Update: Cell wall metabolism and fruit ripeningsoftening

1
Update Cell wall metabolism and fruit
ripening/softening This brief slide show will
give an idea about the things that researchers
are doing currently in order to understand the
relationship of cell wall change to the
terminal development of fruits. New enzymes
have been identified and are being looked
for. Manipulations of the presence of single
enzymes have had limited impact, so the
cooperation of enzymes in polymer breakdown is
explored. Links between ion localization and
polymer solubility are being examined. New
analytical approaches, including studies of
natural and engineered mutants, are being used.
2
Pectin Lyase has been considered to be a
microbial pectin- degrading enzyme, not a plant
enzyme. Things are different now! Marin-Rodriguez
et al. (2002) Pectate lyases, cell wall
degradation and fruit softening. J. Exp. Bot.
532115-2119. (review 9a) Jimenez-Barbudez et
al. (2002) Manipulation of strawberry fruit
ripening by antisense expression of a pectate
lyase gene. Plant Physiol. 128751-759.
(ref.49a) Pectin lyase is an enzyme that breaks
polymer backbone glycosidic linkages in
a-1,4-linked galacturonans. The location of
cleavage is like that for PG, but the reaction
mechanism for the cleavage is trans- or
b-elimination. The reaction creates a
4,5-unsaturated galacturonosyl residue at the
non-reducing end of one of the shorter
galacturonan products generated in the reaction.
3
Where pectinases act on a simple (HG) pectin
4
Earlier work had shown a sequence in ripening
strawberry that encoded a PL. Jimenez-Barbudez
et al. developed a construct of the antisense
orientation of this PL sequence driven by the 35S
CaMV promoter to transform strawberries. The
result was a number of Apel lines that showed
altered PL gene expression.
While the data show some variability, fully ripe
fruit have higher external and internal firmness
than controls (i.e., those not transformed or
transformed only with the GUS-encoding sequence).
Internal firmness readings are based on
measurements at points where epidermal tissue was
peeled away. The greater the fruit size, the
greater the disparity in external firmness
between Apel and control fruit (below).
5
The different Apel lines showed differing degrees
of suppression of PL gene expression, altered
pectin metabolism, and ripe fruit showed less
leaky membranes (based on electrolyte leakage
from excised fruit cylinders, data not shown).
The left panels show Northern analysis (A B)
looking at strawberry PL gene expression and (C)
a Western blot using anti-strawberry PL serum.
The table (above) shows the extent of swelling of
isolated cell walls and the amount of cell wall
pectin isolated from cell walls prepared from
Apel and control lines.
6
The firmness values of the Apel fruit diverge
from values for controls as ripening proceeds
(left figure) and Apel fruit show improved
firmness retention after ripe fruit are held for
4 days at 25C (right figure).
There are PL genes in many species that yield
edible fruits. The list includes grapes, bananas
and tomatoes (in ripening fruit).
7
Studies of the effects of manipulation of genes
encoding presumptive wall-modifying enzymes in
ripening tomato fruits have been extensive (see
review 2a by Brummell and Harpster (2001) Cell
wall metabolism in fruit softening and quality
and its manipulation in transgenic fruits. Plant
Mol. Biol. 47311-340). 1. Antisense
suppression of PG expression during ripening
retards softening, but only once fruit have
reached the red ripe stage. 2. Sense
construct-based suppression of fruit expansin
(Exp) gene expression delays softening in its
early stages. Models of Exp action have led
researchers to conclude that Exp plays a role in
the breakdown of xyloglucan polymers through
cooperation with enzymes like xyloglucan
endotransglucosylase (XET) and endo-b-1,4-glucanas
es (EGases) see review 11, but the effect of
Exp gene suppression in tomato fruit is
accompanied by a somewhat confusing effect on
pectin metabolism during ripening (ref. 13)
. Does the examination of presently accepted
cell wall models help us to understand what is
going on in tissues as cell wall metabolism is
modified?
