article

miRNA and viral infections in vertebrates

Posted: 7 February 2009 | Pål Sætrom, Post Doctoral Fellow, Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology and Jörn Klein, Post Doctoral Fellow, Department of Computer and Information Science, Norwegian University of Science and Technology | No comments yet

For plants and invertebrates, RNA interference is firmly established as an important antiviral mechanism. Even before Fire, Mello, and co-workers described RNA interference (RNAi) in worms in 19981 it was becoming clear that plants have an RNA-dependent pathway that protects against viral infections2. The pathway, then termed post-transcriptional gene silencing (PTGS), helps plants like tobacco recover from initial viral infections and ensures that plants are protected from subsequent infections from the same or similar viral strains3. Subsequent studies have revealed that plant PTGS and Fire and Mello’s RNAi are identical – the triggers are short RNAs derived from long double-stranded RNAs (dsRNA)4. Incorporated into the RNA-induced silencing complex (RISC), RISC cleaves transcripts like viral messenger RNAs (mRNAs) with antisense complementary to the short RNAs.

For plants and invertebrates, RNA interference is firmly established as an important antiviral mechanism. Even before Fire, Mello, and co-workers described RNA interference (RNAi) in worms in 19981 it was becoming clear that plants have an RNA-dependent pathway that protects against viral infections2. The pathway, then termed post-transcriptional gene silencing (PTGS), helps plants like tobacco recover from initial viral infections and ensures that plants are protected from subsequent infections from the same or similar viral strains3. Subsequent studies have revealed that plant PTGS and Fire and Mello’s RNAi are identical – the triggers are short RNAs derived from long double-stranded RNAs (dsRNA)4. Incorporated into the RNA-induced silencing complex (RISC), RISC cleaves transcripts like viral messenger RNAs (mRNAs) with antisense complementary to the short RNAs.

For plants and invertebrates, RNA interference is firmly established as an important antiviral mechanism. Even before Fire, Mello, and co-workers described RNA interference (RNAi) in worms in 19981 it was becoming clear that plants have an RNA-dependent pathway that protects against viral infections2. The pathway, then termed post-transcriptional gene silencing (PTGS), helps plants like tobacco recover from initial viral infections and ensures that plants are protected from subsequent infections from the same or similar viral strains3. Subsequent studies have revealed that plant PTGS and Fire and Mello’s RNAi are identical – the triggers are short RNAs derived from long double-stranded RNAs (dsRNA)4. Incorporated into the RNA-induced silencing complex (RISC), RISC cleaves transcripts like viral messenger RNAs (mRNAs) with antisense complementary to the short RNAs.

In vertebrates, long dsRNA triggers the protein kinase R (PKR) and interferon responses which give general RNA decay and shutdown of protein synthesis. These responses are potent antiviral defences on their own and it is therefore unclear whether RNAi has the same antiviral role in vertebrates as in plants and invertebrates5. What is becoming clear, however, is that a different aspect of the RNAi pathway is important in many vertebrate viral infections.

Animals and plants encode small RNA genes that regulate protein coding genes post transcription. The RNAs, called microRNAs (miRNAs), work through the same protein complexes as dsRNA-derived small interfering RNAs (siRNAs) do. The main difference from siRNAs is that miRNAs are derived from imperfect hairpin structures in single-stranded RNAs. Both siRNAs and miRNAs can regulate targets by either binding perfectly complementary sites anywhere in mRNA or partially complementary sites preferentially located in the 3’ untranslated region (UTR) of mRNA. Small interfering RNAs primarily act through the perfect Watson-Crick sites, however, whereas at least animal miRNAs primarily act through the imperfect sites. The former sites give mRNA cleavage whereas the latter sites give translational suppression and mRNA degradation.

MicroRNAs are essential developmental regulators and their roles in human diseases such as cancer are rapidly being uncovered6. Being general regulators that target the majority of mammalian protein coding genes7, vertebrate miRNAs may also play a role in regulating viral transcripts. Vertebrate viruses do however also encode their own miRNAs that regulate both the virus’ and the host’s gene expression. Here, we will review animal miRNA processing, regulation, and targeting before describing the newest findings of how miRNAs influence the virus-host relationship.

