Rapid vascular regrowth in tumors after reversal of VEGF inhibition

Abstract

Inhibitors of VEGF signaling can block angiogenesis and reduce tumor vascularity, but little is known about the reversibility of these changes after treatment ends. In the present study, regrowth of blood vessels in spontaneous RIP-Tag2 tumors and implanted Lewis lung carcinomas in mice was assessed after inhibition of VEGF receptor signaling by AG-013736 or AG-028262 for 7 days. Both agents caused loss of 50%-60% of tumor vasculature. Empty sleeves of basement membrane were left behind. Pericytes also survived but had less alpha-SMA immunoreactivity. One day after drug withdrawal, endothelial sprouts grew into empty sleeves of basement membrane. Vessel patency and connection to the bloodstream followed close behind. By 7 days, tumors were fully revascularized, and the pericyte phenotype returned to baseline. Importantly, the regrown vasculature regressed as much during a second treatment as it did in the first. Inhibition of MMPs or targeting of type IV collagen cryptic sites by antibody HUIV26 did not eliminate the sleeves or slow revascularization. These results suggest that empty sleeves of basement membrane and accompanying pericytes provide a scaffold for rapid revascularization of tumors after removal of anti-VEGF therapy and highlight their importance as potential targets in cancer therapy.

Figures

Figure 1. Regression and regrowth of tumor vessels after VEGF inhibition.

Confocal micrographs of RIP-Tag2 tumors compare the dense vascularity of an untreated tumor (A) with the sparse vascularity after AG-013736 treatment for 7 days (B). Tumor vessels that survived treatment had a more uniform caliber and less branching than those in untreated tumors. At 2 days after the treatment ended, vascular sprouts (C, arrows) marked the beginning of vascular regrowth. The projections from vessels were confirmed as sprouts by examining multiple optical sections in the stack of confocal images from the 80-μm-thick cryostat section. (D) By 7 days, normalized vessels were replaced by typical tumor vessels. Vascularity of Lewis lung carcinoma (E) was similarly reduced by AG-013736 treatment for 7 days (F). (G) By 7 days after the treatment, vascularity of Lewis lung carcinoma was back to baseline. Bar graphs illustrate changes in area density of CD31-positive vessels in RIP-Tag2 tumors (H) and Lewis lung carcinomas (I) after 7 days of treatment (0 days withdrawal) and after the treatment ended. Treatment reduced tumor vascularity in the 2 models by 61% and 50%, respectively. After AG-013736 treatment was stopped, tumor vascularity was back to baseline by 7 days but from day 7–14 did not increase beyond that present in untreated tumors (H). *P < 0.05 compared with the untreated group. Scale bar (applies to all images): 115 μm (A–D); 55 μm (E–G).

Figure 2. CD31-positive blood vessels labeled with i.v. FITC-lectin in RIP-Tag2 tumors.

Confocal micrographs of tumors: untreated (A); treated with AG-013736 for 7 days (B); and 1 day (C) or 7 days (D) after withdrawal of AG-013736. (A) In the untreated tumor, almost all blood vessels were labeled with lectin, but some CD31-positive endothelial sprouts were not labeled (arrows). (B) After treatment, surviving tumor vessels had lectin labeling but lacked sprouts. At 1 day (C) and 7 days (D) after treatment, nearly all tumor vessels were stained with lectin, but lectin-negative endothelial sprouts were abundant (C and D, arrows). Focal regions of extravasated lectin were present at 7 days (D, green, arrowheads in merged image), suggestive of vessel leakiness. (E) Graph showing similarity of area densities of CD31-positive blood vessels and lectin-stained vessels during 7 days of vascular regrowth, indicative of vessel functionality. *P < 0.05 compared with corresponding values for untreated tumors. Scale bar: 22.5 μm.

Figure 3. Reduction in intensity of VEGFR-2 immunofluorescence in RIP-Tag2 tumor vessels after AG-013736 treatment (i.p. injection) for 7 days and during vascular regrowth.

Fluorescence micrographs comparing the intense VEGFR-2 immunofluorescence and high vascular density in an untreated tumor (A) with the faint VEGFR-2 immunofluorescence and sparse vasculature in a tumor after 7 days of AG-013736 treatment (B). (C) At 7 days after the treatment ended, both VEGFR-2 immunofluorescence and vascular density returned to baseline. (D–F) Changes in height of peaks in surface plots illustrate the reduction in VEGFR-2 immunofluorescence after 7-day treatment and rebound 7 days thereafter (F). (G) Measurements of VEGFR-2 immunofluorescence show the magnitude of the changes. Comparison of VEGFR-2 immunofluorescence and area density of CD31-positive tumor vessels (both values expressed as percent of corresponding untreated baseline value) suggests that the increase in VEGFR-2 expression preceded the increase in vascular density. (H) Further experiments showed that a second round of AG-013736 reduced tumor vascularity as much as the first round, indicating that much of the regrown tumor vasculature was VEGF dependent. *P < 0.05 compared with corresponding baseline value. †P < 0.05, CD31 compared with VEGFR-2. Scale bar: 60 μm (A–C).

Figure 4. Changes in pericyte phenotype during regression and regrowth of blood vessels in RIP-Tag2 tumors.

