RNA pieces in the spliceosome, has a domain
V counterpart, containing a 2-nucleotide
bulge located 5 base pairs away from an AGC
triad (10). Formation of an analogous metal-
binding platform in this region of U6 (11) may
explain the apparent ability of spliceosomal
RNAs to retain catalytic activity in the com-
plete absence of the many protein components
that usually accompany splicing (12). A
domain V-like element could have played a
major role during the RNA world era of evolu-
tion, serving as the catalytic center for RNA
cleavage, transesterification, and polymeriza-
The new structure provides a powerful
starting point for future investigations of
group II introns and the spliceosome. The
lack of electron density for domain VI,
which is important for the first step of splic-
ing in many group II introns, and the
absence of exons from the structure preclude
us from seeing how these elements dock
onto the surface created by domains I to V.
Thus, the structural details of substrate
recognition and catalysis remain undefined.
The nature of the conformational change
known to separate the two steps of splicing
(13) also remains unclear. Finally, it will be
important for our understanding of group II
intron self-splicing to capture the structures
of the other intermediates along the splicing
pathway and to pursue experiments that link
features of these structures with functionally
1. A. M. Pyle, Ribozymes and RNA Catalysis (Royal Society
of Chemistry, Cambridge, UK, ed. 2, 2008).
2. A. M. Pyle, A. M. Lambowitz, The RNA World (Cold Spring
Harbor Laboratory, Cold Spring Harbor, NY, ed. 3, 2006).
3. N. Toor, K. N. Keating, S. D. Taylor, A. M. Pyle, Science
320, 77 (2008).
4. M. Stahley, S. Strobel, Science 309, 1587 (2005).
5. T. Steitz, J. Steitz, Proc. Natl. Acad. Sci. U.S.A. 90, 6498
6. L. Zhang, J. A. Doudna, Science 295, 2084 (2002).
7. R. Sigel et al., Nat. Struct. Mol. Biol. 11, 187 (2004).
8. P. Gordon, J. Piccirilli, Nat. Struct. Biol. 8, 893 (2001).
9. P. Gordon, R. Fong, J. Piccirilli, Chem. Biol. 14, 607
10. G. Shukla, R. Padgett, Mol. Cell 9, 1145 (2002).
11. S. Yean, G. Wuenschell, J. Termini, R. Lin, Nature 408,
12. S. Valadkhan, J. Manley, Nature 413, 701 (2002).
13. G. Chanfreau, A. Jacquier, EMBO J. 15, 3466 (1996).
www.sciencemag.org SCIENCE VOL 320 4 APRIL 2008
CREDITS: (TOP) HANS PAERL/UNIVERSITY OF NORTH CAROLINA; (BOTTOM) SATELLITE PHOTO, DIGITALGLOBE
utrient overenrichment of waters by
urban, agricultural, and industrial
development has promoted the
growth of cyanobacteria as harmful algal
blooms (see the figure) (1, 2). These blooms
increase the turbidity of aquatic ecosystems,
smothering aquatic plants and thereby sup-
pressing important invertebrate and fish habi-
tats. Die-off of blooms may deplete oxygen,
killing fish. Some cyanobacteria produce tox-
ins, which can cause serious and occasionally
fatal human liver, digestive, neurological, and
skin diseases (1–4). Cyanobacterial blooms
thus threaten many aquatic ecosystems,
including Lake Victoria in Africa, Lake Erie in
North America, Lake Taihu in China, and the
Baltic Sea in Europe (3–6). Climate change is
a potent catalyst for the further expansion of
Rising temperatures favor cyanobacteria
in several ways. Cyanobacteria generally
grow better at higher temperatures (often
above 25°C) than do other phytoplankton
species such as diatoms and green algae (7, 8).
This gives cyanobacteria a competitive advan-
tage at elevated temperatures (8, 9). Warming
of surface waters also strengthens the vertical
stratification of lakes, reducing vertical mix-
ing. Furthermore, global warming causes
lakes to stratify earlier in spring and destratify
later in autumn, which lengthens optimal
growth periods. Many cyanobacteria exploit
these stratified conditions by forming intra-
cellular gas vesicles, which make the cells
buoyant. Buoyant cyanobacteria float upward
when mixing is weak and accumulate in dense
surface blooms (1, 2, 7) (see the figure). These
surface blooms shade underlying nonbuoyant
phytoplankton, thus suppressing their oppo-
nents through competition for light (8).
Cyanobacterial blooms may even locally
increase water temperatures through the
intense absorption of light. The temperatures
of surface blooms in the Baltic Sea and in
Lake IJsselmeer, Netherlands, can be at least
1.5°C above those of ambient waters (10, 11).
