Progress and prospects: Foamy virus vectors enter a new age
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Gene Therapy (2010) 17, 1423–1429
& 2010 Macmillan Publishers Limited All rights reserved 0969-7128/10
www.nature.com/gt
REVIEW
Progress and prospects: Foamy virus vectors enter
a new age
O Erlwein and MO McClure
Section of Infectious Diseases, Jefferiss Research Trust Laboratories, Imperial College London, London, UK
Foamy viruses, distantly related to the major subfamily efficient transduction of progenitor cells and an integration
of Retroviruses, Orthoretroviruses that include oncoviruses profile less likely to induce insertional mutagenesis, make
(for example, murine leukemia virus (MLV)) and lentiviruses these viruses attractive as vectors. Long-term reversal of
(human immunodeficiency virus (HIV)), are endemic in disease phenotype in dogs with the genetic defect, leukocyte
mammalian species, but not in human populations. Humans adhesion deficiency, by foamy virus vector therapy strength-
infected by accidental or occupational exposure remain well. ens the case for their clinical exploitation.
The virus is not transmitted to others, nor is it associated with Gene Therapy (2010) 17, 1423–1429; doi:10.1038/gt.2010.95;
any disease. These features added to its broad host range, published online 15 July 2010
Keywords: foamy virus; spumaretrovirus; retroviral vector
In brief
Progress Prospects
Novel approaches to areas that have hitherto proved
Foamy virus infection in animals and humans does to be technically challenging, such as:
not cause disease. producing packaging cell lines.
Foamy virus replication is different from that of other defining the cellular receptor.
retroviruses. increasing vector titre further.
Foamy virus integrates into the host cell in typical Reversal of a single gene defect in a canine model
retrovirus manner. may lead to clinical trial of a similar clinical condition
Foamy viruses have untapped potential as safe in humans.
vectors. Reversal of further single-gene defects.
Genes are delivered by foamy virus vectors in vivo.
Foamy virus vectors cure dogs of a genetic disease.
respective foamy virus and in captivity as many as 100%
Foamy virus infection in animals and can test positive. When Hahn’s group looked for
humans does not cause disease evidence of foamy virus infection in wild equatorial
African chimpanzees by testing for antibody and nucleic
Foamy viruses are the only members of the Spumare-
acid in fecal samples, they confirmed a prevalence of
troviruses, one of two families of the Retroviridae, the
between 44 and 100%.2 Morozov et al.3 found that 12
other and much larger family being the Orthoretro-
of 14 free living chimpanzees in Tai National Park
viruses (ORV). From those, the viral vectors that have
(Côte d’Ivoir) were Semliki forest virus (SFV) infected.
been most exploited in trials are murine leukemia virus
The sensitive assays used by Jones-Engel et al.4 found,
(MLV) (a gammaretrovirus) and human immunodefi-
not only that more than 92% of wild non-human
ciency virus (HIV-1) and EIAV (both lentiviruses). The
primates were SFV positive in Thailand but that some
wild-type viruses from which these vectors are derived
had already been seroconverted by the age of 3. Engel
all cause disease, whereas foamy viruses do not. In the
et al.5 reported 88% of Gibraltar’s macaques to be SFV
infected host, active foamy virus replication seems to be
positive. Thus, foamy virus infection is widespread in
restricted to the oral mucosa,1 suggesting that transmis-
non-human primates.
sion in animals occurs by bites and other wounds.
Superinfection of animals with different foamy virus
In the wild, monkeys are commonly infected with their
strains is possible, as shown by inter-species transmis-
sion of SFV in chimpanzees that prey on colobus
Correspondence: Professor MO McClure, Section of Infectious monkeys. In addition to their own strain, the chimpan-
Diseases, Jefferiss Research Trust Laboratories, Imperial College
London, St Mary’s Campus, Norfolk Place, London W2 1PG, UK.
