.
.This work
represents a continuation of a series. See references for parts
I-III.
ABSTRACT
.Extracellular
matrix (ECM) polymers secreted by the diatoms Achnanthes longipes
Ag. and Cymbella cistula (Ehr.) Kirchn. completely encase
the cell and are responsible for adhesion and other interactions
with the external environment. To preserve details of the highly
hydrophilic ECM in the native state, and to preserve, with a high
degree of fidelity, the intracellular structures involved in synthesis
of extracellular polymers, we applied a suite of cryo-techniques.
The methods included high-resolution visualization of surfaces
using cryo-field emission scanning electron microscopy (cryo-FESEM)
and preservation for TEM observation of thin sections by high
pressure freezing (HPF) and freeze substitution (FS). The extracellular
structures of diatoms plunge-frozen in liquid ethane, etched at
low temperature, and observed on a cryo-stage in the FESEM showed
overall dimensions and shapes closely comparable to those observed
with light microscopy. Cryo-FESEM demonstrated the pervasive nature
of the extracellular polymers and their importance in cell-substratum
and cell-cell associations and revealed details of cell attachment
processes not visible using other SEM techniques or light microscopy.
The layer of ECM coating the frustule and entirely encapsulating
cells of A. longipes and C. cistula
was shown to have a significant role in initial cell adhesion
and subsequent interaction with the environment. Trails of raphe-associated
ECM, generated during cell motility, were shown at high resolution
and consist of anastomoses of coiled and linear strands. Cryo-FESEM
revealed a sheet-like mucilage covering stalks.
HPF/FS
of A. longipes resulted in excellent preservation
of intra- and extra-cellular structures comparable to previous
reports for animals and higher plants, and revealed several organelles
not described previously. Three distinct vesicle types were identified,
including a class closely associated with Golgi bodies and postulated
to participate in formation of the extracellular adhesive structures.
HPF/FS showed a number of continuous diatotepic layers positioned
between the plasma membrane and the silicon frustule and revealed
that extracellular adhesive extrusion through frustule pores during
stalk production was closely related to the diatotepum. The stalks
of A. longipes consist of highly organized, multi-layered,
fine fibrillar materials with an electron dense layer organized
as a sheath at the stalk periphery.
.Key index
words: Achnanthes longipes; adhesives; Bacillariophyceae;
biofouling, cell motility; cryo-field-emission SEM; Cymbella
cistula; diatom; extracellular matrix; high-pressure freezing;
freeze substitution; secretion.
Abbreviations:
CF, chemical fixation; CPD, critical point drying; DIC, differential
interference contrast; DL, diatotepic layers; ECM, extracellular
matrix; FS, freeze substitution; FESEM, field emission scanning
electron microscopy; HPF, high pressure freezing.
Achnanthes
longipes and Cymbella cistula are common fouling diatoms
in marine and freshwater biofilms respectively (Hoagland et al.
1993). In these diatoms, attachment is a sequential process including
initial contact with substrata, reorientation and motility followed
by production of permanent adhesion structures including stalks
(Wang et al. 1997). Motility within a biofilm provides a number
of obvious and important ecological advantages (Hoagland et al.
1993), and several hypotheses concerning the mechanism of diatom
motility involve the synthesis of extracellular polymers (Gordon
and Drum 1970, Edgar and Pickett-Heaps 1984, Gordon 1987, Wetherbee
et al. 1998). Stalks elevate diatoms above the substratum and
confer advantages in competition for nutrients and light within
a biofilm (Johnson et al. 1995). Stalks of A. longipes
and C. cistula usually consist of three regions: a surface-adhered
pad, a collar associated with the frustule at the terminal nodules
(A. longipes) or apical pore field (C. cistula),
and an intervening shaft (Hufford and Collins 1972, Daniel et
al. 1987). Microscopical observations of C. cistula revealed
that motile cells, coated by an organic sheath, eventually produced
one or two stalks from apical pore fields. Ultrastructural study
of the marine fouling diatom Achnanthes subsessilis Kütz
suggested that stalk polymers are secreted from the raphe and
organized into a four layered shaft with a mucilaginous collar
at the valve/shaft interface (Blunn and Evans 1981).
We are
interested in delineating the intra- and extra-cellular events
involved in diatom ECM synthesis, secretion and assembly at the
ultrastructural level. The stalks of A. longipes
and C. cistula are particularly difficult to preserve for
electron microscopic observation due to the high level of hydration
of these structures and the ease with which they can be extracted
with common fixatives and dehydrating agents. Observations of
extracellular structures and intracellular events involved in
their biogenesis are limited by relatively poor preservation with
traditional TEM chemical fixation protocols, artifacts produced
during fixation, dehydration and critical point drying (CPD) for
SEM, and resolution limitations (Edgar and Pickett-Heaps 1982,
Rosowski et al. 1986). High pressure freezing (HPF), in conjunction
with freeze-substitution (FS), has been successfully applied to
plant, fungal and animal cells (Gilkey and Staehelin 1986, Kiss
and Staehelin 1995). Specimen preparation techniques involving
rapid cryo-immobilization and observation at low temperature with
field emission (FE) SEM have been shown to preserve native biological
structure (Chen et al. 1995 and citations within) and include
fast-freezing, freeze-fracture, freeze-sublimation, cryo-coating,
cryo-transfer and cryo-FESEM observation with specimens maintained
at low temperature throughout. Herein, we describe results from
application of high-resolution FESEM at low accelerating voltages
on quick-frozen specimens to study extracellular polymers at the
cell surface of A. longipes and C. cistula. In order
to gain more reliable structural information about diatom cells
and extracellular adhesives, we have applied HPF/FS methods to
dissect dynamic cellular events and reveal several novel features
related to A. longipes extracellular polymer secretion.
Cryo-FESEM and HPF/FS TEM have allowed us to gain a better understanding
of the interaction of components and assembly of the ECM and how
this organization affects the physical characteristics of A.
longipes and C. cistula stalks and associated extracellular
structures.
