Alpha-1 Antitrypsin Inhibits Tumorigenesis and Progression of Colitis-Associated Colon Cancer through Suppression of Inflammatory Neutrophil-Activated Serine Proteases and IGFBP-3 Proteolysis

Abstract

Colitis-associated colon cancer (CAC) accompanies the massive infiltration of neutrophils during tumorigenesis and progression of CAC. Depletion of neutrophils in circulation results in significant inhibition of tumor incidence in CAC. However, the underlying mechanisms are largely unclear. In this study, we provide evidence for the crucial involvement of inflammatory neutrophil-activated serine proteases (NSPs) on the dysregulation of the anti-inflammatory and antitumor IGFBP-3/IGFBP-3R signaling axis in CAC using a chronic AOM/DSS mouse model. We also provide preclinical evidence for α1-antitrypsin (AAT) as a preventive and as a therapeutic for CAC. AAT administration not only prevented colitis-associated tumorigenesis but also inhibited established CAC. AOM/DSS treatment results in the significant activation of NSPs, leading to CAC through increased pro-inflammatory cytokines and decreased anti-inflammatory and antitumor IGFBP-3. Collectively, these data suggest that the NSPs proteolyze IGFBP-3, whereas AAT inhibits chronic colonic inflammation-induced NSP activity and subsequently suppresses IGFBP-3 proteolysis. Therefore, the anti-inflammatory and antitumor functions of the IGFBP-3/IGFBP-3R axis are restored. AAT mimicking small peptides also showed their inhibitory effects on NSP-induced IGFBP-3 proteolysis. These results suggest that targeting the NSP-IGFBP-3/IGFBP-3R axis using NSP inhibitors such as AAT and the AAT mimics and IGFBP-3R agonists could lead to novel approaches for the prevention and treatment of CAC.

Keywords: AAT; IGFBP-3; IGFBP-3R; NSPs; anti-inflammation; antitumor; colitis-associated colon cancer; neutrophil.

Figure 1

Preventive effects of AAT on CAC development. (A) Schematic overview of preventive protocol for AAT administration in AOM/DSS-treated mouse. Mice were injected with AOM (day 2 and day 65) followed by four cycles of 2.5% DSS in drinking water. ¹⁸F-FDG PET/CT image and colon tumors were analyzed on day 91. (B) A representative 3D Maximum Intensity Projections of AOM/4cycle DSS (total 91 days) treated mice with/without AAT administration. Four regions within the yellow circle show high FDG uptake and are prominent in the colon of no AAT administration (left) while showing only one region with low FDG uptake in AAT administrated colon (right) in AOM/DSS-treated mouse. Gamma counting of the entire colon was 8.8% ID/g (left) and 5.6 %ID/g (right). FDG %ID/g = FDG uptake percent injected dose per gram of tissue. (C) Colons were removed and examined for macroscopic changes. Tumors were confined to the middle and distal colon. Tumor multiplicity (D), tumor distribution (E), and average tumor load (F) were determined. Results, mean ± SD [n = 6 (−AAT), n = 7 (+AAT); **, p < 0.001]. (G,H) Anti-inflammatory and antitumor effects of AAT during initiation and progression of CAC were analyzed after treatment of AOM/2cycle DSS (total 47 days) (G) as well as AOM/4cycle DSS (total 91 days) (H). Colon tissue was further fixed and stained with hematoxylin/eosin. top, Control colon tissue without AAT treatment. Bottom, AAT treated colon tissue. In PBS-treated control colon tissue, the mucosa is expanded by glandular hyperplasia, mixed inflammation, and TIS formation. The submucosa is also expanded by inflammation. Original magnification, ×4, ×10 and ×20.

Figure 1

Preventive effects of AAT on CAC development. (A) Schematic overview of preventive protocol for AAT administration in AOM/DSS-treated mouse. Mice were injected with AOM (day 2 and day 65) followed by four cycles of 2.5% DSS in drinking water. ¹⁸F-FDG PET/CT image and colon tumors were analyzed on day 91. (B) A representative 3D Maximum Intensity Projections of AOM/4cycle DSS (total 91 days) treated mice with/without AAT administration. Four regions within the yellow circle show high FDG uptake and are prominent in the colon of no AAT administration (left) while showing only one region with low FDG uptake in AAT administrated colon (right) in AOM/DSS-treated mouse. Gamma counting of the entire colon was 8.8% ID/g (left) and 5.6 %ID/g (right). FDG %ID/g = FDG uptake percent injected dose per gram of tissue. (C) Colons were removed and examined for macroscopic changes. Tumors were confined to the middle and distal colon. Tumor multiplicity (D), tumor distribution (E), and average tumor load (F) were determined. Results, mean ± SD [n = 6 (−AAT), n = 7 (+AAT); **, p < 0.001]. (G,H) Anti-inflammatory and antitumor effects of AAT during initiation and progression of CAC were analyzed after treatment of AOM/2cycle DSS (total 47 days) (G) as well as AOM/4cycle DSS (total 91 days) (H). Colon tissue was further fixed and stained with hematoxylin/eosin. top, Control colon tissue without AAT treatment. Bottom, AAT treated colon tissue. In PBS-treated control colon tissue, the mucosa is expanded by glandular hyperplasia, mixed inflammation, and TIS formation. The submucosa is also expanded by inflammation. Original magnification, ×4, ×10 and ×20.

