The authors state that they have no conflicts of interest.
Multicenter Blinded Analysis of RT-PCR Detection Methods for Paramyxoviruses in Relation to Paget's Disease of Bone†
Article first published online: 8 JAN 2007
Copyright © 2007 ASBMR
Journal of Bone and Mineral Research
Volume 22, Issue 4, pages 569–577, April 2007
How to Cite
Ralston, S. H., Afzal, M. A., Helfrich, M. H., Fraser, W. D., Gallagher, J. A., Mee, A. and Rima, B. (2007), Multicenter Blinded Analysis of RT-PCR Detection Methods for Paramyxoviruses in Relation to Paget's Disease of Bone. J Bone Miner Res, 22: 569–577. doi: 10.1359/jbmr.070103
- Issue published online: 4 DEC 2009
- Article first published online: 8 JAN 2007
- Manuscript Accepted: 3 JAN 2007
- Manuscript Revised: 7 NOV 2006
- Manuscript Received: 18 AUG 2006
- Paget's disease;
Conflicting results have been reported on the detection of paramyxovirus transcripts in Paget's disease, and a possible explanation is differences in the sensitivity of RT-PCR methods for detecting virus. In a blinded study, we found no evidence to suggest that laboratories that failed to detect viral transcripts had less sensitive RT-PCR assays, and we did not detect measles or distemper transcripts in Paget's samples using the most sensitive assays evaluated.
Introduction: There is conflicting evidence on the possible role of persistent paramyxovirus infection in Paget's disease of bone (PDB). Some workers have detected measles virus (MV) or canine distemper virus (CDV) transcripts in cells and tissues from patients with PDB, but others have failed to confirm this finding. A possible explanation might be differences in the sensitivity of RT-PCR methods for detecting virus. Here we performed a blinded comparison of the sensitivity of different RT-PCR–based techniques for MV and CDV detection in different laboratories and used the most sensitive assays to screen for evidence of viral transcripts in bone and blood samples derived from patients with PDB.
Materials and Methods: Participating laboratories analyzed samples spiked with known amounts of MV and CDV transcripts and control samples that did not contain viral nucleic acids. All analyses were performed on a blinded basis.
Results: The limit of detection for CDV was 1000 viral transcripts in three laboratories (Aberdeen, Belfast, and Liverpool) and 10,000 transcripts in another laboratory (Manchester). The limit of detection for MV was 16 transcripts in one laboratory (NIBSC), 1000 transcripts in two laboratories (Aberdeen and Belfast), and 10,000 transcripts in two laboratories (Liverpool and Manchester). An assay previously used by a U.S.-based group to detect MV transcripts in PDB had a sensitivity of 1000 transcripts. One laboratory (Manchester) detected CDV transcripts in a negative control and in two samples that had been spiked with MV. None of the other laboratories had false-positive results for MV or CDV, and no evidence of viral transcripts was found on analysis of 12 PDB samples using the most sensitive RT-PCR assays for MV and CDV.
Conclusions: We found that RT-PCR assays used by different laboratories differed in their sensitivity to detect CDV and MV transcripts but found no evidence to suggest that laboratories that previously failed to detect viral transcripts had less sensitive RT-PCR assays than those that detected viral transcripts. False-positive results were observed with one laboratory, and we failed to detect paramyxovirus transcripts in PDB samples using the most sensitive assays evaluated. Our results show that failure of some laboratories to detect viral transcripts is unlikely to be caused by problems with assay sensitivity and highlight the fact that contamination can be an issue when searching for pathogens by sensitive RT-PCR–based techniques.
