Opening new doors: Hedgehog signaling and the pancreatic cancer stroma
Damhofer, H.
Publication date
2015
Document Version
Final published version
Link to publication
Citation for published version (APA):
Damhofer, H. (2015). Opening new doors: Hedgehog signaling and the pancreatic cancer
stroma.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
CHApTer
AdAM12 is a serum biomarker for
stromal activation and prognosis
in pancreatic cancer patients
Helene Damhofer, Lennart B. van Rijssen, Veronique L. Veenstra, Marc J. van de Vijver, Frederike Dijk, Johanna W. Wilmink, Marc G. Besselink, Tom van Leusden, Jan Paul Medema, Hanneke W.M. van Laarhoven, Maarten F. BijlsmaManuscript in preparation
6
ABsTrACT
Pancreatic cancer is characterized by an abundant desmoplastic stroma that is thought to prevent the delivery of drugs to the tumor and thereby limit the effects of conventional chemotherapy in patients suffering from this disease. Moreover, multiple lines of evidence suggest additional mechanisms to exist by which the stroma aids tumor formation, progression and metastasis. As a consequence, the stroma has come to be seen as a promising therapeutic target in pancreatic cancer patients, but no good markers are available to monitor the effect of stromal targeting therapies or to predict response to them. Here we show that ADAM12 is highly upregulated in the stroma of pancreatic cancer patients and that its expression correlates with stromal marker expression and stromal gene signatures. Bioinformatics as well as in vitro experiments suggest that TGF-β signaling mediates the upregulation of ADAM12. Stromal ADAM12 expression was reflected in the levels found in serum of pancreatic cancer patients as compared to healthy controls, and high serum levels of this protein were associated with poor clinical outcome and complemented existing predictors of prognosis. These findings thus provide a good rationale to consider ADAM12 in future research as a serum marker for stromal activation in pancreatic cancer patients undergoing therapies targeting this compartment.
6
inTrOdUCTiOn
Treatment outcomes for patients diagnosed with pancreatic ductal adenocarcinoma (PDAC), the fourth leading cause of cancer-related death in Western societies, have remained largely unchanged for the last three decades. The median survival time after diagnosis remains less than 6 months. Even with combination of surgery and chemotherapy, the survival is limited and in most of the cases patients will succumb to the disease within 3 to 5 years [1]. Furthermore, approximately 80% of patients with pancreatic cancer are diagnosed with locally advanced or metastatic disease, precluding treatment with curative intent. Treatment options for these patients are rather limited [2] and the majority of clinical trials performed showed only partial benefit in a few patients [3], indicating that better patient selection before start of therapy, and non-invasive monitoring during course of treatment are needed to improve current and future therapeutic approaches.
Initial research efforts on pancreatic cancer pathogenesis mainly focused on the molecular biology of tumor cells themselves. In recent years, however, increasing attention is being paid to the non-malignant cells comprising the tumor microenvironment, also called stroma. One of the hallmarks of pancreatic cancer is an abundance of stroma, which is composed of extracellular matrix proteins such as collagens, hyaluronic acid, and secreted protein acidic and rich in cysteine (SPARC), but also cellular elements including cancer associated fibroblasts (CAFs), endothelial cells, and immune cells [4]. This desmoplastic stroma can make up more than 80% of tumor area and is thought to play a pivotal role in the chemoresistance of PDAC by preventing access of the drugs to tumor cells and additionally exert tumor supportive effects [5].
In recent years, several strategies have been developed to target the tumor stroma to enhance the efficacy of chemotherapy. One of the most promising candidates is nab-paclitaxel, an albumin-bound nanoparticle formulation of paclitaxel that can bind to SPARC and lead to depletion of stromal fibroblasts [6-8]. A recently reported phase III study has shown improved survival with combination treatment nab-paclitaxel plus gemcitabine, as compared to gemcitabine monotherapy in patients with advanced PDAC [9]. Several other stromal targeting therapeutic regimens are currently also being tested in clinical trials. However, at the moment there is no good marker available to select patients for stromal targeting therapies, or to monitor therapeutic effects in a non-invasive manner.
ADAM12 is a member of the family of A disintegrin and metalloproteases (ADAMs) and possesses extracellular metalloprotease and cell-binding functions, as well as intracellular signaling capacities. The protein is involved in cell adhesion via binding to integrins and syndecans as well as proteolytic cleavage of several substrates from the surface of producing cells, a process termed ectodomain shedding [10]. Known substrates of ADAM12 include amongst others epidermal growth factor (EGF) [11, 12], heparin-binding EGF (HB-EGF) [13], notch ligand Delta-like 1 (DLL1) [14], and Sonic Hedgehog (SHH) [15, 16], molecules that were also found to play an important role during tumor development. Upregulated expression of ADAM12 is related to a variety of human malignancies such
6
as glioblastomas [17], breast [18], bladder [19], lung [20], prostate [21], and liver cancer [22], and it was shown to correlate with cancer stage in breast and bladder cancer [18, 19] as well as with prognosis in small cell lung cancer patients [23]. A previous study has shown the upregulation of ADAM12 in cancer associated fibroblasts (CAFs) from pancreatic cancer patients compared to non-neoplastic fibroblasts [24], however the clinical relevance of this expression was not explored.
In the present study, we show that ADAM12 is upregulated in the stroma of pancreatic cancer patients and that it has potential to serve as a serum-borne proxy for stromal content and activation status. Furthermore, we demonstrate that TGF-β ligands can drive ADAM12 expression in CAFs. Serum levels of ADAM12 were elevated in pancreatic cancer patients compared to healthy controls, and these levels predicted poor clinical outcome. This suggests that serum-borne ADAM12 may be useful as a non-invasive marker to monitor stromal recruitment and activation in pancreatic cancer patients, warranting further investigation in treatment regimens targeting the stroma.
