Patient cohorts

The use and collection of all clinical and pathologic data and specimens were performed with approval and in accordance with the guidelines established by the University of Cincinnati Institutional Review Board (IRB 2019–0324).

University of cincinnati medical center cohort surgery-first cohort

Consecutive patients (n = 23) with resectable pancreatic ductal adenocarcinoma (PDAC) undergoing a surgery-first approach from 2014–2019 with available paraffin-embedded tissue blocks were included for analysis. Only patients undergoing a surgery first approach were included for analysis in order to evaluate inherent baseline expression of acid sphingomyelinase within the tumor. Patient charts were reviewed and patient demographic, clinical, operative, and pathologic data were collected.

Human protein atlas cohort

The human protein atlas (proteinatlas.org) is a publicly available data source licensed under the Creative Commons Attribution-ShareAlike 3.0 International License that includes a pathologic atlas of the human cancer transcriptome [25]. The database was queried for mRNA expression of the sphingomyelin phosphodiesterase 1 (SMPD1) gene in patients with pancreatic ductal adenocarcinoma. SMPD1 expression levels and clinical characteristics of 172 patients were collected which included survival and cancer stage. Stage 4 patients were excluded from the cohort for analysis.

The cancer genome atlas and clinical proteomic tumor analysis consortium cohort

The Cancer Genome Atlas (TCGA) and Genotype-tissue expression (GTEx) TCGA is a freely web-based accessible database, which collects NGS data from more than 10,000 tumors across 33 cancer types until 2018 [26]. Gene expression and clinical data were taken into consideration in the present study. Genotype-tissue expression (GTEx) GTEx provides publicly available gene expression data from 53 normal tissue sites across nearly 1000 people by RNA sequencing [27]. The Clinical Proteomic Tumor Analysis Consortium (CPTAC) dataset, publicly available since 2021, is based on whole-genome sequencing (WGS) and whole-exome sequencing (WES) of 140 pancreatic cancers with 67 normal adjacent tissues giving rise to proteogenomic characterization of pancreatic ductal adenocarcinoma [28]. The cBioPortal for Cancer Genomics (http://www.cbioportal.org), GEPIA (gene expression profiling interactive analysis) (http://gepia.cancerpku.cn/), and the UALCAN database (http://ualcan.path.uab.edu) analysis tools were employed.

University of cincinnati neoadjuvant cohort

Consecutive patients (n = 94) with biopsy-proven pancreatic ductal adenocarcinoma that underwent neoadjuvant chemotherapy followed by surgical resection from 2012–2022 were included for analysis. Clinicopathologic data were collected from our institutional database. Medical reconciliation records were reviewed prior to, during, and at the completion of neoadjuvant chemotherapy for all medications. Patients were then categorized based on the use of functional inhibitors of acid sphingomyelinase (FIASM) during treatment [20, 29]. Pathologic response in surgical specimens was graded based on the proportion of viable tumor according to the College of American Pathologists grading system. Grades 0 (complete histologic response, no viable cancer cells) and 1 (near complete response, single cells or rare small groups of cancer cells) specimens were categorized as a good pathologic response while grades 2 (partial response with residual cancer) and 3 (poor or no response with extensive residual cancer) were categorized as a poor pathologic response.

Immunohistochemical Staining

Acid sphingomyelinase staining protocol: Tissue slides were deparaffinized and rehydrated in xylene substitute and gradients of ethanol after heating for 1 h at 65 °C. Permeabilization was done with 1% Triton X-100 for 15 min on a shaker plate and then washed with PBS-Tween buffer. Antigen retrieval was done with Tris–EDTA, pH 9.0 buffer by using Thermo-Electron Corporation Shandon Tissue Wave 2, HIER1 program. Slides were cooled to room temperature, washed with PBS-Tween buffer, and then treated with 3% H2O2 for 1 h at 37 °C to block endogenous peroxidase. Specimens were then incubated with rabbit primary antibody, anti-Smpd1 (Proteintech, Cat# 14,609–1-AP) at 1:200 dilution at 4 °C overnight. The sections were extensively washed with PBS-Tween buffer and then incubated with biotinylated goat anti-rabbit secondary antibody at room temperature for 30 min. The specimens were washed with PBS-Tween buffer and then incubated with Streptavidin-HRP pre-diluted (SAV-HRP, # BD550946, ready to use) for 30 min at room temperature. The specimens were then treated with diaminobenzidine DAB (HRP substrate, ThermoScientific, TA-125-QHDX) for about 1 min and the reaction stopped by transfer to distilled water. The specimens were then counterstained with hematoxylin. Sections were mounted onto slides with Permount (SP15-500, Thermo Fisher Scientific, USA) for review.

