Ultrasound molecular imaging of M2 macrophages for early detection of chronic rejection in heart transplantation

Abstract

Chronic rejection (CR) remains the leading cause of morbidity and mortality in heart transplantation survivors. The primary pathological features of CR encompass cardiac allograft vasculopathy and myocardial fibrosis. Currently, its diagnosis heavily relies on invasive procedures, underscoring the pressing need for non-invasive evaluation methods. This study introduces a novel approach utilizing mannose-modified microbubbles (MB_man) targeting CD206 (mannose receptor) positive M2 macrophages for early CR detection. In vitro experiments demonstrate substantial adhesion of MB_man to M2 macrophages compared to common microbubbles (MB_con). In a CR rat model, MB_man and MB_con are administered at three distinct time points (2 weeks, 4 weeks, and 6 weeks), followed by contrast-enhanced ultrasound images and quantitative analysis using the ultrasound destruction-supplementation method. Starting at 2 weeks and continuing through 6 weeks, MB_man demonstrates significantly higher signal intensity than MB_con in allograft rats. However, this difference is not observed in isograft rats at any of the indicated time points. These findings suggest an increase in M2 macrophage infiltration in allografts compared to isografts. Furthermore, the signal intensity of MB_man positively correlates with the percentage of CD206 in allograft rats. This study proposes a promising approach, simultaneous noninvasive ultrasound molecular imaging, for the early-stage evaluation of CR.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-03672-9.

Introduction

Heart transplantation (HT) stands as a transformative treatment for individuals grappling with end-stage heart failure [1]. However, chronic rejection (CR) remains a major challenge following HT, often leading to graft failure and patient mortality [2, 3]. CR is a gradual or recurring response of the graft to the recipient that occurs months or years after HT. It typically manifests as chronic inflammation, stenosis or occlusion of the cardiac vessels, and myocardial fibrosis, ultimately leading to loss of graft function. This is usually caused by a low-level immune response with specific antibodies in the circulation, leading to perivascular inflammation and sustained endothelial damage in the graft vasculature, accompanied by the proliferation of vascular smooth muscle cells blocking the vessels and ultimately leading to a gradual decline in graft function [4]. The primary pathological features of CR encompass cardiac allograft vasculopathy (CAV) and myocardial fibrosis, with most grafts failing within a decade post-transplantation [5–7].

According to the International Society for Heart and Lung Transplantation, coronary angiography is the cornerstone of CAV diagnosis. Additionally, intravascular ultrasound imaging is also currently one of the most widely used and clinically accepted diagnostic tools [8]. However, both methods are invasive and typically detect CR at the mid-to-late stage [9], risking severe complications such as pericardial tamponade, myocardial infarction, stroke, and malignant arrhythmias [10]. Thus, contemporary research urgently seeks early, non-invasive, and efficient indicators for continuous CR monitoring post-HT.

Ultrasound molecular imaging (UMI) emerges as a promising solution. Microbubbles (MBs) with surface-specific ligands selectively accumulate at the target site via bloodstream circulation, reflecting molecular-level alterations at the lesion site [11]. It is deemed a promising diagnostic technique due to its real-time imaging, non-invasiveness, and absence of radiation risks. Hao et al. developed a novel UMI technique that utilizes biosynthetic gas vesicles (GVs) as carriers to enable real-time monitoring of the epithelial-mesenchymal transition status of tumor cells through the combination of antibodies targeting E-cadherin and N-cadherin, thereby effectively assessing the metastatic potential of tumors. The results suggest that this GVs-based UMI can significantly improve the imaging signals of tumor markers, providing a potentially powerful tool for early diagnosis and treatment of cancer [12]. Recently, Jin et al. developed a method based on targeted UMI to assess acute rejection (AR) by detecting granzyme B expression in a mouse model after HT, providing a potentially non-invasive method of monitoring for early diagnosis of cardiac transplant rejection [13]. The above studies have confirmed the application prospects of UMI in the early diagnosis and treatment of diseases. However, there is limited research about UMI in CR after HT. Hence, the application of UMI to visualize CR is feasible and important in monitoring its dynamic changes.

Macrophages play a pivotal role in graft rejection, and their infiltration closely correlates with the prognosis of grafts, particularly in predicting CR [14, 15]. Under diverse microenvironments, macrophages can differentiate into two distinct cell types: classically activated (M1) and alternatively activated (M2). M2 macrophages are particularly associated with microvascular lesions and myocardial fibrosis following HT [16, 17]. CD206, formally known as macrophage mannose receptor 1, is a type I transmembrane C-type lectin and a member of the mannose receptor (MR) family. It is expressed on the surface of M2 macrophages and can bind specifically to mannose [18]. As a cell surface receptor, it is an important marker of macrophage activation and functional status and plays a key role, especially in M2 macrophages, a class of immune cells with anti-inflammatory, tissue repair, and immunomodulatory functions. High expression levels of CD206 are a characteristic hallmark of M2 macrophages [19, 20]. Chen et al. successfully developed a novel bivalent mannosylated-targeting ligand that selectively targeted M2 macrophages and improved drug delivery efficiency, providing a new strategy for the treatment of related diseases [21]. A study developed a novel mannose-conjugated anti-biofouling magnetic iron oxide nanoparticle for specific targeting and imaging of M2 tumor-associated macrophages (TAMs) expressing CD206, providing an effective tool for studying macrophage dynamics in the tumor microenvironment and developing new therapeutic strategies [22]. All the above researches indicate that the expression level of CD206 can serve as a potential biomarker for assessing macrophage polarization status and monitoring disease processes.

Herein, we have developed M2 macrophage-targeted MBs modified by mannose, and the CR rat model was established after HT. MB_man showed excellent targeted adhesion capability towards M2 macrophages in vitro and in vivo. At the indicated time points, MB_man was injected into recipient rats. Quantitative analyses were performed using the destruction-replenishment method. The signal intensity (SI) of the attached MBs was assessed using online Time-Intensity Curve (TIC) analysis software (Fig. 1). Leveraging the characteristics of early extensive M2 macrophage infiltration during the onset of CR, we report for the first time that using UMI can assess the onset and progression of CR.

