18F-fluorodeoxyglucose positron emission tomography–computed tomography for assessing organ distribu

Red blood cells (RBCs) stressed by high temperature are similar to senescent or damaged RBCs in pathological conditions. RBCs can be efficiently labelled with 18F-fluorodeoxyglucose (FDG). The aim of this study was to assess stressed RBCs erythrophagocytosis and organ distribution in vivo with the application of 18F-FDG PET/CT. RBCs were induced under high temperature (48 °C) to prepare stressed RBCs. Fluorescence-activated cell sorting (FACS) was used to analyse reactive oxygen species (ROS) generation, intracellular Ca2+ concentration and membrane phosphatidylserine (PS) externalization of RBCs. 18F-FDG was used to label RBCs and assess the erythrophagocytosis. Finally, 18F-FDG PET/CT was applied to reveal and measure the organ distribution of stressed RBCs in mice. Compared with untreated RBCs, stressed RBCs decreased in cell volume and increased in ROS level, intracellular Ca2+concentration, and PS exposure. RBCs could be labelled by 18F-FDG. Stressed RBCs tended to be phagocytosed by macrophages via assessment of FACS and radioactivity. 18F-FDG PET/CT imaging showed that stressed RBCs were mainly trapped in spleen, while untreated RBCs remained in circulation system. Thus, stressed RBCs can be effectively labelled by 18F-FDG and tend to be trapped in spleen of mice as assessed by PET/CT.

RBCs are the main component of blood, accounting for half of the total blood volume. Under physiological conditions, RBCs have biconcave shape and exhibit reversible deformability and durability1. Since RBCs have a simple structure, lacking nuclei and internal organelles, they are highly susceptible to various changes in the internal environment, such as oxidative stress, osmotic shock, energy depletion, and cytokines. These changes lead to morphological and functional alterations of RBCs, including lipid oxidation imbalance, aberrant activation of membrane ion channels, phosphatidylserine (PS) externalization, and aggregation or allostery of membrane proteins2,3,4, eventually accelerating the clearance of RBCs. Several studies show that these changes occur in RBCs in some chronic diseases5,6,7. Macrophages remove senescent or damaged RBCs from the circulation by a fundamental physiological process called erythrophagocytosis, which is essential for iron/heme metabolism and homeostasis. It is generally believed that erythrophagocytosis is executed by hepatic and splenic macrophages8,9. However, which kind of macrophages have a key function in RBCs clearance remains elusive. Some studies demonstrated that storage-damaged and senescent RBCs are mainly cleared by splenic red pulp macrophages10,11, while other researchers consider hepatic macrophages as the vital cells in the clearance of aged and stressed RBCs12,13. Therefore, it is of great value to study the distribution of RBCs in pathological state and the main scavenging organs.

As one of the most advanced technologies, positron emission tomography–computed tomography (PET/CT) can obtain metabolic and anatomical data simultaneously. PET/CT uses an injection of short half-time radioactive isotope labelled substance into the body to reflect the status of metabolism activities through the aggregation of radioactivity. 18F-fluorodeoxyglucose (18F-FDG), a radioactively labelled glucose analogue, is the main clinical PET/CT imaging agent. 18F-FDG PET/CT has been widely applied in the detection and diagnosis of diseases with high glucose metabolism including infection/inflammation and cancer, by imaging relative 18F-FDG uptake rates in various tissues14. The energy metabolism of RBCs relies on glycolysis. As a glucose analogue, 18F-FDG can thus be taken up by RBCs to achieve RBC labelling, which was recently shown to be feasible15. Currently, whether stressed RBCs are phagocytosed and where they are distributed in vivo remains unclear.

In this study, we used high temperature-induced stressed RBCs to simulate RBCs in pathologic conditions. 18F-FDG was used to label RBCs to detect erythrophagocytosis and PET/CT was performed to assess organ distribution of stressed RBCs in mice.

