Streptococcus equi subspecies equi M protein antibody titers

Streptococcus equi subspecies equi M antibody titers on all 48 horses at 8 weeks after infection are shown in Figure 1. One (2%) had a titer of 1:400, 7 (15%) had a titer of 1:800, 9 (19%) had a titer of 1:1600, 7 (15%) had a titer of 1:3200, 8 (17%) had a titer of 1:6400, 12 (25%) had a titer of 1:12 800, and 4 (8%) had a titer of ≥1:25 600. The mean age was similar for each SeM antibody titer. Figure 1 shows the distribution of SeM antibody titers of horses with and without any clinical signs of S. equi infection. The median SeM antibody titers for horses with and without clinical signs were 1:12 800 and 1:1600, respectively. A correlation for horses with clinical signs to have SeM antibody titers ≥1:6400 (Pearson's R = 0.76, P < .05) was noted. S. equi M protein antibody titers of affected horses were compared to their specific clinical syndromes (Figure 2). The median SeM antibody titer was 1:6400 for uncomplicated cases and was 1:12 800 for both persistent GP infection and complicated cases. A correlation for horses with persistent GP infection or complicated cases to have SeM antibody titers ≥1:6400 (Pearson R = 0.58, P < .05; Pearson R = 0.45, P < .05, respectively) was also found. Six out of 16 (38%) horses with very high SeM antibody titers ≥1:12 800 had evidence of complicated cases (Figure 2). S. equi M protein antibody titers of nonsurvivors (n = 4) were all ≥1:6400. For this outbreak, sensitivity and specificity for a SeM antibody titer ≥1:12 800 detecting complications were 75% (95% CI 45‐105) and 43% (95% CI 23‐64), respectively. Using a cutoff of ≥1:6400 instead, the sensitivity and specificity for detecting complications were 100% (95% CI 100‐100) and 22% (95% CI 5‐39), respectively. All horses with persistent GP infection had SeM antibody titers ≥1:6400, and 8 of 12 had a SeM antibody titer ≥1:12 800. Sensitivity and specificity for SeM antibody titer ≥1:12 800 detecting persistent GP infection were 67% (95% CI 40‐93) and 42% (95% CI 20‐64), respectively. Using a cutoff of ≥1:6400 instead, the sensitivity and specificity for detecting persistent GP infection were 100% (95% CI 100‐100) and 26% (95% CI 7‐46), respectively.

Of the 8 vaccinated horses, 2 had a titer of 1:800, 1 had a titer of 1:1600, 2 had a titer of 1:3200, 1 had a titer of 1:6400, and 2 had a titer of 1:12 800. The median titer of vaccinated horses (1:3200) compared to unvaccinated horses (1:6400) was not significantly different (P = .19).

At 12 weeks after infection, SeM antibody titers were measured on 18 horses. Previous titers on these horses were 1:800 (n = 1), 1:3200 (n = 1), 1:6400 (n = 3), 1:12 800 (n = 11), and > 1:25 600 (n = 1). At the 12 week time point, 1 had a titer of 1:400, 5 had a titer of 1:3200, 9 had a titer of 1:6400, and 3 had a titer of 1:12 800 (Figure 3). One horse with a titer of 1:12 800 was being treated for metastatic abscess formation, the second horse with a titer of 1:12 800 had a persistent GP infection with no evidence of other complications, and the third horse with a titer of 1:12 800 had no clinical evidence of disease or complications. Two other horses at 12 weeks after infection had a persistent GP infection and had a SeM antibody titers ≤1:6400. Fifteen out of 18 horses at this time point had a decrease in SeM antibody titer. At 28 weeks after infection, SeM antibody titers were measured on 36 horses. This included all horses that were still on the property or in the area. Eight horses were lost to follow‐up and 4 had been euthanized secondary to S. equi infection. At the 28 week time point, 3 of 36 (8%) had a titer of 1:400, 9 (25%) had a titer of 1:800, 18 (50%) had a titer of 1:1600, 4 (11%) had a titer of 1:3200, 2 (6%) had a titer of 1:6400, and 0 (0%) had titers ≥1:12 800 (Figure 3). No horse at this time point had evidence of clinical disease. Thirty out of 36 horses demonstrated a decline in SeM antibody titer. There was a decline (r = −0.44, P = .001) in mean SeM reciprocal antibody titer at each time point: 7292 (range 400‐25 600) at 8 weeks after infection, 6244 (range 400‐12 800) at 12 weeks after infection, and 1744 (range 400‐6400) at 28 weeks after infection (Figure 3).

Since the beginning of modern virology in the 1950s, transmission electron microscopy (TEM) has been one of the most important and widely used techniques for the identification and characterization of new viruses. Two TEM techniques are usually used for this purpose: negative staining on an electron microscopic grid coated with a support film and (ultra) thin section TEM of infected cells, fixed, pelleted, dehydrated, and embedded in epoxy plastic. Negative staining can be conducted on highly concentrated suspensions of purified virus or cell culture supernatants. For some viruses, TEM can be conducted on contents of skin lesions (e.g., poxviruses and herpesviruses) or concentrated stool material (rotaviruses and noroviruses). For successful detection of viruses in ultrathin sections of infected cells, at least 70% of cells must be infected, and so either high multiplicity of infection (MOI) or rapid virus multiplication is required.

