Mineral nutrition in White Spruce Picea
Plant Science 141 (1999) 19 – 27
Mineral nutrition in White Spruce (Picea glauca [Moench] Voss) seeds
and somatic embryos; II. EDX analysis of globoids and Fe-rich
particles
Daryl A. Reid a, John N.A. Lott a,*, Stephen M. Attree b,1, Larry C. Fowke b
b
a
Department of Biology, McMaster Uni6ersity, Hamilton, ON, L8S 4K1, Canada
Department of Biology, Uni6ersity of Saskatchewan, Saskatoon, SK, S7N 5E2, Canada
Received 13 July 1998; received in revised form 26 October 1998; accepted 28 October 1998
Abstract
The elemental compositions of globoids and Fe-rich particles were investigated in white spruce (Picea glauca [Moench] Voss)
somatic embryos, zygotic embryos and female gametophytes using energy dispersive X-ray analysis. Globoids, phytate deposits in
seed protein bodies, were found throughout the female gametophyte of seeds and in the ground meristem and procambium from
all regions of both somatic and zygotic embryos. Iron-rich particles, believed to be Fe-associated phytate deposits in seed
proplastids, were found throughout the female gametophyte of seeds and in the protoderm, ground meristem, and procambium
from all regions of both somatic and zygotic embryos. Globoids in somatic and zygotic embryos ranged from 0.5– 3.0 mm in
diameter, but globoids typically ranged from 1.5– 2.0 mm in diameter in somatic embryos and from 2.0– 3.0 mm in diameter in
zygotic embryos. Globoids in female gametophyte tissue ranged from 0.5– 6.0 mm in diameter. All Fe-rich particles studied from
somatic embryos and from seeds ranged from 0.14– 0.25 mm in diameter. Globoids in somatic embryos and seeds contained high
P, moderate K and Mg with occasional traces of Fe and little if any Ca and Zn. Globoids in the zygotic embryo cotyledon
procambium tissue also contained moderate levels of Fe and had significantly higher Fe:P ratios, which were not found in any
other regions in seeds or in somatic embryos. Iron-rich particles in somatic embryos and seeds contained high P and Fe, moderate
K and Mg, and little if any Ca and Zn. Typically, spectra of Fe-rich particles in somatic embryos had P peaks higher than Fe
peaks and spectra of Fe-rich particles in seeds had P peaks lower than Fe peaks. Overall, the composition of globoids and Fe-rich
particles in somatic embryos and zygotic embryos were very similar with only minimal differences found. © 1999 Elsevier Science
Ireland Ltd. All rights reserved.
Keywords: Somatic embryos; Seeds; White spruce; EDX analysis; Globoids; Fe-rich particles
1. Introduction
In this study, energy dispersive X-ray (EDX)
analysis was used to investigate the mineral nutrient reserves of white spruce seeds and somatic
embryos. This paper will focus on those elements
* Corresponding author. Tel.: + 1-905-5259140 ext. 24589; fax:
+ 1-905-5226066; e-mail: lott@mcmaster.ca.
1
Present address: Pacific Regeneration Technologies Inc., Victoria,
BC, V8T 2W1, Canada.
measured on a whole tissue basis in paper I of this
two paper series. Energy dispersive X-ray analysis
is a convenient tool for studying the elemental
compositions of specific spots or larger areas
within the sample and allows the simultaneous
detection of elements with a high detection sensitivity [1,2]. We used EDX analysis to determine if
phytate reserves in mature somatic embryos of
white spruce have similar element compositions to
those found in mature white spruce seeds. This
research will provide information on mineral nu-
0168-9452/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved.
PII: S 0 1 6 8 - 9 4 5 2 ( 9 8 ) 0 0 2 1 2 - X
20
D.A. Reid et al. / Plant Science 141 (1999) 19–27
trient storage which may be useful to those developing artificial seeds.
Globoids (or globoid crystals) are naturally electron-dense deposits within seed protein bodies and
are composed of phytate. Phytate is a mixed
cation salt of myo-inositol hexa-kisphosphoric
acid or phytic acid [3,4]. Phytate is the typical
mineral nutrient storage compound in seeds. Due
to their natural electron density and relatively
large size, these inclusions are convenient particles
for spectral EDX analysis. Energy dispersive Xray analysis of globoids from different tissues and
species have shown the presence of P, Mg and K
with occasional storage of Ca, Mn, Na, Zn, Ba,
and Fe [5].