This positive feedback provides additional
competitive dominance of buoyant cyanobac-
teria over nonbuoyant phytoplankton.
Global warming also affects patterns of
precipitation and drought. These changes in
the hydrological cycle could further enhance
cyanobacterial dominance. For example,
more intense precipitation will increase sur-
face and groundwater nutrient discharge into
water bodies. In the short term, freshwater dis-
charge may prevent blooms by flushing.
However, as the discharge subsides and water
residence time increases as a result of drought,
nutrient loads will be captured, eventually pro-
moting blooms. This scenario takes place
when elevated winter-spring rainfall and
flushing events are followed by protracted
periods of summer drought. This sequence of
A link exists between global warming and
the worldwide proliferation of harmful
Blooms Like It Hot
Hans W. Paerl
and Jef Huisman
Institute of Marine Sciences, University of North Carolina
at Chapel Hill, Morehead City, NC 28557, USA. E-mail:
Institute for Biodiversity and
Ecosystem Dynamics, University of Amsterdam, 1018 WS
Amsterdam, Netherlands. E-mail: jef.huisman@science.
Undesired blooms. Examples of large water bodies
covered by cyanobacterial blooms include the Neuse
River Estuary, North Carolina, USA (top) and Lake
Victoria, Africa (bottom).
Published by AAAS
on April 4, 2008 www.sciencemag.orgDownloaded from
4 APRIL 2008 VOL 320 SCIENCE www.sciencemag.org
events has triggered massive algal blooms in
aquatic ecosystems serving critical drinking
water, fishery, and recreational needs. At-
tempts to control fluctuations in the discharge
of rivers and lakes by means of dams and
sluices may increase residence time, further
aggravating cyanobacteria-related ecological
and human health problems.
In addition, summer droughts, rising sea
levels, increased withdrawal of freshwater for
agricultural use, and application of road salt as
a deicing agent have led to rising lake
salinities in many regions. Several common
cyanobacteria are more salt-tolerant than
freshwater phytoplankton species (12, 13).
This high salt tolerance is reflected by increas-
ing reports of toxic cyanobacterial blooms in
brackish waters (2, 6).
Some cyanobacteria have substantially
expanded their geographical ranges. For
example, Cylindrospermopsis raciborskii—
the species responsible for “Palm Island mys-
tery disease,” an outbreak of a severe hepati-
tis-like illness on Palm Island, Australia (4)—
was originally described as a tropical/subtrop-
ical genus. The species appeared in southern
Europe in the 1930s and colonized higher lat-
itudes in the late 20th century. It is now wide-
spread in lakes in northern Germany (14).
Similarly, the species was noted in Florida
almost 35 years ago and is now commonly
found in reservoirs and lakes experiencing
eutrophication in the U.S. southeast and mid-
west (2). It is adapted to the low-light condi-
tions that typify eutrophic waters, prefers
water temperatures above 20°C, and survives
adverse conditions through the use of special-
ized resting cells (14). These bloom character-
istics suggest a link to eutrophication and
More detailed studies of the population
dynamics in cyanobacterial blooms are needed.
For example, competition between toxic and
nontoxic strains affects the toxicity of
cyanobacterial blooms (15). Furthermore,
viruses may attack cyanobacteria and mediate
bloom development and succession (16). It is
unclear how these processes are affected by
global warming. What is clear, however, is that
high nutrient loading, rising temperatures,
enhanced stratification, increased residence
time, and salination all favor cyanobacterial
dominance in many aquatic ecosystems. Water
managers will have to accommodate the effects
of climatic change in their strategies to combat
the expansion of cyanobacterial blooms.
1. J. Huisman, H. C. P. Matthijs, P. M. Visser, Harmful
Cyanobacteria (Springer, Dordrecht, Netherlands, 2005).
2. H. W. Paerl, R. S. Fulton III, in Ecology of Harmful Marine
Algae, E. Graneli, J. Turner, Eds. (Springer, Berlin, 2006),
3. I. Chorus, J. Bartram, Toxic Cyanobacteria in Water
(Spon, London, 1999).
4. W. W. Carmichael, Human Ecol. Risk Assess. 7, 1393
5. L. Guo, Science 317, 1166 (2007).
6. S. Suikkanen, M. Laamanen, M. Huttunen, Estuar. Coast.
Shelf Sci. 71, 580 (2007).
7. C. S. Reynolds, Ecology of Phytoplankton (Cambridge
Univ. Press, Cambridge, 2006).