Figure 2

Therapeutic effects of AAT on established CAC. (A) Schematic overview of end-stage disease protocol for AAT administration in AOM/DSS-treated mice. Mice were injected with AOM (day 2 and day 65) followed by four cycles of 2.5% DSS in drinking water. AAT was administrated every 3 days after the last DSS cycle (day 77) for 18 days and tumors were examined on day 98. (B) left, A representative 18F-FDG PET/CT image of AOM/DSS-treated mice before (day 75) and after (day 98) AAT administration. Right, Gamma counting of entire colon was analyzed before and after treatment. The % change of FDG %ID/g of the before and after results between AOM/DSS-treated mice and AOM/DSS-treated mice with AAT administration was further analyzed. Results, mean ± SD [n = 3 (−AAT), n = 5 (+AAT); *, p < 0.05. (C) Macroscopic changes in colon by AAT administration: Tumor multiplicity (D), tumor distribution (E), and average tumor load (F) were determined. Results, mean ± SD [n = 5 (−AAT), n = 5 (+AAT); *, p < 0.05; **, p < 0.01]. (G) Colon tissue was fixed and stained with hematoxylin/eosin on day 98. Top, control colon tissue. AOM/DSS-treated colon tissue without (middle) and with (bottom) AAT administration. AOM/DSS and AAT-treated mice show reduced glandular hyperplasia, mixed inflammation and intramucosal adenocarcinoma formation, which is comparable to no AOM/DSS-treated control mice. Original magnification, ×10.

Figure 2

Therapeutic effects of AAT on established CAC. (A) Schematic overview of end-stage disease protocol for AAT administration in AOM/DSS-treated mice. Mice were injected with AOM (day 2 and day 65) followed by four cycles of 2.5% DSS in drinking water. AAT was administrated every 3 days after the last DSS cycle (day 77) for 18 days and tumors were examined on day 98. (B) left, A representative 18F-FDG PET/CT image of AOM/DSS-treated mice before (day 75) and after (day 98) AAT administration. Right, Gamma counting of entire colon was analyzed before and after treatment. The % change of FDG %ID/g of the before and after results between AOM/DSS-treated mice and AOM/DSS-treated mice with AAT administration was further analyzed. Results, mean ± SD [n = 3 (−AAT), n = 5 (+AAT); *, p < 0.05. (C) Macroscopic changes in colon by AAT administration: Tumor multiplicity (D), tumor distribution (E), and average tumor load (F) were determined. Results, mean ± SD [n = 5 (−AAT), n = 5 (+AAT); *, p < 0.05; **, p < 0.01]. (G) Colon tissue was fixed and stained with hematoxylin/eosin on day 98. Top, control colon tissue. AOM/DSS-treated colon tissue without (middle) and with (bottom) AAT administration. AOM/DSS and AAT-treated mice show reduced glandular hyperplasia, mixed inflammation and intramucosal adenocarcinoma formation, which is comparable to no AOM/DSS-treated control mice. Original magnification, ×10.

Figure 3

Effects of AAT on cytokine and antitumor factor production in CAC. Immunohistochemical staining for PCNA, β-catenin, MPO, IL-6, IGFBP-3 IGFBP-3R identified in colons from AOM/DSS-treated mice versus AOM/DSS and AAT-treated mice using an end-stage disease protocol as in Figure 2. Original magnification, ×20.

