Download Exploring Bioinorganic Pattern Formation in Diatoms. A Story of

Scientific Correspondence
Exploring Bioinorganic Pattern Formation in Diatoms.
A Story of Polarized Trafficking
Chiara Zurzolo and Chris Bowler*
Dipartimento di Biologia e Patologia Cellulare e Molecolare, Università Federico II di Napoli, Naples, Italy
(C.Z.); and Laboratory of Molecular Plant Biology, Stazione Zoologica “Anton Dohrn,” Villa Comunale, I–
80121 Naples, Italy (C.B.)
The world’s oceans cover 70% of the surface of our
planet. The photosynthetic organisms living within
its photic zone are responsible for about one-half of
global primary productivity. The most successful organisms are thought to be photosynthetic prokaryotes (cyanobacteria and prochlorophytes) and a
class of eukaryotic unicellular algae known as diatoms (Norton et al., 1996; Van Den Hoek et al., 1997;
Falkowski et al., 1998).
Diatoms are likely to have arisen around 280 million years ago following an endosymbiotic event between a red eukaryotic alga and a heterotrophic
flagellate related to the Oomycetes (Medlin et al.,
1997, 2000). As a consequence, their only phylogenetic similarity to green algae and higher plants is
derived from the primary endosymbiotic event,
which is thought to have occurred at least 700 million
years ago (Kowallik, 1992). Therefore, diatom cells
have a range of features that make them highly divergent from the classical cellular structure of higher
plants, including:
(a) The use of the brown carotenoid pigment fucoxanthin for light energy transfer within the lightharvesting complexes of photosystems I and II.
(b) The presence of four membranes around their
plastids. The inner two membranes are equivalent to
the membranes that normally surround higher plant
chloroplasts, whereas the second membrane (as
counted from the outside) is thought to be derived
from the endosymbiont’s plasma membrane, and the
outer membrane is continuous with the endoplasmatic reticulum of the host cell.
(c) The presence of a rigid cell wall composed
largely of amorphous silica (i.e. glass). The exquisite
lacework-like patterning of diatom cell walls (see
example in Fig. 1) is reproduced with high fidelity
from generation to generation and is species specific.
For these reasons, it has been used since the last
century for taxonomic classification.
Current phylogenetic trees place diatoms close to
Alveolata lineages and far from the green and red
* Corresponding author; e-mail [email protected]; fax 39 – 081–
764 –1355.
www.plantphysiol.org/cgi/doi/10.1104/pp.010709.
lineages of other photosynthetic eukaryotes (Baldauf
et al., 2000). This has been further confirmed by the
analysis of diatom expressed sequence tags (ESTs),
although more similarity than expected was found
with Metazoan genes (S. Scala, N. Carels, A. Falciatore, M.L. Chiusano, and C. Bowler, unpublished
data). The EST program has revealed the presence of
genes encoding enzymes involved in cAMP metabolism, as well as genes encoding the components of
the animal extracellular matrix, such as fibronectins,
elastins, and tenascins, none of which appear to be
present in higher plants. As a consequence, one could
almost consider diatoms as photosynthetic animals
rather than unicellular plants!
The study of diatom biology therefore promises to
reveal many novel aspects (Scala and Bowler, 2001),
and now that so much information has been generated about the “conventional” metabolism of model
organisms such as yeast (Saccharomyces cerevisiae),
Caenorhabditis elegans, mouse (Mus musculus), and
Arabidopsis, it is possible to divert some attention to
more recalcitrant experimental systems. Given the
extreme ecological importance of diatoms, they
should come close to the top of the list of priorities
for plant biologists, and could become the system of
choice for studying phytoplankton interactions with
the marine environment. The range of molecularlevel technologies that have now been established,
such as genetic transformation (Dunahay et al., 1995;
Apt et al., 1997; Falciatore et al., 1999; Fischer et al.,
1999), protocols to study gene expression (Leblanc et
al., 1999), and the availability of several reporter
genes (Falciatore et al., 1999; Zaslavskaia et al., 2000)
make them all the more attractive.