Viruses can be differentiated by their specific morphology (ultrastructure): shape, size, intracellular location or, for some viruses, from the ultrastructural cytopathology and specific structures forming in the host cell during virus replication. Usually, ultrastructural characteristics are sufficient for the identification of a virus at the level of a family. In certain cases, confirmation can be obtained by immuno-EM performed either on virus suspension before negative staining or on ultrathin sections. This requires virus-specific primary antibodies, which might be not available in the case of a novel virus. For on-section immuno-EM, OsO4 post-fixation must be omitted and the partially dehydrated sample must be embedded in a water-miscible acrylic plastic (usually LR White). The ultrastructure of most common viruses is well documented in good atlases and book chapters and many classical publications of the 1960s, 1970s, and 1980s. Several excellent reviews were recently published on the use of TEM in the detection and identification of viruses.

We quantified viral titers in Hazara virus stocks and pooled influenza A virus-positive throat swabs by quantitative reverse transcription-PCR (qRT-PCR), using previously described assays and standards (29, 30).

We collected respiratory samples from the clinical microbiology laboratory at Oxford University Hospitals NHS Foundation Trust, a large tertiary referral teaching hospital in Southeast England. We worked with anonymized residual material from throat and nose swabs generated as a result of routine clinical investigations between January and May 2018. Samples were collected using a sterile polyester swab inoculated into 1 to 3 ml of sterile viral transport medium (VTM), using a standard approach described on the CDC website (28). During the study, respiratory samples submitted to the clinical diagnostic laboratory were routinely tested by a PCR-based test using the GeneXpert assay (Cepheid) to detect influenza A and B viruses and respiratory syncytial virus (RSV). The workflow is shown in Fig. 1. Samples from patients in designated high-risk locations (hematology, oncology, and critical care) were tested using the BioFire FilmArray (bioMérieux) to detect an expanded panel of bacterial and viral pathogens. Quantitative data (cycle threshold [CT]) were generated by the GeneXpert assay, and we used the influenza virus CT value to estimate the viral titers in clinical samples. Using the GeneXpert assay, up to 40 PCR cycles are performed before a sample is called negative (i.e., positives have a CT value of <40). Quantification was not available for the BioFire results.

For methodological assessment, we focused on four categories of samples, as follows: positive pool, negative pools, individual positive samples, and individual negative samples. For the positive pool, we pooled 19 throat swab samples that had tested positive for influenza A virus in the clinical diagnostic laboratory to provide a large enough sample to assess reproducibility (Fig. 1B). For the negative pools, we generated three pools of throat swab samples that had tested negative for influenza virus (consisting of 24, 38, and 38 individual samples) (Fig. 1B). For the individual positive samples, we included 40 individual samples (35 throat swabs and 5 nasal swabs) that had tested positive for influenza A or B virus, selected to represent the widest range of GeneXpert assay CT values (13.5 to 39.3; valid test result range, 12 to 40). For the individual negative samples, we selected 10 individual throat swab samples that were influenza virus negative.

Thirty‐three percent of horses in this outbreak had SeM antibody titers ≥1:12 800 eight weeks after infection. This is in contrast to typical outbreaks and previous SeM antibody titer interpretation.1, 2 A recent study measured SeM antibody titers 1.5 to 27.5 months after natural infection.6 At 1.5 months after infection, none of the 45 horses tested had a SeM antibody titer ≥1:12 800.6 SeM antibody titers ≥1:12 800 occur after experimental commingling infections.7 SeM antibody titers peak at 5 weeks after experimental infection, in which SeM‐specific IgGa titers rise to 1:38 400 ± 14 021 and SeM‐specific IgGb titers rise to 1:102 400 ± 56 089.7 That experimental study did not describe if horses developed any complications after outbreak.7 In our outbreak, 6 out of 8 horses with complications had SeM antibody titers ≥1:12 800, and 2 of 8 complicated cases had SeM antibody titers of 1:6400. One of these cases with dysphagia caused by severe GP disease was later euthanized. The other case had mild signs of purpura hemorrhagica which resolved without the use of immunosuppressive medications. These cases illustrate that SeM antibody titers after infection might not follow the typical interpretations of a titer ≥1:12 800 supporting a diagnosis of metastatic abscess formation or purpura hemorrhagica1, 2; however, an increased titer would warrant further diagnostics to rule out S. equi complications.

Of the horses in this outbreak with SeM antibody titers ≥1:12 800 at 8 weeks after infection, 5 of 16 (31%) had uncomplicated cases, 5 of 16 (31%) had persistent GP infection, 3 of 16 (19%) had complicated cases, and 3 of 16 (19%) had both persistent GP infection and complicated cases (Figure 2). Based on these data and the calculated sensitivity and specificity, SeM antibody titers might be useful for monitoring for complications and risk of complications after outbreak because of high sensitivity but do not necessarily confirm complications are present because of low specificity. S. equi M protein antibody titers do not correlate with a persistent GP infection.1, 8, 9 In our outbreak, 12 horses had persistent GP infection. At 8 weeks after infection, all of these horses (100%) had SeM antibody titers ≥1:6400 and 8 of 12 (67%) had SeM antibody titers ≥1:12 800. The high SeM antibody titers in these horses likely relate to continual exposure to the organism eliciting an immune response. At a SeM antibody titer ≥1:12 800, the sensitivity and specificity of detecting persistent GP infection were low (67% and 42%, respectively); however, when using the cutoff of ≥1:6400, the sensitivity was 100%, but the specificity was unacceptable (26%). A correlation applicable to other populations or outbreaks cannot be determined based on this outbreak, but an increased SeM antibody titer after outbreak could warrant additional diagnostics including testing for carrier status via nasopharyngeal or GP lavage.

