Cell Metabolism
Letter
Endogenous Hyperinsulinemia Contributes
to Pancreatic Cancer Development
Anni M.Y. Zhang,1 Jamie Magrill,1 Twan J.J. de Winter,1 Xiaoke Hu,1 Søs Skovsø,1 David F. Schaeffer,2 Janel L. Kopp,1,*
and James D. Johnson1,*
1Department of Cellular and Physiological Sciences, Life Sciences Institute, University of British Columbia, Vancouver, BC, Canada
2Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada
*Correspondence: janel.kopp@ubc.ca (J.L.K.), james.d.johnson@ubc.ca (J.D.J.)
https://doi.org/10.1016/j.cmet.2019.07.003
The incidence of cancers, including pancre-
atic ductal adenocarcinoma (PDAC), con-
tinues to rise alongside the global obesity
and diabetes epidemics. This rise sug-
gests that factors modulated by environ-
ment or lifestyle play a key role, but the
mechanisms by which obesity and type 2
diabetes promote cancer remain unclear.
Obesity and type 2 diabetes are associ-
ated with elevations in circulating insulin
and glucose, as well as systemic low-
grade inflammation, which have been pro-
posed to contribute to cancer initiation or
progression. Primary hyperinsulinemia,
defined as circulating insulin in excess of
what is required for glucose homeostasis
and independent of insulin resistance,
can be found in 30% of obese adults
(Tricò et al., 2018). Hyperinsulinemia is
associated with PDAC, independent of
BMI (Tsujimoto et al., 2017). Nevertheless,
the cause and effect relationship between
hyperinsulinemia and cancer remains to
be determined directly.
Caloric restriction, fasting, and low-car-
bohydrate diets have been proposed for
prevention or as an adjunct to cancer ther-
apy based in part on what has come to be
known as the insulin-cancer hypothesis.
Insulin has mitogenic effects on human
pancreatic cells via signaling implicated
in tumorigenesis (Chan et al., 2014), but
there is no direct evidence that hyperinsu-
linemia in vivo can affect the development
of any cancer.
KRAS mutations are found in 90% of
human PDAC and reliably induce preneo-
plastic pancreatic intraepithelial neoplasia
(PanIN) precursor lesions when expressed
in murine pancreatic acinar cells using an
inducible Ptf1aCreER allele (Basturk et al.,
2015; Kopp et al., 2012). To test the
hypothesis that a diet-induced increase in
circulating insulin plays a causal role in
PDAC initiation, Ptf1aCreER;LSL-KrasG12D
mice with reduced insulin gene dosage
were fed a high-fat diet (HFD) from wean-
ing to 1 year of age. HFD causes hyperin-
sulinemia and accelerates PanIN forma-
tion/progression in KrasG12D-expressing
mice (Chang et al., 2017). We employed
controls with a normal complement of
wild-type Ins1 alleles (Ins1+/+) and experi-
mental mice with only a single functional
Ins1 allele (Ins1+/ ) from the same parental
sets (Figure S1A). All mice were maintained
on an Ins2 / background to prevent insu-
lin gene compensation (Mehran et al.,
2012). We previously found that these ge-
netic manipulations of Ins1 and Ins2 result
in a sustained reduction in fasting insulin
modest enough not to chronically impair
glucose homeostasis (Mehran et al.,
2012). On the Ptf1aCreER;LSL-KrasG12D
background, there was the expected mi-
nor reduction in fasting insulin and body
weight in mice with reduced Ins1
compared to controls (Figures S1C–S1F).