zees were also infected with SFV specific for colobus
E-mail: m.mcclure@imperial.ac.uk monkeys.6 Despite having been described in a variety of
Received 16 May 2009; revised 28 April 2010; accepted 28 April mammalian species, including non-human primates,
2010; published online 15 July 2010 cats, cows and horses, sea lions and hamsters, in whichFoamy virus vectors enter a new age
O Erlwein and MO McClure
1424
they cause life-long persistent infections and, despite before budding from the infected cell. Thus, as about 20%
intensive searching, no natural reservoir for foamy virus of virions contain infectious DNA, foamy viruses can be
in humans has ever been documented. Accidental zoonotic viewed as DNA viruses that replicate by an RNA
infections, however, are not unknown for people in close intermediate, rather than hepadnaviruses. The full length
contact with non-human primates, such as animal of the single-stranded DNA genome is 13 Kb, longer than
handlers and bush meat traders.7 Investigating prototype that of MLV (8 kb) or HIV-1 (9 Kb) and, similar to other
foamy virus (PFV) infection in Asians, Jones-Engels et al.8 complex retroviruses, carries additional open reading
identified 2.6% positivity (8 out of 305). The virus persists frames in the 30 region of the genome, giving rise to
in peripheral blood lymphocytes for over 20 years, accessory proteins, in this case, Tas (the transcriptional
suggesting a capability for long-term gene expression in transactivator essential for replication) and, by virtue of
the case of PFV-derived vectors. Infected humans do not a splicing event, Bet, an accessory protein abundantly
transmit the virus horizontally. However, when blood of produced in the infected cell. Unusually, both are
an SFV-infected monkey was transfused into an uninfected transcribed from an internal promoter, located in the
animal, molecular evidence of infection became obvious env open reading frame, rather than from the promoter
after 8 weeks, with seroconversion 1 week later, confirm- in the LTR. The Bet protein, disposable for replication
ing that foamy viruses can be transmitted through blood in vitro, is nonetheless expressed in abundance in the
products.9 In marked contrast to the cytopathic effect infected cell. Its exact function is not clear, although there
induced in vitro (it is the infection-induced vacuolation is a suggestion that it functions in a manner similar to the
that gives cells their ‘foamy’ appearance), no pathology HIV-1 Vif protein in antagonizing cellular APOBEC3
has been assigned to foamy virus infection in any host. proteins.10 It is interesting that Trim-5-alpha can also
Early papers attempting to link infection with a variety of restrict foamy viruses by sequences in the N-terminal
disease states have not withstood the test of time. half of Gag.11
Although it is true to say that the effect of foamy virus As for the ORVs, the LTRs contain cis-acting regulatory
infection in humans under conditions of immunosuppres- elements for viral protein expression. The internal
sion is not known, it is known that primates and cats promoter in env has some basal activity, resulting in the
dually infected with immunodeficiency viruses and foamy expression of the accessory proteins, Tas and Bet. Tas, in
viruses have no different a profile from those that are turn, is a DNA-binding transcriptional activator that
foamy virus free. There is no evidence from many studies enhances gene expression from the internal promoter but
carried out over the last decade in both animals and also from the otherwise silent promoter in the U3 region,
humans that foamy virus infection causes any clinical leading to the expression of the structural proteins. This
condition or deleterious effects. way, PFV can control gene expression in a temporal
manner, something ORVs, such as HIV-1, achieve
through alternative splicing. To express their structural
Foamy virus replication is different from proteins, ORVs first generate Gag–Pol fusion proteins
that are subsequently cleaved into Gag and Pol that can
that of other retroviruses be further processed. The Gag–Pol fusion proteins also
Foamy viruses have a typical retrovirus morphology provide a means by which the ORVs encapsidate their
with the distinguishing feature of extra long (15 nm) Pol protein into virions, as Gag self-assembles. For foamy
spikes by which they gain entry to a wide variety of cell viruses, Pol is expressed independently from Gag from
types in vitro. Indeed, it is challenging to find a cell line its own spliced mRNA. A fraction of the PFV Gag
refractory to PFV infection, something that has consis- proteins is processed by cleaving a 3-kDa peptide from
tently hampered identification of their cellular receptor. the C-terminus, resulting in a double band of about
However, it is known that viruses enter the cell at the low 71/68 kDa in size. This cleavage is essential for reverse
pH of the late endosome and travel along the micro- transcription. However, no analogous proteins to matrix,
tubules towards the microtubule-organizing center. The nucleocapsid or capsid is found in mature PFV virions.
first molecularly cloned isolate originated from a cell As the foamy virus Pol is expressed independently from
culture of a nasopharyngeal carcinoma patient and was Gag and no Gag–Pol fusion protein is made, this poses
designated, the human spumaretrovirus or human the problem of incorporating Pol into the virion. It is
foamy virus. Later it became clear that this was a thought that regions in cis-acting sequences present on
chimpanzee isolate, and human foamy virus is now viral RNA that are involved in RNA encapsidation can
considered to have resulted from cross-species transmis- also bind to Pol. Recently, however, it was shown that
sion. It remains the most studied of the foamy viruses Gag, too, contains C-terminal determinants that facilitate
and has now been designated the PFV. encapsidation of Pol into the virions.12 This independent
Foamy virus genomes share overall organizational expression of Pol from a spliced mRNA may be an
similarity with all complex retroviruses; the open read- advantage for vector development, as a vector produced
ing frames encode three standard retroviral structural from three separate plasmids encoding each of the
proteins for the core, enzymes of replication and virus structural genes is less likely to undergo recombination.