MATERIALS
AND METHODS
Algal
culture and light microscopy. Achnanthes longipes Ag.
(NEIS #330) was cleaned to minimize bacterial contamination and
cultured as described in Wang et al. (1997). Cymbella cistula
(Ehr.) Kirchn. was provided by Dr. Kyle Hoagland and cultured
as described in Wustman et al. (1997). Samples were prepared for
and DIC microscopy was conducted as described in Wang et al. (1997).
Cryo-FESEM.
Formvar-coated gold EM grids mounted on a cover slip were sterilized
under UV light before inoculation. Culture medium (10 mL) containing
clean diatom cells (105 cells mL-1) was
added to the petri dish and cells were observed with light microscopy
for 2 to 24 h under standard culture conditions described above.
Unattached cells were removed with culture media washes prior
to plunge freezing. Sample processing for cryo-FESEM was as in
Chen et al. (1995) as described briefly below. Diatoms attached
to formvar-coated grids were washed with double-distilled water
for 1 s (to minimize salt crystal formation) immediately prior
to plunge freezing in liquid ethane (-190o C) cooled
by liquid nitrogen (LN2). Samples were transferred
under LN2 to a cryo-transfer stage (Model 626, Gatan,
Warrendale, PA). The cryo-stage with specimen was then inserted
into a planar magnetron sputtering device (MED 010, Bal-Tec, Maddlebury,
CT) against a counter flow of dry nitrogen gas. To partially etch
the sample, sublimation was carried out at - 85o C
for variable times up to 30 min. Sputter-coating was performed
under 2 Pa of high purity (>99.99%) argon gas. The 2.5-nm thickness
of chromium was monitored with a quartz thickness monitor (Sycon
100, Sycon, East Syracuse, NY). Diatoms on the Gatan cryo-stage
were transferred under LN2 to the microscope and maintained
at -110o C during observation. The Hitachi S-900 field-emission
scanning electron microscope (FESEM) used was an "in-lens"
type operated at an accelerating voltage of 1.5kV and as described
in Chen et al. (1995) . Images were acquired digitally using Digital
Micrograph (Gatan) and contrast and brightness were adjusted using
Adobe Photoshop (Adobe Systems, Inc., Mountain View, CA).
Chemical
fixation and critical point drying (CF/CPD) for FESEM.
Cells were cleaned and inoculated as described above and diatoms
attached to cover slips were treated with 2% glutaraldehyde for
6 h, washed with seawater (2x) and treated with 0.1% osmium for
10 min. Fixed cells were washed with seawater (3x), put through
a graded series to 100% distilled water, dehydrated with an acetone
series [50%, 75%, 85%, 95% (10 min each) followed by 100% acetone
for 10 min (2x)], and critical point dried (CPD) using optimal
procedures to minimize surface distortion (Ris, 1985). Samples
were coated with a thin layer of Pt deposited by Argon ion-beam
sputtering and viewed at room temperature by FESEM as described
above.
Rapid
cryo-immobilization by high-pressure freezing (HPF). A.
longipes was cultured as described in Wang et al. (1997).
Cells actively attaching and producing ECM were selected and concentrated
by low speed centrifugation. A drop (1 µL) of concentrated
cell suspension was pipetted onto a gold planchette (diameter
3 mm, central cavity about 0.3 mm, Bal-Tec) and frozen in a Balzers
HPM 010 high pressure freezing apparatus at 2100 bar. The frozen
samples were stored in LN2 prior to initiation of freeze
substitution.
Freeze
substitution (FS), infiltration and embedding. HPF frozen
specimens were transferred under LN2 to 1.8 mL cryogenic
vials (Vangard International, Neptune, N. J., U. S. A) filled
with 1% (w/v) osmium tetroxide in sieve-dried acetone (-80o
C) as a substitution medium. Substitution was carried out
at -80o C in a dry ice/alcohol bath for 3 d followed
by 8 h at -20o C, 1 h at 0o C and 1 h at
room temperature. The sample was repeatedly rinsed in fresh sieve-dried
acetone until the acetone appeared clear. The following infiltration
schedule was used: 1:3 Spurr:acetone overnight; 1:1 and 3:1 Spurr:acetone
each for 4 h, and; 100% Spurr resin for 48 h with a change of
resin after 24 h. The Spurr-embedded sample was polymerized at
60o C for 24 h. To improve the contrast of freeze substituted
samples, ultrathin sections were stained with 2% (w/v) uranyl
acetate and then with 1% (w/v) lead citrate each for 10 min. Sections
were observed at 60 kV in a JEOL 100CX STEM (Wang et al. 1997).
Chemical
fixation (CF) and processing for TEM. A. longipes cells
actively attaching and produing ECM were selected and chemically
fixed with glutaraldehyde and osmium, dehydrated in an ethanol
series, infiltrated and embedded in Spurr resin and sectioned
as described in Wang et al. (1997).
RESULTS
.Field
Emission Scanning Electron Microscopy
Achnanthes
longipes – Cells inoculated onto formvar coated grids
performed the adhesion sequence as described in Wang et al. (1997),
including initial association with the substratum in random orientations,
a period of motility followed by cessation of movement, pad production
and shaft formation. Cryo-FESEM preserved ECM coatings on cells
and substrata and revealed detailed substructure of the extracellular
adhesives. In contrast, conventional SEM sample preparation techniques,
consisting of fixation, dehydration and CPD, produced a marked
absence of extracellular polymers and/or significant modification
of retained structures. When motile cells (2 h following inoculation
and observed by light microscopy) were examined with cryo-FESEM,
a ribbon-like material was observed in close association with
the raphe valve and the substratum. This "trail" appeared
variously as tightly coiled strands of varying sizes, alternating
and interconnected with linear bands (Figs.
