Two Divergent Endo-P-I ,4-glucanase Genes Exhibit Overlapping Expression i n Ripening Fruit and Abscising
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
The Plant Cell, Vol. 6, 1485-1493, October 1994 O 1994 American Society of Plant Physiologists
Two Divergent Endo-P-I,4-glucanase Genes Exhibit
Overlapping Expression i n Ripening Fruit and Abscising
Coralie C. Lashbrook, Carmen Gonzalez-Bosch,’ and Alan B. Bennett2
Mann Laboratory, Department of Vegetable Crops, University of California, Davis, California, 95616
Two structurally divergent endo-p-l,Cglucanase (EGase) cDNAs were cloned from tomato. Although both cDNAs (Cell
and Ce12) encode potentially glycosylated, basic proteins of 51 to 53 kD and possess multiple amino acid domains con-
served in both plant and microbial EGases, Cell and Ce12 exhibit only 50% amino acid identity at the overall sequence
level. Amino acid sequence comparisons to other plant EGases indicate that tomato Cell is most similar to bean abscis-
sion zone EGase (680/), whereas Ce12 exhibits greatest sequence identity to avocado fruit EGase (57%). Sequence
comparisons suggest the presence of at least two structurally divergent EGase families in plants. Unlike ripening avocado
fruit and bean abscission zones in which a single EGase mRNA predominates, EGase expression in tomato reflects the
overlapping accumulation of both Cell and Ce12 transcripts in ripening fruit and in plant organs undergoing cell separa-
tion. Cell mRNA contributes significantly to total EGase mRNA accumulation within plant organs undergoing cell separation
(abscission zones and mature anthers), whereas Ce12 mRNA is most abundant in ripening fruit. The overlapping expres-
sion of divergent EGase genes within a single species may suggest that multiple activities are required for the cooperative
disassembly of cell wall components during fruit ripening, floral abscission, and anther dehiscence.
INTRODUCTION
Cell wall disassembly is a significant component of many phys- extensively modified (Gross and Wallner, 1979; Huber, 1983).
iological and developmental processes, including vegetative The solubilization of pectins by the ripening-induced enzyme
growth, fruit ripening, and the abscission of plant organs. The polygalacturonase (PG) has been most extensively studied.
dissolution of cell wall polysaccharides during these events However, PG-mediated pectin degradation alone cannot ac-
requires the activity of multiple cell wall hydrolases, each po- count for the significant changes in tomato fruit texture that
tentially responding to a unique subset of developmental, accompany ripening. In transgenic tomato plants, reduction
hormonal, and environmental cues. The mobilization of cell of PG mRNA levelsto 0.10/0 of wild-type levelssignificantly alters
wall hydrolases for degradative processes is regulated in part the degree of pectin polymerization observed in ripening fruit
via changes in gene expression (Fischer and Bennett, 1991) cell walls (Smith et al., 1990), which affects a variety of fruit
and permits the modulation of architectural constraints required quality parameters (Kramer et al., 1990; Schuch et al., 1991)
for transitions in plant growth and development. In some sys- but does not prevent fruit softening (Smith et al., 1990). Con-
tems, cell wall metabolism has the capacity to generate versely, expression of a chimeric PG transgene in the tomato
biologically active pectic Vong et al., 1986; Brecht and Huber, ripening mutant rin resulted in pectin solubilization and depo-
1988; Campbell and Labavitch, 1991) and hemicellulosic (York lymerization at near wild-type levels but not fruit softening
et al., 1984; Farkas and Maclachlan, 1988; McDougall and Fry, (Giovannoni et al., 1989; DellaPenna et al., 1990). The main-
1988) cell wall fragments capable of further regulatory activ- tenance of normal rheological properties of ripe tomato fruit
ity. Thus, cell wall disassembly has both a structural and in the absence of PG-mediated pectin degradation indicates
informational role in plant development. that textura1changes associated with ripening are dependent
Tomato fruit ripening has proven to be an excellent model upon cell wall dissolution processes catalyzed by other en-
system for studying the contributions of individual hydrolases zymes. Analysis of cell walls isolated from ripening tomato fruit
to plant cell wall modifications. During ripening, both pectic has not detected the release of oligosaccharide fragments de-
and hemicellulosic components of tomato cell walls are rived from crystalline cellulose (Gross and Wallner, 1979;
Ahmed and Labavitch, 1980), whereas solubilization of cell wall
Present address: Departamento de Bioquimica y Biologia Molecu- components believed to be derived from the hemicellulosic
lar, Facultad de Biologicas, Burjassot, Valencia 46100, Spain. matrix is evident (Huber, 1983). The involvement of endo-
* To whom correspondence should be addressed. p-1,4-glucanases (EGases; p (1:4] 4-glucan hydrolase; EC1486 The Plant Cell
3.2.1.4) in hemicellulosedegradation during vegetative growth
and fruit ripening has been implicated (Hayashi et al., 1984;
Hatfield and Nevins, 1986). Although the term “cellulase” has
been widely used to describe these endoglucanases, the term
is misleading in view of the current lack of evidence for EGase-
cata1yzed cellulose degradation.
The presence of EGase activity in ripening tomato fruit (Hall,
1963,1964; Dickinson and McCollum, 1964) and in abscising
tomato flowers (Roberts et al., 1984; Tucker et al., 1984) has
been established, and a role for EGases in cell wall dissolution
processes within bean abscission zones (Tucker et al., 1988;
Tucker and Milligan, 1991) and in avocado fruit (Tucker et al.,
1987; Cass et al., 1990) is suggested by studies of EGase gene
expression. Currently, the nature of the cell wall substrate(s)
modified by these enzymes and thus their physiological func-
tion are unknown. To better assess the physiological function
of EGases in a system amenable to molecular genetic manip-
ulation, we have cloned EGase cDNAs from tomato and have
determined that they belong to a gene family with at least four
divergent members. Unlike ripening avocado fruit and bean
abscission zones, in which a single EGase mRNA predomin-
ates, EGase exprebsion in tomato reflects the overlapping
expression of structurally divergent EGase gene family mem-
bers. The structure and differential expression of two members
of the tomato EGase gene family, EGasel (Cell) and EGase2
(Ce12), are described in this report.