resULTs
AdAM12 is upregulated in the pancreatic cancer stroma
To explore the expression of ADAM12 in pancreatic cancer patients we made use of two published microarray datasets derived from resected tissue specimens and found ADAM12 to be significantly upregulated in tumor tissue compared to normal pancreatic tissue in both datasets investigated (Figure 1A) [25, 26]. Additionally, after dichotomizing patients in either high or low ADAM12 expression by median expression on microarray, we found that high expression of ADAM12 in tumor tissue was associated with worse survival after surgical resection in the Badea cohort (Figure 1B). This suggests a potential relevance of ADAM12 expression for clinical outcome. By comparing the ADAM12 expression levels in tumor tissue to cell lines, we observed that the expression in pancreatic cancer cell lines is generally low or nearly undetectable (Figure 1C), suggesting that the origin of ADAM12 in cancer tissue is not the epithelial fraction, but more likely the tumor microenvironment. Immunohistochemical staining for ADAM12 on human pancreatic cancer tissue revealed strong reactivity with the desmoplastic stroma discernible by fibrillary staining patterns (Figure 1D, black arrow), however some membranous staining was observed in the epithelial cells as well, matching the findings from the gene expression studies that showed low expression levels in cell lines. To more conclusively address the source of ADAM12 expression in the epithelial versus the stromal compartment by molecular profiling, we measured the expression of different ADAM family members in patient-derived xenograft (PDXs) models of pancreatic cancer patients [27]. During outgrowth of the patient tumor in immunocompromised host animals, the human stroma is completely replaced by murine cells, allowing the use of species-specific qPCR to determine gene expression in the epithelial or stromal fraction specifically. Reflecting the observation in cell lines on microarray, human ADAM12 expression was found to be relatively low compared to the
6
high expression of two other ADAM family members, ADAM10 and ADAM17 (Figure 1E). These two metalloproteases are the best characterized sheddases and are often found upregulated in different malignancies including pancreatic cancer [15, 28]. However, an opposite expression pattern of ADAMs was seen in the stromal mouse compartment, with high abundance of ADAM12 transcript compared to the other ADAMs (Figure 1F). Staining
Figure 1. ADAM12 is upregulated in pancreatic cancer patients and expressed in the stroma. A) Box plots
showing log2 transformed gene expression values from two microarray datasets of pancreatic cancer patients comparing normal and tumor tissue. Badea et al. set (GSE15471), n=36 paired biopsies; Pei et al. set (GSE16515); n=15 (normal), n=36 (tumor). **P<0.01, ***P<0.001. B) Kaplan-Meier plot of patients from the Badea set, divided in high or low ADAM12 expressing by median expression value. C) Box plots showing log2 transformed ADAM12 gene expression values from the Badea tumor set compared to expression found in pancreatic cancer cell line sets Maupin (GSE21654), n=22 and Broad Institute (GSE36133), n=43. D) Immunohistochemistry for ADAM12 on resected material from PDAC patient. Black arrows indicate stromal staining. E) Transcript levels for indicated genes were measured on 10 different patient derived xenografts by qPCR using human-specific primers. Box plots show min to max distribution of ADAM transcripts relative to human GAPDH. F) As for panel E, using mouse-specific primers. G) Immunohistochemical staining of ADAM12 on PDX tumor. H) Flow cytometric staining for ADAM12 and mouse MHC class I molecule H2Db on a dissociated PDX tumor. First and second panel show single stainings for ADAM12 and H2Db respectively; third panel shows double staining for both markers.
A
*** **
normal tumor normal tumor
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 cell lines D E F ADAM12 C B survival [days] ADAM12 low ADAM12 high
log rank p-value 0.025 1.0 0.8 0.6 0.4 0.2 0.0 cum survival 0 500 1000 1500 2000 2500 H ADAM12 ADAM12-APC H2Db-Alexa488 100 µm G H2Db ADAM12+H2Db 200 µm ADAM12
gene expression log2
gene expression log2
Badea set
expression relative to hGAPDH expression relative to mGapdh
hADAM10 hADAM17 hADAM12 mAdam10 mAdam17 mAdam12
10-1 10-2 10-3 10-4 10-5 10-6 10-1 10-2 10-3 10-4 100
6
of ADAM12 on PDX tissue showed strong reactivity with peritumoral fibroblasts (Figure 1G) and flow cytometric analysis of dissociated xenograft tissue revealed high reactivity with cells coexpressing the mouse MHC class I molecule H2Db confirming the murine, and therefore stromal, origin of ADAM12 protein (Figure 1H).
Although immunohistochemistry and detailed spatial analysis of more patient samples will be needed to definitively address the exact distribution of ADAM12 expression in the different cell types compromising the complex pancreatic tumor microenvironment, our data do show that upregulation of ADAM12 expression in pancreatic cancer patients originates mainly in the tumor stroma rather than the epithelium.
stromal activation in patient tumors is correlated with AdAM12 expression
To further investigate the association of ADAM12 expression with the pancreatic cancer stroma, we correlated expression levels of ADAM12 in resected tumor specimens with the commonly used stromal activation marker ACTA2 (gene encoding for alpha-smooth muscle actin, α-SMA) and found that tumors with higher ADAM12 levels also displayed higher ACTA2 expression on microarray (Figure 2A). Similar correlations were found for the expression of COL1A1 (Figure 2B), one of the main collagens produced and deposited by activated cancer associated fibroblasts [29, 30], and SPARC (secreted protein, acidic and rich in cysteine), another prominent constituent of the extracellular matrix in pancreatic cancer (Figure 2C). These correlations were confirmed in two independent microarray datasets from pancreatic cancer tissue (Supplementary Figure) and critically, also by qPCR analysis of tumor specimen from the additional AMC cohort of 15 patients (Figure 2 D-F). Interestingly, an inverse correlation with the ductal epithelial marker cytokeratin 19 (KRT19) was observed (Figure 2G) and this was reflected in the negative correlation of ADAM12 expression with the tumor cell percentage score of the tissue used for RNA isolation (Figure 2H). By employing gene set enrichment analysis (GSEA) on patient microarray datasets we found a strong enrichment of an extracellular matrix gene signature in tumors with high compared to low ADAM12 expression levels (Figure 2I). Additionally, genes specifically expressed in the activated pancreatic stroma [31] were highly associated with ADAM12 high tumors (Figure 2J, Supplementary Figure). Together these data prove that ADAM12 can be used as a surrogate marker to assess the stromal content in tumor tissue of pancreatic cancer patients.
TgF-
β
regulates AdAM12 expression in cancer associated fibroblastsOne of the best studied signaling molecules involved in the activation of cancer associated fibroblasts (CAFs) during pancreatic cancer progression is the profibrogenic cytokine transforming growth factor beta (TGF-β), which has been shown to increase the expression of α-SMA, collagens and several other extracellular matrix proteins in pancreatic stellate cells (PSCs) [32-34]. Previously, TGF-β has been shown to induce expression of ADAM12 in hepatic stellate cells [22], the hepatic equivalent of PSCs. Performing GSEA on the pancreatic cancer microarray dataset, we found a significant enrichment of a TGF-β
6
E F D p-value 0.0026 0.00 0.01 0.02 0.03 0.0 0.1 0.2 0.3 0.4 0.5 p-value <0.0001 0.00 0.01 0.02 0.03 0 1 2 3 G H 0.00 0.01 0.02 0.03 0.0 0.5 1.0 1.5 p-value <0.0001 0.00 0.01 0.02 0.03 0.00 0.01 0.02 0.03 0.04 0.05 p-value 0.0155 0.00 0.01 0.02 0.03 0 20 40 60 80scored tumor percentage
p-value 0.0202 A B C 4 6 8 10 9 10 11 12 13 14 p-value <0.0001 4 6 8 10 10 11 12 13 14 p-value <0.0001 4 6 8 10 10 11 12 13 p-value <0.0001 I J ES: 0.808 p-value: 0.0008 0.8 0.7 0.5 0.4 0.3 0.2 0.1 0.0
Enrichment score (ES)
Stroma
ADAM12 high ADAM12 low
0.6 ES: 0.679 p-value: 0.0008 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
Enrichment score (ES)
Extracellular matrix
ADAM12 high ADAM12 low
ADAM12 expression ADAM12 expression
ADAM12 expression ADAM12 expression ADAM12 expression ADAM12 expression (log2) ADAM12 expression (log2) ADAM12 expression (log2)
ACTA2 expression
ACTA2 expression (log2) COLA1A expression (log2) SPARC expression (log2)
COLA1A expression SPARC expression
KRT19 expression
Figure 2. ADAM12 expression correlates with stromal markers in PDAC patient tumors. A) Scatter plot
of ADAM12 in relation to ACTA2 expression in patient tumors (Pei microarray dataset), n=36. Linear regression analysis was used to determine p-value of correlation. B) As for panel A, correlating to COL1A1 and; C) SPARC. D-G) Association of ADAM12 expression with ACTA2, COL1A1, SPARC, and KRT19 in the AMC patient cohort measured by qPCR, n=15. F) Scatter plot of ADAM12 expression in the same patient tumors as for panels D-G in relation to percentage of tumor cell content on tissue used for RNA isolation, scored by a pathologist. I) Gene set enrichment analysis (GSEA) on the Pei dataset using gene ontology set ‘extracellular matrix’ as a gene signature. Patients are divided in either ADAM12 high or ADAM12 low expressing by median. J) GSEA as for panel I using a published pancreatic stromal signature by Binkley et al. ES, enrichment score, p= FDR q-value.
