1 Introduction
Tumor growth and metastasis depend on adequate blood supply. The growth of tumor cells is autonomous and faster than the formation of new blood vessels, leading to inadequate nutrient and oxygen supply for new blood vessels. Thus, some tumor cells have a low blood flow and form a hypoxia necrosis area [
1]. This phenomenon is the theoretical basis for using
Salmonella as a vector for gene transfer in cancer treatment because it may preferentially localize to and proliferate in hypoxic areas [
2,
3]. VNP20009, a genetically modified
S. enterica serovar Typhimurium, was attenuated by deletion of
msbB and
purI and underwent phase I clinical trials for metastatic melanomas [
4–
6]. However, no antitumor activities were reported in the trials [
7,
8]. To enhance the anti-tumor therapeutic efficacy of
Salmonella, scholars have developed applications of tumor-targeting VNP20009 as delivery to overcome the penetration limitation of solid tumor [
8] and maximize the activities of genes of interest while reducing systemic toxicity to the host. Mediating the expression of potential genes can further restrict the accumulation of
Salmonella-delivered anti-tumor agents at the tumor area to achieve the maximum anti-tumor therapeutic efficacy.
Targeting angiogenesis that supports tumor growth rather than tumors themselves may be a promising approach for tumor therapy. Tumstatin is an endogenous angiogenic inhibitor that is generated from type IV collagen [
9,
10]; it can inhibit the proliferation and promote the apoptosis of endothelial cells. Maeshima
et al. [
11] reported that the anti-angiogenic activity center of tumstatin is localized to amino acids 54–132 (Tum5), which predominantly binds to the β3 subunit of avβ3 integrin; they also found that tumstatin can inhibit PI3K, FAK, and PKB cell signal transduction pathways [
12].
Selective targeting of tumors by hypoxia-selective tumoricidal delivery system increases the concentration of therapeutic genes in tumor while minimizing the damage to normal tissues. A controllable delivery system is used to manage the timing and location of protein expression
in vivo. The anaerobic-inducible and constitutive
J23100 promoter can be heterologously expressed in facultative anaerobes and shows excellent stability [
13]. In this perspective, we selected VNP20009 as a carrier to deliver Tum5 under the control of the
J23100 promoter (VNP-Tum5) for treatment of melanoma in a mouse model. The selective accumulation of engineered VNP20009 to express tumstatin may avoid unintended toxicities for normal tissues, such as liver and spleen, thereby enhancing the anti-tumor efficacy. The present study explored the anti-tumor effect and possible mechanisms of VNP-Tum5 on a mouse model of B16F10 melanoma
in vitro and
in vivo.
Salmonella-mediated tumor-targeted tumstatin gene therapy significantly suppressed tumor growth and extended host survival. Results provide a preclinical proof-of-principle that VNP-Tum5 has better anti-tumor and anti-angiogenic effects than VNP
in vitro and
in vivo. Hence, Tum-load
Salmonella is a potential therapeutic approach in oncotherapy. This work is the first to report the use of Tum5-transformed VNP20009 as a specific gene delivery system for tumor treatment.
2 Materials and methods
2.1 Bacterial strains, tumor cell lines, and animals
The bacterial strain VNP20009 and B16F10 melanoma were obtained from ATCC (USA). MUVECs were purchased from Hefei Bomei Biotechnology Co., Ltd. (Hefei, China). All the strains used in this study were grown in Luria-Bertani broth at 37 °C. B16F10 melanoma cells and MUVECs were cultured in 5% CO2 at 37 °C in high-glucose DMEM (Sigma-Aldrich, Shanghai, China) supplemented with 10% fetal bovine serum (Hyclone, USA), streptomycin (100 μg/mL), and penicillin (100 IU/mL). Female C57BL/6 mice, 6–8 weeks old, were obtained from the Comparative Medicine Center of Yangzhou University (Yangzhou, China) and maintained under a sterile environment condition for 1 week before the start of the experiment. This study was approved by the Animal Care and Use Committee of Nanjing University and carried out following the Guide for the Care and Use of Laboratory Animals by the National Research Council.
