Sononeoperfusion effect by ultrasound and microbubble promotes nitric oxide release to alleviate hypoxia in a mouse MC38 tumor model

Graphical abstract

Highlights

• •Sononeoperfusion refers to tumor perfusion enhancement by ultrasound and microbubble.
• •Sononeoperfusion effect is a promising method to alleviate tumor hypoxia.
• •The mechanism of sononeoperfusion may be related to cavitation and related NO release.
• •Sononeoperfusion effect can be easily translated to the clinic.

Abstract

Tumor hypoperfusion not only impedes therapeutic drug delivery and accumulation, but also leads to a hypoxic and acidic tumor microenvironment, resulting in tumor proliferation, invasion, and therapeutic resistance. Sononeoperfusion effect refers to tumor perfusion enhancement using ultrasound and microbubbles. This study aimed to further investigate hypoxia alleviation by sononeoperfusion effect and explore the characteristics and mechanism of sononeoperfusion effect. To stimulate the sononeoperfusion effect, mice bearing MC38 colon cancers were included in this study and diagnostic ultrasound for therapy was set at a mechanical index (MI) of 0.1, 0.3, and 0.5, frequency of 3 MHz, pulse length of 5 cycles, and pulse repetition frequency of 2000 Hz. The results demonstrated that a single ultrasound and microbubble (USMB) treatment resulted in tumor perfusion enhancement at MI = 0.3, and nitric oxide (NO) concentration increased at MI = 0.3/0.5 (P < 0.05). However, there were no significant difference in the hypoxia-inducible factor-1α (HIF-1α) or D-lactate (D-LA) (P > 0.05) levels. Multiple sononeoperfusion effects were observed at MI = 0.3/0.5 (P < 0.05). For each treatment, USMB slightly but steadily improved the tumor tissue oxygen partial pressure (pO₂) during and post treatment. It alleviated tumor hypoxia by decreasing HIF-1α, D-LA level and the hypoxic immunofluorescence intensity at MI = 0.3/0.5 (P < 0.05). The sononeoperfusion effect was not stimulated after eNOS inhibition. In conclusion, USMB with appropriate MI could lead to a sononeoperfusion effect via NO release, resulting in hypoxia amelioration. The tumors were not resistant to multiple sononeoperfusion effects. Repeated sononeoperfusion is a promising approach for relieving tumor hypoxia and resistance to therapy.

1. Introduction

Hypoxia is one of the major characteristics of most advanced solid tumors [1]. With uncontrolled proliferation, there is typically a poorly perfused region in the tumor center. The distance from the tumor cells to the capillaries limits oxygen diffusion. A “hypoxic zone” arises in the tumor due to an imbalance between oxygen consumption and supply [1], [2]. Tissue is considered hypoxic when the oxygen pressure drops below 10 mmHg, compared to 40–60 mmHg in most normal tissues [3]. By adapting to hypoxia, surviving tumor cells gain a hypoxia-resistant phenotype by triggering the expression of certain genes, signaling pathways, and metabolism [4]. An increase in lactate and decrease in pH are associated with hypoxia and hypoperfusion, which further promote tumor progression [5].

Importantly, hypoxia and the related tumor microenvironment are barriers to chemotherapy, radiotherapy, and immunotherapy. Diffusion barriers and drug resistance make chemotherapeutic drugs less effective. Decreased reactive oxygen species (ROS) levels lead to radiotherapy resistance. Infiltration and accumulation of suppressor T cells inhibit immunotherapy [1], [6]. Several methods have been developed to overcome hypoxia and hypoperfusion, including vascular normalization, hypoxia-targeted drugs, hypoxia-activated prodrugs (HAPs), and direct O₂ delivery into tumors [6]. To date, no approach has been approved by the FDA to effectively reverse tumor hypoxia. Further studies are required to resolve hypoxia and hypoperfusion.

We found that a low mechanical index (MI) diagnostic ultrasound (DUS) combined with microbubbles (contrast agent) could improve tumor blood perfusion, called “sononeoperfusion effect” [7]. Enhanced tumor perfusion improves the therapeutic effect. In chemotherapy, the drug concentration in tumor tissue was elevated by up to 3.12-fold in rat Walker-256 breast cancer (doxorubicin) and 2.83-fold in rabbit VX2 tumors (gemcitabine) [7], [8], and tumor growth was inhibited in rabbit VX2 tumors [9]. In immunotherapy, the sononeoperfusion effect resulted in tumor growth inhibition and longer survival in mice with MC38 colon cancer [10]. In radiotherapy, ultrasound combined with microbubbles enhances radiosensitivity of human U251 glioblastoma cells and mouse GL261 tumors [11]. This is a promising and noninvasive approach for tumor therapy. The required ultrasound intensity was extremely low according to the FDA and IEC guidelines. Our previous studies have primarily found that the expression of nitric oxide (NO), endothelial nitric oxide synthase (eNOS), and adenosine-triphosphate (ATP) were increased and hypoxia-inducible factor-1α (HIF-1α) was decreased post treatment [7], [8]. Therefore, the effect of sononeoperfusion may be related to NO release, leading to the hypoxia amelioration. This study investigated whether ultrasound and microbubble (USMB) treatment ameliorated tumor hypoxia.