This reported outbreak is novel because there was a high proportion of horses with SeM antibody titers ≥1:12 800 (33%). There was also a high proportion of horses with complicated disease (29%) compared to reported complication rates (2%‐20%).1, 2, 4 In this outbreak, case fatality in complicated cases (50%) was high compared to reported case fatality in complicated cases (up to 40%).2 Lastly, there was a high proportion of horses developing persistent GP infection (43%); this is higher than previous reports (up to 10%)2 but consistent with a more recent report (up to 40%).1, 4 Possible explanations for the higher complication rates and higher convalescent SeM antibody titers include a higher dose of exposure, a more virulent S. equi strain,10, 11 or a more naïve population. The causative S. equi organism was not sequenced for the M protein gene in this outbreak. The role of possible concurrent equine enteric coronavirus infections contributing to high complication rates in this outbreak is unknown. It could have resulted in delayed detection and quarantine specific for S. equi transmission, or it could have resulted in immunosuppression allowing more sequelae.

Repeat SeM antibody titers revealed immunologic decay in most cases at each time point consistent with resolution of disease, discontinued exposure to the organism, and are consistent with previous reports of experimental infection.7 As late as 12 weeks after infection, 3 horses had SeM antibody titers of 1:12 800. One horse was being treated for metastatic abscess formation, and the titer was declining consistent with response to treatment. At 28 weeks after infection, all horses had SeM antibody titers <1:12 800. This is similar to a recent outbreak that reported 4.2% of horses having SeM antibody titers ≥1:12 800 after outbreak in which serologic testing was performed at 8 months after infection.12 High titers (1:3200 or 1:6400) persisted in 6 horses at 28 weeks after infection which would have been previously interpreted as consistent with 4‐12 weeks after infection.2 Declining but persistently increased serum SeM‐specific IgGb titer levels were previously reported at 28 weeks after experimental infection.7

Limitations of interpreting data from this outbreak are that the exact dates of infection and times of resolution of clinical signs in each horse were unknown and SeM antibody titers were unknown before the outbreak. Peak SeM antibody titers could not have been detected as SeM antibody titers peak at 5 weeks after experimental infection,7 and serologic testing and testing for carrier status was performed 8 weeks after initial infection in this report. Additionally a selection bias likely existed at the 12 week after infection time point at which the majority of horses rechecked had a previous titer of ≥1:6400. Lastly, 13 horses were not tested for carrier status and were presumed not to have persistent GP infection; however, this could have led to misclassification. At initial testing for carrier status, some horses were tested only once with nasopharyngeal lavage largely based on economics; however, the current recommendations for detection of carrier status are endoscopy and GP lavage.1, 4 This limitation could have led to underestimating the number of horses with persistent GP infection in this outbreak. It could have also led to overestimating the sensitivity and specificity of SeM antibody titers for detecting persistent GP infection as all of these horses had a SeM antibody titer ≤1:3200; however, this outbreak and other studies have indicated that SeM titers are not useful in this manner.1, 8, 9

This outbreak illustrates the utility of SeM antibody titers after infection with S. equi. This study demonstrates that a horse may have complications of strangles without a SeM antibody titer ≥1:12 800 and that a horse may have a SeM antibody titer ≥1:12 800 without complications. A convalescent SeM antibody titer ≥1:12 800 warrants additional investigation for complications or persistent GP infection but does not necessarily confirm a horse has complicated disease.

Given the lack of experimental or epidemiological data on nosocomial virus transmission and AGMPs, current guidelines for infection control are based on the precautionary principle. In order to have a more nuanced understanding of the risks associated with difference AGMPs and viruses, we need more research in both clinical and experimental settings. This type of research could take two different forms, retrospective epidemiological studies or on-site sampling and experimental tests. The former was used to assess the risk of AGMPs and SARS-CoV transmission. However, the quality of retrospective data limits these kinds of studies. Control cases are necessary to rule out other sources of nosocomial transmission besides AGMPs, such as direct patient contact and fomite transmission. Moreover, these studies rely on reporting that may be infrequent and/or unreliable.

Air sampling for viruses during AGMPs performed on patients would provide the most clinically relevant data. Multiple air-sampling techniques for viruses now exist. Researchers can use both solid and liquid impactors to sample and recover aerosolized viruses. Personally worn bioaerosal samplers and stationary room samplers were used to detect influenza A virus RNA in an emergency department. Researchers also used aerosol samplers to determine the amount of influenza A (H1N1) RNA in aerosols in the vicinity of patients while AGMPs were being performed. This allowed the authors to determine which procedures were associated with a higher concentration of viral RNA. In order to determine whether viable virus is present in aerosols, virus isolation could be performed from air samples, and further quantification could occur through titrations or plaque assays. One group used a simulated aerosol chamber to demonstrate that viable influenza virus A could be extracted from surgical masks and N95 respirators. Particle sizers may also be used in experimental settings to characterize the size and dispersal of aerosols. One group used particle sizers to measure the size and travel distance of aerosols from patients who underwent AGMPs, such as nebulizer treatment and NIV. Determining the quantity of viable virus expelled from certain patients during AGMPs could help determine phenomena like super-spreading events, while understanding aerosol characteristics such as particle size could elucidate mechanisms of transmission.

While on-site sampling works for current nosocomial transmission events, we can design experiments to gain prospective knowledge. Procedures such as bronchoscopy and intubation are performed on animal models of the high-risk viral diseases we identified, and air sampling during these procedures could determine whether they are aerosol-generating. Experimentally generating virus-laden aerosols of different sizes and under different environmental conditions could also help determine the risks of different viruses based on their stability in aerosols. Researchers could then create risk models for different viruses based on aersol stability, as well as data on the quantity, concentration, travel distance, and size of aerosols formed during AGMPs.