Paget's disease of bone (PDB) is a common condition characterized by focal abnormalities of increased bone turnover, leading to complications such as bone pain, deformity, pathological fractures, and deafness.(1) Genetic factors play an important role in PDB, and mutations have now been identified in four genes that cause PDB and the related syndromes of familial expansile osteolysis, expansile skeletal hyperphosphatasia, and juvenile Paget's disease. There is also evidence to suggest that environmental factors play a role in PDB because the incidence and severity of the disease has fallen over recent years in the United Kingdom and New Zealand.(2,3) Environmental factors that have been implicated as triggers for PDB including dietary calcium deficiency,(4) repetitive mechanical loading,(5) and paramyxovirus infection.(6) Paramyxoviruses were first suggested as a cause of Paget's disease in 1974, when nuclear inclusion bodies were found in Pagetic osteoclasts that were thought to resemble viral particles.(7) Several investigators have reported the presence of these structures in Pagetic osteoclasts,(8) and they have also been found in osteoclasts derived from patients with familial expansile osteolysis.(9) These inclusion bodies are not specific for Paget's disease, however, because ultrastructurally identical inclusions have been reported to be present in macrophages from patients with primary oxalosis.(10) Extensive research has been conducted over the past 30 years to try and detect other evidence of paramyxoviruses in Pagetic tissue by immunohistochemical staining, in situ hybridization (ISH), and RT-PCR–based techniques. Early immunohistochemical studies showed positive staining for measles virus (MV) and respiratory syncytial virus (RSV) in PDB bone samples and cultured cells from patients with PDB,(11) whereas later work using ISH and ISH-RT-PCR found evidence of canine distemper virus (CDV) transcripts in Pagetic bone.(12,13) In contrast, evidence of MV expression has been reported in a subset of peripheral blood cells from pagetic patients by some workers,(14) whereas others have found no evidence of MV in pagetic bone using ISH-RT-PCR.(15) Studies using RT-PCR to detect paramyxovirus expression in bone have also yielded conflicting results. Some workers have found no evidence of paramyxovirus transcripts in PDB bone, bone marrow cells, or peripheral blood cells from patients with PDB.(16,17) One group has predominantly detected CDV transcripts in pagetic bone samples,(18,19) whereas another group has detected MV.(14,20) It is currently unclear whether failure to detect viral transcripts in some laboratories is caused by problems with sensitivity of the RT-PCR assay, as has been suggested by proponents of the viral hypothesis,(21) or whether the positive results reported by others laboratories is caused by external contamination, as suggested by opponents of the viral hypothesis.(22) The aim of this study was to directly compare the sensitivity with which different laboratories could detect paramyxovirus transcripts on samples of bone RNA using RT-PCR–based methods. Evidence of MV and CDV transcripts in RNA derived from bone and peripheral blood cells from patients with PDB were also sought using the most sensitive assays from the comparative study.
MATERIALS AND METHODS
Participating laboratories and study design
Principal investigators from laboratories who had previously published data on RT-PCR detection of viral transcripts for paramyxoviruses in PDB(14,16–18) were approached to determine whether they were interested in taking part in a “blinded” study to compare between laboratory sensitivity of assays for RT-PCR detection of paramyxovirus transcripts. Other participating laboratories included a virology laboratory in Belfast with a special interest in paramyxoviruses (Professor Bert Rima) and the UK reference laboratory for detection of MV at the National Institute for Biological Standards and Controls (NIBSC; Dr Muhammad Afzal). Three of the four research laboratories who had previously published on RT-PCR detection of paramyxoviruses in PDB agreed to take part.(16–18) One of these laboratories (Aberdeen) used the RT-PCR method used by other researchers who had previously published in this area(14,20) but who elected not to take part in this study. All participating laboratories were asked to perform RT-PCR analysis for MV and CDV on samples of Vero cell total RNA (1 μg) that had been spiked with known amounts of nucleocapsid RNA from MV and CDV ranging from 1 to 100,000 copies for MV and 1 to 10,000 copies for CDV. Each laboratory received 10 control samples of Vero cell RNA that did not contain viral transcripts, two samples that contained 1000 copies of MV and CDV spiked into 1 μg of bone RNA, and two samples of Vero cell RNA that had been spiked with 1 or 10 copies of the plasmids used to generate the RNA samples. The test samples were prepared in a single batch, and replicate aliquots were prepared for distribution to the participating laboratories. The test samples were prepared in the tissue typing laboratory in Belfast (Dr Martin Curran) to minimize the possibility of cross-contamination. The tissue typing laboratory is situated in a different building from the Belfast laboratory that analyzed the samples. In all cases, the RT-PCR analysis was run on a blinded basis by the participating laboratories who fed back results to the tissue-typing laboratory that prepared the samples. Sample codes were broken once all the data had been received by the coordinating laboratory. The RNA spiking method used in this study is an accepted approach to compared sensitivity of RT-PCR–based pathogen detection techniques between laboratories.(23)
Extraction of RNA from bone and Vero cells
Total RNA was extracted from the bone samples, peripheral blood mononuclear cells, and cultured cells using RNAzol according to the manufacturer's recommended protocol (Gibco, Paisley, UK). The integrity of the RNA was confirmed by agarose gel electrophoresis followed by ethidium bromide staining, which showed bands corresponding to intact 18S and 28S ribosomal RNA species. Sample integrity was confirmed by RT-PCR analysis for expression of β-actin and β-2-microglobulin (all samples) and osteocalcin, calcitonin receptor, and TRACP (bone samples) as previously described.(16)
Preparation of viral transcripts
Plasmids containing the entire MV and CDV nucleocapsid genes were prepared by the Belfast laboratory using virus rescue systems as previously described.(24) Plasmid pEMC-Na, which encodes the MV N protein, was obtained from Martin Billeter.(25) Plasmid pEMC-N, which encodes the CDV N protein, was generated in the Belfast laboratory.(24) In both plasmids, expression of the MV-N and CDV-N genes are under the control of a T7 promoter. RNA transcripts were prepared in vitro using T7 RNA polymerase according to the manufacturer's instructions. The reaction products were treated with RNase-free DNAase, phenol chloroform extracted, and ethanol precipitated. The amount of transcript present in the resulting samples was assessed by spectrophotometry at 260 and 280 nm, and the quality of the samples was assessed by gel electrophoresis and ethidium bromide staining to ensure that transcripts of the correct size had been generated and that the amounts of truncated transcripts, which may lead to an incorrect estimate of the copy numbers, were low.
