Characterization of murine pancreatic carcinoma models regarding immunosuppressive mechanisms and therapy with bifunctional siRNA targeting galectin-1

Pancreatic ductal adenocarcinoma (PDAC) is a very aggressive tumor
that is characterized by abundant tumor stroma and a potent
immunosuppressive microenvironment. Further studies to clarify why
T cells infiltrate the tumor but are not able to perform effector
functions as well as to find new effective therapies to overcome
immunosuppression are urgently needed. The aims of the present
study were (1) to characterize different murine PDAC tumor models
in regard to their utility for studying novel immunotherapeutic
approaches, (2) to assess the therapeutic efficacy of a novel
bifunctional ppp-siRNA that combines silencing of the
immunosuppressive molecule galectin-1 and RIG-I-mediated immune
activation in murine PDAC and (3) to characterize the
immunosuppressive mechanisms leading to T cell inhibition in the
tumor microenvironment. (1) This study revealed that the pancreatic
cancer cell line T110299, which was developed from a primary tumor
of the KPC mouse model, is a new valuable tool for studying novel
treatment strategies for PDAC. The histological appearance of
T110299 tumors reflects in many aspects the primary tumors in KPC
mice, which harbor mutations in the Kras oncogene and p53, and the
human disease with regard to tumor differentiation, extensive tumor
stroma development, poor vascularization and expression of
immunosuppressive molecules, like indoleamine 2,3-dioxygenase (IDO)
and galectin-1 (Gal-1). In contrast, Panc02 tumors were found to
have a sarcomatoid architecture with very little tumor stroma. In
Panc02 cells, galectin-1 was strongly expressed by the tumor cells,
which differs from the situation found in humans, KPC mice and
T110299 tumors, where galectin-1 is preferentially expressed in
tumor-associated pancreatic stellate cells (PSC). However,
expression of the cytosolic helicase RIG-I was functional in Panc02
cells, but defective in T110299 cells. As RIG-I is expressed in all
human PDAC cell lines tested, the Panc02 model appears to be better
suited to study RIG-I-based immunotherapies (Ellermeier et al.,
2013). Thus, the histological and functional characterization of
the tumor models in this thesis will allow selecting the
best-suited tumor model for addressing specific aspects of
immunotherapy. (2) Treatment studies of PDAC were performed with
the 5’ppp-modified siRNA molecule ppp-Gal-1 in the orthotopic
Panc02 tumor model. The dual activities of this molecule were
confirmed in vitro, leading to (i) reduced galectin-1 expression
via RNAi; and (ii) production of CXCL10 and IFN-, MHC-I
up-regulation and apoptosis of tumor cells via RIG-I activation.
Treatment of mice with orthotopic pancreatic tumors with ppp-Gal-1
significantly prolonged survival, as compared to unmodified
OH-Gal-1 or control RNA. In addition, 20% of the mice completely
rejected their tumors leading to long-term tumor control. Thus,
bifunctional 5’ppp-modified siRNA is a promising treatment strategy
for PDAC deserving further pre-clinical evaluation. (3) Pancreatic
tumor cells employ multiple mechanisms for suppression of T cell
responses. This study identified TGF-β and IDO as two potent
mechanisms leading to inhibition of T cell proliferation. Minute
amounts of PDAC supernatants effectively blocked T cell
proliferation induced by CD3 and CD28 triggering. This could be
partially prevented by SD-208, a small molecule inhibitor of TGF-β
receptor signaling, or by blocking IDO activity with D-1-MT.
Interestingly, tumor supernatants induced up-regulation of IDO mRNA
expression in T cells. Furthermore, blocking IDO activity in T
cells appeared to be more effective than blocking IDO in tumor
cells. This leads to a new hypothesis that factors secreted by the
tumor cells induce IDO expression in T cells, which in turn leads
to auto-intoxication of the T cells via kynurenine production and
eventually T cell apoptosis. Further studies confirming this
hypothesis are warranted.