one of the PPIn- binding sites
could also act as a potential NoLS. Finally, we wanted to investigate whether PIP3 is important
for the regulation of PARP-1 activity upon H2O2 induced DNA damage using MEF cells
harbouring WT or kinase dead version of class I PI3K p110b. However, no differences were
observed when comparing the PAR-intensities in the two cell lines. Taken together, these
results further verify that PARP-1 is a novel PPIn effector protein, however, additional studies
must be performed to map their functional role.

Marie Constance Solheim

Marie Constance Solheim og Kari Fladmark


The algal produced b-N-methylamino-L-alanine (BMAA) is a non-protein amino acid
neurotoxin. High abundance of the toxin producing algae is linked to increased cases of
neurodegenerative diseases and BMAA is therefore considered a health risk. The high incidence
of the Amyotrophic Lateral Sclerosis (ALS)-Parkinson-dementia complex in Guam led to the
discovery of its neurotoxic effect as it was found to be present in brain tissue from the patients.
How BMAA is contributing to neuropathology at the molecular level remains unclear.
Sometimes studies suggest that BMAA acts through overstimulating glutamate receptors.
Others that BMAA is mis-incorporated into proteins and thereby affecting the protein folding.
BMAA exposure is thought to induce endoplasmic reticulum stress (ER), increase oxidative
stress and disturb mitochondrial dysfunction. L-serine has shown promising neuroprotective
effects towards BMAA neurotoxicity and has been suggested as a treatment against
neurodegeneration. At present, L-serine has completed phase I clinical safety trial and has been
approved for a phase II trial for ALS patients. Here, we wanted further to investigate BMAA
neurotoxicity and the effect of L-serine and to compare its impact with known cellular
disturbance related to ALS pathologies.
Stress granules (SG) are inclusions consisting of stalled initiation complexes assembled in the
presence of cellular stress and associated with neurodegeneration, including ALS. Here, we
tested whether SGs could be induced by BMAA. No SGs could be observed in BMAA exposed
cells. Another observation in ALS pathogenesis is inflammation and increased expression of
the High-mobility Group Box 1 protein (HMGB1). The protein is known to translocate to the
cytoplasm during cellular stress, interacting with astrocytic receptors that cause inflammatory
responses. Our group has previously observed an up-regulation of HMGB1 in neurons from
BMAA treated zebrafish larvae. Our study confirmed increased level of HMGB1 in BMAA
exposed neuroblastoma cells, but interestingly the upregulation seemed to be restricted to the
nuclei. This increase seemed to be inhibited by L-serine. Early signs in ALS pathology are
abnormalities in mitochondrial dynamics and biochemistry. Ultrastructural analysis of BMAA
exposed neuroblastoma cells did not show any alterations in mitochondrial morphology.
However, it revealed a higher degree of mitochondrial associated rough ER, which appeared
lower in cells co-treated with L-serine. Lastly, we showed using the Seahorse XFe analyzer that
BMAA altered the mitochondrial metabolism by decreasing basal oxygen consumption rate and
ATP production. L-serine inhibited this decrease.

Elise Førsund

Elise Førsund, Øyvind Halskau, Aurelia Lewis og Espen Bariås

Senescence-Related Changes to the Lipid Composition of SH-SY5Y Cells with Implications for α-Synuclein Misfolding