Herein, we aimed to further investigate the mechanism and characteristics of the sononeoperfusion effect and explore the tumor hypoxia amelioration after single or multiple USMB treatments. Several hypoxic and related factors were detected during and post treatment.

2. Methods

2.1. Ultrasound system

A commercial DUS system with an X4-12L linear array transducer (VINNO70, VINNO Technology Co. Ltd., Suzhou, China) was used for therapeutic ultrasound exposure and ultrasound imaging as described previously [9]. For pre- and post-treatment ultrasound imaging, the contrast-enhanced ultrasound (CEUS) mode is integrated with an analysis software that can evaluate the contrast intensity to represent blood perfusion by area under the curve (AUC), peak intensity (PI), and so on. For ultrasound treatment, the VINNO70 was equipped with a modified flash mode (Vflash mode) for microbubble destruction. The Vflash mode can regulate microbubble cavitation by adjusting several parameters, including frequency, MI, pulse length (PL), pulse repetition frequency (PRF), and destruction (on)/replenishment (off) time. It provides alternative emissions of DUS and treatment ultrasound to ensure sufficient cavitation nuclei for microbubble destruction and ensures that the treatment is conducted in the same plane for treatment accuracy. Additionally, ultrasound could be weakly focused on the region of interest (ROI) to reduce its impact on the surrounding tissue (Figure 1F). A membrane hydrophone (HMB-0500, ONDA Corp., Sunnyvale, CA, USA) was used to measure the peak negative pressure (PNP) at different MI. The probe was placed 2 cm away from the hydrophone in a scanning tank (AIMS III, ONDA Corp., Sunnyvale, CA, USA) filled with degassed water (Supplementary Figure 1).

Fig. 1.

Flowcharts of experimental procedure. A: Single treatment for bilateral tumors and NO detection; B: eNOS inhibition of unilateral tumors before treatment; C: Single treatment for unilateral tumors and follow-up detection; D: Three treatments for unilateral tumors and follow-up detection; E: Real-time tumor pO2 monitor for each treatment. F: Diagnostic ultrasound sequences and therapeutic sequences alternatively emit in VINNO 70 Vflash mode for treatment. Dark blue refers to CEUS, light blue refers to USMB treatment, red refers to eNOS inhibition, yellow refers to tumor pO₂ monitor.

2.2. Contrast agent

A lab-made Zhifuxian microbubble was used as a contrast agent for CEUS, and cavitation nuclei were used for ultrasound treatment [12]. It is a lipid-coated microbubble suspension with a perfluoropropane gas core. The concentration was 2–9 × 10⁹/mL and the average microbubble diameter was 1.92 ± 0.52 μm. Microbubbles were administered through the mice tail vein using a 26G needle. For CEUS, 0.01 mL original microbubble suspension was injected. For treatment, 0.01 mL original microbubble solution was diluted with saline (0.5 mL) and slowly injected for 10 min.

2.3. Animal model

A total of 119 C57 mice (male and female, 6–8 weeks old) with MC38 colon cancer were used in this study. Mice were subcutaneously injected unilaterally or bilaterally with 0.1 mL MC38 cell suspension (1–5 × 10⁷/mL) subcutaneously in the thigh. Transplanted tumors with a maximum diameter of approximately 1 cm were adopted after seven days. All the mice were euthanized with excess CO₂ at the end of the experiment. All animal studies were approved by the Laboratory Animal Welfare and Ethics Committee of Army Medical University.

2.4. Experimental procedure

Mice were randomly divided into three experimental (MI = 0.1, 0.3, and 0.5) and a control group. All the mice underwent CEUS before and after treatment. Therapeutic ultrasound was conducted at 3 MHz frequency with 5 cycles of PL, 2000 Hz of PRF, and 1 s (on)/ 1 s (off). During the 10 min treatment, the tumor center was 2 cm away from the transducer surface using a gel pad. The X4-12L transducer is 3.8 cm long, which could encompass the maximum length of the tumor (approximately 1 cm) in this study. The ROI was extended 1 cm beyond the tumor border to encompass the entire tumor. To ensure proper tumor coverage, the transducer was rotated approximately 45° clockwise every minute.