MicroRNA processing

The canonical animal miRNA processing pathway proceeds through three distinct steps (reviewed in reference no.8). First, the nuclear Microprocessor complex recognises and excises miRNA hairpins from longer primary RNA transcripts. Second, the nuclear transport receptor Exporting-5 exports the approximately 60 nucleotide (nt) long hairpin products, called miRNA precursors (pre-miRNAs), to the cytoplasm. Third, the ribonuclease (RNase) III Dicer processes the pre-miRNAs into an approximately 22 nt long duplex form by removing the hairpin loop before handing the duplex to the RISC complex. There, one strand – the mature miRNA – is preferentially incorporated into the catalytic Argonaute 2 (Ago2) protein.

Importantly, a correct substrate can, in principle, enter the miRNA pathway at any point. Researchers have used this property to for example create transgenic cells that stably express artificial siRNAs – either as short hairpin RNAs (Dicer substrates that rely on nuclear export) or as miRNA mimics (true Microprocessor substrates)9. Note this is also really the property that allows researchers to use artificial siRNAs, as the siRNAs enter the pathway as artificial Dicer products. By the same token, certain classes of miRNAs can skip parts of the processing pathway. One such example is the mirtrons, which are short introns that fold into small pre-miRNA hairpins and thereby bypass Microprocessor processing10. Another example is small nucleolar RNA (snoRNA)-derived miRNAs (sno-miRNAs), which depend on Dicer but may also require other unknown factors for processing11. As the sno-miRNAs identified by the Meister and Rajewsky labs are atypical Exportin-5 and Dicer substrates – both proteins prefer RNA-duplexes with 3’ overhangs12 – one could speculate that any hairpin RNA that enters the cytoplasm may give a functional miRNA.

MicroRNA regulation

MicroRNAs show wide ranges of tissue – and developmental – specific expression patterns13. This can partly be explained by miRNAs sharing the transcriptional complexity of protein coding genes, as RNA polymerase II seems to transcribe most primary miRNA transcripts14. Still, many questions remain about miRNA transcription; for example, why are many miRNAs that are located within introns of known genes apparently independently transcribed15?

The other major regulator mechanism is post-transcriptional regulation of miRNAs’ processing and targeting. RNA editing is one mechanism that can influence miRNA processing and, by changing the mature miRNA sequence, miRNA targeting16. Modulating the levels of the individual proteins in the miRNA processing pathway will, in general, also affect miRNA expression levels. Exportin 5, in particular, may be a key factor, as nuclear export seems to be a rate limiting step in the pathway17. An even more direct and targeted mechanism for regulating miRNA processing, however, was discovered by studying the let-7 miRNA family18,19. Both undifferentiated and differentiated human embryonic stem cells express primary let-7a transcripts, but only differentiated cells express the mature let-7a miRNA. The reason is that the protein lin28 recognises the let-7 pre-miRNA’s distinct loop structure and disrupts its processing into a mature miRNA, possibly by mediating 3’ uridylation of the pre-miRNA, thereby preventing Dicer processing and promoting pre-miRNA degradation18.

Differentially expressed proteins that recognise subtle structural differences between miRNA precursors may indeed be a general mechanism for regulating miRNAs. Ago2, for example, is important in processing certain miRNA precursors, as the enzyme cleaves the passenger strand in the miRNA hairpin prior to Dicer processing20. Changes in Ago2 expression levels should therefore affect the expression levels of such Ago2-dependent miRNAs compared with canonical miRNAs.

MicroRNA targeting

In animal miRNA targeting (reviewed in reference no.21), the so-called seed sequence (or seed) is an essential concept. Ago2 uses the 5’ nucleotides in the mature miRNA as a nucleation signal for recognizing target RNAs and the most important signal for target recognition is the seed at nucleotides 2-8. Perfect Watson-Crick pairing between the miRNA seed and mRNA is sufficient for translational suppression, but additional pairing between the miRNA 3’ end and mRNA can give stronger regulation or compensate for imperfect seed pairing22. The sequence context of the target site is also important. An obvious example of this property is that functional target sites are preferentially located in the 3’ UTR, but the actual location within the 3’ UTR is also important23. Finally, multiple target sites can give synergistic regulation depending on the distance between consecutive sites24.

