24 nt upstream of the splice junction
during NMD). As with SECIS binding, the
helicase activity of eIF4a3 is not required
for NMD either, suggesting that rather
than unwinding, the interaction of eIF4a3
with the sugar-phosphate backbone of
the RNA is of paramount importance for
This work also raises a number of inter-
esting questions. As the brain does not
respond to selenium depletion in the
same way, does this reflect altered rela-
tive levels of eIF4a3 and SBP2, which
allow for production of high levels of
redox-related proteins to protect neurons
from oxidative stress? Does eIF4a3 also
directly compete with rpL30 for binding
to the SECIS element, and does this
ensure that the ribosome is not tempted
of SBP2 binding? The SECIS element is in
the N terminus of GPx1; does the same
mechanism hold if the element is moved
elsewhere in the mRNA? Finally, it will be
important to find out the mechanism
involved in the upregulation of eIF4a3 in
selenium-deficient cells; is there a sele-
nium-dependent repressor protein bound
to the eIF4a3 mRNA that regulates
expression, or do increased protein levels
reflect a change in the rate of eIF4a3
synthesis and/or degradation or import
into the nucleus?
Berry, M.J., Banu, L., Chen, Y.Y., Mandel, S.J.,
Keiffer, J.D., Harney, J.W., and Larsen, P.R.
(1991). Nature 353, 273–276.
Budiman, M.E., Bubenik, J.L., Miniard, A.C.,
Middleton, L.M., Gerber, C.A., Cash, A., and
Driscoll, D.M. (2009). Mol. Cell 35, this issue,
Carlson, B.A., Xu, X.M., Gladyshev, V.N., and
Hatfield, D.L. (2005). J. Biol. Chem. 280, 5542–
Chavatte, L., Brown, B.A., and Driscoll, D.M.
(2000). EMBO J. 19, 306–314.
Cole, S.E., LaRiviere, F.J., Merrikh, C.N., and
Moore, M.J. (2009). Mol. Cell 34, 440–450.
Copeland, P.R., Fletcher, J.E., Carlson, B.A.,
Hatfield, D.L., and Driscoll, D.M. (2000). EMBO J.
Fletcher, J.E., Copeland, P.R., Driscoll, D.M., and
Krol, A. (2001). RNA 7, 1442–1453.
Moriarty, P.M., Reddy, C.C., and Maquat, L.E.
(1998). Mol. Cell. Biol. 18, 2932–2939.
Papp, L.V., Lu, J., Holmgren, A., and Khanna, K.K.
(2007). Antioxid. Redox Signal. 9, 775–806.
Squires, J.E., Stoytchev, I., Forry, E.P., and Berry,
M.J. (2007). Mol. Cell. Biol. 27, 7848–7855.
Unraveling the Secrets of Regulating
Mitochondrial DNA Replication
Michele M. Klingbeil1,* and Theresa A. Shapiro2
1Department of Microbiology, University of Massachusetts, Amherst, MA 01003, USA
2Division of Clinical Pharmacology, Department of Medicine, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA
In this issue, Liu et al. (2009) report that maxicircle DNA copy number in trypanosomes is regulated by
proteolysis of a helicase; the complex kinetoplast DNA system yields a clear view of how mitochondrial
DNA replication can be regulated.
Genomic stability requires that DNA is
replicated only once per cell cycle, and
precise regulation is essential to prevent
the reinitiation of synthesis within the
same S phase. Origins of DNA replication
act as landing pads for numerous initia-
tor proteins, but it is the recruitment,
assembly, and activation of a helicase
that licenses the initiation of replication.
Given their pivotal importance, it is not
surprising that nature has devised diverse
mechanisms for controlling the amount
of active helicase available for DNA repli-
cation. In addition to alterations in its
synthesis, mechanisms include degrada-
tion of helicase loading proteins, regula-
tion by phosphorylation of the helicase
or its loading proteins, and additional
proteins that interact with and inhibit the
loading proteins (Figure 1).
Eukaryotes must also maintain multiple
copies of an extranuclear genome, the
mitochondrial DNA (mtDNA). In most
organisms, mtDNA replication occurs
throughout the cell cycle, yet segregation
must be tightly controlled so that the
mitochondrial genome is properly distrib-
uted between daughter cells. Failure to
maintain mtDNA copy number results
in mitochondrial malfunctions that are
associated with several human diseases
(Copeland, 2008). DNA helicases Pif1p
tenance in yeast and metazoans, respec-
tively, and changes in their activity are
correlated with mtDNA abundance. RNAi
silencing of Twinkle expression results in
sion leads to an increase in mtDNA copy
number, indicating that Twinkle performs
a rate-limiting step in mtDNA replication
(Tyynismaa et al., 2004). However, the
underpinning mechanism for maintaining
mtDNA copy number has remained a
mystery. Englund and colleagues now
Molecular Cell 35, August 28, 2009 ª2009 Elsevier Inc.
demonstrate that a mitochondrial replica-
tive helicase may itself be directly regu-
lated, addressing an important aspect of
mtDNA replication that has been difficult
to assess in other eukaryotic models (Liu
et al., 2009).