Studying mature seeds from eleven Pinus species, small ( 5 0.33 mm in diameter) naturally electron-dense
inclusions
were
found
in
membrane-bound structures resembling proplastids and were termed Fe-rich particles based on
EDX analysis [6]. These particles contained high P
and Fe, moderate levels of K and Mg, and traces
of Ca, Mn, and Zn [6]. Based on the similarities of
EDX analysis spectra of Fe-rich particles, globoids
and precipitated ferric phytate, it was proposed
that these particles could represent storage deposits of Fe-rich phytate [6,7]. The high levels of P
also suggested that these deposits were not phytoferritin inclusions, which is a common iron storage
compound found in chloroplasts [8].
In this study the elemental composition of
globoids and Fe-rich particles in various regions of
white spruce somatic embryos, zygotic embryos
and female gametophytes was investigated using
EDX analysis. This paper will concentrate on the
levels of P, K, Mg, Ca, Fe, and Zn within these
deposits since these nutrients were of focus in part
one of this two paper series. We believe this to be
the first study of globoid and Fe-rich particle
composition in white spruce and the first comparison of their composition in somatic embryos and
seeds of any species.
2. Materials and methods
2.1. Tissue preparation for EDX analysis
White spruce (Picea glauca [Moench] Voss)
seeds were collected from the Big River and
Christopher Lake regions north of Saskatoon,
Saskatchewan. Mature zygotic embryos were excised from artificially dried seeds using fine-tipped
forceps. White spruce somatic embryos were prepared and desiccated as previously described
[9,10]. To prevent the extraction of water-soluble
phytate, while allowing enough tissue expansion
that epoxy resin can penetrate the tissue, a lowwater-content preparation procedure was used
[11]. Tissue samples were initially placed into 80%
ethanol overnight. These samples were then placed
into 100% ethanol for at least 8 h and again in
100% ethanol overnight. To complete the dehydration, the tissue samples were treated with propylene oxide for at least 7 h.
Using a standard hardness Spurr’s epoxy resin,
the samples were infiltrated on a tissue rotator
using a propylene oxide:Spurr’s epoxy resin series
(2:1, 1:2, 0:1, and 0:1) for 24 h in each solution.
Following infiltration, each embryo was sectioned
into cotyledons, hypocotyl and radicle regions and
each region was placed into a separate rubber
mold. Additionally, each female gametophyte was
divided into cotyledon-associated, hypocotyl-associated and radicle-associated regions prior to dehydration and placed into separate molds
following infiltration with Spurr’s resin. All molds
were filled with Spurr’s resin and allowed to
harden in a 70°C oven overnight.
A Reichert OM U2 ultramicrotome (Reichert,
Austria) was used to cut thick sections (1–1.5 mm)
using dry glass knives. Each section was flattened
using eyelashes and cactus spines mounted on
wooden sticks and picked up using 100 mesh
Formvar-carbon coated copper grids. Drops of
100% ethanol were used to help the sections adhere to the Formvar.
2.2. EDX analysis of globoids and Fe-rich
particles
EDX analysis was carried out at 80 kV and a
magnification of 25 000 using the scanning mode
of a JEOL-1200 EX-II TEMSCAN microscope
(JEOL, Tokyo) and a PGT model IMIX-II microanalysis system (Princeton Gamma-Tech., Princeton, NJ). All spectra were collected for 60 s. with
aperture, spot size, tilt and detector distance kept
the same for all analyses. Only those spectra collected with greater than 1000 counts s − 1. and less
than 5% dead time were used.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
For each of five somatic embryos and five
zygotic embryos, five spectra of globoids from
different cells and of different sizes were collected
from both the ground meristem and procambium
in each of the cotyledon, hypocotyl and radicle
regions. Additionally, for each of these same five
embryos, five spectra of Fe-rich particles were
collected from each of the protoderm, ground
meristem and procambium in each of the cotyledon, hypocotyl and radicle regions. Five spectra of
globoids and five spectra of Fe-rich particles were
collected from each of the cotyledon-associated,
hypocotyl-associated and radicle-associated regions from each of five female gametophytes.
Therefore in total, 525 spectra of Fe-rich particles
and 375 spectra of globoids were collected.
X-ray counts for all the spectra were collected
by integrating the peaks at the selected window
widths: Mg, 1153.7–1354.3 eV; P, 1905.3–2120.7
eV; K, 3193.7 –3432.3 eV; Ca, 3568.6–3813.4 eV;
Mn, 5758.5 –6037.5 eV; Fe, 6259.9–6546.1 eV; Cu,
7891.7 –8200.3 eV; and Zn, 8478.9–8795.1 eV [12].
In order to produce the best fit background line
for all elements, points were connected at the
following eV values: 510, 660, 814, 1453, 1730,
2500, 2800, 3020, 4200, 5400, 8400, 9400, and
11 000 eV. Background subtraction was performed
for each element in each spectrum by subtracting
the number of counts in the background from the
total number of counts in the element. A useful
format to express peak values is in the form of
peak-to-background (P/B) ratios, which have been
defined as the number of counts above the background for a peak divided by the number of
background counts [13]. Peak-to-background ratios were calculated for each element in each spectrum by using the total number of counts before
and after subtraction [14]. Ratios of elements to P
were also calculated since P is the main anion to
which the elements are bound in phytate.
