Plant Physiol.

(1998) 118: 83–90

Callose Deposition Is Responsible for Apoplastic

Semipermeability of the Endosperm Envelope of
Muskmelon Seeds1

Kyu-Ock Yim and Kent J. Bradford*

Department of Vegetable Crops, University of California, Davis, California 95616–8631

recycled back to the mother plant via an apoplastic path-

Semipermeable cell walls or apoplastic “membranes” have been way (Oparka and Gates, 1981; Peoples et al., 1985). Strate-
hypothesized to be present in various plant tissues. Although often gically located semipermeable apoplastic membranes may
associated with suberized or lignified walls, the wall component prevent solute movement by mass flow back to the mother
that confers osmotic semipermeability is not known. In muskmelon plant, retaining solutes within the unloading regions of
(Cucumis melo L.) seeds, a thin, membranous endosperm com- developing seeds (Bradford, 1994). The tetrazolium ion,
pletely encloses the embryo, creating a semipermeable apoplastic which is used for vital staining, does not penetrate into
envelope. When dead muskmelon seeds are allowed to imbibe, most grass seeds (Brown, 1907) or through the inner coat of
solutes leaking from the embryo are retained within the envelope, watermelon (Thornton, 1968), tomato, or pepper seeds
resulting in osmotic water uptake and swelling called osmotic dis-
(Beresniewicz et al., 1995b). Similarly, the lanthanum ion, a
tention (OD). The endosperm envelope of muskmelon seeds stained
water-soluble heavy metal, is accumulated at the inner
with aniline blue, which is specific for callose (b-1,3-glucan). Out-
side of the aniline-blue-stained layer was a Sudan III- and IV-
seed coat adjacent to the endosperm in tomato and pepper
staining (lipid-containing) layer. In young developing seeds 25 d (Beresniewicz et al., 1995b; Taylor et al., 1997). In an exam-
after anthesis (DAA) that did not exhibit OD, the lipid layer was ination of seeds of 500 species from more than 40 families,
already present but callose had not been deposited. At 35 DAA, semipermeability was very common, except in the Legu-
callose was detected as distinct vesicles or globules in the en- minosae and some genera of Cistaceae and Cruciferae
dosperm envelope. A thick callose layer was evident at 40 DAA, (Gola, 1905, cited by Kotowski, 1927). The Casparian strip
coinciding with development of the capacity for OD. Removal of in the root endodermis has been considered to be essen-
the outer lipid layer by brief chloroform treatment resulted in more tially impermeable to both water and solutes, but Steudle
rapid water uptake by both viable and nonviable (boiled) seeds, but and Peterson (1998) have proposed that a semipermeable
did not affect semipermeability of the endosperm envelope. The Casparian strip is required to explain water and solute
aniline-blue-staining layer was digested by b-1,3-glucanase, and movement through roots.
these envelopes lost OD. Thus, apoplastic semipermeability of the
Suberized or lignified walls are often associated with
muskmelon endosperm envelope is dependent on the deposition of
apoplastic permeability barriers (O’Brien and Carr, 1970;
a thick callose-containing layer outside of the endosperm cell walls.
Cochran, 1983; Welbaum and Bradford, 1990; Jacobsen et
al., 1992; Welbaum et al., 1992). Suberin has been suggested
as the semipermeable material in seeds of corn (Johann,
The presence of semipermeable apoplastic “membranes” 1942), Johnsongrass (Harrington and Crocker, 1923), to-
has been reported in sugarcane stems (Welbaum et al., mato, and pepper (Beresniewicz et al., 1995a), whereas
1992), developing seeds (Bradford, 1994), and roots cutin may serve this role in leek seeds (Taylor et al., 1997).
(Steudle and Peterson, 1998). For example, sugarcane stems In pollen grains, Heslop-Harrison (1964) suggested that
accumulate high apoplastic Suc concentrations but the xy- callose in the pollen mother cells acts as a “molecular
lem stream within the vascular bundles is virtually free of sieve” to isolate the pollen cells from maternal compounds.
solutes (Welbaum and Meinzer, 1990). Welbaum et al. Nonetheless, although many histochemical studies have
(1992) demonstrated that a semipermeable apoplastic bar- attempted to relate the composition of semipermeable
rier exists between the vascular bundles and storage pa- cell walls to their function, there is no direct demonstra-
renchyma apoplast. Solutes and water are transported into tion that specific wall components are responsible for
developing seeds primarily through the phloem, and high semipermeability.
apoplastic solute concentrations are involved in maintain- In muskmelon (Cucumis melo L.) seeds the embryo is
ing phloem unloading in sink tissues (Wolswinkel, 1992). completely enclosed by a membranous envelope that has
At the same time, excess water delivered via the phloem is been described as consisting of a layer of endosperm cells
and several layers of thick-walled perisperm cells (origi-
1 nating from the nucellus); it was therefore termed the
This work was supported by the U.S. Department of Agricul-
ture Binational Agricultural Research and Development Fund
“perisperm envelope” (Singh, 1953; Welbaum and Brad-
(grant no. US-2422-94).
* Corresponding author; e-mail [email protected]; fax Abbreviations: DAA, days after anthesis; OD, osmotic disten-
1–530 –752– 4554. tion; SEM, scanning electron microscopy.
83
84 Yim and Bradford Plant Physiol. Vol. 118, 1998

