Journal of Experimental Botany, Vol. 53, No. 368, pp. 513–523, March 2002
Organization of cell walls in Sandersonia aurantiaca
floral tissue
Erin M. O’Donoghue1, Sheryl D. Somerfield and Julian A. Heyes
Crop and Food Research Ltd., Private Bag 11 600, Palmerston North, New Zealand
Received 6 August 2001; Accepted 2 November 2001
Abstract
Visual symptoms of the onset of senescence in
Sandersonia aurantiaca flowers begin with fading
of flower colour and wilting of the tissue. When fully
senescent, the flowers form a papery shell that
remains attached to the plant. The cell walls of these
flowers have been examined to determine whether
there are wall modifications associated with the late
stages of expansion and subsequent senescencerelated wilting. Changes in the average molecular size
of pectin were limited through flower opening and
senescence, although there was a loss of neutral
sugar-containing side-branches from pectins in opening flowers, and the total amounts of pectin and cellulose continued to rise in cell walls of fully senescent
sandersonia flowers. Xyloglucan endotransglycosylase activity increased in opening and mature flowers,
but declined sharply as flowers wilted. Concomitantly,
the proportion of hemicellulose polymers of increasing molecular weight increased from flower expansion
up to the point at which wilting occurred. Approximately 50% of the non-cellulosic neutral sugar in
mature flower cell walls was galactose, primarily
located in an insoluble cell wall fraction. Total
galactose in this fraction increased per flower with
maturity, then declined at the onset of wilting.
b-Galactosidase activity was low in expanding tepals,
but increased as flowers matured and wilted.
Key words: Cell walls, flower senescence, galactose,
molecular weight, XET.
Introduction
Flower opening and petal senescence are highly controlled biological events critical in the life cycle of
1
flowering plants. They are also events of commercial
value, contributing to the visual appeal and post-harvest
vase-life of cut flowers. There are questions whether
flower opening and the onset of senescence-indicative
deteriorative changes are likely to be a developmental
continuum (with opening as the first visual symptom of a
senescence programme that has already started), rather
than separate but coexisting processes (senescence starts
only when flower maturity is reached).
There is enormous variation among flowering plants in
the manner and timing of both the flower opening phase
and the onset of petal senescence. As an example, Asiatic
lily opening is achieved by petal cell expansion, with
uneven expansion of the different epidermal layers of the
petal contributing to the development of the open flower
shape (Bieleski et al., 2001). By contrast, the fabric of
the morning glory flower has formed before opening,
and opening is achieved by unfolding the massed petal
tissue, although this also appears to require selective cell
expansion (Phillips and Kende, 1980). Studies of the
factors controlling cell expansion have focused primarily
on cells of growing shoots. In these tissues, expansion is
controlled by cellular turgor pressure and stress on the
cell wall. Expansins, non-enzymic proteins which may
disrupt the tight arrangement of hemicelluloses with
cellulose microfibrils, are thought to be a fundamental
factor in the ability of the wall to extend (Cosgrove, 1998,
1999), but modification of covalent linkages within
existing cell wall polymers as well as deposition of new
cell wall material ensure the wall retains strength during
expansion (Cosgrove, 1997). Gene expression relating to
the presence of expansins has been identified in young pea
petals (Michael, 1996) and in tomato flowers (Brummell
et al., 1999), and polygalacturonase gene expression has
been detected in kiwifruit flowers at anthesis (Wang et al.,
2000). This suggests that the cell wall may be similarly
modified in opening petals although the extent is likely to
To whom correspondence should be addressed. Fax: q64 6 351 7050. E-mail: [email protected]
ß Society for Experimental Biology 2002
514
O’Donoghue et al.
be strongly reliant on the nature of the polysaccharides
present.
Factors associated with petal senescence have been
reviewed most recently by Rubinstein who focused on the
increasing molecular evidence supporting programmed
cell death in petal senescence, determined by specific gene
expression changes and signalling events (Rubinstein,
2000). Petal wilt is associated with the onset of senescence
in many flowers and is related to loss of membrane integrity, changes in turgor and ion leakage. Recent studies
indicate that factors controlling these changes are initiated prior to visual wilt and involve new protein synthesis
(Celikel and van Doorn, 1995; van Doorn et al., 1995;
Panavas and Rubinstein, 1998).
The loss of cell wall integrity is closely associated
with tissues undergoing senescence, most notably in the
ripening-related softening of many fruits. Catabolic processes such as solubilization and breakdown of cell wall
polymers (Fischer and Bennett, 1991) dominate in this
situation, although there are reports of new cell wall synthesis also occurring (Mitcham et al., 1989). There are also
a small number of studies describing significant cell wall
modification in flower petals that are undergoing changes
to texture during senescence, such as quantitative loss
of hemicellulose in senescing morning glory (WiemkenGehrig et al., 1974) and increased pectin solubilization
and loss of large-sized hemicellulose polymers in senescing petals of cut carnation flowers (de Vetten and Huber,
1990; de Vetten et al., 1991). Panavas et al. found increasing activity of cell wall hydrolases in senescing daylily,
indicating the potential for wall breakdown (Panavas
et al., 1998). Up-regulated expression of a polygalacturonase gene has also been found in senescing kiwifruit
flowers (Wang et al., 2000). With this in mind the cell
walls and some cell wall-related hydrolases of Sandersonia
aurantiaca flowers have been analysed during opening
and subsequent senescence, with the primary aim of
determining whether there are specific elements of wall
modification that occur during these events, especially
with reference to changes in petal texture.
