Airway microparticles, also known as microvesicles and ectosomes, are membrane-bound vesicles of 100-1000 nm in size, produced from various cell types by budding from cell surface [1]. MPs carry cell surface molecules on their surface and a number of cytoplasmic contents. They play important roles in various physiologic processes and pathological conditions [2]. In addition to these roles, we have previously shown that MPs from human BAL could bind and inhibit influenza virus infection through their surface sialic acid, which is the viral receptor [3]. Because MPs in airway secretion are originated from airway epithelial cells and should carry many epithelial cell surface molecules, including those used by respiratory viruses as docking or entry receptors, MPs may inhibit various respiratory viruses using viral receptor on their surface to bind and trap virions in similar fashion to MP sialic acid and influenza virus.
Respiratory syncytial virus or human RSV is a common human-specific virus capable of causing severe respiratory illnesses especially in young children. There is no efficacious vaccine and anti-viral treatment available. RSV can modulate specific immune responses and evade innate immune responses. Complex interactions between viral factors and host adaptive and innate responses allow the virus to infect without inducing solid protective immunity [4]. Understanding the innate defense against this virus may help improve treatment and prevention. We therefore tested whether respiratory tract MPs contribute to the innate defense against RSV.
We first tested whether human BAL MPs carried RSV receptors. A number of molecules have been described as candidates for RSV receptors including heparin sulfate, TLR4, nucleolin, and CX3CR1 [5, 6]. Among these, CX3CR1 has been shown to be the viral entry receptor in human primary differentiated airway epithelial cells [5]. To test whether this receptor is available on MPs in human airway secretion, human BAL samples were stained with anti-CX3CR1 and Annexin V and analyzed by flow cytometry. The 10 BAL samples were collected from individuals who underwent bronchoscopy and BAL for investigation of suspected lung cancer. BAL was performed on non-lesional lung segments. Collection of the samples was approved by the Ethics Committee of the Faculty of Medicine Siriraj Hospital under protocol COA No. Si 191/2016. All the human subjects provided written informed consent. Majority of gated particles were stained positive by Annexin V, which can stain MPs, exosomes and apoptotic bodies. Because of the gating was done using a size marker of 1.34 mM latex bead, most particles in the gate had a size range that was compatible with MPs (Fig. 1A). To distinguish MPs from apoptotic bodies, we double-stained the BAL samples with Annexin V and propidium iodide. BAL samples showed positive staining for only Annexin V whereas a control for the apoptotic body from the supernatant of H2O2-treated cells showed double-staining for both Annexin V and propidium iodide (Fig. 1B). We therefore concluded that majority of the events in the dot plot in Fig. 1C represent MPs. Although there was some variation in the levels of CX3CR1 positive staining among BAL samples, the number of MPs with CX3CR1 positive ranged from 181 to 849 (510 ± 282) particles/ml (N = 10) (Table 1). This indicates that the RSV receptor, CX3CR1, is available on the surface of MPs from human BAL and may be able to trap RSV virions and inhibit the infection. We have previously shown that majority of MPs in BAL were derived from respiratory epithelial cells as determined by their staining for keratin sulfate and surfactant protein D [3]. However, only around 16% of MPs were stained positive for CX3CR1. This is probably because CX3CR1 is expressed only on ciliated epithelial cell while MPs can be released from ciliated, non-ciliated and alveolar epithelial cells.
With the presence of CX3CR1 on BAL MPs, it is likely that they may be able to bind, trap RSV virions and inhibit the infection. The 10 human BAL samples were tested for anti-RSV activity. Figure 2A shows representative data from 10 BAL samples in duplicate. Undiluted BAL samples showed around 80% inhibitory activity against RSV, and serially diluted samples showed declining inhibitory activity in a dose-response fashion. The decline of the inhibitory activity with the dilution did not show the usual end-point type of titration curve. This suggested that the measured inhibitory activity was already below the end point of the optimal activity possibly because the BAL sample was already diluted by the collecting process. However, the anti-RSV activity of BAL samples did not show significant correlation with the number of CX3CR1-positive MPs, indicating that MP was not the only anti-RSV mechanism of BAL. Although the BAL samples showed significant inhibitory activity against RSV, the activity was mostly at moderate level suggesting that the BAL anti-RSV activity may provide limited defense. Whether the BAL anti-RSV activity plays important role in protection against RSV infection requires further studies.
Bronchoalveolar lavage may contain various soluble anti-viral factors. To provide further evidence for MPs anti-RSV activity, MPs and exosomes were partially purified from BAL samples by serial centrifugation. Figure 2B shows anti-RSV activity and MP content of fractions after the serial centrifugation of pooled BALs. Almost all MPs were purified in the 13,000g fraction. Both 13,000g and 100,000g pellets showed anti-RSV activity suggesting that both MP and exosome could inhibit RSV infection, while the pellet washes did not have any anti-RSV activity indicating that the inhibitory activity in the pellet was not due to leftover supernatant. The soluble fraction in 100,000g supernatant also contained anti-RSV activity indicating that BAL contains soluble factors capable of inhibiting RSV infection. A number of soluble antiviral factors have been reported to exhibit anti-RSV activity. Beta-defensin has been shown to inhibit RSV, whereas there are conflicting data regarding the role of surfactant protein A on RSV infection [7-9].
We further tested whether the MP anti-RSV activity was brought about by the presence of the viral receptor CX3CR1 on MPs. To do this, BAL and MP preparation were preincubated with anti-CX3CR1 monoclonal antibody or isotype control before incubating with the virus. Preincubating BAL and MP with anti-CX3CR1 significantly reduced the RSV inhibitory activity (Fig. 3), whereas exosome and soluble fractions did not show any reduction of anti-RSV activity by anti-CX3CR1. This indicates that CX3CR1 on MP contributed to the MP anti-RSV activity. Since the preincubation with anti-CX3CR1 did not completely abolish the anti-RSV activity, it is possible that there were other mechanisms for the anti-RSV activity, for example other RSV receptors might be present on MPs and contribute to the inhibitory activity.
Microparticles are released from the cell surface and therefore inherit cell surface molecules from the plasma membrane of their cellular origin. Most MPs in BAL are derived from respiratory epithelial cells. Those cells carry receptors for various respiratory viruses on their surface. These receptors are likely to be transferred to MP surface making them capable of trapping viruses specific to those receptors. Here we show that BAL MPs can inhibit RSV in a receptor-dependent manner. This further supports the hypothesis that MPs contribute to innate anti-viral defense against various respiratory viruses.