It is not surprising that the study of silica cell wall
biogenesis has been a major focus of diatom research
for the last 150 years. However, most information is
descriptive, being based on microscopic observations, and only recently have biochemical and molecular techniques been incorporated, due largely to the
pioneering work of Nils Kröger and colleagues (for
review, see Kröger and Sumper, (1998)). The current
review aims to summarize the basic aspects of the
silica biomineralization process and to indicate the
major gaps in current understanding that can now be
addressed using knowledge and tools derived from
better studied organisms.
Plant Physiology, December 2001, Vol.
Downloaded
127, pp. from
1339–1345,
on Junewww.plantphysiol.org
15, 2017 - Published by
© www.plantphysiol.org
2001 American Society of Plant Biologists
Copyright © 2001 American Society of Plant Biologists. All rights reserved.
1339
Scientific Correspondence
Figure 1. Electron micrograph of the diatom Campylodiscus sp. at
890⫻ magnification. Photograph courtesy of Masahiko Idei.
SILICA AND THE DIATOM CELL CYCLE
Diatoms lie within the Heterokont division of the
class Bacillariophyceae (also known as Stramenopiles). They usually exist as single cells of between 5
␮m and 5 mm, depending on the species, although
some can form chains or colonies (Van Den Hoek et
al., 1997; Lee, 1999). Each cell is surrounded by a
unique type of cell wall (known as frustule) which
consists of two halves of amorphous polymerized
silica resembling a box with an overlapping lid (Fig.
2). The inner frustule is known as the hypotheca and
the outer one is denoted the epitheca. In addition to
the main valves (hypovalve and epivalve), each frustule often contains several ring-like silica structures
denoted girdle bands (Fig. 2).
There are two major diatom groups: the centric and
the pennate diatoms. These can be most simply distinguished from each other on the basis of cellular
symmetry: Centric diatoms are radially symmetrical
(resembling a petri dish), whereas pennate diatoms
are elongated and bilaterally symmetrical.
Diatom cell division typically proceeds through
asexual mitotic divisions. However, because the rigid
siliceous cell walls largely preclude cell growth
through expansion (except unilaterally by the addi-
tion of new girdle bands), the two daughter cells
must be formed inside the parent cell (Fig. 3; PickettHeaps et al., 1984, 1990; Van Den Hoek et al., 1997).
One sibling cell uses the epitheca of the parent cell as
a size guide to generate a new hypotheca, whereas
the other uses the parental hypotheca, which therefore becomes the epitheca of the daughter cell (Fig.
3). As a consequence, mitotically dividing populations decrease in mean size over time. In most species, size restoration occurs through sexual reproduction once a critical size threshold (typically 30%–40%
of the maximum size) has been reached (Mann, 1993;
Van Den Hoek et al., 1997).
The strict requirement for silica in frustule formation has led to the evolution of silica-dependent
checkpoints in diatom mitosis, which presumably
ensure that cell division does not occur unless there
are sufficient bio-available silica for frustule formation. Checkpoints have been found to occur in diatoms at both the G1/S and G2/M transitions (Brzezinski et al., 1990; Martin-Jezequel et al., 2000) and it
has been speculated that diatom DNA polymerases
may be silica dependent. This has not yet been confirmed at the molecular level.
It is known that diatoms possess silica uptake transporters that are able to accumulate silica at intracellular concentrations up to 250 times higher than in the
surrounding media (Martin-Jezequel et al., 2000).
Where it is localized intracellularly is still open to
debate (see “Transport to the SDV”). These transporters have novel primary structures, and by expression
in Xenopus laevis oocytes it has been found that they
are highly selective for the soluble form of silica,
Si(OH)4 (Hildebrand et al., 1997, 1998). Homologous
sequences do not exist in the Arabidopsis genome,
although it is possible that they will be found in rice,
which is known to have silicified hairs (Raven, 1983).
FORMATION OF A NEW FRUSTULE
Following mitosis, two daughter cells form inside
the parent cell, one of them bound by the parent
Figure 2. Schematic overview of the siliceous
components of diatom cell walls. Drawing by
Ian Nettleton.