ford, 1990). However, as will be shown here, the envelope Zeiss) with a fluorescein isothiocyanate filter (excitation,
surrounding the muskmelon embryo is composed entirely 470 nm; emission, 515 nm; beam splitter, 505DCLP).
of a single layer of endosperm cells covered by a thick, For SEM, seeds were freeze-dried for 3 d and mounted
noncellular external layer. We will therefore refer to this on aluminum stubs. For cross-sectional views, endosperm
tissue as the “endosperm envelope.” When dead musk- envelopes fractured during freeze-drying were positioned
melon seeds are allowed to imbibe, solute leakage from the with a fracture plane on an aluminum stub. Samples were
embryo generates an osmotic gradient across the semiper- sputter coated with 30-nm gold (SEM coating system, Bio-
meable endosperm envelope, resulting in water uptake and Rad) and observed with a scanning electron microscope
swelling, known as OD (Welbaum and Bradford, 1990). (model DS 130, International Scientific Instruments, Top
The intact endosperm envelope can shrink and swell re- Con Technologies, Inc., Paramus, NJ) at 10 kV.
peatedly in response to external osmotic conditions with-
out significant loss of solutes (Welbaum and Bradford,
1990). A semipermeable envelope is present in many spe- OD and Water-Imbibition Kinetics
cies of Cucurbitaceae and Compositae (e.g. lettuce), as To induce OD, decoated seeds were killed by boiling
shown by the development of OD in hydrated dead seeds them in water for 3 min, and were then incubated on
(Hill and Taylor, 1989). water-saturated blotter paper. To observe the imbibition
In muskmelon, endosperm envelope semipermeability kinetics, viable or boiled decoated seeds with or without
was maintained after various treatments that killed the chloroform treatment (see below) were incubated on water-
embryos (freezing and thawing the hydrated seeds, boil- saturated blotter paper in Petri dishes at 30°C. At frequent
ing, autoclaving, aging, and 100% methanol), whereas the intervals, seeds were briefly blotted on lint-free tissues for
semipermeability was lost after strong acid or alkaline 30 s, weighed, and returned to the Petri dish. Observations
treatments (Welbaum and Bradford, 1990). A Sudan IV- were terminated when live seeds completed germination
staining (lipid-containing) layer was detected on the outer (radicle emergence).
surface of the endosperm envelope and was assumed to be
involved in semipermeability, in analogy with the pre-
sumed role of suberized cell walls in other tissues (Wel- Chloroform and Enzyme Treatment
baum and Bradford, 1990). However, no direct evidence To remove the outer lipid-containing layer, decoated
was available to determine which component of the en- seeds were dipped in chloroform for 3 min, rinsed in water,
dosperm envelope is responsible for semipermeability. We and incubated on water-saturated germination paper over-
demonstrate here that an extracellular layer composed pri- night. To treat with b-1,3-glucanase, decoated seeds exhib-
marily of callose is entirely responsible for the semiperme- iting OD (with or without chloroform treatment) were
able properties of the muskmelon endosperm envelope. incubated on germination paper saturated with 3 3 1024
units of endo-b-1,3-glucanase (purified from Helix pomatia;
MATERIALS AND METHODS Fluka, Buchs, Switzerland) in 0.1 m citrate-phosphate
buffer (pH 5.5). After the seeds had lost OD, they were
Muskmelons (Cucumis melo L. cv Top Mark) were field examined using light microscopy or SEM as described
grown at the University of California (Davis) in 1995 and above.
1996, and seeds were harvested, dried, and stored at 220°C
(Welbaum and Bradford, 1988). For the seed-development
study, flowers were tagged at anthesis and harvested at 5-d RESULTS
intervals from 25 to 60 DAA.
Anatomy of the Endosperm Envelope
Hand-sectioned muskmelon seeds showed dark-blue
Anatomy
staining (aniline blue) of the endosperm envelope below
Intact seeds were hand sectioned, and decoated seeds the spongy tissue of the seed coat (Fig. 1A, arrow). Thin
were embedded in paraffin (Jensen, 1962) and thin sec- sections of decoated, paraffin-embedded seeds showed
tioned (10 mm) with a microtome. Decoating was done specific gray-blue staining with aniline blue associated
manually with forceps without damaging the endosperm with the endosperm envelope outside of a single layer of
envelope or embryo. Sections on glass slides were stained rectangular endosperm cells (Fig. 1B). The specificity of the
with 0.05% aniline blue in 0.1 m phosphate buffer (pH 8.2) blue staining for callose observed by light microscopy was
or a Sudan III and IV mixture (equal volumes of saturated confirmed by fluorescence microscopy using synthetic an-
solutions in 70% ethanol) for approximately 5 min and iline blue fluorochrome (Stone et al., 1984). The endosperm
observed using light microscopy. To decrease unspecific envelope showed yellow fluorescence only when stained
staining by aniline blue, toluidine blue O (0.5% in 0.1 m with this fluorochrome (Fig. 1C, arrow), whereas the testa
phosphate buffer, pH 7.0) was used after aniline blue stain- and thin lipid layer of the endosperm envelope showed
ing. To confirm the specificity of the blue staining for autofluorescence in the presence or absence of the dye
callose, mature seeds were hand sectioned, stained with (data not shown). Only triploid nuclei were found when
0.001% synthetic aniline blue fluorochrome (Sirofluor, Bio- the nDNA contents of entire envelope tissues were ana-
supplies Australia, Parkville, Australia) for 20 min, and lyzed using flow cytometry, demonstrating that all living
observed with a fluorescence microscope (Axiovert 100, cells originated from the endosperm (data not shown).
 Callose and Apoplastic Semipermeability 85