Among them, 28 mice with bilaterally transplanted tumors (eight in each experimental group and four in the control group) underwent CEUS on both sides of the tumor. One of the two tumors was selected as the therapeutic side and the other as the control side. The control side was blocked with tinfoil, and DUS could not visualize the tumor. NO concentration in the tumor was measured after the experiment (Figure 1A). Fifteen mice with unilaterally transplanted tumors (experimental group only, five in each group) underwent CEUS and treatment post eNOS inhibition (Figure 1B). Twenty mice with unilaterally transplanted tumors (five mice in each group) received single treatment and HIF-1α, ATP, and D-lactate (D-LA) were detected 1 h post treatment (Figure 1C). Fifty-six mice with unilaterally transplanted tumors (14 mice per group) underwent CEUS and treatment thrice on the first, third, and fifth day. Among 56 mice, eight mice in each group were detected with HIF-1α and D-LA 1 h after the last treatment (three mice in each group underwent hypoxyprobe immunofluorescence staining), and six mice in each group had real-time tumor oxygen partial pressure (pO₂) measurements (two of them continuously monitored pO₂ 40 min after the end of treatment) (Figure 1D-E).

2.5. Assessment of tumor perfusion

One-minute CEUS was conducted on the same section before and after treatment. The tumor area was selected to obtain the time-intensity curve (TIC) using VINNO70 built-in analysis software. PI and AUC were computed automatically. The tumor perfusion area was computed using FIJI (2.3.0/1.53 t, Fiji: an open-source platform for biological-image analysis) (tumor perfusion area = contrast agent perfusion area/tumor total area) according to the CEUS image of the PI (Supplementary figure 2A-D). According to the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB) guidelines, the AUC and PI computed from the TIC are qualified to represent the tumor perfusion condition [13]. The perfusion increase ratio reflects the increase in tumor perfusion and is expressed as X increase ratio = (Xpost - Xpre) / Xpre (X: AUC, PI, tumor perfusion area).

At the end of the study, we summarized the characteristics of sononeoperfusion effect. One hundred and forty-eight sononeoperfusion effects in MC38 tumors (37 sononeoperfusion effects in each group) were evaluated, including 28 bilateral tumor-bearing mice (therapeutic tumors of the experimental group, n = 24, and bilateral tumors of the control, n = 8) and 52 unilateral tumor-bearing mice (20 tumors with single treatment and 32 tumors with multiple treatments, enumerated as 96 sononeoperfusion effects, n = 116). A PI ratio above 1.1 was considered a significant enhancement, between 1 and 1.1 is considered a mild enhancement, and less than 1 was basically unchanged.

2.6. NO/ATP detection and inhibition

The tumor tissues were ground and lysed 1 h post treatment, and the total NO or ATP content in the tumor tissues was measured using a Total NO Assay Kit (Beyotime Biotechnology, Shanghai, China) or a Mouse ATP ELISA Kit (Jiangsu Meimian Industrial Co., Ltd, Jiangsu, China). For NO inhibition, 75 μg/kg N^G-Nitro-L-arginine Methyl Ester, Hydrochloride (L-NAME, Beyotime Biotechnology, Shanghai, China) was administered intraperitoneally injection (IP) 30 min prior to the study to inhibit eNOS.

2.7. Tumor hypoxia detection

We used a hypoxia-inducible factor, hypoxia fluorescence probe, and pO₂ probe to detect tumor hypoxia.

Tumor tissues were homogenized in saline for 1 h post treatment. The particulates were removed by centrifugation and assayed. Using mouse HIF-1α ELISA Kit (Jiangsu Meimian Industrial Co., Ltd, Jiangsu, China), the standard curve and HIF-1α content were gained.

The Hypoxyprobe™-1 Plus Kit (Hypoxyprobe, Inc., Burlington, USA), containing the hypoxic marker pimonidazole HCl and an anti-FITC secondary reagent, was used for the immunochemical detection of cell and tissue hypoxia. The mice were intravenously infused with pimonidazole HCl solution (20 mg/mL, 3 mL/kg) before treatment and perfused with saline 1 h after the end of the last treatment. Immunofluorescence was conducted on formalin-fixed, paraffin-embedded tumor sections using an anti-FITC secondary antibody, CD31 (Wuhan Servicebio Technology Co. Ltd, Wuhan, China), and Hoechst 33,342 (Wuhan Servicebio Technology Co. Ltd, Wuhan, China).