Sandersonia aurantiaca (Hook.) is a liliaceous monocotyledon with a bell-shaped corolla formed from fused
petals and sepals (tepals). Sandersonia flowers have been
used as models for the study of senescence in ethyleneinsensitive floral systems (Eason and Webster, 1995;
Eason and de Vré, 1995; Eason et al., 1997, 2000a, b). The
progression of bud appearance, flower opening and senescence in the sandersonia flower proceeds in a manner
that provides reliable visual indicators for flower selection. These developmental stages have been linked to
quantitative measurements of intensifying and fading
colour, loss of chlorophyll and protein, and changes to
fresh weight, dry weight and starch content (Eason
and Webster, 1995; Eason et al., 1997). Pollination does
not affect the onset or progression of senescence in
sandersonia (JR Eason, personal communication). New
protein synthesis, central to the co-ordination of senescence events in sandersonia, is thought to be linked to
changes in gene expression that begin prior to the visual
symptoms of deterioration (Eason et al., 1997, 2000a, b).
Materials and methods
Plant material
Sandersonia tubers were grown in a plastic house as described
previously (Eason and Webster, 1995). Flowers were selected
according to the criteria described earlier (Eason and Webster,
1995) and are shown in Fig. 1A, I–V. In brief, flowers in bud
(included for some experiments) have aperture tips that are
pointed, green and closed. Opening flowers are defined by an
open, green aperture with the tips turned back, while mature
flowers are completely opened and are fully orange with no
trace of green coloration. The subsequent wilting phase affects
the turgidity of the whole flower and there is a loss of colour
intensity. Senesced flowers are pale brown, do not resist compression but are not brittle. Subsequent to this the flowers
become dry and shrivelled but still remain attached to the
pedicel. Eason and Webster indicated that for sandersonia
the average developmental time from bud was 2.5 d (opening),
5.5 d (mature), 9.5 d (wilted), and 11.5 d (senesced) (Eason and
Webster, 1995).
Tissue preparation
Fresh tissues from the bell region of sandersonia flowers were
fixed in ethanol : acetic acid : formalin : water (10 : 1 : 2 : 7; by
vol.), equilibrated in ethanol, followed by Histoclear, then
embedded in paraffin wax. Thin sections were stained with
toluidine blue (1%) and examined using light microscopy.
For other procedures, the stem, reproductive tissues and
nectaries were discarded and the remaining tepal tissue (encompassing the bell and aperture regions) of each flower was
weighed, frozen in liquid N2 then stored at 80 8C. Frozen tissue
was subsequently used for cell wall preparations or freeze-dried.
Tensile and compressive strength
The strength of fresh tepal tissue was measured by measuring
the force required to pull tepal strips apart at a 908 angle to the
vascular tissue. Tepal strips (approximately 15 mm long, 4 mm
wide) were excised from the bell region of each of 10 flowers.
Adhesive tape was attached to the sides of the strip, and the strip
width and thickness measured using digital callipers. The taped
edges were clamped into an Instron Universal Testing Machine
(Model 4301) and the tepal pieces were pulled apart at a crosshead speed of 50 mm min1. Load and displacement data were
calculated using Instron Series IX software (Version 5.34.00).
The strength required to pull the tepal tissue apart was expressed
on a per strip (N mm1) and per area (N mm2) basis. In both
cases the width of the strip was taken into account, whereas for
strength per area calculations the thickness of the tepal strip was
also included. Flower firmness was measured as the force (N)
required to compress whole flowers through a distance 20%
of the widest diameter (according to the method of Eason
et al., 1997).
Sandersonia cell walls
515
buffered phenol to inactivate wall-bound enzymes (Huber,
1992). The EIR was air-dried, then ground to a powder and
stored at 80 8C.
The EIR was assayed for cellulose content (Dische, 1953;
Updegraff, 1969), and for total pectin (Ahmed and Labavitch,
1977). Pectin esterificiation was estimated using the protocol of
Wood and Siddiqui with volumes adjusted for microassays
in 96-well microtitre plates (Wood and Siddiqui, 1971).
Soluble pectins were sequentially extracted from EIR
by 50 mM trans-1,2-cyclohexanediamine-N,N,N9,N9-tetraacetic
acid (CDTA) in 50 mM NaOAc, pH 6.0, followed by 50 mM
Na2CO3 with 20 mM NaBH4. The proportion of EIR to
extractant was 1 mg : 0.5 ml. Both extractions were at room
temperature for 4 h with constant shaking. Hemicelluloses were
extracted from EIR in 6 N NaOH containing 0.13 mM NaBH4,
using the method described earlier (O’Donoghue et al., 1997).
For all the extractions, uronic acids were determined by
the method of Blumenkrantz and Asboe-Hansen (Blumenkrantz
and Asboe-Hansen, 1973), and total carbohydrates were assayed
using the phenol : sulphuric acid method (Dubois et al., 1956).
Neutral sugar analysis
Neutral sugars in the EIR were hydrolysed in 2 N trifluoroacetic
acid and then derivatized to alditol acetates (Blakeney et al.,
1983). The alditol acetates were quantified by gas chromatography using a BPX-70 capillary column (0.53 mm 3 25 m,
SGE) and flame ionization detection. Allose was used as an
internal standard. The temperature programme was as follows:
170 8C (2 min), 10 8C min1 to 190 8C, then 7 8C min1 to
240 8C (0.25 min).
In order to quantify galactose in cell wall fractions, sequential extractions of EIR, using CDTA, Na2CO3 and 6 N NaOH
were prepared. CDTA and Na2CO3 extracts were dialysed
against 9.0 l water (2 3 12 h) prior to derivatization. The 6 N
NaOH-soluble material was not passed through anion exchange
columns prior to analysis. Insoluble extraction residues were
washed with water until neutral, then freeze-dried. All extracts
and insoluble extraction residue from each developmental stage
were analysed for galactose content by gas chromatography,
using the derivatization method described above.