8
Review 11, based on the considerable work done on
wall metabolism and softening and tomato,
proposes a model for enzyme cooperation in the
metabolism of cell walls during growth and their
disassembly during fruit ripening. The model
suggests that the timing of metabolism of the
cellulose-xyloglucan and pectin networks is not
over-lapping.
9
The cell wall has two coextensive
polysaccharide networks they are the cellulose -
xyloglucan network and the pectin network.
The cell wall also has porosity. That is, the
gaps not filled by wall polymers or the water
shells that surround them define the sizes of
polymers that can diffuse through the wall.
10
We should also be aware of the structures of the
molecules that might (1) contribute to fruit
texture and (2) be metabolized to result in fruit
softening during ripening.
HGA is a simple pectin. Its backbone bears no
branches. Its building-block, backbone sugars
are all galacturonosyl (GalUA) residues that are
linked a-1,4- to neighbors. The polymer will be
digested by PG or PL, but the presence of methyl
esters on GalUA carboxyl groups can influence
this. HGAs are thought to be held in the wall
via bridges through Ca2 ions to other HGAs.
11
The typical cell wall also contains more complex,
RG-type pectins. RG-I is shown. Its backbone
has alternating residues of GalUA and the neutral
sugar Rhamnose.
RG-I also has a large array of possible side
chains. They contain the 6 carbon neutral sugar
Galactose (Gal) the 5 carbon neutral sugar
Arabinose (Ara). Breaking the RG-I backbone
would require a Rhamno-galacturonase (RGase) but
the RGase could only act if it has access to the
backbone. Access might be provided if
a-arabinosidase and/or b-galactosidase were to
remove some side chains. These enzymes could also
act to increase wall porosity
12
RG-II is even more complex. Its backbone might
be cleaved by PG or PL, but many enzymes would be
required to provide access to that backbone.
13
In the past several years researchers have begun
to examine the roles of enzymes that cooperate
in the digestion of pectins to see if greater
understanding and control of ripening-related
softening would result. Smith et al. (2002)
Down-regulation of tomato b-galactosidase 4
results in decreased fruit softening. Plant
Physiol. 1291755-1762 (ref. 85a). Sozzi et al.
(2002) Gibberellic acid, synthetic auxins, and
ethylene differentially modulate
a-arabinofuranosidase activities in antisense
1-aminocyclopropane-1-carboxylic acid synthase
tomato pericarp disks. Plant Physiol. 129930-940
(ref. 86b). These papers recognize that many
glycosidase enzymes are present in isoforms
encoded by gene families. Smith, Gross et al.
have already characterized the tomato
b-galactosidase gene family. The work of Sozzi in
our lab has just recently made clear that there
is a family of a-arabinosidases in tomato fruits
and we (Powell, Lurie, Brummell, Greve and
Labavitch) are developing a research proposal to
characterize them.
14
In Smith et al. (2002), expression of the
ripening fruit b-galactosidase (b-Gal II) was
suppressed with an antisense construct. This is
the only tomato b-galactosidase that will act on
a naturally-occurring tomato galactan polymer
that is, most likely,a polymer that is attached
as a side chain on an RG-I pectin.
The Northern analysis (above) shows that b-Gal II
comes up in normal fruits (A P) but the mRNA
level is relatively reduced in some of the
transgenic lines 3 days after breaker stage (B3),
but often not reduced 7 days after breaker (B7).
The activity against a model substrate (right,
panel A)is often greater in transgenics than
controls, but, at B3, the activity against the
side chain mimic substrate (panel B) is
substantially reduced not, however, at B7.
15
The overall reduction in cell wall Gal content
that is seen in normally ripening tomatoes is
seen in the transgenic lines also. The
manipulations clearly do not alter the general
metabolism of Gal-containing wall molecules in
ripening fruit. Nevertheless, a few of the
transgenic lines did have firmer fruit, based on
compression tests.