Evaluation of immunohistochemical staining in pathologic specimens

All slides were reviewed and scored by a pathologist (JW) blinded to the identity of the samples. The pathologist is an expert in pancreatic cancer. Only staining within the invasive tumor component was considered when grading the staining. The slides were then scored on an ordinal scale based on the intensity of acid sphingomyelinase staining: 0 = negative or no staining, 1 +  = weak staining, 2 +  = intermediate/moderate staining, and 3 +  = strong staining intensity. Patients were then categorized as ASM low expression (staining intensity of 0 or 1 +) or ASM high expression (staining intensity of 2 + or 3 +) for comparison.

Orthotopic pancreatic cancer murine model

All animal experiments were approved by the University of Cincinnati Ethic Committee and the Institutional Animal Care and Use Committee. Eight-week-old, wild-type male, C57BL/6 J mice were purchased from Jackson Labs (000,664, Jackson Labs, USA). Acid sphingomyelinase-deficient mice (Smpd1−/−) were bred and genotyped at the animal facility of the University of Cincinnati. We used young acid sphingomyelinase-deficient mice (maximum age 10 weeks) to avoid sphingomyelin accumulation. Mice were anesthetized using 120 mg/kg ketamine plus 20 mg/kg xylazine. Orthotopic injection was performed as described by Tepal et al. [30]. In detail, a left subcostal incision was made just below the rib cage and the pancreas was identified. The tumor cell suspension was created by mixing 25 µL of Matrigel with 25 µL of Pan02 cells (National Cancer Institute-Frederick Cancer Research and Development Center, Frederick, MD, USA) containing 1 × 106 cells. Pan02 cells were cultured in DMEM + 10% FBS medium, under 37 ℃ and 5% CO2; no antibiotic was added. Cells were washed, trypsinized, and washed in PBS at least 3 times prior to injection. The tumor suspension was slowly injected into the pancreas and the needle left in place for 60 s to allow the Matrigel to set. After ensuring hemostasis, the abdomen was closed in 2 layers using 3–0 silk suture. Wild-type mice were then randomly divided into 2 groups. One group was treated with 10 mg/kg of the antidepressant and functional inhibitor of the acid sphingomyelinase amitriptyline or vehicle, i.e., 0.9% NaCl. Amitriptyline or 0.9% NaCl was injected into the peritoneal cavity every other day. The size of the pancreas cancer was determined in all groups 15 days after tumor initiation.

Flank tumor model

A total of 25,000 KPC tumor cells were injected subcutaneously into the flank of wild-type or acid sphingomyelinase-deficient C57BL/6 mice. Tumor growth was monitored daily using calipers. Mice were sacrificed at day 14 after tumor cell injection; the tumors were removed and shock frozen in liquid nitrogen for further analysis.

Transfection of siRNA targeting acid sphingomyelinase

KPC (20 × 106 cells) were incubated with commercial siRNA targeting acid sphingomyelinase (Santa Cruz Inc., # sc-41651-SH) or a control construct for 15 min on ice. Cells were then electroporated at 500 V, 5 impulses with each 3-ms duration, left on ice for 15 mice, transferred into 10 mL MEM supplemented with 10 mM HEPES (pH 7.4; Carl Roth GmbH, Karlsruhe, Germany), 2 mM l-glutamine, 1 mM sodium pyruvate, 100 μM nonessential amino acids, 100 U/mL penicillin, 100 μg/mL streptomycin (all from Invitrogen), and 10% FCS (PAA Laboratories GmbH, Coelbe, Germany) and grown for 2 days. Dead cells were removed and positive cells were selected using 3.5 μg/mL puromycin. Stable transfected cells were used as bulk cultures. The activity of the acid sphingomyelinase in transfected cells was determined using a biochemical assay. To this end, cells were lysed in 250 mM sodium acetate (pH 5.0) and 0.2% NP40 for 5 min, removed from the plate and 270 μL of the lysates were incubated with 30 μL of 0.05 μCi [14C]sphingomyelin (52 mCi/mmol; ARC). Prior to addition, the substrate [14C]sphingomyelin was dried for 10 min in a SpeedVac, resuspended in 250 mM sodium acetate (pH 5.0) and 0.1% NP40 and sonicated for 10 min in a bath sonicator. The samples were incubated for 30 min at 37 °C with shaking at 300 rpm. The enzymatic reaction was terminated by extraction in 4 volumes of CHCl3:CH3OH (2:1, v/v), the samples were centrifuged, and an aliquot of the upper aqueous phase was scintillation-counted to determine the release of [14C]phosphorylcholine from [14C]sphingomyelin.