Due to peak overlaps, correction factors were
used to calculate the actual Ca, Fe, and Zn counts
in each spectrum [7]. Correction factors were used
prior to background subtraction. The Kb peak of
K overlaps the Ka peak of Ca and therefore a
correction factor of 8.80% of the total X-ray
counts for the K Ka peak was subtracted from the
total counts in the Ca window to give the actual
counts for Ca. Similarly, the Kb peak of Mn
overlaps the Fe Ka peak and therefore a correction
factor of 11.60% of the total X-ray counts for Mn
21
Ka peak was subtracted from the total counts in
the Fe window to give a corrected Fe value.
Finally, the Kb peak of Cu overlaps the Ka peak
of Zn and therefore 2.00% of the net Cu Ka counts
was subtracted from the total Zn counts to give a
corrected Zn value.
2.3. Statistical analysis
The statistical significance between means was
determined by using the MINITAB analysis of
variance test. When a difference was found,
Tukey’s test was used to determine which of the
means were statistically different at P \0.05 [15].
Due to the use of manually predetermined background points and relatively short sampling times,
several negative counts were obtained for Ca, Fe
and Zn P:B values as well as for the Ca:P, Fe:P,
and Zn:P ratios. For each P/B value and ratio of
P/B values, all the negative values from all the
spectra collected were averaged and any means
equal to or less than the absolute value of this
mean were considered to be not significantly different from the background line and were assigned
a value of zero. Peak-to-background values and
ratios of P/B values for individual embryos were
analysed to see if there were any significant variations from embryo-to-embryo.
The MINITAB two-sample t-test for the difference between two means at P \0.05 was used to
compare EDX analysis values for somatic embryos to values from zygotic embryos. In this
comparison, for each P/B value and ratio of P/B
values all the globoids in somatic embryos were
pooled together and compared to all the globoids
pooled together for zygotic embryos. Each element
P/B value and each ratio of an element to P in
somatic embryo globoids were compared to the
corresponding value in zygotic embryo globoids.
This was then repeated for all the EDX analysis
values for Fe-rich particles.
3. Results
3.1. General obser6ations
Naturally electron-dense globoids were found
throughout the female gametophyte tissue and in
the ground meristem and procambium in the
cotyledons, hypocotyl and radicle regions of so-
22
D.A. Reid et al. / Plant Science 141 (1999) 19–27
Figs. 1 – 4. Scanning transmission electron micrographs of
thick sections (1– 1.5 mm) from white spruce somatic embryos,
zygotic embryos and female gametophytes. Fig. 1. Naturally
electron-dense globoids (G) variable in size and a smaller
Fe-rich particle (F) in ground meristem tissue from the radicle
of a somatic embryo. Scale bar, 2 mm.
matic and zygotic embryos. Globoids were not
found in the protoderm of either somatic or
zygotic embryos. Typically, globoids in somatic
embryos were found to be 0.5 –2.0 mm in diameter
(Figs. 1 and 2), but were occasionally observed to
be 3.0 mm in diameter. Globoids in zygotic embryos were also found to be between 0.5 and 3.0
mm in diameter (Fig. 3), but typically they were
found to be between 2.0 and 3.0 mm in diameter.
Female gametophyte tissue was found to contain
globoids that ranged in size from 0.5 –6.0 mm in
Fig. 2. Hypocotyl ground meristem tissue of a somatic embryo showing a small Fe-rich particle (F) and protein bodies
(P), which contain globoids (G). Scale bar, 2 mm.
Fig. 3. Iron-rich particles (F) and protein bodies (P), which
contain globoids (G), are shown in the hypocotyl ground
meristem tissue of a zygotic embryo. Scale bar, 2 mm.
diameter (Fig. 4). Globoids in the ground meristem of somatic and zygotic embryos tended to be
larger than those found in the procambium of the
same embryo.
Generally only one globoid was observed per
section of a single protein body in somatic and
zygotic embryos, but occasionally as many as
three were visible. Protein bodies from female
gametophyte tissue varied in the number of
globoids per section, with some protein bodies
containing one globoid and another protein body
within the same cell containing four or more
globoids. Globoids were most frequent per unit
area in female gametophyte tissue and least frequent per unit area in somatic embryos.