Multiple environmental factors influence the viability of aerosolized viruses, including relative humidity, temperature, UV radiation, and gas composition of the air. These factors affect viruses differently, and so it is important to consider the environments where emerging viruses exist and where AGMPs are being performed. The aforementioned AGMPs include those that are both ubiquitous, such as CPR or manual ventilation, as well as those that are limited to advanced health-care settings, such as bronchoscopy. The risk of transmission from AGMPs may be very different in a field-based treatment unit than from a tertiary care hospital. Additionally, some health-care settings, such as hospitals with biocontainment units, include additional air-handling systems that can limit aerosol exposure to HCWs. Therefore, the risk of nosocomial transmission via AGMPs varies greatly based on the environment, and this must be considered when designing experiments.

Results of the serological assay of 709 bat sera by ELISA against the three viruses are shown in S2 Fig with 88 of them being further confirmed by WB (S3 Fig). Since no standard bat sera (either positive or negative) were available, the highest coincidence rate (CR) between WB and ELISA was used to determine OD492 ELISA positive cut-off values: 0.10, 0.10 and 0.11 at the highest CR value for each virus (87.5%, 86.4% and 86.4%, respectively, for LAIV, XSV and SEOV) (see S4 Fig). With such cut-offs, the κ test showed high levels consistence between the two methods with Z values being 7.0862 for LAIV, 6.8255 for XSV, and 6.9270 for SEOV (all p<0.0001), and the high κ values being 0.7260–0.7505. These results indicate that the established ELISA was valid to test the bat sera. Using these cut-offs, 131 of 709 (18.5%) bat sera were found to be HV antibody positive. Fig 3 shows the distribution of OD492 readings of the 131 positive bat sera, with most sera having OD492 readings between the cut-off and 0.30. To further determine antibody titers, the positive sera were 4-fold diluted from 100× to 1,600× and retested by ELISA. Results showed that most positive sera had titers of 100×, yet 18 sera reached 400×, with the H anti-SEOV at 1,600× (Fig 4). Of 131 positive sera, 55 (7.76%) showed cross-reactivity to all three viruses, 19 (2.7%) to both of LAIV and XSV, 9 (1.3%) to both XSV and SEOV, and 7 (1.0%) to both LAIV and SEOV, whereas sera reacted exclusively with one virus were only 9 (1.3%) with LAIV, 10 (1.4%) with XSV and 22 (3.1%) with SEOV. This further showed that seroprevalence of HVs in bats widely existed in four provinces (in Guangdong bat sera were not collected). As shown in Fig 5, among 13 cities with bat serum collection 12 were seropositive with levels from 5.5% to 35.9%. Of 16 bat species tested 13 had seropositive rates ranging from 4.8% to 50.0%.

We have successfully expressed rRVFV-N protein in E. coli and purified the protein to near homogeneity by his-tag based affinity chromatography under native conditions (Fig. 1). The purified rRVFV-N protein reacted with RVFV rabbit hyper immune serum and mouse anti-histidine antibody in Western blot (Fig. 2). Most of the expressed rRVFV-N protein was in the soluble form and the purification process was done under native condition without using any detergent, and thereby omitting the tedious process of refolding a denatured protein. The expression and purification procedures described in this study provide a simple and efficient way to obtain pure rRVFV-N protein in large quantity.

Recombinant RVFV N protein based ELISA systems have been reported in the diagnosis of RFV infection in humans and animals [19–24]. Fafetine et al., reported indirect IgG and indirect IgM ELISA systems for domestic sheep, goats and cattle and compared a recombinant N protein based indirect IgG ELISA with an IgG sandwich ELISA using sera collected from small domestic ruminants. Jansen van Vuren et al. reported indirect IgG and indirect IgM ELISA system for humans and sheep using recombinant RVFV N protein as coating antigen. Paweska et al. evaluated recombinant N protein based indirect IgG ELISA for humans and African buffalos [21, 22]. Williams et al developed an IgM capture ELISA for domestic animals. In all these reports, rRVFV-N protein works well as an antigen, proving the usefulness of rRVFV-N protein. But most of the reports have used indirect IgG and IgM ELISA systems where the coating reagent is the assay antigen, which in turn is allowed to react with the test serum supposedly containing the primary antibody followed by an enzyme-labeled secondary antibody. Recombinant RVFV-N protein-based IgG sandwich ELISA and IgM capture ELISA systems for humans have not been reported. Unlike the indirect ELISAs, the IgG sandwich and IgM capture ELISAs have antibodies as coating reagents to capture the assay antigen and the IgM from test sera, respectively. It has been shown in small ruminants that IgG sandwich ELISA is more sensitive than IgG indirect ELISA. For human diagnosis, indirect IgM ELISA suffers from potential false-positive results because of the existence of rheumatoid factor in some individuals and from false-negative results because of the competition from the IgG antibody. IgM capture ELISA is recommended for more sensitive and more specific detection.

Our evaluation on the reactivity of the rRVFV-N protein in the inactivated virus-based IgG sandwich ELISA and IgM capture ELISA using 12 human serum samples for each ELISA system showed that among the 12 samples tested for IgG (Fig. 3) and another 12 samples for IgM (Fig. 4), the results using the rRVFV-N protein perfectly matched with those using the inactivated virus antigen. This indicates that N protein is an important immunoreactive component of the RVFV.

