Lack of evidence for human infection with Xenotropic murine leukemia virus-related virus in the Brazilian Amazon basin

Introduction: This study confi rmed the absence of natural infection with Xenotropic murine leukemia virus-related virus (XMRV) or XMRV-related disease in human populations of the Brazilian Amazon basin. We demonstrated that 803 individuals of both sexes, who were residents of Belem in the Brazilian State of Pará, were not infected with XMRV. Methods: Individuals were divided into 4 subgroups: healthy individuals, individuals infected with human immunodefi ciency virus, type 1 (HIV-1), individuals infected with human T-lymphotrophic virus, types 1 or 2 (HTLV-1/2), and individuals with prostate cancer. XMRV infection was investigated by nested PCR to detect the viral gag gene and by quantitative PCR to detect pol. Results: There was no amplifi cation of either gag or pol segments from XRMV in any of the samples examined. Conclusions: This study supports the conclusions of the studies that eventually led to the retraction of the original study reporting the association between XMRV and human diseases.

Xenotropic murine leukemia virus-related virus (XMRV) was first identified by Urisman et al. 1 , who studied the correlation between XMRV infection and the R462Q mutation of the ribonuclease L (RNase L) antiviral protein in familiallinked cases of prostate cancer (PC).Schlaberg et al. 2 detected XMRV proteins in malignant epithelial cells, although no correlation was observed between XMRV infection and the R462Q mutation.A similar lack of association was found in other studies performed in Mexico 3 and the United States 4 .
The study of XMRV in blood samples from chronic fatigue syndrome (CFS) patients by Lombardi et al. 5 revealed that 67% of patients tested positive for the virus, compared to 3.7% in the control population.XMRV infection has also been investigated in patients with respiratory tract infections, and approximately 10% of immunodefi cient patients were infected with XMRV, compared to 3.2% of immunocompetent patients.In addition, salivary viral ribonucleic acid (RNA) concentrations in infected individuals, although relatively low, were suggestive of a potential transmission route 6 .
Xenotropic murine leukemia virus-related virus proteins were detected in 4.2% of Japanese people infected with human T-lymphotrophic virus, type 1 (HTLV-1), although this result may have been an artifact of the high degree of similarity between gp21 in HTLV-1 and p15E in XMRV.Positive serological fi ndings from previous studies may have also been the result of cross reactions 7 ; however, antibodies specifi c to the virus were apparently detected 8,9 .Other studies showed a prevalence of between 1-3% in the general (i.e., healthy) population, which indicated a possible role of blood transfusions in the dissemination of the virus 10,11 .
New information about XMRV has raised controversy regarding the prevalence of the virus.Potential confounding factors include: a) the diffi culty of reproducing results obtained for both the general population and patients with PC or CFS; b) the origin of the virus in relation to endogenous murine leukemia virus; c) the contamination of human blood and tissue samples with murine deoxyribonucleic acid (DNA); and d) the contamination of commercial polymerase chain reaction (PCR) reagents with potential sources of false-positive results [12][13][14] .
The results of subsequent studies involving PC and CFS patients failed to show detectable XMRV or found extremely low levels of proviral DNA in the samples, indicating that PCR detection of the virus is not always feasible or reliable 13,[15][16][17] .

METHODS
In addition, the lack of an association between viral infection, new of PC, and the RNase L R462Q mutation further questioned the potential of XMRV as a human oncovirus 14,18 .
The present available data led to the retraction of the original data reporting an association between XMRV and human diseases [19][20][21] .A wide variety of infectious agents, many of which are capable of infecting humans, are endemic to the Brazilian Amazon.Thus, the present study aimed to determine the presence or absence of natural XMRV infection in human populations in the Amazon and its potential associations with diseases.

Sample collection
DNA samples acquired in previous studies conducted at the Virus Laboratory of the ICB/UFPA were obtained from 803 individuals of both sexes who were residents of the City of Belém, State of Pará, Brazil.The subjects were divided into 4 groups as described in Table 1: healthy individuals (n = 290), individuals infected with human immunodeficiency virus, type 1 (HIV-1) (n = 292), individuals infected with HTLV-1/2 (n = 107), and individuals with prostate cancer (n = 114).

DNA extraction
DNA samples were extracted using the Pure Link Total DNA Kit (Life Technologies, USA), according to the manufacturer's instructions.Samples from the healthy individuals and HIV-1 and HTLV-1/2-infected individuals were obtained from leukocytes.DNA from patient PC samples was extracted from paraffi n blocks and frozen tumor tissues.

Molecular analysis
Infection by XMRV was investigated using nested PCR to amplify a 413-bp fragment of the gag gene, as described previously 10 .In addition, the XMRV pol gene was detected by quantitative polymerase chain reaction (qPCR), using the StepOnePLUS™ Real-Time PCR System (Life Technologies, USA), as described by Schlaberg et al. 2 .Detection of the human glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene was used as an endogenous control for qPCR reactions.The AO-H4 and AM2-9 plasmids, encoding XMRV cDNA derived from viral RNA isolated from a prostate cancer specimen (VP62), were used as positive controls 1 .TaqMan FAM-labeled probes were designed to target conserved regions of both the XMRV gag and pol genes and in the GAPDH gene.Plasmid samples were kindly provided by Dr. Nicole Fisher of the Institute for Medical Microbiology, Virology and Hygiene, at the Hamburg-Eppendorf University Medical Center in Hamburg, Germany.