Fig. 1. (A) Sandersonia flowers used in this work, selected according to
the developmental stages described previously (Eason and Webster,
1995). I, bud (aperture tips closed, pointed and green); II, opening
(aperture open, tips folded back), III, mature (fully open, completely
orange); IV, wilted (loss of turgidity in bell, faded colour); V, senesced
(pale brown, no resistance to compression). (B) Transverse sections of
fixed, paraffin-embedded tepal tissue at developmental stages corresponding to those in (A) stained with toluidine blue. Arrows indicate
areas of parenchyma disruption; vt, vascular tissue. Scale bar ¼ 100 mm.
Cell wall compositional analyses
Ethanol-insoluble residue (EIR) was prepared by homogenizing
frozen tepal tissue in ethanol and then chilling at 20 8C for
16 h. The homogenates were vacuum-filtered through several
layers of organza fabric (obtained locally), then treated with
Molecular size distribution
Molecular size distribution of extracted cell wall fractions was
evaluated using size-exclusion chromatography. CDTA- and
Na2CO3-soluble pectin extracts (approximately 500 mg uronic
acid ml1) were applied to a Superose 6HR size exclusion column
(10 mm 3 300 mm, Pharmacia) operating in 30 mM NaOAc,
pH 6.5, containing 20 mM NaCl and 10 mM EDTA. Flow rate
was 0.5 ml min1 and 0.5 ml fractions were assayed for uronic
acids (Blumenkrantz and Asbose-Hansen, 1973) and total
carbohydrates (Dubois et al., 1956) in 96-well microtitre plates.
Molecular size profiles are representative of four extractions of
CDTA- and Na2CO3-soluble pectins from each developmental
stage.
Hemicelluloses were similarly separated using a Superose
6HR column except that the elution buffer was 50 mM NaOAc,
pH 5.0, containing 10 mM NaCl. Fractions were assayed for
total carbohydrates (Dubois et al., 1956) and for xyloglucan
using an iodine binding assay (Kooiman, 1960) in 96-well
microtitre plates. Hemicellulose molecular size profiles are
representative of two extractions from each developmental
stage.
516
O’Donoghue et al.
Analysis of cell wall enzyme activity
b-Galactosidase: Freeze-dried tissue was extracted in 50 mM
NaOAc, pH 4.5 ("1 M NaCl) and extracts were assayed for
b-galactosidase activity using p-nitrophenyl-b-D-galactopyranoside as substrate. The method of Tanimoto and Igari was used
(Tanimoto and Igari, 1976), with modifications (O’Donoghue
et al., 1998). Activity is expressed as U flower1 where
1 U ¼ 1 mmol nitrophenol released h1. Assays were performed
in the presence of up to 30 mM galactose that co-solubilized in
the extract.
Pectinmethylesterase (PME): Freeze-dried tissue (20 mg) was
extracted in 50 mM NaOAc pH 6.0, with 1 M NaCl, and
aliquots were incubated with 0.2% citrus pectin and extraction
buffer without NaCl for 16 h at 40 8C. Methanol released was
assayed (Wood and Siddiqui, 1971) according to the method
adapted for a microtitre plate. PME activity is expressed as U
flower1 where 1 U ¼ 1 nmol methanol released h1.
water loss. The firmness of whole tepals, measured
objectively by compressibility, increased as the flowers
opened and was greatest in the mature flowers (Fig. 2B).
Firmness of wilted flowers was only 10% of that at
maturity. There was no difference between the firmness of
wilted and senesced flowers.
Tepal tissue strength per tissue strip was highest in
expanding and fully opened, mature flowers, but declined
as flowers began to wilt (Table 1). The most fragile tissues
were those of senesced tepals. Tepal tissue from postmature flowers was thinner than that of opening or mature
flowers (data not shown). When the tepal thickness is
accounted for in the strength calculation, there was no
Xyloglucan endotransglycosylase (XET): Freeze-dried tissue
(20 mg) was extracted with 1.2 ml buffer (250 mM NaOAc,
400 mM NaCl, pH 5.8) with 20 ml of 625 mg ml1 dithiothreitol and 30 mg polyvinylpolypyrrolidone. XET was assayed
using basic procedures (Fry et al., 1992; Redgwell and Fry,
1993), with the exception that the substrate consisted of
asparagus xyloglucan and tritiated asparagus-derived xyloglucan oligomers buffered with 1 M NaOAc, pH 5.8, prepared
and used as described previously (O’Donoghue et al., 2001).
Activity was measured by the incorporation of the tritiated
xyloglucan oligomer into the xyloglucan polymer using scintillation counting, and is expressed as U flower1, where
1 U ¼ Bq (kBq supplied. h)1.
Results
Cross-sections of toluidine blue-stained sandersonia tepal
tissue are shown in Fig. 1B, I–V, alongside photographs
of whole flowers at matching developmental stages in (A).
Flowers at the bud stage (B, I) had a clearly defined layer
of parenchyma cells between the upper and lower
epidermis. This layer became progressively disordered as
the flowers began to open (B, II), reached full size (B, III)
and wilted (B, IV), with large spaces becoming apparent.
In fully senesced flowers (B, V) the parenchyma structure
was completely collapsed and tepal tissue appeared to
consist of an upper and lower epidermis, with intact
vascular tissue existing within a compressed strip of the
remains of the parenchyma cells.
The total fresh weight of developing and senescing
sandersonia flowers is shown in Fig. 2A, with the contributory water content and dry matter flower1 embedded
in the same bar for each developmental stage, and the
ratio of water : dry matter per flower as a line graph. The
maturity-related increase, as well as the wilting-related
decrease in average fresh weight flower1 was due to
changes in both water and dry matter content of
sandersonia flowers. The proportions of water and dry
matter per flower did not change from opening to wilting,
but a change in this ratio occurred in senescing flowers
where dry matter loss flower1 did not keep pace with
Fig. 2. Sandersonia flower characteristics at each developmental
stage. (A) Bar graph of sandersonia flower fresh weight components
(mg flower1); full bar, total flower weight (LSD ¼ 21.1); hatched area,
water content (LSD ¼ 18.5); solid area, dry matter content (LSD ¼ 3.0).