Using the rRVFV-N protein, we then developed our own IgG sandwich ELISA and IgM capture ELISA for human serum and compared with the inactivated virus-based ELISA systems. After applying them to 96 healthy human serum samples collected during the RVF surveillance programme in Kenya or to 93 serum samples collected from RVF-suspected patients during the 2006–2007 RVF outbreak, we showed that the rRVFV-N protein-based IgG sandwich ELISA (Table 1) and IgM capture ELISA (Table 2) were in 100% concordance to the inactivated virus-based ELISA systems with a sensitivity and specificity of 100%. These findings clearly demonstrated the usefulness of the rRVFV-N protein-based ELISA systems for reliable clinical diagnosis of RVFV infection in humans.

Compared to the inactivated virus based-ELISAs, which were used by most researchers for the RVF diagnosis, the rRVFV-N protein-based ELISAs used in this study offer several distinct advantages such as the elimination of the use of infectious virus in the antigen production, the easier way to standardize the test because variable factors (virus multiplicity of infection, virus strains, cell line, cell condition) associated with the preparation of inactivated virus are no longer present, the short period of time (within 1 week after cloning) and the low costs required for the preparation of the recombinant antigen. It would be especially useful in cases of large-scale epidemiological investigation and for application in developing countries.

The concept of viruses developed from the observations of Ivanovsky and Beijerinck of “filterable agents”, with the discovery of the causative agent of tobacco mosaic in the 1890s. Yellow fever virus and dengue virus were the first two arboviruses to be isolated early in the 20th century. The pioneering work of Alexis Carrel on the development of many cell and tissue culture methods in the 1910s at the Rockefeller Institute, and later refinements by Maitland, Eagle, and Enders, led to the widespread use of various culture systems as indispensable tools for virus studies. Although these tools have since been used extensively for the in vitro characterization of viruses, they were inadequate for the identification and classification of a flood of novel viruses collected through the YFV surveillance program supported by the Rockefeller Foundation. Jordi Casals, among others, led the use of the complement fixation (CF) test in order to study viruses affecting the central nervous system. The CF test exploits the unique affinity of complement for antigen–antibody complexes. The original assay was developed in the 1920s for the serologic study of YFV, was improved in the 1930s, and its sensitivity and specificity improved again in the 1950s. Using CF tests, Casals and his colleagues were able to classify viruses into antigenic groups. However, the inherent complexity and labour consuming aspects of the assay (titrations of antigen, complement, and hemolysin for optimal outcomes), technical demands (accurate interpretation of outcomes) and the development of alternative assays (see below), have restricted its applicability in laboratories worldwide.

Hirst observed in 1941 that chicken erythrocytes agglutinated in the presence of the influenza A virus and that virus-specific antibodies inhibited agglutination, forming the foundation for the hemagglutination inhibition (HI) test. A decade later and on Theiler’s suggestion, Casals showed that many arboviruses also agglutinated erythrocytes, establishing the HI test as a diagnostic tool for arbovirus infection and identification. The gold standard for arbovirus identification, the plaque reduction neutralization test (PRNT), has its origins in the observations of Stokes and colleagues, dating back in the 1920s when monkeys could be protected against YFV by the inoculation of convalescent sera from patients who had recovered from the disease. By the early 1930s, Max Theiler had adapted the assay for use in mice, in which mixed serum and virus was inoculated intracerebrally. The cell culture adaptation of the test was first demonstrated by Itoh and Melnick in 1957 for studying the seroconversion of Chimpanzees infected with echoviruses, and a year later by Henderson and Taylor to detect antibodies to the eastern equine encephalitis virus. Versions of this assay are now widely used, including the microPRNT, the virus reduction neutralization test (VRNT), the focus reduction neutralization test (FRNT), the rapid fluorescent inhibition test (RFFIT), the flow-cytometry neutralization, the colorimetric micro-neutralization assay (CmNt), and the reporter virus particle-based neutralization assays.

Historically, these methods (virus isolation, HI, CF, and neutralization tests) served as the basis of arbovirus diagnosis for many years, augmenting electron microscopy (see section below), which allowed the visualization of viruses in infected tissues and cell cultures. However, an inherent limitation of the serology-based assays has been their inability to determine whether antibodies in the examined serum were the result of a recent or past infection. This conundrum was solved by determining whether antibodies were IgM (recent infection) or IgG (past infection), using the enzyme-linked immunosorbent assay (ELISA). The introduction of ELISA revolutionized the field by also offering increased specificity and sensitivity for the accurate detection of many viruses.

Overall, although identification of pathologic agents through serologic assays is quite straightforward, there are instances where accurate identification may not be possible due to cross-reactivity. For example, cross-reactivity among flaviviruses poses a challenge in their identification, even when the “gold standard” of PRNT for arbovirus detection is applied, especially in hyper-endemic settings of flavivirus circulation. Several diagnostic labs faced this challenge during the recent emergence and explosive spread of the Zika virus in the Americas. A similar challenge is also common for the serologic diagnosis of bunyavirus infections, which is attributed to their ability to reassort. In this scenario, a novel bunyavirus may be misidentified as a known pathogen due to the presence of the M segment (contributed by the known pathogen), which encodes the immune-reactive envelope proteins (reviewed in).