1, 2). Cryo-FESEM of the cells and associated adhesive structures
compared closely with high-resolution DIC images, showing similar
size and appearance of shaft, collar and pad (Figs.
3, 4) and revealing a distinct, undifferentiated, smooth ECM
layer fully encapsulating the silicon frustule (cover photo).
The shaft diameter is 7-10 µm when calculated from either
cryo-FESEM or DIC. A surface film was found on the substratum
6-24 h following inoculation and was interrupted by cleared "paths"
terminating in stalk pads (Fig.
3). CF/CPD resulted in loss of substratum surface film and
in "clean" frustules with little polymer evident on
the cell surface (data not shown).
Cryo-FESEM
revealed an expanded, net-like mucilaginous collar covering the
frustule at the point of shaft emergence (Fig.
5, and cover
photo). In cells preserved by CF/CPD, the three-dimensional
appearance of the collar was lost as it was collapsed into a thin
layer that was closely associated with the frustule at the raphe
terminus (Fig.
6). Cryo-FESEM of shaft surfaces indicated a sheath covering
the shaft (Fig.
7) whereas CF/CPD resulted in a textured shaft surface interrupted
by cracks that appeared to be drying artifacts (Fig.
8) and a reduction of shaft diameter to ~ 4 µm. Details
of the substratum attachment site revealed by cryo-FESEM showed
cells attached to the substrate via a pad structure composed of
fluffy, fibrillar materials (Fig.
9), in contrast to condensed pad material attached at substrata
observed in CF/CPD prepared samples (Fig.
10).
Cymbella
cistula - Alcian staining of this freshwater diatom showed
the typical extracellular attachment structures including pads
and stalks (Fig.
11). Cryo-FESEM showed the surface of the band- or ribbon-like
stalk, which was 1.5 - 2 µm in diameter (matching DIC observations),
covered with a tightly associated ECM layer (Fig.
12) continuous over the cell and stalk surface (Figs.
12-14). Depending on the length of sublimation time (see Materials
and Methods), this coating varied from a relatively smooth, continuous
sheet to fibrillar strands representing aggregation of polymers
from dried mucilage (Figs.
12-14). This coating was not observed in CF/CPD treated cells
(Fig.
15). Marked distortion and shrinkage of stalks was observed
after CF/CPD treatment. Discrete collars are not shown, although
they were occasionally observed.
.Transmission
Electron Microscopy of A. longipes
Excellent
ultrastructural preservation was achieved for the vast majority
of cells by high pressure freezing in seawater, followed by freeze
substitution. Observation of HPF preserved cells in freeze fractured
preparations confirmed the effectiveness of this technique (data
not shown). The ultrastructure of HPF/FS diatom cells differed
from that of conventionally prepared cells in several significant
aspects as shown in Figs. 16 and 17. Cells prepared by HPF/FS
had smooth membranes and more turgid membrane-bound compartments
(Fig.
16), whereas large "storage inclusions" (membrane-less
osmiophilic bodies) predominated in samples prepared with CF-TEM
(Fig.
17). HPF/FS revealed a ground cytoplasm with a regular, evenly
distributed matrix, H-shaped, lobed chloroplasts which appeared
turgid, with smooth surface contours and an amorphous dense stroma
and tubular type mitochondria present throughout the cytoplasm
(Fig.
16). In HPF/FS, the nucleus, suspended in the center
of the cell, was spherical, the nuclear envelope appeared smooth
and was traversed with regularly spaced nuclear pores with distinct
substructure (Figs.
16, 18).
There were at least ten Golgi bodies in each cell, always located
in a perinuclear position and consisting of just a few cisternae.
Membranes of the distal cisternae appeared coated and were associated
with tubular networks and vesicles at the cisternal periphery
were filled with electron dense products (Fig.
18).
Three distinct
vesicle types containing electron dense materials were observed
in HPF/FS cells (Figs.
16, 19).
Vesicle type one (v1) occurred only in the immediate vicinity
of the nucleus and Golgi (Figs.
16, 19)
and their contents appeared similar to some ECM deposits. Irregular,
tubular, membrane delimited extensions of v1 were present, usually
with greater electron density than the main portion (Fig.
19). The second vesicle type (v2) was spherical, electron
dense, variously sized and contained dark granular contents delimited
by a dense membrane (Figs.
16, 19).
Vesicle v2 was distributed throughout the cytoplasm (Fig.
16). Type three (v3) was electron translucent and usually
associated with the chloroplast and the plasma membrane (Fig.
16). In contrast, in CF-TEM, none of above vesicles were discernable
(Fig.
17).
In HPF/FS,
the layer between the PM and the frustule, the diatotepum, totally
encompassed the protoplasm, appeared thick and multi-laminate,
and was usually closely appressed to the frustule (Figs.
16, 20,
21). The diatotepum consisted of three regions, an electron
translucent area adjacent to the PM, a thicker, more electron
dense region and a distal electron translucent zone (Figs.
20, 21). There were no observed discontinuities in the diatotepum
which would allow direct protoplasmic contact with the frustule
or external spaces (Figs.
16, 20,
21). An electron transparent region was consistently found
between the diatotepum and PM in the area of ECM secretion (Figs.
20, 21). Fibrillar material appeared to be directly derived
from the most distal layer of the diatotepum, transversed frustule
pores and was continuous with stalks (Fig.
20). In addition, diatotepic associated mucilage was seen
in the raphe canal (Fig.
21). In CF-TEM, the diatotepum was thinner, appeared discontinuous
in some areas and both the PM and the diatotepum were undulate
and separated from the frustule (Fig.
17).
Stalk structure
in HPF/FS (Fig.
22) was distinct from that observed in CF-TEM prepared cells
(Fig.
23). Fibril diameter appeared much smaller and the fibril
distribution was more homogeneous in HPF/FS (Fig.