RESULTS
Two Tomato Endoglucanases Are
Structurally Divergent
-68%
cDNA clones encoding two tomato EGase genes were isolated BAC
using oligonucleotide probes correspondingto amino acid se-
quence domains conserved in bean and avocado EGases
(Tucker et al., 1987; Tucker and Milligan, 1991). In Figure l A ,
comparison of the deduced amino acid sequences of tomato
. 57%
Avocell
Cell and Ce12 revealed severa1 common features. Cell and
Ce12 cDNAs have a hydrophobic N-terminal signal sequence
characteristic of proteins targeted via the endomembrane path-
way to locations including the cell surface and contain within
their coding regions N-glycosylation consensus motifs (Asn-
X-Serflhr). Cell and Ce12 cDNAs encode mature proteins rang-
ing in predicted size from 51 to 53 kD, with basic predicted
pls of 7.3 and 8.2, respectively. Cell and Ce12 cDNAs also con-
tain a number of short, highly conserved domains located in
equivalent positions of all plant EGases sequenced thus far.
Nevertheless, an overall sequence comparison of Cell and
Ce12 reveals only 50% amino acid sequence identity. This is
the lowest identity yet observed between two plant EGases
and is consistent with our observation that Cell and Ce12
cDNAs do not cross-hybridize, even under conditions of re-
laxed stringency.Tomato Endoglucanase Gene Family 1487
Figure 1B illustrates the sequence comparisons between 0.045- a
tomato Cell and Cel2, bean abscission zone EGase, and < 0.04- • Cell
avocado fruit EGase. This dendrogram suggests that the struc- 1
£
0.035.
0.03.
Q Cel2
e
tural divergence of EGases into the two families shown
o 0.025.
occurred prior to the evolutionary divergence of major dicoty-
ledonous families. The comparison also suggests that both 1 0.02.
avocado and bean could possess members of both families. 1
<
0.015.
0.01'
£&
fc\«
>
>
Genomic DMA gel blot analysis of both species suggests that z
cr 0.005.
this is the case (Tucker and Milligan, 1991). -Tu«-w«caJlL $
Genomic DMA gel blot analysis shown in Figure 2 indicates Cell ^ ^ ^ ^ *
that Cell and Cel2 are transcribed from distinct genes of low
copy number. At both moderate (20°C below the melting tem-
perature; Tm -20°C) and high (Tm -5°C; data not shown)
*- 04 *•
stringencies, Cel2 hybridizes to a small set of large genomic o o o a i I
fragments, whereas Cell hybridizes to a different array of
smaller fragments. Restriction fragment length polymorphism Ripening Stage
analysis has assigned Cell to tomato chromosome 8 and Figure 3. Cell and Cel2 mRNA Expression over the Course of Tomato
Ce/2 to chromosome 9 (C.C. Lashbrook, K.B. Alpert, and J.W. Fruit Development and Ripening.
deVerna; unpublished data). Neither Cell nor Ce/2 maps to
Specific transcript levels, expressed as the percentage of poly(A)+
a previously defined locus.
mRNA present in tomato fruit, were quantified by densitometric anal-
ysis of autoradiographs of protected fragments from RNase protection
assays relative to transcript-specific standard curves. A direct com-
Cell and Cel2 mRNAs Accumulate in Fruit in parison of image intensities was not possible because of different
Response to Developmental and Hormonal Signals exposure requirements for the Cell and Cel2 signals.
RNase protection assays were used to assay the accumula- of tomato fruit development. Figure 3 indicates that Cell mRNA
tion of mRNAs corresponding to Cell and Cel2 over the course transiently accumulates in immature green (IG) fruit, suggesting
a possible role for Cell in fruit cell expansion, the liquefaction
of locules, or other fruit maturation events. It is likely that Cell
E = O g (5 o c1488 The Plant Cell
commenced (Figure 3), their presence subsequently becom-
ing evident at the breaker stage of ripening. Thus, the timing
of both Cell and Cel2 mRNA accumulation follows the initia-
tion of ethylene synthesis. The relationship between Cell
mRNA induction and ethylene synthesis is complicated by the
transient expression of Cell mRNA during the late phases of
fruit maturation, prior to ethylene production. The low levels
of Cell mRNA detected in the MG1 and MG2 stages may rep-
resent low levels of transiently expressed Cell mRNA from
the IG2 stage persisting into the earliest stages of fruit ripen-
Cell
ing. Thus, Cell appears to be developmentally as well as
hormonally regulated.
Cel2
To further assess the potential regulation of EGase expres-
sion by ethylene, ripening tomato fruit were treated with the
ethylene action inhibitor 2,4-norbornadiene (NBD), a compet-
itive inhibitor of the ethylene receptor (Sister and Yang, 1984).
Figure 4 shows that the accumulation of mRNAs from both
Cell and Cel2 was severely inhibited in NBD-treated fruit tested Figure 5. Tissue Distribution of EGase mRNAs.
at two ripening stages, suggesting that the accumulation of Cell and Cel2 mRNA levels in various tomato tissues were quantified
Cell and Cel2 mRNAs is regulated by ethylene. The inhibi- by RNase protection analysis. Fruit RNA was isolated from the pericarp
tory effect of 2000 nL L~1 NBD on Cell and Cel2 mRNA at three stages of ripening. Vegetative tissues were harvested from
accumulation was reversed by treatment with 1000 u.L L"1 eth- young field-grown tomato plants. Floral tissues were isolated from field-
ylene (data not shown). grown tomato flowers on the day of anthesis. Abscission zones were
isolated from floral explants that abscised (+ ab) or failed to abscise
(- ab) after being held 4 days in water-filled tubes. mRNA abundance
Cell and Cel2 mRNAs Are Differentially Expressed in is expressed as the percentage of poly(A)* mRNA present in each
Fruit and Flowers tissue.
