Abbreviations
CTCLcutaneous T-cell lymphomas
HTShigh-throughput sequencing
MCPyVMerkel cell polyomavirus
MFmycosis fungoides
PTLVprimate T-cell lymphotropic virus
TO THE EDITOR
The involvement of infectious, mainly (retro)viral, agents in cutaneous T-cell lymphomas (CTCL) remains a debated issue (
Manzari et al., 1987
; Hall et al., 1991
; Pancake et al., 1995
; Bazarbachi et al., 1997
; Wood et al., 1997
; Morozov et al., 2005
). A recent study using a method detecting virtually all known primate T-cell lymphotropic viruses (PTLVs) failed to identify any PTLV-related sequence in a series of CTCL patients, confirming previous negative reports (Courgnaud et al., 2009
). However, the issue of an implication of a new or already known virus is still pending and deserves additional investigations. In this perspective, the recently developed and highly powerful high-throughput sequencing (HTS) approach is of particular interest, as it may identify new, currently unknown nucleic sequences in biological samples that may be related to infectious agents and viruses in particular. A major recent achievement of this “broad approach” method, free of any underlying hypothesis regarding a particular subset of infectious agents, is the identification of Merkel cell polyomavirus (MCPyV) in an aggressive neuroendocrine skin tumor (Feng et al., 2008
; Shuda et al., 2008
). However, no such experiment has been conducted in CTCL to date.To address this issue, the whole transcriptome of cutaneous lesions from six patients with CTCL (three with stage Ia–IIa mycosis fungoides (MF) and three with Sézary syndrome with blood involvement fulfilling ISCL criteria) was analyzed for the presence of viral transcripts from known or unknown species by HTS in a pilot study. More detailed information about the investigated patients is available as Supplementary Material online. To ensure results homogeneity, blood sample from SS patients was not analyzed. Written informed consent was obtained from all patients and the experiments were conducted in accordance with the Declaration of Helsinki Principles. Total RNA was extracted with Trizol, retrotranscribed, and randomly amplified as recently described (
Cheval et al., 2011
). High-throughput sequencing was subcontracted to GATC Biotech AG (Konstanz, Germany). The single end sequencing was performed with an Illumina (San Diego, CA) GA II on two channels and was conducted with an average depth of 5.8 million reads per sample and a length of 96 nucleotides per read. Sequences were first selected or trimmed according to their quality scores. The human genome was filtered with SOAPaligner (http://soap.genomics.org.cn) using the Homo sapiens hg19 reference. This host filtering step eliminated an average of 91.6% reads per sample. Remaining reads (ranging from 249,570 to 776,095) were assembled in contigs using CLC Genomics Workbench (http://www.clcbio.com). On average, 5,668 contigs with a length of 100 nucleotides or more were generated (with a minimum of 830 contigs and a maximum of 15,770 per sample). A comparison of the single reads and contigs with available genomic and taxonomic data was made on the generalist nucleotidic (nt) and proteic (nr) databases maintained locally. The aforementioned databases were scanned using the BlastN and BlastX algorithms provided by Paracel Blast (Striking Development, Los Angeles, CA), a software capable of executing searches on multiple nonshared memory processors simultaneously. Binning (or taxonomic assignment) was based on the best hit among reads with a significant e-value (below 10−3). Nonassignated sequences corrresponded to 2–13% of the contigs (maximal size 922nt) and 5–22% of the reads.- Supplementary Table
As disclosed in Table 1, no known viral transcript was detected in CTCL cutaneous lesions, except for rare hits against endogenous retroviruses that are common in all tissue samples we have examined till now, regardless of their origin (not shown). Moreover, no singleton (nonassembled read) exhibited a significant homology with a known virus (not shown), whereas singletons from housekeeping genes (β-actin, β2 microglobulin) were detected in each of the six samples (not shown). A number of bacterial transcripts of various origins were also detected, partially related to a contamination from the skin or the environment as recently described by us in a metagenomic study of skin surface (
Foulongne et al., 2012
).Table 1Viral and bacterial contigs derived from the transcriptome of six patients
| Number of contigs | Average contig identity (%) | Number of assembled reads | |
|---|---|---|---|
| Eukaryotic viruses | |||
| Human endogenous retrovirus | 2 | 99.6 | 5 |
| Human endogenous retrovirus K | 1 | 98.0 | 2 |
| Human endogenous retrovirus HERV-K (I) | 1 | 97.3 | 5 |
| Bacteria | |||
| Actinobacteria | 19 | 47.9 | 317 |
| Bacteroidetes | 44 | 51.1 | 1,236 |
| Candidate division WWE1; Candidatus Cloacamonas | 1 | 26.3 | 13 |
| Chlamydiae | 2 | 67.3 | 16 |
| Chlorobi | 1 | 48.4 | 6 |
| Chloroflexi | 1 | 40.9 | 14 |
| Cyanobacteria | 40 | 67.5 | 12,837 |
| Deferribacteres | 4 | 53.1 | 23 |
| Deinococcus thermus | 2 | 35.7 | 208 |
| Elusimicrobia | 1 | 66.3 | 3 |
| Environmental samples | 10 | 90.9 | 1,099 |
| Firmicutes | 155 | 59.2 | 2,047 |
| Fusobacteria | 11 | 52.0 | 91 |
| Lentisphaerae | 1 | 39.4 | 18 |
| Planctomycetes | 4 | 44.1 | 18 |
| Proteo.Alphaproteobacteria | 22 | 53.3 | 199 |
| Proteo.Betaproteobacteria | 19 | 56.3 | 1,471 |
| Proteo.Deltaproteobacteria | 4 | 59.5 | 16 |
| Proteo.Epsilonproteobacteria | 24 | 55.0 | 322 |
| Proteo.Gammaproteobacteria | 85 | 59.8 | 1,770 |
| Spirochetes | 21 | 65.0 | 188 |
| Tenericutes | 38 | 69.8 | 15,312 |