Participating laboratories used previously published methods for the detection of paramyxovirus transcripts by RT-PCR as described in more detail below. For the Aberdeen laboratory, detection of MV was by nested PCR using primers MV1 and MV4 for the first round amplification, followed by primers MV2 and MV3 (Table 1). The Aberdeen laboratory also used the primers and reaction conditions previously used by Reddy et al.(20) to detect MV in PDB (MV7 and MV8). For detection of CDV, the Aberdeen laboratory used primers CDV1 and CDV4 in the first round PCR and CDV2 and CDV3 in the nested PCR. Each test sample was assayed on one occasion with each primer pair, and for each sample analyzed, control reactions were included in which the RT step was omitted (one reaction) and the cDNA was replaced by ultrapure water (three reactions). In all cases, the test RNA samples (5 μl) were reverse transcribed using superscript RT (Gibco) and random hexamer primers (2 μg per reaction; Boehringer, Lewes, East Sussex, UK) according to the manufacturer's instructions. The PCR step was carried out in a 25-μl volume using 2.5 μl of cDNA, 0.6 U of Taq polymerase (Advanced Biotechnologies, Epsom, Surrey, UK), and 1.5 mM MgCl2. The thermal cycling protocol consisted of 1 cycle at 94°C for 4 minutes; 39 cycles at 94°C for 50 s, the annealing temperature for 60 s and 72°C for 90 s; and 1 cycle of 72°C step for 10 minutes. For primers MV1 and MV4, MV2 and MV3, and MV7 and MV8, the annealing temperature was 60°C. For CDV1 and CDV4 and CVD2 and CDV3, the annealing temperature was 55°C. Where nested primers were used, 1 μl of the PCR product from the first-round PCR was used as a template in the nested PCR reaction using the same thermal cycling protocol as in the first round PCR. Where no nested primers were available (MV7 and MV8), a second round PCR reaction was set up using 1 μl of product and the same primers as for the first round. The PCR products were analyzed by gel electrophoresis through a 1.5% agarose gel, stained with ethidium bromide. When PCR products were obtained, the bands were retrieved from the gel, the products were purified, and their identity was confirmed by automated DNA sequencing using the forward and reverse PCR primers as the sequencing primers. Sequences obtained were compared with known viral sequences using the BLAST program. In nested PCR reactions, such samples were reamplified using the nested primers and additional no-template controls were added. We never observed any products in no template controls. All pipetting steps were carried out using filter tips, and all sample manipulations were carried out in a class 2 laminar flow hood. In nested reactions, the template was added in a dedicated “post-PCR” area in the laboratory. The class 2 laminar flow hood area where PCR reactions were set up was never allowed to be used for any work that involved the handling of post-PCR DNA products.