Both Cell and Cel2 mRNA transcripts are present in a wide
variety of tomato plant tissues, although at quite different levels of abundance. As shown in Figure 5, ripening tomato fruit have
a high level of total EGase mRNA, which is contributed mainly
by Cel2. In contrast, abscission zones and anthers also have
significant levels of total EGase mRNA, comprised mainly of
Cell mRNA. Floral abscission and anther dehiscence are func-
Q Q Q tionally similar events, involving the separation of cell layers,
m ffi CD and it is possible that Cell-catalyzed cell wall degradation is
z z z
1
an integral part of both processes. Alternatively, Cell may par-
+ +
I I I ticipate in cell wall degradative processes associated with
earlier stages of anther development. In sweet pea, two EGase
f
Probe: Cell
isoforms purified from developing anthers appeared to be iden-
tical to isoforms purified from abscising flower buds (Sexton
et al., 1990). Because sweet pea EGase accumulation preceded
pollen grain maturation, a potential role for the enzyme in the
breakdown of walls including those surrounding the tapetum
was suggested.
m. Ethylene Accelerates EGase mRNA Accumulation and
Probe: Cel2 Flower Abscission
Figure 4. Effect of NBD on EGase Expression during Fruit Ripening. Cell and Cel2 mRNA levels were assayed over the course of
MG fruit of equivalent internal ethylene content were treated with 2000 floral abscission to determine the temporal relationship exist-
nL L~1 NBD (+) until nontreated controls (-) reached the ripening ing between the abscission process and the expression of
stages Brk + 4d (4 days after breaker stage) and overripe (4 days af- specific EGases. Maximum abscission of tomato flowers held
ter full red stage). Total RNA isolated from fruit was subjected to RNase in the presence or absence of 10 u,L L~1 ethylene occurred
protection analysis. after 4 days in air or after 2 days in ethylene as shown in FigureTomato Endoglucanase Gene Family 1489
6. This is consistent with numerous observations in other spe-
Table 1. Effect of NBD on EGase mRNA Accumulation in
cies that ethylene accelerates the abscission process (reviewed Tomato Flower Abscission Zones
by Sexton et al., 1985). Figure 6 indicates that Cell mRNA
was completely absent at anthesis but accumulated within EGase mRNA
the flower abscission zones held in air as cell separation pro- Abundance
Abscission
(°/o Control)
gressed. As complete abscission was approached, Cell mRNA Qata
levels declined. Cel2 mRNA, which was detectable at anthe- Treatment (% Abscised) Cell Cel2
sis, accumulated further at the onset of abscission and, like Air (control) 66 100 100
that of Cell, declined in abundance as complete abscission Air and NBD, 9 1 0
was approached. Ethylene accelerated the abscission of tomato 2000 nl I'1
flowers and accelerated the rate of accumulation of both Cell C2H4, 15 uL L-', and 71 107 100
and Cel2 mRNAs to their maximum levels. Although ethylene NBD, 2000 uL L-1
treatment of flowers accelerated the timing of EGase mRNA Flowers were subjected to designated gas treatments in sealed jars
accumulation, it did not increase the maximum level of Cell for 4 days and scored for abscission by inflection of pedicels at their
or Cel2 mRNA ultimately accumulated. abscission zones. Specific EGase transcript levels in mRNA prepared
Table 1 provides additional evidence for the dependence from at least 100 abscission zones were quantitated by RNase pro-
of Cell and Cel2 EGase expression on ethylene in abscission tection analysis.
zones. Tomato flower abscission was significantly reduced, and
the accumulation of both Cell and Cel2 mRNA was virtually
abolished in abscission zones of flowers treated with 2000 nL Tomato EGase mRNA Levels Are Induced by Cell
L~1 NBD. Both abscission and EGase mRNA abundance in Separation and Differentially Localized in Abscission
the presence of NBD were subsequently restored to control Zones and Flanking Pedicel Tissue
levels by the addition of 15 u,L L~1 ethylene, demonstrating
that NBD effects were completely reversible by ethylene. Taken In the abscission-related experiments described above, we ex-
together, the data in Figure 6 and Table 1 indicate that both amined abscission zone EGase mRNA levels isolated from
Cell and Cel2 mRNA levels are regulated by ethylene and are populations of flowers at defined time points, with each popu-
temporally correlated with the abscission process. lation containing differing proportions of both abscised and
nonabscised members. To determine the temporal relation-
ships between tomato Cell and Cel2 mRNA accumulation and
the actual separation of the flower abscission zone cell layers,
•*» -+ CEL1 »- specific EGase transcript levels within abscised and nonab-
scised abscission zones selected from a flower population
exhibiting ~50% abscission were compared in Figure 7. In
-» GEL 2 *- * ^» «, ^» addition, to determine if tomato Cell and Cel2 mRNA accumu-
100
lation extended beyond the region of cells actually undergoing
separation, mRNA levels were quantitated in noncontiguous
pedicel tissue flanking the abscission zone. Cell and Cel2
mRNAs were present at basal levels in nonabscised tissue
§ but increased ~20- and 30-fold, respectively, in the abscis-
5 sion zones of abscised flowers. Cel2 mRNA accumulation was
40
jl 40 primarily limited to abscission zones, where it constituted up
20
to 0.003% of the mRNA. Cel2 mRNA was also detectable in
% 20 proximal pedicel tissue, although at levels ~10-fold less than
in the abscission zone. In contrast, Cell mRNA in abscised
0 1 2 3 4 0 0.5 1.0 1.5 2.0 abscission zones constituted over 0.01% of the mRNA pres-
DAYS IN AIR DAYS IN ETHYLENE ent. Notably, comparably high levels of Cell mRNA were