Ethical considerations
The present study was submitted to and approved by the Ethics Committee of the Federal University of Pará (Number: 0024.0.324.324-11).
Analysis by qPCR indicated that the pol segments of XMRV were not amplified in any of the samples tested (Table 1 and Figure 1), although positive control templates (i.e., XMRV cDNA plasmids) and endogenous controls (i.e., GAPDH) were detected.Nested PCR also failed to amplify the targeted gag segment in patient samples, but it was amplifi ed in the controls.
Xenotropic murine leukemia virus-related virus was initially regarded as a virus that was capable of infecting humans and being transmitted through similar routes to those used by other human retroviruses.It was additionally thought to occur in co-infections, for example with HIV-1 or HTLV, as proposed by Groom et al. 16 and Qiu et al. 9 Although a previous study described a high rate of HIV-1/HTLV-1/2 co-infection 22 , the present study did not fi nd evidence for co-infection with XMRV among people infected with HIV-1 or HTLV-1/2, confi rming previous results from HIV-1 carriers in Germany 23 .We also found no evidence of XMRV infection in prostate tumor samples, contradicting the results of previous studies that found an apparent association between this retrovirus and prostate cancer [1][2][3][4] .Similar negative results were attributed to the failure of PCR procedures 13,[15][16][17] .However, it is important to note that in the present study, nested and real time PCR were used, and both procedures were effective in detecting proviral DNA of HIV-1 and HTLV-1/2 in several different groups, even when viral copy numbers were extremely low [24][25][26][27] .

Rev Soc Bras Med
Several factors can interfere with the diagnosis of XMRV infections in humans.The most likely explanation for false positive results is contamination of commercial PCR reagents [12][13][14] .Additionally, Tuke et al. 28 proposed that a specifi c Taq DNA polymerase may contain monoclonal mouse antibodies with varying quantities of murine DNA.This DNA is detectable in some batches and can thus generate false positive results.The results of the present study did not support this conclusion; the reagents used were provided by several different manufacturers, including the manufacturer used by Tuke et al. 28 , and no false positives were observed.It thus seems more likely that contamination of reagents, human blood or tissue samples with murine genomic DNA was restricted to only a few research laboratories, which would account for the divergence of results.Paprotka et al. 29 proposed that XRMV originated from a single recombination event between 1993 and 1996, involving 2 endogenous MLVs (i.e., PreXMRV-1 and PreXMRV-2) derived from a xenograft of the 22Rv1 cell line in nude mice.The possibility that this recombination event occurred in multiple locations is highly unlikely, given the marked similarity of sequences isolated in different studies.XMRV was therefore A B considered a virus that originated from a single recombination event and did not have a natural human reservoir 13,30 .It thus seems likely that all the subsequent isolates of the virus were derived from this unique recombination event because of crosscontamination in various laboratories when handling 22Rv1 and other susceptible cell lines.From 1999 onwards, 22Rv1 cell lines were distributed worldwide and were largely used in biological studies of prostate cancer 29 .
Clearly, there was no contamination in the present study.Furthermore, the absolute lack of XMRV in the various groups investigated herein challenges the results of previous studies that showed a high prevalence of XMRV in human populations and alleged an association with prostate cancer.This study supports the conclusions of studies that eventually led to the retraction of the original study reporting the association between XMRV and human diseases 19-21 .

FIGURE 1 -
FIGURE 1 -A: Dose-dependent qPCR detection of the XMRV plasmid control (red) at concentrations of 333.6, 83.4,20.8, 5.2, 1.3, 0.33, and 0.08ng/uL and qPCR detection of the GADPH gene (black).The GAPDH gene was used as an endogenous control for qPCR reactions, and XMRV cDNA encoded in the AO-H4 and AM2-9 plasmids was used as a positive control for pol detection.All patient samples were XMRV-negative as assessed by real time PCR and nested PCR.B: Ethidium bromide staining of a 1.5% agarose visualized under UV light, showing nested PCR results for the amplifi cation of a 413-bp fragment of the gag gene.Lanes: 1: 100-bp DNA ladder; 2-3, plasmid controls; 4-17, representative negative XMRV patient samples.ΔRn (Delta Rn) is the normalization of Rn (fl uorescence of the reporter dye divided by the fl uorescence of a passive reference dye) obtained by subtracting the baseline (ΔRn = Rn -baseline).qPCR: quantitative polymerase chain reaction; XMRV: xenotropic murine leukemia virus-related virus; GAPDH: glyceraldehyde 3-phosphate dehydrogenase; cDNA: complementary DNA; UV: ultraviolet.