Along with an increasing life expectancy in the population comes more incidences of ageassociated
disorders like Parkinson’s disease (PD), affecting 1% of the population over the
age of 60 to as high as 5% over the age of 85. At the molecular level, PD is characterized by
the gradual loss of dopaminergic neurons and the presence of protein aggregates in the
affected surviving cells. These aggregates consist of misfolded α-Synuclein (α-Syn), a protein
located in presynaptic terminals of neurons, existing in an equilibrium between soluble and
membrane-bound form. It is proposed that the membrane-bound form of the protein also
modulates its aggregation propensity, making the lipid membrane an interesting focus of PD
research. As aging is the most well-established risk factor for developing the disease, this
thesis aimed to search for age-related changes to the lipid composition of neurons, which
could help explain why old cells are more prone to α-Syn misfolding and aggregation.
To mimic the molecular characteristics of the aging cell, cellular senescence was induced in
the neuroblastoma SH-SY5Y cell line using the cytotoxic anticancer agent etoposide. Lipid
profiling was achieved using a combination of solution-phase 31P NMR and LC-MS/MS. As a
result, interesting findings were made, including a 37% and 19% senescence-related increase
in the abundance of phosphatidylinositol (PI) and phosphatidylserine (PS), which are
negatively charged lipids implicated in α-Syn binding. Additionally, LC-MS/MS detected a
senescence-related decrease of 10.4% in the overall polyunsaturated fatty acid (PUFA)
content, in particular the polyunsaturated omega-6 fatty acid arachidonic acid (AA, 20:4), a
fatty acid involved in the growth and repair mechanisms of neurons. Other interesting
findings regarding the fatty acid composition were a 14.0% decrease in the stearic/oleic acid
(18:0/18:1) ratio, as some researchers now claim that oleic acid may worsen α-Syn pathology.
Some slight changes were also observed in the overall length of the fatty acids, of which short
acids of 10-12 carbons were observed in the senescence-induced cells and not in the untreated
cells, a result which also could be of relevance for PD as misfolding of α-Syn has reported to
only occur in the presence of short hydrocarbon chains. Overall, our findings were in line
with existing knowledge regarding age-induced lipid changes. Although few of these findings
were significant or alone can explain why aging poses the greatest risk factor for developing
PD, they do suggest that aging changes the lipid environment in neurons, possibly to one that
is more susceptible for the neuronal toxicity mechanisms associated with PD.

Shaoor Ahmad Khan

Shaoor Ahmad Khan, Aurelia Lewis, og Andrea Papdiné Morovicz

Role of PIP binding motifs on the nucleolar localization and function of ErbB3 binding protein 1

Sub-cellular localization is key to the specific function of proteins. Proteins can be recruited to different cell compartments via their interaction with the signaling lipids, polyphosphoinositides (PPIn). While their actions have been comprehensively documented in cytoplasmic membranes, these lipids are also present in membrane-less compartments within the nucleus. To understand the function of PPIn in the nucleus, we thought to identify nuclear PPIn binding proteins using quantitative mass spectrometry combined with PPIn affinity pull down. Using this approach, we identified ErbB3-binding protein 1 (EBP1), known to contribute to many cellular functions through interactions with RNA, DNA as well as other proteins. Using biochemical and biophysical approaches, we have demonstrated a direct interaction between EBP1 and PPIns via two lysine rich motifs located in the N- and C-termini. The C-terminal motif was shown to be required for the localization of EBP1 in nucleoli. A frameshift tumor mutant which introduced additional basic residues in the C-terminal motif led to an increase in PPIn binding and nucleolar localization of EBP1. Here we showed that the nucleolar presence of EBP1 correlated with the presence of rRNA for the WT and the frameshift tumor mutant but not the C-term PPIn-binding mutant. This suggests a molecular link between EBP1’s localization and its subsequent effect on nucleolar processes thereby playing a role in cells transformation. We also performed CRISPR/CAS genome editing of exon 2 on PA2G4 gene in HEK293T cells however it didn’t yield any fruitful outcomes. The clones analyzed by sequencing were wild type with no expected change, though few clones showed reduction at protein level with one clone suggesting an in-frame deletion

Kristianne Hjorth Viken

Kristianne Hjorth Viken,Odd Andre Karlsen, Anders Goksøyr og Roger Lill-Langøy

Functional characterization of the ligand binding properties of the Atlantic cod (Gadus morhua) peroxisome proliferator-activated receptor a1 and a2