Tissue pO₂ detection instrument (Tow-Int Technology Co., Ltd., Shanghai, China) uses a chemiluminescent microsensor to detect tissue pO₂. Before the first CEUS, a sensor-protecting needle was inserted into the center of the tumor and guided by ultrasound (Supplementary figure 1E-F). The sensor fiber was then screwed out. The average pO₂ value for 20 s was recorded as the primary tumor pO₂. The tumor pO₂ curve was continuously recorded during the 10-minute treatment. After the second CEUS scan, the average pO₂ value for 20 s was recorded as post-pO₂. Additionally, the two mice were continuously monitored for tumor pO₂ for 40 min post treatment.

2.8. Tumor lactate detection

The tumors were harvested at the end of the third treatment period. Tumors were homogenized with saline, and the supernatant was extracted to detect the D-LA content in tumor tissues using a Mouse D-LA ELISA Kit (Jiangsu Meimian Industrial Co., Ltd., Jiangsu, China).

2.9. Statistics

Statistical analyses were conducted using SPSS software (version 26.0; SPSS, Inc., Chicago, IL, USA). The results were considered significantly different when P < 0.05. Data were demonstrated as mean ± SD, and the graphs were demonstrated as bar charts or box plots. Independent data were evaluated using one-way analysis of variance (ANOVA) if the variance was homogeneous and followed a normal distribution, and, if not, using a nonparametric statistical method (Kruskal-Wallis test). Paired data were evaluated using a paired t-test if they followed a normal distribution or a nonparametric paired t-test if they did not. Repeated measurement data were evaluated using repeated-measures measures ANOVA.

3. Results

There were no significant differences in tumor size, AUC, PI, or perfusion area before treatment among the groups (P > 0.05). The respective PNP was 0.16 MPa (MI = 0.1), 0.43 MPa (MI = 0.3), and 0.7 MPa (MI = 0.5).

3.1. Single and multiple sononeoperfusion effects

For the single sononeoperfusion effect of bilaterally transplanted tumors, the mice received a single USMB treatment on a random side of the tumors. When MI = 0.3, for the treated side, the PI increased by 27.15 %, the AUC increased by 29.85 %, and the perfusion area increased by 64.58 % post treatment (P < 0.05); the PI increased by 13.55 %, the AUC increased by 15.86 % of the control side post treatment (P < 0.05), and no significant difference was observed in the perfusion area of the control side (P > 0.05). When MI = 0.1/0.5 in the control group, there was no statistically significant difference between the two sides for the AUC, PI, and perfusion area before and after treatment (P > 0.05) (Figure 2A-D). For tumor perfusion enhancement between the two sides, the AUC and PI increase ratios of the treated side were higher than those of the control side at MI = 0.3 (P < 0.05). The perfusion area increase ratio of the treated side was higher than that of the control side at MI = 0.5 (P < 0.05). No statistical differences were found among the other groups (P > 0.05) (Figure 2E-G).

Fig. 2.

Tumor perfusion and NO changes post single treatment for bilateral MC38 tumors (n = 28, 8 in each treatment group, 4 in the control group). A: B-Mode and CEUS images of tumors Pre- and Post-treatment. Red circle refers to tumor of the treated side, yellow circle refers to tumor of the control side. Post treatment, tumor perfusion of the treated side is increased especially at MI = 0.3, and the control side of treatment groups also increases slightly. B-D: Box plots of PI, AUC and perfusion area indicates that tumor perfusion is significantly enhanced post treatment for both sides at MI = 0.3. E-G: PI, AUC and perfusion area increase ratio indicate that the sononeoperfusion effect of the treated side is more obvious when MI = 0.3 compared with that of the control side. H: NO concentration is elevated post treatment at MI = 0.3/0.5. *P < 0.05, **P < 0.01.

For the single sononeoperfusion effect on unilaterally transplanted tumors, mice received a single USMB treatment. When MI = 0.3, the PI increased by 23.59 %, the AUC increased by 20.16 %, and the perfusion area increased by 112.5 % post treatment, indicating a sononeoperfusion effect (P < 0.05). No significant differences were observed among the other groups (P > 0.05) (Figure 3A-E). For tumor perfusion enhancement among the groups, the AUC increase ratio, PI increase ratio, and perfusion area increase ratio at MI = 0.3 were higher than those at MI = 0.5, and the control (P < 0.05) and perfusion area ratio at MI = 0.3 was higher than that at MI = 0.1 (P < 0.05). The AUC ratio at MI = 0.1 was higher than that at MI = 0.5 and the control (P < 0.05) (Figure 3F-H).

