Discussion
As it causes devastating losses to the swine industry, coronavirus PEDV has aroused wide mounting concern worldwide (Choudhury, Dastjerdi, Doyle, Frossard, & Steinbach, 2016; C. Lee, 2016). PEDV infection mainly cause diarrhea, dehydration and even death in suckling piglets. Although several reports has indicated that piglets developed viremia after PEDV infection (Chen et al., 2016; Jung et al., 2014), the mechanism of how PEDV spreads in the blood remains unclear. Our previous study has indicated that PEDV could be carried by CD3+T cells to enter the blood (Li et al., 2018). In addition to CD3+ T cells, could PEDV bind to other cell types to facilitate transmission? In the present study, we demonstrated that the coronavirus PEDV could hijack RBCs and spread in the blood circulation, causing intestine infection.
Red blood cells are the most abundant cell in the bloodstream, serving for oxygen transport. Despite of loss of nuclei and organelles, RBCs still express numerous cell surface receptors to contact with the exogenous agents in the blood, performing several additional immunological functions (Baum, Ward, & Conway, 2002). However, several receptors are employed by pathogens for binding RBCs to spread in the blood. The most notorious pathogen is the malaria parasite, which infects RBCs by binding to receptors on the cell surface, such asPlasmodium vivax and Plasmodium knowlesi attaching Duffy Ag receptor for chemokines (DARC) (Boddey & Cowman, 2013; Horuk et al., 1993), and Plasmodium falciparum adhering glycophorin A (GYPA) (Farrow et al., 2011). Moreover, RBCs are considered as possible vehicles for virus transmission, such as HIV-1 (Beck et al., 2009), Dengue virus (Sutherland et al., 2016) and West Nile virus (WNV) (Rios, Daniel, Chancey, Hewlett, & Stramer, 2007). In this study, we verified that PEDV could bind to RBCs and maintain its viability and infectivity for 12 h.
Mounting researches have indicated that there is a small amount of immature RBCs in the circulation of fetuses and newborns, containing a nucleus and cell surface transferrin receptor (CD71) (Elahi, 2014, 2019). The CD71+ RBCs are enriched during pregnancy or newborns, and have distinctive effect on suppressing innate immune responses (Delyea et al., 2018; Dunsmore et al., 2017; Elahi, 2014; Elahi et al., 2013). Similar to previous researches, CD71 was only present in the RBCs of newborn piglets, but not of 40-120 days old piglets. As a transmembrane receptor for guaranteeing iron (Gammella et al., 2017), CD71 is also considered as a preferred entry for human pathogenic arenaviruses (Abraham, Corbett, Farzan, Choe, & Harrison, 2010; Radoshitzky et al., 2011), hepatitis C virus (HCV) (Martin & Uprichard, 2013) and malaria parasite (Gruszczyk, Huang, et al., 2018; Gruszczyk, Kanjee, et al., 2018). Moreover, our previous study has proved that CD71 was considered as a co-receptor for Transmissible gastroenteritis virus (TGEV) (Zhang, Hu, Yuan, & Yang, 2018) and PEDV infection (unpublished), which could be caught by spike protein of PEDV. In our study, the presence of CD71 exerted a significant effect on promoting PEDV adhering neonatal RBCs. CD71 is employed to internalize iron through clathrin-mediated endocytosis. Meanwhile, clathrin-mediated endocytosis also played an important role in facilitating PEDV internalizing into RBCs, just as HIV-1, Influenza and Ebola virus (EBOV) invade target cells (Aleksandrowicz et al., 2011; Mercer & Helenius, 2009). Therefore, PEDV could exploit CD71 and clathrin-mediated endocytosis to internalize in neonatal RBCs.