Figure 6. EGase mRNA Accumulation during Tomato Flower observed in abscised abscission zones and in the distal pedicel
Abscission. tissue that was not undergoing cell separation. The abscis-
sion of tomato flowers has been shown to be initiated in
Floral explants harvested within a day of anthesis were held in water-
parenchymatous cells distal to cells that later form the sepa-
filled tubes for 4 days in air or 2 days in ethylene. At designated times,
the percentage of abscission of 100 flowers was determined by inflec- ration layer in both ethylene-treated (Roberts et al., 1984) and
tion of floral pedicels at their abscission zones. mRNA prepared from non-ethylene-treated tomato flowers (Jensen and Valdovinos,
these populations, containing both abscised and nonabscised flowers, 1967). In tomato flowers, significant EGase enzyme activity has
was subjected to RNase protection analysis. A direct comparison of been measured in extracts of abscission zones and distal tis-
signal intensities was not possible because of different exposure re- sues but not in preparations from proximal tissue (Tucker et
quirements for the Cell and Cel2 signals. al., 1984). The association of Cell, but not Cel2, mRNA with1490 The Plant Cell
0.0141 those species are quite small (Tucker et al., 1987; Tucker and
Milligan, 1991). Gene cloning strategies based upon the de-
~ 0.012
tection of conserved plant EGase domains have not been used
T 0.01 in bean and avocado; thus, it is not known whether other diver-
c ^Distal D gent EGase mRNAs also accumulate in these species,
3 0.008
c ^ Abscission although genomic DNA gel blot analysis is consistent with the
•9 0.006 A Proximal presence of two divergent gene families in avocado (Tucker
and Milligan, 1991). Given the failure of tomato Cell andCe!2
0.004
to cross-hybridize with each other, genomic DNA gel blot anal-
0.002- ysis using either probe might greatly underestimate the
complexity of the tomato EGase gene family. On the basis of
substrate specificity and chromatographic separation behavior,
ft five distinct tomato EGase activities have been described in
Cel2 ripening tomato fruit pericarp and locules (Maclachlan and
Cell
Brady, 1992). At least four divergent EGase genes are pres-
Figure 7. Expression of Cell and Cel2 in Abscised and Nonabscised ent in tomato (reviewed in Brummell et al., 1994), with Cell
Tomato Flowers. and Cel2 representing major genes expressed in abscission
Several hundred tomato flowers were held in water-filled tubes until zones and fruit, respectively.
~50% had abscised as measured by inflection of each floral pedicel The high sensitivity of the RNase protection assay used in
at its abscission zone. RNA was prepared from abscission zones and this study allowed the detection of both Cell and Cel2 mRNAs,
flanking pedicel tissue of the abscised (+) and nonabscised (-) mem- both present at levels below the limit of detection by RNA gel
bers of the population, and EGase transcript levels were quantified blot hybridization. It is notable that in tomato the combined
using RNase protection assays. amount of Cell and Cel2 mRNAs rarely exceeds 0.05% of the
mRNA population. In contrast, both ripening avocado fruit and
bean abscission zones exhibit much higher levels of EGase
mRNA (Christoffersen et al., 1984; Tucker et al., 1988). Tomato
distal tissue sections suggests that the enzyme activity previ- EGase enzyme activity is also considerably lower than that
ously reported may correspond to the Cell protein product. observed in avocado fruit (Lewis et al., 1974), which is consis-
tent with the low mRNA levels that we detected.
The precise physiological role of EGases in plant develop-
ment is not known, although their involvement in cell wall
DISCUSSION
disassembly occurring during fruit ripening and abscission has
been frequently proposed. The detection of Cell and Cel2 in
We have described the primary structure and pattern of mRNA a variety of tissues throughout the tomato plant suggests that
accumulation of Cell and Cel2, two members of a multigene EGases may be associated with additional developmental
EGase family in tomato. Both Cell and Cel2 mRNAs accumu- processes that involve rearrangements in cell wall architec-
late in ripening fruit and in flower abscission zones as well ture. The overlapping accumulation of Cell and Cel2 mRNA
as at lower levels in other vegetative and floral organs. Although in ripening fruit and abscission zones suggests the possibil-
their pattern of mRNA accumulation is overlapping, Cell mRNA ity that the protein product of each mRNA has distinct
predominates in abscission zones and anthers, whereas Cel2 biochemical functions required to cooperatively bring about
mRNA predominates in ripening fruit. Cell and Cel2 exhibit required changes in cell wall architecture. Alternatively, each
low sequence identity and fail to cross-hybridize at moderate EGase may have identical properties but differ primarily in spa-
to low stringency. The structural divergence of these two tomato tial and temporal expression, sharing some overlap perhaps
EGases may reflect the evolution of functional differences. The as a built-in redundancy of function. The latter interpretation
presence of elevated levels of Cell mRNA in physiological con- is not supported by the differential EGase substrate specifici-
texts involving cell separation may be consistent with the ties that have been observed in distinct tomato fruit EGase
observation of significant amino acid sequence identity (68%) activities separated chromatographically (Maclachlan and
between tomato Cell and the bean abscission zone EGase. Brady, 1992) or by the significant differences in primary struc-
Cel2, in contrast, is most abundant in ripening fruit and ex- ture between Cell and Cel2.
