In this study, we identified a small number of people who possessed Abs against XMRV Gag CA, regardless of gender or disease condition (PC and CFS), but none of the individuals in the three tested populations retained strong Ab responses to multiple XMRV proteins. We were unable to isolate XMRV from the blood of PC patients and detected no XMRV genes in the blood of any CFS patients.
We screened blood donors and patients with PC and CFS for XMRV Abs using a similar method to that developed as our in-house confirmatory test for human T-lymphotropic virus (HTLV)-1 infection in Japanese blood donors in the late 1980s, as no XMRV-positive human plasma was available to validate XMRV Ab tests. Unlike HTLV and HIV infection, XMRV-positive plasma bound only to Gag CA proteins in our study. However, in feline gammaretrovirus infections, immune responses are not always strong enough to induce a detectable amount of Abs [21
]. In an animal study of XMRV infection, Qiu and colleagues [22
] found that rhesus macaques intravenously inoculated with 3.6 ×106
50% tissue culture infective dose of XMRV showed good Ab responses against Env SU, Env transmembrane subunit (TM), and Gag proteins. In this animal model, transient viremia was observed for less than 2 weeks, but the Ab responses prolonged over 100 days post-inoculation and declined thereafter without boosting, despite high-dose viral inoculation [22
]. These data suggest that XMRV replication is relatively limited in vivo
to induce lasting immune responses compared with HIV and HTLV infection. Alternatively, the anti-Gag CA Abs we observed could account for cross-reactivity with other immunogens, although seven of 11 Ab-positive plasma samples showed high specificity to XMRV Gag (Figure and Table ). In addition, Western blotting of 2262 blood donors by Qiu and colleagues identified two blood donors positive for anti-p30 (CA) Ab and one positive for anti-gp70 (Env SU) [22
]. These Ab-positive blood donors showed no multiple reactivities to viral antigens, as observed in the present study, but the prevalence of the single antigen-reactive donor was much lower than that in our current result (0.13% vs. 1.6%, respectively). It is possible that the positive reaction to CA protein might include more cross-reactivity in our study. Further investigation of human plasma collected from individuals clearly infected with XMRV is required to verify our Ab screening results.
At the beginning of our study, the presence of XMRV in the blood of PC patients had not been reported; however, we speculated that XMRV might infect blood cells similar to the infection of PBMCs by other gammaretroviruses [23
]. We obtained positive nested RT-PCR results on plasma collected from the Ab-positive PC patient only with extensive PCR conditions of 50 cycles using outer and inner primer pairs (Figure , P24). We were, however, unable to consistently detect the XMRV gene in the same patient 4 and 15 months later using freshly collected blood samples. Co-cultivation of activated PBMCs by Concanavalin A and IL-2 with the LNCap-FGC cell line, which is highly susceptible to XMRV [17
], gave rise to devastating LNCap-FGC cell death (data not shown), and we were unable to detect XMRV genes in the cell culture (Figure ). Our data suggest that P24 was perhaps infected with XMRV or some related viruses, but viral replication in the blood was somewhat limited. If this is the case, the prevalence of XMRV in PC patients (one of 67 patients) would be relatively close to that previously reported [5
]. We cannot, however, exclude the possibility that the positive P24 signal in the PCR assays was caused by contamination, as discussed recently [24
]. We did not PCR-amplify mouse-derived genetic materials [24
] because of the lack of remaining P24 test sample that tested positive for XMRV PCR, although we did use a hot start Taq polymerase that is inactivated not by anti-Taq mouse mAbs but by chemical modification in our RT-PCR test [26
We were unable to detect XMRV DNA or RNA in CFS patients, in accordance with the results of some previous studies [8
]. It is unlikely that our detection procedures caused such a big difference from those studies that reported a prevalence of 67% or 86.5% [2
], because all studies employed highly sensitive PCR methods. The difference may instead be explained by the characteristics of patient populations. All CFS patients in our study met the Centers of Disease Control and Prevention (CDC) diagnostic criteria [27
]; however, the currently employed diagnosis of CFS is not based on objective and quantitative measures but on the claims of patients and some authorized criteria.
Although our results of Ab screening are ambiguous, we conclude that XMRV infection is not involved in the onset and/or progression of PC and CFS in the population we screened. Even if the Abs we detected, or at least the XMRV-specific ones, were caused by XMRV infection, there was no statistically significant difference in the serological prevalence of XMRV among the three populations of the study. Moreover, the negative or inconsistent PCR results in the Ab-positive patients can be explained by the limited replication of XMRV in vivo
. Alternatively, by assuming that the Ab reaction is attributable to cross-reactivity, the negative PCR results likely indicate the absence of XMRV infection in patients. In either case, our results do not support an association between XMRV and CFS, in line with previous findings [8
Retroviral integration is theoretically harmful to the host cell because it disrupts the host genome. To reduce the risk of XMRV infection during blood transfusion, a reliable screening strategy should be established. The implementation of such a screening or inactivation protocol for blood products, however, will be influenced by the evaluation of the prevalence of XMRV by a universal test with high sensitivity and specificity, which must be urgently developed.