2.2 Plasmid construction and bacterial transformation
Tum5 (tumstatin, 54–132 aa) was amplified by primers Tum-sense and Tum-antisense. The primer sequences were as follows: Tum-sense, 5′-cggcagcggaggtggaggcagcggaaatgaacaagcccatgga-3′; Tum-antisense, 5′-ctggtactagtggatccttaaatcgtaggaccttcacagaca-3′. The Tum5 sequence was cloned into the prokaryotic expression vector pTh01 (maintained in our laboratory) by using a one-step cloning kit (Vazyme Biotech, Nanjing, China) under the control of the J23100 promoter to obtain plasmid pTh21-Tum5. The positive clones were confirmed by DNA sequencing (Sangon Biotech, Shanghai, China). The pTh01 and pTh21-Tum5 plasmids were transformed into VNP20009 by using Gene Pulser Xcell™ (Bio-Rad, CA, USA) at 25 μF, 1.8 kV, and 400 Ω and plated on LB agar containing 50 μg/mL kanamycin. Positive VNP20009 was named VNP and VNP-Tum5.
2.3 Construction of mouse tumor model
Female wild-type C57BL/6 (6–8 weeks old) mice were inoculated subcutaneously with 5×10
5 B16F10 melanoma cells in 0.1 mL of PBS on the mid-right flank. Mice with B16F10 melanoma and tumor volume of about 100 mm
3 (at day 7 post tumor inoculation) were randomly divided into three groups: VNP-Tum5, VNP, and PBS. Mice in the VNP-Tum5 and VNP groups were treated intraperitoneally with
S. typhimurium strains VNP-Tum5 or VNP at a dose of 1 × 10
6 CFU in 0.1 mL of PBS, respectively [
14,
15]. Mice in the PBS group were injected with 0.1 mL of PBS and used as control (
n = 12 in each group). The width and length of the tumor were recorded every 3 days. The number and dates of death of mice were recorded to calculate survival rate. Tumor volume was determined using the formula: tumor volume = length × width
2 × 0.52. Assuming that changes in the tumor volume are exponential, a regression line, lg
y = lga + b
x (
x, days after the baseline radiologic image;
y, tumor volume), was calculated by nonlinear square regression. The doubling time (DT) of tumor volume was defined as (lg2)/b. DT was calculated from the baseline image and historical image obtained before. Tumor growth delay (TGD) was calculated as the time taken by each individual tumor to reach 800 mm
3 in the treatment groups compared with that in the untreated controls. The mice were sacrificed at day 13 post tumor inoculation (
n = 4 in each group). Tumor tissues, spleen, and liver were collected.
2.4 Staining and microscopy
An B16F10 melanoma mouse model was used to understand the treatment properties of VNP-Tum5. After bacterial treatment for 6 days, mice in the VNP-Tum5, VNP, and PBS groups were sacrificed to determine the mechanism of the recombined bacteria. The tumor, spleen, and liver tissues were fixed in 10% paraformaldehyde (PFA) overnight, paraffin-embedded, sectioned, fixed on slides (thickness of 5 μm), and prepared according to the standard manufacturing procedures for paraffin sections. Hematoxylin–Eosin (HE) staining was carried out by HE Staining Kit (Solarbio Life Science, Beijing, China). TdT-mediated dUTP Nick-End Labeling (TUNEL) assays were carried out by detecting the apoptotic nuclei on TUNEL BrightGreen Apoptosis Detection Kit (Vazyme Biotech, Nanjing, China) following the manufacturer’s instructions. Ki67 mouse mAb (#9449) and CD31 Rabbit mAb (#77699) were acquired from CST Technology. After staining, the relative fluorescence intensity was quantified by ImageJ software (NIH, Bethesda, MD, USA).