hibits a greater amino acid sequence identity to avocado fruit On the basis of primary sequence comparisons, tomato
EGase (57%). Cell and Cel2 can be assigned to the E family of cellulases
The existence of multiple, divergent EGase genes in tomato (Henrissat et al., 1989; Beguin, 1990), a group of bacterial and
invokes the question of whether similar gene families may exist plant EGases whose structural similarities are believed to
in other plants. Analysis of the EGase genomic organization underlie commonalities in catalytic function. It has been sug-
in avocado and bean suggests that EGase gene families in gested that plant and microbial members of this EGase familyTomato Endoglucanase Gene Family 1491
evolved from a common ancestor (Navarro et al., 1991), yet Plant Materials and Tissue Preparation
important distinctions have evolved between prokaryotic and
eukaryotic cellulases of the E family. Perhaps most significant Fruit and flowers were harvested from field-grown tomatoes (Lycopef-
is that plant EGase sequences identified to date lack the cel- sicon esculentum cv Castlemart). Fruit were harvested at the mature
green (MG) stage and allowed to ripen in humidified20-L bottles sub-
lulose binding domains and cellulosome anchoring motifs
jected to continuous flow (20 L hr-l) of air or 10 WLL-l ethylene at
o i their microbial counterparts. Thus, there is at present no 2OOC. Air tanks contained open vials of mercuric perchlorate (250 mM
structural evidence that plant endoglucanasesare capable of in perchloric acid) to absorb ethylene. Fruit were harvested at various
participating in degradation of crystalline cellulose. It has been stages of ripening and scored by color using U.S. Departmentof Agricul-
difficult to clearly define the biochemical function of purified ture standards (U.S. Department of Agriculture, 1975) as well as by
plant EGases, possibly becauseof the unrecognized existence ethylene production iate using agas chromatographfitted with aflame
of multiple EGases in each tissue. There is evidence that ionization detector. Fruit selected for expression analysis exhibitedthe
EGases from different sources act on different hemicellulosic following ethylene production rates: 0.02 to 0.1 (mature green 1; MGl),
substrates (Hayashi et al., 1984; Hatfield and Nevins, 1986; 0.2 to 0.3 (MG2), 0.4 to 0.5 (MG3), 0.6 to 0.7 (MG4), 1 to 3 (breaker),
Durbin and Lewis, 1988). Havingsevera1EGase genes cloned 3 to 5 (turning), 5 to 8 (pink), 6 to 9 (light red), and 2 to 6 nL per g
fresh weight of tissue per hr (red). MG fruit within these defined ethyl-
from a single source now provides a unique opportunity to as-
ene evolution ranges also exhibited interna1 anatomical features
sess their specific biochemicalfunction by expression of single corresponding to stages MG1 to MG4 (Su et al., 1984). lmmature green
genes in heterologous systems. In addition, experiments to 1 (IG1) fruit was defined as fruit with an average fresh weight 25%
assess the physiological function of each tomato EGase in that of MG; IG2 fruit weight was .u50°/o that of MG. Interna1anatomy
transgenic plants with altered expression of each gene are of IG1 and IG2 fruit was surveyed to confirm immaturity of seed and
underway. locules. Overripe fruit was defined as fruit held 4 days past the attain-
ment of full red color. In all fruit samples, seeds and locular material
were removed, and pericarptissue was frozen in liquid nitrogen prior
to storage at -8OOC for subsequent analysis.
METHODS
Floral explants harvested within a day of anthesis were bright yel-
low with no evidence of flower or pedicel senescence. Floral explants
consisting of a stem bearing one.tothree flowers were trimmed of leaves
Tomato Endo-P-1,Gglucanase Cloning
and buds and then placed in water-filled tubes and treated in air or
ethylene as described above. For time course experiments, abscis-
Degenerate oligonucleotide hybridizationprobes corresponding to two
sion zones (2 to 2.5 mm) and equivalently sized dista1 and proximal
conserved domains of avocado and bean endo-P-1,4-glucanases
tissue sections located 2 to 2.5 mm away from the abscission zone
(EGases), GGYYDAGDN and CWERPEDMD(liucker et al., 1987; Tucker
were harvestedand frozen in liquid nitrogen and then stored at -8OOC.
and Milligan, 1991), were synthesized with the sequences 5'-TTA-
Abscission was scored on an absolute basis following the inflection
(G)TCICCIGCA(G)TCA(G)TAA(G)TAlCC-3' and 5'-TCCATA(G)T-
of each floral pedicel at its abscission zona
CT(C)TCIGGICGT(C)TCCCAA(G)CA-3: respectively, where I is deox-
For experiments with 2,4-norbornadiene(NBD; bicycloheptadiene;
yinosine. Oligonucleotides were 32P-radiolabeledat their 5' termini
Aldrich Chemicals), fruits or flowers were placed in sealed 20-L bot-
with T4 polynucleotide kinase (Bethesda Research Laboratories) to
tles fitted with septum-covered ports into which appropriate amounts
a specific activity exceeding 5 x 108 dpm vg -l and were used to
of liquid NBD and/or 100% ethylene were introduced to yield speci-
screen a ripe tomato fruit cDNA library in the plasmid vector pArc
fied final concentrations. Solid NaOH served as a COn scavenger.
(DellaPenna et al., 1986) using screening methods described in the
Bethesda Research Laboratories newsletter Focus (1984; volume 6,
pages 1 to 3) and elsewhere (Hanahan and Meselson, 1980). Hybridiz- RNA lsolation
ing colonies were purified, and the cloned inserts partiallysequenced.