Fig. 1.

Schematic diagram of mannosylated microbubbles targeting M2 macrophages for monitoring chronic rejection (CR) in heart transplantation (HT). Mannose receptor (CD206) targeted microbubbles (MB_man) feature the conjugation of mannose to microbubbles via DSPE-PEG-Mannose synthesis, allowing for selective targeting. Through ultrasound molecular imaging, MB_man was used to monitor the surface receptor CD206 of M2 macrophages during the period of 2 to 6 weeks following HT, indicating the migration of M2 macrophages involved in CR

Materials and methods

Materials

The lipids, including 1,2-distearoyl-sn-glycero-3-phosphatidylcholine (DSPC), 1,2-distearoyl-sn-glycero-3-hosphoethanolamineN-[methoxy(polyethyleneglycol)−2000] (DSPE-PEG₂₀₀₀), were purchased from Sigma-Aldrich (St. Louis, MO, USA). DSPE-PEG-NH₂ was obtained from Aladdin (Shanghai, China). α-D-Mannopyranosylphenyl isothiocyanate (MPITC) was purchased from Chemsynlab (Beijing, China). Anhydrous dimethyl sulfoxide (DMSO) was purchased from Solarbio (Beijing, China). Dulbecco’s modified Eagle medium (DMEM) was purchased from Gibco (California, USA). IL-4 (214-14) and IL-13(210-13) were purchased from Peprotech (Rocky Hill, NJ, USA). APC anti-CD206 antibody (141708) was purchased from BioLegend (California, USA). Phosphate Buffer Saline (PBS) was obtained from Servicebio (Wuhan, China). 4’,6-diamidino-2-phenylindole (DAPI) was purchased from Beyotime (Shanghai, China). Cell Counting Kit-8 (CCK8) was purchased from Biosharp (Guangzhou, China). The rabbit monoclonal anti-CD206 antibody (224595) used in immunofluorescence was purchased from Cell Signaling Technology (Beverly, MA, USA). The rabbit polyclonal antibody anti-CD68 antibody (BA3638) was acquired from Boster Biological Technology (Wuhan, China). The rabbit monoclonal anti-CD31 antibody (ab182981) and anti-Zonula Occludens-1 (ZO-1) tight junction protein antibody (ab221547) were obtained from Abcam (Cambridge, UK).

Preparation of microbubbles

The common MBs (MB_con) were prepared as a control utilizing thin-film hydration. A phospholipid mixture of DSPC and DSPE-PEG₂₀₀₀ was dissolved in chloroform at a molar ratio of 9:1. Then, the chloroform in the phospholipid mixture was volatilized by a rotary evaporator under nitrogen flow, and a uniform thin phospholipid film was formed on the interior wall of the test tube. Subsequently, the excess chloroform was removed under vacuum for over 2 h. Following this, the film was hydrated by adding 5.5 mL aqueous solution to the test tube. The suspension was dispersed by sonication (PS40, Jeken, Dongguan, China) for 10 min at 60 ℃. Afterward, 1 mL of the solution was aliquoted into a vial and sealed with a lid. Perfluoropropane (C₃F₈) was used to replace the air in the vial. The vial was then agitated using a mechanical oscillator (AM-1, Monitex) for 30 s to generate bubbles. Next, the solution was centrifuged to isolate the supernatant MBs from the smaller nanobubbles settling in the lower layer. The centrifugation parameter was 300 rpm for 3 min.

To prepare mannosylated MBs (MB_man), mannose was selected as the ligand to be conjugated to the MBs. DSPE-PEG-Mannose was initially synthesized. DSPE-PEG-NH₂ was reacted with MPITC in DMSO overnight under the action of a magnetic stirrer at room temperature. The resultant liquid was dialyzed (MWCO 1000 Da) against ultrapure water for 24 h at room temperature to eliminate free MPITC, followed by freeze-drying. DSPE-PEG-Mannose’s conjugated structure was analyzed using ¹H nuclear magnetic resonance (NMR) spectroscopy (Bruker AV400, Switzerland) after dissolution in DMSO-d6. Then, DSPC, DSPE-PEG₂₀₀₀, and DSPE-PEG-Mannose prepared in the previous step were dissolved at molar ratios of 9:0.5:0.5 or 8.5:0.5:1 in chloroform. The subsequent preparation method of MB_man was the same as for MB_con. Store vials containing MBs at 4 °C before conducting the follow-up experiments.

DiI-labeled MBs and DiR-labeled MBs were prepared by adding 10 µL DiI or DiR (1.0 mg/mL) to the initial phospholipid mixture, with the entire preparation process conducted in the absence of light.

Characterization of MB_con and MB_man

The particle diameters and zeta potentials of both MB_con and MB_man were assessed by phase analysis light scattering (Nanobrook, Brookhaven Instruments, USA). The concentration of MBs was calculated with a hemocytometer. The appearance of images of MBs was obtained using an inverted microscope with phase contrast (Olympus IX73, Japan). The agarose mold was made to obtain ultrasound images in vitro. Different concentrations (1 × 10⁵, 5 × 10⁵, 1 × 10⁶, 5 × 10⁶, 1 × 10⁷/mL) of MB_con and MB_man with different DSPE-PEG-Mannose proportions (5%, 10%) were added into the wells of agarose mold. Then, the ultrasound images of MBs were taken with a commercial LOGIQ E9 ultrasound system (GE Healthcare, American). The most optimal proportion of DSPE-PEG-Mannose was identified as 5%; therefore, this proportion was used in the following experiments. The ultrasound imaging signal intensity of both MB_con and MB_man at a concentration of 1 × 10⁷/mL was quantitatively analyzed at 0, 10, 30, and 60 min to evaluate their contrast ability. The particle diameters and concentrations of MB_man were measured at 0 min, 30 min, 1 h, 2 h, 4 h, and 12 h post-oscillation to evaluate its stability.

