Viral RNA (vRNA) detection based on RT-PCR techniques using various clinical samples or antibody detection using serum samples from patients in the acute and convalescent phases of diseases is commonly used for diagnosis. For vRNA detection, the methods for the complete inactivation of clinical samples are relatively established. For instance, samples containing RNA viruses or virus-infected cells are mixed with the commercially available guanidinium isothiocyanate buffers such as AVL and RLT buffers of the RNA extraction kits provided by QIAGEN (Valencia, CA), and then ethanol is added for complete chemical inactivation [16-18]. Conversely, protocols for the chemical inactivation of serum samples used for serological tests are not determined, since the chemical reagents interfere with the quality in serological tests. Typically, serum samples are treated first by heat inactivation at 56°C for 30 min to minimize the effects of complement on the results of antibody-based assays, and this heating step was thought to eliminate the infectious virus. Although the heating of serum at 56°C for 30 min is enough to eliminate the infectivity of dengue virus [19], previous studies have demonstrated that a longer incubation time (e.g., 60 min) at 56°C, or incubation at a higher temperature, is required for complete inactivation of many RNA viruses because of their various thermal stabilities [20-24]. In the current study, we demonstrated that the conventional procedure for inactivating complement in serum (56°C for 30 min) almost eliminated the infectivity of NiV in pooled human serum. This suggests that the conventional procedure of heating significantly lowers the risk of acquiring accidental infection through manipulation of serum samples, potentially including infectious NiV. However, it was notable that we detected a residual of the live virus once with the titer corresponding to the limit of detection after the heat treatment (Fig. 2), whereas no infectious NiV was detected under strict conditions (56°C for 60 min and 60°C for 30 min) in the triplicate test (Fig. 2 and data not shown). This might indicate that the latter conditions were required to ensure complete inactivation of the serum samples. It might be a rare case that the serum samples of patients contained more than 1 × 107 TCID50/mL of NiV, compared with the levels of viral load reported in an animal model of nonhuman primate [25, 26]. Even with the strict conditions for heat inactivation, a single step of heat treatment would not guarantee complete inactivation. Recently, we reported a similar examination for the effective inactivation of a virus, severe fever with thrombocytopenia syndrome virus (SFTSV) [27]. In the report, CPEs were accidentally observed in some trials of virus isolation from clinical serum samples after heating to 60°C for 60 min, although we had already demonstrated that the same condition fully eliminated the infectivity of SFTSV in the triplicate test of condition check. This inconsistency may be attributed to the difficulty in heating sample tubes evenly with conventional heating devices (water bath or heat blocker), as water often condenses on the lid of the tubes during heating. Thus, we think that the combination of another method with heat treatment would ensure complete inactivation compared with seeking the best condition for the single-step treatment of heating.
In addition to heat inactivation, UV and gamma irradiation can be frequently used for virus inactivation [27-29]. Special apparatuses are required for effective inactivation using UV and gamma irradiation, and these devices are not necessarily equipped in low-containment laboratories or regional bases for testing. Transilluminators for DNA detection (with a long wavelength, 312 nm or 365 nm) are normally found in regional laboratories, although a specific transilluminator with a shorter wavelength (e.g., 254 nm) or a UV crosslinker with adjustable related parameters is required to destroy viral genomes effectively by UV irradiation. Therefore, in this study, a transilluminator with a wavelength of 312 nm was employed to examine the suitable conditions for the effective inactivation of NiV in low-containment laboratories (Fig. 1). Moreover, to maximize the delivery of UV irradiation into serum samples, a clear polypropylene tube with high transparency for UV light was selected as a container for serum samples. Furthermore, sample tubes placed on the transilluminator were all covered with aluminum foil (Fig. 1) to reflect the transmitted and scattered UV lights to the target sample [15]. Whereas UV irradiation for 10 min effectively inactivated the infectious NiV spiked in the growth medium (Fig. 3b), the virus in pooled human serum was not eliminated by the 10-minute irradiation (more than 1 × 104 TCID50/mL of NiV remained inactivated) (Fig. 3a). This result suggests that excess serum proteins prevent the virus from being UV-irradiated. However, 30-minute UV irradiation of the pooled human serum spiked with infectious NiV was effective for virus inactivation (Fig. 3a). Although we demonstrated that UV irradiation is an easy and effective method for the inactivation of NiV, we should consider that each viral genome of the mixture partially damaged by UV light could compensate for their damage to each other via RNA recombination. In paramyxovirus, events of RNA recombination have been reported in natural infection [30, 31]. Thus, the combination of other methods with UV inactivation would ensure the efficacy of inactivation.
Based on the sensitivities of NiV to heating and UV irradiation, we designed a recommended protocol for the effective inactivation of the virus in serum samples, which can be applied for the safe processing of samples in low-containment laboratories (Fig. 1). Using this protocol, we succeeded in inactivating 6.0 ×105 TCID50 of NiV spiked in pooled human serum. In the triplicate test for virus isolation and subsequent blind passages, no CPEs were observed. Moreover, it was shown that the protocol combining heating and UV irradiation could completely eliminate virus infectivity (Table 1). In the confirmation of full inactivation, we performed UV- irradiation step at first before heating, to exclude the possibility that the residual water on the surface of tubes after heating with water bath could affect efficacy of the UV-radiation. But, for treatment of clinical serum samples, heating step could be performed before UV- irradiation. On the other hand, if the virus load in the serum of patients is thought to exceed 1 × 107 TCID50 /mL, the heating condition could be replaced with the stricter condition (60°C for 30–60 min). Importantly, heating serum samples even at 60°C for 30 min never affected or interfered with ELISA titers against SFTSV in our prior report, whereas heating at 60°C for 60 min slightly decreased their titers [27]. Moreover the 10-min UV treatment using the same transilluminator employed in the current study, with which infectious SFTSVs were completely inactivated in tested serum samples, did not at all decrease the ELISA antibody titer.
Additionally, we quantified the virus RNAs in the serum sample with spiked NiV containing 6.0×105 TCID50 (200uL, 90% human serum), by SYBR real-time RT-qPCR using the reported N primer set (142 bp-long target) [32]. Then, the average copy number of the triplicate samples without the inactivation was approximately 8.5 × 105, whereas that of the triplicate samples after treatment with the recommendation protocol was approximately 9.6 × 104 as shown in Additional File 1 (Supplementary Figure 1, and Supplementary Materials and Methods). This may indicate that the serum sample after the inactivation method is still available for vRNA detection, with decrease in sensitivity.
While the protocol for the effective inactivation of NiV was determined in the study, we need to evaluate the risks of processing serum samples in clinical settings, especially in the space outside the bio-safety cabinet. Regarding NiV, it is known that the virus is relatively stable in the liquid phase (growth media, urine, and fruit juice) at the environmental temperature [14]. Thus, NiV in serum samples with a large number of serum proteins is expected to be stable. In this study, NiV spiked in pooled human serum survived for a long time (Fig. 4). The infectious titer was only reduced by 1.08 log10 after leaving it to stand for 1 week at room temperature in the liquid phase. In contrast, the infectivity of NiV at the spot of the virus solution completely dried in the solid phase and quickly decreased within 15–20 h. This might mean that keeping the space for processing samples dried continuously decreases the risk of infection of NiV in laboratories used for serological tests.