There is great diversity in mtDNA struc-
ture and modes of replication, with the
most unusual example found in trypano-
somes. These protists posses kinetoplast
DNA (kDNA) that is composed of thou-
sands of minicircles and tens of maxi-
circles catenated into a single network
cation requires the proper duplication of
all circles to maintain minicircle sequence
diversity and mitochondrial function, and
in synchrony with nuclear S phase. The
massive and topologically complex struc-
ture of kDNA provides a highly informative
reporter system for studying DNA metab-
olism in vivo. Vivid lesions in the network
and its replicating monomers provide
multiple independent lines of evidence
that pinpoint with confidence the site of
a mechanistic defect. These advantages
are clearly evident in the report from
Englund and colleagues, who utilize strik-
ing changes in the kDNA network, visible
even by light microscopy, to identify
TbPIF2 as the helicase for replication of
kDNA maxicircles (Liu et al., 2009).
The Pif1p helicase subfamily was origi-
nally described in yeast based on two
paralogs, Pif1p and Rrm3p, both of which
are dually targeted to the nucleus and
mitochondria (Boule ´ and Zakian, 2006).
Remarkably, TbPIF2 is just 1 of 6 PIFs tar-
geted to the mitochondrion of trypano-
somes. This seeming redundancy is not
unprecedented, as at least six DNA poly-
somerases are exclusively targeted to
the Trypanosoma brucei mitochondrion.
However, Liu et al. (2009) have found
that silencing TbPIF2 expression resulted
in rapid loss of maxicircles with little effect
on minicircles and that the five other
mitochondrial PIFs could not compen-
sate, suggesting that trypanosomes have
evolved to utilize the multiple PIFs in
nonredundant roles. Conversely, overex-
pression of TbPIF2 increased maxicircle
copy number, again with significantly
less change in minicircles. The selective
control of maxicircle abundance corre-
lated well with PIF abundance and
mimicked the pattern described for the
helicase Twinkle. Lastly, overexpression
of an inactive mutant of TbPIF2 acted in
a dominant-negative fashion to cause
rapid maxicircle loss, indicating that it
is helicase catalytic activity that is essen-
tial for maintaining maxicircle abundance.
This series of experiments identifies
TbPIF2 as the rate-limiting step in maxi-
in kDNA replication.
Most importantly, Liu et al. (2009)
TbPIF2 is controlled through proteolytic
degradation by TbHslVU, thus providing
the first example of a mechanism for
regulating mtDNA replication licensing.
recently characterized as a mitochondrial
proteasome in T. brucei that dramatically
affected kDNA copy number (Li et al.,
2008). However, no substrates for this
protease were identified. TbPIF2 now
provides a missing link between TbHslVU
and kDNA replication. Given the paucity
of knowledge on maxicircle DNA replica-
tion machinery, Liu et al. (2009) have
also added essential tools to study maxi-
circle replication by showing that the
ters are gapped, similar to the well-
characterized minicircle replication inter-
mediates. These gapped progeny may
represent another level of control to pre-
vent rereplication of the numerous kDNA
now left with even more questions and
avenues to pursue. Does TbHslVU pro-
teolysis regulate other kDNA replication
proteins? Almost certainly yes, since
overexpression of TbPIF2 does not mimic
the massive increase in minicircles seen
when TbHslVU expression is silenced
(Li et al., 2008). Does TbPIF2 interact
selectively with 1 of the 6 mtDNA poly-
merases in a processive maxicircle repli-
some, while another TbPIF is specific for
minicircles? Three mtDNA polymerases
are essential for maintaining kDNA copy
number, but the division of labor at repli-
cation forks is still unclear (Chandler
et al., 2008). How are proteins targeted
for degradation by TbHslVU? Does the
TbHslVU proteasome play a role in the
synchrony of nuclear and mitochondrial
replication in trypanosomes? Does an
HslVU-like system regulate replicative
helicases in other mitochondria? No
single regulatory mechanism common to
all organisms or even to all tissues in the
same organism probably exists, but
some variation on helicase turnover may
be a general mechanism for governing
mtDNA copy number.
Figure 1. Comparisons of Regulatory Mechanisms for Select Replicative Helicases
The licensing of DNA replication origins is regulated to prevent rereplication in the same cell cycle. Prere-
plication complexes are assembled during the M-G1 transition, and Cdt1 plays a key role in nuclear DNA
licensing. Mechanisms for regulating DNA licensing vary widely, with additional regulatory mechanisms
apparent as the complexity of the genome increases. For reviews on replication licensing and cell-cycle
regulation of DNA replication, see Blow and Dutta, 2005; Nielsen and Løbner-Olesen, 2008; and Sclafani
and Holzen, 2007.