envelope (Gag, Pol and Env, respectively), as well as Foamy virus capsids are restricted to the cytoplasm in
accessory viral proteins, all of which are flanked by long the absence of its cognate envelope protein and, thus,
terminal repeat (LTR) sequences (Figure 1). The 50 end particles are not released into the supernatant in the
dimerization signal suggests that the pre-genomic single- absence of it or in the provision of a different envelope.
stranded diploid RNA is packaged into the virion, but This points to a specific interaction between Gag and its
whereas the RNA is characteristically reverse-transcribed cognate Env, making it difficult to pseudotype foamy
into DNA, before it is integrated into the host genome, virus capsids with Env proteins of other viruses without
the process can also occur late in the viral life cycle, just loss of infectivity. The foamy virus replication cycle
Gene TherapyFoamy virus vectors enter a new age
O Erlwein and MO McClure
1425
Promoter IP
pol orf-1
U3 R U5 gag env orf-2 U3 R U5
DLS
===
CASI CASII
tas
bet
Figure 1 The PFV genome. Genomic organization of PFV with highlighted characteristics relevant to vector generation. Two sequences
necessary for vector transduction are indicated, the cis-acting sequences, CasI and CasII. These sequences are important for encapsidation,
they carry the dimer linkage structure (DLS), and they also interact with the Pol protein. In addition to the promoter in the LTR, there is an
internal promoter (IP) in the Env ORF, which drives the expression of the accessory proteins Tas and Bet.
Budding and
Cell to cell transfer release
Golgi
Late RT
apparatus
Microtubules ER
Receptor
Early Assembly
RT
Uncoating
MTOC
Integration
PIC Transcription
Nucleus
Figure 2 PFV replication cycle. PFV attaches to an unknown cellular receptor, thereafter the viral core travels along microtubules towards
the microtubule-organizing centre and early reverse transcription occurs. The PFV protease cleaves the Gag protein and triggers disassembly
of the core at the microtubule-organizing center. Following integration into the host genome, viral mRNAs and proteins are produced and the
PFV virions assemble in the cytoplasm. An endoplasmic reticulum (ER) retrieval signal targets Env to the ER and without its cognate Env, no
PFV budding occurs. Late reverse transcription can take place before viral budding, resulting in about 20% of virions containing infectious
DNA. (Figure kindly provided by Dr Gillian Wills)
(shown in Figure 2) bears resemblance to that of the over recent years and genome-wide analysis of integra-
hepadnaviruses.13 tion has shown that insertion into the host genome is not
as random an event as was once thought, but influenced
by target DNA sequences. Lentiviruses, such as HIV-1,
Foamy virus integrates into the host cell in typical strongly favour integration into active transcription
retrovirus manner units. However, the fact that MLV targets transcription
Integration of the proviral DNA into the host genome is start sites and CpG islands, in part explains the
an essential part of the life cycle of every retrovirus and leukaemia that developed subsequent to the gene
foamy viruses, despite being uniquely different in many therapy trial for severe combined immunodeficiency-
respects, are no exception to this. The integration process X1, when the MLV vector integrated near the proto-
for retroviruses has been the subject of several studies oncogenes, LMO2 and MDS/Evi1.14 For foamy viruses, a
Gene TherapyFoamy virus vectors enter a new age
O Erlwein and MO McClure
1426
few reports describe the systematic investigation of be silenced by host factors, switching off continuous
foamy virus integration sites in cellular DNA. In one expression. This, however, has not occurred in animal
study, a total of 2829 different insertion events in human models transplanted for over two years with PFV
CD34+ cells and unselected human fibroblasts were vectors.24
investigated. Foamy virus integrations could be found on
all human chromosomes with clusters and gaps, but
overall did not show a preference for integration within Foamy viruses have untapped potential as
genes, unlike MLV or HIV. There was one hotspot (that is,
four or more integrants within a 50 kb region) 470 kb
safe vectors
from the Evi1 locus, but the relevance of this finding With the definition of sequences necessary for packaging
with respect to activation of this locus is not clear. For and regulation of gene expression it became possible to
the LMO2 gene activated in MLV trials, foamy virus generate foamy virus vectors with minimal cis-acting
integration was found in 700 kb distance. A similar sequences, that were replication defective, self inactivat-
pattern was found in a second study, where 628 ing and could carry transgenes of up to 9.2 kb in length.
