Discussion
To aid the establishment of MPXV serology in other laboratories, in the present work methods are described for the detection of OPXV IgG and IgM antibodies as well as neutralising antibodies. To this aim, a small panel of sera from patients after MPXV or CPXV infection or VACV/IMVANEX vaccination was used and IgG and IgM detection was tested by IFA and ELISA and the presence of neutralising antibodies by NT. Furthermore, MPXV, CPXV and VACV were employed as sources of antigens in the IFA and the NT test. The aim was to test discrepancies or agreements between different assays and different antigens for sera with different infection or immunization backgrounds. It could be confirmed that IFAs and NTs based on different OPXV species can be used for the detection of antibodies in sera from individuals who had an infection with MPXV or CPXV or were vaccinated against poxviruses, due to the well-known antibody cross-reactivity between different OPXV species [1, 10].
Some sera (M3 and V3 to V5, Figure 1B) were not detected by all OPXV IFAs as a result of their very low titre (≤20) at or below the limit of detection of the IFA. Additionally, after MPXV infection some sera showed higher titres on MPXV-infected slides, indicating higher reactivity against homologue virus strains as previously described [25].
The greatest differences in reactivity to the different OPXV were observed in the NT. These titre differences could result from the biological variation of the NTs since different OPXV species were used and the tests were not optimised for each virus individually. Since cell culture supernatant as well as virus particles from cells were used, the virus stocks contained a mixture of the two poxvirus particle forms: the extracellular enveloped virions (EEV) and the intracellular mature virions (IMV). For EEV and IMV, different antibodies are needed for neutralization since the epitopes on the virion surface differ. Hence, the ratio of EEVs to IMVs could have been different in the different OPXV stock preparations, leading to larger titre discrepancies. Furthermore, although the amount of virus particles used was set to be comparable between the different OPXV (1000 TCID50/mL), fewer virus particles were detected in the back-titration of the VACV stock (supporting table S1). This could explain the fact that the observed neutralising antibody titres were highest in the VACV NT and not in the NT corresponding to the “source virus” of the antibodies contained in the different samples. However, except for this minor technically induced variation, the NT results corresponded to the expected results: CPXV sera showed the highest titres in the CPXV NT and MPXV sera in MPXV NT. Finally, the larger differences between the titres observed in the different OPXV NTs, as compared to the different OPXV IFAs, could also be due to the generally much lower neutralising antibody titres as compared to binding antibody titres and hence larger variation between individual measurements.
Although serological assays have been described which are able to discriminate between vaccination and infection with MPXV to some extent, those assays rely on pre-absorption with viral antigen [25] or combinations of specific peptides [17] while there is a large overlap of cross-reactive epitopes between the closely related OPXVs. It is known that individuals who received the first-generation poxvirus vaccine against smallpox still have cross-reactive neutralising antibodies against MPXV even 40 years post vaccination [26]. However, in accordance with a recent study, these neutralising antibody titres in individuals with one or two vaccination doses are rather low [12].
The newly optimised ELISA protocol is more suited for the detection of acute infections due to its higher dynamic range which results from quantification over standard curves and its ability to detect IgM antibodies. The ELISA enables a higher throughput as compared to IFA and NT, with 20 sera measured per ELISA plate if two serum dilutions are tested. Testing two serum dilutions (1:100 and 1:1000) is advisable to capture the higher dynamic range of acute infections in patients as compared to seroprevalence studies for which the ELISA was initially established at a single 1:100 dilution. One of the two sera with IFA titres of 1:20, which gave higher ELISA signals than the sera with IFA titres of 1:80, was highly positive for PPV, indicating a possible minor cross-reactivity with the VACV-specific lysate used in the ELISA, while the other serum was initially titrated to a titre of 1:320, indicating some ambiguity regarding the exact titre (Figure 3). Some overlap between ELISA results falling into different IFA titres indicates some method-specific discrepancies; yet overall, both assays correlate well, especially for samples with higher IFA titres. Differentiation between lower IgM titres was not possible, which in part could be due to cross-reactivity of the used detection antibodies, as one serum with a high IgM reading by ELISA had an IFA IgM titre of only 1:20 but was highly positive for IgG with a titre of 1:20,480.
It has been shown that during MPXV pathogenesis two viraemic phases occurred that enabled virus dissemination in the infected individual [3]. In contrast, viraemia in vaccinees seems to be rarer [27]. Viraemia in MPXV-infected individuals also seems to be rare [28] and is potentially also clade specific as differences in dissemination are seen between clade I and clade II MPXV. However, viraemia/DNAaemia has been observed during MPXV infection in a prairie dog model, and in two animals viable virus was also found in the blood [29]. Moreover, DNAaemia was also shown to occur in CPXV-infected individuals [30]. This leads to the conclusion that viraemia cannot be excluded in MPXV-infected individuals. Hence, a safe handling procedure was established for BSL-2 laboratories with which at least a five log-level depletion of infectious poxvirus particles could be accomplished in simulated serum samples. The protocol using Triton, Tween and heat can be used to inactivate sera for IFA and ELISA. Due to cytotoxicity, the inactivation protocol is not applicable for lower dilution steps in NT, but NTs based on VACV or other suitable OPXV may be performed in a BSL-2 laboratory under a class II biosafety cabinet.