Through autotransfusion with PEDV-loaded RBCs, we found that PEDV could cause intestinal infection in newborn piglets. Until now, blood transfusions are still the route of transmission of many virus, as is the case for HIV-1 (D. Lee, 2006), Hepatitis E Virus (Izopet, 2018), DENV (Pozzetto, Memmi, & Garraud, 2015), HCV (Schuch, Thimme, & Hofmann, 2015) and WNV (David & Abraham, 2016). The essential characteristics of transmission through blood transfusion are that the virus could survive or maintain viability in the blood or blood components and cause infection through the blood vessels route (Glynn et al., 2013). Although coronavirus PEDV persists in RBCs and maintains its viability and infectivity for a relatively short time (about 12 h), but it is enough for PEDV to hijack RBCs to the target tissues. After 1 h of transfusion, PEDV appeared in the spleen, which is the only lymphoid organ that filters the blood. Macrophage in the spleen could phagocytose RBCs that bind to pathogens (Minasyan, 2016), such as RBCs infected by Plasmodium (de Back et al., 2014). However, after 48 h, transfusion with PEDV-loaded RBCs caused intestine infection and PEDV could colonize in intestinal epithelial cells. PEDV could hitchhike RBCs to reach the small intestine and colonize in IECs (Li et al., 2018). Since RBCs are thought to be not able to penetrate the vascular endothelium under normal physiological condition, we speculated that the RBCs could be recognized and transferred the virus to certain types of immune cells in the blood.
As a sentry in the peripheral blood, peripheral blood mononuclear cells (PBMCs) play a very important role in immune response against both infectious invaders. After co-culture, we verified that PEDV could transmit from RBCs to PBMCs within a short time. Further, CD3+T cells were determined as the main cells types in PBMCs for capturing virus, and could form conjugation to acquire PEDV from RBCs within 1 h. Moreover, PEDV could continue to hitchhike CD3+ T cells and migrate to the intestine, causing typical diarrhea (Li et al., 2018). Several reports have hypothesized that RBCs might be exploited by pathogen as a “Trojan horse” to carry them to the target cells (Baum et al., 2002; Beck et al., 2009). This would be similar with HIV-1 binding to DCs, which could transfer the virus to the susceptible cells in the lymph nodes (Harman, Kim, Nasr, Sandgren, & Cameron, 2013; Manches, Frleta, & Bhardwaj, 2014). However, the detailed mechanism of PEDV transfer from RBCs to T cells needs to be validated.
In general, RBCs circulate in the blood vessels, and are difficult to directly contact with pathogens from outside. The previous study in our group have indicated that PEDV could infect piglets through airborne transmission and develop a transient nasal epithelium infection (Li et al., 2018). Moreover, numerous capillaries are distributed under the nasal epithelial cells (NECs), and some are even adjacent to the NECs. In addition, the permeability of the capillaries in the lamina propria of the nasal mucosa was very high than that of other tissues (Watanabe, Saito, Watanabe, & Mizuhira, 1980). Through intranasal incubation with PEDV, PEDV-loaded RBCs were found in the capillary adjacent to the NECs. Although the mechanism by which PEDV could enter the nasal capillary after infection with NECs has poorly been clarified, nasal capillary might be the possible entry for PEDV binding RBCs. Further, the nasal cavity is a contacting room between respiratory system and environment air. Cells in nasal cavity are exposed in relatively high oxygen condition with 19-21 % oxygen partial pressure (pO2), but are exposed in lower oxygen condition with 6-8 % pO2 (Carreau et al., 2011). And increasing evidences have indicated that oxygen concentration exerted significant effect on virus infection (Ebbesen & Zachar, 1998; Mazzon et al., 2013; Rastelli et al., 2016). In the present study, the relatively high oxygen concentration promoted the affinity of PEDV to RBCs. Besides, the temperature is in nasal cavity is cooler than in other tissues, with an average temperature of 33 °C (Bailey, Casey, Pawar, & Garcia, 2017; Forero et al., 2017), but reducing temperature had no effect on PEDV binding RBCs. Therefore, the relatively higher oxygen concentration in nasal cavity might be a promoting factor for PEDV binding RBCs.
In summary, our study has demonstrated for the first time that the coronavirus PEDV could hijack RBCs and spread throughout the body, causing intestine infection. PEDV could bind and be endocytosed into neonatal RBCs through CD71 and clathrin-mediated endocytosis (Fig 6). Subsequently, PEDV-loaded RBCs could be recognized and transferred the virus to CD3+ T cells in PBMCs by form conjugation. Moreover, the capillary adjacent to the nasal epithelium was speculated as the entry for PEDV binding RBCs. Our studies put forward a new insight into the role of RBCs in coronavirus PEDV as potential cells for virus transmission, which might be helpful to reveal the transmission and pathogenesis of viruses with the same characteristics, even though the detailed mechanisms remain to be clarified.