2.5 In vitro cell migration assay
MUVECs were seeded into six-well plates at 2 × 105 cells/well and cultured with 10% FBS overnight. After washing twice with PBS, the monolayers were wounded in a line across the well with a 10 μL standard pipette tip and then incubated with VNP and VNP-Tum5 in the absence of serum media. Photographs of migration into open wounds were taken at different time points until the scratches were almost closed.
2.6 Western blot analysis
The tissues were lysed in RIPA buffer (Santa Cruz Biotechnology) supplemented with 1 μL of protease inhibitor cocktail (Cell Signaling, USA) and incubated on ice for about 30 min. Protein concentration was calculated using the BCA Protein Quantification Kit (Vazyme Biotech, Nanjing, China). About 50 μg of total protein was used for Western blot analysis following standard conditions with primary antibodies as follows: diluted 1:1000 for VEGF-A (#19003-1-AP, Proteintech Group, Wuhan, China), 1:1000 for GAPDH (#2819, Santa, Shanghai, China), Flag (#18230, Abcam, Cambridge, USA), VEGFR-2 (#9698, CST, Shanghai, China), AKT (#4085S, CST, Shanghai, China), p-AKT (#9275S, CST, Shanghai, China), PI3K (#4292S, CST, Shanghai, China), p-PI3K (#17366S, CST, Shanghai, China), caspase 3 (#9662S, CST, Shanghai, China), and p-caspase 3 (#17366S, CST, Shanghai, China). Detection was performed with ECL plus Western blot detection system (Tanon, Shanghai, China). The tumor tissues were homogenized in lysis buffer and incubated on ice for about 30 min. The sample was centrifuged, and the supernatant was used for Western blot analysis. Intensity was quantitatively analyzed by ImageJ software.
2.7 Flow cytometry analysis of cell apoptosis
2 ×105 B16F10 cells were seeded in a 12-well plate for about 12 h to reach a density of 70%–80% and treated with different bacteria for 2 h and 4 h. Cells were harvested by trypsinization, washed with cold PBS, and stained with EGFP-annexin V in binding buffer for 15–20 min and propidium iodide for 5 min. The results were analyzed using Flow Cytometer (BD FACSCanto II). Apoptosis rate was analyzed with FlowJo analysis software (Tree Star, Ashland, OR, USA). Three independent experiments were conducted, and each experiment was repeated twice.
2.8 Cell toxicity assay
MTT (Thiazolyl Blue Tetrazolium Bromide) assay was used to determine anti-proliferative effects on endothelial cells [
16]. In brief, 5 × 10
3 MUVECs were seeded in a 96-well plate for 12 h to reach a density of 70%–80% and then treated with
Salmonella for 2 h at different MOI (1:10, 1:50, 1:100). Absorbance was recorded at 490 nm.
2.9 Statistical analysis
All data were expressed as mean ± SD after analysis with GraphPad Prism 8.0 (Graph Pad Software, San Diego, CA, USA). The effect of treatment on survival time was determined using log-rank test. Paired Student’s t-test analysis was used to estimate statistical differences among groups. P < 0.05 was considered to indicate statistical significance.
3 Results
3.1 VNP-Tum5 inhibited tumor growth and prolonged survival time
In bacteria-based anticancer therapy, maintaining functional gene expression in the hypoxic tumor regions is essential. In our previous work, we constructed the pTh01 plasmid with a strong promoter J23100, a bacterial signal peptide that ensures protein expression, and a Flag tag, which is a kanamycin resistance selection marker (Fig.1). As shown in Fig.1, the Tum5-Flag protein was successfully expressed in melanoma tissues.