8. K. D. Jöhnk et al., Global Change Biol. 14, 495 (2008).
9. J. A. Elliott, I. D. Jones, S. J. Thackeray, Hydrobiologia
559, 401 (2006).
10. M. Kahru, J.-M. Leppänen, O. Rud, Marine Ecol. Prog.
Ser. 101, 1 (1993).
11. B. W. Ibelings, M. Vonk, H. F. J. Los, D. T. van der Molen,
W. M. Mooij, Ecol. Appl. 13, 1456 (2003).
12. L. Tonk, K. Bosch, P. M. Visser, J. Huisman, Aquat. Microb.
Ecol. 46, 117 (2007).
13. P. H. Moisander, E. McClinton III, H. W. Paerl, Microb.
Ecol. 43, 432 (2002).
14. C. Wiedner, J. Rücker, R. Brüggemann, B. Nixdorf,
Oecologia 152, 473 (2007).
15. W. E. A. Kardinaal et al., Aquat. Microb. Ecol. 48, 1 (2007).
16. M. Honjo et al., J. Plankton Res. 28, 407 (2006).
n 2006, Yamanaka and colleagues (1) dis-
covered that mouse fibroblasts could be
reprogrammed to a pluripotent, embry-
onic stem (ES) cell–like state by the simple
introduction of four transcription factors,
Oct4, Sox2, Klf4, and c-Myc. This finding has
since been reproduced (2–6) and extended to
human fibroblasts using the same cocktail of
genes (7, 8) or one composed of Oct4, Sox2,
Nanog, and Lin28 (9). These so-called “in-
duced pluripotent stem cells” (iPS cells)
appear similar to ES cells in that they can give
rise to all the cells of the body and display fun-
damental genetic and morphologic ES cell
characteristics (see the figure). The concept of
an iPS cell brings together decades of work in
the fields of ES cell biology and nuclear
reprogramming that predicted it might be pos-
sible to impose pluripotency upon a somatic
cell (10). iPS cells not only have the potential
to produce patient-specific stem cells, but
they also provide a platform to study the biol-
ogy of pluripotency and cell reprogramming.
In Science Express, Aoi et al. (11) broaden the
application of iPS cell methodology to murine
epithelial cell types, highlighting differences
when compared with reprogramming of
fibroblasts. And on page 97 of this issue,
Viswanathan et al. (12) address the role of one
of the reprogramming factors, Lin28, in regu-
lating microRNAs (miRNAs) in ES cells. The
findings of Viswanathan et al., and recent
work by Benetti et al. (13) and Sinkkonen et al.
(14), advance our knowledge of the little-
understood roles of miRNAs in ES cells.
Collectively, these studies take us closer to
understanding how ES cells maintain an
undifferentiated, self-renewing, and pluripo-
tent state, and to defining how pluripotency
can be imposed on other cell types.
To date, fibroblasts and mesenchymal
stem cells have been used to generate iPS cells
(1–9). A next step is to determine whether
other cell types are susceptible to reprogram-
ming. Toward this end, Aoi et al. produced iPS
cells from two epithelial cell populations,
adult mouse hepatocytes and gastric epithelial
cells, by expressing Oct4, Sox2, Klf4, and
c-Myc. Like iPS cells generated from fibro-
blasts (iPS-fibroblast), those from primary
hepatocytes (iPS-Hep) and gastric epithelial
cells (iPS-Stm) were pluripotent and gave rise
to adult and germline chimeras. However,
iPS-Hep and iPS-Stm differ from iPS-fibro-
blast cells in several important respects, indi-
cating that the dynamics of reprogramming
may not be equivalent in these cell types. For
instance, although c-Myc was used, iPS-Hep
and iPS-Stm cell–derived chimeric mice did
not display the c-Myc–dependent tumori-
genicity observed in iPS-fibroblast–derived
chimeric mice. In addition, iPS-Hep and iPS-
Stm cells could be generated using less strin-
gent selection conditions. Thus, epithelial cell
types may be more prone to reprogramming
How do these differences inform us about
the mechanism of reprogramming? Given that
ES cells are an epithelial population, charac-
terized by cell adhesion (mediated by the
membrane protein E-cadherin), one possibil-
ity is that epithelialization is an event required
The requirements for reprogramming different
somatic cell types to a pluripotent state may
not be equivalent.
Anne G. Bang and Melissa K. Carpenter
Novocell Inc., 3550 General Atomics Court, San Diego, CA
92121, USA. E-mail: email@example.com
Published by AAAS
on April 4, 2008 www.sciencemag.orgDownloaded from