M2 macrophage attachment studies

The Raw264.7 cells utilized in this study were procured from the Chinese Academy of Sciences Cell Bank and cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin at 37 °C in a humidified atmosphere with 5% CO₂. When cultured to 90% confluence, the cell suspension was planted in a 6-well plate at a density of 1 × 10⁶ cells per well. Following 24 h incubation, cells in the 6-well plate were incubated with 20 ng/mL IL-4 and 20 ng/mL IL-13 for 48 h to induce macrophage polarization. Afterward, the stimulated cells were gathered, and the level of the M2 macrophage marker (CD206) was assessed using flow cytometry with APC anti-CD206 antibody.

The prepared M2 macrophages were utilized to further evaluate the attachment capability of MB_man. In brief, the culture medium was aspirated, and either 3 × 10⁷ MB_man or MB_con was introduced onto the cell monolayer, followed by a 30 min incubation. Subsequently, unbound MBs were eliminated with a PBS wash. The attachment of MBs to M2 macrophages was assessed using an optical microscope (Olympus IX73, Japan). Additionally, confocal laser scanning microscope (CLSM) imaging was performed using DiI-MB_man or DiI-MB_con following the same incubation procedure. After being washed with PBS three times, the cells underwent fixation with 4% paraformaldehyde for 10 min. Next, the nuclei were stained with DAPI and scanned with a CLSM (LSM 780, Zeiss, Germany). To further confirm the conjugation rate of the MB_man to M2 macrophages, IL-4 + IL-13-stimulated cells were exposed to either 3 × 10⁷ DiI-MB_man or DiI-MB_con for 30 min. The cells were then subjected to PBS washes three times and analyzed by flow cytometry.

Heterotopic heart transplantation model

Male F344 and Lewis rats weighing between 200 and 250 g were obtained from Vital River Laboratory (Beijing, China) for this study. All procedures involving animals adhered to ethical standards for experimental research and were approved by the Institutional Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology.

The allograft group comprised donor hearts transplanted from F344 rats to Lewis rats, while the isograft group involved grafts from one Lewis rat to another. Recipient rats were fasted for 8 h before surgery. Heterotopic HT models were performed as previously described [23]. Anesthesia was induced in the rats with 3.5% isoflurane and maintained with 2% isoflurane via a mask. The donor rat received an intravenous heparin sodium saline dose of 2 mL (125 U/mL). Afterward, a thoracotomy was performed to expose the donor’s heart, followed by ligating the superior and inferior vena cava and dissecting the ascending aorta and pulmonary artery. The excised heart was immediately preserved in hypothermic saline. The recipient rat underwent a similar anesthesia procedure. An abdominal midline incision was performed, and the inferior vena cava and abdominal aorta were isolated and temporarily clamped. The donor pulmonary artery and ascending aorta were anastomosed with the recipient’s abdominal aorta and inferior vena cava, respectively. Successful graft implantation was confirmed by observing regular cardiac contractions postoperatively, followed by standard abdominal closure. To mitigate AR, recipient rats were administered FK506 (tacrolimus) prophylaxis consisting of daily subcutaneous injections (1 mg/kg body weight) for the initial 2-week postoperative period.

In vivo UMI

UMI was performed at three specific time points following HT: specifically, at 2 weeks, 4 weeks, and 6 weeks. The imaging procedure utilized a commercial ultrasound system (LOGIQ E9, GE Healthcare, American), along with a linear array transducer (ML6-15), securely held in place by a clamp holder. This setup was optimized initially, and imaging parameters were consistently kept during the entire process. Both contrast-mode and B-mode images were obtained side-by-side, with a mechanical index (MI) of 0.08. The destruction-replenishment method was employed to distinguish ultrasound signals originating from MBs adhering to local regions versus those in circulation within the bloodstream [24, 25]. 3 × 10⁸ MB_man or MB_con were bolus injected into the transplanted rat via the tail vein. At the 3 min mark, all the MBs in the graft were destroyed by a high-power flash pulse (MI = 0.25) for 1 s. Images lasting 10 s before and after the flash pulse were captured for further analysis. MB_man or MB_con was injected into the same rat in a randomized order. A 30-minute interval was kept between the injection of the two kinds of MBs to allow for the clearance of previously injected MBs.

Quantitative ultrasound image analysis

The ultrasound imaging data were analyzed using online software for TIC analysis. Under the guidance of B-mode images, the region of interest (ROI) was manually delineated, and the TIC was generated. The average ultrasound signal intensity (SI, expressed in dB) of both adherent and circulating MBs preceding the destruction pulse was calculated. After the flash, the average SI of circulating MBs was also analyzed. Ultimately, the SI of targeted MBs was determined as the difference between the average SI before and after the flash.

Ex vivo fluorescence imaging

DiR-MBs were used to investigate whether the concentration of MB_man was enhanced in transplanted hearts. 4 weeks after HT, 3 × 10⁸ DiR-MB_man or DiR-MB_con were injected into 4 allograft recipients and 4 isograft recipients through the tail vein. Afterward, the rats were euthanized 3 min after injection to harvest the grafts and main organs (hearts, livers, spleens, lungs, and kidneys). Following this, ex vivo imaging was conducted by a small animal living imaging system (IVIS, PerkinElmer Inc., USA) equipped with a 750 nm excitation filter and a 780 nm emission filter.

Histological assay

Cardiac grafts were obtained after finishing contrast imaging at the indicated time points for pathological assay. The sections were fixed in formalin and embedded in paraffin. Hematoxylin and eosin (H&E), Masson, and Elastica Van Gieson (EVG) staining were conducted to assess CR. The evaluation indices included the intima-to-media (I/M) ratio and fibrotic area ratio, as previously described [26]. The planimetric areas within the internal elastic lamina (IEL), the external elastic lamina, and the lumen were meticulously outlined and measured with an image analysis system (Image J 1.51). The I/M ratio was computed using the formula: I/M= (IEL area-lumen area)/(external elastic lamina area-IEL area). The fibrotic area ratio in grafts after transplantation was calculated using the following formula: (fibrotic areas/whole area) ×100% using the same image analysis system [26].