Fluorescence microscopic images of tumors stained for CD31 (green, endothelial cells) and α-SMA (red, pericytes) comparing untreated vasculature (A), regression after AG-013736 for 7 days (B), and regrowth for 7 days after treatment (C). (D) Area density measurements show that AG-013736 reduced CD31 more than α-SMA, but both returned to baseline during 7 days of regrowth. (E and F) By comparison, PDGFR-β immunoreactivity was not changed by the 7-day treatment. (G) Measurements showed that the area density of PDGFR-β–positive pericytes, unlike that of α-SMA–positive pericytes, remained relatively constant. (G) The ratio of α-SMA immunoreactivity to PDGFR-β immunoreactivity, illustrated by the amount of colocalization of the 2 markers, decreased after treatment but returned to baseline by 7 days after treatment ended. The stability of the PDGFR-β–positive pericyte population suggests that the short-term change in α-SMA immunoreactivity reflects a change in pericyte phenotype rather than in pericyte number. *P < 0.05 compared with control. †P < 0.05 compared with PDGFR-β. Scale bar (applies to all images): 85 μm (A–C); 100 μm (E and F).

Figure 5. Selective inhibition of VEGFR signaling by AG-028262 causes reversible changes in endothelial cells and pericytes in RIP-Tag2 tumors.

Fluorescence micrographs of tumors stained for CD31 (green) comparing tumor vasculature without treatment (A) with that after AG-028262 treatment for 7 days (B) and 7 days after drug withdrawal (C). (D) Area density measurements show a significant decrease in CD31-positive endothelial cells after AG-028262 treatment and complete regrowth after withdrawal for 7 days. (E) α-SMA immunoreactivity decreased 54% during AG-028262 treatment for 7 days (0 days withdrawal) but returned to baseline by 7 days after the treatment ended. By comparison, PDGFR-β immunoreactivity changed little during treatment or regrowth. (F and G) In RIP-Tag2 tumors treated with AG-013736 followed by 2 days withdrawal, some sleeves of basement membrane contained pericytes but not CD31-positive endothelial cells (region between the arrows). (F and G) Pericytes marked by α-SMA immunoreactivity. (G) Basement membrane marked by type IV collagen immunoreactivity. *P < 0.05 compared with control. Scale bar (applies to all images): 160 μm (A–C); 15 μm (F and G).

Figure 6. Stability of vascular basement membrane during regression and regrowth of blood vessels in RIP-Tag2 tumors.

Fluorescence microscopic images of tumors showing CD31-positive endothelial cells (green), type IV collagen–positive basement membrane (red), and merged images at baseline (A), after AG-013736 treatment for 7 days (B) and after 7-day treatment and 7-day withdrawal (C). The 2 markers colocalized almost completely in untreated RIP-Tag2 tumors (A), but after AG-013736 treatment (B), CD31-positive vessels were sharply reduced but basement membrane was not, as reflected by abundant red strands (B, Merged). (C) By 7 days after treatment ended, tumor vascularity recovered, and most type IV collagen again colocalized with CD31-positive vessels (Merged). (D) Graph of area densities of CD31 and type IV collagen showing that basement membrane remained relatively constant during regression and regrowth of endothelial cells. (E) Graph showing similar rates of tumor vessel regrowth, reflected by CD31 area density, expressed as percent reduction compared with the value for untreated tumors, and disappearance of empty basement membrane sleeves, reflected by the fact that type IV collagen was unaccompanied by CD31, during 14 days after the end of AG-013736 treatment. *P < 0.05 compared with the untreated group. †P < 0.05 compared with the corresponding value for type IV collagen. Scale bar: 120 μm (A–C).

Figure 7. Vascular basement membrane as a scaffold for regrowth of tumor vessels.

(A–C) Vascular regression and regrowth in RIP-Tag2 tumors. (A) Staining with FITC-lectin and CD31 showed a patent vessel with nonpatent sprouts (arrows) 1 day after AG-013736 treatment ended. (B) Type IV collagen marks loose basement membrane (green). Intense type IV collagen staining next to sprouts (arrows) reflects matrix heterogeneity around regrowing vessel. (C) CD31-positive filopodia (arrows) penetrated basement membrane sleeve. (D–F) Paired images showing that MMP inhibitor AG3340 increased AG-013736–induced vascular regression (D and E) but did not remove basement membrane sleeves (E) or slow revascularization after AG-013736 treatment ended (F). (G) CD31 and type IV collagen area densities indicated no inhibition of regrowth by MMP inhibition. Groups: (I) vehicle, 7 days; (II) AG-013736 plus AG3340, 7 days; (III) AG-013736 plus AG3340, 7 days, then no treatment, 7 days; (IV) AG-013736 plus AG3340, 7 days, then AG3340 alone, 7 days. (H) Area densities of CD31 and type IV collagen immunoreactivities indicated no inhibition of regrowth by antibody HUIV26. Groups: (I) vehicle, 7 days; (II) AG-028262, 7 days; (III) AG-028262 plus HUIV26, 7 days; (IV) AG-028262, 7 days, then no treatment, 7 days; (V) AG-028262 plus HUIV26, 7 days, then HUIV26 alone, 7 days. (I and J) Sites of VEGF accumulation in basement membrane sleeves (I and J, arrows) and tumor cells (I, arrowheads) in RIP-Tag2 tumor after AG-013736 treatment (i.p. injection) for 7 days. *P < 0.05 compared with vehicle group. Scale bar (applies to all images): 17 μm (A and B); 13 μm (C); 160 μm (D–F); 20 μm (I and J).