6
signaling pathway signature in patients with high ADAM12 levels (Figure 3A), suggesting a potential role of TGF-β signaling in the regulation of ADAM12 in pancreatic cancer.
To functionally address the impact of TGF-β on the regulation of ADAM12 expression in fibroblasts, we isolated and cultured primary CAFs from a PDX. Treatment of CAFs with TGF-β resulted in the marked upregulation of Adam12 transcript levels, and this effect could be blocked by co-administration of the TGF-β type I receptor inhibitor A83-01 [35] (Figure 3B). Co-culture of CAFs with tumor cells from the same PDX led to significant upregulation of Adam12 in the fibroblasts compared to (untreated) CAF monocultures, which could be blocked by treatment with A83-01 or the TGF-β blocking antibody 2G7 (Figure 3C), clearly showing that TGF-β ligand is present in these co-cultures and able to activate Adam12 expression in CAFs. Having established the association of ADAM12 expression with stromal quantity, we next proceeded to assess the use of ADAM12 as a biomarker in pancreatic cancer patients.
AdAM12 is upregulated in the serum of pancreatic cancer patients and associates with poor prognosis
Previous reports have described the use of ADAM12 as a serum marker during pregnancy to detect fetal genetic aberrations during the first trimester [36], and the increase in ADAM12 in the serum of small cell lung cancer patients has been shown to indicate disease progression [23]. To explore the use of the ADAM12 as a serum marker in pancreatic cancer patients, we determined the concentration of ADAM12 in the serum of 149 patients diagnosed with resectable or irresectable PDAC before therapeutic intervention. Serum of 38 healthy individuals was used as a control. PDAC patients had a significant upregulation of ADAM12 serum levels (median 372 ± 625 pg/ml IQR; interquartile range) compared to healthy volunteers (median 154 ± 169 pg/ml IQR; p-value <0.0001) (Figure 4A).
Figure 3. TGF-β regulates ADAM12 expression in cancer-associated fibroblasts. A) GSEA using a
KEGG-derived TGF-β pathway gene signature on patients with high or low ADAM12 expression levels. B) Mouse
Adam12 gene expression relative to mouse Gapdh measured by qPCR in CAFs isolated from a PDX treated
with 5 ng/ml TGF-β or TGF-β+ TGF-β receptor I inhibitor A83-01. n=3, **P<0.01, ***P<0.001. C) Expression
of Adam12 in CAFs co-cultured with primary human tumor cells from the same PDX, measured by qPCR.
Co-cultures were additionally treated with A83-01, isotype control antibody or TGF-β blocking antibody
2G7; n=3, **P<0.01, ***P<0.001. B C 0.00 0.05 0.10 0.15 ** *** 0.00 0.05 0.10 0.15 0.20 *** *** ** A ES: 0.453 p-value: 0.025 0.45 0.40 0.30 0.25 0.20 0.15 0.10 0.05
Enrichment score (ES)
KEGG: TGF-β signaling pathway
ADAM12 high ADAM12 low
0.35
0.00
mAdam12 expression mAdam12 expression
co-culture CTR CTR TGF-β TGF-β + A83-01 + anti-TGF-β Ab + A83-01 + ctr Ab
6
To investigate the association of ADAM12 serum levels with clinical parameters, we dichotomized using 206 pg/ml (first quartile) as a cut-off value. No statistically significant correlation of circulating ADAM12 levels was found with age, gender, tumor size, stage of disease, or resectability (Table 1, chi-square test). ADAM12 levels were also not found to correlate with age or tumor size when investigated as continuous variables (p=0.96 and p=0.11, respectively; Spearman’s rank test). When examining the serum levels in all the different stages directly, no significant differences were found between any of the stages (Figure 4B). However, when the impact of high ADAM12 serum levels on overall survival was analyzed by Kaplan–Meier survival curves and log-rank test, patients with elevated ADAM12 serum levels were found to show significantly worse prognosis than patients with low ADAM12 serum levels (Figure 4C, p=0.010). The median survival was 337 days (265-401, 95% CI) in patients with high ADAM12 and 842 days (estimated) in patients with low-levels. An exact median survival in the ADAM12 low group could not be calculated due to the majority of patients still being alive at the time of analysis. Univariate Cox proportional-hazard regression model showed that high stage, no resection of the primary tumor, and high ADAM12 serum levels were statistically significant poor prognostic factors for the whole PDAC cohort (Table 2A). The hazard ratio (HR) of high serum ADAM12 levels for mortality was 2.13 (1.2-3.8 95% CI ; p=0.011). In the multivariate Cox proportional-hazard regression model, only surgical resection remained a significant predictor of poor clinical outcome (Table 2B). Similar results were obtained for all the analysis by using median (372 pg/ml) as a cut-off value to split patients in ADAM12 high and low expressing group (data not shown).
As the full PDAC cohort is quite heterogeneous with regards to the treatment received, and surgical resection of the primary tumor was found to be the main predictor for outcome, we performed a subgroup analysis in the resected and unresected cohorts.
Table 1. Correlation between ADAM12 serum levels and clinicopathological parameters of PDAC patients
(n=149). Patients were divided into high and low ADAM12 serum level using 206 pg/ml as a cutoff value.
