In this work, RNase A-PbS QDs as a specific bone image contrast was firstly injected through the tail vein of the Balb/C nude mouse to achieve imaging of bone structures. As shown in Movie S1 and S2 in Supporting Information, as the QDs spreads throughout the whole body of the mouse via blood circulation, specific structures of the skeletal system were imaged by detecting the fluorescence signals from the QDs. Moreover, it is noteworthy that multiple structures of the skeletal system could be clearly identified and distinguished from either a posterior (prone position) or anterior (supine position) view. Thus, characteristics of both the posterior and anterior view of the Balb/C nude mouse after QDs injection were fully evaluated through SWIR in vivo imaging movies in a real-time manner.
Figure 1a-h was the posterior view of the mouse. After 15 min post-injection, fluorescence signals were firstly detected from the liver, spleen and heart of the Balb/C nude mouse (Fig. 1b and 1c). Then, as shown in Fig. 1d-g, a variety of bone structures gradually started to be observed since 30 min post-injection. Specifically, from axial bones (the parietal bone, maxilla, spine, pelvis) to limb bones (the humerus, radius and ulna, metacarpus, femur, tibiofibula, calcaneus, phalanges), from upper to lower extremities, from proximal to distal bones as well as major joints (the elbow, wrist, hip, knee, ankle), morphology of bone structures all above were identified in a fixed order with labelling by SWIR fluorescence. During the longitudinal observation, the fluorescence intensity (FL) measured from specific bone structures reached a peak around 1 h post-injection, and then slowly faded with increasing observation time (Figure S1 in Supporting Information), which indicated that it would be the optimal observation time for skeletal system imaging of the nude mouse. After 168 h, fluorescence signals could not be detected from structures of the skeletal system any more, while metabolic organs such as the liver and the spleen still showed weak fluorescence signals (Fig. 1h).
Figure 1. (a) Photograph of Balb/C nude mouse. (b-i) Real-time SWIR fluorescence imaging of Balb/C nude mouse by RNase A-PbS QDs (posterior view) with images acquired at 5 min, 15 min, 30 min, 60 min, 4 h, 24 h, 72 h and 168 h post-injection, respectively. (i) X-ray photo and (j) SWIR image of bones in an overall view (prone position). (k) X-ray photo and (l) SWIR image of bones in an overall view (supine position). (m-q) Zoomed images of (i-l) respectively. (r) SWIR image of the skull.
Notably, from posterior view, as shown Figure S2 in Supporting Information, FL intensity in both the skull and the spine was higher than that in the other structures of the skeletal system at 1 h post-injection, which showed both the skull and the spine has a superior imaging among other observed bone structures from a posterior view. this phenomenon is ascribed to the anatomical and structural characteristics around the skull and the spine (Figure S2b in Supporting Information). Anatomically, both of the two bone structures had a relatively abundant source of blood flow from the carotid artery, aorta and common iliac artery, while an extraordinarily large vascularity named Adamkiewicz artery from the thoracolumbar segment provided mostly blood supply for the spinal cord[22, 23]. On the other hand, as the vertebral vein had no valve, the blood could flow in both directions, increasing the dwelling time of QDs and facilitating high FL intensity of the spine. Interestingly, from anterior view (Figure S3 in Supporting Information), although similar results of imaging were obtained comparable to the posterior view, the lower extremities were imaged quicker than the upper extremities, due to the fact that the anterior side of the lower extremities was covered by fewer muscles than the posterior side. Furthermore, those in vivo imaging evaluations were also repeated in the Balb/C mouse in order to confirm the above results. No obvious differences between the Balb/C mouse and the Balb/C nude mouse were observed in SWIR imaging (Figure S4, S5 and Movie S3, S4 in Supporting Information).
To investigate the features of in vivo SWIR fluorescence imaging, a comparison with X-ray imaging was made and shown in Fig. 1i-r. From a posterior view (prone position), the pelvis overlapped with other structures and could not be fully recognized under X-ray examination (Fig. 1i and Fig. 1n). Compared with X-ray imaging, morphology of the pelvis could be clearly distinguished with a full shape of both the ilium and the sacrum in SWIR fluorescence imaging (Fig. 1j and Fig. 1o). While from an anterior view (supine position), the sternum could be recognized and differentiated as a segmental structure, which integrated with the background under X-ray imaging (Fig. 1k, l, p and q). Notably, regarding the skull, both the left and right lateral ventricles could be observed from the SWIR images, which could not be distinguished with X-ray images (Fig. 1m and Fig. 1r). Unlike traditionally two-dimensional anatomical structures under X-ray (Fig. 1i and Fig. 1k), our QDs-based SWIR fluorescence imaging was capable of generating highly-magnified and well-resolved images for frequent and longitudinal studies of bones without radiation in the future.