Immunofluorescence staining with anti-CD206 antibody and anti-CD68 antibody was conducted to identify the infiltration of M2 macrophages in the cardiac grafts. Quantification of CD206 was performed by calculating the percentage of positive area. Meanwhile, immunofluorescence staining was conducted with an anti-CD31 antibody to label vascular endothelium in the transplanted heart. Additionally, immunofluorescence staining with anti-ZO-1 tight junction protein antibody was performed. DAPI was used for nuclear staining. Mean fluorescence intensity (MFI) was calculated to quantify ZO-1. All the quantification was conducted with an image analysis system (Image J 1.51).

Safety evaluation

Cell viability assessment

Following the manufacturer’s instructions, cell viability was detected utilizing the CCK8. The H₉C₂ cell suspension was seeded in a 96-well plate at a density of 5.0 × 10³ cells per well in 100 µL. Different amounts of MB_man or MB_con (MBs/cells ratio = 20, 100, 500, 2000) were added into the wells. For each test ratio, three replicate wells were utilized. Following treatment for 12-24 h, 10 µL of CCK-8 reagent was added to each well and incubated for an additional 2 h. After that, the absorbance of each well at the wavelength of 450 nm was determined by a multimode microplate reader (Spark^®, TECAN, Switzerland), with wells without cells serving as blanks for analysis. Cell viability was calculated according to the formula in the manual.

In vivo biosafety evaluation

The experimental group of healthy SD rats received injections of either 3 × 10⁸ MB_man or MB_con, while an equivalent volume of PBS was injected into the control group. To conduct the blood routine examination, blood was collected in Ethylene Diamine Tetraacetic Acid (EDTA) spray-coated tubes 1 day and 7 days after injection, respectively.

Simultaneously, non-anticoagulant blood was gathered and centrifuged at 3000 rpm for 10 min to obtain plasma for biochemical analysis. The biochemical indices included alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (UREA), and creatinine (CREA). Additionally, primary organs (heart, spleen, kidney, liver, lung) were harvested for histological assessment through H&E staining, 1 day and 7 days after injection, respectively.

Statistical analysis

All data were expressed as mean ± standard error (SE) and analyzed by GraphPad Prism 8.0 software. Comparisons between groups were performed using Student’s t-test. One-way analysis of variance (ANOVA) was used for multiple groups. The Spearman correlation was applied to perform the correlation analysis. To ensure robustness and reliability, each assay was performed with at least 3 replicates. *P < 0.05, **P < 0.01, and ***P < 0.001 indicated that the difference is statistically significant.

Results

Characterization of MB_man and MB_con

In this study, mannose was conjugated to DSPE-PEG₂₀₀₀-NH₂ via a reaction between an isothiocyanate and the amine group to obtain DSPE-PEG-Mannose (Fig. 2A), enabling it to function as a targeted ligand for MR present in M2 macrophages. Figure 2B shows the ¹H-NMR spectrum of DSPE-PEG-NH₂. The ¹H NMR spectrum exhibited characteristic signals of DSPE-PEG-Mannose. The peaks observed at 7.0-7.5 ppm corresponded to the phenyl group in MIPTC (Fig. 2C). To identify the optimal proportion of DSPE-PEG-Mannose in MB_man, the contrast effect in vitro for MBs was assessed with varying DSPE-PEG-Mannose proportions, including 0, 5%, and 10%. Compared to MBs without DSPE-PEG-Mannose, the SI of MBs decreased with increasing DSPE-PEG-Mannose proportion (Fig. S1). Consequently, MBs with a 5% proportion of DSPE-PEG-Mannose were chosen for this study, as their contrast capacity closely resembled that of MBs without DSPE-PEG-Mannose.

Fig. 2.

Synthesis of DSPE-PEG-Mannose and characterization of MBs. (A) Schematic of synthetic DSPE-PEG-Mannose. (B) The ¹H-NMR spectrum of DSPE-PEG-NH₂. (C) The ¹H-NMR spectrum of DSPE-PEG-Mannose. (D) Bright-field photograph of MB_con and MB_man. Scale bar = 20 μm. (E) Particle size (n = 5) and (F) zeta potential (n = 3) of MB_con and MB_man. (G) The changes of the contrast effect at different time points (0 min, 10 min, 30 min, 1 h) and (H) the quantitative analysis of SI indicated that the two types of MBs exhibited no significant difference in contrast effect at different time points (n = 6)

Figure 2D depicts a photomicrograph of MB_con and MB_man under an inverted microscope with phase contrast, illustrating that both maintained the same spherical shape without conspicuous interconnection of particles. The particle diameters of MB_con and MB_man were 2.31 ± 0.089 μm and 2.04 ± 0.096 μm, respectively (Fig. 2E). The zeta potentials of MB_con and MB_man were -23.39 ± 1.02 mV and -20.67 ± 1.09 mV, respectively (Fig. 2F). The contrast effect of MB_con and MB_man was stable during 1 h (Fig. 2G and H). The initial concentration of MB_man was 6.1 × 10⁹/mL. To assess the stability of MB_man, we measured particle sizes and concentrations over time. The particle sizes of MB_man remained stable for 12 h. The concentrations of MB_man could be maintained at 80% after 12 h (Fig. S2). All the above results indicated the successful construction of MB_man.

Targeted adhesion capability of MBs towards M2 macrophages

Raw264.7 cells were stimulated with IL-4 + IL-13 for 48 h. As depicted in Fig. S3 A and B, the level of CD206 expression in the stimulated cells (17.63%, about 14.1-fold higher) was significantly higher than the unstimulated cells (1.25%), showing the effective polarization of Raw264.7 cells to M2 macrophages. To assess the targeted adhering ability of MBs to the M2 macrophages in vitro, the stimulated cells were incubated with MB_con or MB_man. Figure 3A revealed that there are a large number of MB_man adhere to the IL-4 + IL-13-stimulated cells. However, almost no MB_con adhesion can be observed around the cells. The CLSM images showed the same result (Fig. 3B). Subsequently, the adhesion rate of MB_man to M2 macrophages was quantitatively estimated using flow cytometry. As illustrated in Fig. 3C and D, M2 macrophages incubated with DiI-MB_man exhibited a higher percentage of fluorescence-positive rate compared to the DiI-MB_con group (38.52 ± 2.76% versus 1.99 ± 0.32%, P < 0.001). All the above results confirm that mannose coupling on the surface of MBs effectively promotes the cell adhesion of M2 macrophages in vitro.