The exact mechanism underlying the synergism remains unknown. As translational suppression is a relatively slow process compared with Ago2’s catalytic cleavage25, however, multiple RISC complexes bound at closely spaced target sites may cooperatively stabilise each other at the sites or possibly accelerate the regulatory process. This could explain why miRNAs prefer targets in 3’ UTRs, as ribosomes would displace RISC from sites in the coding sequence (CDS) before RISC could effect translational suppression. At the same time, one would then also expect that miRNAs that successfully regulate mRNAs through sites in CDS have multiple, closely spaced sites in the target CDS. This prediction seems to hold for the few genes that currently have verified miRNA target sites in CDS26.

It is also worth mentioning that at least a few miRNAs have a role in transcriptional regulation27. Although much of the exact mechanism remains unclear, the miRNAs seem to regulate transcription by targeting transcripts that are sense or antisense to the regulated gene. Destruction of these presumably non-coding transcripts through miRNA targeting has opposite effects on epigenetic regulatory patterns – sense targets give transcriptional repression whereas antisense targets give transcriptional activation28. Given the abundance of natural antisense transcripts29, it is likely that other miRNAs use this regulatory mechanism as well. If so, it would be surprising if no viruses have evolved to use the mechanism.

MicroRNAs and virulence

Removing essential components of the RNAi pathway can give an indication of miRNAs’ general importance in viral infections. Otsuka and colleagues produced Dicer-deficient mice and infected them with the rhabdovirus vesicular stomatitis virus (VSV), resulting in a 5- to 10-fold increase in viral titre in peritoneal macrophages30. Similarly, HIV-1 replicates faster in cells where Dicer and the Microprocessor protein Drosha have been depleted by siRNAs31. In HIV-1’s case, the enhanced replication can at least partly be explained by host miRNAs down-regulating histone acetylase PCAF, which is an important co-factor for HIV-1 gene expression.

MicroRNAs and viral latency

Viral latency, often the cause of chronic infections, is due to the persistence of silent viral DNA in the host cell. As the major function of cellular miRNAs is to inhibit protein translation, the involvement of miRNAs in viral latency is expected. Especially the role of miRNAs in the regulation of human immunodeficiency virus (HIV) latency and persistence has attracted interest in the recent years. This is because the current highly potent anti-HIV therapies work during active viral replication, but are ineffective in the state of viral latency33. Huang et al.34 have shown that the 3′ end of HIV messenger RNA are targeted by a cluster of cellular miRNAs, which are over expressed in resting CD4+ T cells as compared to activated CD4+ T cells. Introducing antimiRs against the miRNAs over expressed in the resting T-lymphocytes lead to the release of viral particles, indicating the role of the miRNAs in latency and the possibility of an antimiR-based therapeutic approach.

In addition to using host miRNAs to induce latency, viruses encode miRNAs that may contribute to latency, as shown for the HIV-1 trans-activation response (TAR) element. The TAR hairpin structure is found at the 5′ end of the viral RNA and is recognised and processed by the cellular miRNA pathway into a mature miRNA35. Although the potential role of the TAR-derived miRNA is still unclear, other viruses, such as the Epstein-Barr virus (EBV), Hepatitis B virus (HBV), Human herpes virus 8 (HHV8), and Marek’s disease virus (MDV), encode miRNAs with demonstrated roles in virus latency36-40.

MicroRNAs and viral oncogenesis

That certain viruses are oncogenic is well established, but recent reports indicate that miRNAs have a role in virus-induced oncogenesis. Especially for EBV, which is associated with several lymphoproliferative disorders such as Hodgkin’s disease and Burkitt’s lymphoma, has the important role of miRNAs been demonstrated38,39,41. Particularly, the upregulation of the cellular miRNA miR-155 in EBV infected cells36,39,42,43 seems to play a role in oncogenesis. By direct targeting and by down-regulating multiple transcriptional regulatory genes, miR-155 alters gene expression and may thereby down-regulate important tumour suppressors43. Also for other members of the Herpesviridae, like Human herpesvirus 8 (HHV-8)44 or Marek’s Disease virus40,45, for Human papilloma virus46, and for some Retroviridae47 do miRNAs contribute to tumour induction.

MicroRNAs and tissue tropism

Viruses vary greatly not only in their host range, but also in their affinity to various tissues within a host (tissue tropism). As the expression patterns of cellular miRNAs also differ between cell types, miRNAs may contribute to the tissue tropism of viruses. One example is the liver-specific miRNA miR-122, which interacts with a target sequence in the 5′ non-coding region of the hepatitis C virus (HCV) genome and stimulates virus replication48.