1340
Downloaded from on June 15, 2017 - Published by www.plantphysiol.org
Copyright © 2001 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 127, 2001
Scientific Correspondence
Figure 3. Schematic overview of mitotic cell division and hypovalve and girdle band formation. For further details see text.
N, Nucleus; MC, microtubule center; SDV, silica deposition vesicle. Drawing by Ian Nettleton.
epitheca on one side and the other bound by the
hypotheca on the other side (Fig. 3). Each daughter
cell must then synthesize a hypotheca ex novo before
they can separate. This process has been observed
extensively in numerous diatom species at the ultrastructural level, such that a rather complete description of the process is now available (Pickett-Heaps et
al., 1990). The key events are shown in Figure 3 and
can be summarized as follows:
(a) The nucleus of each daughter cell moves to the
side of the cell where the hypotheca will be formed.
(b) A microtubule center (MC) positions itself between the nucleus and the plasma membrane above
which the hypotheca will eventually be placed.
(c) A specialized vesicle known as the silica deposition vesicle (SDV) forms between the MC and the
plasma membrane in a region that becomes the “pattern center.”
(d) The SDV elongates into a tube and then spreads
out perpendicularly to eventually form a huge vesicle along one side of the cell.
(e) A new valve is formed within the SDV by the
targeted transport of silica, proteins, and polysaccharides. During this process, the SDV becomes acidic as
a result of the silica polymerization process (Vrieling
et al., 1999). Some of the organic components eventually form a coat around the silica framework,
whereas others are involved in silica deposition.
(f) Once valve biogenesis is complete, it is
exocytosed by fusion of the SDV membrane (the
Plant Physiol. Vol. 127, 2001
silicalemma) with the plasma membrane. As a consequence, the inner face of the silicalemma is
thought to become the new plasma membrane.
(g) Following separation, the daughter cells can
expand unidirectionally along the cell division axis
by the biogenesis of girdle bands. These structures
are also formed within SDVs in a manner analogous
to that described above.
These events are very similar in both pennate and
centric diatoms, although the timing can be different
(Pickett-Heaps et al., 1990; Van Den Hoek et al.,
1997). Very little is known about the underlying
mechanisms controlling these events. Most intriguing is how such intricate structures can be generated
from a shapeless inorganic substrate. Although it has
been proposed that much of the early biomineralization process can proceed via physico-chemical space
filling of silica (Gordon and Drum, 1994), this is
certainly not sufficient to explain the exquisite
species-specific designs that are ultimately generated. The fact that frustule design is faithfully reproduced from parent to daughter cells indicates that
there is a significant genetic basis underlying bioinorganic pattern formation.
One strategy that appears to be important in pattern formation is known as macromorphogenesis,
whereby silica deposition is “molded” into a pattern
by the presence of organelles such as mitochondria
spaced at regular intervals along the cytoplasmic side
of the SDV (Schmid, 1994). Therefore, these or-
Downloaded from on June 15, 2017 - Published by www.plantphysiol.org
Copyright © 2001 American Society of Plant Biologists. All rights reserved.
1341
Scientific Correspondence
ganelles are thought to physically restrict the targeting of silica from the cytoplasm, to ensure the laying
down of a correctly patterned structure (for models,
see Schmid, 1994).
The importance of macromorphogenesis will only
be clear when it is known how silica and silicaassociated proteins and polysaccharides are deposited into the SDV. This is thought to occur, at least
partially, by their transport to the SDV in vesicles,
which is described as membrane-mediated morphogenesis (Pickett-Heaps et al., 1990).
ORGANIC COMPONENTS OF DIATOM
CELL WALLS
Much of the biochemical studies of frustule composition have been done in the pennate diatom Cylindrotheca fusiformis (Kröger and Sumper, 1998). This
work has led to the discovery of novel peptides
known as silaffins that may participate in the basic
biomineralization process within the SDV (Kröger et
al., 1999). These cationic peptides of various sizes are
generated from precursor polypeptides and are characterized by the presence of Lys-Lys repeat elements
that are covalently modified by the posttranslational
addition of novel chemical units, such as oligo-Nmethyl propylamines (Kröger et al., 2001). It is interesting that silaffins can promote the formation of
nanoscale silica spheres in vitro (Kröger et al., 1999).