22). The overall diameter of the stalk in HFS/FS was approximately
8 µm (matching cryo-FESEM and DIC data), whereas, it was
reduced to approximately 5 µm in CF-TEM. HPF/FS TEM revealed
a highly dense layer of crystallized ECM distributed along the
stalk periphery (Fig.
22).
DISCUSSION
The presence
of a smooth coating encompassing the frustule of A. longipes
and C. cistula was obvious with cryo-FESEM, although CF/CPD
removed this layer or altered it extensively. Although significant
frustule coatings have been visualized by CF/CPD for some diatoms
in certain instances (Hoagland et al. 1993), coatings may have
been overlooked in CF/CPD observations and their significance
as mediators of cell-cell, cell-substratum and cell-external environment
interactions underestimated. Recent work on diatom adhesion mechanisms
has shown the significance of this coating layer in passive attachment/first
contact adhesion, especially on hydrophobic surfaces. Investigations
of adhesion in Stauroneis decipiens Hustedt revealed that
cells first contact the substratum via the girdle region of the
frustule (Lind et al. 1997). The cells were slightly attached
at this point (Wetherbee et al. 1998) and this initial interaction
was mediated by the frustule or frustule coatings. The first association
of A. longipes with surfaces was a random orientation frustule-substratum
contact (Wang et al. 1997) and this association is mediated by
the frustular coating. In both cases, passive attachment was followed
by more permanent adhesion fostered by ECM polymers generated
from the raphe valve(s) and involved in motility. Frustule coatings
are also involved in cell-cell adhesion in A. longipes,
where cells may remain attached to one another following several
mitotic events, forming long chains of cells (Wang 1995, Wang
et al. 1997). These intercellular adhesive ECM components have
been localized with alcian staining, and lectins and mono- and
poly-clonal antibodies that co-localize stalk components (Wustman
et al. 1997,1998) indicating that, in A. longipes, the
primary components of frustular coatings are not SDV and PM remnants
from silicification processes. An additional ECM component is
produced during cell dispersion, when motility is initiated and
cells detach from the stacks (Wustman et al. 1997, 1998). The
source of frustule coatings in A. longipes and C. cistula
is not known, although cryo-FESEM also showed (cover
photo) a previously described ECM component that escaped from
between girdle bands, expanded rapidly and labeled with a monoclonal
antibody to A. longipes adhesives (Wustman et al. 1998).
When observed
with cryo-FESEM after 2-6 h incubation, ECM "trails"
were found in close association with A. longipes cells
observed to be motile immediately prior to freezing. Light microscopic
observations of A. longipes indicated cable-like trails
of material which accumulated behind motile cells and were attached
to the surface at irregularly spaced intervals (Wang et al. 1997,
Wustman et al. 1998). Video microscopy showed elastic characteristics
of this trail during cell motility (data not shown). High resolution
cryo-FESEM provides detailed "snapshots" of these trails
generated during cell motility in A. longipes. Most interestingly,
individual trails are composed of multiple discrete strands that
are either coiled about one another or span long stretches in
a linear format. Lind et al. (1997) reported raphe-derived ECM
mediated active adhesion in S. decipiens, and used video
microscopy to describe adhesive trails associated with the diatoms
as they glided over a substratum. Models of raphe-associated motility
consistently invoke extrusion of ECM components as a part of the
motive force (Edgar and Pickett-Heaps 1984, Gordon 1987, Cohn
and Disparti 1994, Wetherbee et al. 1998) with the observation
of trails a common, although not consistently observed, feature.
Part of the reason for this may be the lability of the trails,
as they have been reported to be highly soluble and dispersed
a short time after deposition and during traditional fixation
for electron microscopy (Edgar and Pickett-Heaps 1982, Edgar 1983,
Edgar and Pickett-Heaps 1984, Hoagland et al. 1993). ECM associated
with motility has been shown to be critical in epipelic diatom
migration and survival (Smith and Underwood 1998). We are currently
performing detailed evaluations of trail structure correlated
with the various motility modes we have observed in A. longipes,
in an attempt to discern structural relationships between trail
morphology and observed diatom behavior. Six hours following inoculation,
cryo-FESEM revealed an extensive reticulum of material that appeared
to be composed of trails and other components distributed across
the areas traversed by diatoms during the initial adhesion events.
The fact that this matrix is absent from CF/CPD prepared material
suggests that this contribution to biofilm formation may have
been underestimated or overlooked in previous investigations of
establishment of the biofilm consortium.
Permanent
adhesion in the diatoms was achieved by stalk production (Wang
et al. 1997). The preservation of stalks of A. longipes
and C. cistula was facilitated by cryo-FESEM and the technique
resulted in structures that correlated very well with that observed
in living cultures with high resolution DIC. High-resolution cryo-FESEM
of A. longipes indicated that the pad and collar consist
of fine, irregularly oriented fibrils and supported previous cytochemical
staining, lectin labeling, and DIC observations that they originated
from the same material (Daniel et al. 1987, Wang et al. 1997,
Wustman et al. 1997). Cryo-SEM reveals a sheet-like mucilage covering
stalks of A. longipes and C. cistula in a more directly
convincing manner than previously achieved (Hufford and Collins
1972, Hasle 1974). This sheath obscures the fibrillar nature of
the underlying stalk observed in TEM studies. The correspondence
of features of the diatom stalk surface to the results from HPF/FS
TEM suggests that these represent "true" features of
stalk surfaces. Daniel et al. (1987) and Wang et al. (1997) investigated
stalk strcuture in A. longipes using CF-TEM and reported
the arrangement of fibrils within the various layers of the stalk.
In the current study, using HPF/FS sample preparation techniques,
we found the same basic fibrillar organization pattern as reported
previously. However, HPF/FS did yield several unique observations,
including: 1) the stalk diameter is more consistent with results
of cryo-FESEM and DIC analysis; 2) the constituent fibrils are
of smaller diameter, with sub-patterning revealed, and; 3) a layer
of electron dense material consistently appears as a sheath at
the stalk periphery, which was not visible using CF. The A.
longipes shaft sheath may be closely related to mucilage secreted
during cell motility, based on lectin and antibody-FITC labeling
of an amorphous mucilage which covered the collar, shaft and pad
(Wustman et al. 1997, 1998).