The anti-tumor effects of Salmonella strain VNP and VNP-Tum5 were determined on B16F10 mouse model. In mice treated with VNP-Tum5 (1 × 106 CFU), tumor growth was suppressed (Fig.1) and survival was prolonged (Fig.1). At sacrifice time, B16F10 melanoma (day 13) in the VNP-Tum5 group reached an average tumor volume of 737 mm3, while that in the VNP group was 1235 mm3 (1.67-fold). The tumor sizes were lower in the treatment groups than in the control group (treated with PBS). The average tumor weight in the VNP-Tum5 group (1.35 g) was reduced compared with that in the VNP group (2.45 g, Fig.1). Moreover, VNP-Tum5 significantly prolonged the overall survival (Fig.1). The tumor DT (7.3 days, Fig.1) and TGD (23.2 days, Fig.1) were significantly increased in mice treated with VNP-Tum5 than in the VNP group (4.6 days and 18.4 days, respectively). Overall, VNP-Tum5 led to stronger tumor growth inhibition than VNP.
3.2 Colonization and toxicological risk assessment of VNP-Tum5 in tumor-bearing mice
As shown in Fig.2, the colony formation assays of VNP-Tum5 in the tumor, spleen, and liver tissues demonstrated that the exact number of tumor-colonized bacteria was not improved in the VNP group. The relative tumor specificity of VNP-Tum5 to the liver was improved to 2500:1; a previous study reported that VNP20009 had a tumor-to-liver ratio of bacterial colonization of about 1000:1 [
17]. The results demonstrated that the expression system under the control of the hypoxia-induced
J23100 promoter could enhance the targeted efficacy.
As shown in Fig.2 and 2C, the Tum5-Flag protein was successfully expressed in the spleen and liver. Treatment with VNP-Tum5 did not change the weight of the body (Fig.2), spleen (Fig.2 and 2E), and liver (Fig.2 and 2G) compared with VNP. However, the weights of the spleen (Fig.2) and liver (Fig.2) in the VNP-Tum5 and VNP groups were higher than those in the PBS group. The body weight of tumor-bearing mice was then examined on day 5 after bacterial therapy. A steady decrease of approximately 15% in body weight was found in mice treated with Salmonella (Fig.2) compared with PBS. Additionally, the HE staining of the liver and spleen (Fig.2 and 2I) displayed no necrosis areas in the VNP-Tum5 and VNP groups. Hence, VNP-Tum5 has no potential toxic side effects compared with VNP.
3.3 VNP-Tum5 induced the apoptosis and necrosis of melanoma in vivo and in vitro
The histochemistry scores (Fig.3 and 3B) showed that all groups had tumor necrosis region, while VNP-Tum5 formed more necrosis areas (63.5%) than VNP (39.8%). The TUNEL analysis revealed that the tumor tissues in the VNP-Tum5 group had higher apoptosis rate (36.5%) than that in the VNP group (24.2%, Fig.3 and 3E). VNP-Tum5 could induce more melanoma cells to undergo apoptosis. Tumor proliferation was also assessed by immunofluorescence using an antibody against murine Ki67 (Fig.3). VNP-Tum5 can inhibit tumor proliferation rate by 17.2% compared with VNP (P < 0.01, Fig.3 and 3D).
After B16F10 cells were treated with PBS, VNP, and VNP-Tum5 for 2 h and 4 h, the rates of apoptosis and necrosis were analyzed by flow cytometry (Fig.3). The apoptosis and necrosis rates were significantly increased by 19.2% (P < 0.001, Fig.3) and 3.5% (P < 0.05, Fig.3), respectively, in the VNP-Tum5 group at 4 h compared with those in the VNP group. Statistical analysis showed no significant difference in the apoptosis and necrosis rates after treatment at 2 h (Fig.3 and 3H). Hence, VNP-Tum5 could induce more apoptosis and necrosis of melanoma in vivo and in vitro. Moreover, the synergistic tumor inhibition effects of VNP-Tum5 could be partially attributed to the enhanced apoptosis and necrosis of melanoma cells.