integration sites in human 293 cells were mapped. These With no packaging cell line available, strategies to
authors described the direct hit of seven proto-oncogenes produce foamy virus vectors depend on transient
(for example, familial breast/ovarian cancer gene 2, co-transfection of up to three individual expression
BRCA, or epidermal growth factor receptor) and 108 cassettes for the structural proteins Gag, Pol and Env
events if analyzed for the insertion within a distance of and a plasmid encoding the transgene.25 This minimizes
0.5 Mb from an oncogene. In a canine in vivo model the generation of replication-competent PFV through
described recently,15 no foamy virus vector integration recombination, improving safety. Replication-defective
sites were found near the LMO2 oncogene and two were foamy virus vectors can now be produced consistently at
present in the MDS1-EviI gene locus. This was signi- titres of 107 ml after concentration, sufficient for ex vivo
ficantly less than for gammaretroviral vectors. In gene therapy applications, and without detectable helper
ontological studies, the authors found only one class of virus. Similar self-inactivating feline foamy virus virus
genes, involved with catabolism, to be overrepresented vectors have been produced that are capable of long-
for insertion of the PFV vector. Again, the relevance term transduction in cell lines.23 Foamy virus vectors are
of this finding is unclear. Taken together, a comparison more efficient than gammaretroviral vectors in transdu-
of lentivirus and foamy virus integration16 reveals that cing quiescent cells, but unlike lentiviral vectors, they fail
lentiviruses integrate preferentially into units of active to transduce truly resting cells in which foamy viruses
transcription and that, although all retrovirus vectors accumulate close to the centrosome, but uncoating is
will integrate close to proto-oncogenes, PFV does not impaired. On stimulation, however, disassembly and
have a preference for sites within genes, but only a viral infection proceeds, indicating that uncoating is the
modest preference for transcription start sites and for rate-limiting step for productive foamy virus infection of
CpG islands When the detailed mechanism of PFV growth-arrested cells.26 For example, quiescent human
integration is known, this may offer a further possible CD34+ progenitor cells can be transduced by PFV and
advantage over ORVs. The 3D structure of the PFV HIV vectors at a frequency of 40–50%, when transduction
integrase (IN) core domain has recently been published was assayed when cells were allowed to cycle. Under
and in vitro characterization of the enzyme has high- these conditions, however, transduction by MLV was
lighted the usefuleness of the PFV integrase to detailed only about 5%.
structural studies of integration. Although the PFV IN Vectors derived from PFV have been used to deliver
has little sequence homology to the HIV IN, it was still transgenes efficiently and transduce a diverse range of
sensitive to lentiviral strand transfer inhibitors, suggest- cells, including hematopoietic cells in vitro. For example,
ing that these reagents target highly conserved regions of Sun et al.27 inhibited hepatitis B virus replication by small
IN–DNA complexes17 The mechanism of the IN–DNA interfering RNAs delivered by PFV vectors. Taylor et al.28
recognition event is emerging as the functional LTR used PFV vectors to produce anti-HIV-1 transgenes and
nucleotides involved become known18 and the solubility block virus replication in primary macrophage derived
and catalytic efficiency of the PFV IN is established.19 from transduced peripheral blood CD34 cells and
No integrating vector system can be regarded as lymphoid cell lines. Rothenaigner et al.29 succeeded in
completely ‘safe’ based on its integration profile, and transducing human neural progenitor cells by PFV
safeguards are needed. To this end, an assay capable of vectors, leading to gene expression in a differentiation-
detecting activation of neighbouring genes could be dependent manner, which might prove to be important
useful.20,21 The potential risk of transformation might in targeting neural cells that differentiate in different
be reduced by introducing physiological promoters to areas of the brain. Gharwan et al (2007)30 described the
replace native or other retroviral promoters.22 There is efficient transduction (14–48%) of human and macaque
an additional risk with ORV vectors of a read-through embryonic stem cells when infected with a low multi-
into host sequences, resulting in fusion proteins with plicity of infection (MOI ¼ 1). Successful transduction of
unforeseen characteristics, if the termination of transcrip- the non-human primate cells indicates that foamy virus
tion at the 30 end of the viral genome is not tightly vectors are not encountering post-entry blocks in
controlled. In feline foamy virus the read-through from monkey cells, as was found for lentiviral vectors. In
the 30 LTR into neighbouring genes is a more stringent earlier experiments, designed to compare different viral
process, resulting in termination at the 30 end23 and this vectors and using non-obese diabetic/severe combined
may hold for PFV as discussed by Rethwilm.13 Once immunodeficiency mice repopulating human CD34+
integration is complete, subsequent gene expression cord blood, gene transfer levels of up to 84% had been
