for changes in cccDNA induced by LTβR.
Depletion of either APOBEC3A or APO-
BEC3B from infected liver cells abolished
cccDNA degradation elicited by IFN-α or
How might APOBEC3 enzymes selectively target cccDNA? Lucifora et al. implicate the HBV core protein, which binds to
both cccDNA and APOBEC3A, thereby
bringing them into close contact to promote
cccDNA deamination and degradation (see
the figure). Host genomic DNA, as well as
HBV DNA that has integrated into the host
genome, are spared.
What do these new findings mean for
understanding HBV infection and improv-
ing treatment? In terms of resolution of
infection, moderate cccDNA deamination
occurred even without IFN-α, providing a
possible explanation for the low frequency
of spontaneous clearance seen in infected
patients (2% of patients per year) (7). Still,
the reason for the disappointing response to
IFN-α–based HBV therapy remains unclear.
The high concentrations of IFN-α used by
Lucifora et al. to achieve maximal cccDNA
degradation might suggest that clinically
acceptable concentrations are not sufficient.
Alternatively, the finding that the expression
of APOBEC3A is only transient after IFN-α
treatment in patients underscores the fact
that hepatocytes rapidly become refractory
to the cytokine (8). To explain the unexpected
effect of APOBEC3 on double-stranded
DNA, rather than on single-stranded DNA,
the authors posit that transcription of viral
RNA generates single-stranded HBV DNA
that is accessible to APOBEC3 action. In
this scenario, the repressive effect of IFN-α
on HBV transcription (9) could antagonize
APOBEC3-mediated cccDNA deamina-
tion and efficient elimination. Lucifora et
al. also show that addition of potent nucleo-
side or nucleotide analogs that suppress viral
replication enhance the effect of IFN-α on
cccDNA. However, this does not mesh with
the disappointing, albeit limited, clinical
trial results obtained thus far (10). It may
be that preexisting cccDNA is more resis-
tant to that action of APOBEC compared to
cccDNA that is newly generated by reverse
transcription. Alternatively, the extent of
deamination in response to IFN-α may be
insufficient to irreversibly damage every
molecule of cccDNA. However, combining
newer analogs with IFNs that do not elicit a
refractory state (8), or with LTβR agonists,
may be worth testing.
Nonetheless, there is much about the
IFN-α/LTβR-APOBEC3 mechanism that
is mysterious. The amount of APOBEC3A
or APOBEC3B expression (at the transcript
level) did not always positively correlate
with cccDNA degradation. It is also unclear
why deaminated cccDNA is degraded rather
than repaired. In addition, hepatocytes lack-
ing an APOBEC3B allele associated with
more frequent HBV persistence (11) did not
show markedly diminished cccDNA degra-
dation in response to LTβR agonists. How-
ever, these in vitro results were obtained
with hepatocytes derived from a limited
number of donors.
The findings of Lucifora et al. add to
other studies showing that IFN-α acts
on multiple levels of the HBV life cycle.
Further defining the precise mechanisms
of the anti-HBV effect of IFN-α and other
cytokines on cccDNA degradation may
open up potential new avenues for more
effective HBV elimination. LTβR agonists,
if proven safe, could be an alternative to
IFN-α treatment, with enhanced cure rates.
Human studies are needed, but one hopes
that we may be getting closer to making the
notion of “once HBV, always HBV” a vanishing memory.
1. G. Dusheiko, Liver Int. 33 (suppl. 1), 137 (2013).
2. J. Lucifora et al., Science 343, 1221 (2014); 10.1126/
3. L. G. Guidotti et al., Science 284, 825 (1999).
4. P. Turelli, B. Mangeat, S. Jost, S. Vianin, D. Trono, Science
303, 1829 (2004).
5. R. Suspène et al., Proc. Natl. Acad. Sci. U.S.A. 102, 8321
6. M. Bonvin et al., Hepatology 43, 1364 (2006).
7. J. Liu et al., Gastroenterology 139, 474 (2010).
8. Z. Makowska, F. H. Duong, G. Trincucci, D. F. Tough, M. H.
Heim, Hepatology 53, 1154 (2011).
9. L. Belloni et al., J. Clin. Invest. 122, 529 (2012).
10. H. L. A. Janssen et al., Lancet 365, 123 (2005).
11. T. Zhang et al., Hum. Mol. Genet. 22, 1262 (2013).
Society, Where None Intrudes
H. Charles J. Godfray
A field study reveals an intricate tropical
community of plants, herbivores, and
Explaining the diversity and structure of networks of interacting species is one of the main challenges in mod-
ern ecology. Even building the food webs
needed to understand how these networks
are structured can be a massive undertaking,
especially in the tropics, where diversity is
highest and understanding of the taxonomy
of the species involved is often poor. How-
ever, progress has been made with commu-
nities of plant-eating insects and the special-
ized insects that feed on them, in particular
parasitic wasps. On page 1240 of this issue,
Condon et al. (1) combine field ecology and
molecular biology to describe the structure
of a self-contained community of tropical
plant-eating flies and the wasps that attack
them. The results reveal an unexpectedly
intricate community structure.
The number of insect herbivore species
on Earth is not known, but conservative esti-
mates are in the millions, with peak diver-
sity in the tropics (2). Lepidoptera (butter-
flies and moths) and beetles are the most
common types of plant-eating insects; sev-
eral groups of true flies have also evolved
herbivory. Nearly all of these species are
attacked by parasitic wasps, which typically
lay their eggs in or on their host’s body. The
developing wasp larva requires just a single
host (which it invariably kills) for its full
development. Parasitic wasps and ecologi-
cally similar insects are often called para-
sitoids, because their life history is interme-
diate between predators and true parasites.
Because parasitoids take some time to kill
their hosts, it is much easier to construct
plant-host-parasitoid food webs than is the
case for predators, which must be observed
in the act of killing and eating the prey to
establish a trophic link. Thus, some of the
best examples of large tropical food webs
resolved at the species level involve this type
of interaction (3– 5).
One of the major structuring forces
in community ecology is competition for
resources: The more distinct resources—
such as host-plant species—exist, the more
species of consumers can coexist. But very
often, a single host-plant species supports
Department of Zoology, University of Oxford, South Parks
Road, Oxford OX1 3PS, UK. E-mail: charles.godfray@zoo.