Importantly, this reduction in insulin
in Ptf1aCreER;LSL-KrasG12D;Ins1+/ ;Ins2 /
mice did not affect fasting glucose in fe-
males over 1 year of age (Figure S1G). In
male mice from the same litters, we
observed impaired fasting glucose, and in
some cases frank diabetes (Figure S1H),
consistent with known sex differences
in insulin sensitivity. Therefore, comparing
female Ptf1aCreER;LSL-KrasG12D;Ins1+/ ;
Ins2 / mice to female Ptf1aCreER;LSL-
KrasG12D;Ins1+/+;Ins2 / controls offered
us the unique opportunity to test the role
of insulin in pancreatic cancer initiation,
in the absence of changes in fasting
glucose.
To test our primary hypothesis that
decreasing endogenous insulin production
would affect the initiation of KrasG12D-
driven PDAC, we measured the percent
of total pancreatic area occupied by
PanINs and tumor. At 57 weeks of age,
abundant ductal lesions with histologic
and molecular characteristics of low-grade
PanINs, including highly acidic mucin con-
tent indicated by Alcian blue staining (Fig-
Cell Metabolism 30, September 3, 2019 ª 2019 Elsevier Inc. 403
ures S1I and S1J), were found in both
groups. Only one mouse from each geno-
type developed PDAC, which expressed
the ductal marker Cytokeratin 19 (Figures
S1K and S1M). Remarkably, Ptf1aCreER;
LSL-KrasG12D;Ins1+/ ;Ins2 / mice had
approximately half the area covered by
PanINs or tumor (12.7% ± 3.4%)
compared with Ptf1aCreER;LSL-KrasG12D;
Ins1+/+;Ins2 / control mice (25.4% ±
3.8%; p = 0.03) (Figure S1M). Without
including tumor area, experimental mice
still had approximately half the area
covered by PanINs (12.3% ± 3.1%; p =
0.01) compared to controls (24.7% ±
3.3%) (Figure S1N). Mice with reduced in-
sulin also had significantly fewer high-
grade PanINs per section (Figures S1L
and S1O) or per whole pancreas area.
There was no significant difference be-
tween the genotypes when the number of
high-grade PanINs was normalized to the
total PanIN area. At 1 year of age, PanIN
plus tumor area correlated with fasting in-
sulin, but not glucose levels (Figures
S1P–S1R), arguing against a prominent
role for glycemia in PanIN formation in
our model. Although we were unable to
follow them for 1 year, male mice with
reduced Ins1 had less PanINs (Figures
S1S–S1U). Together, our data implicate
endogenous increases in insulin, but
not glucose, in HFD-mediated promotion
of PanIN development and potentially
pancreatic cancer.
We investigated possible underlying
mechanisms associated with insulin-
driven PanINs formation, starting with the
mitogenic effects of insulin (Chan et al.,
2014). We were unable to measure statisti-
cally significant differences in proliferation
rate in any cell types, although variation
was high between mice within groups (Fig-
ures S1V and S1W). Our data cannot
exclude the possibility that insulin provided
mitogenic and anti-apoptotic signals to
cells expressing oncogenic KrasG12D prior

to the time points we studied. PDAC is
characterized by an intense desmoplastic
reaction, and the increased inflammation
associated with PanIN formation could
drive the desmoplasia. Studies have sug-
gested that resulting cancer-associated
stroma could support tumor survival and
growth (Sousa et al., 2016). Insulin has
also been shown to enhance the prolifera-
tion of pancreatic stellate cells and aug-
ments their production of extracellular
matrix proteins through PI3K (Yang et al.,
2016). By staining the collagen with Sirius
Red, we found that mice with reduced in-
sulin had significantly less collagen depo-
sition than controls (20.1% ± 4.3% and
41.3% ± 5.1%, respectively; p = 0.007)
(Figures S1X and S1Y). This result is
consistent with the idea that hyperinsuline-
mia promotes PanIN development in part
by contributing to the fibrogenesis associ-
ated with PanINs, or that increased fibrosis
is caused by the greater number of
PanINs.