from the integrated genes transported by the vector can shown for foamy virus vectors, at least as efficiently as
Gene TherapyFoamy virus vectors enter a new age
O Erlwein and MO McClure
1427
lentiviral vectors and more efficiently than gammaretro- 6 weeks, but re-inoculation with vector was necessary to
viral vectors. However, the transition from in vitro to maintain the restored phenotype for another 6–7 weeks,
in vivo success has not been seamless, one example possibly due to genetic silencing. Si et al.34 reported on
of which is highlighted. the long-term repopulating activity of Fanconi anaemia
Fancc/ stem cells in mice after exposure to foamy
virus vector, and showed repopulation of primary and
Genes delivered by foamy virus vectors secondary recipients. This study successfully used a
short (8–14 h) vector exposure time without pre-stimula-
in vivo tion, compared with the 4-day gammaretroviral vector
Chronic granulomatosis disease is a diverse group of protocol that perpetuates a time-dependent increase in
hereditory diseases caused most commonly by a defect apoptosis and a reduction in myeloid progenitors and
in the phagocyte nicotinamide adenine dinucleotide repopulating ability. Integrated retroviral vectors may
phosphate oxidase (PHOX). The affected gene on the change the activity of certain genes, not just by directly
X chromosome codes for the gp91 protein, gp91-PHOX. inserting into these genes, but also by integration in close
When Sca1+ haematopoeitic stem cells (HSCs) isolated proximity to them. Hendrie et al.20 developed a plasmid-
from gene-deficient mice are transduced by PFV vector based assay to detect activation of a reporter gene by
carrying the PHOX transgene, 40–50% of the cells were inserted proviral vectors based on HIV, MLV and FV. In
transduced, a success rate similar to that achieved by this setting, PFV vectors had a lower propensity to
lentiviral vectors. However, replacing the transduced activate the reporter gene than vectors based on HIV or
haematopoeitic stem cells in gp91 PHOX-deficient gammaretrovirus, indicating that PFV vectors may be
mice was unsuccessful both in our own laboratory and safer in terms of failing to trans-activate neighbouring
in those of others. This was possibly due to either proto-oncogenes.
insufficient vector titre, viability of transfused cells
or toxicity when the vector was concentrated. However,
in an encouraging single experiment, CD34+ cells Foamy virus vector cure dogs of a genetic
transduced with PFV vector containing an enhanced
green fluorescent protein (EGFP) expression cassette
disease
we found that 15–20% of CD34+ cells expressed the The progression of foamy virus vectors towards the goal
marker 6 weeks after engrafting them into a non-obese of therapeutic exploitation will owe much to Russell’s
diabetic/severe combined immunodeficiency mouse. ground-breaking work of the last few years that has
Since those problematic early experiments, a number provided the proof of principle needed pour encourager les
of more successful in vivo applications have been autres to consider seriously the therapeutic potential
reported. When bone marrow cells, transduced with of these vectors. In the first place, his group showed
PFV vector expressing the DNA repair protein that when CD34+ cells fractionated from PBMCs were
O6-methylguanine DNA methyltransferase were engrafted transduced by foamy virus vectors expressing an EGFP
into mice, complete myeloablative therapy caused death marker gene and intravenously infused into two dogs,
in a proportion of animals. However, mice treated with a the dogs maintained long-term transgene expression of
sub-myeloablative therapy survived, and up to 55% of more than 450 days and 650 days, respectively, in all cells
them expressed O6-methylguanine DNA methyltransfer- of haematopoietic lineage. Approximately 19% of cells
ase in 50% of the progeny bone marrow cells for at least expressed the marker gene.24 When foamy and lentiviral
one year after transplantation.31 The fact that foamy vectors were compared in two dogs for their ability to
virus-mediated gene expression in quiescent stem cells is transduce long-term haematopoietic repopulating canine
only possible when the cells are subsequently stimulated cells in vivo, both dogs showed rapid neutrophil and
to divide raised questions about how effective PFV multi-lineage engraftment of transduced cells and a
vectors might be in post-mitotic tissue, such as the brain. similar level of transgene expression for more than
Caprariello et al.32 addressed this by comparing equal 700 days after engraftment.35 In one dog the transduction
titres of foamy and lentiviral vectors in vivo and found efficiency of the foamy virus vector was 4.5 and 4.7% for
that foamy virus vectors were significantly better at lymphocytes and granulocytes, respectively. For the
transducing brain parenchyma than the lentiviruses. One lentivirus vector, initial transduction efficiency was
week after transduction, the PFV transduced area higher (9.5% in lymphocytes and 9.3% in granulocytes),
measured 20.5 mm3 versus 1.17 mm3 for the lentiviral but by day 939 after transplantation had dropped to
vector. However, after 8 weeks, the transduced volume levels of the foamy virus vector. For the second dog it
of the brain was reduced to 0.521 mm3 and 0.367 mm3 was found that both vectors transduced lymphocytes
for the PFV vector and the lentiviral vector, respectively. and granulocytes at a similar rate of about 2% each.