Other major organic constituents of diatom biosilica are putrescine-derived long-chain polyamines.
Like the silaffins, these can also induce rapid silica
precipitation in vitro (Kröger et al., 2000), forming
spheres between 100 nm and 1 ␮m depending on the
polyamine used. Different diatoms have different
complements of silaffins and frustulins, which confer
species-specific differences to in vitro silica precipitation. It is remarkable that some of these combinations generate silica blocks in addition to silica
spheres. This observation infers that silaffins and
polyamines therefore may be involved in generating
the basic building blocks that are then modeled into
diatom frustules, and that combinatorial interactions
between them may be responsible for the underlying
genetic basis of species-specific pattern formation.
However, the fact that they are both bound tightly to
the silica scaffold of diatom cell walls begs the question: “How did they get there?” If they are bound to
a preformed silica framework, then how was it
formed? On the other hand, if silaffins and polyamines are responsible for silica pattern formation,
how are they themselves directed into a network that
determines the final pattern? To find out, it is necessary to understand the mechanisms underlying the
targeting of silaffins, polyamines, and silica to
the SDV.
Other proteinaceous components of diatom cell
walls include frustulins and pleuralins (formerly denoted HEP proteins; Kröger et al., 1997; Kröger and
1342
Wetherbee, 2000). Like the silaffins, pleuralins are
also tightly bound to silica and can only be removed
from diatom cell walls following the solubilization of
silica with hydrogen fluoride. Pleuralins are encoded
by a small multigene family and are characterized by
the presence of repeated amino acid motifs. It is
interesting that the localization of one of them,
pleuralin-1, is precisely restricted to the most terminal girdle band (known as the pleural band) of the
epitheca (Fig. 2), in the region of overlap with the
hypotheca. During cell division it also becomes associated with the pleural band of the hypotheca, at a
time when the parental hypotheca is functionally
converted into an epitheca of one of the daughter
cells (see Fig. 3; Kröger and Wetherbee, 2000). This is
a very significant observation, because it demonstrates that hypotheca-epitheca differentiation is under strict developmental control.
Pleuralin-1 is not targeted to the SDV but is directly
secreted into the cleavage furrow that forms between
the two daughter cells (Kröger and Wetherbee, 2000).
Association with the terminal girdle band of the hypotheca therefore occurs in the extracellular space.
Secretion is thought to occur by general mechanisms
of exocytosis, but it is nonetheless an intriguing example of polarized transport specifically toward the
cleavage furrow. It will be interesting to determine
how many other wall-associated proteins avoid the
SDV during their secretion and incorporation into
diatom frustules.
Frustulins are much more loosely associated with
diatom cell walls than are silaffins and pleuralins,
and can be extracted with EDTA. To date, five different frustulins have been described, all of which are
glycoproteins able to bind calcium due to the presence of EF hands (Kröger et al., 1994). They also
contain characteristic acidic Cys-rich domains, the
function of which is not yet known. Frustulins are not
thought to be involved in silica deposition, but constitute the outermost protein coat of the cell wall (van
de Poll et al., 1999).
TRANSPORT TO THE SDV
Both the silaffins and the frustulins must be targeted to the SDVs. Because the SDV is such an unusual but fundamental structure for the diatom cell,
understanding its biogenesis and dissecting the targeting mechanisms for silica and protein import are
of major biological interest. Furthermore, because the
frustule is formed on only one-half of the cell, it is a
dramatic example of cell polarity.
Silaffins, frustulins, and pleuralins are all synthesized as precursor proteins containing aminoterminal presequences that resemble typical signal
sequences for the co-translational import of proteins
into the endoplasmatic reticulum. Other unidentified
sequences are likely to be present within the silaffin
and frustulin precursors that target them from the
Downloaded from on June 15, 2017 - Published by www.plantphysiol.org
Copyright © 2001 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 127, 2001
Scientific Correspondence
Golgi apparatus to the SDV. The silaffin precursor is
cleaved into specific peptides that are found associated with silica in the frustule (Kröger et al., 1999).