Fig. 4. Female gametophyte tissue containing globoids (G)
variable in size within protein bodies (P). Scale bar, 6 mm.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
23
Table 1
Mean peak-to-background ratios and ratios of Fe:P in globoids and Fe-rich particles from resin embedded thick sections for five
white spruce female gametophytesa
Region
Particle
P
K
Mg
Fe
Fe:P
Cotyledon-associated
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Globoid
3.39 1.0a
11.4 9 1.9b
3.3 9 1.1a
12.09 1.9b
4.2 91.7a
11.7 92.2b
2.3 9 0.8a
8.7 92.4b
2.5 9 0.9a
10.39 1.4b
2.79 1.4a
9.99 2.2b
0.9 90.3a
3.9 91.1b
0.9 90.2a
4.5 9 0.7b
0.99 0.3a
4.79 1.1b
9.8 93.3a
0.5 9 0.3b
10.99 4.6a
0.3 9 0.2b
13.79 5.9a
0.29 0.2b
3.09 0.6a
0.1 90.0b
3.2 90.6a
0.0 9 0.0b
3.49 0.8a
0.09 0.0b
Hypocotyl-associated
Radicle-associated
a
Each mean ( 9S.D.) was calculated using 25 values; each value in a single column followed by the same letter is not
significantly different at P\0.05.
Iron-rich particles were found in the protoderm,
ground meristem and procambium from the
cotyledons, hypocotyl and radicle regions of somatic and zygotic embryos (Figs. 1–3). Iron-rich
particles were also found throughout the female
gametophyte tissue, but they tended to be very
difficult to locate. Typically Fe-rich particles were
0.14 –0.25 mm in diameter and were also naturally
electron-dense. Occasionally Fe-rich particles were
found in clusters in embryo tissues, but never in
female gametophyte tissue.
3.2. EDX analysis of globoids and Fe-rich
particles
Typical EDX analysis spectra (based on the
mean P/B values given in Tables 1–3) of globoids
from somatic embryos, zygotic embryos and female gametophytes were similar in composition
and a representative spectrum is shown in Fig. 5.
Typical globoids contained high levels of P, moderate levels of Mg and K, occasional traces of Fe
and little if any Ca and Zn. One exception to this
typical composition was that globoids from the
procambium of zygotic embryo cotyledons contained moderate levels of Fe and slightly reduced
K and Mg (Fig. 6). Spectra of globoids from the
same tissue region in somatic embryos showed no
significant levels of Fe and more typical K and Mg
levels.
Iron-rich particles in somatic embryos were
found to be similar in composition to those found
in zygotic embryos and female gametophytes. All
Fe-rich particles contained high P and Fe with
moderate levels of K and Mg, and little if any Ca
and Zn. Iron-rich particles sampled from somatic
embryos typically had P peaks higher than Fe
peaks (Fig. 7); whereas, Fe-rich particles from
seeds typically had P peaks lower than Fe peaks
(Fig. 8). Although Fe-rich particles in seeds and
somatic embryos had means that differed in the
proportions of P to Fe, particles of similar P and
Fe proportions were found in both seeds and
somatic embryos. Comparison of a typical globoid
spectrum (Fig. 5) with Fe-rich particle spectra
(Figs. 7 and 8), show that K and Mg peaks were
slightly lower in Fe-rich particles.
Mean P/B values and the ratio of Fe:P for
female gametophytes, zygotic embryos, and somatic embryos are given in Tables 1–3, respectively. Since little if any Ca or Zn were detected in
globoids and Fe-rich particles, these values are not
given in these tables. No significant differences
were found in P/B values from different female
gametophytes or from embryo-to-embryo for either somatic or zygotic embryos. The majority of
the variation was found to be from particle-toparticle.
Comparison of P/B ratios and ratios of Fe:P in
globoids and Fe-rich particles tested in various
regions of female gametophyte tissue showed no
significant variation from region-to-region (Table
1). The P P/B values were between 2.7 and 3.6
times higher in globoids than in Fe-rich particles.
Globoids also had between 3.2 and 4.5 times
higher K and 4.3 to 5.2 times higher Mg than
Fe-rich particles. Iron P/B values were between
19.6 and 68.5 times higher in Fe-rich particles than
in globoids. Globoids in the female gametophyte
tissue had trace amounts of Fe and therefore very
low Fe:P ratios. The Fe:P ratio was 3.0 and higher
in Fe-rich particles.