The authors suggested that there was a transient phase of However, the piece de resistance was the reversal of
virus entry and gene expression but a lack of permanent disease phenotype by PFV-mediated gene transfer to
integration for the foamy virus vector resulting in only repopulating cells in a canine model. Canine leukocyte
2.5% of the initially transduced volume maintained. By adhesion deficiency is a rare autosomal recessive condi-
comparison, 31.6% of the volume transduced with the tion affecting Irish setters and similar to leukocyte
lentiviral vector maintained long-term expression. Liu adhesion deficiency in humans. A missense mutation in
et al.33 were of a similar opinion after injecting PFV vector the beta-2 integrin subunit gene ITGB2 (CD18) resulting
expressing glutamic acid decarboxylase into dorsal in a cysteine to serine change at position 36 of the CD18
root ganglion neuronal cells to attenuate below-injury subunit of leukocyte adhesion proteins, prevents leuko-
level central neuropathic pain, after spinal cord injury. cyte surface expression of the CD11/Cd18 complex.
Symptoms were reversed after 7 days, reversal lasted for Cell–cell adhesion events are consequently disrupted
Gene TherapyFoamy virus vectors enter a new age
O Erlwein and MO McClure
1428
and granulocytic dysfunction ensues. Dogs with this References
genetic immunodeficiency die at an early age of severe
recurrent bacterial infections, despite administration of 1 Murray SM, Picker LJ, Axthelm MK, Hudkins K, Alpers CE,
massive doses of antibiotics. Treatment of 4–11-week-old Linial ML. Replication in a superficial epithelial cell niche
canines by infusion of autologous CD34+ haematopoietic explains the lack of pathogenicity of primate foamy virus
stem cells transduced by foamy virus vector carrying the infections. J Virol 2008; 82: 5981–5985.
canine CD18 gene corrected the disease state in four out 2 Liu W, Worobey M, Li Y, Keele BF, Bibollet-Ruche F, Guo Y et al.
of five dogs without apparent side effects (the death of Molecular ecology and natural history of simian foamy virus
the fifth dog was unrelated to the treatment). It is infection in wild-living chimpanzees. PLoS Pathog 2008; 4:
impressive that transgene expression was at least three e1000097.
times higher than that in a similar experiment using 3 Morozov VA, Leendertz FH, Junglen S, Boesch C, Pauli G,
gammaretroviruses. The healthy state has been main- Ellerbrok H. Frequent foamy virus infection in free-living
chimpanzees of the Taı̈ National Park (Côte d’Ivoire). J Gen
tained for over 3 years, the abnormal lymphocyte
Virol 2009; 90: 500–506.
proliferation and neutrophil adhesion characteristics of
4 Jones-Engel L, Steinkraus KA, Murray SM, Engel GA, Grant R,
the disease state are now normal, and analysis of virus
Aggimarangsee N et al. Sensitive assays for simian foamy
integration indicates a lower risk of integration near viruses reveal a high prevalence of infection in commensal,
oncogenes compared with gammaretrovirus integra- free-ranging Asian monkeys. J Virol 2007; 81: 7330–7337.
tion.15 These dramatic results were the first demonstra- 5 Engel GA, Pizarro M, Shaw E, Cortes J, Fuentes A, Barry P et al.
tion of cure of a genetic disease by foamy virus vector Unique pattern of enzootic primate viruses in Gibraltar
therapy in a clinically relevant model and will surely macaques. Emerg Infect Dis 2008; 14: 1112–1115.
blaze the trail towards clinical trials of human disease. 6 Leendertz FH, Zirkel F, Couacy-Hymann E, Ellerbrok H,
Morozov VA, Pauli G et al. Interspecies transmission of simian
foamy virus in a natural predator-prey system. J Virol 2008; 82:
Prospects 7741–7744.