Fig. 3.

Cell binding of MB_con or MB_man to M2 macrophages and enhanced accumulation of DiR-MB_man to allografts. (A) More MB_man were observed binding to the surface of M2 macrophages compared to MB_con in optical microscopy. Scale bar = 50 μm. (B) The representative CLSM images of DiI-MB_con and DiI-MB_man adhesion to M2 macrophages were consistent with the findings observed in optical microscopy. Scale bar = 50 μm. (C and D) Representative flow cytometric plots and quantitative analysis of the adhesion of DiI-MB_con and DiI-MB_man to M2 macrophages. ***P < 0.001, n = 5. (E) Ex vivo fluorescence imaging of grafts and major organs from heart transplanted rats. “a” represents the graft; “b” represents the heart; “c” the liver; “d” the spleen; “e” the lung; and “f” the kidney. (F) Quantitative fluorescence analysis of the grafted heart. *P < 0.05, **P < 0.01. n = 4

Targeted imaging necessitates the specific and adequate accumulation of MB_man in the cardiac graft. Therefore, the accumulation of MB_man in the transplanted heart was investigated. Minimal fluorescence was observed in both the allograft and isograft after injection of DiR-MB_con. On the contrary, obvious fluorescence aggregation was observed in the allograft after injection of DiR-MB_man (Fig. 3E). These results were further confirmed by quantitative analysis. The fluorescence intensity of DiR-MB_man in allografts increased significantly, which was 1.74- and 1.76-fold higher compared to that of DiR-MB_con in allografts and DiR-MB_man in isografts, respectively (Fig. 3F), suggesting that injection of DiR-MB_man could enhance the accumulation in allografts.

In vivo UMI

Since alterations in the intima may be detected as early as 1 or 2 weeks after HT, the lesion at this stage presents with mild intimal thickening and mild fibrosis [27]; we perform in vivo UMI early at 2 weeks after HT. UMI was conducted at 2 , 4 ,and 6 weeks post-transplantation, respectively. Following a bolus injection of MBs, the myocardium exhibited swift enhancement devoid of localized perfusion abnormalities, reaching a peak within a short period. Subsequently, both MB_man and MB_con exhibited a gradual decline in intensity. An instantaneous flash pulse was performed to destroy MBs in the graft 3 min after injection. The SI of MBs notably decreased after flash and rebounded within a few seconds due to the gradual reflow of MBs. The SI of adherent MBs was obtained by calculating the difference between the average SI before and after the flash.

Figure 4A, B, and C displayed the representative contrast-enhanced ultrasound (CEUS) images and the TIC in the same recipient following the injection of MB_con and MB_man. 2 weeks after transplantation, the quantitative analysis demonstrated that MB_man showed a noteworthy rise in the SI in allograft rats compared with MB_con (1.99 ± 0.23 dB versus 0.91 ± 0.23 dB, P < 0.05). Nevertheless, no obvious discrepancy was observed between MB_man and MB_con in isograft rats (1.02 ± 0.18 dB versus 1.04 ± 0.21 dB, P > 0.05) (Fig. 4D). The SI of the MB_man was also much higher than MB_con in allograft rats at 4 weeks after transplantation (4.10 ± 0.53 dB versus 1.55 ± 0.37 dB, P < 0.01). The SI of MB_man and MB_con in isograft rats also exhibited no significant difference (1.44 ± 0.28 dB versus 1.31 ± 0.38 dB, P > 0.05) (Fig. 4E). Finally, the statistical analysis was conducted on the TIC at 6 weeks post-transplantation. The results remained consistent with those observed previously. MB_man also showed a significantly higher SI compared to MB_con in allograft rats (3.38 ± 0.60 dB versus 1.38 ± 0.33 dB, P < 0.01). Similarly, no statistical difference was observed between the SI of MB_man and MB_con in isograft rats (1.32 ± 0.36 dB versus 0.55 ± 0.10 dB, P > 0.05) (Fig. 4F). These results suggest that M2 macrophage-targeted UMI can achieve early quantitative evaluation of CR as early as 2 weeks post-HT and enable dynamic monitoring of CR.

Fig. 4.

UMI in vivo at different time points. (A, B, and C) The representative CEUS images (B mode, before flash, the first frame post flash, 10 s after flash, and TIC) in one isograft rat (top two lines) and one allograft rat (bottom two lines) following the injection of MB_man or MB_con at 2 W, 4 W, and 6 W, respectively. The boundary of the myocardium was delineated based on the B-mode images. The areas highlighted in the two yellow lines were regions of interest. (D, E, and F) Quantitative analysis of the average SI of adherent MBs at 2 W, 4 W, and 6 W, respectively. *P < 0.05, **P < 0.01, ***P < 0.001, n = 6

Histological results

As shown in Fig. 5, the graft myocardium was stained with H&E, EVG, and Masson. To evaluate graft vasculopathy, the I/M ratio of graft vessels was analyzed in each group. The results indicated that there was mild neointimal hyperplasia in the allografts at the 2-week mark post-transplantation. The I/M ratio was higher in allografts than in isografts (0.313 ± 0.041 versus 0.149 ± 0.036, Fig. 5A and B). Figure 5D shows that neointimal hyperplasia developed characteristically in allografts at 4 weeks. Conversely, the histological characteristics mentioned above were not present in the isografts. The I/M ratio demonstrated a notably higher value in allografts compared to that in isografts. (0.570 ± 0.175 versus 0.168 ± 0.031, Fig. 5E). Similarly, the neointimal hyperplasia could be significantly observed in allografts at 6 weeks, while there was no intimal hyperplasia in isografts (Fig. 5G). The I/M ratio was also notably higher in allografts compared to isografts. (0.626 ± 0.068 versus 0.166 ± 0.029, Fig. 5H). Furthermore, myocardial fibrosis of the grafts was evaluated by Masson staining. The results indicated that myocardial fibrosis occurred as early as 2 weeks post-transplantation in the allograft group (Fig. 5A). Quantitative analysis revealed that the fibrotic areas were significantly larger than those observed in the isograft group (9.522% ± 1.058% versus 5.353% ± 0.518%) (Fig. 5C). A similar situation was observed at the 4-week mark (11.999% ± 1.415% versus 4.838% ± 0.990%, Fig. 5D and F) and 6-week mark (13.527% ± 1.120% versus 3.967% ± 0.336%, Fig. 5G and I). The aforementioned findings suggest that CR likely commenced at the 2-week mark after HT and persisted over time, showing ongoing progression.