Figure 1. A, Cross-section of a dry muskmelon seed stained with 0.05% aniline blue and 0.5% toluidine blue O. The testa
(TE) is at the top, covering a spongy layer. The endosperm envelope (arrow) below the spongy layer shows an outer
blue-stained layer. The bar represents 10 mm. B, Thin section of paraffin-embedded, decoated dry muskmelon seed stained
with 0.05% aniline blue and a saturated mixture of Sudan III and IV. An aniline blue-staining layer (gray-blue) is adjacent
to a layer of rectangular endosperm cells (EN), and a Sudan-staining layer (orange-red) is present on top of the aniline-
blue-staining layer. The bar represents 10 mm. C, Similar to B, but stained with synthesized aniline blue fluorochrome and
viewed by fluorescence microscopy with a fluorescein isothiocyanate filter. The bright-yellow fluorescent band (arrow)
identifies the location of callose in the endosperm envelope. The testa (TE) shows strong autofluorescence that was not
dependent on the presence of the fluorochrome (data not shown). The bar represents 10 mm. D, Surface of the endosperm
envelope of muskmelon seeds with OD viewed by SEM. The bar represents 10 mm. E, Freeze-fractured cross-section of the
muskmelon endosperm envelope viewed by SEM. Above the inner rectangular endosperm cells (EN) is a globular, callosic
layer having a “foamy” appearance (CA). Some crushed cells (CC) appear to be present in the outer lipid-containing layer
(LL). The bar represents 2 mm. F, Removal of the lipid-containing layer by chloroform treatment. Decoated muskmelon seeds
were dipped in chloroform for 3 min and hand sectioned. Sections were stained with aniline blue and Sudan III and IV. The
Sudan-staining layer was removed but the aniline-blue-staining layer was intact after chloroform treatment (compare with
Fig. 2D). The bar represents 10 mm. G, Freeze-fractured cross-section of the chloroform-treated muskmelon endosperm
envelope viewed by SEM. The globular layer remained but the waxy outer layer was removed by the chloroform treatment.
The bar represents 2 mm.

When viewed using SEM, the surface of the endosperm containing component (Fig. 1B), we tested whether this
envelope was smooth and waxy in appearance without lipid layer could be removed by dipping decoated seeds in
obvious cellular structure (Fig. 1D), in contrast to the cel- chloroform. This treatment removed the Sudan-staining
lular outlines clearly visible in back-illuminated light- layer (Fig. 1F, arrow; compare with Fig. 2D, arrow), but did
microscope surface views (Welbaum and Bradford, 1990). not affect aniline blue staining of the callose layer (Fig. 1F).
In cross-view (Fig. 1E), the outer lipid layer appeared to This was also evident in SEM, in which the chloroform
consist of plate-like or waxy sheets and possibly one or two treatment removed the outer layers and exposed the glob-
layers of crushed cells. Beneath this was a thick layer ular, foam-like layer covering the endosperm cells (Fig. 1G;
composed of globules that had a “foamy” appearance and compare surface with that in Fig. 1D). Because suberin is
no cellular compartmentation or contents. Below the glob- not soluble in chloroform, the Sudan-staining material is
ular layer was a single layer of rectangular endosperm apparently not composed of suberin. Thus, the endosperm
cells. envelope of muskmelon contains a single line of en-
Because the outermost layer of the envelope stained dosperm cells covered by a thick, globular callosic layer,
orange-red with Sudan III and IV, indicating a lipid- which in turn is covered by what appears to be a waxy
86 Yim and Bradford Plant Physiol. Vol. 118, 1998