Line graph above indicates water : dry matter ratio per flower (k) at
each stage (LSD ¼ 0.7). In each case, df ¼ 15, a ¼ 0.05. (B) Compression
strength (N) of sandersonia flowers at each developmental stage
(LSD ¼ 0.063, df ¼ 15, a ¼ 0.05).
Table 1. Tensile strength of sandersonia tepal tissue
Tensile strength per strip (N mm1) and corrected for tepal thickness
(strength area1, N mm2). Tensile strength was measured by pulling a
strip of tepal tissue apart at right angles to the vascular tissue, using an
Instron Universal Testing Machine. Strip dimensions were measured
with digital calipers.
Opening
Mature
Wilted
Senesced
LSD
Strength per strip
(N mm1)
Strength per area
(N mm2)
0.205
0.207
0.180
0.156
0.026
0.840
0.753
0.724
1.670
0.168
Sandersonia cell walls
Fig. 3. Sandersonia tepal cell wall components, (mg flower1) at each
developmental stage. (A) Ethanol-insoluble residue (n ¼ 3). (B) Cellulose
(n ¼ 3). (C) Total pectin (n ¼ 3). (D) CDTA-soluble pectin (n ¼ 8). (E)
Na2CO3-soluble pectin (n ¼ 8). Standard errors of the means are
indicated by two-way error bars.
517
change in the tissue strength per unit area (fixed width
of the strip w4 mmx 3 thickness) until flowers were fully
senesced, at which point the strength per unit area more
than doubled.
The amount of EIR flower1 increased as flowers
became fully mature (Fig. 3A), but decreased as flowers
began to wilt and finally senesced. Starch levels in these
preparations were ;8 mg mg1 EIR in opening flowers
and -3 mg mg1 EIR in the succeeding stages (data not
shown). The amount of cellulose and pectin (flower1
basis) increased as flowers expanded to maturity, remained
at these levels in wilted flowers, but increased again in
fully senesced flowers (Fig. 3B, C). There were no changes
in esterified pectin content (% total pectin basis) until
flowers became fully senescent. Levels decreased from
35.9% ("0.6) at the onset of wilting to 30.4% ("0.6)
in the senesced tepals. Hemicellulose levels ranged from
1–2 mg per flower and no distinct trends in hemicellulose
content were apparent in the developmental stages
sampled (data not shown).
The quantity of CDTA-soluble pectin flower1 was
similar in all developmental stages until flowers were fully
senescent, when there was an almost 25% loss of this
pectin fraction (Fig. 3D). On the other hand, the amount
of Na2CO3-soluble pectin flower1 declined gradually in
successive stages with a significant drop in this fraction as
flowers began to wilt (Fig. 3E). The pectin accumulation
in senesced flowers must be due to pectins other than
those soluble in CDTA or Na2CO3.
The predominant non-cellulosic neutral sugar in cell
walls of opening and mature sandersonia flowers was
galactose (46–50 mol%, Table 2). The levels of arabinose,
xylose and glucose were lower (between 10–12 mol%
each). As the flowers began to wilt, the most notable
feature was the reduction in the proportion of galactose
present. This trend continued as the flowers approached
complete senescence.
Throughout sandersonia flower development the cell
wall galactose was primarily located in polysaccharides
that were insoluble in CDTA, Na2CO3 or 6 N NaOH
(Table 3). The amount of galactose in this insoluble
residue (flower1 basis) almost doubled in mature flowers,
but decreased significantly as the flowers wilted and
Table 2. Major non-cellulosic neutral sugars (mol% basis) in sandersonia tepal cell walls
Non-cellulosic neutral sugars were determined by gas chromatographic analysis of alditol acetates derivatized from trifluoroacetic acid hydrolysates of
cell wall preparations made from opening, mature, wilted, and senesced sandersonia tepals. Rha, rhamnose; Fuc, fucose; Ara, arabinose; Xyl, xylose;
Man, mannose; Gal, galactose; Glc, glucose. Results are the means of three derivatizations, expressed in mol% and with standard errors.
Opening
Mature
Wilted
Senesced
Rha
Fuc
Ara
Xyl
Man
Gal
Glc
7.2"0.6
9.4"0.6
11.3"0.5
18.5"3.0
1.6"0.2
1.7"0.2
2.1"0.3
2.7"0.7
11.0"0.8
10.5"0.7
10.5"0.7
9.9"1.2
12.1"3.3
12.1"1.2
14.2"1.3
23.9"1.3
10.3"3.3
5.2"0.6
8.0"0.9
9.7"1.1
46.1"2.4
50.6"1.0
38.3"2.8
20.0"1.7
11.8"1.4
10.4"1.1
15.6"5.5
15.4"1.6
518
O’Donoghue et al.
senesced. Cell wall galactose was partitioned in much
lower amounts into polymers soluble in CDTA, Na2CO3
and 6 N NaOH. As sandersonia flowers matured there
was some accumulation of galactose into CDTA-soluble
polymers, but an almost 35% loss of galactose from
Na2CO3-soluble polymers. The amount of galactose
decreased in all three solubilized fractions with the onset
of wilting and senescence, but these losses did not
approach the amount of galactose lost from insoluble
residues at the same stages.
Molecular size distribution
The average molecular size distribution profiles of
CDTA-soluble pectins extracted from sandersonia tepals
are shown in Fig. 4. In opening flowers, most of the
Table 3. Partitioning of galactose (mg flower1) in isolated
fractions of sandersonia EIR
Galactose content was determined by gas chromatographic analysis
of alditol acetates derivatized from trifluoroacetic acid hydrolysates of
CDTA, Na2CO3 and 6 N NaOH extracts and residue of EIR from
opening, mature, wilted, and senesced sandersonia flowers. Results are
the means of three separate extractions, expressed in mg flower1 and
with standard errors.