Fig. 5.

The histopathologic changes of cardiac allografts. (A, D, and G) The representative images of H&E staining, EVG staining, and Masson staining of the grafts at 2, 4, and 6 weeks after transplantation. Scale bar = 20 μm. (B, E, and H) Quantitative analysis of the I/M ratio in each group at 2, 4, and 6 weeks. (C, F, and I) Quantitative analysis of fibrotic areas in each group at 2, 4, and 6 weeks. *P < 0.05, ***P < 0.001, n = 6

In the present study, we specifically investigated CD206 ⁺ M2 macrophage infiltration during CR. Immunofluorescence staining targeting CD206 was systematically performed on graft tissues. The results revealed significant CD206 expression in the myocardium of the allograft group as early as 2 weeks, with minimal expression in isografts (Fig. 6A). This was further confirmed by quantitative analysis (0.343% ± 0.043% versus 0.117% ± 0.034%, Fig. 6B). The result suggests that M2 macrophage infiltration occurs at a very early stage after transplantation. Meanwhile, we also observed significant CD206 expression at 4 weeks (0.627% ± 0.075% versus 0.156% ± 0.055%, Fig. 6C and D) and 6 weeks (0.452% ± 0.096% versus 0.110% ± 0.012%, Fig. 6E and F) in allografts. Quantitative assessment demonstrated a marked increase in the proportion of CD206 + cells at 4 weeks compared to the 2-week (P < 0.05). However, no statistically significant difference was observed in CD206 + cell proportion between 4- and 6-week, suggesting a stabilization in M2 macrophage accumulation beyond the 4-week phase (Fig. 6G). Figure 6H demonstrated that the SI of MB_man was positively correlated with the percentage of CD206 (R² = 0.8179, P < 0.0001). However, the correlation was not observed for MB_con in Fig. 6I (R² = 0.1772, P>0.05).

Fig. 6.

Representative images of immunofluorescence staining in grafts. (A, C, and E) There was evident CD206 infiltration in allografts at 2, 4, and 6 weeks after transplantation, whereas only a few were seen in isografts. Meanwhile, immunofluorescence images illustrated the spatial proximity of CD206 (FITC) and CD31 (CY3) in allografts at all indicated time points. Scale bar = 10 μm (main), 2 μm (Zoomed insets). (B, D, and F) Quantitative analysis of CD206% in each group at 2, 4, and 6 weeks. (G) Quantitative analysis of CD206% in allografts at 2, 4, and 6 weeks. (H) Correlation analysis between the SI of MB_man and the percentage of CD206. (I) Correlation analysis between the SI of MB_con and the percentage of CD206. *P < 0.05, **P < 0.01, ***P < 0.001, n = 6 at 2 W, 4 W and 6 W

To delineate the spatial distribution of CD206⁺ macrophages within pathological vessels, we performed immunofluorescence staining for CD31 (an endothelial cell-specific marker) to investigate the anatomical relationship between CD206 ⁺macrophage infiltration and vascular structures in cardiac grafts. As depicted in Fig. 6A, C, and E, the CD206 + macrophages were frequently observed in close spatial proximity to CD31 + endothelial cells lining the vessels.

To verify that CD206-targeted cells are indeed macrophages and to avoid false-positive results due to off-target effects, we performed double immunofluorescence staining on cardiac allograft tissues to evaluate the cellular identity of CD206 ⁺ cells in CR microenvironments. As shown in Figure S4, FITC-labeled CD206 exhibited strong spatial overlap with CD68 ⁺ macrophages (Cy5), confirming the predominant targeting of M2 macrophages in vivo. Increased vascular permeability is regarded as a prominent characteristic of pathologic angiogenesis [28]. ZO-1 is a protein that regulates endothelial cell-cell junctions [29]. To understand the endothelial cell permeability in CR, immunofluorescence staining of ZO-1 was performed. As shown in Fig. S5A, C, and E, the allografts exhibited weaker green fluorescence than the isografts at all the indicated time points. The results were further confirmed by quantitative analysis. The mean fluorescence intensity (MFI) of ZO-1 in allografts decreased significantly by 39.97%, 22.59%, and 27.06% at 2 weeks, 4 weeks, and 6 weeks, respectively (Fig. S5B, D, and F), indicating increased permeability of endothelial cell monolayers.

Biosafety assessment in vitro and in vivo

The CCK8 assay revealed that there was no significant reduction in rat myocardial cell viability after incubating with MB_con or MB_man at different concentrations for 12 h and 24 h (Fig. S6). Potential side effects were evaluated 1 day and 7 days after injection of MB_con or MB_man. No significant tissue necrosis or morphological changes were observed in the main organs across all groups, further confirming the biocompatibility of MBs (Fig. 7A). Biochemistry analysis revealed normal levels of ALT, AST, CRE, and BUN in all groups, indicating that liver and kidney function remained normal after injection (Fig. 7B). Additionally, the counts of red blood cells, white blood cells, platelets, lymphocytes, neutrophils, and monocytes were not significantly varied after injection (Table S1).

Fig. 7.