Figure 2. A, Cross-section of developing musk-

melon seeds at 25 DAA. The seeds were de-
coated carefully, sectioned, and stained with
aniline blue and Sudan III and IV. Only Sudan
staining (arrow) was detected on the endosperm
envelope. B, Freeze-fractured cross-section of
endosperm envelope 25 DAA viewed by SEM.
Only the plate-like or waxy appearance was
evident; the globular layer present in mature
envelopes was absent. C, Cross-section of de-
veloping muskmelon seeds at 35 DAA. De-
coated seeds were hand sectioned and stained
with aniline blue. Aniline-blue-staining vesicles
were observed in the endosperm envelope.
These vesicles were not present in other devel-
opmental stages. D, Cross-section of mature
muskmelon seeds (55 DAA). Decoated seeds
were hand sectioned and stained with aniline
blue and Sudan III and IV. The orange Sudan-
staining lipid layer (arrow) was observed on top
of the thick, aniline-blue-staining layer. The bar
in each panel represents 5 mm.

outer coating and some crushed cell remnants possibly accompanied by additional water uptake associated with
derived from the perisperm or testa. embryo growth (Fig. 3A). Boiled seeds exhibited identical
initial water- absorption kinetics, but did not attain a
water-content plateau; instead, they continued to absorb
Relationship between Semipermeability and Callose water and became osmotically distended (Fig. 3). Removal
Deposition in Developing Muskmelon Seeds of the outer lipid layer from viable decoated seeds by
dipping them in chloroform hastened the initial uptake of
Early in development (25 DAA), muskmelon seeds are water and subsequent radicle emergence (Fig. 1, F and G),
not capable of OD, but their endosperm envelopes become but did not affect the plateau water content (Fig. 3). When
semipermeable at approximately 40 DAA (Welbaum and boiled seeds were treated with chloroform, water absorp-
Bradford, 1990; data not shown). However, endosperm tion was even more rapid and OD was achieved within 4 to
envelopes at 25 DAA can be stained with Sudan III and IV 6 h, 12 h earlier than for boiled seeds with the lipid layer
(Fig. 2A, arrow) but not with aniline blue (Fig. 2A). Using present (Fig. 3). Thus, the outer lipid layer slows the rate of
SEM, the globular layer was not observed at 25 DAA (Fig. initial water uptake, but is not required for semipermeabil-
2B). At 30 DAA, the endosperm envelope did not stain with ity of the endosperm envelope.
aniline blue (data not shown), but at 35 DAA, aniline
blue-staining vesicles were evident (Fig. 2C). By 40 DAA, a
thick aniline blue-staining callosic layer was present (data
not shown), coinciding with the acquisition of semiperme- Role of the Callose Layer in Semipermeability
ability, and at 50 DAA, virtually all seeds exhibited both Callose (b-1,3-glucan) is hydrolyzed by endo-b-1,3-
semipermeability and the anatomical characteristics of ma- glucanase, so seeds exhibiting OD were incubated in com-
ture seeds (Fig. 2D). mercially purified enzyme to determine whether the cal-
lose layer would be degraded and whether this would
Role of the Lipid Layer in Water Absorption affect semipermeability. Decoated melon seeds were boiled
and Semipermeability and incubated on water-saturated blotting paper to induce
OD, which was maintained for at least 7 d regardless of
Decoated muskmelon seeds showed a typical triphasic whether the outer lipid layer was present (Fig. 4). When
imbibition time course, with an initial phase of rapid water allowed to imbibe on solutions containing b-1,3-glucanase,
uptake (0–12 h) followed by a plateau phase of relatively seeds began to lose OD after 5 d, and by 7 d of incubation,
constant water content (12–18 h) (Fig. 3A). Radicle emer- only 20% of the seeds maintained semipermeability (Fig. 4).
gence occurred between 18 and 22 h (Fig. 3B), and was A chloroform dip before b-1,3-glucanase treatment accel-
 Callose and Apoplastic Semipermeability 87