CDTA
Na2CO3
6 N NaOH
Residue
Opening
Mature
Wilted
Senesced
214"20
442"28
336"31
857"133
277"23
289"42
316"33
1543"95
172"6
144"5
260"29
1161"147
95"19
77"5
99"9
489"86
uronic acid content eluted as one peak, beginning very
close to the void volume of the column and with extensive
tailing (Fig. 4A). Analysis of the fractions for total
carbohydrates using phenol : sulphuric acid, revealed
an extra dimension to this pectin profile (Fig. 4B).
Phenol : sulphuric acid detects both neutral and acidic
sugars, although it is more sensitive to the neutral sugar
component. A second peak, eluting between the 500 kDa
and 70 kDa size markers was evident, but not fully
resolved from the first, largest peak. This may represent
the neutral sugar branch component of a small group
of pectins that were water-soluble and co-extracted in
CDTA. A third peak eluted very close to the inclusion
volume of the column. Neutral sugar analysis of the
CDTA-soluble extracts, without prior trifluoroacetic
acid hydrolysis, confirmed the presence of sugar monomers that would have eluted in this peak. The mol% of
neutral sugars ranged from 80% glucose in opening
flowers to 50% glucoseu25% rhamnose in senesced flowers.
To investigate whether there could be neutral sugarcontaining oligomers present in CDTA-soluble extracts
(suggested by the broadness of the third peak), fractions
corresponding to this peak were collected, concentrated,
then separated on a Superdex Peptide column (Pharmacia;
carbohydrate exclusion limit, 10 kDa, inclusion limit
;0.5 kDa). The majority of carbohydrate eluted at the
inclusion volume, but a small proportion from extracts of
all developmental stages did fractionate (data not shown)
suggesting the presence, albeit minor, of carbohydrate
oligomers.
Fig. 4. Molecular size distribution of CDTA-soluble pectin extracted from opening (A), mature (B), wilted (C), and senesced (D) sandersonia tepals.
Pectins were fractionated on Superose 6HR, and fractions were assayed for uronic acid (m) and total carbohydrates (k). Distributions are expressed
as percentage units of the total detected. Tick marks at the top of each graph indicate the elution position of dextran size markers in the order
5 3 103 kDa, 500 kDa, 73 kDa, 40 kDa, 9.3 kDa, and glucose.
Sandersonia cell walls
519
Fig. 5. Molecular size distribution of Na2CO3-soluble pectin extracted from opening (A), mature (B), wilted (C), and senesced (D) sandersonia tepals.
Pectins were fractionated on Superose 6HR, and fractions were assayed for uronic acid (m) and total carbohydrates (k). Distributions are expressed
as percentage units of the total detected. Tick marks at the top of each graph indicate the elution position of dextran size markers in the order
5 3 103 kDa, 500 kDa, 73 kDa, 40 kDa, 9.3 kDa, and glucose.
The size distribution of the uronic acid component
of CDTA-soluble pectins from sandersonia did not alter
appreciably with increasing maturity, wilting or advanced
senescence of the tepals. On the other hand there were
significant changes in both the mid- and smallest-sized
polymer groups whose distributions became more prominent as flowers matured and wilted, and broaderuflatter
as flowers approached full senescence.
The molecular size profiles of polymers soluble in
Na2CO3 were distinctly different from CDTA-soluble
pectins (Fig. 5). The uronic acid component of these
fractions eluted in one very broadly distributed peak. The
shape of the leading edge of this peak suggests that
there were void-volume-sized polymers present in opening flowers (Fig. 5A), whereas in subsequent stages
(Fig. 5B–D) there were proportionately more polymers
in the lower size ranges. This was confirmed by the
phenol : sulphuric acid analysis of the column fractions
which showed the gradual disappearance of the earliesteluting polymers of relatively large molecular size. A
second peak, as a shoulder to the main body of the
distribution, became less evident in the same way. A
comparison of total carbohydrate and uronic acid distributions of this Na2CO3-soluble extract suggests that
the group of relatively large polymers contained a
high proportion of neutral sugars, presumably as sidebranches to the main chain of uronic acids. In senesced
sandersonia flowers, Na2CO3-soluble pectins had similar total carbohydrate and uronic acid distributions,
suggesting that neutral sugars were either equivalently
Fig. 6. Molecular size distribution of total hemicelluloses (A) and
xyloglucan (B) in opening (m), mature (k), wilted (j) and senesced (h)
sandersonia tepals, fractionated on Superose 6HR. Distributions are
expressed as percentage units of the total detected. Tick marks at the top
of each graph indicate the elution positions of dextran size markers, in
the order 5 3 103 kDa, 500 kDa, 73 kDa, 40 kDa, and 9.3 kDa.
dispersed over the pectin chains, or that they were
absent altogether.
The molecular weight distribution of hemicelluloses
extractable in 6 N NaOH is shown in Fig. 6. There
were two predominant size groups of hemicelluloses in
sandersonia, the largest eluting between the column void
volume and the 500 kDa dextran marker and the other
eluting between the 40 and 9.3 kDa size markers. A third,
less distinct, size cluster was centred around the 73 kDa
520
O’Donoghue et al.
marker. As flowers developed from opening to wilting,
the peak containing the largest-sized polymers moved to
a position of larger average molecular size. The elution
position of the smallest-sized group remained unchanged,
although the relative amounts present were altered. The
intermediate cluster became less distinguishable in wilted
flowers. In fully senesced flowers there was a downshift in
the peak containing the largest-sized polymers so that
the elution profile resembled that of hemicelluloses from
mature tepals. The xyloglucan molecular size distribution
closely followed that of the large-sized hemicelluloses
(Fig. 6B), with average molecular size increasing until
flowers wilted, then decreasing in flowers that were fully
senesced.
sandersonia flowers swelled and the aperture expanded.