Factor n=149 total n
AdAM12 serum level
p-value low high Age <67 70 18 52 0.956 ≥67 79 20 59 Gender male 79 20 59 0.96 female 70 18 52 Stage I+II 59 19 40 0.129 III+IV 90 19 71 Resection no 89 18 71 0.072 yes 60 20 40 tumor size ≤2cm 14 5 9 0.32 >2cm 127 20 97
6
Figure 4. ADAM12 is upregulated in the serum of PDAC patients and predicts poor prognosis. A) Serum
ADAM12 levels measured by ELISA of healthy individuals (n=38) or patients diagnosed with PDAC (n=157). Graph shows scatter dot plots indicating median with interquartile range. ***P<0.0001; Mann-Whitney U-test. B) ADAM12 serum levels in PDAC patients divided by stage of disease. No significant difference was found between any of the stages using Kruskal-Wallis test. Indicated is median with interquartile range. C) Kaplan-Meier plot of all AMC PDAC patients divided into ADAM12 high or ADAM12 low by median serum levels of the total cohort. Log-rank test used to determine significance. D) Kaplan-Meier survival plot of the subgroup of resected patients in relation to high or low ADAM12 serum levels.
A B survival [days]600 800 100012001400 400 200 0 p = 0.010 ADAM12 high n=108 ADAM12 low n=37 C 1400 1200 1000 800 600 400 200 0 p = 0.035 ADAM12 high n=38 ADAM12 low n=19
all PDAC patients D resected PDAC patients
survival [days]
I (7) IIA (16) IIB (36) III (42) IV (48)
1 10 100 1000 10000 100000 se ru m A D AM 12 p g/ m l 1.0 0.8 0.6 0.4 0.2 0.0 Cum Survival 1.0 0.8 0.6 0.4 0.2 0.0 Cum Survival healthy (38) PDAC (149) 1 10 100 1000 10000 100000 *** se ru m A D AM 12 p g/ m l
We found no difference in survival in patient with irresectable disease (p=0.45, log-rank test). However, when analyzing the association of ADAM12 serum levels with survival in 58 patients undergoing resection, we found that patients with high circulating ADAM12 levels had reduced overall survival (Figure 4D; p=0.035), corroborating the prognostic value of ADAM12 RNA expression in the tumor of resected patients from an independent cohort (Figure 1B). Univariate Cox proportional-hazard regression model showed that
6
Table 2. Cox proportional hazard regression model. A) Univariate analysis for overall survival in PDAC
patients (n=145). B) Multivariate analysis with factors stage, resection and ADAM12 serum level.
A Factor Hr 95% Ci p-value Age <67 1 0.93-2.30 0.10 ≥67 1.46 Gender male 1 0.67-1.63 0.85 female 1.044 Stage I+II 1 1.51-4.30 0.00047 III+IV 2.545 Resection yes 1 2.17-6.83 0.000004 no 3.85
ADAM12 serum low 1 1.19-3.82 0.011
high 2.13 B Factor Hr 95% Ci p-value Stage I+II 1 0.53-2.17 0.852 III+IV 1.07 Resection yes 1 0.16-0.72 0.002 no 3.35
ADAM12 serum low 1 0.91-2.97 0.101
high 1.64
in our relatively small patient cohort, only high ADAM12 serum level was a significant prognostic factor, with a HR of 3.66 for overall survival (95% CI 1.01-13.2; p=0.048) whereas none of the other known factors for poor prognosis such as high stage, poor differentiation grade, tumor size >2cm, high lymph node ratio (LNR)[37] or non-radical resection were significant prognostic factors. This could be due to the limited number of patients analyzed and a short follow up time, but could also be suggestive of a very high prognostic value of ADAM12 in these patients.
disCUssiOn
One of the hallmarks of PDAC is a marked stromal fibroblast proliferation and deposition of ECM components in response to tumor derived stimuli, a phenomenon known as the ‘desmoplastic reaction’ that seems to support tumor growth and invasion [38]. In the present study we investigated ADAM12 expression in pancreatic cancer patients and PDX models and found expression typically in the stromal compartment. This was reflected in the correlation of ADAM12 expression with stromal activation and ECM markers ACTA2, COL1A1 and SPARC, as well as a high enrichment of genes associated with extracellular matrix and activated pancreatic stroma in tumors with high ADAM12 levels. These findings
6
point to ADAM12 as a potential serum-borne marker for stroma in pancreatic cancer patients. Reports in literature regarding the source of ADAM12 in different cancer types are not unambiguous, and sometimes contradictory. Peduto and colleagues showed the expression of ADAM12 by in situ hybridization to exits mainly in the stroma of commonly used murine models for breast, prostate, and intestinal cancer [21]. Another study in hepatocellular carcinoma revealed ADAM12 staining throughout the stroma, while tumor cells were negative [22]. However, studies performed in bladder and breast cancer patients showed expression in tumor cells rather than the adjacent stroma [19, 39], although additional staining in the tumor endothelial cells was observed in breast and ovarian cancer [11, 40]. It thus appears that the source of ADAM12 expression can differ greatly between malignancies. To address if ADAM12 could be used as a general stromal marker in other tumors with reported stromal activation, more in depth immunohistochemical analysis of patient tumor material is needed.
Furthermore, we also found that high expression of ADAM12 in pancreatic tumor tissue, and more specifically in CAFs, can be a result of activated TGF-β signaling. This fits with the well-described role of TGF-β in stromal activation during malignant progression. TGF-β mediated ADAM12 upregulation was also seen in co-culture experiments with tumor cells, and this could be blocked by coadministration of either a TGF-β receptor
Table 3. Univariate Cox proportional hazard regression model for overall survival in resected PDAC patients
(n=58). Factor Hr 95% Ci p-value Age <67 1 0.51-4.05 0.491 ≥67 1.44 Gender male 1 0.14-1.40 0.167 female 0.445
radical resection yes 1 0.75-5.91 0.156
no 2.11 Grade well/moderate 1 0.76-7.84 0.136 poor 2.44 Stage I+II 1 0.17-3.52 0.75 III+IV 0.783 tumor size ≤2cm 1 0.42-5.39 0.54 >2cm 1.496 LNR <0.18 1 0.39-3.08 0.87 >0.18 1.09
adjuvant therapy yes 1 0.95-9.79 0.06
no 3.056
ADAM12 serum low 1 1.01-13.2 0.048
6
inhibitor or a TGF-β blocking antibody, suggesting a paracrine mediated induction of ADAM12 expression. If this induction is directly caused by TGF-β ligands derived from the tumor cells, or other factors that would in turn induce production of TGF-β in the CAFs themselves leading to autocrine signaling pathway activation, remains to be determined.