Figure 4

AAT suppresses increased serum neutrophil serine proteases PR3 and NE and subsequent IGFBP-3 proteolysis in AOM/DSS mice. (A) Representive WIB of sera from 4 AOM/DSS treated mice, 3 AOM/DSS+AAT treated mice and 2 control mice. AOM/DSS treated mice show increased PR3, NE and IGFBP-3 proteolysis in circulation whereas AAT administration results in significant inhibition of serum PR3, NE and increase of intact IGFBP-3 in circulation. (B) AAT administrated AOM/DSS mice show decreased level of oxidized-AAT but increased total AAT in circulation. ***, p < 0.001. (C) NSPs proteolyze IGFBP-3. Recombinant non-glycosylated human IGFBP-3 protein (30 nM) was proteolyzed by treatment with 100, 250 and 500 nM of PR3, 100 and 250 nM NE or CG for 20 min. at 37 °C. (D–F): AAT and AAT mimic inhibit NSP-induced IGFBP-3 proteolysis. Based upon the sequence identity among NSPs at a putative binding site of AAT and elafin, a 7 amino acid-long peptide (LIRCAML) was generated (D). IGFBP-3 proteolysis by PR-3 (left figure) and NE, CG (right figure) was inhibited by AAT (20 μM) and different concentrations of an AAT mimic (0.05, 0.1, 0.5 and 1 μM) (E). Few IGFBP-3 proteolytic fragments were detected in 20 μM AAT treated or 1 μM AAT mimic treated samples. Anti-protease activity of L- and D-form of N- and C-terminal modified AAT mimics (left figure) (F). Nonfunctional mutant lost anti-protease activity (right figure).

Figure 4

AAT suppresses increased serum neutrophil serine proteases PR3 and NE and subsequent IGFBP-3 proteolysis in AOM/DSS mice. (A) Representive WIB of sera from 4 AOM/DSS treated mice, 3 AOM/DSS+AAT treated mice and 2 control mice. AOM/DSS treated mice show increased PR3, NE and IGFBP-3 proteolysis in circulation whereas AAT administration results in significant inhibition of serum PR3, NE and increase of intact IGFBP-3 in circulation. (B) AAT administrated AOM/DSS mice show decreased level of oxidized-AAT but increased total AAT in circulation. ***, p < 0.001. (C) NSPs proteolyze IGFBP-3. Recombinant non-glycosylated human IGFBP-3 protein (30 nM) was proteolyzed by treatment with 100, 250 and 500 nM of PR3, 100 and 250 nM NE or CG for 20 min. at 37 °C. (D–F): AAT and AAT mimic inhibit NSP-induced IGFBP-3 proteolysis. Based upon the sequence identity among NSPs at a putative binding site of AAT and elafin, a 7 amino acid-long peptide (LIRCAML) was generated (D). IGFBP-3 proteolysis by PR-3 (left figure) and NE, CG (right figure) was inhibited by AAT (20 μM) and different concentrations of an AAT mimic (0.05, 0.1, 0.5 and 1 μM) (E). Few IGFBP-3 proteolytic fragments were detected in 20 μM AAT treated or 1 μM AAT mimic treated samples. Anti-protease activity of L- and D-form of N- and C-terminal modified AAT mimics (left figure) (F). Nonfunctional mutant lost anti-protease activity (right figure).

Figure 5

Anti-proliferative and anti-inflammatory effects of the IGFBP-3/IGFBP-3R axis in colon cancer cells. (A) Basal expression of IGFBP-3R at mRNA (top) and protein (bottom) levels analyzed by Quantitative RT-PCR and WIB, respectively. The mRNA data are presented as fold of change compared to Caco-2. The values of Caco-2 were considered as 1. The cell lysates were subjected to WIB for IGFBP-3R expression. The data represent mean ± SE, n = 3 in duplicate. Heterogeneous expression of IGFBP-3R was observed. (B) Data demonstrating IGFBP-3-induced apoptosis in HT-29 cells (top). Cell death assay was performed two days after the infection. Apoptotic cell death was measured using the Cell Death Detection ELISA. The value for Ad:EV at each MOI was considered as 1. The data represent mean ± SE, n = 3 in duplicate. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Representative immunoblot analysis for IGFBP-3 and IGFBP-3R after infection of Ad:IGFBP-3 (Bottom). Ad:EV: adenoviral plasmids with empty vector; Ad:IGFBP-3: with IGFBP-3 cDNA; Ad:IGFBP-3^GGG: with IGFBP-3^GGG mutant cDNA. Inhibitory effects of IGFBP-3 on TNF-α-induced NF-κB activity (C) and subsequent induction of ICAM-1 expression (D) in HT29 colon cells. n = 3. (E) Protective role of IGFBP-3 on LPS- or TNF-α-induced disruption of colonic epithelial barrier function. Measurement of TEER across confluent intestinal monolayers treated with LPS (1mM) or to TNF-α (100 ng/mL) for 2 days in the presence/absence of IGFBP-3 (30 nM) using a Millicell-ERS voltohmmeter. n = 3 in duplicate.

Figure 6

Central hypothesis: Colonic inflammation-induced NSPs proteolyze IGFBP-3 in circulation, thereby inhibiting endocrine/paracrine/autocrine-derived antitumor and anti-inflammatory actions of IGFBP-3 in CAC. IGFBP-3 activates IGFBP-3R and induces caspase-dependent apoptosis as well as inhibiting colitis-induced NF-κB-signaling, thereby resulting in suppression of CAC progression.