XET activity was highest in fully mature flowers, followed
by sharp declines in wilted and senesced flowers.
Discussion
Both colorimetric and molecular size analyses by sizeexclusion chromatography revealed an inability of fresh
tissue extracts from any of the sandersonia samples to
hydrolyse either partially de-esterified citrus pectin or
carboxymethylcellulose, indicating very low or no endopolygalacturonase and b-1,4-glucanase-type activity, and
this was not pursued further.
b-Galactosidase activity (flower1 basis, Table 4), low
in buds and opening flowers, was significantly higher in
mature flowers. There was a further, almost doubling of
activity as tepals wilted, then a sharp decrease as the
flowers entered the final stage of senescence. With 1 M
NaCl included in the extraction buffer, this activity pattern was retained, but all levels were elevated, indicating the presence of a proportion of tightly wall-bound
enzyme.
PME activity was low in sandersonia flowers and
assays required long incubation times in order to obtain
detectable activity. Nevertheless, PME activity (flower1
basis, Table 5) was greatest in opening and mature flowers
then declined as flowers wilted. The predominance of
XET activity (flower1 basis, Table 5) was also associated
with opening of sandersonia flowers. There was low XET
activity in buds, but an almost 5-fold increase occurred as
Some of the cell wall characteristics of sandersonia flowers
have been described as the flowers open and then senesce.
It is interesting that the opening of sandersonia flowers
is accompanied by considerable disruption to the cell
organization in the parenchyma layer of the tepal tissue,
similar to opening daylily (Panavas et al., 1998). This
‘looseness’ may permit lateral expansion of the flower
tissue. The parenchyma cells also appear to break down
considerably, so that in senesced flowers only epidermal
cells and vascular tissue are intact.
The main interest was in determining whether there
were distinctive changes, occurring after Sandersonia
flowers reached maturity, that might contribute to the
textural changes associated with wilting, and whether
there were elements of cell wall modification, initiated
during flower opening, that continued unchecked through
to senescence. The cell walls of only two other species,
daylily and carnation, have been studied in great detail
with respect to this period in flower development.
Sandersonia, daylily and carnation flowers differ greatly
in the combinations of ethylene sensitivity, visual senescence symptoms and speed of deterioration after opening.
There is new evidence that there are few common features
with respect to cell wall characteristics of their opening
and senescence.
There was little indication of pectin hydrolysis in
sandersonia flowers, and there was detectable endopolygalacturonase activity at any developmental stage.
Increasing polygalacturonase activity was reported in
senescing daylily petals (Panavas et al., 1998), indicating
the potential for pectin backbones to be hydrolysed in
these flowers, but carnation petals have little pectin
hydrolysis during senescence (de Vetten and Huber,
1990). The carnations of that study, however, were cut
Table 4. b-Galactosidase activity in sandersonia flowers
Table 5. PME and XET activities in sandersonia flowers
b-Galactosidase activity was assayed in freeze-dried tissues of bud,
opening, mature, wilted, and senesced sandersonia tepals and detected
by the release of nitrophenol from p-nitrophenyl-b-D-galactopyranoside.
Results are expressed as U flower1 ("standard error, n ¼ 3) where
1 U ¼ 1 mmol nitrophenol released h1.
PME activity was detected by the release of methanol from citrus pectin.
Results are expressed as U flower1 ("standard error, n ¼ 3) where
1 U ¼ 1 nmol methanol released h1. XET activity was detected by the
incorporation of tritiated XG oligomers into XG polymers and results
are expressed as U flower1 ("standard error, n ¼ 3) where 1 U is 1 Bq
of tritiated polymer formed (kBq supplied. h)1.
Cell wall hydrolases
Bud
Opening
Mature
Wilted
Senesced
Extraction
buffer only
Extraction
bufferq1 M NaCl
5.1"1.2
8.6"0.7
21.4"4.3
39.0"3.6
17.9"2.0
9.2"0.5
19.2"0.6
38.4"2.2
52.1"4.4
19.1"3.0
Bud
Opening
Mature
Wilted
Senesced
PME
XET
233"22
362"81
409"96
266"23
167"22
52.9"0.9
252.6"8.6
310.2"0.5
43.5"0.2
6.2"0.3
Sandersonia cell walls
and then supplied with sucrose solution rather than
senescing naturally, attached to the parent plant.
A sharp increase was found in XET activity associated
with sandersonia opening. It is likely that, as with growing shoots, XET is part of a secondary response of wall
modification that supports the initial stress relaxation in
expanding cell walls, and reforms linkages to retain wall
strength. The increase in the average molecular size of
xyloglucan is likely to be due to progressive XET action
on existing polymers, since extractable hemicellulose
levels did not change appreciably. Between wilting and
full senescence, the xyloglucan of sandersonia flowers
began to degrade, but this was quite limited in comparison to the extensive enzyme-mediated degradation
of hemicelluloses from carnation that starts even prior
to the attainment of full bloom (de Vetten et al., 1991).
No evidence was found of b-1,4-glucanase activity
(carboxymethylcellulose hydrolysis) at any point in
sandersonia floral development, unlike daylily where
cellulase activity was greatest in opening petals (Panavas
et al., 1998).
Galactose formed about half the non-cellulosic neutral
sugar component of pre-senescent sandersonia flowers.
The reduction in the mol% of galactose in wilted flowers
paralleled the increase in b-galactosidase activity. Daylily
too, has increased b-galactosidase activity specific to the
onset of senescence-related deterioration (Panavas et al.,
1998). Although galactose is not the predominant noncellulosic neutral sugar in carnation flowers compared
to sandersonia, galactose loss is also associated with
senescence (de Vetten and Huber, 1990). The loss of
galactose from the cell walls of many ripening fruit
species has been reported repeatedly and there have been
wide-ranging comparative studies of the occurrence of
this phenomenon (Gross and Sams, 1984; Redgwell et al.,
1997), the significance of which has not yet been satisfactorily resolved. Most galactose from ripening fruit
cell walls is lost from polymers soluble only in strong
(4 N) alkali or from insoluble, highly branched pectic
polysaccharides (Redgwell et al., 1997). This also appears
to be the case in sandersonia, but with the major
quantities of galactose located in, and lost from,
uncharacterized polymers that are insoluble by standard
means.