Peak-to-background ratios and ratios of Fe:P
for globoids and Fe-rich particles in zygotic embryos (Table 2) were very similar to those for
female gametophyte tissue (Table 1). Generally,
D.A. Reid et al. / Plant Science 141 (1999) 19–27
24
Table 2
Mean peak-to-background ratios and ratios of Fe:P in globoids and Fe-rich particles from resin embedded thick sections for five
white spruce zygotic embryosa
Region
Tissue
Particle
P
K
Mg
Fe
Fe:P
Cotyledons
Protoderm
Ground meristem
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
4.99 0.9c
5.1 9 1.3c
11.3 9 2.3a
4.7 9 1.4c
6.9 92.2b
3.8 9 1.1c
3.5 9 1.0c
11.8 9 2.0a
3.5 91.1c
10.2 92.5a
4.0 91.9c
3.5 90.9c
12.0 9 1.3a
3.6 9 1.0c
10.5 91.9a
2.5 9 1.0a
2.79 0.9a
7.69 2.5cd
2.9 9 1.0a
5.69 1.2d
2.2 91.0a
2.3 90.7a
9.7 94.0b
2.7 9 1.0a
7.39 2.2cd
2.79 1.9a
2.19 0.8a
9.39 3.4bc
2.7 9 0.9a
7.39 1.5cd
1.19 0.4a
1.3 9 0.3a
4.29 0.8cd
1.09 0.3a
3.29 2.4b
4.79 3.4c
2.39 2.8a
4.4 90.9c
0.8 90.3a
3.6 9 0.7d
1.3 9 1.2a
0.99 0.3a
4.79 0.7c
0.9 90.2a
4.1 9 0.8cd
17.09 4.8bc
18.69 4.9b
0.59 0.4a
15.29 3.0bcd
7.49 5.7e
13.29 4.3cd
14.5 9 5.1bcd
0.2 90.2a
12.8 93.6cd
0.3 90.1a
11.0 94.1d
13.3 9 3.7cd
0.19 0.1a
13.8 93.9cd
0.2 90.1a
3.59 0.8c
3.8 9 0.9c
0.09 0.0a
3.39 0.7c
1.19 0.8b
3.59 0.7c
4.19 0.8c
0.0 90.0a
3.8 90.9c
0.0 9 0.0a
3.3 9 1.5c
3.99 0.8c
0.09 0.0a
4.0 90.7c
0.0 90.0a
Procambium
Hypocotyl
Protoderm
Ground meristem
Procambium
Radicle
Protoderm
Ground meristem
Procambium
a
Each mean ( 9S.D.) was calculated using 25 values; each value in a single column followed by the same letter is not
significantly different at P\0.05.
Fe-rich particles from zygotic embryos had significantly smaller P/B values for P, K and Mg than
globoids. Globoids sampled in the cotyledon procambium tissue in zygotic embryos had significantly higher Fe values and smaller P, K and Mg
values than other globoids within these embryos.
Globoids from the cotyledon procambium tissue
of zygotic embryos had Fe:P ratios above 1.0.
With the exception of this moderate Fe in globoids
from the cotyledon procambium tissue of zygotic
embryos, the composition of globoids and Fe-rich
particles were very consistent from tissue-to-tissue.
Table 3 illustrates the P/B ratios and ratios of
Fe:P for globoids and Fe-rich particles in somatic
embryos. As found with zygotic embryos, the
composition of these deposits did not differ greatly
from tissue-to-tissue. Globoids had significantly
higher P, K and Mg than Fe-rich particles and
globoids in the cotyledon procambium of somatic
embryos did not have the moderate levels of Fe
found in zygotic embryos.
Comparison of globoids and Fe-rich particles in
zygotic and somatic embryos using the MINITAB
two-sample t-test (Table 4) found that, in general,
globoids in zygotic embryos had slightly higher
P/B values for P, K, Fe and Fe:P than globoids in
somatic embryos. Iron-rich particles from zygotic
embryos had slightly higher Fe:P values and significantly lower P and K values than Fe-rich particles in somatic embryos. Overall, globoid and
Fe-rich particle compositions were very similar in
somatic and zygotic embryos.
4. Discussion
The focus of this study was phytate-containing
globoids that are found inside protein bodies [16].
The low-water-content preparation procedures
used here were designed to retain the phytate in
the tissue. It has been shown that K-phytate and
Na-phytate are very water-soluble, while phytates
containing divalent and trivalent cations tend to
be less soluble in water [17,18]. Aqueous fixation
and thin sectioning on a water-filled microtome
boat can result in major losses of P, Mg and K
[19]. Prolonged fixation and washing can cause the
complete extraction of globoids [20]. Thicker sections cut on dry knives prevent the shattering of
globoids as well as eliminating extraction problems during microtomy. Since globoids are naturally electron dense, they can be readily identified
in these thicker sections and the addition of electron-dense stains can be omitted.