Potential roles for stromal ADAM12 upregulated during malignant transformation and progression can be manifold: The well characterized function of ADAM12 as a sheddase of growth factor receptor ligands such as EGF and HB-EGF could provide a nurturing environment for tumor cells by enhancing proliferation and offering protection from apoptosis [41, 42]. Furthermore, ADAM12 was shown to participate in matrix remodeling by cleavage of extracellular matrix proteins gelatin, type IV collagen, and fibronectin [18], suggesting involvement of this metalloprotease in cancer invasion and metastasis formation. Transgenic expression of ADAM12 in the murine mammary tumor virus-PyMT breast cancer model accelerated tumor growth and increased aggressiveness [43]. Inversely, genetic deletion of ADAM12 in the same breast cancer model, or the W10 prostate cancer mouse model delayed tumor development and progression compared to littermate controls [21, 39]. These findings addressing both ‘loss of function’ and ‘gain of function’ of ADAM12 in different murine cancer models support an overall tumor promoting function of this metalloprotease. The notion of ADAM12 as a tumor promoting/ enhancing factor is reflected in the worse survival of high ADAM12-expressing pancreatic cancer patients, on the RNA level but also as measured in the serum of our own cohort. Further studies, similar to the ones performed for breast and prostate cancer, are still needed to elucidate the functional role of ADAM12 in pancreatic cancer pathobiology.
The observation from this study that ADAM12 levels are a prognostic factor in patients undergoing surgical resection of the primary tumor is promising. A study addressing the direct association with systemic ADAM12 levels and stromal content in tumor of resected pancreatic cancer patients is being conducted at the moment, and will more conclusively allow the link of ADAM12 with stromal status in these patients. Future prospective studies with more patients as well as longer follow up are needed to fully address the value of serum ADAM12 levels as an independent prognostic marker.
Furthermore, recent therapeutic efforts are focused on using the tumor-promoting effects of the abundant stroma as the Achilles’ heel of pancreatic cancer to develop more effective treatment combinations. Such stromal targeting strategies include agents such as Shh pathway inhibitors, agonistic CD40 antibodies, platelet derived growth factor receptor (PDGFR) inhibitors, hyaluronidase, and the SPARC-mediated cytotoxic agent nab-paclitaxel. Many of these compounds have entered clinical development for treatment of pancreatic cancer patients and especially nab-paclitaxel has shown promising results in a phase III clinical trial [9, 44]. For these treatments, ADAM12 could serve as a valuable serum-borne surrogate stromal marker in pancreatic cancer patients, either to select patients for treatment or to monitor response over the course of treatment in a non-invasive manner. Anecdotal support for such analyses comes from the measurement of serum ADAM12 levels in a PDAC patient with locally advanced disease, receiving four cycles of the combination regimen gemcitabine
6
plus nab-paclitaxel (Abraxane, Celgene). A nearly 3-fold reduction of systemic ADAM12 levels (628 pg/ml to 215 pg/ml) was observed after treatment, showing proof of principle. However, more nab-paclitaxel treated patients should of course be followed over the course of treatment to formally validate this observation.
In summary, we provide data that show a specifically stromal origin of ADAM12 expression in pancreatic cancer, and demonstrate that systemic levels of ADAM12 correlate with poor prognosis, warranting additional studies on ADAM12 as a biomarker with potentially both prognostic and predictive value.
MATeriAL And MeTHOds
pancreatic cancer patients and samples
Collection of the patient material was approved by the institute’s relevant ethics committee, and written informed consent was received from all participants. All PDAC patients treated in the Academic Medical Center Amsterdam (period 2011-2014) were diagnosed histopathologically using biopsies or surgical specimens. Surgical resection specimens were histologically inspected and processed according to national and international guidelines [45]. Microscopic assessment was performed by an experienced pathologist and the final diagnosis was set in accordance with the WHO Classification [46], and the pTNM classification of malignant tumors [47]. Clinicopathological data were obtained from medical records and included age, gender, tumor diameter, differentiation grade, lymph node ratio (positive lymph nodes/total number of lymph nodes examined), therapies received, and tumor-node-metastasis (TNM) staging. Blood samples for serum collection were obtained from patients undergoing resection perioperatively, or before the start of treatment in case of irresectable patients. Venous blood of 38 healthy individuals without any indication of malignancy was collected as a control group.
establishment of patient-derived xenografts and primary cell lines
Freshly excised tumor pieces (approx. 3x3x3 mm) originating from the primary tumor or liver metastasis were washed several times in PBS containing 10 µg/ml gentamycin (Lonza) and 1% penicillin-streptamycin. Pieces were grafted subcutaneously into the flank of immunocompromised NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice (JAX 005557) with Matrigel (BD). Animals were bred and maintained at the local animal facility according to the legislation and ethical approval was obtained for the establishment of patient derived xenografts (PDX). After outgrowth of patient tumor and reaching a size of approximately 500 mm3, PDX tumors were harvested and passaged, and/or used to establish in vitro cultures. Harvested xenografts were minced with sterile scalpel knifes, placed in 8% FBS containing IMDM with collagenase IV (0.5 mg/ml, Sigma) in a tube and incubated at 37°C for 60 min with vortexing every 10 min. The dissociated suspension was passed through a 70 µm cell strainer, washed with culture medium and was grown in IMDM containing 8% FBS and 50 µM β-mercaptoethanol. During the first 5-10 passages, cultures were
6
mixed containing islands of human epithelial cells and murine fibroblasts. This culture was used for the blocking antibody and chemical inhibition experiments. A culture without epithelial component, confirmed by flow cytometry using anti-EpCAM antibody staining (DAKO, F0860 at 1:100), was used for stimulation experiments.
Flow cytometry
Dissociated PDX tissue was washed in FACS buffer (PBS containing 1% FBS) by resuspension and centrifugation for 4 min at 1300 rpm in a chilled centrifuge. Antibodies were diluted in FACS buffer and incubated 30 min at 4°C. Concentrations used were: rabbit anti-ADAM12, 1:500 (Proteintech); anti-H2Db biotin, 1:100 (BD Biosciences); secondary APC-labeled anti-rabbit antibody (southern biotechnology) or streptavidin-Alexa488 conjugate (Invitrogen) were diluted 1:500. After washing cells were resuspended in FACS buffer containing 1ug/ml Propidium Iodide (PI) (Sigma) and acquired on a FACSCanto II (BD, Franklin Lakes, NJ). Data were analyzed with FlowJo 7 (Tree Star, Ashland, OR).
