Project
Study on the role of beta cell serotonin during pregnancy in the mouse.
Pregnancy is period in which the maternal body has to undergo several adaptations to meet the
demands of the fetus, for example in pancreatic beta cells. Pregnancy hormones stimulate prolactin
receptors (PRLRs) on beta cells to increase their capacity to produce insulin. In the last years, several
studies have been conducted on these changes in beta cells. Pancreatic islets of non-pregnant and
pregnant mice (at different time points) were compared for their mRNA expression profile. Two
genes were found to be highly induced during pregnancy in islets, Tph1 and Tph2, encoding the
enzymes tryptophan hydroxylase 1 and 2. These enzymes are responsible for the rate-limiting step
of serotonin biosynthesis, leading to the production and secretion of beta cell serotonin.
Remarkable about this serotonin is that not all beta cells produce serotonin to the same extent, but
that it displays a heterogeneous production pattern. Questions raised about the function of
serotonin in beta cells. A first study showed that serotonin acts locally to increase beta cell
proliferation during pregnancy via the 5-HTR2B receptor. Another study claimed that serotonin is
responsible for the stimulation of glucose-induced insulin secretion via the 5-HTR3A receptor.
In this thesis, pregnancy-related beta cell adaptations were investigated with a focus on the role of
serotonin in the process. In a first part, it was shown that both female and male islets are capable
of undergoing pregnancy-induced changes. Using an in vitro and an in vivo approach, it was
demonstrated that islets of both genders upregulate the same set of genes, produce serotonin and
have an increased beta cell proliferation rate. Furthermore, it was shown that the upregulation of
the set of genes was caused by stimulation of prolactin receptors by placental lactogens.
The main finding of this thesis was that neither beta cell serotonin nor 5-HTR2B are needed for the
increase in beta cell proliferation during pregnancy, which is in contrast with the results of another
research group. The percentage of proliferating beta cells was assessed in three different mouse
models in non-pregnant and pregnant condition. A first model was the total Tph1KO mouse, the
second model a newly discovered spontaneous “beta cell-specific” Tph1betaKO model. Both models
underwent a normal increase in proliferation compared to control mice. The third studied model
was the Htr2b deficient mouse strain, in which, again, no difference was found between Htr2b+/+
and Htr2b-/- mice regarding beta cell proliferation. In the future, more experiments are needed to
unravel the function of serotonin. In the Tph1betaKO mice, the influence of local serotonin on insulin
secretion and on islet blood flow will be investigated.
The discovery of the spontaneous “beta cell-specific” Tph1betaKO mouse is an important step
forward in the search to the role of beta cell serotonin. However, the path towards a useful model
was highly convoluted and full of hurdles and pitfalls. At first, the Cre/LoxP technology was used to
generate a mouse model. In total, four different models were tested (Pdx1-CreLate, Ngn3-Cre, Rip-
CreERTR26LacZ and Tph1fl/fl) and in three out of four models, problems were found. The Pdx1-CreLate
mouse contained the full sequence of the human growth hormone gene (hGH) and this led to a
pregnancy phenotype in non-pregnant mice. The Ngn3-Cre father mice were demonstrated to have
aspecific recombination in testes, giving rise to total Tph1KO mice among the offspring. Finally, the
Tph1fl/fl mouse turned out to be no good due to poor design of the floxed allele, as a fully active
truncated TPH1 enzyme was expressed in the Tph1betaKO mice. Therefore, Cre/LoxP technology
could not be used for the study. The discovery of all these artefacts in just one study suggests that
these artefacts sneak into the design of the conditional knock-out mice in other scientific fields,
beside the beta cell field.
The role of 5-HTR2B was already studied in islets, however, its highest mRNA signal was found in
placenta. Moreover, the signal for Htr2b mRNA was the second highest of all serotonin receptors in
the mouse body. The possibility of an endocrine communication axis between islets and placenta
was studied. In this model, beta cell serotonin would stimulate 5-HTR2B on placental cells providing
the sufficient production of pregnancy hormones. On their turn, pregnancy hormones stimulate
PRLRs on beta cells. Htr2b deficient placentas were compared with wild-type placentas for their
gene expression levels. None of the expressed genes, including Prl genes, in placenta were altered
in Htr2b-/- mice. Furthermore, placentas looked normal on macroscopic level. The function of 5-
HTR2B in placenta remains an open question and requires further research.