| Thermotogae | 3 | 81.0 | 32 |
| Verrucomicrobia | 1 | 38.2 | 132 |
HTS is a highly comprehensive method based on random sequencing of nucleic acids present in a given sample and likely to ensure the detection of every possible sequence from microorganisms present in the target tissue, either previously known or not. This pilot study is, to our knowledge, a previously unreported attempt to use HTS in search for infectious agents and, more particularly, for viral sequences in CTCL. We have chosen to sequence all RNAs, without any selection for polyA+ RNAs, as a hallmark of the presence of all pathogens including bacteria. As recently described (
Cheval et al., 2011
), our pipeline is able to detect viruses present in databases with a level of sensitivity roughly equivalent to quantitative PCRs and to acquire full-length genomes (Bouquet et al., 2012
; Foulongne et al., 2012
). It can also identify unknown viruses (Sauvage et al., 2011a
, Sauvage et al., 2011b
) even when the viral species defines a new genus within a family. This powerful procedure is thus particularly suited to situations where involvement of an infectious agent is suspected but none has been detected/found. In the present study, HTS yielded no sequence corresponding to a known or unknown viral agent. More particularly, no sequence from the newly described polyomaviruses was identified, but most of them are probably of limited distribution except for MCPyV and a random effect cannot be ruled out. Nevertheless, methodological limitations may have reduced the import of these negative results. First, this study was conducted in a limited number of patients all originating from a particular geographical area (southern France) and no definitive conclusion can be drawn. Different results might be obtained in other areas by analogy, with observations made for MCPyV in MCC (Garneski et al., 2009
). Second, the selected patients were not representative of all subsets of CTCL as folliculotropic MF and CD30+ lymphoproliferative disorders were not represented. Despite these limitations, these data clearly contribute to a body of evidence that argues against infectious, more particularly viral agents being an etiologically relevant general characteristic in these patients, owing to the high investigational power of the molecular tool. It remains to determine whether or not this negative result is shared by all subsets of CTCL, especially in folliculotropic MF, as follicles are well-known sanctuaries for diverse viruses such as HPV.ACKNOWLEDGMENTS
The platform “Genotyping of Pathogens and Public Health” is supported in part by the Institut de Veille Sanitaire (Saint-Maurice, France). This study was partly supported by grants from both region Ile de France and the Programme Hospitalier de Recherche Clinique of the Montpellier University Hospital (AOI 2008, Protocole UF8425).
SUPPLEMENTARY MATERIAL
Supplementary material is linked to the online version of the paper at http://www.nature.com/jid
REFERENCES
- Mycosis fungoides and Sezary syndrome are not associated with HTLV-I infection: an international study.Br J Haematol. 1997; 98: 927-933
- Identical consensus sequence and conserved genomic polymorphism of hepatitis E virus during controlled interspecies transmission.J Virol. 2012; 86: 6238-6245
- Evaluation of high-throughput sequencing for identifying known and unknown viruses in biological samples.J Clin Microbiol. 2011; 249: 3268-3275
- Absence of HTLV-related sequences in skin lesions and peripheral blood of cutaneous T-cell lymphomas.J Invest Dermatol. 2009; 129: 2520-2522
- Clonal integration of a polyomavirus in human Merkel cell carcinoma.Science. 2008; 319: 1096-1100
- Human skin microbiota: high diversity of DNA viruses identified on the human skin by high throughput sequencing.PLoS One. 2012; 7: e38499
- Merkel cell polyomavirus is more frequently present in North American than Australian Merkel cell carcinoma tumors.J Invest Dermatol. 2009; 129: 246-248
- Deleted HTLV-I provirus in blood and cutaneous lesions of patients with mycosis fungoides.Science. 1991; 253: 317-320
- HTLV-V: a new human retrovirus isolated in a Tac-negative T cell lymphoma/leukemia.Science. 1987; 238: 1581-1583
- Mycosis fungoides in European Russia: no antibodies to human T cell leukemia virus type I structural proteins, but virus-like sequences in blood and saliva.Intervirology. 2005; 48: 362-371
- The cutaneous T cell lymphoma, mycosis fungoides, is a human T cell lymphotropic virus-associated disease. A study of 50 patients.J Clin Invest. 1995; 95: 547-554
- Human polyomavirus related to African green monkey lymphotropic polyomavirus.Emerg Infect Dis. 2011; 17: 1364-1370
- Identification of the first human gyrovirus, a virus related to chicken anemia virus.J Virol. 2011; 85: 7948-7950
- T antigen mutations are a human tumor-specific signature for Merkel cell polyomavirus.Proc Natl Acad Sci USA. 2008; 105: 16272-16277
- No evidence of HTLV-I proviral integration in lymphoproliferative disorders associated with cutaneous T-cell lymphoma.Am J Pathol. 1997; 150: 667-673
Article info
Footnotes
The authors state no conflict of interest.
Identification
Copyright
© 2013 The Society for Investigative Dermatology, Inc. Published by Elsevier Inc.
User license
Elsevier user license | How you can reuse 
Elsevier's open access license policy
Elsevier user license
Permitted
For non-commercial purposes:
- Read, print & download
- Text & data mine
- Translate the article
Not Permitted
- Reuse portions or extracts from the article in other works
- Redistribute or republish the final article
- Sell or re-use for commercial purposes
Elsevier's open access license policy