Characteristics of stressed RBCs

After heating for 30 min at 48 °C, RBCs changed in morphology. Forward scatter can reflect the relative volume of RBCs16, the forward scatter width (FSC-W) of untreated RBCs was higher than FSC-W of stressed RBCs, indicated that cell volume was smaller in stressed RBCs than in untreated RBCs (Fig. 1A,B). The membrane molecular distribution and intracellular ion concentration of stressed RBCs also differed from untreated RBCs. ROS level and intracellular Ca2+ concentration was described as mean fluorescence intensity (MFI), PS externalization was expressed as Annexin-positive percentage. The DCFH-DA MFI (Fig. 1C,D), Flou4 MFI (Fig. 1E,F) and Annexin-positive percentage (Fig. 1G,H) of stressed RBCs were higher than that of untreated RBCs. These results indicated that heating resulted in elevated ROS levels, increased calcium influx, and higher PS exposure in stressed RBCs.

Stressed RBCs tended to be phagocytosed

Since FACS is the most common method to measure erythrophagocytosis, we used fluorescence probe TER-119-labelled RBCs to detect this process. The FITC-dependent fluorescence intensity of macrophages reflected the phagocytosis of RBCs. The mean fluorescence intensity of macrophages incubated with stressed RBCs was higher than that of macrophages incubated with untreated RBCs (Fig. 2A,B). Consistent with the FACS results, immunofluorescence indicated that stressed RBCs were more likely to be engulfed by macrophages than untreated RBCs. Bright field showed Raw 264.7 cells (Fig. 2C,D). Fluorescence field showed that many stressed RBCs were located in macrophages, indicating they were phagocytosed (Fig. 2F). Untreated RBCs were less phagocytosed by macrophages, and less untreated RBCs adhered to macrophage membranes (Fig. 2E). The mean fluorescence intensity value of macrophages treated with stressed RBCs was higher than that of untreated RBCs (Fig. 2G).

18F-FDG labelling efficiency of untreated and stressed RBCs was 96.13 ± 1.81% and 95.71 ± 2.46% (Fig. 3A). Difference between the labelling efficiency of stressed RBCs and untreated RBCs was not significant. The post-labelling stability of untreated RBCs were 98.91 ± 0.079%, 98.2 ± 0.183%, 98.34 ± 0.079% at 1 h, 2 h and 3 h respectively, and the labelling stability of stressed RBCs were 97.68 ± 0.452%, 96.56 ± 0.375%, 96.71 ± 0.478% (Fig. 3B). The radioactivity of the unphagocytic RBCs and macrophages were measured to calculate the ratio of phagocytosis. The ratio of stressed RBCs incubated macrophages was higher than the ratio of untreated RBCs incubated macrophages (Fig. 3C), which meant stressed RBCs tended to be phagocytosed by macrophages compared with untreated RBCs.

Stressed RBCs were trapped in spleen

Upon injection with untreated RBCs, the cardiac radioactivity was high while the radioactivity of other organs were low (Fig. 4A). Images showed that untreated RBCs mainly distributed in cardiac (Fig. 4C). Instead, upon injection with stressed RBCs, the splenic radioactivity was high (Fig. 4B) and images revealed that spleen showed a strong accumulation of 18F-FDG-labelled stressed RBCs (Fig. 4D). In short, injected stressed RBCs were trapped in spleen while untreated RBCs kept in cardiovascular system.

Subsequently, H&E staining of spleen from stressed RBCs-injected mice showed many RBCs in splenic sinuses (Fig. 5B), but less RBCs in spleen of untreated RBCs-injected mice (Fig. 5A). Immunofluorescence also revealed that stressed RBCs gathered in spleen (Fig. 5C–E). These results demonstrated that stressed RBCs trapped in spleen, which were consistent with the data of in vivo organ distribution measured by PET/CT.

There is controversy about the main organ responsible for RBC clearance10,11,12,13. Our study used PET/CT to confirm the accurate organ distribution of stressed RBCs, which could reflect their clearance site. First, we demonstrated that stressed RBCs showed similar characteristics to RBCs under pathological state (Fig. 1). Second, we proved that 18F-FDG can label RBCs effectively and there were no statistical differences between the labelling efficiency of stressed and untreated RBCs (Fig. 3A). Thus, 18F-FDG labelling of RBCs is suitable to assess erythrophagocytosis and in vivo organ distribution of stressed RBCs and untreated RBCs. Then, our data demonstrated increased erythrophagocytosis in stressed RBCs (Figs. 2, 3C). Finally, PET/CT imaging showed that splenic radioactivity increased rapidly in mice injected with stressed RBCs (Fig. 4D), and these results were confirmed by H&E staining and immunofluorescence of spleen sections (Fig. 5).