Energy dispersive X-ray analysis of individual
globoids in white spruce somatic embryos, zygotic
embryos and female gametophyte tissue revealed
the presence of P, K, and Mg with occasional
traces of Fe and little if any Ca or Zn. All
globoids sampled from somatic embryos, zygotic
D.A. Reid et al. / Plant Science 141 (1999) 19–27
25
Table 3
Mean peak-to-background ratios and ratios of Fe:P in globoids and Fe-rich particles from resin embedded thick sections for five
white spruce somatic embryosa
Region
Tissue
Particle
P
K
Mg
Fe
Fe:P
Cotyledons
Protoderm
Ground meristem
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
4.4 9 1.1d
5.5 9 1.6d
10.3 9 2.1ab
4.8 9 1.4d
10.2 9 1.7ab
4.8 9 1.1d
4.7 9 1.5d
11.1 9 2.1a
4.7 9 1.3d
8.7 9 2.5bc
4.8 9 1.3d
5.2 9 2.0d
10.6 9 1.9a
4.9 9 2.0d
7.4 9 2.2c
2.6 90.8e
3.19 1.2e
6.1 91.4bc
2.79 1.1e
6.49 1.9b
3.0 90.8e
3.69 1.1de
8.89 2.5a
3.0 90.8e
6.3 91.7bc
3.49 1.0e
3.4 91.4e
8.39 2.0a
3.29 1.5e
4.9 91.9cd
0.9 90.3d
1.2 9 0.4d
4.2 9 1.0b
1.1 90.3d
4.49 1.0ab
1.0 90.2d
1.0 90.3d
4.69 0.7ab
1.0 9 0.4d
3.5 91.1c
1.0 9 0.2d
1.1 90.4d
4.89 1.0a
1.0 9 0.4d
3.2 90.7c
12.89 3.6b
14.69 4.4b
0.2 90.2a
13.99 4.9b
0.19 0.1a
14.0 93.9b
13.6 9 4.5b
0.29 0.2a
14.1 95.3b
0.1 90.1a
14.79 4.6b
16.49 6.6b
0.19 0.2a
14.4 96.1b
0.1 90.2a
2.9 90.6bc
2.69 0.4c
0.09 0.0a
2.9 90.5bc
0.0 90.0a
2.9 9 0.5bc
2.9 90.4bc
0.0 90.0a
2.99 0.6bc
0.0 9 0.0a
3.19 0.5b
3.1 90.4b
0.0 90.0a
3.09 0.4b
0.0 9 0.0a
Procambium
Hypocotyl
Protoderm
Ground meristem
Procambium
Radicle
Protoderm
Ground meristem
Procambium
a
Each mean ( 9S.D.) was calculated using 25 values; each value in a single column followed by the same letter is not
significantly different at P\0.05.
embryos and female gametophytes had this typical
composition except for globoids from the cotyledon procambium of zygotic embryos, which had
Fe peaks similar in height to K. Energy dispersive
X-ray analysis of globoids for white spruce, as
investigated in this study, showed similar compositions to globoids from seeds of various Pinus
species [6]. As found here for white spruce, the
protein bodies in Pinus also contained one or more
globoids [6]. In zygotic and somatic embryos of
white spruce, protein bodies often contained one
globoid, but as many as three were observed in a
Figs. 5 – 8. Energy dispersive X-ray analysis spectra of
globoids and Fe-rich particles from thick sections of white
spruce zygotic and somatic embryos. The energy lines for
each element illustrated are as follows: Mg (Ka = 1.2 keV); P
(Ka =2.0 keV); Cl (Ka =2.6 keV); K (Ka = 3.3 keV; Kb = 3.6
keV); and Fe (Ka = 6.4 keV; Kb =7.1 keV). Note the peaks at
8.0 and 8.9 keV are the Cu Ka and Cu Kb peaks, respectively,
from the copper grids used for holding the sections. Fig. 5.
Typical EDX analysis spectrum of a globoid from the
hypocotyl ground meristem tissue of a somatic embryo.
single section.
It has been proposed, based on studies of angiosperm seeds, that the size and frequency of
globoids was related to the ratio of divalent
cations (Mg2 + and Ca2 + ) to monovalent cations
(K + ), as measured per g dry weight [21]. According to this hypothesis, seed tissues with a high
(Mg +Ca):K ratio have larger and more frequent
globoids; whereas, a low ratio is correlated to
smaller and less frequent globoids. A lower ratio
would result in more K-phytate, which is more
water-soluble and therefore would be present
throughout the proteinaceous matrix rather than
in discrete globoids [21,22]. In part I of this two
paper series, atomic absorption spectroscopy was
used to measure the concentrations of K, Mg and
Ca. From these results the (Mg +Ca):K ratios for
Fig. 6. Energy dispersive X-ray analysis spectrum of a globoid
from the cotyledon procambium tissue of a zygotic embryo
showing moderate Fe levels.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
26
Fig. 7. Typical EDX analysis spectrum of an Fe-rich particle
from the radicle ground meristem of a somatic embryo showing the P peak higher than the Fe peak.