Most importantly, as shown in Fig. 1m and Fig. 1r, both the maxillary and the parietal of the skull have much clearer and more stratified morphology than that in X-ray images. The result showed that the substructure and biological activities of the skull could be analyzed by QDs-based SWIR imaging. Interestingly, simultaneous with bone imaging, the contours of the liver, spleen and heart could only be clearly demonstrated by SWIR, not by X-ray. The additional information exhibited by SWIR imaging had great potential to offer a comprehensive diagnosis of organs injuries accompanied with acute trauma, especially for the organs with abundant blood supply or neighboring fracture zones. Therefore, compared with X-ray imaging, RNase A-PbS QDs-guided SWIR in vivo imaging not only demonstrated a multi-spatial resolution imaging of the skeletal system based on tissue penetration and viewing angle without radiation exposure, but also provided a precise overall risk evaluation for bone disorders, especially for trauma in clinical practice.
To further confirm the in vivo results above, ex vivo imaging of the isolated bone tissues of both Balb/C nude mouse and Balb/C mouse after QDs injection was obtained following in vivo observation (Fig. 2). Comparable with the in vivo results, the FL intensity was obviously higher in the skull, the spine and four limbs than other bone structures due to blood supply. However, as shown in Fig. 2a and Fig. 2b, the FL intensity of bone structures quickly decreased from 3 h to 72 h in both Balb/C nude mouse and Balb/C mouse. It was much quicker than that in in vivo imaging observation, because the fluorescence of bone structures could preserve for 168 hours in in vivo imaging. The quick fading of fluorescence detected in the isolated bone tissues was probably due to the exposure of QDs with various factors such as surface ligands, pH value of the solvent, photooxidation in the environment and so on. Moreover, after tissue isolation from the living body, cells died over time in vitro, leading to decomposition of QDs accompanied with a rapid decrease of FL intensity[25, 26]. Therefore, the binding capability of RNase A-PbS QDs could depend on cell states including living cells and died cells, providing extra information about the biological activities of bone tissues.
As the fluorescence of bone structures could be partially interfered with soft tissues, the skeletal structures were isolated and examined by SWIR fluorescence imaging and X-ray imaging ex vivo. As shown in Fig. 2c-j, strong fluorescence signals were detected from the sternum (Fig. 2g and Fig. 2i, green arrow), the ribs (Fig. 2i, white arrow) and the lumber vertebrae (Fig. 2h and Fig. 2j), which agreed with previous SWIR in vivo imaging. Particularly, Figure S6 in Supporting Information demonstrated the segmented bone structures of rib and sternum in details with the SWIR images, while it was not visualized under X-ray. Moreover, these small bone pieces with a diameter less than 0.1 cm in the sternum could be clearly identified in SWIR images and X-ray images. This result suggested that SWIR imaging would be an alternative approach of X-ray for bone microstructure in future. Therefore, SWIR imaging would provide a dynamic monitoring for both physiological and pathological processes of bone, benefiting for the diagnosis of skeletal disease in future.
As shown in Fig. 3a, to further study the underlying mechanism of SWIR QDs-guided bone imaging, four cell types related with bone metabolism including osteoblasts (MC3T3-E1), fibroblasts (NIH3T3), BMSC and Macrophages (RAW264.7) were selected[27–31]. It was reported that QDs could non-specifically bind with cell membranes before entering cells[32–35], In this work, the cells were rinsed by PBS twice to remove those non-specifically QDs bound to cell membranes. As shown in Fig. 3b, fluorescence signals were detected in the four types of cells after 0.5 h, 3 h, 24 h and 48 h of incubation. Among the four, the FL intensity of MC3T3-E1 cells reached a peak after three hours of incubation, which suggested that MC3T3-E1 cells were more easily to uptake QDs than the other three cells types (NIH3T3, BMSCs and RAW264.7). Moreover, similar results were also detected from those four cell types before rinsing (Figure S7 in Supporting Information). Moreover, the FL intensity reached a peak when the QDs concentration was around 250 µg mL− 1 (Figure S8 in Supporting Information), and the MC3T3-E1 cells still showed the highest FL intensity in comparison with the other tested cell types at the same concentration, which further proved the unique binding relationship between QDs and MC3T3-E1 cells.
To further study the relationship between QDs and the cell lifecycle, these QDs used in the first round of cell culture were collected and recycled as a medium additive for the second round of cell culture). Surprisingly, no fluorescence signals were detected after 48 hours of incubation, suggesting that only freshly prepared QDs was able to maintain its fluorescence in the first round of cell incubation, while second-hand QDs could lose fluorescence after cell metabolism in the second round of cell incubation. Therefore, as shown in Fig. 3a, the QDs’ fluorescence could be quenched after cell metabolism and would eventually be eliminated from the body without disturbing SWIR imaging in the next time.
Although in vivo results revealed that QDs in bone tissues were eliminated up to 80% after 72 hours, while most were cleared out from internal organs (such as the liver and the spleen) after 168 hours. Therefore, as shown in Fig. 3c and Figure S9 in Supporting Information, bone tissues and main organs in the tested mouse showed no obvious inflammation and abnormality after QDs injection, indicating our prepared QDs had excellent biocompatibility, in agreement with our previous studies[36, 37].