Fig. 3.

Tumor perfusion and hypoxic factors changes post single treatment for unilateral MC38 tumors (n = 20). A: B-Mode images and CEUS images of one tumor from the treatment group at different time points pre- and post-treatment. The preliminary perfusion is similar. Post treatment, tumor perfusion is significantly enhanced. B: B-Mode and CEUS images of tumors Pre- and Post-treatment. Red circle refers to tumors. Tumor perfusion increases especially at MI = 0.3, while the control did not increase. C-E: Box plots of PI, AUC, and perfusion area indicate that tumor perfusion is significantly enhanced post treatment at MI = 0.3. F-H: PI, AUC, and perfusion area increase ratio indicate that the sononeoperfusion effect is more obvious when MI = 0.3 compared with the other two treatment groups and the control group. I-K: There is no significant difference for ATP, D-LA, and HIF-1α level post treatment. *P < 0.05, **P < 0.01.

For multiple sononeoperfusion effects, mice with unilateral tumors were treated thrice with USMB. As the tumor progressed, the tumor center gradually developed hypoperfusion. At MI = 0.3, PI increased by 12.71 % and 13.50 %, AUC increased by 15.46 % and 18.35 %, and perfusion area increased by 7.87 % and 24.32 % after the first and third treatments, respectively (P < 0.05). At MI = 0.5, PI increased by 8.68 %, AUC increased by 9.63 % post the first treatment, and AUC after the second treatment significantly increased by 7.46 % (P < 0.05). In the control, the perfusion area after the first treatment, AUC after the second treatment, and PI after the third treatment were significantly lower than those before treatment (P < 0.05) (Figure 4A-D). For tumor perfusion enhancement, compared with the control, the PI increase ratio, AUC increase ratio, and perfusion area increase ratio were higher after every treatment at MI = 0.3 (P < 0.05). The PI increase ratio was higher after the third treatment at MI = 0.1 (P < 0.05). The perfusion area increase ratio was higher after the first and third treatments at MI = 0.5 (P < 0.05) (Figure 4E-G).

Fig. 4.

Tumor perfusion changes post three unilateral MC38 tumor treatments (n = 32). A: B-Mode and CEUS images of tumors Pre- and Post-treatment. Red circle refers to tumors. Tumor center tend to be hypoperfused with development. When MI = 0.3 or 05, tumor perfusion increases several times, while the control does not increase. B-D: Box plot of PI, AUC, and perfusion area indicate that tumor perfusion is significantly enhanced post treatment at MI = 0.3 or 0.5. E-G: For three treatments, PI, AUC, and perfusion area ratio indicate that the sononeoperfusion effect is more obvious when MI = 0.3 or 0.5 compared with the control. *P < 0.05, **P < 0.01, ***P < 0.001.

3.2. Characteristics of sononeoperfusion effect

The sononeoperfusion effect can be characterized as: 1) an increase in tumor perfusion area post treatment, 2) enhancement of tumor vasculature post treatment, and 3) an increase in tumor CEUS intensity post treatment. Although there was no obvious perfusion area enhancement at the CEUS peak intensity, the brightness increased. 4) Delay in CEUS fading post treatment. TIC decreased slowly (Figure 5A). The scatter plot demonstrated that the lower the initial perfusion area, the more significant the perfusion enhancement in the blood flow (Figure 5B).

Fig. 5.

Characters of sononeoperfusion effect. A: Four types of sononeoperfusion effect. Red arrow refers to the tumor vessel. Blue arrow indicates the time-intensity curve post treatment decrease more slowly. B. Scatter diagram of Perfusion area and PI ratio. Tumors with less initial perfusion are more easily to be enhanced. C: Heat Map for PI, AUC, and perfusion area ratio of all the MC38 tumors. Tumor perfusion enhancement is more obvious at MI = 0.3.

At MI = 0.3, 59.46 % (22/37) of the tumors were significantly enhanced, 29.73 % (11/37) were mildly enhanced, and 10.81 % (4/37) remained basically unchanged. When MI = 0.1, 32.43 % (12/37) of the tumors were significantly enhanced, 24.32 % (9/37) were mildly enhanced, and 43.24 % (16/37) remained basically unchanged. When the MI was 0.5, 32.43 % (12/37) were significantly enhanced, 40.54 % (15/37) mildly enhanced, and 27.03 % (10/37) remained basically unchanged. In the control group, 13.51 % (5/37) were significantly enhanced, 27.03 % (10/37) were mildly enhanced, and 59.46 % (22/37) remained basically unchanged (Figure 5C).