3.4 VNP-Tum5 targeted tumor vessels and inhibited angiogenesis by downregulating VEGF-A
Tumor necrosis is usually caused by insufficient blood supply; in this regard, whether VNP-Tum5 affects tumor angiogenesis should be investigated. The blood vessel density of melanoma tumor tissue was evaluated by immunofluorescence using an antibody against mouse CD31, a well-recognized marker of angiogenesis (Fig.4). The vascular density of tumors treated with VNP-Tum5 was significantly decreased by about 6% compared with that of tumors treated with VNP (Fig.4). To determine the mechanism underlying the lower microvascular densities induced by VNP-Tum5, we detected the vascular endothelial growth factor (VEGF-A) levels of the tumor (Fig.4 and 4D). As expected, the expression of VEGF-A was decreased by about 1.9-fold by VNP-Tum5 treatment (P < 0.05, Fig.4).
3.5 VNP-Tum5 inhibited the migration and proliferation of MUVECs
A series of experiments was conducted on MUVECs to investigate the antiangiogenic activity of VNP-Tum5 in vitro (MOI = 10:1). First, we examined the effect of VNP-Tum5 on MUVEC motility by wound healing scratching assay (Fig.4). Quantitative analysis revealed that VNP-Tum5 effectively hindered the healing process by 47.8%, which was higher than that in the VNP group (22.6%, P < 0.001, Fig.4).
The effect of VNP-Tum5 and VNP on the proliferation of MUVECs was assessed using MTT assay. As shown in Fig.4, the inhibition rates of VNP-Tum5 at MOIs of 100:1, 50:1 and 10:1 were increased by 9.1% (P < 0.001), 7.3% ( P < 0.01), and 0.7% ( P > 0.01), respectively, compared with VNP. In summary, the inhibition of MUVEC proliferation was stronger in the VNP-Tum5 group than in the VNP group. Moreover, the inhibition rate increased in a dose-dependent manner with increasing MOI. Collectively, these results suggest that VNP-Tum5 could significantly inhibit the migration and proliferation of MUVECs in vitro.
3.6 VNP-Tum5 induced apoptosis by inhibiting the VEGFR2/PI3K/AKT pathway
The vascular endothelial growth factor (VEGF-A) levels in tumor tissues treated with VNP were measured to determine the mechanism of lower microvascular densities caused by VNP-Tum5. As expected, VEGF-A in the tumor tissues was inhibited by VNP-Tum5 therapy (Fig.4), suggesting the potential regulatory role of VEGF-A by VNP-Tum5. To determine the signaling pathways associated with VEGF-A in mediating apoptosis, we examined the expression of VEGFR2 and its downstream pathway molecules such as PI3K and AKT (Fig.5). Treatment with VNP-Tum5 obviously decreased the phosphorylation of PI3K and AKT proteins (Fig.5). The levels of PI3K and AKT were not changed or even reduced, and the total protein expression levels were kept unchanged (Fig.5), suggesting that VEGFR2 was responsible for the inhibition of the phosphorylation of PI3K and AKT molecules. In addition, the protein expression levels of p-PI3K/PI3K and p-AKT/AKT were significantly reduced in the VNP-Tum5 group.
4 Discussion
Tumor development may be determined by the relative levels of pro-angiogenic factors and anti-angiogenic factors, which are in a balanced state under normal physiologic conditions [
18]. The transformation from
in situ tumor to malignancy may involve a transformation of neovascularization, so gene-regulated endogenous angiogenesis inhibitors at the physiologic level may serve as a checkpoint for tumor growth [
19]. Tumstatin is a potential antitumor drug due to its dual mechanism of anti-tumor angiogenesis and inhibition of tumor cell proliferation [
20,
21]. Compared with common anti-tumor drugs, tumstatin has the following advantages. (1) Tumstatin has fewer side effects because it is a small fragment of endogenous protein (28 kDa) and is a highly conserved molecule that does not normally produce toxicity [
22,
23]. (2) Tumstatin has less drug resistance because it targets vascular endothelial cells where genes are more stable than tumor cells [
11]. (3) Tumstatin has high specificity because it specifically inhibits tumor angiogenesis without affecting physiologic angiogenesis [
24]. (4) Tumstatin has high anti-tumor effect because it can target blood vessels to induce tumor tissue necrosis and prevent tumor metastasis to achieve better anti-tumor effect [
25,
26]. Thus, tumstatin is a promising therapeutic candidate in the control of tumor angiogenesis and growth.