For the Liverpool Center, MV transcripts were detected using the following primer pairs (Table 1): MV1 + MV4, MV2 + MV3, MV5 + MV6, and MVf + MVr, and CDV transcripts were detected using the CDVf + CDVr primers pairs. Primers that were able to detect both MV and CDV were also used as previously described.(17) Each test sample was assayed on three occasions by PCR with each primer pair with identical results. The RNA samples (5 μl) were reverse transcribed using omniscript RT (QIAGEN) and random decamer primers (2 μl per reaction; ABgene) according to the manufacturer's instructions. A PCR master mix was made to give a reaction volume of 10 μl containing 0.4 μl of cDNA, 0.5 U of HotStarTaq polymerase (QIAGEN), and 1.5 mM MgCl2. The thermal cycling protocol was 1 cycle at 95°C for 15 minutes; 39 cycles at 94°C for 15 s, the annealing temperature for 30 s and 72°C for 40 s; and a final cycle of 72°C for 5 minutes. Annealing temperature for primers MV1 + MV4, MV2 + MV3, MV5 + MV6, and MVf + MVr was 60°C, and for CDVf + CDVr primers, it was 56°C. Control reactions were carried out by replacing the cDNA with ultrapure water. The PCR products were analyzed by gel electrophoresis through a 2% agarose gel, stained with ethidium bromide, and visualized under a UV lamp. Selected PCR products were excised and purified from the gel, and their identity was confirmed by automated DNA sequencing using the forward and reverse PCR primers as the sequencing primers. Sequences obtained were compared with known viral sequences using the BLAST program. All pipetting steps were carried out in a class 1 lamina flow hood.
The RT-PCR method used by the Manchester center was similar to that as previously described by Gordon et al.(19) and Hoyland et al.(18) Each test sample was assayed on one occasion with each primer pair. The samples were heated to 95° for 50 s, 55°C for 1 minute (65°C for measles primers), and 1 minute at 72°C for 40 cycles. Products were analyzed on agarose gels stained with ethidium bromide and visualized under UV light. Selected PCR products were excised and purified from the gel, and their identity was confirmed by automated DNA sequencing using the forward and reverse PCR primers as the sequencing primers. Sequences obtained were compared with known viral sequences using the BLAST program.
For the Belfast laboratory, MV transcripts were detected using primer pairs MV1198–1218 and MV1724–1702 and CDV transcripts using primer pairs CDV482–501 and CDV719–701 as shown in Table 1. Each test sample was assayed in on three occasions by PCR with each primer pair with identical results. The same oligonucleotides were used to prime synthesis of cDNA using AMV-RT (Promega, Southampton, UK) according to the manufacturer's instructions, using 20% of the volume of the test samples using essentially the same method as described by Gassen et al.(24) The cDNA synthesis step was omitted for assessment of plasmid DNA contamination. PCR was performed using Taq DNA polymerase (In Vitrogen, Paisley, UK) according to the manufacturer's instructions. The DNA was amplified with 30 cycles using an annealing temperature of 55°C and an elongation time of 2.5 minutes at 72°C. Products were visualized by ethidium bromide staining after gel electrophoresis on 2% agarose gels.
For the measles virus reference laboratory at NIBSC, two methods of MV detection were used. In the blinded study, a RT-PCR, followed by nested PCR, was carried out using primers specific for the N gene region of the measles virus genome (MV-N RT-PCR).(26) In the study of patient samples, the MV-N RT-PCR assay was also used, but the samples were also tested with a quantitative PCR assay specific for the M gene region of the measles virus (MV-M qPCR).(27) The sensitivity limits of the MV-N RT-PCR assay are 2.5 >10−4 plaque forming units of measles virus per reaction.(28) The sensitivity limit of MV-M qPCR assay was established with the M gene–specific synthetic RNA templates that were originally produced from the SSPE measles virus strain as reported for the N gene–specific in vitro–derived RNA transcripts.(28) The detection limit was 40 copies when the template were titrated in a 10-fold serial dilution profile and 10 copies when they were assayed in the 2-fold serial dilution profile.(27)
The MV-N RT-PCR procedure applied was based on the single-step RT-PCR amplification approach described previously. Briefly, in each reaction, 3 μl of RNA was reverse transcribed and amplified with the measles virus N gene–specific primers, MV1/MV2, and other essential reagents supplied in the EZrTth RNA PCR kit (Perkin Elmer, Norwalk, CT, USA). The cDNA synthesis step was carried out at 60°C for 30 minutes on a thermal cycler (Cyclogene; Techne), after which the reaction mixture was subjected to the following assay conditions: 1 cycle of 94°C for 2 minutes, 40 cycles of 94°C for 45 s and 60°C for 45 s; and 1 cycle of 60°C for 7 minutes. For the nested PCR procedure, 2 μl of primary PCR product was subjected to reamplification with MV3/MV4 primer set, dNTP mixture, and AmpliTaq DNA polymerase (0.5 unit/reaction). All nested PCR amplifications were carried out for a single round of 30 cycles of 94°C for 1 minute; 50°C for 0.5 minutes; and 72°C for 1 minute. The reaction products were resolved in 1% agarose gels that were stained with ethidium bromide and visualized under the UV lamp. The MV-M qPCR was carried out on an ABI Prism 7000 sequence detection system using one step EZ RT-PCR kit (Applied Biosystems) as described.(28) A standard 25-μl reaction volume contained 1>EZ buffer, 2.5 mM MnOAc2, 200 nM of each M gene primer (MF1/MR1), 100 nM of probe (MVmp), 0.1 U of rTth polymerase, 0.01 U of AmpErase, and 3.0 μl of template RNA solution that was subjected to the following assay conditions: 50°C for 2 minutes, 58°C for 30 minutes, 95°C for 5 minutes; and a single round of 40 cycles (94°C for 20 s, 60°C for 1 minute). The assay data were analyzed using computer software provided with the instrument.