3.3. Tumor hypoxia improvement

For HIF-1α, there was no statistically significant difference among the group post a single treatment (P > 0.05) (Figure 3 K). However, the HIF-1α of three treatment groups significantly decreased after three treatments compared with the control group, MI = 0.3 (42.22 ± 9.06 pg/mL) < MI = 0.5 (50.41 ± 13.49 pg/mL) < MI = 0.1 (50.60 ± 10.97 pg/mL) < Control (62.39 ± 11.89 pg/mL) (P < 0.05) (Figure 6A).

Fig. 6.

Real-time and subsequent hypoxia changes among groups. A-B: Box plot of HIF-1α and D-LA. HIF-1α decrease in all three groups. D-LA decrease when MI = 0.3/0.5. C: Line chart of tumor tissue pO₂. When MI = 0.3/0.5, the hypoxia is slightly reversed but with no significance. D-E: Schematic and histogram of real-time pO₂ changes for three treatments. For all the treatment groups, pO₂ of tumor tissue significantly increase during 10-minute treatment, while pO₂ of the control remain the same or slightly decrease. F: Tumor tissue pO₂ changes for three treatments are significant. G: Tumor tissue pO₂ continuously increase post treatment. *P < 0.05, **P < 0.01, ***P < 0.001.

The immunofluorescence fluorogram demonstrated that areas with more surviving cells (Hoechst 33342, green) were less hypoxic (hypoxy probe, yellow), and areas with fewer surviving cells were more hypoxic. The hypoxyprobe fluorescence was more obvious at MI = 0.1 and the control compared with MI = 0.3 and 0.5 (P < 0.05), indicating that hypoxia was more serious in the control and MI = 0.1 groups. There was no significant difference in the mean fluorescence intensity of CD31 among the groups (P > 0.05) (Figure 7A-C).

Fig. 7.

A-B: Overview and details of triple staining immunofluorescence. A: 1.3×, B: 10×. CD31 (red), hypoxyprobe (yellow), and Hoechest 33,342 (green) are merged to represent location of endothelial cell, hypoxia, and tumor cells. When MI = 0.3/0.5, yellow is less than the other, it indicates hypoxia improvement. C. Mean fluorescence intensity of hypoxyprobe. *P < 0.05, **P < 0.01.

Tumor pO₂ changes were detected using a tissue pO₂ detection instrument in real time. The initial pO₂ in each group was not significantly different (P > 0.05). During the 10 min treatment, the tumor pO₂ gradually increased in each treatment group. Tumor pO₂ in the control group remained basically unchanged or decreased (Figure 6D). When MI = 0.3, the tumor pO₂ increased significantly for each of the three treatments. At MI = 0.1/0.5, an evident increase was observed in the first and second treatments. The pO₂ of the control group decreased significantly during the first treatment (P < 0.05) (Figure 6E). Post treatment, the tumor pO₂ continuously increased for another 40 min and was maintained at a higher level (Figure 6G). For the changes in tumor pO₂, all three treatments at MI = 0.3, the second and third treatments at MI = 0.5, and the third treatment at MI = 0.1 were significantly higher than those of the control (P < 0.05) (Figure 6F). During the three treatments, the tumor pO₂ gradually decreased in the MI = 0.1 or the control groups and gradually increased at MI = 0.3/0.5, but the difference was not statistically significant (P > 0.05) (Figure 6C).

3.4. Decrease in tumor lactate metabolite

There was no statistically significant difference in the D-LA among the groups post a single treatment (P > 0.05) (Figure 3 J). However, the D-LA at MI = 0.3/0.5 significantly decreased after three treatment compared with the control group (P < 0.05), MI = 0.3 (55.43 ± 17.09 μmol/L) < MI = 0.5 (56.35 ± 15.81 μmol/L) < MI = 0.1 (68.47 ± 11.13 μmol/L) < Control (75.22 ± 8.52 μmol/L) (Figure 6B).

3.5. NO/ATP release and sononeoperfusion effect

At the end of the single treatment, NO or ATP concentrations were measured on both sides of the bilateral tumor-bearing mice. When MI = 0.3 and 0.5, the NO concentration in the treated tumors was significantly higher than that in the control tumors (P < 0.05), while there was no significant difference between the treated and control sides in the MI = 0.1 and control group (P > 0.05) (Figure 2H). Inhibition of eNOS was before treatment. There were no significant improvements in the PI, AUC, or perfusion area post treatment in any treatment group (P > 0.05) (Figure 8A-D). Tumor pO₂ hardly increased post treatment (Figure 8E) and remained unchanged for 30 min (Figure 8F).