Treatment of primary cultures
Primary murine cancer associated fibroblasts (CAFs) were seeded in 12 well culture plates and upon reaching confluence, pre-starved overnight in 0.5% FBS containing medium and treated for 24h with 5 ng/ml recombinant human TGF-β 1 (Prepotech) in the presence or absence of 1 µM ALK4/5/7 inhibitor A83-01 (Tocris Bioscience). Co-cultures containing cancer cells and CAFs from the same PDX were grown in 12 well culture plates, pre-starved and treated additionally with 50 µg/ml of TGF-β neutralizing antibody 2G7 or mouse IgG2b isotype control antibody (R&D) for 24h before RNA isolation.
rnA isolation and quantitative real-time pCr
Small pieces of PDX tumor were homogenized using an Ultra-Turrax tissue homogenizer T8 (IKA-Werke, Germany) in 1ml of Trizol (Invitrogen). Primary cells were lysed in Trizol and RNA isolation was performed according to the manufacturer’s protocol. Snap frozen patient tumor samples were embedded in Tissue-Tek® OCT (Sakura, Fine Tech, Japan) and 30 slices of 20 µm thickness each were cut on a cryotome. Cut tissue slices were immersed in 1 ml of RNA-bee, homogenized and RNA isolation was performed according to manufacturer’s protocol. For tumor percentage scoring a 10 µm thick slide was kept before the tissue was cut in thicker slices, mounted on a microscope slide (Superfrost plus, Fisher), dried at room temperature and used for H&E staining. Scoring of tumor percentage was performed by an experienced pathologist. cDNA was synthesized using Superscript III (Invitrogen) and random primers (Invitrogen). Real-time quantitative RT-PCR was performed with SYBR green (Roche, Basel, Switzerland) on a Lightcycler LC480 II (Roche). Relative expression of genes was calculated using the comparative threshold cycle (Cp) method and values were normalized to reference gene Gapdh or Actb.
6
gene set enrichment Analysis (gseA)
Gene set enrichment analysis (GSEA, v2.0.14) software was downloaded from the Broad institute (http://www.broadinstitute.org/gsea) and used for analysis according to the author’s guidelines [48, 49]. The median ADAM12 expression was used to divide patients in high and low expressing group. Gene set for KEGG TGF-β signaling pathway and GO term ‘extracellular matrix’ were downloaded from the Molecular Signature Database (MSigDB; Version 4.0); the pancreatic stroma signature was published by Binkley et. al [31]. 1000 phenotype permutations were performed to determine significance of the enrichment score.
immunohistochemical stainings
Tumor tissue was fixed in 4% formalin prior to paraffin embedding. Sections of 5 µm were prepared on a microtome. Tissue sections were deparaffinized and antigen retrieval was performed using 0.5% pepsin solution in 5 mM HCl. Endogenous peroxidase activity was inhibited with 0.3% hydrogen peroxide in PBS, and unspecific staining was blocked using 5% normal goat serum (Abcam) for 20 min at room temperature. Primary anti-ADAM12 antibody (Proteintech) was diluted 1:100 in normal antibody diluent (KliniPath), applied on tissue sections and incubated overnight at 4°C in a humidified chamber. For amplification of signal Brightvision+ post antibody block (Immunologic) was used prior to the addition of the secondary antibody poly-HRP-anti-Ms/Rt/Rb IgG (Immunologic) both for 30 min at room temperature. Visualization was performed using Vector® NovaRED™ (Vector Labs) according to manufacturer’s protocol, counterstained with 30% haematoxylin and tissue sections were mounted with non-aqueous medium. Pictures were taken on an Olympus BX51 microscope equipped with a DP70 digital camera system.
gene name species forward primer 5’-3’ reverse primer 5’-3’
GAPDH human GAAGGTGAAGGTCGGAGTC TGGAAGATGGTGATGGGATT
ACTB human CAGAAGGATTCCTATGTGGGCGA TTCTCCATGTCGTCCCAGTTGGT
ADAM10 human TTCGATGCAAATCAACCAGA TTCCTTCCCTTGCACAGTCT
ADAM12 human TTTCCACCACCCTCTCAGAC GCCTCTGAAACTCTCGGTTG
ADAM17 human GGGAACATGAGGCAGTCTCT ACCGAATGCTGCTGGATATT
ACTA2 human CAAAGCCGGCCTTACAGAG AGCCCAGCCAAGCACTG
COL1A1 human CACACGTCTCGGTCATGGTA AAGAGGAAGGCCAAGTCGAG
SPARC human GAAAGAAGATCCAGGCCCTC CTTCAGACTGCCCGGAGA
KRT19 human CCTGGAGTTCTCAATGGTGG CTAGAGGTGAAGATCCGCGA
Gapdh mouse CTCATGACCACAGTCCATGC CACATTGGGGGTAGGAACAC
Adam10 mouse AAGATGGTGTTGCCGACAGT TGGTCCTCATGTGAGACTGC
Adam12 mouse GCTTTGGAGGAAGCACAGAC CGCATCAACGTCTTCCTTTT
6
AdAM12 eLisA
Serum from patient blood samples was collected after centrifuging for 10 min at room temperature at 1300 g and stored at -80ºC as serum fractions until analysis. ADAM12 levels in serum were determined using commercial human ADAM12 DuoSet ELISA kit (R&D systems), according to manufacturer’s recommendations. Briefly, after coating the 96-well plates (Nunc MaxiSorp, Greiner) with capture antibody overnight at room temperature, and blocking the plate with 1% BSA solution the following day, 80 µl of serum samples was added in duplicate. After incubation for two hours at room temperature and mild washing steps, samples were incubated with biotinylated detection antibody for additional two hour followed by a 20 min incubation step with horse-radish peroxidase-labeled streptavidin and development with tetramethylbenzidine substrate solution (TNB) for additional 20 min. Absorbance was measured at 450 nm and 570 nm with a microplate reader (BioTek Synergy) after addition of 1 M H2SO4 stop solution. For wavelength correction, the 570 nm readings were subtracted from the 450 nm values before further analysis. The serum concentration of protein was determined with 4-parametric logistic regression analysis using GraphPad Prism 5 software.
statistics
GraphPad Prism 5 software was used to perform linear regression analysis of gene expression in patient tumors and Student’s t-tests to compare differences of in vitro treatments and gene expression between tumor and normal tissue. SPSS statistics software package 22 (IBM) was used for chi-square testing, Spearman’s rank correlation, Cox proportional-hazard regression modelling, and Kaplan-Meier survival analysis using log-rank test. Patients who died within 30 days after operation were excluded from survival analysis. Each test was two-sided and p <0.05 was considered statistically significant.
Author contributions
H.D and M.F.B coordinated the project and wrote the manuscript. H.D produced, collected and analyzed the data and made figures. V.L.V performed RNA isolation of patient tumor samples. M.J.vd V and F.D provided tissue material and gave pathological advice. J.W.M. provided serum samples of irresectable patients. M.C.B provided blood and tissue specimen from patients undergoing surgery. L.B.v R collected clinical data from patients. T.v L established patient derived xenografts and collected serum samples. J.P.M and H.W.M.vL supervised the study and critically revised the manuscript.
6
reFerenCe LisT
1 Hidalgo M: Pancreatic cancer. N Engl J Med
2010, 362:1605-1617.
2 Ryan DP, Hong TS, Bardeesy N: Pancreatic
adenocarcinoma. N Engl J Med 2014, 371:1039-1049.
3 Oberstein PE, Olive KP: Pancreatic cancer: why
is it so hard to treat? Therapeutic advances in
gastroenterology 2013, 6:321-337.
4 Mahadevan D, Von Hoff DD: Tumor-stroma
interactions in pancreatic ductal adenocarcinoma.
Mol Cancer Ther 2007, 6:1186-1197.