The compressibility test described here quantifies the
firmness of the whole flower, objectively confirming the
increasing turgidity of mature flowers and onset of wilt.
This test does not provide information on the strength of
the tepal fabric and an attempt was made to quantify
this by measuring the force required to pull apart tissue
strips of fixed dimensions, at right angles to the vascular
strands. Tissue strips were composed of differing total
amounts of water and dry matter at each developmental
stage, with the cell wall being an increasing proportion
of dry matter in post-mature flowers. In addition, while
521
tissue strip length and width dimensions were fixed, it was
found that tissue from post-mature flowers was reduced
in thickness, presumably due to tissue shrinkage. It is
likely that the similar proportions of water and dry matter
in opening and mature flowers were responsible for the
similar tensile strength in these strips, even when tissue
thickness is taken into account. As flowers opened to full
maturity there was a decreasing proportion of carbohydrate associated with long-chain Na2CO3-soluble
pectin. The concurrent loss of galactose from this pectin
fraction suggests a loss of side-branches during opening,
but this change did not impact on the strength of the
tepal fabric.
The strength of tepal strips was reduced in wilting and
senesced sandersonia flowers compared to those of fully
mature flowers. On a per unit area basis (accounting for
the differences in tepal thickness), however, the tensile
strength of the tissue remained unchanged in wilting
flowers and significantly increased in senesced flowers.
Significant cell wall polymer changes accompanied these
alterations to tepal strength. There was significant loss
of galactose per flower during wilting and senescence,
most notably from an insoluble cell wall fraction, and the
polymers undergoing this change may play a key role in
cell wall cohesiveness. There were also increased amounts
of cellulose and pectin flower1 as the sandersonia flowers
progressed from wilting to fully senesced, and while there
was no evidence of substantial solubilization or degradation of cell wall pectins there was a slight downshift in the
average molecular size of xyloglucan in this last stage.
Cell walls in senescing flowers may also be stiffened in
other ways (e.g. increased phenolic-based linkages, nonvascular lignification) so that the tepal shell that remains
was, with thickness accounted for, much stronger even
than that of the opening flower. This accumulation of cell
wall material in the tepal tissue may be reflective of the
development of the sandersonia tepal tissue as a protective cover for the developing ovary, and it would be
interesting to identify whether it was a shared feature of
the large group of monocotyledonous flowers that senesce
without petal abscission (van Doorn, 2001).
Finally, as to whether the changes occurring in the cell
walls of senescing sandersonia flowers are distinct events
or consequences of the considerable expansion required
for full opening, it was found that only one change, that of
galactose mobilization, spanned the opening–senescence
spectrum, although different polymers of the cell wall were
involved at different times. There was continuous loss
of galactose from Na2CO3-soluble pectins, but the major
loss of galactose occurred as flowers began to wilt,
originating from highly insoluble, but as-yet unidentified,
polysaccharides. The characterization of these polymers,
as well as the wilting-associated role of b-galactosidase
activity in galactose mobilization, will be addressed in
future work.
522
O’Donoghue et al.
Acknowledgements
Our thanks to Jocelyn Eason, Bub Vizi and Greg Kent for
providing the flowers for this work, Leigh de Vré and Jason
Johnston for assistance with the tensile strength tests and
Wilhelmina Borst and Katie Jarmai-Graf for assistance with
histology. This work was supported by the New Zealand
Foundation for Research, Science and Technology.
References
Ahmed AE, Labavitch JM. 1977. A simplified method for
accurate determination of cell wall uronide content. Journal of
Food Biochemistry 1, 361–365.
Bieleski R, Elgar J, Heyes J. 2001. Mechanical aspects of
rapid flower opening in Asiatic lily. Annals of Botany
86, 1175–1183.
Blakeney AB, Harris PJ, Henry RJ, Stone BA. 1983. A simple
and rapid preparation of alditol acetates for monosaccharide
analysis. Carbohydrate Research 113, 291–299.
Blumenkrantz N, Asboe-Hansen G. 1973. New method of
quantitative determination of uronic acids. Analytical
Biochemistry 54, 484–489.
Brummell DA, Harpster MH, Dunsmuir P. 1999. Differential expression of expansin gene family members during
growth and ripening of tomato fruit. Plant Molecular Biology
39, 161–169.
Celikel FG, van Doorn WG. 1995. Solute leakage lipid
peroxidation and protein degradation during the senescence
of Iris tepals. Physiologia Plantarum 94, 515–521.
Cosgrove DJ. 1997. Assembly and enlargement of the primary
cell wall in plants. Annual Review of Cell and Developmental
Biology 13, 171–201.
Cosgrove DJ. 1998. Cell wall loosening by expansins. Plant
Physiology 118, 333–339.
Cosgrove DJ. 1999. Enzymes and other agents that enhance cell
wall extensibility. Annual Review of Plant Physiology and
Plant Molecular Biology 50, 391–417.
de Vetten NC, Huber DJ. 1990. Cell wall changes during the
expansion and senescence of carnation (Dianthus caryophyllus) petals. Physiologia Plantarum 78, 447–454.
de Vetten NC, Huber DJ, Gross KC. 1991. Endoglycanasecatalyzed degradation of hemicelluloses during development
of carnation (Dianthus caryophyllus L.) petals. Plant
Physiology 95, 853–860.