Perfluoroalkylated substances (PFASs) are persistent in nature, resisting both biotic and abiotic degradation, resulting in the bioaccumulation of these substances in organisms, as well as biomagnification upwards the trophic levels in both terrestrial and marine food webs. These properties has made PFASs an environmental concern.
The peroxisome proliferator-activated receptors (PPARs) are ligand-activated transcription factors and members of the superfamily of nuclear receptors. The PPAR subfamily consists of three members: PPARa, PPARd, and PPARg. The different subtypes have distinct roles in the regulation of the lipid homeostasis, differing in tissue specific expression, ligand specificity, and target genes. Atlantic cod (Gadus morhua) has four different gmPpars; gmPpara1, gmPpara2, gmPpard and gmPparg. gmPpara1 has been demonstrated to be activated by some exogenous compounds including the PPARa model-agonist WY-14643, and four PFASs, including PFOA, PFNA, PFHxA, and PFHxS. Although the gmPpara2 subtype is also responsive to WY-14643, it is not activated by any of the PFASs. When aligning the protein sequence of gmPpara1 and gmPpara2, an extension of 14 additional amino acids (AAs) in the hinge region of gmPpara2 is observed. Moreover, two AAs important for the binding of WY-14643 in a second allosteric binding site identified in the human ortholog (hPPARa) differ in gmPpara2 compared to gmPpara1. We hypothesized that these differences play a role in the observed discrepancies in the activation profiles of gmPpara1 and gmPpara2.
In this study, site-directed mutagenesis was used to remove the 14 additional AAs in the hinge region, and mutate the two AAs in the putative second binding site in gmPpara2. A luciferase-based reporter gene assay in COS-7 cells was then used to study ligand activation of the gmPpars (wild types and mutants) when exposed to WY14643 and selected PFASs. Protein immunoblotting were used to confirm the synthesis and presence of the gmPpar variants in the COS-7 cells, and a cytotoxicity assay was used to monitor the cytotoxicity of the ligands.
Four different mutants were successfully constructed. Notably, the removal of the 14 AAs in the hinge region had a large impact on the activation profile with WY-14643, producing a 15 and 6 times higher fold activation in comparison to the wild type gmPpara2, and gmPpara1, respectively. Mutants containing the AA substitutions in addition to the deletion also increased the WY-14643-mediated activation, but slightly less compared to the gmPpar variant containing only the 14 AA deletion. Intriguingly, none of the mutants were activated by any of the three PFASs tested.

Nikara Pedersen

Nikara Pedersen, Elisabeth Ougnedal, Ingvild Aukrust og Bjørn Ivar Haukanes

Functional analysis of BRCA1 variants of unknown significance (VUS) detected in patients with breast or ovarian cancer

Breast cancer is the most diagnosed cancer among women. While most cancer cases are sporadic in origin, an estimated 5–10% are due to pathogenic germline variants. In the tumor suppressor BRCA1, such variants confer a high risk of developing breast and ovarian cancer and are associated with hereditary breast and ovarian cancer syndrome. BRCA1 variants are interpreted using a five-tier system that scales from class 1 (benign) to class 5 (pathogenic). However, for many variants, evidence is limited or conflicting. These variants are therefore classified as variants of uncertain significance (VUS, class 3), complicating risk assessment and representing a major source of uncertainty for both patients and clinicians. This especially applies to missense variants as their effects are more difficult to interpret, necessitating supporting evidence from functional assays. With the increasing use of genetic testing, VUS represent a current and future challenge. Many VUS are found in the BRCT domain of BRCA1, which is involved in transactivation of gene expression in vivo. In this work, a fusion protein of yeast GAL4 DNA binding domain and human BRCA1 BRCT was used to characterize the transactivation activity of BRCA1 variants by means of a dual luciferase reporter assay. Our analysis included six BRCA1 class 3 variants of interest, four (likely) benign/pathogenic controls, and eight inter-laboratory controls that were previously analyzed at other hospitals in Norway. Based on our functional analysis we propose the re-classification of one variant, p.(Asp1692Ala), as class 4 (likely pathogenic), and eight variants as class 2 (likely benign). We also identify a variant that displays poor concordance as an inter-laboratory control and merits additional investigation: p.(Gly1709Arg). Our findings expand the knowledge base of BRCA1 class 3 variants. but further studies are warranted to support the definitive classification of these VUS.

Sindre Hole

Sindre Hole, Aurelia Lewis og Diana Türcu

Early studies on a novel phosphoinositide 3-kinase p110 isoform, and its localisation