5 Whatcott C, Han H, Posner RG, Von Hoff DD:
Tumor-stromal interactions in pancreatic cancer.
Crit Rev Oncog 2013, 18:135-151.
6 Yardley DA: nab-Paclitaxel mechanisms of action
and delivery. Journal of controlled release:
official journal of the Controlled Release Society
2013, 170:365-372.
7 Desai N, Trieu V, Damascelli B, Soon-Shiong P:
SPARC Expression Correlates with Tumor Response to Albumin-Bound Paclitaxel in Head and Neck Cancer Patients. Transl Oncol 2009, 2:59-64.
8 Alvarez R, Musteanu M, Garcia-Garcia E,
Lopez-Casas PP, Megias D, Guerra C, Munoz M, Quijano Y, Cubillo A, Rodriguez-Pascual J, et al: Stromal disrupting effects of nab-paclitaxel in pancreatic cancer. Br J Cancer 2013, 109:926-933.
9 Von Hoff DD, Ervin T, Arena FP, Chiorean EG,
Infante J, Moore M, Seay T, Tjulandin SA, Ma WW, Saleh MN, et al: Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N
Engl J Med 2013, 369:1691-1703.
10 Kveiborg M, Albrechtsen R, Couchman JR, Wewer UM: Cellular roles of ADAM12 in health and disease. The international journal
of biochemistry & cell biology 2008,
40:1685-1702.
11 Frohlich C, Klitgaard M, Noer JB, Kotzsch A, Nehammer C, Kronqvist P, Berthelsen J, Blobel C, Kveiborg M, Albrechtsen R, Wewer UM: ADAM12 is expressed in the tumour vasculature and mediates ectodomain shedding of several membrane-anchored endothelial proteins.
Biochem J 2013, 452:97-109.
12 Horiuchi K, Le Gall S, Schulte M, Yamaguchi T, Reiss K, Murphy G, Toyama Y, Hartmann D, Saftig P, Blobel CP: Substrate selectivity of epidermal growth factor-receptor ligand sheddases and their regulation by phorbol esters and calcium influx. Mol Biol Cell 2007, 18:176-188.
13 Asakura M, Kitakaze M, Takashima S, Liao Y, Ishikura F, Yoshinaka T, Ohmoto H, Node K, Yoshino K, Ishiguro H, et al: Cardiac hypertrophy is inhibited by antagonism of ADAM12 processing of HB-EGF: metalloproteinase inhibitors as a new therapy. Nat Med 2002, 8:35-40.
14 Dyczynska E, Sun D, Yi H, Sehara-Fujisawa A, Blobel CP, Zolkiewska A: Proteolytic processing of delta-like 1 by ADAM proteases. J Biol Chem 2007, 282:436-444.
15 Damhofer H, Veenstra VL, Tol JA, van Laarhoven HW, Medema JP, Bijlsma MF: Blocking Hedgehog release from pancreatic cancer cells increases paracrine signaling potency. J Cell Sci 2014. 16 Ohlig S, Farshi P, Pickhinke U, van den Boom J,
Hoing S, Jakuschev S, Hoffmann D, Dreier R, Scholer HR, Dierker T, et al: Sonic hedgehog shedding results in functional activation of the solubilized protein. Dev Cell 2011, 20:764-774. 17 Kodama T, Ikeda E, Okada A, Ohtsuka T, Shimoda M, Shiomi T, Yoshida K, Nakada M, Ohuchi E, Okada Y: ADAM12 is selectively overexpressed in human glioblastomas and is associated with glioblastoma cell proliferation and shedding of heparin-binding epidermal growth factor. Am J
Pathol 2004, 165:1743-1753.
18 Roy R, Wewer UM, Zurakowski D, Pories SE, Moses MA: ADAM 12 cleaves extracellular matrix proteins and correlates with cancer status and stage. J Biol Chem 2004, 279:51323-51330.
19 Frohlich C, Albrechtsen R, Dyrskjot L, Rudkjaer L, Orntoft TF, Wewer UM: Molecular profiling of ADAM12 in human bladder cancer. Clin Cancer
Res 2006, 12:7359-7368.
20 Rocks N, Paulissen G, Quesada Calvo F, Polette M, Gueders M, Munaut C, Foidart JM, Noel A, Birembaut P, Cataldo D: Expression of a disintegrin and metalloprotease (ADAM and ADAMTS) enzymes in human non-small-cell lung carcinomas (NSCLC). Br J Cancer 2006, 94:724-730.
21 Peduto L, Reuter VE, Sehara-Fujisawa A, Shaffer DR, Scher HI, Blobel CP: ADAM12 is highly expressed in carcinoma-associated stroma and is required for mouse prostate tumor progression.
Oncogene 2006, 25:5462-5466.
22 Le Pabic H, Bonnier D, Wewer UM, Coutand A, Musso O, Baffet G, Clement B, Theret N: ADAM12 in human liver cancers: TGF-beta-regulated expression in stellate cells is
6
associated with matrix remodeling. Hepatology 2003, 37:1056-1066.
23 Shao S, Li Z, Gao W, Yu G, Liu D, Pan F: ADAM-12 as a diagnostic marker for the proliferation, migration and invasion in patients with small cell lung cancer. PLoS One 2014, 9:e85936. 24 Yu J, Walter K, Omura N, Hong SM, Young A,
Li A, Vincent A, Goggins M: Unlike pancreatic cancer cells pancreatic cancer associated fibroblasts display minimal gene induction after 5-aza-2’-deoxycytidine. PLoS One 2012, 7:e43456.
25 Badea L, Herlea V, Dima SO, Dumitrascu T, Popescu I: Combined gene expression analysis of whole-tissue and microdissected pancreatic ductal adenocarcinoma identifies genes specifically overexpressed in tumor epithelia.
Hepatogastroenterology 2008, 55:2016-2027.
26 Pei H, Li L, Fridley BL, Jenkins GD, Kalari KR, Lingle W, Petersen G, Lou Z, Wang L: FKBP51 affects cancer cell response to chemotherapy by negatively regulating Akt. Cancer Cell 2009, 16:259-266.
27 Damhofer H, Ebbing EA, Steins A, Welling L, Tol JA, Krishnadath KK, van Leusden T, van de Vijver MJ, Besselink MG, Busch OR, et al: Establishment of patient-derived xenograft models and cell lines for malignancies of the upper gastrointestinal tract manuscript in preparation 2015.
28 Saftig P, Reiss K: The “A Disintegrin And Metalloproteases” ADAM10 and ADAM17: novel drug targets with therapeutic potential? European
journal of cell biology 2011, 90:527-535.
29 Shintani Y, Hollingsworth MA, Wheelock MJ, Johnson KR: Collagen I promotes metastasis in pancreatic cancer by activating c-Jun NH(2)-terminal kinase 1 and up-regulating N-cadherin expression. Cancer Res 2006, 66:11745-11753. 30 Bachem MG, Schunemann M, Ramadani M,
Siech M, Beger H, Buck A, Zhou S, Schmid-Kotsas A, Adler G: Pancreatic carcinoma cells induce fibrosis by stimulating proliferation and matrix synthesis of stellate cells. Gastroenterology 2005, 128:907-921.
