Figure 1.
Collagen increases histone H3K9 and H3K27 acetylation.
A, B. Human pancreatic tissue microarrays (TMAs) containing 24 specimens were immunostained with IgG control antibody or for histone H3K9 and histone H3K27 acetylation (Ac). The TMAs were also trichrome stained to assess for fibrosis. The insets show higher magnification images of staining with control IgG, and for histone H3K9Ac and histone H3K27Ac. C. Quantification of histone H3K9Ac- and histone H3K27Ac-positive cells was performed using Adobe Photoshop CS3 software. *, p<0.01 relative to sections with low fibrosis. D. Panc1 and CD18 cells were grown on tissue culture plastic or in three-dimensional collagen gels for 24 hours. Cells were lysed and immunoblotted for histone H3K9Ac and H3K27Ac using α-tubulin as loading control. The results are representative of at least four independent experiments.
Figure 2.
HMGA2 regulates collagen-induced H3K9 and H3K27 acetylation.
A. Panc1 and CD18 cells were grown on tissue culture plastic or in three-dimensional collagen gels for 24 hours and immunoblotted for HMGA2. The results are representative of three independent experiments. B–D. Panc1 and CD18 cells were transfected with control siRNA or with 2 different HMGA2 siRNAs, allowed to recover overnight and then embedded in three-dimensional collagen for 24 hours. Lysates were then immunoblotted for HMGA2 (B), histone H3K9Ac (C) and histone H3K27Ac (D) using α-tubulin as loading control. The results are representative of at least three independent experiments. E. Human pancreatic TMAs were immunostained for HMGA2 (left), histone H3K9Ac (middle) and histone H3K27Ac (right). F. The relationship between HMGA2 and histone H3K9Ac or H3K27Ac was assessed by Fisher's exact test.
Figure 3.
Collagen increases expression of p300, PCAF and GCN5 histone acetyltransferases.
A. Panc1 and CD18 cells were grown on tissue culture plastic or in three- dimensional collagen gels for 24 hours. Cells were lysed and immunoblotted for p300, PCAF and GCN5 histone acetyltrasferases (HATs) using α-tubulin as loading control. The results are representative of three independent experiments B–D. Panc1 and CD18 cells were transfected with control siRNA or with combination of siRNAs against p300, PCAF and GCN5 (HAT siRNA); allowed to recover overnight and then embedded in three-dimensional collagen gels for 24 hours. The lysates were immunoblotted for the corresponding HAT proteins (B), and histone H3K9Ac (C) and H3K27Ac (D). The results are representative of at least three independent experiments. E, F. Panc1 and CD18 cells were transfected with control siRNA or with individual siRNAs against p300, PCAF and GCN5; allowed to recover overnight and then embedded in three-dimensional collagen gels for 24 hours. The lysates were immunoblotted for histone H3K9Ac (E) and H3K27Ac (F). The results are representative of three independent experiments.
Figure 4.
HMGA2 mediates collagen-induced p300, PCAF and GCN5 HAT expression.
Panc1 and CD18 cells were transfected with control siRNA or with 2 different HMGA2 siRNAs, allowed to recover overnight and then plated in three-dimensional collagen gels for additional 24 hours. The lysates were then analyzed for p300 (A), PCAF (B) and GCN5 (C) HATs using α-tubulin as loading control. The results are representative of at least three independent experiments.
Figure 5.
HMGA2 and HATs mediate protection against gemcitabine in three-dimensional collagen gels.
A. CD18 cells grown on plastic or in three-dimensional collagen gels were left untreated or treated with gemcitabine for 24 hours. The cells were then trypsinized or extracted out of collagen by collagenase treatment. The cells were replated on tissue culture plastic or in three-dimensional collagen gels at low density (left). The cells were photographed 5 days later and counted (right). **, p<0.001. The results are representative of at least three independent experiments. B, C. CD18 cells were transfected with control siRNA, HMGA2 siRNA (B) or combination of siRNAs against PCAF, GCN5 and p300 (C), allowed to recover overnight and then plated in collagen gels for 24 hours. The cells were then treated with gemcitabine for 24 hours, extracted out of collagen by collagenase treatment and then replated in three-dimensional collagen gels at low density. The cells were photographed 5 days later and counted. *, p<0.05. The results are representative of at least 3 independent experiments.
Figure 6.
Regulation of gemcitabine resistance by HMGA2 and HATs.
Previously, we had published that the collagen microenvironment induced HMGA2 expression [Dangi-Garimella S et al, [6]]. We now demonstrate that PDAC cells in the collagen microenvironment induce HMGA2-dependent HAT expression and histone H3 acetylation, suggesting that the collagen microenvironment promotes euchromatin formation. This signaling pathway allows PDAC cells to resist the effects of gemcitabine in the collagen microenvironment.