Molecular Cell 35, August 28, 2009 ª2009 Elsevier Inc.
Finally, is TbPIF2 a useful target for Download full-text
therapy? African trypanosomes cause
sleeping sickness in humans, a fatal dis-
ease if not treated. Available therapies
are widely acknowledged to be anti-
Screens of large chemical libraries have
recently yielded helicase inhibitors with
potent anti-infective activity (De Clercq,
2008). Perhaps TbPIF2 helicase, which
catalyzes an essential and rate-limiting
step in kDNA replication, will also prove
to be a vulnerable target for the develop-
ment of antitrypanosomal therapy.
Blow, J.J., and Dutta, A. (2005). Nat. Rev. Mol. Cell
Biol. 6, 476–486.
Boule ´, J., and Zakian, V.A. (2006). Nucleic Acids
Res. 34, 4147–4153.
Chandler, J., Vandoros, A.V., Mozeleski, B., and
Klingbeil, M.M. (2008). Eukaryot. Cell 7, 2141–
Copeland, W.C. (2008). Annu. Rev. Med. 59, 131–
De Clercq, E. (2008). Expert Opin. Emerg. Drugs
Li, Z., Lindsay, M.E., Motyka, S.A., Englund, P.T.,
Liu, B., Wang, J., Yaffe, N., Lindsay, M., Zhao, Z.,
Zick, A., Shlomai, J., and Englund, P.T. (2009).
Mol. Cell 35, this issue, 490–501.
Nielsen, O., and Løbner-Olesen, A. (2008). EMBO
Rep. 9, 151–156.
Sclafani, R.A., and Holzen, T.M. (2007). Annu. Rev.
Genet. 41, 237–280.
Tyynismaa, H., Sembongi, H., Bokori-Brown, M.,
Granycome, C., Ashley, N., Poulton, J., Jalanko,
A., Spelbrink, J.N., Holt, I.J., and Suomalainen, A.
(2004). Hum. Mol. Genet. 13, 3219–3227.
Ribosome Shifting or Splitting:
It Is All Up To the EF-G
Brooke Christian,1Emdadul Haque,1and Linda Spremulli1,*
1Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-3290, USA
Mitochondria possess two elongation factor Gs: one with translocation activity (EF-G1mt) and the other with
but, rather, in ribosome recycling.
Protein biosynthesis is a crucial process
for all cells. In eukaryotes, protein syn-
thesis occurs not only in the cell cyto-
plasm, but also in mitochondria. Mam-
polypeptides, all of which are essential
components of the electron transfer chain
and ATP synthase required for energy
system found in mammalian mitochondria
hasa number of
including ribosomes composed of smaller
rRNAs and more proteins than their
prokaryotic counterparts (Sharma et al.,
2003). Further, mitochondrial mRNAs are
generally leaderless or have only a few
nucleotides 50to the start codon. In
contrast, as illustrated in Figure 1, this
system has preserved a number of
translation (Spremulli et al., 2004). In the
initiation step, IF2mt:GTP promotes the
binding of fMet-tRNA to the small ribo-
somal subunit (28S), and IF3mtbinds to
and maintains a pool of free 28S small
subunits. In the elongation step, EF-
Tumt:GTP delivers the aminoacyl-tRNA
to the A site of the ribosome. EF-Tsmt
regenerates the EF-Tumt:GTP from EF-
drial system) catalyzes both the transloca-
tion of peptidyl-tRNA from the A site to the
P site and the movement of the mRNA to
expose the next codon in the A site. When
a stop codon, a release factor (RF1amt) in-
duces the hydrolysis of the newly formed
polypeptide (Soleimanpour-Lichaei et al.,
2007). Following the action of the release
factor, the 55S complex carrying the
deacylated tRNA and mRNA is targeted
by RRFmt(Rorbach et al., 2008). How the
tRNA and the mRNA dissociate from the
ribosome is not clear.
The activities of Escherichia coli EF-G
are well studied, and, in most bacteria,
this protein acts during both polypeptide
chain elongation and termination. How-
ever, database searches have indicated
that two forms of EF-G are present in
mitochondrial systems. A striking differ-
ence between traditional bacterial ribo-
some recycling and the mammalian
mitochondrial system is now revealed in
this issue of Molecular Cell by the work
of Tsuboi et al. (2009). These authors
provide convincing evidence that the
translocation but is inactive in the ribo-
with RRFmt to release the tRNA and
mRNA and to dissociate the monosome
into subunits. In contrast to EF-G1mt,
EF-G2mt is not active in translocation.
Thus, EF-G2mtis not biologically equiva-
lent to a traditional translocase, and
Tsuboi et al. (2009) now propose that
this factor be renamed RRF2mtto reflect
its role in ribosome recycling.
Molecular Cell 35, August 28, 2009 ª2009 Elsevier Inc.