Therefore, it is conceivable that this maturation event
occurs in a post-Golgi compartment (e.g. within the
transport vesicles or in the SDV itself after specific
targeting of the precursor protein to the SDV).
Bearing in mind the above considerations, it seems
likely that the Golgi apparatus is connected to the
SDV via transport vesicles. Such a system would also
provide a convenient and economical means for silicalemma expansion. Although clouds of vesicles emanating from the Golgi have often been observed
during cell wall deposition (Pickett-Heaps et al.,
1990), vesicle fusion to the SDV has not been clearly
established. However, because many specific mechanisms of vesicle targeting to organelles have been
studied intensively in other eukaryotes, parallels can
be easily searched for. One approach would be to
look for the presence of specific protein pairs such as
V-SNARE and T-SNARE, which are known to mediate vesicle docking (Chen and Scheller, 2001; Pelham,
2001). Furthermore, the use of drugs such as brefeldin A, which blocks forward vesicle trafficking from
the Golgi apparatus (Lippincott-Schwartz, 1983; Pelham, 1991), could clarify the role of Golgi vesicles in
SDV formation.
We envisage two possible mechanisms of vesicle
targeting from the Golgi apparatus to the SDV. Analysis of the silaffin and frustulin sequences reveals
stretches of amino acids highly enriched in acidic
residues, which is also typical of granins, secretory
proteins that accumulate in post-trans-Golgi network
(TGN) secretory granules in neuroendocrine cells
(Huttner et al., 1991). Therefore, these SDV-targeted
proteins could follow a pathway similar to regulated
secretion occurring at the plasma membrane of specialized secretory cells (Miller and Moore, 1990;
Tooze, 1998). In this case, peptides would accumulate
in the Golgi-derived vesicles by a process requiring
protein aggregation, due to acidic and Pro-rich
amino acid sequences (Benedum et al., 1987; Castle et
al., 1992; Rindler, 1998), and would then be targeted
to the SDV in response to a specific signal in a manner analogous to the regulated secretory pathway of
mammalian cells. A report of small vesicles surrounding the nascent SDV, but not yet fusing (Li and
Volcani, 1985), might support this model.
Another possibility would be a mechanism similar
to the Man 6-phosphate receptor-mediated pathway
that hydrolytic enzymes follow from the TGN to the
lysosomes (von Figura, 1991). In this case, a transmembrane receptor protein should be present in the
TGN, where it would bind to signals within the
SDV-targeted proteins separating them from other
proteins in the lumen and concentrating them in
coated vesicles that would then be specifically targeted to the silicalemma. The fact that some Golgiderived vesicles in diatoms seem to possess a coat
Plant Physiol. Vol. 127, 2001
(Gordon and Drum, 1994) could support this
hypothesis.
Conversely, how silica is targeted to the SDV is also
a mystery. In principle, it could follow the same route
followed for cell wall-associated proteins or could be
targeted independently. It is not known if it is already complexed with organic molecules during its
transport through the cell or even whether it is sequestered into vesicles. The existence of silica transport vesicles has been proposed (Schmid and Schulz,
1979), although evidence is poor. What is clear is that
if silica was already precipitated inside targeting vesicles, it would have been observed in electron micrographs. The fact that it has not (Pickett-Heaps et al.,
1990; Martin-Jezequel et al., 2000) indicates that silica
is only likely to encounter silaffins and polyamines
once inside the SDV. Furthermore, if silica is packaged into vesicles, it must first be targeted into them.
There is no evidence that this happens. Although
both silica transporters and ionophores have been
identified in diatoms (Bhattacharya and Volcani,
1983; Martin-Jezequel et al., 2000) it is not known
whether they are localized on vesicles inside the cell.
Furthermore, it is not clear why a specific vesicle
transport mechanism should be needed to transport
silica to the SDV because the presence of silica transporters on the silicalemma could be sufficient. The
availability of antibodies against the silica transporters would make it possible to determine whether
they are present on the silicalemma as well as on the
plasma membrane. This is one of the most urgent
questions that requires an answer.