Recently, Kelly and co-workers used miRNA-directed tissue tropism for tissue-specific destabilisation of viral genomes in a therapeutic setting49. More specifically, they inserted target sequences complementary to muscle-specific miRNAs into the 3′ untranslated region (UTR) of a coxsackie virus A21 (CVA21), a potently oncolytic picornavirus. Oncolytic viruses can infect and lyse cancer cells, while leaving normal cells unharmed. The resulting recombinant virus propagated in subcutaneous tumors, causing total regression and sustained viremia, but could not replicate in cells expressing complementary miRNAs. This miRNA-guided tissue tropism provides a unique method for generating new vaccines and improving the safety of existing vaccines.

Concluding remarks

As outlined above, both host- and virus-encoded miRNAs closely interact in regulating multiple aspects of vertebrate viral infections. One should therefore expect signs of co-evolution between the miRNAs and their target sequences. Most animal miRNAs are evolutionary conserved50, but miRNAs are also frequently gained and lost51, which may be an indication of co-evolution with viruses. In contrast, viral miRNAs are as variable as the viral genome itself, due to genetic drift, error prone replication, and bottleneck events. Evolutionarily conserved viral microRNA can be found, however52.

An intriguing hypothesis for the role of miRNAs in HIV pathogenesis and immune evasion has been proposed by Linda Ludwig53. She describes a model of intrinsic RNA silencing by the ‘HIV-1 antisense initator promoter element’ (HIVaINR), a genome region of the virus forming a duplex with the U3R sense viral RNA54. The element seems to use the RNAi pathway to target the immune relevant interleukin- 2 gamma chain55.

In this context it appears likely that there are RNA viruses which have evolved to manipulate and use the cellular miRNA pathway in a Microprocessor independent way. So far it has been hypothesised that only DNA viruses (with the exception of Poxviridae) or Retroviruses encode viral miRNAs, as viral transcripts must originate from or enter the nucleus for Microprocessor processing. Animals have at least two types of Microprocessor-independent miRNAs, however. We therefore expect that at least some RNA viruses have bypassed the Microprocessor-dependence and encode for their own viral miRNAs. We also expect that the increasing understanding of miRNAs’ roles in viral infections will give new therapeutic approaches to fight viral disease.