has been reported. Extracted callose has a high water-

holding capacity (Barskaya and Balina, 1971; Vithanage et
al., 1980), and low permeability to small molecules of
water-swollen callose (Eschrich and Eschrich, 1964, cited
by Stone and Clarke, 1992). Here we present evidence that
a thick deposit of callose covering the endosperm envelope
of muskmelon seeds serves as a semipermeable molecular
filter that readily allows movement of water but not of
solutes.
The thick outer wall of the muskmelon endosperm en-
velope can be stained with both aniline blue dye and
aniline blue fluorochrome (Fig. 1, A–C, F), which are spe-
cific for b-1,3-glucans, and is also virtually completely
digested by b-1,3-glucanase (Fig. 5, A–C). Thus, the thick
globular layer of the outer endosperm wall (Fig. 1, E and G)
is composed largely of callose, although the presence of
other components in addition to callose cannot be ex-
cluded. The b-1,3-glucanase treatment also causes loss of
OD (Fig. 4), providing direct evidence that the callose layer
is responsible for semipermeability. This conclusion is fur-
ther supported by the simultaneous deposition of callose,
the development of semipermeability (Welbaum and Brad-
ford, 1990), and the increase in the energy required to
penetrate the endosperm envelope (Oluoch, 1996) at
around 40 DAA during seed development.
Callose generally occurs in plant cells as a component of
Figure 3. A, Water-absorption kinetics of muskmelon seeds. De- specialized wall or wall-associated structures at certain
coated seeds were incubated on water-saturated blotters at 30°C with
stages of growth. Callose has been detected in or on cell
or without chloroform treatment (3-min dip) and/or boiling. Dotted
walls of various tissues, including cell plates, cotton seed
lines for control and chloroform-treated seeds indicate germinated
seeds. F, Control seeds; f, chloroform-treated seeds; E, boiled fibers, pollen grain cell walls, innermost pollen tube walls,
seeds; M, boiled and chloroform-treated seeds. The bars represent 6 endosperms, sieve plates, and abscission zones (Esau, 1948;
SE (n 5 20). B, Percentages of germination (control [F] and Currier, 1957; Heslop-Harrison, 1964; Scott et al., 1967;
chloroform-treated [f]) and OD (boiled [E] and boiled and Morrison and O’Brien, 1976; Waterkeyn, 1981; Stone and
chloroform-treated [M]) of decoated muskmelon seeds.

erated the loss of OD (Fig. 4), suggesting that the lipid layer
restricts access by the enzyme to the callose layer.
Endosperm envelopes from seeds that had been treated
with both chloroform and b-1,3-glucanase and had lost OD
did not stain with aniline blue or Sudan III and IV (Fig. 5A;
compare with Fig. 2D). The b-1,3-glucanase treatment di-
gested the globular layer outside of the endosperm cells,
leaving only a thin, porous network in the surface view
(Fig. 5B). A fractured cross-section of the b-1,3-glucanase-
treated endosperm envelope showed that the callose layer
had essentially disappeared (Fig. 5C, arrow), leaving only a
few isolated regions where the globular material could still
be found (Fig. 5C, CA). The effects of the enzyme treatment
on OD (Fig. 4) and on callose degradation (compare Fig. 1,
A, B, E, and G, with Fig. 5, A–C) leave little doubt that the
callose layer is responsible for the semipermeability of the Figure 4. Loss of OD after b-1,3-glucanase treatment. Decoated
endosperm envelope. muskmelon seeds were killed in boiling water for 3 min and incu-
bated overnight on blotters saturated with distilled water to induce
OD. For chloroform treatments, boiled seeds with OD were dipped
DISCUSSION into a chloroform solution for 3 min and rinsed with water. Seeds in
buffer (M) or in buffer after chloroform treatments (E) maintained OD
Although it was hypothesized long ago that callose may for at least 7 d. The presence of b-1,3-glucanase in the buffer (L)
act as a “molecular filter” in plants by altering the gel- initiated the loss of OD at 5 d of incubation or after only 1 d in seeds
filtration properties of the cell wall (Heslop-Harrison, pretreated in chloroform (Œ). The bars represent 6 SE from three
1964), no direct in vivo evidence to support this hypothesis independent experiments of 20 seeds each.
88 Yim and Bradford Plant Physiol. Vol. 118, 1998