Fig. 8. Typical EDX analysis spectrum of an Fe-rich particle
from the radicle ground meristem of a zygotic embryo showing the P peak lower than the Fe peak.
white spruce somatic embryos, zygotic embryos
and female gametophytes, were 0.38, 0.49 and
0.62, respectively. White spruce somatic embryos,
with the smallest (Mg + Ca):K ratio, had smaller
globoids and less frequent globoids than in either
zygotic embryos or female gametophytes. White
spruce female gametophyte tissue, with the highest
(Mg + Ca):K ratio, had the largest globoids and
the highest frequency of globoids. These results
support the theory proposed for angiosperm seeds
[21] and this is the first report of this ratio for
conifer seeds or somatic embryos. The relationship
of this ratio to globoid size and frequency therefore appears to be valid in both gymnosperm and
angiosperm seed tissues.
The second type of storage material studied here
was Fe-rich particles. The composition of Fe-rich
particles in white spruce seeds and somatic embryos are consistent with those reported for Pinus
seeds [6]. Recently, seeds from nine genera in the
Pinaceae family were studied for the presence and
composition of Fe-rich particles [23]. It was found
that Fe-rich particles are a common characteristic
of conifers in the family Pinaceae and that they
have similar compositions with the exception of
lower Fe:P ratios in Cedrus and Abies [23]. From
EDX analysis studies and comparisons to prepared phytate and phytoferritin deposits, it was
proposed that Fe-rich particles represent stores of
Fe-rich phytate [6,7,23]. The findings of this current study are consistent with that proposal.
The current study has found that phytate stores
within white spruce somatic embryos are very
similar in location and composition to those found
in white spruce zygotic embryos. It was found that
the (Mg+Ca):K ratio for whole tissue described
for angiosperm seeds holds true for this conifer
species. The fact that somatic embryos are produced using prepared media holds the possibility
of altering the media composition of Mg, Ca, and
K to study their effects on phytate biosynthesis. It
may be possible to influence globoid size and
frequency by optimizing the ratio of these elements within the media. In addition, since a fully
mature white spruce somatic embryo now closely
resembles a mature white spruce zygotic embryo,
this system of culturing white spruce somatic embryos may be very useful in studying the processes
of phytate biosynthesis. It can be very difficult to
obtain large quantities of conifer embryos at vari-
Table 4
MINITAB two-sample t-test comparison of mean peak-to-background values and ratios of Fe:P in globoids and Fe-rich particles
from white spruce somatic and zygotic embryos. For each value, all the globoids in somatic embryos were compared to all the
globoids in zygotic embryos and all the Fe-rich particles in somatic embryos were compared to all the Fe-rich particles in zygotic
embryosa
Particle type
Embryo type
P
K
Mg
Fe
Fe:P
Globoid
Somatic
Zygotic
Somatic
Zygotic
9.7 9 2.4a
10.5 9 2.6b
4.9 91.5c
4.1 9 1.3d
6.89 2.3a
7.89 2.9b
3.19 1.1c
2.59 1.1d
4.19 1.1a
3.9 91.1a
1.0 90.3c
1.09 0.5c
0.1 9 0.2a
1.4 93.5b
14.39 5.0c
14.4 94.7c
0.09 0.0a
0.29 0.5b
2.9 90.5c
3.79 0.9d
Fe-rich particle
a
Each mean ( 9S.D.) for globoids was calculated using 150 values and each mean ( 9S.D.) for Fe-rich particles was
calculated using 225 values; each value in a single column followed by the same letter is not significantly different at P\0.05.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
ous stages of development to perform detailed
biochemical and molecular studies. Somatic embryos offer a virtually unlimited supply of embryos at every stage and these embryos may be
useful in discovering some of the key aspects of
phytate production, which could then be further
explored in zygotic embryos. Somatic embryos
offer the potential to learn about aspects of mineral nutrition that have, for various reasons, been
very difficult to study in natural zygotic embryos.
Acknowledgements
This work was supported by Natural Sciences
and Engineering Research Council of Canada
(NSERC) grants awarded to John N.A. Lott and
Larry C. Fowke and an NSERC/industrial grant
with Pacific Regeneration Technologies Inc. (Victoria, BC) awarded to Larry C. Fowke and
Stephen M. Attree. We would like to thank Marcia West, Klaus Schultes and Tim Dament for
their advice and technical assistance in this study.
References
[1] J.N.A. Lott, E. Spitzer, X-ray analysis studies of elements stored in protein body globoid crystals of Triticum
grains, Plant Physiol. 66 (1980) 494 – 499.