Diatom
stalks have been reported to be an anastomosis of extremely fine
fibrils extruded through the siliceous cell wall (Drum 1969, de
Francisco and Roth 1977, Gibson 1979, Blunn and Evans 1981, Daniel
et al. 1987). HPF/FS preservation of A. longipes revealed
fine fibrillar material derived from the outer layers of the diatotepum
which passed through frustule pores and was integrated into the
stalk. We found a multi-laminate diatotepum that was closely associated
with the frustule and the PM and that appeared much more voluminous
than in CF treated specimens, where a very condensed, discontinuous
diatotepic layer was usually widely separated from the frustule
and PM. A-M. Schmidt (personal communication) has also observed
a diatotepum composed of several layers in A. longipes.
In A. subsessilis (Blunn and Evans 1981), the diatotepum
is also laminated, which contrasts with the homogeneous appearance
of this layer in Nitzachia alba (Kütz) W. Smith (Lauritis
et al. 1968), Gomphonema parvulum Kütz. (Dawson
1973) and in Amphipleura pellucida Kütz (Stoermer
et al. 1965). Cytochemistry of the diatotepum indicates acidic
polysaccharides and the only chemical analysis conducted on Phareodactylum
tricornutum Bohlin yielded a sulfated glucuronomannan (Ford
and Percival 1965). Schmidt (1994) describes the deposition of
this layer from dense vesicles (dv) following the completion of
frustule formation, effectively sealing the siliceous chambers
from the cytoplasm in Coscinodiscus wailessi Gran and Augst.
and A. longipes. The diatotepum of A. longipes encapsulates
the protoplast without apparent discontinuity, which contrasts
with isolated reports of diatotepic layers with openings at the
labiate and strutted processes, raphe, and ocelli in other diatoms
(von Stosch 1981, Schmid 1994). Our observation of continuous
diatotepic layers in the area of stalk extrusion and derivation
of stalk fibrils from this layer provided evidence for direct
participation of this layer in stalk biogenesis. Although in stalk
producing cells we found the diatotepum continuous across the
raphe, with diatotepum associated material present in the raphe
canal (Fig.
21), we have not critically examined motile cells to determine
a potential role of the diatotepum in cell motility. We are currently
examining cells confirmed as actively motile by light microscopy
immediately before freezing with special reference to their diatotepic
layers. The present of a continuous diatotepic layer over the
raphe during cell motility would have profound implications on
current diatom motility models which suggest direct cytoskeletal
mediated movement of ECM within the raphe canal (see Hoagland
et al. 1993, Wetherbee et al. 1998 for reviews).
Large numbers
of perinuclear Golgi bodies were observed in HPF/FS A. longipes
cells, correlating with reports of Drum et al. (1966) of up to
twenty Golgi in one ultrathin section of A. longipes. Golgi
derived vesicles have been implicated in mucilage secretion for
cell locomotion (Drum and Hopkins 1966, Edgar and Pickett-Heaps
1982, Edgar and Pickett-Heaps 1984), and in formation of the mucilage
sheath or stalk when present. In A. subsessilis, vesicles
with fibrillar contents, resembling stalk material, were believed
to be derived from the maturing (trans) face of the Golgi bodies
(Blunn and Evans 1981). It is logical to assume that, in A.
longipes, the appearance of reticular vesicles or individual
small vesicles associated with the Golgi bodies may be part of
a transport system for ECM components.
Daniel
et al. (1980) reported irregularly shaped vesicles (v1) throughout
the peripheral and central cytoplasm of Amphora veneta
Kütz that contained dispersed fibrous material similar in
morphology to that of the ECM and postulated to be part of the
ECM secretion pathway. In A. longipes, vesicle (v1) contained
material that appeared similar to that of ECM components, were
primarily distributed in the perinuclear space and were often
irregular in outline. Different vesicles types from G.
parvulum Kütz., Encyonema caespitosum
Kütz., Diatoma tenue var. elongatum and Encyonema
ventricosum were hypothesized to be involved in mucilage-stalk
production and secretion (Crawford 1973, Dawson 1973, Daniel et
al. 1980, Sicko-Goad et al. 1989). Attempts to match vesicle types
observed in HPF/FS A. longipes with those described in
previous reports, based on morphology and distribution, were generally
not successful. Although the structure and distribution of v1
in A. longipes suggests a function in extracellular
adhesive production and secretion, we expect results of monoclonal
antibody (Wustman et al. 1998) immuno-gold localization studies
to define the role of v1 in ECM production in A. longipes.
To our
knowledge, this is the first ultrastructural study of diatoms
using HPF/FS techniques. Application of these procedures to A.
longipes gave major improvements in ultrastructural preservation
and allowed us to distinguish new intra- and extra-cellular structures
not visible in conventionally fixed and dehydrated diatom cells.
This study also demonstrates the utility and power of cryo-FESEM
for high-resolution observation and characterization of extracellular
adhesives in diatoms. Plunge freezing in liquid ethane provided
vitrification (freezing with lack of ice crystal formation) of
A. longipes and C. cistula samples to a depth sufficient
to preserve the structure of the diatom cells and associated extracellular
structures. By monitoring and controlling the process, we were
able to achieve sublimation of "unbound" water from
the specimen, while maintaining hydration in structures that contained
something less labile than "pure" water (i.e. ECM polymers).
Sublimation was immediately followed by deposition of a thin film
(2.5 nm) of chromium by planar-magnetron sputtering, resulting
in even distribution and fine grain that gave increased signal
without obscuring features on sample surfaces (Centonze et al.