example, callose deposits in clover seed coats apparently

prevent water absorption (Bhalla and Slattery, 1984), in
contrast to the high water permeability of the muskmelon
endosperm envelope (Fig. 3A).
Lipid-containing or suberized cell walls (stained with
Sudan dye) are frequently associated with semipermeable
regions (Fig. 1B; Harrington and Crocker, 1923; Johann,
1942; O’Brien and Carr, 1970; Cochran, 1983; Welbaum and
Bradford, 1990; Jacobsen et al., 1992; Welbaum et al., 1992;
Beresniewicz et al., 1995a). However, the lipid-containing
outer layer of the melon endosperm envelope slows water
uptake but is not responsible for semipermeability (Fig. 3).
Whether callose is also associated with other “suberized”
apoplastic membranes, such as the Casparian strip of the
root endodermis or the sheaths of sugarcane vascular bun-
dles, is unknown. Cucumber, zucchini, watermelon, and
barley seeds, all of which exhibit semipermeability, also
have a thick aniline-blue-staining layer inside of the seed
coat (data not shown). However, some seeds have semi-
permeable layers that do not contain callose (Beresniewicz
et al., 1995a; K.-O. Yim and K.J. Bradford, unpublished
results for lettuce endosperm envelopes). Because embryos
often leak solutes upon initial imbibition (Simon and Mills,
1983), a semipermeable envelope would prevent loss of
solutes to the environment until the embryo is capable of
reabsorbing them before initiation of radicle growth. There
are many additional locations in the plant where the ability
to restrict solute movement in the apoplast while permit-
Figure 5. A, Cross-section of chloroform- and b-1,3-glucanase- ting water continuity and flow would be advantageous,
treated muskmelon seeds. Seeds with OD were treated with chloro- particularly in developing seeds (Bradford, 1994), in the
form and incubated for 36 h on blotters saturated with solution root (Steudle and Peterson, 1998), and in the vascular sys-
containing b-1,3-glucanase. Seeds that lost OD were hand sectioned tem (Canny, 1995).
and stained with aniline blue and Sudan III and IV. The Sudan- The large amounts of callose present and its specific
staining lipid layer and the aniline-blue-staining callose layer were deposition only on the outer side of the endosperm enve-
removed (arrow; compare with Fig. 2D). B, Surface of chloroform- lope (Fig. 1, B, C, and E) raise intriguing questions about
and b-1,3-glucanase-treated muskmelon endosperm envelope the mechanism of callose synthesis in these cells. Synthesis
viewed by SEM (compare with Fig. 1G). C, Freeze-fractured cross-
of callose via a well-characterized plasma membrane-
section of chloroform- and b-1,3-glucanase-treated muskmelon en-
bound callose synthase is usually activated temporally by
dosperm envelope viewed by SEM. The cross-section shows that the
globular layer is removed (arrow; compare with Fig. 1G), leaving specific signals such as wounding, infection, or other
only a few remnants in some areas (CA). The bar in each panel stresses (for review, see Delmer and Amor, 1995). The
presents 5 mm. plasma membrane callose synthase is activated by micro-
molar concentrations of Ca21 and b-glucoside (Hayashi et
al., 1987), suggesting that transient increases in Ca21 con-
Clarke, 1992). Callose deposition is also induced locally by centration caused by cell perturbation, such as by microbe
wounding, stress, and fungal or viral infection (Currier, infection or wounding, induce callose deposition outside of
1957; Esau and Cronshaw, 1967; Coffey, 1976). Despite the the plasma membrane. Although less well characterized,
widespread occurrence of callose, its general function is not there is some evidence for a Golgi-vesicle-mediated
well understood (for review, see Stone and Clarke, 1992). It callose-synthesis system in developmentally regulated
may serve as a matrix for deposition of other cell wall callose-rich tissues such as pollen tube walls, pollen tube
materials, as in developing cell plates and sieve-plate plugs, and developing cell plates. In pollen tubes callose
pores; as a cell wall-strengthening material, as in cotton synthase activity was associated with Golgi vesicles (Hel-
seed hairs (Maltby et al., 1979) and pollen; as a sealing or sper et al., 1977) and showed little dependence on Ca21
plugging material at the plasma membrane of pit fields, (Schlüpmann et al., 1993; Li et al., 1997). Callose was de-
plasmodesmata, and sieve-plate pores (Eschrich, 1975); as a tected by immunogold labeling in Golgi vesicles, which
mechanical obstruction to growth of fungal hyphae; or as a were concentrated at the tips of germinating pollen tubes of
special permeability barrier, as in pollen mother cell walls camellia (Hasegawa et al., 1996). However, Li et al. (1997)
and muskmelon endosperm envelopes (Heslop-Harrison, proposed that inactive callose synthase is secreted in ves-
1964; this report). In addition, the degree of polymeriza- icles at the tips of tobacco pollen tubes and is activated
tion, age, and thickness of the deposits may vary the phys- upon insertion into the plasma membrane. When callose
ical properties of callose (Stone and Clarke, 1992). For accumulation occurred at about 35 DAA in the muskmelon
 Callose and Apoplastic Semipermeability 89