References

  1. A. Fire, S. Xu, M. K. Montgomery et al., Nature 391 (6669), 806 (1998).
  2. F. Ratcliff, B. D. Harrison, and D. C. Baulcombe, Science (New York, N.Y 276 (5318), 1558 (1997).
  3. F. G. Ratcliff, S. A. MacFarlane, and D. C. Baulcombe, Plant Cell 11 (7), 1207 (1999).
  4. A. J. Hamilton and D. C. Baulcombe, Science (New York, N.Y 286 (5441), 950 (1999); P. D. Zamore, T. Tuschl, P. A. Sharp et al., Cell 101 (1), 25 (2000).
  5. B. R. Cullen, Nat Immunol 7 (6), 563 (2006).
  6. H. S. Soifer, J. J. Rossi, and P. Saetrom, Mol Ther 15 (12), 2070 (2007).
  7. R. C. Friedman, K. K. Farh, C. B. Burge et al., Genome research (2008).
  8. V. N. Kim, Nature reviews 6 (5), 376 (2005).
  9. P. J. Paddison, A. A. Caudy, R. Sachidanandam et al., Methods in molecular biology (Clifton, N.J 265, 85 (2004); Y. Zeng, E. J. Wagner, and B. R. Cullen, Molecular cell 9 (6), 1327 (2002).
  10. E. Berezikov, W. J. Chung, J. Willis et al., Molecular cell 28 (2), 328 (2007); K. Okamura, J. W. Hagen, H. Duan et al., Cell 130 (1), 89 (2007); J. G. Ruby, C. H. Jan, and D. P. Bartel, Nature 448 (7149), 83 (2007).
  11. C. Ender, A. Krek, M. R. Friedlander et al., Molecular cell 32 (4), 519 (2008).
  12. A. Vermeulen, L. Behlen, A. Reynolds et al., RNA (New York, N.Y 11 (5), 674 (2005); Y. Zeng and B. R. Cullen, Nucleic acids research 32 (16), 4776 (2004).
  13. P. Landgraf, M. Rusu, R. Sheridan et al., Cell 129 (7), 1401 (2007).
  14. A. Marson, S. S. Levine, M. F. Cole et al., Cell 134 (3), 521 (2008).
  15. F. Ozsolak, L. L. Poling, Z. Wang et al., Genes & development 22 (22), 3172 (2008).
  16. Y. Kawahara, B. Zinshteyn, T. P. Chendrimada et al., EMBO reports 8 (8), 763 (2007); Y. Kawahara, B. Zinshteyn, P. Sethupathy et al., Science (New York, N.Y 315 (5815), 1137 (2007); W. Yang, T. P. Chendrimada, Q. Wang et al., Nature structural & molecular biology 13 (1), 13 (2006).
  17. D. Grimm, K. L. Streetz, C. L. Jopling et al., Nature 441 (7092), 537 (2006).
  18. I. Heo, C. Joo, J. Cho et al., Molecular cell 32 (2), 276 (2008).
  19. S. R. Viswanathan, G. Q. Daley, and R. I. Gregory, Science (New York, N.Y 320 (5872), 97 (2008); M. A. Newman, J. M. Thomson, and S. M. Hammond, RNA (New York, N.Y 14 (8), 1539 (2008); A. Rybak, H. Fuchs, L. Smirnova et al., Nature cell biology 10 (8), 987 (2008).
  20. S. Diederichs and D. A. Haber, Cell 131 (6), 1097 (2007).
  21. W. Filipowicz, S. N. Bhattacharyya, and N. Sonenberg, Nat Rev Genet 9 (2), 102 (2008).
  22. J. Brennecke, A. Stark, R. B. Russell et al., PLoS biology 3 (3), e85 (2005).
  23. A. Grimson, K. K. Farh, W. K. Johnston et al., Molecular cell 27 (1), 91 (2007).
  24. P. Saetrom, B. S. Heale, O. Snove, Jr. et al., Nucleic acids research 35 (7), 2333 (2007).
  25. B. Haley and P. D. Zamore, Nature structural & molecular biology 11 (7), 599 (2004).
  26. Y. Tay, J. Zhang, A. M. Thomson et al., Nature 455 (7216), 1124 (2008); J. J. Forman, A. Legesse-Miller, and H. A. Coller, Proceedings of the National Academy of Sciences of the United States of America 105 (39), 14879 (2008).
  27. D. H. Kim, P. Saetrom, O. Snove, Jr. et al., Proceedings of the National Academy of Sciences of the United States of America 105 (42), 16230 (2008); R. F. Place, L. C. Li, D. Pookot et al., Proceedings of the National Academy of Sciences of the United States of America 105 (5), 1608 (2008).
  28. K. V. Morris, S. Santoso, A. M. Turner et al., PLoS genetics 4 (11), e1000258 (2008).
  29. A. Werner, RNA biology 2 (2), 53 (2005).
  30. M. Otsuka, Q. Jing, P. Georgel et al., Immunity 27 (1), 123 (2007).
  31. R. Triboulet, B. Mari, Y. L. Lin et al., Science (New York, N.Y 315 (5818), 1579 (2007).
  32. M. N. Krishnan, A. Ng, B. Sukumaran et al., Nature 455 (7210), 242 (2008).