Panel A shows data from a flat plate compression
test, panel B shows data from a spherical
indenter compression test. If polymer cross
links affect cell wall strength and wall strength
relates to fruit firm-ness, how can integrity of
wall Gal-containing polymers affect firmness?
After all, the polymers with Gal are present as
side chains that are not cross linking. Could
the galactans influence wall porosity or the
access of pectin backbones to enzymes that might
otherwise alter the backbones and cross-links,
and, hence, wall integrity? We (Powell, Brummell,
Greve and Labavitch with Smith and Gross) are
currently developing a plan to suppress the
expression of both b-gal II and PG in tomatoes to
see if the effect reported in Smith et al. can be
enhanced.
16
Gabriel Sozzi, a Ph.D. student at the University
of Buenos Aires, came to our lab with an interest
in tomato fruit a-arabinosidases (a-Afs). a-Afs
cleave terminal Ara residues from polysaccharides
and glycoproteins therefore they might act to
trim side chains of Ara residues from backbones
of polymers like RG-I.
Gabriel showed that a-Af activity (fresh weight
basis) declined over the course of fruit growth
(left figure) and then rose again as the fruit
ripened (right figure, note y-axis scale
difference). He studied both wild-type tomatoes
and those with antisense suppression of ACC
synthase (ACC-S). Activity of a-Af did not
increase in ACC-S fruit unless they were treated
with ethylene (ref. 86a).
17
He was able to separate, based on their size,
three distinct a-Af activities from extracts of
early ripening fruits. The three represented
only a small amount of the extracted protein
(upper left panel).
Again using WT and ACC-S fruit, he showed that
a-Af-1(triangles) was low through development,
a-Af-2 fell steadily as fruit grew to full size
(squares), and that (lower right panel) a-Af-3
appeared early in ripening and rose steadily in
WT fruit (filled diamonds), but not in ACC-S
fruit (open diamonds) unless they were treated
with ethylene ().
18
Gabriel adapted the excised pericarp disc fruit
ripening system that we have used for several
years for use with the ACC-S fruit and then
tested the effects of synthetic auxins, GA3, and
ethylene on the a-Af isoforms. The disks of
ACC-S fruit and discs did not change color
(ripen) unless they were treated with ethylene
(lower left panel). Changes in the activities of
a-Af -1, 2 and 3 in MG disks treated with the
hormones are shown in panels A, B C,
respectively. The different responses to hormones
and different patterns of change in growing and
ripening fruits convince us that there are at
least 3 a-Afs in tomato and that a-Af-3 is
ripening-specific (ref. 86b).
19
Manipulation of changes in a-Af gene expression
could affect softening in ripening tomato fruit,
just as manipulation of the expression of
b-galactosidase gene expression does. The main
roles of these enzymes may be in cooperating with
pectin polymer backbone-digesting enzymes in more
effective digestion of crucial bonds in pectin
polymers. One reason that attempts to genetically
modify the extent of ripening-related fruit
softening have not been more successful could be
that both wall polymer systems (cellulose-XG and
pectins) are involved in conferring wall strength
and so the metabolism of both must be considered
when trying to modify wall weakening. A group of
researchers in Vegetable Crops (Gurrieri,
Bennett, Powell), Food Science (Kalamaki), and
Pomology (Brummell, Greve, Struijs, Van linden
and Labavitch) has been examining the effects of
simultaneous down-regulation of PG and Exp in
tomato fruit on fruit texture characteristics.
The effect on fruit softening is not clear
(relative to the lines suppressed in PG or Exp
gene expression), but processed tomato products
from the doubly suppressed fruits have improved
consistency.
20
You may recall that strawberry fruit ripening was
associated with solubilization and size reduction
of hemicelluloses, including XG. Woolley et al.