endosperm envelope, discrete vesicles or globules were Esau K (1948) Phloem structure in the grapevine, and its seasonal
stained with aniline blue (Fig. 2C), consistent with the changes. Hilgardia 18: 217–296
globular appearance of the callose layer (Fig. 1, E and G). Esau K, Cronshaw J (1967) Relation of tobacco mosaic virus to the
host cells. J Cell Biol 33: 665–678
However, more detailed studies of the callose synthesis Eschrich W (1975) Sealing systems in phloem. In MH Zimmer-
and deposition process in muskmelon endosperm enve- mann, JA Milburn, eds, Encyclopedia of Plant Physiology, Vol 1:
lopes are required to determine the mechanism by which Transport in Plants. I. Phloem Transport. Springer-Verlag, Ber-
the callose layer is formed. lin, pp 39–56
In conclusion, we have demonstrated that the semiper- Eschrich W, Eschrich B (1964) Das verhalten isolierter callose
gegenüber wässrigen lösungen. Ber Dtsch Bot Ges 77: 329–331
meability of the muskmelon endosperm envelope is caused Gola G (1905) Richerche sui rapporti tra i tegumenti seminali e le
by a callose-containing layer deposited outside of the outer soluzioni saline. Annali di Botanica 3: 59–100
walls of the endosperm cells. A chloroform-soluble waxy Harrington GT, Crocker W (1923) Structure, physical character-
layer outside of the callosic layer delays water uptake but istics, and composition of the pericarp and integument of
is not required for semipermeability. These results directly Johnson grass seed in relation to its physiology. J Agric Res 23:
193–222
confirm, for the first time to our knowledge, the long- Hasegawa Y, Nakamura S, Nakamura N (1996) Immunocyto-
standing hypothesis that callose may act as a molecular chemical localization of callose in the germinated pollen of
sieve. They also show that the presence of Sudan-stainable Camellia japonica. Protoplasma 194: 133–139
material does not in itself imply impermeability to water. Hayashi T, Read SM, Bussell J, Thelen M, Lin FC, Brown RM,
Delmer DP (1987) UDP-glucose:(1,3)-b-glucan synthases from
The existence of semipermeable apoplasts resulting from the
mung bean and cotton. Differential effects of Ca21 and Mg21 on
deposition of callose (or other materials) in the cell wall enzyme properties and on macromolecular structure of the glu-
provides options for controlling water and solute transfer in can product. Plant Physiol 83: 1054–1062
many parts of the plant and at critical stages of development. Helsper JPFG, Veerkamp JH, Sassen MMA (1977) b-Glucan syn-
thetase activity in Golgi vesicles of Petunia hybrida. Planta 133:
303–308
Heslop-Harrison J (1964) Cell walls, cell membranes and proto-
ACKNOWLEDGMENTS plasmic connections during meiosis and pollen development. In
HF Linskens, ed, Pollen Physiology and Fertilization. North-
The authors express their appreciation to Dr. Deborah Delmer Holland, Amsterdam, pp 29–47
(University of California, Davis) for critical reading and construc- Hill HJ, Taylor AG (1989) Relationship between viability, en-
tive comments on the manuscript and for technical assistance with dosperm integrity and imbibed lettuce seed density and leakage.
the fluorescence microscopy. The assistance of Dr. Sunitha Gurus- HortScience 24: 814–816
inghe with the flow cytometry of endosperm nuclei is gratefully Jacobsen JS, Fisher DG, Maretzki A, Moore PH (1992) Anatomy
acknowledged. of sugarcane stem in relation to phloem unloading and sucrose
storage. Bot Acta 18: 959–969
Jensen WA (1962) Botanical Histochemistry: Principles and Prac-
Received February 17, 1998; accepted June 14, 1998 tice. WH Freeman, San Francisco, CA, pp 35–99
Copyright Clearance Center: 0032–0889/98/118/0083/08. Johann H (1942) Origin of the suberized semipermeable mem-
brane in the caryopsis of maize. J Agric Res 64: 275–281
Kotowski F (1927) Semipermeability of seed coverings and stim-
ulation of seeds. Plant Physiol 2: 177–186
LITERATURE CITED Li H, Bacic A, Read SM (1997) Activation of pollen tube callose
Barskaya EI, Balina NV (1971) The role of callose in plant anthers. synthase by detergents. Evidence for different mechanisms of
Fiziol Rast 18: 716–721 action. Plant Physiol 114: 1255–1265
Beresniewicz MM, Taylor AG, Goffinet MC, Koeller WD (1995a) Maltby D, Carpita NC, Montezinos D, Kulow C, Delmer DP
Chemical nature of a semipermeable layer in seed coats of leek, (1979) b-1,3-Glucan in developing cotton fibers. Structure, local-
onion (Liliaceae), tomato and pepper (Solanaceae). Seed Sci ization and relationship of synthesis to that of secondary wall
Technol 23: 135–145 cellulose. Plant Physiol 63: 1158–1164
Beresniewicz MM, Taylor AG, Goffinet MC, Terhune BT (1995b) Morrison IN, O’Brien TP (1976) Cytokinesis in the developing
Characterization and location of a semipermeable layer in seed wheat grain: division with and without a phragmoplast. Planta
coats of leek and onion (Liliaceae), tomato and pepper (So- 130: 57–67
lanaceae). Seed Sci Technol 23: 123–134 O’Brien TP, Carr DJ (1970) A suberized layer in the cell walls of
Bhalla PL, Slattery HD (1984) Callose deposits make clover seeds the bundle sheath of grasses. Aust J Biol Sci 23: 275–287
impermeable to water. Ann Bot 53: 125–128 Oluoch MO (1996) Effects of priming and stage of development on
Bradford KJ (1994) Water stress and the water relations of seed vigor and longevity of muskmelon (Cucumis melo L.) seeds. PhD
development: a critical review. Crop Sci 34: 1–11 dissertation. Virginia Polytechnic Institute and State University,
Brown R (1907) On the existence of a semi-permeable membrane Blacksburg, Virginia
enclosing the seeds of some of the Gramineae. Ann Bot 21: 79–87 Oparka KJ, Gates P (1981) Transport of assimilates in the devel-
Canny MJ (1995) Apoplastic water and solute movement: new oping caryopsis of rice (Oryza sativa L.): the pathways of water
rules for an old space. Annu Rev Plant Physiol Plant Mol Biol 46: and assimilated carbon. Planta 152: 388–396
215–236 Peoples MB, Pate JS, Atkins CA, Murray DR (1985) Economy of
Cochran MP (1983) Morphology of the crease region in relation to water, carbon and nitrogen in the developing cowpea fruit. Plant
assimilate uptake and water loss during caryopsis development Physiol 77: 142–147
in barley and wheat. Aust J Plant Physiol 10: 473–491 Schlüpmann H, Bacic A, Read SM (1993) A novel callose synthase
Coffey MD (1976) Flax rust resistance involving the K gene: an from pollen tubes of Nicotiana. Planta 191: 470–481
ultrastructural survey. Can J Bot 54: 1443–1457 Scott PC, Miller LW, Webster BD, Leopold AC (1967) Structural
Currier HB (1957) Callose substance in plant cells. Am J Bot 44: changes during bean leaf abscission. Am J Bot 54: 730–734
478–488 Simon EW, Mills LK (1983) Imbibition, leakage and membranes.
Delmer DP, Amor Y (1995) Cellulose biosynthesis. Plant Cell 7: In C Nozzolillo, PJ Lea, FA Loewus, eds, Mobilization of Re-
987–1000 serves in Germination. Plenum Press, New York, pp 9–27
90 Yim and Bradford Plant Physiol. Vol. 118, 1998