The results of the between-laboratory comparison are summarized in Table 2. The sensitivity for detection of MV transcripts using the MV-N RT-PCR assay was greatest for the NIBSC, where the sample with 100 MV transcripts was positive. Further dilution studies showed positive results down to 16 transcripts but negative results at 10 transcripts. The Aberdeen and Belfast laboratories had a detection limit of 1000 MV transcripts in Vero cell RNA and bone RNA. The detection limit was also 1000 copies using the method previously reported by Reddy et al.(20) to detect MV in PDB samples. The limit of detection for MV at the Liverpool and Manchester laboratories was 10,000 transcripts. The Manchester laboratory detected CDV transcripts in two of the samples that contained MV transcripts and also detected CDV in one of the negative controls. The sensitivity for detection of CDV transcripts was greatest for the Aberdeen, Belfast, and Liverpool laboratories, where the limit for detection was 1000 transcripts in Vero cell RNA. The Belfast laboratory detected 1000 CDV transcripts in bone RNA, but the Aberdeen laboratory failed to detect signal in this sample. The Aberdeen and Belfast laboratories were able to detect 10 copies of plasmid DNA containing the CDV nucleocapsid, whereas no signal was detected at 1 copy. The Manchester laboratory had a limit for detection of 10,000 CDV transcripts but no signal was detected at copy number below this. The NIBSC laboratory did not assay the samples for CDV.
Clinical samples from 12 patients with PDB and 4 control subjects derived from the United Kingdom were analyzed for the presence of MV and CDV transcripts. The studies for MV detection were performed at NIBSC, and the studies for CDV detection were done at Aberdeen (Table 3). These studies showed no evidence of MV transcripts in the Paget's samples or the control samples using assays with a sensitivity of 16 (MV-N-RT-PCR) and 10 copies (MV-M-qPCR), which according to the results presented in this paper, are between 62.5 and 100 times more sensitive that previous assays used to detect MV in PDB.(14,16,17,20) Similarly, no CDV transcripts were detected using an assay with a sensitivity of 1000 copies, which is 10 times more sensitive than a previous assay that has been reported to detect CDV in PDB.(18,19) All of the patient samples were strongly positive for expression of the housekeeping genes β-actin and β-2-microglobulin, and the bone samples were also positive for expression of the calcitonin receptor, TRACP, and osteocalcin, indicating that the RNA extraction procedure had been effective in obtaining osteoblast and osteoclast RNA. We also analyzed the clinical samples and for MV using the MV7 and MV8 primer sets and reaction conditions used by Reddy et al.(20) but detected no products either on a first round PCR of 40 cycles or a nested PCR, in which one tenth of the products of the first reaction were reamplified in a second reaction of 40 cycles (data not shown).
This study showed that differences in the ability of different laboratories to detect paramyxoviruses in PDB is unlikely to be caused by differences in sensitivity of the RT-PCR assays used for virus detection. In this study, the participating centers used essentially the same assays as had previously been used to screen for evidence of paramyxovirus transcripts in PDB,(14,16–18,20) but the assay sensitivity for laboratories that failed to detect paramyxoviruses in PDB were equal to or better than those that did detect viral transcripts. For example, comparison of the assay sensitivities between laboratories showed that the RT-PCR method used by the Aberdeen and Liverpool laboratories was 10 times more sensitive than that used by the Manchester laboratory for detection of CDV. For MV, the Aberdeen laboratory was 10 times more sensitive than the Liverpool and Manchester laboratories for MV and had equivalent sensitivity for MV as an assay previously used to detect MV transcripts in bone marrow and blood samples from patients with PDB.(14,20) The assay used by NIBSC in the blinded study was 62.5 times more sensitive at detecting MV than the next most sensitive assay.