  33. E. M. Tedaldi, J Addict Dis 27 (2), 83 (2008); L. Shen and R. F. Siliciano, J Allergy Clin Immunol 122 (1), 22 (2008); M. D. Marsden and J. A. Zack, J Antimicrob Chemother (2008).
  34. J. Huang, F. Wang, E. Argyris et al., Nat Med 13 (10), 1241 (2007).
  35. J. Son, P. D. Uchil, Y. B. Kim et al., Biochem Biophys Res Commun 374 (2), 214 (2008); Y. P. Liu, J. Haasnoot, O. ter Brake et al., Nucleic acids research 36 (9), 2811 (2008); Z. Klase, P. Kale, R. Winograd et al., BMC Mol Biol 8, 63 (2007).
  36. F. Lu, A. Weidmer, C. G. Liu et al., J Virol 82 (21), 10436 (2008).
  37. R. H. Edwards, A. R. Marquitz, and N. Raab-Traub, J Virol 82 (18), 9094 (2008); P. J. Fannin Rider, W. Dunn, E. Yang et al., Curr Top Microbiol Immunol 325, 21 (2008); W. B. Jin, F. L. Wu, D. Kong et al., Comput Biol Chem 31 (2), 124 (2007).
  38. S. Barth, T. Pfuhl, A. Mamiani et al., Nucleic acids research 36 (2), 666 (2008); J. E. Cameron, Q. Yin, C. Fewell et al., J Virol 82 (4), 1946 (2008); X. Cai, A. Schafer, S. Lu et al., PLoS Pathog 2 (3), e23 (2006).
  39. N. Motsch, T. Pfuhl, J. Mrazek et al., RNA biology 4 (3), 131 (2007).
  40. J. Burnside, E. Bernberg, A. Anderson et al., J Virol 80 (17), 8778 (2006).
  41. T. Xia, A. O’Hara, I. Araujo et al., Cancer Res 68 (5), 1436 (2008); A. Navarro, A. Gaya, A. Martinez et al., Blood 111 (5), 2825 (2008); L. Xing and E. Kieff, J Virol 81 (18), 9967 (2007); N. Kim do, H. S. Chae, S. T. Oh et al., J Virol 81 (2), 1033 (2007).
  42. J. E. Cameron, C. Fewell, Q. Yin et al., Virology 382 (2), 257 (2008); G. Gatto, A. Rossi, D. Rossi et al., Nucleic acids research 36 (20), 6608 (2008); J. Kluiver, E. Haralambieva, D. de Jong et al., Genes Chromosomes Cancer 45 (2), 147 (2006); N. Rahadiani, T. Takakuwa, K. Tresnasari et al., Biochem Biophys Res Commun 377 (2), 579 (2008).
  43. Q. Yin, J. McBride, C. Fewell et al., J Virol 82 (11), 5295 (2008).
  44. E. Murphy, J. Vanicek, H. Robins et al., Proceedings of the National Academy of Sciences of the United States of America 105 (14), 5453 (2008); L. V. McClure and C. S. Sullivan, Cell Host Microbe 3 (1), 1 (2008); R. L. Skalsky, M. A. Samols, K. B. Plaisance et al., J Virol 81 (23), 12836 (2007); M. A. Samols, R. L. Skalsky, A. M. Maldonado et al., PLoS Pathog 3 (5), e65 (2007).
  45. M. Chen, W. S. Payne, H. Hunt et al., Virology 377 (2), 265 (2008); J. Burnside, M. Ouyang, A. Anderson et al., BMC Genomics 9, 185 (2008); H. Xu, Y. Yao, Y. Zhao et al., J Virol Methods 149 (2), 201 (2008); Y. Yao, Y. Zhao, H. Xu et al., J Virol 82 (8), 4007 (2008); J. Burnside and R. W. Morgan, Cytogenet Genome Res 117 (1-4), 376 (2007); Y. Yao, Y. Zhao, H. Xu et al., J Virol 81 (13), 7164 (2007).
  46. . Schwartz, Front Biosci 13, 5880 (2008); I. Martinez, A. S. Gardiner, K. F. Board et al., Oncogene 27 (18), 2575 (2008); X. Cai, G. Li, L. A. Laimins et al., J Virol 80 (21), 10890 (2006).
  47. G. B. Beck-Engeser, A. M. Lum, K. Huppi et al., Retrovirology 5, 4 (2008); S. Kim, N. Kim, B. Dong et al., J Virol 82 (20), 9964 (2008).
  48. C. L. Jopling, S. Schutz, and P. Sarnow, Cell Host Microbe 4 (1), 77 (2008).
  49. E. J. Kelly, E. M. Hadac, S. Greiner et al., Nat Med 14 (11), 1278 (2008).
  50. A. Grimson, M. Srivastava, B. Fahey et al., Nature 455 (7217), 1193 (2008); M. Marz, T. Kirsten, and P. F. Stadler, J Mol Evol (2008); C. Mascaux, J. F. Laes, G. Anthoine et al., Eur Respir J (2008); U. Technau, Nature 455 (7217), 1184 (2008).
  51. J. Lu, Y. Shen, Q. Wu et al., Nature genetics 40 (3), 351 (2008).
  52. G. J. Seo, L. H. Fink, B. O’Hara et al., J Virol 82 (20), 9823 (2008).
  53. L. B. Ludwig, Retrovirology 5, 79 (2008).
  54. L. B. Ludwig, J. L. Ambrus, Jr., K. A. Krawczyk et al., Retrovirology 3, 80 (2006).
  55. N. L. Alves, F. A. Arosa, and R. A. van Lier, Immunol Lett 108 (2), 113 (2007).