Preliminary safety evaluation. (A) H&E staining of major organs resected from rats injected with PBS, MB_con, and MB_man on 1 day and 7 days after injection, respectively (scale bar = 50 μm). (B) Typical hematological parameters (n = 3)

Discussion

The present study represents the first report on the method to evaluate M2 macrophage infiltration within cardiac grafts using UMI, establishing a non-invasive and dynamic monitoring technique for CR. By performing UMI at 2-week intervals and conducting continuous dynamic monitoring in vivo, we aimed to achieve the purpose of early evaluation. Our findings establish a groundwork for future research and clinical implementation of M2 macrophage-targeted UMI for early non-invasive monitoring of CR.

Current clinical evaluation techniques for CR lack molecular-level assessment. Only a limited number of molecular imaging studies have been conducted for CR. Marcus et al. pioneered the first successful antibody-mediated molecular imaging of cardiac CR by demonstrating specific accumulation of monoclonal antibody F8 in a heterotopic HT model using immuno-PET [30]. In parallel, alternative molecular evaluation approaches for CR have been investigated by other researchers, though these methods are associated with radiation exposure [31, 32]. Our previous work demonstrated that UMI targeted at T lymphocytes can identify AR following HT [23]. In this study, our results show that MB_man enables real-time molecular assessment and continuous dynamic monitoring of pathological processes, with early detection of M2 macrophages serving as a prognostic indicator for CR onset. This advancement may facilitate the future development of multi-target and multimodal ultrasound molecular probes, thereby advancing precision evaluations for CR [33, 34].

It has been previously reported that macrophage serves as the principal cellular mediators of graft rejection, and their infiltration is intricately linked to the prognosis of grafts, particularly in predicting CR [35]. Huibers et al. demonstrated that the majority of macrophages infiltrating the vascular intima in CR were M2 macrophages [36]. M2 macrophages, conventionally regarded as anti-inflammatory regulators, may drive CR through cytokine-mediated fibrosis and matrix metalloproteinases (MMPs)-dependent vascular injury [37]. Previous studies demonstrate that graft-infiltrating M2 macrophages primarily contribute to chronic cardiac rejection by releasing anti-inflammatory and repair-promoting substances like IL-10, along with pro-fibrotic agents such as TGF-β and VEGF, which encourage fibrosis and vascular lesion formation [17, 38]. Crucially, M2-associated MMPs (e.g., MMP-2, MMP-9) play a critical role in the degradation and remodeling of the extracellular matrix during CR, thereby facilitating tissue remodeling and intimal thickening of vessels [39–41]. Numerous studies have suggested that M2 macrophages may serve as potential targets for disease diagnosis and treatment For example, TAMs are the predominant infiltrating cells in the tumor microenvironment and typically exhibit an M2 phenotype [42]. Some reports also indicated that M2 macrophages are crucial in the regression of atherosclerotic plaques [43, 44]. These results offer the possibility of non-invasively targeting and monitoring CR by tracing M2 macrophages.

CD206 is a macrophage-specific surface marker that shows high expression on M2 macrophages but minimal expression on the M1 phenotype across species, including humans, rats, and monkeys [45]. It can be used to locate M2 macrophages, creating opportunities for non-invasive targeted monitoring in clinical settings [21, 46]. Ye et al. developed a chlorogenic acid (CHA)-encapsulated mannosylated liposome, which could selectively transport CHA to TAMs and encourage the transition of M2 macrophages to the M1 phenotype, thereby inhibiting tumor growth [47]. Here, we developed CD206-targeted MB_man to specifically bind M2 macrophages. In vitro experiments showed MB_man’s superior adhesion to M2 macrophages versus control MB_con, as quantified by flow cytometry, fulfilling the essential requirement for in vivo ultrasound imaging. The evaluation of MB_man’s performance under physiologically relevant conditions is also significant. Previous studies have consistently demonstrated targeted microbubble adhesion under physiological shear stress conditions: Guenther et al. reported P-selectin-targeted MBs maintained 3.4% adhesion efficiency at 40 dynes/cm² [48], cRGDfK-modified MBs showed significantly enhanced α_Vβ₃ binding affinity [49], and Yan’s team confirmed superior performance of LyP-1-conjugated MBs in microfluidic systems [50]. Collectively, these findings support the principle that appropriately designed targeted microbubbles can achieve specific binding under hemodynamic shear stress. While direct experimental validation under simulated flow was not undertaken in our work, the significant accumulation of MB_man within the allograft, demonstrated by ex vivo fluorescence imaging, serves as functional evidence for its stability and targeting efficiency in a physiological hemodynamic environment.

In the realm of HT, UMI is primarily utilized in monitoring rejection after HT, assessing graft inflammatory responses, and implementing targeted therapeutic interventions [51, 52]. Gregory et al. successfully detected rat acute cardiac rejection through UMI using intercellular adhesion molecule-1 targeted MBs [53]. Afterward, an increasing number of studies focusing on UMI for assessing cardiac allograft rejection have emerged [13, 23, 54]. However, there is limited research on UMI in CR after HT. Unlike AR, CR poses evaluation challenges, especially in its early stages [55]. CR is a progressive process involving pathophysiological changes at multiple time points. For a comprehensive understanding of the process, it is crucial to choose time points that are representative of the different developmental stages of CR. Serial histological analyses unveiled the infiltration of mononuclear cells and the thickening of the intima, commencing in the majority of grafts at the 2-week post-transplantation [56]. Therefore, the choice of observation from the early post-transplantation period provides insight into the vasculopathy and the initial stage of macrophage infiltration and its subsequent dynamics. Based on this, we set the starting point for early detection of UMI to be 2 weeks, with an interval of every two weeks, and ending the imaging at 6 weeks. Our histological findings revealed that the intimal changes in allografts exhibited thickening at 2 weeks, which coincided with upregulation of CD206 expression starting at 2 weeks, suggesting the onset of M2 macrophage infiltration at this time point. Importantly, UMI demonstrated a notably higher SI of MB_man compared to MB_con in the allograft at 2 weeks, with consistent and stable targeting observed at the 4th and 6th weeks. Moreover, the ultrasound imaging signal of MB_man exhibited a positive correlation with the percentage of M2 macrophages, further confirming the significant value of MB_man in detecting CR. These findings highlight the potential of UMI for early assessment of CR.

