The role of the stem cell marker OLFACTOMEDIN-4 and the microRNA generator DICER1 in colorectal carcinogenesis

The intestine is a pivotal organ which is divided into two
anatomical parts: the small intestine and the large intestine
(colon and rectum). Both parts are made up of single layered
epithelium. This epithelium is composed of villi (protrusions) –
found only in the small intestine - and crypts (invaginations)
leading to an increase of the surface of the intestinal lumen
whereby the uptake of nutrients and water is improved. Every five
days, the intestinal epithelium is renewed whereby both, crypts and
eventually villi, are filled up with new cells. The homeostasis of
the crypts/villi relies on adult stem cells (SCs), especially crypt
base columnar (CBC) cells, which are located at the base of the
crypts. These are regulated by an active Wnt signaling pathway. A
deregulation of the Wnt signaling pathway leads to cancer formation
found in humans almost exclusively in the colon and rectum.
Colorectal cancer (CRC) is worldwide the third most common cause
for cancer related deaths. In the majority of CRC, origin and
progress are caused by mutations in the adenomatous polyposis coli
(APC) gene which encodes an essential component of the β-catenin
destruction complex that is the central element of the Wnt
signaling pathway. As a consequence of these mutations, the
executor of the Wnt signaling pathway, β-catenin, which is in this
context a transcription factor, cannot be downregulated any more.
As a consequence target genes of β-catenin are expressed in an
unregulated manner. These target genes regulate features of stem
cell biology which confer cancer stemness, metastasis, EMT
(epithelial-mesenchymal transition), chemoresistance and other
characteristics to colorectal tumor cells. Interestingly, APC
mutations have only an effect when they occur in the adult stem
cells. Thus, the descendend tumor cells show characteristics of
these cells and have been termed cancer stem cells (CSCs). Like
adult stem cells in the normal crypt CSCs are the origin of cancer
and are characterized by an activated - here deregulated - Wnt
signaling pathway and thus, by the aforementioned features.
Clinically, cancer death is caused in most cases by metastasis
which is treated by chemotherapy from which most if not all CRCs
escape by the development of chemoresistance which is an intrinsic
feature of the CSCs. Therefore, CSC specific targeted therapies
might be a promising therapeutic tool for a successful treatment of
CRCs. One possibility is the interference of CSC sustaining
molecules as these molecules are involved in the induction and
maintenance of CSCs. Here, a promising molecule is olfactomedin-4
(OLFM4) which was discussed to be a CSC marker. But the role of
OLFM4 as a CSC marker and important factor for tumorigenesis has
been controversially described. Therefore, I investigated in the
first part of my thesis the role of OLFM4 in CRC cells. I
demonstrate that OLFM4 was expressed only in two out of 14 CRC cell
lines. The assumption that OLFM4 was only expressed in cells with
characteristics of CSCs and thus, was not detected in the cell
lines as they possess only a small proportion of CSCs, was not
confirmed. I found that CSCs showed a reduced OLFM4 expression and
thus, OLFM4 was not coexpressed with other SC markers. These
results indicate that OLFM4 is not a marker of CSCs in CRC. In
order to analyze the functional role of OLFM4 in CRC cells, I
overexpressed OLFM4 lentivirally. However, the overexpression of
OLFM4 and thus, high OLFM4 protein levels did not influence the
expression of CSC, EMT or differentiation marker. Likewise, OLFM4
did not play a functional role for proliferation, stemness and
metastatic features. Therefore, this study demonstrates that OLFM4
is not a CSC marker and has no functional role for the driving
activity in the process of colorectal carcinogenesis. Additionally,
I evaluated in the second part of my thesis the role of the
microRNAome (miRNAome) in colorectal carcinogenesis, the influence
on CSC features and whether the miRNAome might be a tool for
specific CSC targeted therapies. microRNAs (miRNAs) are generally
downregulated in tumors whereby the miRNA loss promotes
tumorigenesis. As the majority of the CRC cases are driven by an
APC mutation in the SC compartment, I used for my investigations a
mouse model with a conditional Apc knockout in CBC cells which
develops efficiently intestinal adenomas. This mouse model was
crossed with another mouse model harboring a conditional knockout
of the essential miRNA generator Dicer1 to investigate the role of
a loss of the miRNAome in murine Wnt driven intestinal tumors. In
this part of my study I demonstrated that hetero- and homozygous
deletion of Dicer1 in CBC cells, in combination with an Apc
knockout, enhances significantly the number of adenomas. Moreover,
deletion of Dicer1 resulted in smaller adenomas caused by reduced
proliferation. Further analysis of DICER1 deletion in human CRC
cell lines revealed that loss of DICER1 and thus, miRNAs led
likewise to a decreased proliferation. Additionally, I showed that
loss of miRNAs increased the expression/protein levels of CSC
markers and CSC features indicating that loss of DICER1 promotes
tumorigenesis. Moreover, I translated these mouse model/cell
culture results into human colonic normal and tumor tissue as well
as CRC. In a collection of different tissues (normal tissue,
adenomas and cancers of stages I to IV), increased DICER1 levels
were seen from normal tissue to adenomas followed by decreased
levels during carcinoma progression. Increased levels of DICER1
were also found in the murine Wnt driven adenomas. In support with
this I provided finally evidence that DICER1 expression is
regulated by the Wnt signaling pathway thus already early in the
beginning of the colorectal tumorigenesis. Thus, this data showed
that DICER1 is a tumor suppressor in intestinal cancer and the loss
of DICER1 and hence, of the miRNAome, influences CSC marker
expression and marker protein levels as well as proliferation and
CSC features. Therefore, the miRNAome might possibly become a
therapeutic target for CSC targeted therapy.