Our study is the first to separate the
causal role of hyperinsulinemia from hy-
perglycemia and directly test the insulin-
cancer hypothesis in vivo. Limitations of
our study include the analysis of PanIN
at a single time point. Future studies
should provide a more detailed time
course of the early initiation and address
the role of insulin in progression to
PDAC. Notwithstanding, our findings are
consistent with our recent study demon-
strating that a reduction of fasting insulin,
so mild that it did not significantly increase
fasting glucose, was sufficient to signifi-
cantly extend mean and maximum life-
span of mice (Templeman et al., 2017).
Insulin-lowering interventions such as ex-
ercise, diet, and metformin warrant further
investigation as strategies to prevent
cancer or limit its progression.
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at
https://doi.org/10.1016/j.cmet.2019.07.003.
ACKNOWLEDGMENTS
The authors thank members of the Johnson and
Kopp labs for discussions. This study was funded
404 Cell Metabolism 30, September 3, 2019
by a grant from the Cancer Research Society
(Canada), an Ideas Grant from the Pancreas
Centre BC (Canada), and an Innovation Grant (IN-
NOV16-1) from the Canadian Cancer Society
Research Institute (Canada) to J.D.J. and J.L.K.
A.M.Y.Z. was supported by Canada Graduate
Scholarships – Master’s program, a Four-Year
Fellowship from the University of British
Columbia, and Frederick Banting and Charles
Best Canada Graduate Scholarships.
AUTHOR CONTRIBUTIONS
A.M.Y.Z. designed and managed the study; ac-
quired, analyzed, and interpreted all data unless
otherwise noted; and wrote the manuscript. J.M.
acquired, analyzed, and interpreted data (CK19,
Ki67, and Alcian blue staining). T.J.J.d.W. ac-
quired, analyzed, and interpreted data (male PanIN
staining). X.H. managed the mouse colony and
performed mouse husbandry. S.S. provided expert
advice on breeding and study design. D.F.S. pro-
vided expert advice on pancreatic cancer pathol-
ogy and study design. J.L.K. designed, analyzed,
and interpreted experiments; edited the manu-
script; obtained funding; and supervised the study.
J.D.J. analyzed and interpreted data, edited the
manuscript, obtained funding, supervised the
study, and conceived the study concept and
design.
SUPPORTING CITATIONS
The following references appear in the Supple-
mental Information: Alejandro et al. (2011); Basturk
et al. (2015); Chan et al. (2014); Chang et al. (2017);
Kopp et al. (2012); Lee et al. (2018); Mehran et al.
(2012); Pan et al. (2013); Sousa et al. (2016);
Templeman et al. (2017); Tricò et al. (2018); Tsuji-
moto et al. (2017); Tuveson et al. (2004); Yang
et al. (2016).
REFERENCES
Alejandro, E.U., Lim, G.E., Mehran, A.E., Hu, X.,
Taghizadeh, F., Pelipeychenko, D., Baccarini, M.,
and Johnson, J.D. (2011). Pancreatic b-cell Raf-1
is required for glucose tolerance, insulin secre-
tion, and insulin 2 transcription. FASEB J. 25,
3884–3895.
Basturk, O., Hong, S.M., Wood, L.D., Adsay, N.V.,
Albores-Saavedra, J., Biankin, A.V., Brosens, L.A.,
Fukushima, N., Goggins, M., Hruban, R.H., et al.;
Baltimore Consensus Meeting (2015). A revised
classification system and recommendations from
the Baltimore Consensus Meeting for Neoplastic
Precursor Lesions in the Pancreas. Am. J. Surg.
Pathol. 39, 1730–1741.
Chan, M.T., Lim, G.E., Skovsø, S., Yang, Y.H.,
Albrecht, T., Alejandro, E.U., Hoesli, C.A., Piret,
J.M., Warnock, G.L., and Johnson, J.D. (2014).
Effects of insulin on human pancreatic cancer pro-
gression modeled in vitro. BMC Cancer 14, 814.