1995). Although we have taken advantage of the high degree of
fidelity resulting from sample preservation and observation using
these techniques, we are just now beginning detailed, high resolution
analysis of the macromolecular structure of the adhesive polymers
and freeze-fracture imaging using related technologies.
SUMMARY
In this
study, we have refined our understanding of the ultrastructure
of the diatoms A. longipes and C. cistula, especially
as related to extracellular adhesive structure and function, using
state-of-the-art electron microscopy techniques. Our comparative
approach reveals several novel aspects of diatom ultrastructure
heretofore undescribed. High resolution cryo-FESEM demonstrated
ECM components which were lost during processing for CF/CPD SEM.
We observed a thin, undifferentiated ECM frustule coating associated
with A. longipes and C. cistula that serves as a
mediator for initial contact and transient adhesion to substrata
in both marine and freshwater environments. Cryo-FESEM and HPF/FS
also revealed an amorphous sheath of ECM encompassing stalks of
A. longipes and C. cistula. When imaged at high
resolution following cryo-preservation, the trails resulting from
raphe valve associated diatom movement across substrata appear
as an anastomoses of coiled and linear strands.
HPF/FS
demonstrated novel intra- and extra-cellular features of A.
longipes and produced ultrastructure with better definition
and order and with fewer extraction artifacts. A. longipes
Golgi architecture is comparable to well defined results in higher
plants and three distinct types of intracellular vesicles, v1,
2, 3 have been identified, with v1 postulated to participate in
ECM production. A characteristic diatotepum, tightly associated
with the PM and appressed to the frustule in A. longipes,
consists of multiple layers. The diatotepum continuously encompasses
the protoplast and ECM such as stalks appear to be derived from
this structure and extruded through frustule pores.
This first
application of cryo-FESEM and HPF/FS to diatoms demonstrates the
unique capabilities and indicates the value and potential uses
of these techniques in the study of diatom ultrastructure. Both
cryo-FESEM and HPF/FS TEM provide new insights into the processes
involved in diatom settlement and adhesion, and thereby create
more reliable structural information for studying production,
secretion and assembly of extracellular adhesives. Accurate structural
characterization of the ECM as revealed by high resolution cryo-FESEM
and correlation with LM and HPF/FS will greatly assist in appreciation
of the function and importance of these extracellular polymers
in cell-substratum and cell-cell interactions, motility and stalk
production in marine and freshwater diatoms.
.This research
was supported by Office of Naval Research Grants N00014-91-J-1108,
N000014-94-1-0273 and N00014-94-1-0766, National Institutes of
Health Grant RR 00570, and a MTU Research Excellence Fund award.
This work represents a portion of a dissertation by Y.W. to be
submitted in partial fulfillment of the requirements for the Ph.D.
degree from Biological Sciences at MTU. We thank Dr. Jean-Claude
Mollet and Mr. Owen Mills for expert assistance with fixation
protocols and TEM operation and Prof. Kyle Hoagland for providing
Cymbella cistula. Special appreciation is accorded Prof.
Hans Ris for optimization of critical point drying protocols.
.REFERENCES
Blunn,
G.W. and Evans, L.V. 1981. Microscopical observations on Achnanthes
subsessilis, with particular reference to stalk formation.
Bot. Mar. 24:193-9.
Centonze,
V.E., Chen, Y., Severson, T.F., Borisy, G.G. & Nibert, M.L.
1995. Visualization of individual reovirus particles by low-temperature,
high-resolution scanning electron microscopy. J. Struct.
Biol. 115:215-25.
Chen, Y.,
Centonze, V.E., Verkhovsky, A. & Borisy, G.G. 1995. Imaging
of cytoskeletal elements by low-temperature high-resolution scanning
electron microscopy. J. Microsc. 179:67-76.
Cohn, S.A.
& Disparti, N.C. 1994. Environmental factors influencing diatom
cell motility. J. Phycol. 30:818-28.
Crawford,
R.M. 1973. The protoplasmic ultrastructure of the vegetative cell
of Melosira varians C. A. Agardh. J. Phycol.
9:50-61.
Daniel,
G.F., Chamberlain, A.H.L. & Jones, E.B.G. 1980. Ultrastructural
observations on the marine fouling diatom Amphora. Helgol.
Wiss. Meeresunters 34:123-49.
Daniel,
G.F., Chamberlain, A.H.L. & Jones, E.B.G. 1987. Cytochemical
and electron microscopical observations on the adhesive materials
of marine fouling diatoms. Br. Phycol. J. 22:101-18.
Dawson,
P.A. 1973. Observations on the structure of some forms of Gomphonema
parvulum Kutz. II. The internal organization. J.
Phycol. 9:165-75.
de Francisco,
A. & Roth, L.E. 1977. The marine diatom, Striatella unipunctata.
I. Cytoplasmic fine structure with emphasis on the Golgi apparatus.
Cytobiologie 14:191-206.
Drum, R.W.
1969. Electron microscope observation of diatoms. Osterr.
Bot.J. 116:321-30.
Drum, R.W.
& Hopkins, J.T. 1966. Diatom locomotion: an explanation.
Protoplasma 62:1-33.
Drum, R.W.,
Pankratz, H.S. & Stoermer, E.F. 1966. Electron microscopy
of diatom cells. In Diatomeenschalen im elektronenmikroskopischen
Bild, Vol. 6. Helmke, J.G. & Kriego, W. [Eds.], J. Cramer
-Verlag, Lehre
Edgar,
L.A. 1983. Mucilage secretions of moving diatoms. Protoplasma
118:44-8.
Edgar,
L.A. & Pickett-Heaps, J.D. 1982. Ultrastructural localization
of polysaccharides in the motile diatom Navicula cuspidata.
Protoplasma 113:10-22.
Edgar,
L.A. & Pickett-Heaps, J.D. 1984. Diatom locomotion. Prog.
Phycol. Res. 3:47-88.