If silica transporters are present on the silicalemma,
they could be targeted to it by endocytic vesicles
derived from the plasma membrane. This process
would both increase the size of the SDV and create
the conditions for concentrating silica within this
organelle. Fusion of Golgi-derived vesicles carrying
silaffins and polyamines with the SDV would then
initiate the process of silica deposition.
SDV FORMATION
Based on the above considerations, it is possible
that the SDV has an endocytic origin similar to the
early endosomal compartment present in mammalian cells (Woodman, 2000). The initial formation of
this organelle could be strictly linked (or driven)
by the positioning of the MC between the nucleus
and the plasma membrane immediately after cytokinesis. The fact that microtubule depolymerization
affects valve formation supports a role for microtubules in SDV formation (see Pickett-Heaps et al.,
1990, and refs. therein).
The observation that the acidity of the SDV increases with silica deposition (Vrieling et al., 1999)
can be paralleled with the maturation hypothesis of
mammalian cells, which postulates that late endosomes/lysosomes derive from maturation of early
Downloaded from on June 15, 2017 - Published by www.plantphysiol.org
Copyright © 2001 American Society of Plant Biologists. All rights reserved.
1343
Scientific Correspondence
endosomes (Mullins and Bonifacino, 2001). The identification of markers specific for early and late endosomes and lysosomes (e.g. Rab5, Rab7, and Rab9,
respectively; Zerial and McBride, 2001) on the silicalemma therefore could provide some clues to the
origin of the SDV.
An additional, albeit speculative, connection between silica polymerization and lysosomes comes
from the recent finding that a lysosomal cathepsin
l-like hydrolase is the major silica binding protein
within the silica-rich spicules of sponges (Shimizu et
al., 1998).
EXOCYTOSIS
The final stage of the valve formation process is
exocytosis of the SDV. There are at least four models
for how this could occur (Pickett-Heaps et al., 1990;
van de Poll et al., 1999). It is most likely that the SDV
fuses at specific sites, perhaps the valve edge. To
drive fusion at specific sites, landmarks for vesicle
docking must be present. Furthermore, timing must
be precise. In the case of mammalian and plant cells,
specific vesicle fusion is driven by the pairing of Vand T-SNAREs (Chen and Scheller, 2001; Pelham,
2001). However, additional factors define cognate
fusion and docking sites. One example is represented
by the exocyst in yeast, a multiprotein complex involved in vesicle targeting and docking at the plasma
membrane (TerBush et al., 1996). The exocyst is specifically localized to sites of active exocytosis and its
proper localization is thought to be important for
spatial regulation of secretion in yeast and mammalian cells (TerBush et al., 1996; Zahraoui et al., 2000).
It has been found recently that the small GTPbinding protein Rho1 is involved in regulation of
exocyst localization at the plasma membrane (Guo et
al., 2001). It would be interesting to analyze whether
a similar multiprotein complex exists at specific sites
of the SDV or plasma membranes of diatoms.
fundamental features of their intracellular transport,
and could reveal where different silica transporters
are localized and whether they are mobile during
SDV formation and silica deposition.
Of major importance is the identification of other
components regulating SDV biogenesis, frustule formation, and exocytosis. One extremely valuable approach is the random sequencing of expressed genes
to generate ESTs. Several thousand ESTs have been
generated in the pennate diatom Phaeodactylum tricornutum, and sequences encoding proteins homologous to a range of GTP-binding proteins such as Rabs
and Rhos have already been identified, as have COP
coatomer proteins, dynamin, cathespsins, and clathrin adaptor proteins (S. Scala, N. Carels, A. Falciatore, M.L. Chiusano, and C. Bowler, unpublished
data). Therefore, these provide important new substrates for initiating molecular and cellular studies in
diatoms.
The species-specific patterns of silica nanofabrication indicates that there is a considerable genetic
basis underlying pattern formation. The isolation of
pattern mutants will be an important approach to
develop insights into how this works. The two diatom species for which most molecular tools have
been developed, P. tricornutum and C. fusiformis (Scala and Bowler, 2001), represent simple experimental
systems that can be used for such approaches.