Two of the cDNAs with significant sequence identityto EGases of bean Total RNA was isolated from floral or vegetative tissues or samples
and avocado were designated tomato Cell and Ce12. The Cell clone of at least 100 abscission zones or noncontiguousflanking tissue sec-
isolated from the primary library screen contained a complete coding tions as previously described (Chomczynski and Sacchi, 1987) with
sequence. The initial clone of Ce12 was only partia1 length and was two modifications. Water-hydratedpolyvinylpolypyrrolidonewas present
used to rescreen the library, yielding a full-length clone. Full-length in the initial grinding buffer at 0.06 g of thick slurry per mL and was
Cell and Ce12 cloned inserts released from the pArc vector by Smal subsequently removed by centrifugation prior to phenol extraction of
digestion were subcloned into pBluescript (SK+ and KS+,respec- the aqueous medium. In addition, LiCl precipitation (2 M LiCl final
tively) vectors for use in subsequent experiments. concentration)was added at the end of the procedure to remove car-
Each strand of Cell and Ce12 was sequenced by the dideoxy chain bohydrates and other A230-absorbingcontaminants. RNA purity was
termination method (Sanger et al., 1977) using sequence-specific assessed by ratios of AZe0to AZe0,which routinely exceeded 1.8. Ra-
primers complementary to domains of the cDNA or using vector se- tios of AZM)to Apgoroutinely exceeded 2.0.
quence primers for sequencing into cDNA deletions generated by Total RNA from fruit was isolated from 5 to 10 g fresh weight of fro-
exonuclease 111. Sequence data was recorded using the Microgenie zen pericarp. Tissue ground in an electric coffee grinder containing
software program (Beckman Instruments, Palo Alto, CA) and analyzed dry ice chips was added to a solution containing one part ultrapure
with the PC Gene system (IntelliGenetics, Inc., Mountain View, CA). phenol saturated with 1 M Tris, pH 9, one part chloroform-isoamylal-
Oligonucleotides were synthesized by Operon Technologies (Alameda, cohol (24/1 [v/v]), and two parts 1 M Tris, pH 9, 5% mercaptoethanol
CA) or the Universityof California-Davis Protein Structure Laboratory. (v/v) at a ratio of 0.25 g fresh weight of tissue per mL, homogenized1492 The Plant Cell
for 1 min at high speed with a Tissuemizer (Tekmar Inc., Cincinnati, data prior to publication, Dr. David Brummell for assistance with se-
OH), and centrifuged for 10 min at 10009 in a benchtop centrifuge quence analysis, Kurt Toenjes for technical assistance, and Garry
or at 12,0009 in a Sorvall centrifuge (Du Pont Co., Wilmington, DE): Pearson for providing field-grown tomato flowers. Restriction fragment
the aqueous phase was retained. Following back extraction of the or- length polymorphism analysis was conducted using facilities of
ganic phase, nucleic acids in the pooled aqueous phases were Campbells Research Institute, Davis, CA. This researchwas supported
precipitatedwith sodium acetate and ethanol. After resuspension in by U.S. Department of Agriculture National Research lnitiative Com-
0.1 M Tris, 0.1% SDS, pH 8, the crude RNA was extracted with phe- petitive Grants Program Grant No. 91-37304-6508 and a grant from
nol-chloroform (saturated in 0.1 M Tris, pH 8) and then chloroform; Unilever/ Van den Bergh Foods.
extraction was followed by sequential precipitationwith 2 M LiCl and
NHAc-EtOH; resuspension was in 300 pL of RNase-freewater. One-
half volume of absolute ethanol(l50 pL) was added, and the suspen- Received May 27, 1994: accepted August 16, 1994.
Sion was incubated 30 min on ice before microcentrifugation at full
speed for 10 min. This yielded a pellet of polysaccharide contaminants,
which was discarded. The ethanolic supernatant, containing total RNA,
was precipitated when 5 M NH4Ac (140 pL) and additional absolute REFERENCES
ethanol(900 pL) were added. Following resuspension of the pellet in
300 pL of H20,the polysaccharideremoval procedurewas repeated.
Total RNA was resuspendedin 300 pL of diethyl pyrocarbonate-treated Ahmed, A.E., and Labavitch, J.M. (1980). Cell wall metabolism in
water and was quantified spectroscopically. Ratios of AZM)to AZe0rou- ripening fruit. I. Cell wall changes in ripening “Bartlett” pears. Plant
tinely exceeded 1.8, and ratios of AZM)to A230typically exceeded 2.0. Physiol. 65, 1009-1013.
Babbitt, J.K., Powers, M.J., and Patterson, M.E. (1973). Effects of
growth-regulators on cellulase, polygalacturonase, respiration, color
Genomlc DNA Gel Blot Analysis and texture of ripening tomatoes. J. Am. SOC.Hort. Sci. 98, 77-81.
Beguin, F! (1990). Molecular biology of cellulose degradation. Annu.
Genomic DNA isolated as previouslydescribed (Murray and Thompson, , Rev. Microbiol. 44, 219-248.
1980; modified by Bernatzky and Tanksley, 1986) was digested with
Bernatzky, R., and Tanksley, S.D. (1986). Genetics of actin-related
the indicated restrictionenzymes, electrophoresced on 0.8% agarose
sequences in tomato. Theor. Appl. Genet. 72, 314-321.
gels, and capillary blotted (Southern, 1975)to HyBond-Nmembranes
(Amersham). Probes were radiolabeled with =P-dATP by random Brecht, J.K., and Huber, D.J. (1988). Products released from enzymi-
priming. Hybridization at 20% below the melting temperature for Cell cally active cell wall stimulate ethylene production and ripening in
(T, -20% for Cell) was for 48 hr at 30% in 50% formamide, 6 x preclimacteric tomato (Lycopersicon esculentum Mill.) fruit. Plant
SSPE (1 x SSPE is 0.15 M NaCI, 10 mM sodium phosphate, 1 mM Physiol. 88, 1037-1041.
EDTA, pH 7.4), 5 x Denhardt’s solution (1 x Denhardt’s solution is Brummell, D.A., Lashbrook, C.C., and Bennett, A.B. (1994). Plant
0.02% Ficoll, 0.02% WP, 0.02% BSA), 0.2% SDS, and 0.05 M sodium endo-l,4-P-D-glucanases: Structure, properties and physiological
phosphate containing 100 pg mL-l base-denatured salmon sperm function. Am. Chem. SOC.Symp. Ser., in press.