During the 2013–2016 outbreak of EVD, the Centers for Disease Control and Prevention (CDC) updated its guidelines regarding precautions to prevent the transmission of EBOV in health-care settings. The updated guidelines further emphasized proper personal protective equipment (PPE) and isolation when performing AGMPs on EVD patients. Currently, in addition to standard PPE, the CDC recommends the use of eye protection, airborne infection isolation rooms, and N95 or higher respirators when performing AGMPs on patients with VHFs, SARS-CoV, and avian or pandemic influenza A viruses. The CDC made similar recommendations as an interim guidance for MERS-CoV, and we were unable to find additional CDC AGMP guidelines for Nipah virus, Hendra virus, or hantaviruses. More evidence of nosocomial transmission events of these and other viruses due to AGMPs is likely to prompt future guidelines. Therefore, our understanding of the risks associated with AGMPs and nosocomial virus transmission is not static, and we must continue to improve our knowledge to develop appropriate precautions.

The ambiguity of which procedures and viruses require additional protective measures during AGMPs may lead to breaches in protocol. During many of the cited nosocomial transmission events, HCWs did not use proper eye or respiratory protection. Even when aware of the need for respiratory protection, HCWs may mistakenly wear surgical masks or unfitted N95 respirators, which do not provide proper protection. Additionally, HCWs may not have access to approprtiate PPE depending on the health-care environment. Therefore, while determining the risks of certain viruses and procedures is essential, communicating their respective precautions and providing resources is equally important. Likewise, proper patient triage and diagnosis are the first steps to ensuring that precautions are undertaken when performing AGMPs.

Overall, more research and communication about the risks of certain viruses and AGMPs are necessary to resolve the uncertainty surrounding their role in nosocomial virus transmission. Although we identified certain viruses and procedures that could be high risk and should be experimentally or clinically tested, emerging viruses or novel procedures may also play significant roles. If we are to design proactive infection control guidelines and understand the underlying biology of viral transmission, we must conduct collaborative clinical and scientific research on nosocomial virus transmission and AGMPs.

For each specimen, assays for ten common and newly identified viruses were performed. Briefly, WUPoyV and BOV were tested using monoplex PCRs described previously.18,19 Other viruses were tested using the Luminex platform and multiplex xTAG™ respiratory viral panel assay (RVP Assay) according to the manufacturer's instructions.20 All multiple infection samples were retested. If there was discordance between two tests, the sample was confirmed by monoplex PCR.

Nasopharyngeal aspirates (NPA) were obtained by trained personnel following standard operating procedures within 24 hour after admission. The specimens were transported immediately to the laboratory by sterile viral transport media, then divided into aliquots and immediately frozen at −80°C until further processing.

Total viral nucleic acids (DNA and RNA) were extracted from 200 μl of NPA specimen using the AxyPrep Body Fluid DNA/RNA Miniprep Kit (Axygen, Union City, CA, USA), according to the manufacturer's instructions. Purified DNA and RNA were stored at −80°C in aliquots for further PCR analysis.

Demographic characteristics and possible transmission routes are presented as numbers (percentages). Viral load for RT-PCR are presented as mean ± SEM values. Differences among groups regarding to occupation, working location and vaccination were determined by using Fisher exact test or univariate logistic regression. Odds ratio (OR) values were calculated with 95% confidence intervals (CI). Student's t-test was used for statistical comparisons between continuous variables. P value of less than 0.05 was statistically significant. All analyses were performed using SPSS 13.0 (SPSS Inc., Chicago, IL, USA).

Of 88 bat sera tested by WB, 48 with sufficient volume were further tested for neutralizing antibody (NAb) titers against SEOV by the fluorescent antibody virus neutralization test (FAVNT). Results showed that nine bat sera (18.8%) from four provinces had NAb titers ranging from 32× to 128× (Fig 4), of which five were both WB and ELISA positive, with the other four negative for both. The positive control (H anti-SEOV) had an NAb titer of 513× (Fig 4). Of interest is that one serum from Rousettus leschenaultii bat in Xishuangbanna, BN78, had the highest NAb titer (128×) against SEOV and was WB and ELISA positive only for SEOV, not for LAIV or XSV. Results also showed that some WB and ELISA double-positive bat sera against the three viruses did not neutralize SEOV, including samples BN19, CZ63, CZ26, NP39 and ZJ61.

Recombinant RVFV-N is a safe and affordable antigen for RVF diagnosis. However, the current evaluation used a small number of samples thus, further in-depth validation of the assays are required before these could be used to replace the currently accepted assays based on whole antigen. The newly established rRVFV-N-based IgG sandwich and IgM capture ELISA systems may offer safe and reliable tools for diagnosis of RVFV infection in humans and are especially useful in large-scale epidemiological investigation and for application in developing countries.

The HotNet2 (HotNet diffusion-oriented subnetworks) algorithm is based on a heat diffusion kernel algorithm that considers the heats of individual genes as well as the topology of gene-gene interactions. Because the HotNet2 algorithm can reduce the false positive rate, can identify subnetworks with high biological relevance, and can be sensitive to both real and simulated data, it was used to find significant subnetworks associated with various diseases.

To further screen genes for JEV infection, we applied the HotNet2 algorithm to identify the genes that may contribute to JEV infection. According to the SNP-to-gene mapping method, we mapped the single nucleotide polymorphisms (SNPs) in the phenome-wide association study (PheWAS) data to genes to identify potential genes associated with encephalitis, which exhibits similar symptoms to those of JEV infection. To recognize the gene-interaction networks related to encephalitis, we used the p-values derived from PheWAS data and the HotNet2 algorithm to calculate the subnetwork. We obtained 16 subnetworks that involved 64 genes associated with encephalitis (Table S3). It should be noted that four genes among the three subnetworks belong to the ubiquitin proteasome system (UPS) (Figure 4), which agrees with the results that encephalitis-related viruses, including JEV, West Nile Virus (WNV), and Venezuelan equine encephalitis virus (VEEV), could utilize the UPS to promote viral entry, replication, and release. In addition, the proteins (TAP1, TAP2, TAPBP) interacting with PSMB8 and PSMB9 belong to antigen-loading components that were important in the antiviral innate immune response. The protein ADAR in the subnetwork was reported to inhibit hepatitis C virus (HCV) replication through eliminating HCV RNA by adenosine to inosine editing. These results confirmed that the genes identified by the HotNet2 algorithm were important in JEV infection.