31 Binkley CE, Zhang L, Greenson JK, Giordano TJ, Kuick R, Misek D, Hanash S, Logsdon CD, Simeone DM: The molecular basis of pancreatic fibrosis: common stromal gene expression in chronic pancreatitis and pancreatic adenocarcinoma. Pancreas 2004, 29:254-263. 32 Truty MJ, Urrutia R: Basics of TGF-beta and
pancreatic cancer. Pancreatology 2007, 7:423-435.
33 Apte MV, Haber PS, Darby SJ, Rodgers SC, McCaughan GW, Korsten MA, Pirola RC, Wilson JS: Pancreatic stellate cells are activated by proinflammatory cytokines: implications for pancreatic fibrogenesis. Gut 1999, 44:534-541. 34 Kordes C, Brookmann S, Haussinger D,
Klonowski-Stumpe H: Differential and synergistic effects of platelet-derived growth factor-BB and transforming growth factor-beta1 on activated pancreatic stellate cells. Pancreas 2005, 31:156-167.
35 Tojo M, Hamashima Y, Hanyu A, Kajimoto T, Saitoh M, Miyazono K, Node M, Imamura T: The ALK-5 inhibitor A-83-01 inhibits Smad signaling and epithelial-to-mesenchymal transition by transforming growth factor-beta. Cancer Sci 2005, 96:791-800.
36 Laigaard J, Christiansen M, Frohlich C, Pedersen BN, Ottesen B, Wewer UM: The level of ADAM12-S in maternal serum is an early first-trimester marker of fetal trisomy 18. Prenat
Diagn 2005, 25:45-46.
37 Tol JA, Brosens LA, van Dieren S, van Gulik TM, Busch OR, Besselink MG, Gouma DJ: Impact of lymph node ratio on survival in patients with pancreatic and periampullary cancer. The British
journal of surgery 2014.
38 Bardeesy N, DePinho RA: Pancreatic cancer biology and genetics. Nat Rev Cancer 2002, 2:897-909.
39 Frohlich C, Nehammer C, Albrechtsen R, Kronqvist P, Kveiborg M, Sehara-Fujisawa A, Mercurio AM, Wewer UM: ADAM12 produced by tumor cells rather than stromal cells accelerates breast tumor progression. Mol Cancer Res 2011, 9:1449-1461. 40 Sasaroli D, Gimotty PA, Pathak HB, Hammond R,
Kougioumtzidou E, Katsaros D, Buckanovich R, Devarajan K, Sandaltzopoulos R, Godwin AK, et al: Novel surface targets and serum biomarkers from the ovarian cancer vasculature. Cancer Biol
Ther 2011, 12:169-180.
41 Ardito CM, Gruner BM, Takeuchi KK, Lubeseder-Martellato C, Teichmann N, Mazur PK, Delgiorno KE, Carpenter ES, Halbrook CJ, Hall JC, et al: EGF receptor is required for KRAS-induced pancreatic tumorigenesis. Cancer Cell 2012, 22:304-317.
42 Yamanaka Y, Friess H, Kobrin MS, Buchler M, Beger HG, Korc M: Coexpression of epidermal growth factor receptor and ligands in human pancreatic cancer is associated with enhanced tumor aggressiveness. Anticancer Res 1993, 13:565-569.
6
43 Kveiborg M, Frohlich C, Albrechtsen R, Tischler V, Dietrich N, Holck P, Kronqvist P, Rank F, Mercurio AM, Wewer UM: A role for ADAM12 in breast tumor progression and stromal cell apoptosis. Cancer Res 2005, 65:4754-4761. 44 Heinemann V, Reni M, Ychou M, Richel DJ,
Macarulla T, Ducreux M: Tumour-stroma interactions in pancreatic ductal adenocarcinoma: rationale and current evidence for new therapeutic strategies. Cancer Treat Rev 2014, 40:118-128. 45 Westra WH: Surgical pathology dissection : an
illustrated guide. 2nd edn. New York: Springer;
2003.
46 Bosman FT, World Health Organization., International Agency for Research on Cancer.:
WHO classification of tumours of the digestive system. 4th edn. Lyon: International Agency for
Research on Cancer; 2010.
47 Sobin LH, Gospodarowicz MK, Wittekind C, International Union against Cancer.: TNM
classification of malignant tumours. 7th edn.
Chichester, West Sussex, UK ; Hoboken, NJ: Wiley-Blackwell; 2010.
48 Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, Mesirov JP: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005, 102:15545-15550.
49 Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J, Puigserver P, Carlsson E, Ridderstrale M, Laurila E, et al: PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet 2003, 34:267-273.
6
sUppLeMenTArY inFOrMATiOn
p-value 0.0056 E F I Badea Perez (ICGC) 7.2 7.4 7.6 7.8 8.0 8.2 8.4 10 12 14 16 7.2 7.4 7.6 7.8 8.0 8.2 8.4 8 10 12 14 16 7.2 7.4 7.6 7.8 8.0 8.2 8.4 11 12 13 14 15 16 p-value 0.0045 p-value 0.061 A B C 4 6 8 10 10 11 12 13 14 p-value 0.0021 4 6 8 10 10 11 12 13 14 4 6 8 10 11 12 13 p-value 0.0004 p-value 0.0008 D H ES: 0.617 p-value: 0.037 ES: 0.710 p-value: 0.029 0.5 0.4 0.3 0.2 0.1 0.0Enrichment score (ES)
Stroma
ADAM12 high ADAM12 low
0.6 0.7 0.5 0.4 0.3 0.2 0.1 0.0
Enrichment score (ES)
Stroma
ADAM12 high ADAM12 low
0.6
ACTA2 expression (log2)
ADAM12 expression (log2) ADAM12 expression (log2) ADAM12 expression (log2)
ADAM12 expression (log2) ADAM12 expression (log2)
ADAM12 expression (log2)
ACTA2 expression (log2) COL1A1 expression (log2)
COL1A1 expression (log2)
SPARC expression (log2)
SPARC expression (log2)
10
supplementary Figure. ADAM12 expression correlates with stromal markers in PDAC patient tumors.
Scatter plot of ADAM12 in relation to ACTA2 (A), COL1A1 (B), and SPARC (C) expression in patient tumors (Badea microarray dataset), n=36. Linear regression analysis was used to determine p-value of correlation. D) Gene set enrichment analysis (GSEA) on the Badea dataset using a published pancreatic stromal signature by Binkley et al. Patients are divided in either ADAM12 high or ADAM12 low expressing by median. E-H) Same as for panel A-C but on the Perez et al. (ICGC) pancreatic cancer dataset (GSE36924), n=91. I) GSEA as in panel (D) on the Perez (ICGC) dataset. ES, enrichment score, p= FDR q-value.