DNA. Blots were sequentially washed at T, -2OOC for Cell (2 x Campbell, A.D., and Labavltch, J.M. (1991). lnduction and regula-
SSPE, 0.1% SDS, O.OlO/o, sodium pyrophosphate, 54%) and T, -5% tion of ethylene biosynthesis and ripening by pectic oligomers in
for Cell (1 x SSPE, 0.1% SDS, 0.01% sodium pyrophosphate, 64%). tomato pericarp discs. Plant Physiol. 97, 706-713.
Hybridization was detected by autoradiography on preflashed film. Cass, L.G., Kirven, K.A., and Christoffersen, R.E. (1990). lsolation
and characterizationof a cellulase gene family member expressed
mRNA Quantltation during avocado fruit ripening. MOI. Gen. Genet. 223, 76-86.
Chomczynski, P., and Sacchi, N. (1987). Single-stepmethod of RNA
Cell and Ce12 mRNA transcript levels in 5 to 20 pg of total ANA were isolation by acid guanidinium thiocyanate-phenolchloroformextrac-
quantified by RNase protection assays (RPA II kit; Ambion, Austin, tion. Anal. Biochem. 162, 156-159.
TX) with two modifications: RNase digestion was performed at 30% Christoffersen, R.E., Tucker, M.L., and Laties, G.G. (1984). Cellu-
instead of 37% and gels of protected fragments were fixed for 30 min lase gene expression in ripening avocado fruit: The accumulation
in 15% ethanol, 10% acetic acid and dried on a gel dryer (Bio-Rad) of cellulase mRNA and protein as demonstrated by cDNA hybrid-
prior to autoradiography on preflashed film. Signals within the linear ization and immunodetection. Plant Moi. Biol. 3, 385-391.
range of a standard curve were quantified by densitometry using the DellaPenna, D., Alexander, D.C., and Bennett, A.B. (1986). Molecular
Biolmage program (Millipore, Ann Arbor, MI) attached to a Sun Sparc cloning of tomato fruit polygalacturonase: Analysis of polygalac-
Workstation (Sun Microsystems Inc., Mountain View, CA). turonase mRNA levels during ripening. Proc. Natl. Acad. Sci. USA
Cell and Ce12 sense strand transcripts used to generate standard 83, 6420-6424.
curves for RNase protection assays were synthesized in vitro via the
DellaPenna, D., Kates, D.S., and Bennett, A.B. (1987). Polygalac-
incorporation of 3H-UTPusing T3 or T7 RNA polymerase. Shorter 32P-
turonase gene expression on Rutgers, rin, nor, and Nr tomato fruits.
labeled Cell and Ce12 antisense probes were transcribed from sub-
Plant Physiol. 85, 502-507.
clones in pBluescript (SK- and KS+, respectively) vectors.
DellaPenna, D., Lashbrook, C.C., Toenjes, K., Glovannoni, J.J.,
Fischer, R.L., and Bennett, A.B. (1990). Polygalacturonase iso-
ACKNOWLEDGMENTS zymes and pectin depolymerization in transgenic rin tomato fruit.
Plant Physiol. 94, 1881-1886.
We are grateful to Dr. Mark Tucker (U.S. Department of Agriculture, Dickinson, D.B., and McCollum, D.P. (1964). Cellulase in tomato fruits.
Beltsville, MD) for providing bean abscission zone EGase sequence Nature 203, 525.Tomato Endoglucanase Gene Family 1493
Durbin, M.L., and Lewis, L.N. (1988). Cellulases in Phaseolus vul- Sanger, F., Nicklen, S., and Coulson, A.R. (1977). DNA sequencing
garis. Methods Enzymol. 160, 342-351. with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74,
Farkas, V., and Maclachlan, G. (1988). Stimulation of pea 1,4-p-glu- 5463-5467.
canase activity by oligosaccharides derived from xyloglucan. Sato, T., Kusaba, S., Nakagawa, H., and Ogura, N. (1985). Polygalac-
Carbohydr. Res. 184, 213-219. turonase mRNAof tomato: Size and content in ripe fruits. Plant Cell
Fischer, R.L., and Bennett, A.B. (1991). Role of cell wall hydrolases Physiol. 26, 211-214.
in fruit ripening. Annu. Rev. Plant Physiol. Plant MOI. Biol. 42, Schuch, W., Kanczler, J., Hobson, G., Tucker, G., Grierson, D.,
675-703. Bright, S.,and Bird, C. (1991). Fruit quality characteristicsof trans-
Glovannoni, J.J., DellaPenna, D., Bennett, A.B., and Fischer, R.L. genic tomato fruit with altered polygalacturonaseactivity. HortScience
(1989). Expression of a chimeric polygalacturonase gene in trans- 26, 1517-1520.
genic rin (ripening inhibitor) tomato fruit results in polyuronide Sexton, R., Lewis, L.N., Trewavas, A.J., and Kelly, P. (1985). Ethyl-
degradation but not fruit softening. Plant Cell 1, 53-63. ene and abscission. In Ethyleneand Plant Development, J.A. Roberts
Gross, K.C., and Wallner, S.J. (1979). Degradation of cell wall poly- and G. Tucker, eds (London: Butteworths), pp. 173-196.
saccharides during tomato fruit ripening. Plant Physiol. 63, 117-120.
Sexton, R., De1 Campillo, E., Duncan, D., and Lewis, L.N. (1990).
Hall, C.B. (1963). Cellulase in tomato fruits. Nature 200, 1010-1011. The purification of an anther cellulase (p(l:4)4-glucan hydrolase)
Hall, C.B. (1964). Cellulase activity in tomato fruits according to por- from Lathyrus odoratus L. and its relationshipto the similar enzyme
tion and maturity. Bot. Gaz. 125, 156-157. found in abscission zones. Plant Sci. 67,169-176.