(ref. 97b) have purified an endo-b-1,4-glucanase
(EGase) from strawberry fruit and cloned its gene
(cel1). The protein will catalyze hydrolysis of
tamarind XG and the artificial cellulase
substrate, CM-cellulose. They purified the EGase
using cellulose affinity chromatography, meaning
that they bound extracted proteins to a cellulose
matrix and then eluted it by washing the column
with buffer containing cellobiose. This protocol
did not allow them to collect the EGase encoded
by cel2. Antisense and sense suppression of
cel1 expression was accomplished but there was no
impact on overall extractable EGase (assayed with
the CM-cellulose substrate) activity or fruit
softening. Presumably the product of cel2
expression was able to compensate for the
down-regulation of one EGase gene. Strawberry
fruit also express a ripening-related Expansin
gene (ref. 24).
21
Harpster et al. (2002, ref 40b) identified an
EGase in pepper fruits. The gene, CaCel1, was
normally expressed with ripening and expression
was suppressed by driving expression of a
truncated version of the CaCel1 gene (sense
suppression under conditions where no expression
of a functional gene is possible). A few lines
with suppressed sense mRNA (lower panel) and
decreased CaCel1 protein and enzyme activity
(assays against CMC ) in red ripe fruit (upper
panels) were identified. Fruit softening was
not affected by the manipulation and there was no
reduction in the normal ripening-related
digestion of hemicellulose polymers (next slide).
22
Pepper fruit pectins
Base-soluble polysaccharides
Neither the pectin size change nor the digestion
of hemicelluloses was altered in the reduced
EGase- peppers (panels A D, green B E,
mature red and C F, over-ripe fruit)
The panel on the left shows the impact of
incubating proteins from control and CaCel1
over-expressing peppers purified using cellulose
affinity columns with the base-soluble
polysaccharides from green pepper fruits. Assays
(Y-axis values) show total sugar distributions
over fractions (panels A C) or xyloglucan size
distributions (panels B D, iodine-binding
assay). Size distributions for undigested
substrates are shown in the upper panels, the
lower panels show distributions following a 6 h
enzyme treatment.
23
I have commented earlier that focusing on cell
wall metabolism, only, probably should not
provide total control of fruit softening because
turgor pressure has an important part in the
process. However, I suspect that we dont really
know all of the enzymes that might be important.
Recall that PL has just been discovered as a
plant enzyme that plays a potentially important
role in fruit softening. In a later lecture you
will learn why my lab is interested in fruit
pectin metabolism prior to the start of ripening.
There are many tomato fruit PG ESTs, some
present during fruit development, but only the
ripening-related fruit PGs have received
attention. Furthermore, PGs may not be the only
relevant tomato pectin polymer-cleaving enzyme.
Shown are SEC separations of pectins isolated
from green tomato fruits. The blue dot plot
shows the undigested polymers and the open
circles show the result of incubating the
polymers with a protein extract from small green
fruit, fruit that have no PG activity. The
Y-axis shows the relative uronic acid content of
fractions. We think that this digestion reflects
the action of several fruit enzymes, including
RGase, an enzyme of fungi that has been rarely
reported in plants (see ref. 37, like PL,
formerly).
24
In early lectures we discussed the roles of
inorganic elements in the structure of cell
walls. For pectins, the roles of Ca2 and borate
as cross linking (presumable wall
integrity-conferring) components seems clear.
Pectin solubility (whether or not the pectin
backbones are cleaved) is a feature of wall
change during ripening of many fruits. We are
now asking whether boron dynamics are important
in ripening. Do RG-II - B complexes (dimers of
RG-II) exist in fruits. Does the proportion of
dimers change with ripening? Is there a chemical
or biochemical way that cross-linking of RG-II
monomers could be altered during ripening? These
issues have not been systematically
addressed. Food scientists and processors know
that addition of Ca2 to fruits will increase
their firmness and my lab (Mignani et al., ref.