Singh B (1953) Studies on the structure and development of seeds Waterkeyn L (1981) Cytochemical localization and function of the
of Cucurbitaceae. Phytomorphology 3: 224–239 3-linked glucan callose in the developing cotton fibre cell wall.
Steudle E, Peterson CA (1998) How does water get through roots? Protoplasma 106: 49–67
J Exp Bot 49: 775–788 Welbaum GE, Bradford KJ (1988) Water relations of seed devel-
Stone BA, Clarke AE (1992) Chemistry and Biology of (133)-b- opment and germination in muskmelon (Cucumis melo L.) I.
Glucans. La Trobe University Press, Victoria, Australia, pp Water relations of seed and fruit development. Plant Physiol 86:
355–429 406–411
Stone BA, Evans NA, Bonig I, Clarke AE (1984) The application of Welbaum GE, Bradford KJ (1990) Water relations of seed devel-
Sirofluor, a chemically defined fluorochrome from aniline blue opment and germination in muskmelon (Cucumis melo L.) IV.
for the histochemical detection of callose. Protoplasma 122:
Characteristics of the perisperm during seed development. Plant
191–195
Physiol 92: 1038–1045
Taylor AG, Beresniewicz MM, Goffinet MC (1997) Semiperme-
able layer in seeds. In EM Black, AJ Murdoch, TD Hong, eds, Welbaum GE, Meinzer FC (1990) Compartmentation of solutes
Basic and Applied Aspects of Seed Biology. Kluwer Academic and water in developing sugarcane stalk tissue. Plant Physiol 93:
Publishers, Dordrecht, The Netherlands, pp 429–436 1147–1153
Thornton ML (1968) Seed dormancy in watermelon Citrullus vul- Welbaum GE, Meinzer FC, Grayson RL, Thornham KD (1992)
garis Shrad. Proceedings of the Association of Official Seed Evidence for and consequences of a barrier to solute diffusion
Analysts 58: 80–84 between the apoplast and vascular bundles in sugarcane stalk
Vithanage HIMV, Gleeson PA, Clarke AE (1980) The nature of tissue. Aust J Plant Physiol 19: 611–623
callose produced during self-pollination in Secale cereale. Planta Wolswinkel P (1992) Transport of nutrients into developing seeds:
148: 498–509 a review of physiological mechanisms. Seed Sci Res 2: 59–73