Fig. 8.

A: B-Mode and CEUS images of tumors Pre- and Post-treatment post eNOS inhibition. Red circle refers to tumors. Tumor perfusion do not increase by USMB post eNOS inhibition. B-D: Box plot of PI, AUC, and perfusion area indicate that USMB cannot enhance tumor perfusion post eNOS inhibition. E. Tumor tissue pO₂ remained basically unchanged during 10-minute treatment post eNOS inhibition. F. Tumor tissue pO₂ remains basically unchanged for 30 min post eNOS inhibition.

4. Discussion

In this study, we applied a novel approach using DUS and microbubbles to enhance tumor blood perfusion, called “sononeoperfusion effect.” It improved pO₂ of the tumor tissue, decreased HIF-1α and D-LA levels, and improved the hypoxic microenvironment by releasing NO (Table 1). Our previous studies found that sononeoperfusion can enhance tumor perfusion and therapeutic outcomes in several animal models, including rat Walker-256 breast cancer, rabbit VX2 tumor, nude mouse PANC-1 pancreatic cancer, and mouse MC38 colon cancer [7], [8], [10], [14]. Remarkably, we found that HIF-1α concentration decreased post USMB in rabbit VX2 tumor [8]. Therefore, this study further investigated hypoxia improvement post single or multiple USMB treatments in real time and over a longer period. The sononeoperfusion effect may enhance tumor therapy by improving perfusion and ameliorating hypoxia.

Table 1.

Overall comparisons.

		MI = 0.1	MI = 0.3	MI = 0.5	Control
Single USMB Treatment	Perfusion	↑		↑	○
	HIF-1α	○	○	○	○
	D-LA	○	○	○	○
	Real-time pO2	○		↑	○
	NO	○	↑	↑	○
Multiple USMB Treatments	Perfusion	○		↑	○
	HIF-1α	↓		↓	○
	D-LA	○	↓	↓	○
	pO2 Trends		○	○
	Hypoxyprobe	○			○
NO Inhibition	Perfusion	○	○	○	○
	pO2	○	○	○	○

Summary of all changes observed post single or multiple USMB treatments. For perfusion, indicates significant increase in all PI, AUC, and perfusion area. ↑ indicates significant increase in any of the PI, AUC, or perfusion area. For the other indexes, ↑ or ↓ indicates P < 0.05, or indicates P < 0.01, ○ indicates P > 0.05, and indicates the decreasing trend with no significance.

In a previous study, USMB at MI = 0.3 could enhance tumor perfusion and increased drug concentration in nude mice bearing pancreatic cancer among three different MIs (MI = 0.3, 0.7, 1.1) [14]. CEUS (MI = 0.08) can sometimes stimulate tumor perfusion enhancement; hence, we included DUS of lower and higher intensity around MI = 0.3 to further explore the appropriate parameter. The results demonstrated that sononeoperfusion effect could be stimulated at an MI as low as 0.1 (0.16 MPa). However, we observed a more stable sononeoperfusion effect at an MI of 0.3 (0.43 MPa). The sononeoperfusion effect may be related to stable cavitation with no thermal effect at an MI under 0.5 [15]. With an increase in the ultrasound intensity, the microbubble oscillation becomes more obvious. Ultrasound intensity from 0.8 to 6.9 MPa led to cavitation increase in phantom experiments for 50 times [16]. At MI = 0.5, stable cavitation occurred, but with stronger oscillations and bioeffects, which may not be conducive to perfusion enhancement. Higher ultrasound > 3.0 MPa significantly depletes the neovasculature and reduces tumor perfusion [17]. Therefore, an appropriate intensity for a suitable cavitation is important for the sononeoperfusion effect. These results are consistent with previous studies that demonstrated that low-intensity ultrasound (below 0.4 MPa) enhanced drug delivery [18]. A clinical trial using DUS at MI = 0.2 enhanced gemcitabine treatment of pancreatic cancer, which increased survival and quality of life [19].

In addition to cavitation, USMB enhances perfusion via other mechanisms. Lindner found that NO release is related to muscles sonoreperfusion [20]. Our previous study also found increased NO and eNOS post USMB [7]. This study demonstrated that the NO concentration significantly increased post USMB treatment (Figure 2H). Post eNOS inhibition, USMB did not stimulate sononeoperfusion or ameliorate hypoxia (Figure 8). Thus, NO release by the USMB induced the sononeoperfusion effect. The sononeoperfusion effect is closely related to hypoxia amelioration. Interestingly, the NO concentrations and tumor perfusion on the control side in the treatment group were slightly higher than those in the control group, and the control side also demonstrated perfusion enhancement (Figure 2E-H). This indicated that the bioeffect of sononeoperfusion effect influenced not only the ultrasound exposure loci, but also the downstream. To avoid downstream effects, we used unilateral tumor-bearing mice.