Phosphoionositide 3-kinase (PI3K) is a family of enzymes that are part of the PI3K signalling
pathway, which is one of the most often altered pathways in human cancer. The PI3K
family is divided into three classes, where the class I produce phosphatidylinositol (3,4,5)-
triphosphate (PIP3), a secondary messenger that acts upon several pathways and is linked
to cancer. Of the four distinct isoforms of class I PI3K, the p110, and p110 forms are
ubiquitously expressed in all tissues. Most research has been conducted on p110, due to the
frequent observations of mutations in cancers. Less research has been done on p110, but
it has been shown to exhibit transforming properties when overexpressed. Both the p110
and p110 isoform exhibit high sequence homology as well as signalling properties. One
dierence is the subcellular localisation of these isoforms, p110 being the only one found
in the nucleus. When our group determined the nuclear localisation of p110 in a panel
of endometrial cancer cells; an additional nuclear protein of lower mass was detected using
antibodies targeting dierent sites of p110. The focus of this thesis was to investigate and
explain the occurrence of the lower nuclear p110 form.
During this study, we validated the detection of this p110 form using dierent antibodies.
Its existence as a theoretical short form of p110 became a plausible explanation due to the
following results. The theoretical form starts from an initiation codon at amino acid (aa)
number 555, contains an insert of 5 aa, compared to canonical p110, and is 521 aa long.
In order to validate the existence of the short p110 isoform, the insert was detected by
RT-PCR, and the theoretical ATG site for the short form was found to have a stronger
Kozak sequence then the canonical ATG. The validity of the short form ATG to function as
a start codon was tested by mutating either the canonical ATG or the short form ATG. This
experiment was inconclusive; therefore, it needs to be repeated.
Interestingly the p110 short form was found to be cytoplasmic, not nuclear. It also
exhibited autophagy like patterning when overexpressed, at ve times higher frequency than
the canonical p110. The canonical form as been implicated in autophagy, but these studies
are controversial. Additional studies on the p110 short form may clarify theses ndings.

Helene Lorgen Leithe

Helene Lorgen Leithe, Ole Jacob Nøstebakken og Lise Madsen


Dioxins and dioxin-like polychlorinated biphenyls (dl-PCBs) are chemical congeners formed as by-products from incomplete thermic reactions, found in the environment because of improper disposal from industrial processes and combustions. Dioxins and dl-PCBs are of human concern, as they are persistent to degradation and bioaccumulates in marine food-webs. Salmon (Salmo salar) is recommended to be a part of the human diet because of healthy marine nutrients including omega-3 fatty acids, but is also one of the main routes of dioxin and dl-PCBs exposure. Dioxins and dl-PCBs which binds to the aryl hydrocarbon receptor (AHR) and elicits toxic responses, are incorporated in the toxic equivalency factor (TEF)-system. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is indicated to be the most potent congener, because of its high binding affinity to the AHR. Earlier research has found that the toxicity of TCDD alters the male reproductive system by disrupting spermatogenesis, thereby reducing sperm quantity and sperm motility, decreasing reproductive organs weights, and causing a delay in puberty. In 2018, the European food safety authority (EFSA) estimated a new tolerable weekly intake (TWI) of dioxins and dl-PCBs in food, reducing it from 14 to 2 pg toxic equivalency quotient (TEQ)/kg body weight/week. Male reproduction was one of the pivotal effects when estimating the new TWI, where the daily exposure in adults should be kept below 0.25 pg TEQ/kg bw/day.
The aim of the study was to investigate if continuous doses of TCDD impacts the reproductive system, the spermatogenesis, and sperm quantity in male Wistar HAN rats. In addition, we measured whether inclusion of salmon in the diet could possibly reduce the toxicity of TCDD. Observations made throughout the animal experiment, including body mass development, feed efficiency, haematology tests, and the TCDD-levels in the liver upon sacrifice was analysed to get a wider view of possible effects between the experimental groups. An investigation on male reproduction was performed, assessing the concentration of gonadotropins participating in spermatogenesis, sperm quantity and motility, and the morphology of testis. Lastly, proteomics was performed on testis to obtain a closer look if TCDD or salmon impacts protein abundance and regulation, possibly related to spermatogenesis. In addition, a cell study using gonadotropin releasing hormone-TAG 1-7 cells were performed, assessing the toxicity of various dioxins, furans and PCB126 at doses ranging from 48pM – 1.55nM at equivalent TEQ.
The results from the animal experiment analyses indicate that the toxicity of the utilised TCDD dose did not affect the reproductivity in male Wistar HAN rats, and the inclusion of salmon did not affect the toxicity. However, the results from the testis proteomics revealed that proteins related to spermatogenesis are regulated differently, but still present, when exposed to TCDD compared to the control. The conclusion is that the toxicity from the applied dose of TCDD do not disrupt male Wistar HAN reproduction in a crucial state, as no significant differences related to spermatogenesis or sperm quantity was found. The results from the in vitro cell study confirms that the applied congeners induced cells death in a dose-dependent manner, with varying toxicity even when the doses had an equivalent TEQ.