7 Boneva RS, Switzer WM, Spira TJ, Bhullar VB, Shanmugam V,
To summarize, foamy virus vectors are not being Cong ME et al. Clinical and virological characterization of
offered as the panacea for all that is wrong with other persistent human infection with simian foamy viruses. AIDS Res
vectors and, indeed, we have highlighted some of Hum Retroviruses 2007; 23: 1330–1337.
their disadvantages; an unknown receptor, no available 8 Jones-Engel L, May CC, Engel GA, Steinkraus KA, Schillaci MA,
packaging cell lines, an inability to transduce non- Fuentes A et al. Diverse contexts of zoonotic transmission of
dividing cells without previous stimulation and cyto- simian foamy viruses in Asia. Emerg Infect Dis 2008; 14:
toxic proteins at high concentrations, which has curtailed 1200–1208.
the production of concentrated vector and has possibly 9 Brooks JI, Merks HW, Fournier J, Boneva RS, Sandstrom PA.
been responsible for low engraftment. However, their Characterization of blood-borne transmission of simian foamy
advantages would seem to outweigh their current virus. Transfusion 2007; 47: 162–170.
disadvantages and their development since the first 10 Perkovic M, Schmidt S, Marino D, Russell RA, Stauch B,
replication-defective vector less than two decades ago Hofmann H et al. Species-specific inhibition of APOBEC3C by
has been staggering, given the handful of research the prototype foamy virus protein bet. J Biol Chem 2009; 284:
groups world wide dedicated to their investigation. In 5819–5826.
particular, their safety record and enhanced transduction 11 Yap MW, Lindemann D, Stanke N, Reh J, Westphal D,
of haematopoietic, neuronal and mesenchymal13 pro- Hanenberg H et al. Restriction of foamy viruses by primate
genitor cells bodes well for future clinical exploitation. Trim5alpha. J Virol 2008; 82: 5429–5439.
There is no doubt that the work from Russell’s laboratory 12 Lee EG, Linial ML. The C terminus of foamy retrovirus Gag
contains determinants for encapsidation of Pol protein into
showing that animals can be cured of a genetic defect
virions. J Virol 2008; 82: 10803–10810.
using gene therapy based on foamy virus vectors has
13 Rethwilm A. Foamy virus vectors: an awaited alternative to
given a much needed boost to the field. Other labora- gammaretro- and lentiviral vectors. Curr Gene Ther 2007; 7:
tories will now be encouraged to put effort into 261–271.
addressing the challenging questions of foamy virus 14 Hacein-Bey-Abina S, Garrigue A, Wang GP, Soulier J, Lim A,
biology knowing that clinical exploitation is an achiev- Morillon E et al. Insertional oncogenesis in 4 patients after
able goal. Over the next couple of years, we can expect a retrovirus-mediated gene therapy of SCID-X1. J Clin Invest 2008;
two-pronged approach to foamy virus vector develop- 118: 3132–3142.
ment. In the first place, given their safety record and 15 Bauer Jr TR, Allen JM, Hai M, Tuschong LM, Khan IF, Olson EM
given the similarities between canine leukocyte adhesion et al. Successful treatment of canine leukocyte adhesion
deficiency and the human disease, it would not be deficiency by foamy virus vectors. Nat Med 2008; 14: 93–97.
surprising to see extrapolation from the canine studies 16 Beard BC, Keyser KA, Trobridge GD, Peterson LJ, Miller DG,
into clinical trials. It remains to be seen whether vector Jacobs M et al. Unique integration profiles in a canine model of
titres and the lack of a packaging cell line will be long-term repopulating cells transduced with gammaretrovirus,
limiting. Simultaneously, a number of laboratories, world lentivirus, or foamy virus. Hum Gene Ther 2007; 18: 423–434.
–wide, are working towards a greater understanding of 17 Valkov E, Gupta SS, Hare S, Helander A, Roversi P, McClure M
foamy virus biology that is likely to overcome both et al. Functional and structural characterization of the integrase
disadvantages. from the prototype foamy virus. Nucleic Acids Res 2009; 37:
243–255.