It has previously been shown that RT-PCR assays for detection of paramyxovirus transcripts can vary by up to 1000-fold,(23) and so the differences in sensitivity between laboratories noted in this study are not altogether surprising. Although it was beyond the scope of this study to try and pinpoint the sources of these differences, a number of possible factors may be responsible including differences in primer binding efficiency, the number of PCR cycles used, and differences in the amount of target cDNA used. Indeed, the lower sensitivity of the Liverpool assay for MV and CDV detection could have been partly because of the fact that they used 5-fold less cDNA than the other laboratories.
Some researchers have repeatedly isolated MV transcripts from bone marrow cells and peripheral blood cells derived from patients with PDB using RT-PCR–based methods.(14,20,29,30) We have previously suggested(22) that these results might be caused by external contamination on the basis that the sequences thus far isolated have all been of the same MV strain (Edmonston), rather than any of the other 22 known MV genotypes. Moreover, the mutations that were reported in the N gene involve strictly conserved residues that are essential for viral function and are not mutated even in the prototypical MV slow virus disease SSPE. These researchers elected not to take part in this study and we were therefore unable to directly compare their performance in MV transcript detection, although when we carried out RT-PCR analysis using the same primer sets and reaction conditions reported by this group,(20) we found that sensitivity was comparable with the Aberdeen and Belfast laboratories and 62.5 times less sensitive than the NIBSC method.
This study has highlighted the potential problems with false-positive RT-PCR results occurring as the result of external contamination. This was shown by the findings reported by the Manchester laboratory, where CDV transcripts were amplified and sequenced from two samples that contained MV transcripts and from one negative control sample. This might be because of the fact that the Manchester group have conducted extensive studies on the role of CDV in PDB using various techniques such as viral infection of cells, RT-PCR, and ISH,(13,18,19,31,32) which would increase the likelihood of cross-contamination with even trace amounts of PCR products or plasmids within the laboratory. It is of interest, however, that these results occurred even though the present studies were performed in a different laboratory and building from those reported by Gordon et al.(12,19) It should be noted, however, that the previous RT-PCR studies reported by Gordon et al. included some negative controls (Vero cells), and no contamination was reported.(19) Also Gordon et al. analyzed RNA samples from three patients with bone disease other than Paget's and found no evidence of MV or CDV transcripts. In this study, participants were required to analyze a total of 23 samples, including 10 negative controls on a blinded basis. The appearance of contamination in only a few of several samples analyzed is consistent with low-level contamination,(33) because with higher levels of contamination, one would have expected positive results in the control samples also.
The assays with the highest sensitivity for CDV detection in this study were in Belfast and Aberdeen, where 1000 nucleocapsid transcripts could be detected. For MV, the most sensitive assay by nearly two orders of magnitude was the nested RT-PCR assay used for nucleocapsid RNA detection, which had a sensitivity of 16 copies, and the quantitative RT-PCR used for M gene detection, which has a sensitivity of 10 copies.(27) Analysis of bone and peripheral blood samples from patients with PDB with the highly sensitive quantitative PCR assays for MV detection revealed no evidence of paramyxovirus nucleic acids in any sample, which argues against a role for persistent MV infection in the patient sample studied. Similarly, we did not detect evidence of CDV transcripts in PDB clinical samples, although the assay methods were much less sensitive than for MV.
Our study has several limitations. One of these is that we did not evaluate other methods of paramyxovirus detection such as ISH and ISH-RT-PCR, which have been suggested by the Manchester group to be more sensitive than RT-PCR at detecting paramyxovirus transcripts in bone.(18) One group has reported that peripheral blood cells from patients with PDB contain MV transcripts by ISH,(29) but others failed to detect measles transcripts by ISH-RT-PCR or ISH in bone samples from patients with PDB.(15,16) Another limitation of our study is that we also did not evaluate factors such as efficiency of RNA extraction from bone which has been previously been reported by the Manchester group to influence the ability to detect paramyxovirus transcripts by RT-PCR.(18) Although it could be argued that this might have influenced the results of the studies on patient samples, it would not have affected the results of the comparative study of RT-PCR detection between centers. Moreover, the bone samples that were analyzed for the presence of MV and CDV transcripts were positive for expression of osteoclast-specific genes (TRACP, calcitonin receptor) and osteoblast-specific genes (osteocalcin), indicating that the RNA extraction procedure was effective at obtaining bone cell RNA.