Dische Z. 1953. Qualitative and quantitative colorimetric
determination of heptoses. Journal of Biological Chemistry
204, 983–997.
Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. 1956.
Colorimetric method for determination of sugars and related
substances. Analytical Chemistry 28, 350–356.
Eason JR, de Vré L. 1995. Ethylene-insensitive floral senescence
in Sandersonia aurantiaca (Hook.). New Zealand Journal of
Crop and Horticultural Science 23, 447–454.
Eason JR, de Vré LA, Somerfield SD, Heyes JA. 1997.
Physiological changes associated with Sandersonia aurantiaca
flower senescence in response to sugar. Postharvest Biology
and Technology 12, 43–50.
Eason JR, Johnston JW, de Vré L. 2000a. Reversal of glyphosate
inhibition of Sandersonia aurantiaca flower senescence with
aromatic amino acids. Postharvest Biology and Technology
18, 81–84.
Eason JR, Johnston JW, de Vré L, Sinclair BK, King GA. 2000b.
Amino acid metabolism in senescing Sandersonia aurantiaca
flowers, cloning and characterization of asparagine synthetase
and glutamine synthetase cDNAs. Australian Journal of Plant
Physiology 27, 389–396.
Eason JR, Webster D. 1995. Development and senescence of
Sandersonia aurantiaca (Hook.) flowers. Scientia Horticulturæ
63, 113–121.
Fischer RL, Bennett AB. 1991. Role of cell wall hydrolases in
fruit ripening. Annual Review of Plant Physiology and Plant
Molecular Biology 42, 675–703.
Fry SC, Smith RC, Renwick KF, Martin DJ, Hodge SK,
Matthews JM. 1992. Xyloglucan endotransglycosylase, a
new wall loosening enzyme activity from plants. Biochemical
Journal 282, 821–828.
Gross KC, Sams EC. 1984. Changes in cell wall neutral sugar
composition during ripening: a species survey. Phytochemistry
23, 2457–2461.
Huber DJ. 1992. The inactivation of pectin depolymerase
associated with isolated tomato fruit cell wall: implications
for the analysis of pectin solubility and molecular weight.
Physiologia Plantarum 86, 25–32.
Kooiman P. 1960. A method for the determination of amyloid in
plant seeds. Recueil des Travaux Chimiquese de Pays-Bas
(Journal of the Royal Netherlands Chemical Society)
79, 675–678.
Michael, AJ. 1996. A cDNA from pea petals with sequence
similarity to pollen allergen, cytokinin-induced and genetic
tumour-specific genes: identification of a new family of related
sequences. Plant Molecular Biology 30, 219–224.
Mitcham EJ, Gross KC, Ng TJ. 1989. Tomato fruit cell wall
synthesis during development and senescence. Plant
Physiology 89, 477–481.
O’Donoghue EM, Somerfield SD, de Vré LA, Heyes JA. 1997.
Developmental and ripening-related effects on the cell wall of
pepino (Solanum muricatum) fruit. Journal of the Science of
Food and Agriculture 73, 455–463.
O’Donoghue EM, Somerfield SD, Sinclair BK, Coupe SA. 2001.
Xyloglucan endotransglycosylase: a role after growth cessation in harvested asparagus. Australian Journal of Plant
Physiology 28, 349–361.
O’Donoghue EM, Somerfield SD, Sinclair BK, King GA. 1998.
Characterization of the harvest-induced expression of
b-galactosidase in Asparagus officinalis. Plant Physiology
and Biochemistry 36, 721–729.
Panavas T, Rubinstein B. 1998. Oxidative events during
programmed cell death of daylily (Hemerocallis hybrid)
petals. Plant Science 133, 125–138.
Panavas T, Reid PD, Rubinstein B. 1998. Programmed cell
death of daylily petals: activities of wall-based enzymes and
effects of heat shock. Plant Physiology and Biochemistry
36, 379–388.
Phillips Jr HL, Kende H. 1980. Structural changes in flowers
of Ipomoea tricolor during flower opening and closing.
Protoplasma 102, 199–215.
Redgwell RJ, Fischer M, Kendal E, MacRae EA. 1997.
Galactose loss and fruit ripening: high-molecular-weight
arabinogalactans in the pectic polysaccharides of fruit cell
walls. Planta 203, 174–181.
Redgwell RJ, Fry SC. 1993. Xyloglucan endotransglycosylase
activity increases during kiwifruit (Actinidia deliciosa) ripening. Plant Physiology 103, 1399–1406.
Rubinstein B. 2000. Regulation of cell death in flower petals.
Plant Molecular Biology 44, 303–318.
Tanimoto E, Igari M. 1976. Correlation between b-galactosidase
and auxin-induced elongation growth in etiolated pea stems.
Plant and Cell Physiology 17, 673–682.
Updegraff DM. 1969. Semimicro determination of cellulose in
biological materials. Analytical Biochemistry 32, 420–424.
Sandersonia cell walls
van Doorn WG. 2001. Categories of petal senescence
and abscission: a re-evaluation. Annals of Botany 87,
447–456.
van Doorn W, Harkema H, Song JS. 1995. Water relations and
senescence of cut Iris flowers—effects of cycloheximide.
Postharvest Biology and Technology 5, 345–351.
Wang ZY, MacRae EA, Wright MA, Bolitho KM, Ross GS,
Atkinson RG. 2000. Polygalacturonase gene expression in
523
kiwifruit: relationship to fruit softening and ethylene production. Plant Molecular Biology 42, 317–328.
Wiemken-Gehrig V, Wiemken A, Matile Ph. 1974. Mobilization
von Zellwandstoffen in der welkenden Blüte von Ipomoea
tricolor Cav. Planta 115, 297–307.
Wood PM, Siddiqui IR. 1971. Determination of methanol and its
application to measurement of pectin ester content and pectin
methyl esterase activity. Analytical Biochemistry 39, 418–428.