18 Kang SY, Ahn DG, Lee C, Lee YS, Shin CG. Functional
nucleotides of U5 LTR determining substrate specificity of
Conflict of interest prototype foamy virus integrase. J Microbiol Biotechnol 2008; 18:
The authors declare no conflict of interest. 1044–1049.
Gene TherapyFoamy virus vectors enter a new age
O Erlwein and MO McClure
1429
19 Delelis O, Carayon K, Guiot E, Leh H, Tauc P, Brochon JC et al. 28 Taylor JA, Vojtech L, Bahner I, Kohn DB, Laer DV, Russell DW
Insight into the integrase-DNA recognition mechanism. et al. Foamy virus vectors expressing anti-HIV transgenes
A specific DNA-binding mode revealed by an enzymatically efficiently block HIV-1 replication. Mol Ther 2008; 16: 46–51.
labeled integrase. J Biol Chem 2008; 10: 27838–27849. 29 Rothenaigner I, Kramer S, Meggendorfer M, Rethwilm A,
20 Hendrie PC, Huo Y, Stolitenko RB, Russell DW. A rapid and Brack-Werner R. Transduction of human neural progenitor cells
quantitative assay for measuring neighboring gene activation by with foamy virus vectors for differentiation-dependent gene
vector proviruses. Mol Ther 2008; 16: 534–540. expression. Gene Ther 2009; 16: 349–358.
21 Bokhoven M, Stephen SL, Knight S, Gevers EF, Robinson IC, 30 Gharwan H, Hirata RK, Wang P, Richard RE, Wang L, Olson E
Takeuchi Y et al. Insertional gene activation by lentiviral and et al. Transduction of human embryonic stem cells by foamy
gammaretroviral vectors. J Virol 2009; 83: 283–294. virus vectors. Mol Ther 2007; 15: 1827–1833.
22 Zychlinski D, Schambach A, Modlich U, Maetzig T, Meyer J, 31 Cai S, Ernstberger A, Wang H, Bailey BJ, Hartwell JR, Sinn AL
Grassman E et al. Physiological Promoters Reduce the Genotoxic et al. In vivo selection of hematopoietic stem cells transduced at a
Risk of Integrating Gene Vectors. Mol Ther 2008; 4: 718–725. low multiplicity-of-infection with a foamy viral MGMT(P140K)
23 Bastone P, Romen F, Liu W, Wirtz R, Koch U, Josephson N et al. vector. Exp Hematol 2008; 36: 283–292.
Construction and characterization of efficient, stable and 32 Caprariello AV, Miller RH, Selkirk SM. Foamy virus as a gene
safe replication-deficient foamy virus vectors. Gene Ther 2007; transfer vector to the central nervous system. Gene Ther 2009; 16:
7: 613–620. 448–452.
24 Kiem HP, Allen J, Trobridge G, Olson E, Keyser K, Peterson L 33 Liu W, Liu Z, Liu L, Xiao Z, Cao X, Cao Z et al. A novel human
et al. Foamy-virus-mediated gene transfer to canine repopulating foamy virus mediated gene transfer of GAD67 reduces neuro-
cells. Blood 2007; 109: 65–70. pathic pain following spinal cord injury. Neurosci Lett 2008; 432:
25 Wiktorowicz T, Peters K, Armbruster N, Steinert AF, Rethwilm A. 13–18.
Generation of an improved foamy virus vector by dissection of 34 Si Y, Pulliam AC, Linka Y, Ciccone S, Leurs C, Yuan J et al.
cis-acting sequences. J Gen Virol 2009; 90: 481–487. Overnight transduction with foamy viral vectors restores the
26 Lehmann-Che J, Renault N, Giron ML, Roingeard P, Clave E, long-term repopulating activity of Fancc/ stem cells. Blood
Tobaly-Tapiero J et al. Centrosomal latency of incoming foamy 2008; 112: 4458–4465.
viruses in resting cells. PLoS Pathog 2007; 3: e74. 35 Trobridge GD, Allen JM, Peterson L, Ironside CG, Russell D,
27 Sun Y, Li Z, Li L, Li J, Liu X, Li W. Effective inhibition of hepatitis Kiem HP. Foamy and Lentiviral Vectors Transduce Canine Long-
B virus replication by small interfering RNAs expressed from term Repopulating Cells at Similar Efficiency. Hum Gene Ther
human foamy virus vectors. Int J Mol Med 2007; 19: 705–711. 2009; 5: 519–523.
Gene TherapyYou can also read