Chang, H.H., Moro, A., Takakura, K., Su, H.Y., Mo,
A., Nakanishi, M., Waldron, R.T., French, S.W.,
Cell Metabolism
Letter
Dawson, D.W., Hines, O.J., et al. (2017).
Incidence of pancreatic cancer is dramatically
increased by a high fat, high calorie diet in
KrasG12D mice. PLoS One 12, e0184455.
Kopp, J.L., von Figura, G., Mayes, E., Liu, F.F.,
Dubois, C.L., Morris, J.P., 4th, Pan, F.C.,
Akiyama, H., Wright, C.V., Jensen, K., et al.
(2012). Identification of Sox9-dependent acinar-
to-ductal reprogramming as the principal mecha-
nism for initiation of pancreatic ductal adenocarci-
noma. Cancer Cell 22, 737–750.
Lee, A.Y.L., Dubois, C.L., Sarai, K., Zarei, S.,
Schaeffer, D.F., Sander, M., and Kopp, J.L.
(2018). Cell of origin affects tumour development
and phenotype in pancreatic ductal adenocarci-
noma. Gut. Published online January 23, 2018.
https://doi.org/10.1136/gutjnl-2017-314426.
Mehran, A.E., Templeman, N.M., Brigidi, G.S., Lim,
G.E., Chu, K.Y., Hu, X., Botezelli, J.D., Asadi, A.,
Hoffman, B.G., Kieffer, T.J., et al. (2012).
Hyperinsulinemia drives diet-induced obesity inde-
pendently of brain insulin production. Cell Metab.
16, 723–737.
Pan, F.C., Bankaitis, E.D., Boyer, D., Xu, X., Van de
Casteele, M., Magnuson, M.A., Heimberg, H., and
Wright, C.V. (2013). Spatiotemporal patterns of
multipotentiality in Ptf1a-expressing cells during
pancreas organogenesis and injury-induced facul-
tative restoration. Development 140, 751–764.
Sousa, C.M., Biancur, D.E., Wang, X., Halbrook,
C.J., Sherman, M.H., Zhang, L., Kremer, D.,
Hwang, R.F., Witkiewicz, A.K., Ying, H., et al.
(2016). Pancreatic stellate cells support tumour
metabolism through autophagic alanine secretion.
Nature 536, 479–483.
Templeman, N.M., Flibotte, S., Chik, J.H.L., Sinha,
S., Lim, G.E., Foster, L.J., Nislow, C., and Johnson,
J.D. (2017). Reduced circulating insulin enhances
insulin sensitivity in old mice and extends lifespan.
Cell Rep. 20, 451–463.
Tricò, D., Natali, A., Arslanian, S., Mari, A., and
Ferrannini, E. (2018). Identification, pathophysi-
ology, and clinical implications of primary insulin
hypersecretion in nondiabetic adults and adoles-
cents. JCI Insight 3, https://doi.org/10.1172/jci.
insight.124912.
Tsujimoto, T., Kajio, H., and Sugiyama, T. (2017).
Association between hyperinsulinemia and
increased risk of cancer death in nonobese and
obese people: a population-based observational
study. Int. J. Cancer 141, 102–111.
Tuveson, D.A., Shaw, A.T., Willis, N.A., Silver, D.P.,
Jackson, E.L., Chang, S., Mercer, K.L., Grochow,
R., Hock, H., Crowley, D., et al. (2004).
Endogenous oncogenic K-ras(G12D) stimulates
proliferation and widespread neoplastic and devel-
opmental defects. Cancer Cell 5, 375–387.
Yang, J., Waldron, R.T., Su, H.Y., Moro, A., Chang,
H.H., Eibl, G., Ferreri, K., Kandeel, F.R., Lugea, A.,
Li, L., and Pandol, S.J. (2016). Insulin promotes
proliferation and fibrosing responses in activated
pancreatic stellate cells. Am. J. Physiol.
Gastrointest. Liver Physiol. 311, G675–G687.