59) has shown that addition of Ca2 to tomato
fruit disks at almost any stage of ripeness will
increase their firmness, affecting both pectin
solubility and turgor pressure. Are there
mechanisms in fruits that might disrupt
homogalacturonan- Ca2 bridges as ripening
procedes? What effect would a change in pectin
methyl esterification have on the extent of Ca2
bridging? Could the appropriate changes be
accomplished in the apoplast?
25
Huxham et al. (1999) have used electron-energy-los
s spectroscopy (EELS) to image the presence and
distribution of various elements within fruit
tissues (ref. 46a).
The figure above shows the firmness of apples
from 2 orchards in the UK. Measurements were
made at harvest and at monthly intervals while
fruit were stored at 3.5C in low O2 high CO2.
In the images at the right, Ca2 distribution in
walls is seen (trust me!) using EELS. Before
analysis the tissue is frozen and the water
distilled away to limit movement of the ion.
Although high concentrations of Ca2 are often
seen in middle lamellas and at cell junctions,
this is not always the case.
26
Surprisingly, the fruit with the more substantial
Ca2 presence were not the firmest fruit. The
measurements made by Huxham et al. indicated that
N presence was more tightly correlated with
firmness. What cell wall components would be
rich in N?
Analysis of wall hydroxyproline content in fruit
showed that both at harvest (Expt. 2) and after 6
months storage (Expt. 1) the fruit from orchard
A (clear bars) had higher Hyp. The relationship
of fruit firmness to wall protein presence,
integrity and distribution has not been examined,
in general.
Orchard A had the firmer fruit, orchard B fruit
had the higher Ca to C ratio and lower N to C
ratio.
27
As in much of todays plant biology, study of
mutant phenotypes can reveal a great deal about
the biochemistry and cell biology of
developmental processes. The best mutants to use,
from the standpoint of making clear
interpretations, are those with a specific lesion
affecting the expression of a single function at
the level of wall metabolism. However, most
ripening mutations are pleiotropic, affecting
production of transcription factors (tomato rin),
ethylene perception (tomato Nr) etc. A newer
tomato ripening mutant that has received
attention is Cnr. A lot of work on the
disruption in ripening-related wall metabolism
has been done with Cnr. (See refs. 63a, 63b, 89a)
Panel A shows WT fruit (left columns) and Cnr
fruit (right columns) as they ripen (top to
bottom rows). Some red color is seen inside the
Cnr fruit. In panel B the skin is peeled back to
show how little lycopene forms in Cnr flesh.
Panel C shows fruit after treat-ment with 100
mLL-1 ethylene. Only slight external red color
development is seen. Panel D shows pairs of
seedlings on the left are Cnr, next are WT, then
Cnr after exposure to ethylene, then (on the
right) Nr seedlings not responding to ethylene.
28
Walls/tissues of Cnr fruit are stronger before
and after ripening.
When Cnr fruit ripen cell wall preps do not tend
to swell relative to those of WT (A). Redgwell
(ref. 69a) has reported that walls from ripe
fruit swell in water, perhaps because of reduced
wall cross-links that could limit water swelling
in response to re-hydration. When fractured,
Cnr tissues tend to split between cells, rather
than through cells (B). This is a characteristic
of fruit that become mealy during ripening.
29
Use of some of the antibodies we know (and a new
one, PAM1 that recognizes blocks of unesterified
HGA)show differences in wall polymer
distributions. Focus on panels A-D. PAM1 lights
up corners of a triangular cell junction point in
WT walls, not in Cnr (A B). If the sections are
first base-treated to remove Me-esters and then
PAM1 is used, a prominent band of HGA is seen in
the WT middle lamella (C), but not in Cnr (D).
One conclusion would be that Cnr lacks the normal
amount and distribution of wall HG.
LM6, which binds a-1,5-arabinans shows that the
distribution of arabinans is relatively uniform
in WT walls (A), but spotty in the walls of Cnr
fruit (B). The arrow shows a membrane-bound
vesicle filled with LM6-positive material just
inside of the plasmalemma of the Cnr cell.