The eventual elucidation of the mechanisms used
by diatoms to transform soluble silica into sturdy
intricate structures at ambient temperatures and
pressures that way exceed current human capabilities could eventually allow materials scientists to
generate micrometer scale silica structures for a
whole range of nanotechnological applications
(Mann and Ozin, 1996; Morse, 1999; Parkinson and
Gordon, 1999). This, combined with the intrinsic scientific interest of understanding such a process,
makes diatom cell wall biogenesis an attractive research topic that is in desperate need of intensive
molecular and cellular-based studies.
FUTURE PROSPECTS
There is clearly a great need for the development of
pharmacological, biochemical, molecular, and genetic tools to address the novel aspects of diatom
frustule biogenesis such as SDV formation, targeting
to the SDV, silica biomineralization, bioinorganic
pattern formation, and exocytosis. A major bottleneck has been the inability to purify the SDV. However, a technique for fluorescently labeling the SDV
in vivo has been described using rhodamine 123,
which binds silica and which fluoresces brightly at
the acidic pH of the SDV (Li et al., 1989; Brzezinski
and Conley, 1994). It is possible that such labeling
combined with modern fractionation techniques
could be a useful approach. Furthermore, the use of
green fluorescent protein to fluorescently tag SDVdestined proteins such as frustulins could reveal the
1344
ACKNOWLEDGMENTS
We would like to thank Angela Falciatore for critical
reading of the manuscript and Ian Nettleton for the illustrations. We apologize that due to size restrictions it has
not been possible to discuss all the original work relating to
diatom cell wall formation.
Received August 10, 2001; returned for revision August 21,
2001; accepted August 27, 2001.
LITERATURE CITED
Apt KE, Kroth-Pancic PG, Grossman AR (1997) Mol Gen
Genet 252: 572–579
Baldauf SL, Roger AJ, Wenk-Siefert I, Doolittle WF (2000)
Science 290: 972–977
Downloaded from on June 15, 2017 - Published by www.plantphysiol.org
Copyright © 2001 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 127, 2001
Scientific Correspondence
Benedum UM, Lamouroux A, Konecki DS, Rosa P, Hille
A, Baeuerle PA, Frank R, Lottspeich F, Mallet J, Huttner
WB (1987) EMBO J 6: 1203–1211
Bhattacharya P, Volcani BE (1983) Biochem Biophys Res
Commun 114: 365–372
Brzezinski MA, Conley DJ (1994) J Phycol 30: 45–55
Brzezinski MA, Olson RJ, Chisholm SW (1990) Mar Ecol
Prog Ser 67: 83–96
Castle AM, Stahl LE, Castle JD (1992) J Biol Chem 267:
13093–13100
Chen YA, Scheller RH (2001) Nat Rev Mol Cell Biol 2:
98–106
Dunahay TG, Jarvis EE, Roessler PG (1995) J Phycol 31:
1004–1012
Falciatore A, Casotti R, Leblanc C, Abrescia C, Bowler C
(1999) Mar Biotechnol 1: 239–251
Falkowski PG, Barber RT, Smetacek V (1998) Science 281:
200–206
Fischer H, Robl I, Sumper M, Kröger N (1999) J Phycol 35:
113–120
Gordon R, Drum RW (1994) Int Rev Cytol 150: 243–372
Guo W, Tamanoi F, Novick P (2001) Nat Cell Biol 3:
353–360
Hildebrand M, Dahlin K, Volcani BE (1998) Mol Gen
Genet 260: 480–486
Hildebrand M, Volcani BE, Gassmann W, Schroeder JL
(1997) Nature 385: 688–689
Huttner WB, Gerdes HH, Rosa P (1991) Trends Biochem
Sci 16: 27–30