In summary, this study has largely excluded the possibility that differences in sensitivity of RT-PCR detection techniques can explain the differences in rates of detection of paramyxovirus transcripts in PDB between centers, although we did not look at issues such as efficiency of RNA extraction or compare other detection techniques such as ISH and ISH-RT-PCR. We failed to detect MV or CDV transcripts in RNA samples extracted from bone and blood cells in patients with PDB using the most sensitive techniques available.
Apart from the RT-PCR studies discussed above, other lines of evidence have been presented to suggest that paramyxoviruses might be involved in the pathogenesis of PDB. These include the resemblance of nuclear inclusions found in pagetic osteoclasts with paramyxovirus nucleocapsids(7,8); immunohistochemical staining for measles and respiratory syncytial virus in pagetic osteoclasts(11); and positive hybridization for paramyxovirus RNA in clinical samples from patients with PDB.(14,18) However, other workers have failed to detect evidence of paramyxovirus RNA or protein in pagetic samples using ISH, ISHRT-PCR, or immunostaining.(15,16) Moreover, it has also been reported that the inclusion bodies found in pagetic osteoclasts differ from those observed in the prototypical slow virus infection subacute sclerosing panencephalitis,(16) whereas inclusions identical to those in PDB osteoclasts are found in the macrophages from patients with the genetic diseases hereditary oxalosis(10) and familial expansile osteolysis.(9)
There is good evidence that experimental infections with paramyxoviruses such as CDV can affect osteoclast activity,(31) and it has also been shown that MV infection of bone marrow cells from transgenic mice that express the measles virus receptor (CD46) increases osteoclast formation.(34) Overexpression of the MV nucleocapsid in osteoclast precursors has also been shown to increase bone turnover in mice.(35) Whereas these experiments are of interest, infection of cells with viruses other than MV and CDV are also known to induce expression of host genes that cause osteoclast activation,(36) and some viral proteins have been shown to directly affect bone cell function.(37) One of the best examples is the human T-cell leukemia virus type 1 tax gene, which results in a state of generalized increased bone turnover when expressed in mice with features highly characteristic of PDB, including giant osteoclasts, increased osteoblast activity, woven bone, and marrow fibrosis.(37) These experiments show that paramyxoviruses and paramyxovirus proteins can affect osteoclast activity in vitro and in vivo under specific experimental conditions, but they do not address the issue of whether paramyxoviruses are present in PDB tissue or contribute to the pathogenesis of PDB.
Slow virus infections have been implicated in the pathogenesis of a wide variety of chronic diseases on the basis of PCR-based techniques, including autism,(38) inflammatory bowel disease,(39) type 1 diabetes,(40) and multiple sclerosis.(41) However, in most instances, these findings have not been independently replicated.(26,27,42,43) One of the best examples is the study by Ehrlich et al.,(42) who performed a large-scale blinded analysis using PCR to study the possible relationship between HTLV-1 infection and multiple sclerosis, involving >1000 samples from 16 research centers. Extensive safeguards were put in place to avoid false-positive results. The investigators in this study were unable to detect convincing evidence of HTLV-1 infection in multiple sclerosis, contrary to previous reports, and failed to confirm evidence of HTLV-1 infection in patients that were positive for HTLV-1 in a previous study.(41)
We believe that it would be of great value if further independent prospective studies were performed on a blinded basis using some of the highly sensitive quantitative PCR techniques described here and elsewhere(27) to determine whether paramyxovirus transcripts can truly be detected in peripheral blood cells and bone tissue from patients with PDB.
The authors thank the National Association for Relief of Paget's disease for supporting this study and Dr Uta Gassen for preparing the RNA samples for distribution to the participating laboratories, Dr Peter Wilson for undertaking PCR analysis in Liverpool, and Grace Taylor for assisting with the RT-PCR analysis in Aberdeen.
- 11992 Pathophysiology and Treatment of Paget's Disease of Bone, 2nd ed. Martin Dunitz, London, UK.