DENV2 New Guinea C strain (NGC) was propagated in Aedes albopictus C6/36 cells, clarified by centrifugation and stored at −80°C. Virus titers were measured by standard plaque assays on BHK cells. Direct and ADE-mediated infections were performed based on a published protocol with minor modifications. Murine 2H2 and 4G2 antibodies were purchased from Millipore. Briefly, DENV2 was incubated in RPMI containing 2% FBS with or without enhancing titers of 2H2 (20 ng for MOI 1 and 200 ng for MOI 10) or 4G2 antibodies (∼20 ng for both MOI 1 and 10) in a total volume of 250 µl for 30 min at 37°C. 2×105 K562 cells in a similar volume of media in 24-well plates were then infected similarly to pseudovirus infection procedures. After incubation for 1.5 h at 37°C, cells were then washed in DPBS and incubated for ∼24 h before harvesting cells for flow cytometry or supernatant for plaque assay. Cells were assayed for productive infection by intracellular prM staining and analyzed by flow cytometry. The supernatant was clarified by centrifugation, frozen at −80°C, before plaque assays were performed with BHK cells. The infection protocol was slightly modified for adherent J774A.1 cells. 2×105 cells were seeded overnight in 12-well plates and infected with a similar total volume (500 µ l) of virus or virus-antibody complex in DMEM containing 2% FBS the next day. After infection and washing, cells were incubated for ∼2 days before harvesting for flow cytometry. Enhancing titers of 2H2 were based on a previous study while enhancing titers of 4G2 were empirically determined in a pilot experiment using serial 10-fold dilutions of the antibody complexed with DENV2 at MOI 1 or MOI 10. The dilution that gave the best enhancement in infection of K562 cells was subsequently used in all experiments. All statistical analysis was performed using a one-tailed Student's t test. P<0.05 was considered statistically significant.

The unpaired Student’s t test was used for data analyses as indicated, and a value of p < 0.01 was considered very significant (**).

To further evaluate the above findings that bortezomib has the potential ability to inhibit JEV infection, we established a mouse model of JEV infection. Four-week-old BALB/c mice were randomly divided into four groups: a PBS group; a JEV-infected group; a bortezomib-treated group; and a JEV-infected and bortezomib-treated group. The mice in the infected groups were intraperitoneally injected with 106 PFU of the JEV P3 strain. We administered bortezomib intravenously once every day for the first two days and then administered it every two days (Figure 5a). As anticipated, most mice in the untreated infected group died of JEV infection with a mortality rate of 90%. In contrast, the mortality rate of the bortezomib-treated infected group was 40% (Figure 5b). All of the mice in the bortezomib and PBS groups survived until the end of the experiment, indicating that bortezomib has the ability to protect mice from death caused by JEV infection.

To verify the effects of bortezomib on clinical symptoms of JEV, we scored the clinical behavior of mice during the experiment. The JEV-infected mice showed different behavior than noninfected mice, including movement limitations, frequent blinking, body stiffening, and hind limb paralysis. The clinical behavior of the bortezomib-treated infected group was alleviated compared with the untreated infected group (Figure 5c), indicating that bortezomib treatment prevented the JEV-infected mice from pain. The mice in the bortezomib and PBS groups did not show any alterations in behavior, suggesting that bortezomib has the potential to alleviate the suffering caused by JEV infection.

Moreover, to further explore the protection of bortezomib against JEV infection in brains, we collected the brain tissues for hematoxylin-eosin (H&E) staining on day 6 and day 23 post infection. As is shown in Figure 5d, the mice in the JEV-infection group suffered from significant meningitis, vacuolar degeneration, and glial nodules, while the symptoms of mice in the bortezomib-treated group were remarkably alleviated. The mice without JEV infection did not show any histological changes, regardless of whether the mice were treated with bortezomib or not. The mice in all groups showed no evidence of meningitis on day 23 post infection. This result indicated that bortezomib could significantly reduce the damage in brains caused by JEV infection. These results further suggested the ability of bortezomib in the treatment of flavivirus infection and confirmed the crucial role of UPS in the lifecycle of flaviviruses. However, as an anticancer agent, bortezomib has many side effects, such as numbness, erythematous plaques or nodules, purpuric eruptions, and folliculitis. Therefore, it is necessary to control the dose in clinical treatment and pay attention to the reaction of patients after taking bortezomib.

EnzChek® Caspase-3 Assay Kit #1, Z-DEVD-AMC Substrate from Invitrogen (Grand Island, NY, USA) was used to test caspase-3 activity, which was performed according to the manufacturer’s instruction.

All data were analyzed using GraphPad Prism software (GraphPad Software, San Diego, CA). Viral titer data were analyzed with the nonparametric Mann-Whitney test. All differences not specifically stated to be significant were insignificant (p>0.05).

Cells were lysed with 1% NP-40 (Thermo Scientific) and Western blot analysis was performed as previously described. C-myc-tagged IFITM proteins were detected by a murine monoclonal anti-c-myc antibody (9E10, Santa Cruz Biotechnology). Endogenous IFITM protein expression was detected by polyclonal rabbit anti-IFITM1 (FL-125, Santa Cruz Biotechnology) or rabbit anti-IFITM2 (12769-1-AP, Proteintech Group, cross reacts with IFITM3 protein). Anti-tubulin antibodies (Sigma) were used as a loading control.