Spleen is a part of reticuloendothelial system which related to RBCs clearance. Stressed RBCs are trapped in spleen means stressed RBCs are likely removed by spleen. This could explain the anaemia in some diseases. Besides, injected stressed RBCs trapped in spleen indicate low efficiency of transfusion since RBCs must be in cardiovascular system to function properly. RBCs in cardiovascular system indicate that RBCs could perform their physiological functions, which means that transfusion could achieve its purpose. Thus, 18F-FDG PET/CT can contribute to investigate the transfusion efficiency.

In our study, we use 18F-FDG to label RBCs. As a glucose analogue, 18F-FDG could be taken up by RBCs. After 18F-FDG converted into 18F-FDG-P, it can’t continue the following glycolysis process until 18F undergoes radioactive decay17. The post-labelling stability of 18F-FDG labelled RBCs was high, which means 18F-FDG labelled RBCs was suitable for displaying RBCs distribution.

Our research used 48℃ treated RBCs (stressed RBCs) to simulate RBCs under pathological state. Utoh et al. proposed that high temperature treatment would lead to haemolysis and elevated osmotic fragility of RBCs, which were thought to derive from the changes in composition of membrane18. Besides, heat treatment could also result in energy depletion19, it was related to the activation of Janus kinase 3 (JAK3) which contributes to the plasma membrane scrambling20. Recent research revealed that the effect of temperature was related to accelerated RBC damage/aging21. Our research showed that intracellular ROS, Ca2+concentration and PS externalization were increased in stressed RBCs (Fig. 1C–H), these changes were also observed in RBCs of patients with various chronic diseases, such as diabetes22, autoimmune haemolytic anaemia23, haemolytic uremic syndrome5. Nuclei of RBCs are extruded before they enter blood circulation, and there are no organelles such as endoplasmic reticulum and mitochondria in mature RBCs. Thus, RBCs are highly sensitive to changes of internal environment. In some diseases5,22,23,24, hyperosmolar and oxidative stress as well as energy depletion induce changes in RBC membrane surface molecules and intracellular components2. Oxidative stress leads to an imbalance of lipid oxidation, intracellular peroxide accumulation, and elevated ROS level25. Concurrently, ion channels are aberrantly activated and Ca2+ influx increased, resulting in increased intracellular Ca2+concentration, which is related to the translocation of PS from inner leaflet of the cell membrane to the RBC surface26. Thus, we chose ROS, Ca2+ concentration, and PS exposure to represent the characteristics of stressed RBCs. In fact, the changes in stressed RBCs also resemble changes observed in senescent6 and long-term stored RBCs27,28,29.

RBCs under pathological state tend to be phagocytosed by macrophages6. Erythrophagocytosis mainly depends on three pathways. Enhanced PS and band-3 clustering on erythrocytes are the predominant pro-phagocytic signals30,31. CD47 acts instead as a “do-not-eat-me” signal, mediated via interacting with SIRPα32. In the present study, stressed RBCs showed increased PS externalization, which indicated that erythrophagocytosis of stressed RBCs could be mediated by the PS pathway. The in vitro experiment also demonstrated that stressed RBCs were more likely to be phagocytosed by macrophage (Figs. 2, 3C).

This study used high temperature-induced RBCs (stressed RBCs) to investigate the clearance of red blood cells under pathological conditions, but this does not fully represent the RBCs clearance in vivo. In subsequent studies, we will continue to study the in vivo removal of RBCs, such as long-term stored RBCs, and investigate transfusion efficiency of RBCs since the transfusion efficiency of long-term stored RBCs remains poorly understood.

We used 18F-FDG labelled RBCs to measure erythrophagocytosis and organ distribution in vivo, confirmed that stressed RBCs tend to be engulfed and spleen is the main organ for stressed RBCs aggregation. Since stressed RBCs simulate damaged RBCs in pathologic conditions, stressed RBCs trapped in spleen may explain the anaemia of some diseases like diabetes and autoimmune haemolytic anaemia22,23. Besides, this research laid a foundation for future study on stored RBCs transfusion efficiency in clinical.