Ford, C.W.
& Percival, E. 1965. Carbohydrates of Phaeodactylum tricornutum.
Part II. A sulphated glucuronomannan. J. Chem. Soc.
7042-6.
Gibson,
R. 1979. An ultrastructure study of Pseudohimantidium pacificum
Hust. & Krasske (Bacillariophyceae: protoraphidaceae) with
special reference to the labiate processes. Nova Hedwigia
64:147-60.
Gilkey,
J.C. & Staehelin, L.A. 1986. Advances in ultrarapid freezing
for the preservation of cellular ultrastructure. J.
Electron Microsc. Tech. 3:177-210.
Gordon,
R & Drum, R. W. 1970. A capillarity mechanism for diatom gliding
locomotion. Proc. Nat. Acad. Sci. 67: 338-44.
Gordon,
R. 1987. A retaliatory role for algal projectiles, with implications
for the mechanochemistry of diatom gliding motility. J.
Theor. Biol. 126:419-36.
Hasle,
G.R. 1974. The "mucilage pore" of pennate diatoms.
Beih.Nova Hedwigia 45:167-86.
Hoagland,
K.D., Rosowski, J.R., Gretz, M.R. & Roemer, S.C. 1993. Diatom
extracellular polymeric substances: function, fine structure,
chemistry, and physiology. J. Phycol. 29:537-66.
Hufford,
T.L. & Collins, G.B. 1972. The stalk of the diatom Cymbella
cistula: SEM observations. J. Phycol. 8(2):208-10.
Johnson,
L.M., Hoagland, K.D. & Gretz, M.R. 1995. Effects of bromide
and iodide on stalk secretion in the biofouling diatom Achnanthes
longipes (Bacillariophyceae). J. Phycol. 31:401-12.
Kiss, J.Z.
& Staehelin, L.A. 1995. High pressure freezing. In
Severs, N.J. & Shotton, D.M. [Eds.] Rapid Freezing, Freeze
Fracture and Deep Etching. Wiley-Liss, Inc. New York, pp.
89-104.
Lauritis,
J.A., Coombs, J. & Volcani, B.E. 1968. Studies on the biochemistry
and fine structure of silica shell formation in diatoms IV. Fine
structure of the apochlorotic diatom Nitzschia alba Lewin
and Lewin. Arch. Mikrobiol. 62:1-16.
Lind, J.L.,
Heimann, K., Miller, E.A., van Vliet, C., Hoogenraad, N.J. &
Wetherbee, R. 1997. Substratum adhesion and gliding in a diatom
are mediated by extracellular proteoglycans. Planta
203:213-21.
Ris, H.
1985. The cytoplasmic filament system in critical point-dried
whole mounts and plastic-embedded sections. J. Cell
Biol. 100:1474-87.
Rosowski,
J.R., Hoagland, K.D. & Aloi, J.E. 1986. Structural morphology
of diatom-dominated stream biofilm communities under the impact
of soil erosion. In Evans, L.V. & Hoagland, K.D. [Eds.]
Algal Biofouling. Elsevier, Amsterdam, pp. 247-99.
Round,
F.E., Crawford, R.M. & Mann, D.G. 1990. The Diatoms: Biology
& Morphology of the Genera. Cambridge University Press,
New York, pp. 1-747.
Schmid,
A.M. 1994. Aspects of morphogenesis and function of diatom cell
walls with implications for taxonomy. Protoplasma
181:43-60.
Sicko-Goad,
L., Kociolek, J.P. & Stoermer, E.F. 1989. Patterns of mucilage
production and secretion in pennate diatoms. In Bailey,
G.W. [Ed.] Proceedings of the 47th Annual Meeting of the Electron
Microscopy Society of America. San Fransisco Press,Inc., San
Fransisco, pp. 1016-7.
Smith,
D.J. & Underwood, G.J.C. 1998. Exopolymer production by intertidal
epipelic diatoms. Limnol. Oceanogr. 43:1578-91.
Stoermer,
E.F., Pankratz, H.S. & Bowen, C. 1965. Fine structure of the
diatom Amphipleura pellucida II. Cytoplasmic fine structure
and frustule formation. Amer. J. Bot. 52(10):1067-78.
von Stosch,
H-A. 1981. Structural and histochemical observations on the organic
layers of the diatom cell wall. Proc. Diatom Symp.
6:231-51.
Wang, Y.
1995. A study of adhesion mechanisms of the marine biofouling
alga, Achnanthes longipes (Bacillariophyceae). M.S.
thesis. Michigan Technological University. Houghton. pp. 1-73.
Wang, Y.,
Lu, J., Mollet, J.-C., Gretz, M.R. & Hoagland, K.D. 1997.
Extracellular matrix assembly in diatoms (Bacillariophyceae).
II. 2,6-dichlorobenzonitrile inhibition of motility and stalk
production in the marine diatom Achnanthes longipes. Plant
Physiol. 113:1071-80.
Wetherbee,
R., Lind, J.L., Burke, J. & Quatrano, R.S. 1998. The first
kiss: establishment and control of initial adhesion by raphid
diatoms. J. Phycol. 34:9-15.
Wustman,
B.A., Gretz, M.R. & Hoagland, K.D. 1997. Extracellular matrix
assembly in diatoms (Bacillariophyceae). I. A model of adhesives
based on chemical characterization and localization of polysaccharides
from the marine diatom Achnanthes longipes and other diatoms.
Plant Physiol. 113:1059-69.
Wustman,
B.A., Lind, J.L., Wetherbee, R. & Gretz, M.R. 1998. Extracellular
matrix assembly in diatoms (Bacillariophyceae). III. Organization
of fucoglucuronogalactans within the adhesive stalks of Achnanthes
longipes. Plant Physiol. 116:1431-41.
Wustman,
B.A. 1998. Extracellular matrix assembly in diatoms. Ph.D. dissertation.