Multiple treatments stimulated sononeoperfusion effects, indicating that the tumors were not resistant to multiple sononeoperfusion effects. Clinically, tumors often require multiple treatments. USMB can be conducted depending on the duration and frequency of treatment. This indicates that the sononeoperfusion effect is conducive to clinical translation and personalized treatment.

Locally advanced solid tumors tend to develop hypoperfusion owing to insufficient development and functional abnormalities of the vasculature. Tumor hypoperfusion not only impedes therapeutic drug delivery and accumulation but also leads to a hypoxic and acidic tumor microenvironment [4], [21]. Hypoxia leads to tumor proliferation, invasion, low pH, and therapeutic resistance [2]. In this study, each sononeoperfusion effect induced a slight but steady increase in the tumor tissue pO₂. The tumor pO₂ continuously increased post USMB treatment and remained at a higher level. This may be due to enhanced tumor perfusion, resulting in an increase in oxygen supply to the original hypoperfused area (Figure 9). The sononeoperfusion effect can last for more than 4 h [7], which elucidates continuous pO₂ increase. Beside the direct index pO₂, HIF-1α is a regulator of cellular responses to hypoxia and plays a crucial role in tumor angiogenesis. These findings have prognostic implications for clinical outcome [22]. D-LA- and pimonidazole-labeled hypoxic immunofluorescence reflect the hypoxic microenvironment [5], [23]. Single sononeoperfusion effect demonstrated little effect on HIF-1α, D-LA. However, the three sononeoperfusion effects led to a decrease in HIF-1α, D-LA and immunofluorescence. Multiple USMB treatments can improve the therapeutic outcomes. USMB slightly reversed the tumor pO₂ trend, but with no significance (Table 1). Increased USMB treatment might further ameliorate tumor hypoxia.

Fig. 9.

Schematic illustration of tumor hypoxia amelioration by sononeoperfusion effect. Red cells refer to normal tumor cells, yellow cells to hypoxic tumor cells, and blue circles to microbubbles. Tumor hypoperfusion lead to the imbalance of oxygen demand and supply, resulting in tumor hypoxia. Post ultrasound and microbubble treatment, microbubble oscillation and NO release induce sononeoperfusion effect. Tumor perfusion and vascular permeability enhancement increase oxygen supply.

Increasing pO₂ in hypoxic tumor regions is the most direct approach for tackling hypoxia. Several methods have been used, including increasing the oxygen carrier hemoglobin concentration, erythropoietin, and hyperbaric oxyge. However, none of these methods have achieved satisfactory clinical outcomes [2]. Furthermore, studies demonstrated that hyperthermia (40–45 ℃) could enhance tumor perfusion and thus increase oxygen delivery [24], [25]. However, it is difficult to achieve cytotoxic temperatures, and normal tissues are not thermotolerant. It is difficult to transfer hyperthermia to the clinic. Similarly, sononeoperfusion effect can improve tumor perfusion and overcome hypoxia. Moreover, it is required within FDA guidelines and clinical microbubbles, which are easy to translate clinically with fewer side effects. Therefore, repeated sononeoperfusion effect is a promising approach for reversing tumor hypoxia.

5. Conclusion

In conclusion, this study has demonstrated that DUS at appropriate MI = 0.3 and free microbubbles could improve tumor blood perfusion (called “sononeoperfusion effect”) and ameliorate tumor hypoxia through NO release. This combination treatment could be a novel and promising approach for improving clinical efficacy and overcoming resistance to tumor therapy.

CRediT authorship contribution statement

Yi Zhang: Conceptualization, Formal analysis, Investigation, Methodology, Resources, Validation, Visualization, Writing – original draft, Writing – review & editing. Jing Zhang: Conceptualization, Data curation, Formal analysis, Investigation, Software. Tingting Luo: Formal analysis, Methodology, Resources. Zhiping Cai: Data curation, Software. Guoliang Yang: Investigation, Resources. Hui Li: Investigation, Software. Junshuai Wei: Investigation, Validation. Qiong Zhu: Formal analysis, Funding acquisition. Peijing Li: Funding acquisition, Project administration, Supervision. Xiaoxiao Dong: Funding acquisition, Project administration, Resources, Supervision, Validation, Writing – review & editing. Zheng Liu: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.