Throughout the 6-month surveillance period, the number of patients with ILI who visited our hospital for treatment peaked in early September and declined significantly after a national vaccination program was initiated in Taiwan on November 1, 2009 (Figure 2). HCWs reporting H1N1 infection occurred most frequently in September, which was compatible with the peak number of patients with ILI in early September (Figure 2). Similarly, the number of HCWs with H1N1 infection declined after national vaccination was introduced. There were no fatal cases in the cohort analysis.

In conclusion, much remains to be understood about the pathogenesis of CCHFV. However, the toolset for studying CCHFV has been steadily improving in recent years with the development of mouse and non-human primate models to a reverse genetics system for CCHFV that will facilitate dissection of the host and viral determinants of CCHFV pathogenesis. These tools will also allow the development and evaluation of novel therapies that reduce or prevent CCHFV-induced morbidity and mortality. Hopefully, the collaboration of multiple institutions in countries around the world toward the development of vaccines against CCHFV will lead to safe and effective vaccines for CCHFV being deployed in populations at risk for CCHFV infection. Lastly, the role of the tick vector in transmission and pathogenesis needs more attention, as does the role of livestock and other animal species in maintaining and transmitting CCHFV. This could lead to more effective measures to block CCHFV transmission.

In addition to vaccines, investigations of antivirals against CCHFV have been conducted. Ribavirin, a nucleoside analog, is suggested by the World Health Organization for the treatment of CCHFV. However, clinical data supporting the use of ribavirin to treat CCHF are inconsistent; some studies report benefits whereas others report no benefit, and meta-analyses of multiple studies suggest that the efficacy of ribavirin is poor or inconclusive

69–

71. Notably, a placebo-controlled study failed to identify a clinical benefit of ribavirin treatment in patients with CCHF

72. A recent meta-analysis demonstrated that ribavirin treatment needs to be started soon after symptom onset (<48 hours) to reduce odds of death

73. In the recent emergence of CCHFV in Spain, although treatment of an infected nurse with ribavirin had mutagenic effects on CCHFV

in vivo and a reduction in viral titers was coincident with treatment start

74, ultimately ribavirin treatment was discontinued because of suspected hemolytic anemia

75, a potential complication of ribavirin treatment

76,

77. The inconsistent data on the clinical benefit of ribavirin for the treatment of CCHFV and the potential for adverse events with ribavirin treatment have caused significant debate in the field

78–

81. Ethical considerations of placebo-controlled studies will likely make further studies of this type difficult

82, preventing definitive conclusions on the efficacy of ribavirin in patients with CCHF.

Studies in mouse models have also shown inconsistent efficacy of ribavirin. Two studies have shown that even early treatment with ribavirin (<6 hours PI) was unable to prevent lethal disease following infection with two distinct clinical isolates of CCHFV

53,

54. However, another study showed that ribavirin could protect against lethal disease following CCHFV strain 10200 infection when administered early and that protection diminished when treatment was delayed or challenge dose was increased

25. However, results from our lab showed that although early ribavirin treatment extended the mean time to death, ribavirin was unable to prevent death in 10200-infected mice

54. The reason for the distinct outcomes in ribavirin-treated strain 10200-infected mice seen between our study

54 and that by Bente

et al.

25 is unknown but could be due to differences in mouse strain (IFNAR

−/− versus STAT1

−/−), challenge dose, or the time treatment was started after infection. Cumulatively, data in humans and mice suggest that while ribavirin may have limited clinical benefit in patients with CCHF, treatment likely needs to be started early in the course of disease to have clinical benefit. This may prove difficult, as the early symptoms of CCHF are non-specific and can progress rapidly to severe, hemorrhagic manifestations

8 and therefore patients may not present to health-care providers until exhibiting the more serious symptoms of CCHF.

Favipiravir is approved in Japan for the treatment of influenza virus infections

83 but has shown promise against other highly pathogenic RNA viruses, including Ebola

84 and Lassa

85,

86. Two studies have evaluated favipiravir against CCHFV

in vivo. In a study by Oestereich

et al., favipiravir treatment was effective in suppressing viral replication and preventing mortality following CCHFV infection, even when treatment was started 48 hours PI

53. Similarly, work by our group has shown that favipiravir treatment could be delayed until 6 days PI, a time point at which mice were exhibiting advanced disease, including death, and still offer significant clinical benefit to CCHFV-infected mice

54. These data suggest that favipiravir may be an effective antiviral for the treatment of advanced CCHF. Furthermore, Oestereich

et al. demonstrated that favipiravir and ribavirin could synergistically inhibit CCHFV

in vitro, allowing lower doses of both drugs to be used

in vivo with clinical efficacy, suggesting that combination therapies in humans may be effective in treating CCHF while reducing unwanted side effects

53. A similar approach has been used in Lassa fever cases

87. In addition, a high-throughput screen using recombinant CCHFV identified a compound, 2′-deoxy-2′-fluorocytidine, with inhibitory activity superior to that of favipiravir or ribavirin

in vitro

88.

In vivo animal studies will be needed to evaluate how this compound performs in animal models of CCHF. Lastly, monoclonal antibodies have shown efficacy against CCHFV

in vivo

43, and several clones were shown to neutralize divergent CCHFV strains

89, suggesting that they may have promise for the treatment of CCHF.