Intriguingly, our findings revealed that the proportion of CD206 at 4 weeks exhibited a statistically significant increase relative to that at 2 weeks. No statistically significant increase was observed at 6 weeks. This observation aligns with established mechanisms of macrophage-driven fibrosis in chronic allograft rejection. The initial CD206⁺ macrophage surge marks the shift from acute injury (ischemia-reperfusion, alloimmunity) to tissue remodeling, driven by Th2 cytokines (IL-4) and profibrotic factors (TGF-β), promoting monocyte recruitment and M2 polarization [57, 58]. After 4 weeks, CD206 ⁺ levels stabilize due to: (1) macrophage-to-myofibroblast transition (MMT) reducing detectable CD206⁺ cells while enhancing ECM deposition, and (2) diminished pro-M2 signals. By 6 weeks, fibrosis becomes self-sustaining through TGF-β/Smad3-mediated MMT rather than new macrophage recruitment [59]. This biphasic pattern (from expansion to stabilization) reflects macrophage plasticity transitioning from inflammatory to fibrogenic phenotypes via MMT [60]. The plateau signifies advanced fibrosis driven by MMT-derived myofibroblasts, not disease arrest. While CD206 is expressed on multiple cell types, previous studies have demonstrated that CD206⁺ M2 macrophages are the dominant infiltrating immune population within allografts in CR [37, 39]. Our results of double immunofluorescence staining also confirmed that CD206 ⁺cells in CR lesions are primarily macrophages. This suggests that the MB_man signal in our UMI primarily reflects M2 macrophage infiltration.

MBs serve as blood pool contrast agents in a variety of ultrasound-related medical applications. As a result, if neither stasis, endothelial injury, or hypercoagulability occurs within the bloodstream, MBs will tend to be positioned within the luminal region of the vessel [61]. Wong et al. demonstrated increased endothelial cell hyperpermeability in both the intima and media of coronary arteries afflicted with CAV [62]. A study on the ultrastructure of early endothelial damage in coronary arteries after cardiac allografts in rats revealed that endothelial cell disruption, including large cellular gaps, missing cells, and exposed extracellular matrix areas, occurs early after transplantation. The findings suggested that the vascular barrier is partially open after HT due to endothelial damage and increased vascular permeability. This opening may facilitate the infiltration of inflammatory cells and macromolecules that play a key role in the development of graft vasculopathy [63]. Consequently, MBs could traverse the vascular endothelium and reach the affected intimal region during CEUS, which was confirmed by our results of ZO-1 immunofluorescence staining. There was a significant decrease in ZO-1 expression in allografts, indicating increased endothelial cell permeability. All the above findings establish the groundwork for subsequent research and clinical implementation of M2 macrophage-targeted UMI in the noninvasive monitoring of CR.

As of now, UMI has been deployed in clinical settings for various diseases, demonstrating commendable accuracy and safety profiles [11, 64, 65]. During this study, no instances of mortality were recorded following the injection of MBs. Furthermore, we established that the injection of MB_man did not compromise allograft survival or exacerbate the injury, thus affirming the safety of MB_man injection. Nevertheless, several limitations warrant acknowledgment. The diagnostic efficacy of MB_man for CR in humans requires further investigation. In this research, we used a linear probe of 6-15 MHz to detect the heterotopic graft. We are preparing to apply MB_man for CR using a phased array of 1.5-3.5 MHz, which will contribute to the clinical transformation of this technology. Furthermore, it is important to note that while the preferential expression of CD206 on the surface of M2 macrophages has made it renowned as a marker for these cells, other cell types prevalent in the tissues, such as dendritic cells, have also been shown to express this receptor [66]. Thus, the specificity of MB_man for the detection of CR needs to be further explored. CD163 is another important marker of M2 macrophages. In the further, we are designing dual-targeted MBs to detect M2 macrophages infiltrated in CR. We infer that simultaneous detection of CD206 and CD163 by UMI would greatly enhance the specificity of CR evaluation. Clinical translation challenges persist, including interspecies differences, safety confirmation, diagnostic utility against gold standards, and standardized manufacturing. Although rodent studies indicate preclinical potential [67, 68], significant interspecies differences in immune responses, vascular biology, and macrophage-targeting mechanisms necessitate comprehensive evaluation in a non-human primate model. Further investigation into such a model is essential to validate the efficacy and safety of MB_man. The administered doses of MB_man and positive standards should be rigorously confirmed. Additionally, a significant challenge of the targeted ultrasound microbubbles lies in ensuring consistent batch-to-batch quality during large-scale manufacturing.

Conclusion

In summary, our study demonstrated a noninvasive and quantitative method of monitoring M2 macrophage infiltration in cardiac allografts with CR utilizing M2 macrophage-targeted UMI. This approach represents a promising and valuable method for the continuous and dynamic monitoring of CR. Attaining success in the early monitoring of CR may signify a notable landmark in cardiovascular research, providing valuable insights into the potential strategies for patient management.

Author contributions

Conceptualization, Methodology, Writing-original draft, Data analysis: J.X.,C.D.,W.W., and T.G.Heterotopic heart transplantation model: M.H. and W.F.Software, Investigation: X.Z. and J.Q.In vivo ultrasound molecular imaging: J.X., C.D., Y.B. Ex vivo fluorescence imaging: R.W., J.X. Methodology: Y.C. and Q.J.Project supervision: L.Z., Q.L. and M.X.Review and editing, Funding acquisition: W.W.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82371990, 82001852) and the Nature Science Foundation of Hubei Province (No.2023AFB772, 2024AFB032).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

All procedures involving animals adhered to ethical standards for experimental research and were approved by the Institutional Animal Care and Use Committee (IACUC) of Tongji Medical College, Huazhong University of Science and Technology (IACUC number: 3791).

Consent for publication

All authors agreed to submit this manuscript.

Competing interests

The authors declare no competing interests.