Immunohistochemistry of tumor sections for acid sphingomyelinase and CD45

Tumor tissue sections were prepared as above, dewaxed, and rehydrated. To retrieve the antigens, sections were treated with pepsin (Digest All; #003,009, Invitrogen) for 30 min at 37 °C, washed and blocked for 10 min at room temperature with PBS, supplemented with 5% fetal calf serum (FCS). Samples were washed once in PBS and stained with polyclonal anti-human acid sphingomyelinase (1:100 dilution in PBS + 1% FCS, #AF5348, R&D Systems) for 45 min, washed 3 times in PBS plus 0.05% Tween 20 and once with PBS. Samples were then stained with FITC-F(ab)2 fragments against goat IgG (1:500 dilution, Jackson Immunoresearch, #705–096-147) for 30 min in PBS + 1% FCS and washed again 3 times in PBS plus 0.05% Tween 20 and once with PBS. Finally, the samples were stained with PE-coupled mouse anti-human CD45-antibodies (clone 2D1; 1:500 dilution, R&D Systems, FAB1430P) for 30 min in PBS + 1% FCS. Specimens were washed again 3 times in PBS plus 0.05% Tween 20 and embedded in Mowiol. Samples were evaluated by confocal microscopy on a Leica TCS-SL confocal microscope equipped with a 40 × lens, and images were analyzed with Leica LCS software (Leica Microsystems, Mannheim, Germany). All comparative samples were measured at identical settings. Control stainings were performed with FITC-labeled secondary antibodies only or with irrelevant goat IgG followed by staining FITC-labeled FITC-F(ab)2 anti-goat IgG. These controls revealed very weak stainings and confirmed the specificity of the antibody stainings. Fluorescence intensities were quantified using Image J.

Ceramide and sphingosine measurements in isolated pancreatic ductal adenocarcinoma

Tumors were isolated, homogenized in H20 and extracted in CHCl3/CH3OH/1N HCl (100:200:1, v/v/v). The lower phase was dried and resuspended in a detergent solution consisting of 7.5% (w/v) n-octyl glucopyranoside, 5 mM cardiolipin in 1 mM diethylenetriamine-pentaacetic acid. The kinase reaction was initiated by addition of 0.001 units sphingosine kinase in 50 mM HEPES (pH 7.4), 250 mM NaCl, 30 mM MgCl2 10 μM ATP and 10 μCi [32P]γATP or 10 μL diacylglycerol (DAG) kinase (GE Healthcare Europe, Munich, Germany), 0.1 M imidazole/HCl (pH 6.6), 0.2 mM DTPA, 70 mM NaCl, 17 mM MgCl2, 1.4 mM ethylene glycol tetraacetic acid, 2 mM dithiothreitol, 1 μM adenosine triphosphate (ATP), and 5 μCi [32P]γATP. Samples were incubated for 30 min at 37 °C with shaking (350 rpm) and the kinase reaction was terminated by extraction of lipids. The sphingosine kinase reaction was terminated by addition of 20 μL 1N HCl, 800 μL CHCl3:CH3OH:1N HCl (100:200:1, v/v/v), and 240 μL each of CHCl3 and 2 M KCl. Phases were separated; the lower phase was collected, dried, dissolved in 20 μL CHCl3:CH3OH (1:1, v/v), and separated on Silica G60 TLC plates with CHCl3:CH3OH:acetic acid:H2O (90:90:15:5, v/v/v/v) as developing solvent. The ceramide kinase reaction was terminated by extraction with 1000 μL CHCl3:CH3OH:1N HCl (100:100:1, v/v/v), 170 μL buffered saline solution (135 mM NaCl, 1.5 mM CaCl2, 0.5 mM MgCl2, 5.6 mM glucose, 10 mM HEPES; pH 7.2) and 30 μL of 100 mM EDTA, lipids were extracted, separated by Silica G60 TLC plates with chloroform/acetone/ methanol/acetic acid/H2O (50:20:15:10:5, v/v/v/v/v), as developing solvent. The TLC plates were analyzed with a phosphorimager. Sphingosine and ceramide levels were determined with a standard curve of C18-sphingosine or C16 and C24 ceramide.

Statistics

Clinicopathologic data were obtained from the electronic medical records and our institutional prospectively maintained pancreatic cancer database. Variables include patient age, gender, histologic grade, lymphovascular invasion, perineural invasion, stage, lymph node status, and oncologic status. Statistical analyses were performed using JMP Pro v15 (SAS, Cary, NC) and SigmaPlot 14.5 (Systat Software, Inc., Palo Alto, CA). Data were classified as categorical and continuous variables. Categorical variables were described with counts and percent with comparison made with Pearson’s chi-squared and Fisher’s exact test, as appropriate. Continuous variables were examined using Student’s t-test. Survival analyses were performed using Kaplan–Meier methodology with log-rank test for group comparison. p-values < 0.05 were considered statistically significant.