Michigan Technological University, Houghton. pp. 1-227.
.Figure
Legends
.Figs.
1-4. Achnanthes longipes cell and extracellular adhesives.
Fig. 1. Cryo-FESEM of mucilage trails generated during cell motility
showed interconnected, coiled and linear bands. F= frustule. Note
that very little surface film was deposited on the surface during
the first 2 h after inoculation. Scale bar = 2 µm. Fig.
2. Magnified trail material. Scale bar = 0.5 µm. Fig. 3.
Stalk composed of collar, shaft and pad viewed with cryo-FESEM.
Note the surface film (SF) deposited on the substratum and "paths"
terminating in stalk pads. Scale bar = 10 µm. Fig. 4. DIC
view of the cell and stalk. The dimensions of the stalks were
similar in both cryo-FESEM and DIC microscopy. Scale bar =10 µm.
Figs. 5-10.
Ultrastructure of the stalk of A. longipes as seen by FESEM
following cryo-preservation (Figs. 5, 7, 9) and CF/CPD (Figs.
6, 8, 10). Scale bar = 2 µm. Figs. 5-6. The collar (C) obscured
details of the shaft (Sh)/ frustule interface at the raphe (R)
terminus. Fig. 5. Collar ECM (C) was discontinuous from that encapsulating
the frustule (frustule coating = FC). The collar had a fibrillar
substructure which was not obvious for the frustule coating. Fig.
6. CF/CPD removed frustule coatings and collars were retained,
although markedly structurally modified. Figs. 7-8. High magnification
of shaft surface. Fig. 7. Cryo-preservation revealed an ECM layer
coating the shaft which was confirmed by HPF/FS TEM observations
(See Fig. 22). Fig. 8. CF/CPD yielded a shaft surface with obvious
drying and extraction artifacts including condensed polymers and
cracks in the substructure. Figs. 9-10. The shaft emerged from
a pool of pad ECM directly associated with the substratum. Cryo-FESEM
preserved structure comparable to that observed with DIC, whereas,
with CF/CPD, the pad was collapsed and distorted.
Figs. 11-15.
Cymbella cistula. Fig. 11. Cell attachment to substratum
by extracellular polymers in the form of a belt-like stalk and
pad as viewed by DIC microscopy. Stalk and pad were stained with
alcian blue (pH 1.0). Scale bar = 5 µm. Figs. 12-14. Cell
and stalk were encapsulated by a layer of ECM which appeared to
be composed of fine fibrils when cryo-preserved and viewed by
cryo-FESEM (arrows). Extended sublimation times converted the
coating to large coalesced fibrils (arrowheads). Fig.12. Valve
view. Scale bar = 1.5 µm. Fig. 13. Girdle view. Scale bar
= 1.5 µm. Fig.14. Stalk. Scale bar = 0.5 µm. Fig.
15. CF/CPD completely removed the mucilaginous sheath surrounding
the frustule and stalk. Scale bar = 2 µm.
Figs. 16-17.
TEM of A. longipes cells. Scale bar = 1 µm. Fig.
16. Cross section of cell preserved by HPF/FS. Distinct features
such as the frustule (F), nucleus (N) and perinuclear Golgi (G),
chloroplast (C), vesicles v1, v2, v3, plasma membrane (P) and
laminated diatotepic layers (DL) were well preserved. Fig. 17.
CF-TEM preparation yielded large inclusions (arrows), rearrangement
of thylakoids and disruption of the plasma membrane and diatotepic
layers. Perinuclear Golgi were observed although vesicles v1,
v2 and v3 could not be identified.
Figs. 18-19.
HPF/FS TEM of A. longipes . Fig.18. Golgi bodies (G) are
proximal to the nucleus (N) and distal cisternal membranes appear
coated and associated with a tubular network (TN). The nuclear
envelope with a large perinuclear space is punctuated by closely
spaced nuclear pores (P). Scale bar = 0.3 µm. Fig. 19. Vesicles
v1 co-localized with Golgi and were often extended into irregular
tubular structures (T), enclosing material more electron dense
than the majority of the v1 matrix. Vesicles v2, contained granular
electron dense material and with a densely staining membrane,
were distributed throughout the cell and were often seen both
in the peripheral and central cytoplasm (Fig. 16). Scale bar =
0.5 µm.
Fig. 20-23.
TEM observation of secretion and organization of extracellular
adhesives in A. longipes. Scale bar = 1 µm. Fig 20.
HPF/FS TEM showed cross section of cell and associated stalk.
A thick, laminate diatotepum (DL) appressed to the silicon frustule
and composed of alternating regions of varying electron density.
Fibrillar material tightly associated with the most distal layer
of the DL was extruded through frustule pores and integrated into
the stalk (S). Note electron translucent area (arrowhead) between
the PM and the DL during stalk polymer secretion. Fig.21. Extracellular
mucilage (arrowhead) appeared in the raphe (R) canal and was closely
associated with the diatotepum (DL). Figs. 22-23. Longitudinal
cross-sections of the stalk revealed a layered fibrillar organization.
The layers consisted of a central ribbon (CR) parallel to the
long axis of stalk, fibers (arrow) radiating from central ribbon
and oriented perpendicularly to the long axis, multi-layered fibers
(MF) parallel to the long axis in the outer region of stalk, and
amorphous material (AM) in the outermost regions. Note that stalks
preserved by HPF/FS (Fig. 22) contained electron dense mucilage
(MU) at the very outer layer that was not visible in CF-TEM (Fig.
23).
ON THE
COVER: The marine biofouling diatom Achnanthes longipes attaches
to substrata via a proteoglycan stalk secreted through the frustule
directly following cessation of motility. The extracellular mucilage
layer completely encompassing the frustule is revealed in this
cryo-field emission SEM image. This layer has a significant role
in initial cell adhesion and interaction with the environment.
False color was added with Adobe Photoshop. (See article by Wang
et al. in this issue).