Kowallik KV (1992) Origin and Evolution of Plastids from
Chlorophyll-a⫹c-Containing Algae: Suggested Ancestral
Relationships to Red and Green Algal Plastids. Chapman
and Hall, New York
Kröger N, Bergsdorf C, Sumper M (1994) EMBO J 13:
4676–4683
Kröger N, Deutzmann R, Bergsdorf C, Sumper M (2000)
Proc Natl Acad Sci USA 97: 14133–14138
Kröger N, Deutzmann R, Sumper M (1999) Science 286:
1129–1132
Kröger N, Deutzmann R, Sumper M (2001) J Biol Chem
276: 26066–26070
Kröger N, Lehmann G, Rachel R, Sumper M (1997) Eur
J Biochem 250: 99–105
Kröger N, Sumper M (1998) Protist 149: 213–219
Kröger N, Wetherbee R (2000) Protist 151: 263–273
Leblanc C, Falciatore A, Bowler C (1999) Plant Mol Biol 40:
1031–1044
Lee RE (1999) In RE Lee, ed, Heterokontophyta, Bacillariophyceae. Cambridge University Press, Cambridge, UK,
pp 415–458
Li C-W, Chu S, Lee M (1989) Protoplasma 151: 158–163
Li C-W, Volcani BE (1985) Protoplasma 124: 10–29
Plant Physiol. Vol. 127, 2001
Lippincott-Schwartz J (1983) Trends Cell Biol 3: 81–88
Mann DG (1993) Hydrobiologia 269/270: 11–20
Mann S, Ozin GA (1996) Nature 382: 313–318
Martin-Jezequel V, Hildebrand M, Brzezinski MA (2000)
J Phycol 36: 821–840
Medlin LK, Kooistra WCH, Schmid A-MM (2000) In A
Witkowski, J Sieminska, ed, A Review of the Evolution of
the Diatoms: A Total Approach Using Molecules, Morphology and Geology. W. Szafer Institute of Botany,
Polish Academy of Sciences, Cracow, pp 13–35
Medlin LK, Kooistra WHCF, Gersonde R, Sims PA, Wellbrock U (1997) Nova Hedwigia 65: 1–11
Miller SG, Moore HP (1990) Curr Opin Cell Biol 2: 642–647
Morse DE (1999) Trends Biotechnol 17: 230–232
Mullins C, Bonifacino JS (2001) Bioessays 23: 333–343
Norton TA, Melkonian M, Andersen RA (1996) Phycologia 35: 308–326
Parkinson J, Gordon R (1999) Trends Biotechnol 17:
190–196
Pelham HR (1991) Cell 67: 449–451
Pelham HR (2001) Trends Cell Biol 11: 99–101
Pickett-Heaps J, Schmid A-MM, Edgar LA (1990) In FE
Round, DJ Chapman, eds, The Cell Biology of Diatom
Valve Formation, Vol 7. Biopress Ltd., Bristol, UK, pp
1–168
Pickett-Heaps JD, Schmid A-MM, Tippit DH (1984) Protoplasma 120: 132–154
Raven JA (1983) Biol Rev 58: 179–207
Rindler MJ (1998) J Biol Chem 273: 31180–31185
Scala S, Bowler C (2001) Cell Mol Life Sci 58: 1666–1673
Schmid A-MM (1994) Protoplasma 181: 43–60
Schmid A-MM, Schulz D (1979) Protoplasma 100: 267–288
Shimizu K, Cha J, Stucky GD, Morse DE (1998) Proc Natl
Acad Sci USA 95: 6234–6238
TerBush DR, Maurice TM, Roth D, Novick P (1996)
EMBO J 15: 6483–6494
Tooze SA (1998) Biochim Biophys Acta 1404: 231–244
Van Den Hoek C, Mann DG, Johns HM (1997) Algae: An
Introduction to Phycology. Cambridge University Press,
Cambridge, UK
van de Poll WH, Vrieling EG, Gieskes WWC (1999) J
Phycol 35: 1044–1053
von Figura K (1991) Curr Opin Cell Biol 3: 642–646
Vrieling EG, Gieskes WWC, Beelen TPM (1999) J Phycol
35: 548–559
Woodman PG (2000) Traffic 1: 695–701
Zahraoui A, Louvard D, Galli T (2000) J Cell Biol 151:
F31–F36
Zaslavskaia LA, Lippmeier JC, Kroth PG, Grossman AR,
Apt KE (2000) J Phycol 36: 379–386
Zerial M, McBride H (2001) Nat Rev Mol Cell Biol 2:
107–117
Downloaded from on June 15, 2017 - Published by www.plantphysiol.org
Copyright © 2001 American Society of Plant Biologists. All rights reserved.
1345