Anaerobic
Salmonella can selectively colonize tumors, inhibit tumor growth, and prolong survival after systemic infection in animal tumor models [
6,
27,
28]. The effectiveness of engineered attenuated
S. typhimurium strain, which was employed as live delivery vectors of various antitumor therapeutic agents or combined with other therapies, has been evaluated in a large number of animal experiments [
2,
29,
30]. For example, the well-known genetically engineered
S. typhimurium strain VNP20009 (
purI−/msbB−) was attenuated by more than 10 000-fold compared with the wild-type strain [
14,
17,
31,
32]. VNP20009 has an excellent safety profile in rodents and dogs and was thus subjected to phase I clinical trials, which documented the safety of the bacterial regimen in patients with cancers [
7,
33]. However, VNP20009 showed insufficient tumor colonization and weak biological effects, and none of the patients experienced obvious tumor regression [
7].
In this study, we investigated whether tumstatin could exert high antiangiogenesis effects when used with VNP20009 as a delivery system to achieve better inhibitory effect on tumor growth in melanoma-bearing mice. We developed a bacteria-mediated system under the control of the hypoxia-induced
J23100 promoter to express the amino acids 54–132, where the antiangiogenic activity of the tumstatin protein is located [
11].
Treatments using 106 CFU engineered Salmonella induced toxicity (Fig.2). The combination of tumstatin and VNP20009 could be a safe option and has no serious side effects. Therefore, using VNP20009 as a vector to deliver tumstatin may provide a promising therapeutic regimen for melanoma.
The results of the examination of tumor volume and histological changes at macroscopic and microscopic levels showed that Tum5 inhibited tumor growth due to the transgenic expression of Tum5. The vector group VNP failed to exhibit similar effects. The TUNEL and Ki67 assays showed that VNP-Tum5 was associated with significant apoptotic activities and suppressed the proliferation of melanoma cells in vivo and in vitro, which could contribute to the tumor inhibitory effects and survival benefit observed in this study.
The VNP20009 delivery of Tum5 markedly restrained the tumor angiogenesis. We demonstrate the possibility of significantly reducing CD31 levels in tumor tissues. To confirm the mechanisms of apoptosis induction, proliferation inhibition, and anti-angiogenesis activity by VNP-Tum5, we designed a series of experiments in vitro. VNP-Tum5 induced the apoptosis of B16F10 cells. The wound healing and MTT assays on MUVECs showed that VNP-Tum5 inhibited cell migration and proliferation. Thus, VNP-Tum5 exerted an antitumor effect via several mechanisms: inducing tumor apoptosis and suppressing tumor migration. In addition, we evaluated the expression of apoptotic effector caspase 3 and the changes in VEGFR2, p-PI3K, P13K, p-AKT, and AKT on tumor tissues. The expression of cleaved caspase 3 increased, and those of VEGFR2, p-AKT/AKT, p-PI3K/P13K decreased. These results suggest that the antitumor effects of the engineered bacteria may be at least partially mediated by the VEGF-A/VEGFR-2 and P13K-AKT signaling pathways.
To our knowledge, this study is the first to report the tumor suppression and anti-angiogenic effects of VNP20009 used as a delivery vector for the protein tumstatin in vivo and in vitro. VNP-Tum5 therapy could serve as a prototype for further development of a feasible and effective option for treatment of melanoma.
PDF
(4492KB)