Hanahan, D., and Meselson, M. (1980). Plasmid screening at high Sisler, E.C., and Yang, S.F. (1984). Anti-ethylene effects of cis-Pbutene
colony density. Gene 10, 63-67. and cyclic olefins. Phytochemistry 23, 2765-2768.
Hatfield, R., and Nevins, D.J. (1986). Characterization of the hydro- Smith, C.J.S., Watson, C.F., Morris, P.C., Bird, C.R., Seymour, G.B.,
lytic activity of avocado cellulase. Plant Cell Physiol. 27, 541-552. Gray, J.E., Arnold, C., Tucker, G.A., Schuch, W., Hardlng, S.,
Hayashi, T., Wong, Y.-S., and Maclachlan, G. (1984). Pea xyloglu- and Grierson, D. (1990). lnheritance and effect on ripening of an-
can and cellulose. II. Hydrolysis by pea endo-p-1,4-glucanases. Plant tisense polygalacturonasegenes in transgenic tomatoes. Plant MOI.
Physiol. 75, 605-610. Biol. 14, 369-379.
Henrissat, B., Claeyssens, M., Tomme, P., Lemesle, L., and Mornon, Southern, E.M. (1975). Detection of specific sequences among DNA
J.-P. (1989). Cellulase families revealed by hydrophobic cluster anal- fragments separated by gel electrophoresis.J. MOI. Biol. 98,503-517.
ysis. Gene 81, 83-95.
Speirs, J., Lee, E., Brady, C.J., Robertson, J., and McGlasson, W.B.
Hobson, G.E. (1968). Cellulase activity during the maturationand ripen- (1989). Endopolygalacturonase: Messenger RNA, enzyme and
ing of tomato fruit. J. Food Sci. 33, 588-592. softening in the ripening fruit of a range of tomato genotypes. J.
Huber, D.J. (1983). Polyuronide degradation and hemicellulose modifi- Plant Physiol. 135, 576-582.
cations in ripeningtomatofruit. J. Am. SOC.Hort. Sci. 108,405-409. Su, L.Y., McKeon, T., Grierson, D., Cantwell, M., and Yang, S.F.
Jensen, T.E., and Valdovlnos, J.G. (1967). Fine structure of abscis- (1984). Development of 1-aminocyclopropane-I-carboxylic acid syn-
sion zones. I. Abscission zones of the pedicels of tomato flowers thase and polygalacturonase activities during the maturation and
at anthesis. Planta 77, 298-318. ripening of tomato fruit. HortScience 19, 576-578.
Kramer, M., Sanders, R.A., Sheehy, R.E., Melis, M., Kuehn, M., Tong, C.B.S., Labavltch, J.M., and Yang, S.F. (1986). The induction
and Hiatt, W.R. (1990). Field evaluation of tomatoes with reduced of ethylene productionfrom pear cell culture by cell wall fragments.
polygalacturonase by antisense RNA. In Horticultural Biotechnol- Plant Physiol. 81, 929-930.
ogy, A.B. Bennett and S.D. ONeill, eds (New York: Alan R. Liss),
Tucker, G.A., Schindler, C.B., and Roberts, J.A. (1984). Flower ab-
pp. 347-355.
scission in mutant tomato plants. Planta 160, 164-167.
Lewls, L.N., Llnkins,A.E., OSullivan, S.,and Reld, P.D. (1974).Two
forms of cellulase in bean plants. In Proceedings of the 8th Interna- Tucker, M.L., and Milligan, S.B. (1991). Sequence analysis and com-
tional Conference of Plant Growth Substances (Tokyo: Hirokawa parison of avocado fruit and bean abscission cellulases. Plant
Publishing Co.), pp. 708-718. Physiol. 95, 928-933.
Maclachlan, G., and Brady, C.J. (1992). Multiple forms of 1,4-p-glu- Tucker, M.L., Durbin, M.L., Clegg, M.T., and Lewis, L.N. (1987).
canase in ripeningtomato fruits include a xyloglucanase activatable Avocado cellulase: Nucleotide sequence of a putative full-length
by xyloglucanoligosaccharides. Aust. J. Plant Physiol. 19,137-146. cDNA clone and evidence for a small gene family. Plant MOI. Biol.
9, 197-203.
McDougall, G.J., and Fry, S.C. (1988). lnhibition of auxin-stimulated
growth of pea stem segments by a specific nonasaccharide of Tucker, M.L., Sexton, R., de1 Camplllo, E., and Lewis, L.N. (1988).
xyloglucan. Planta 175, 412-416. Bean abscission cellulase: Characterization of a cDNA clone and
Murray, M., and Thompson, W.F. (1980). Rapid isolation of high mo- regulation of gene expression by ethylene and auxin. Plant Phys-
lecular weight DNA. Nucl. Acids Res. 8, 4321-4325. iol. 88, 1257-1262.
Navarro, A., Chebrou, M.-C., Beguin, P., and Aubert, J.-P. (1991). U.S. Department of Agriculture (1975). Color classification require-
Nucleotide sequence of the cellulase gene celfof Clostridium ther- ments in United States standards for grades of fresh tomatoes. U.S.
mocellum. Res. Microbiol. 142, 927-936. Department of Agriculture visual aid (poster) TM-L-1.
Roberts, J.A., Schlndler, C.B., and Tucker, G.A. (1984). Ethylene- York, W.S., Dawill, A.G., and Albersheim, P. (1984). lnhibition of 2,4-
promoted tomato flower abscission and the possible involvement dichlorophenoxyacetic acid-stimulated elongation of pea stem seg-
of an inhibitor. Planta 160, 159-163. ments by a xyloglucan oligosaccharide. Plant Physiol. 75,295-297.You can also read