[2] J.A. Chandler, X-ray microanalysis in the electron microscope, in: A.M. Glauert, (Ed.), Practical Methods in
Electron Microscopy, Elsevier/North-Holland Biomedical Press, Amsterdam, The Netherlands, 1977, pp. 327 –
518.
[3] D.J. Cosgrove, The chemistry and bio-chemistry of inositol polyphosphates, Rev. Pure Appl. Chem. 16 (1966)
209 – 224.
[4] L.F. Johnson, M.E. Tate, Structure of ‘phytic acids’,
Can. J. Chem. 47 (1969) 63 – 73.
[5] J.N.A. Lott, Accumulation of seed reserves of phosphorus and other minerals, in: D.R. Murray (Ed.), Seed
Physiology, vol. 1 Development, Academic Press, Australia, 1984, pp. 139 – 166.
[6] M.M. West, J.N.A. Lott, Studies of mature seeds of
eleven Pinus species differing in seed weight. II. subcellular structure and localization of elements, Can. J. Bot. 71
(1993) 577 – 585.
[7] J. Pittermann, M. West, J.N.A. Lott, Characterization of
globoids and iron-rich particles in cotyledons of Pinus
banksiana seeds and seedlings, Can. J. For. Res. 26
.
27
(1996) 1697– 1702.
[8] B.B. Hyde, A.J. Hodge, A. Kahn, M.L. Birnstiel, Studies
on phytoferritin. I. identification and localization, J.
Ultrastruct. Res. 9 (1963) 248 – 258.
[9] S.M. Attree, M.K. Pomeroy, L.C. Fowke, Production of
vigorous, desiccation tolerant white spruce (Picea glauca
[Moench.] Voss.) synthetic seeds in a bioreactor, Plant
Cell Rep. 13 (1994) 601 – 606.
[10] S.M. Attree, M.K. Pomeroy, L.C. Fowke, Development
of white spruce (Picea glauca (Moench.) Voss) somatic
embryos during culture with abscisic acid and osmoticum, and their tolerance to drying and frozen storage, J. Exp. Bot. 46 (1995) 433 – 439.
[11] J.N.A. Lott, D.J. Goodchild, S. Craig, Studies of mineral
reserves in pea (Pisum sati6um) cotyledons using low-water-content procedures, Aust. J. Plant Physiol. 11 (1984)
459 – 469.
[12] P. Beecroft, J.N.A. Lott, Changes in the element composition of globoids from Cucurbita maxima and Cucurbita
andreana cotyledons during early seedling growth, Can.
J. Bot. 74 (1996) 838 – 847.
[13] I. Ockenden, J.N.A. Lott, Beam sensitivity of globoid
crystals within seed protein bodies and commercially
prepared phytates during X-ray microanalysis, Scanning
Microsc. 5 (1991) 767 – 778.
[14] A. Stewart, H. Nield, J.N.A. Lott, An investigation of
the mineral content of barley grains and seedlings, Plant
Physiol. 86 (1988) 93 – 97.
[15] J.H. Zar, Biostatistical analysis, Prentice-Hall, Englewood Cliffs, NJ, 1984.
[16] J.N.A. Lott, Protein bodies in seeds, Nord. J. Bot. 1
(1981) 421 – 432.
[17] E.C. Brown, M.L. Heit, D.E. Ryan, Phytic acid: an
analytical investigation, Can. J. Bot. 39 (1961) 1290–
1297.
[18] D.E.C. Crean, D.R. Haisman, The interaction between
phytic acid and divalent cations during the cooking of
dried peas, J. Sci. Food Agric. 14 (1963) 824 – 833.
[19] H.R. Skilnyk, J.N.A. Lott, Mineral analyses of storage
reserves of Cucurbita maxima and Cucurbita andreana
pollen, Can. J. Bot. 70 (1992) 491 – 495.
[20] T. Wada, E. Maeda, An improved method for the retention of globoids in aleurone grains in light microscopy
and electron microscopy, Jpn. J. Crop Sci. 48 (1979)
206 – 213.
[21] J.N.A. Lott, P.J. Randall, D.J. Goodchild, S. Craig,
Occurrence of globoid crystals in cotyledonary protein
bodies of Pisum sati6um as influenced by experimentally
induced changes in Mg, Ca and K contents of seeds,
Aust. J. Plant Physiol. 12 (1985) 341 – 353.
[22] C.A. Prattley, D.W. Stanley, Protein-phytate interactions
in soybeans. I. localization of phytate in protein bodies
and globoids, J. Feed Biochem. 6 (1982) 243 – 245.
[23] D.A. Reid, H.C. Ducharme, M.M. West, J.N.A. Lott,
Iron-rich particles in embryos of seeds from the family
Pinaceae, Protoplasma 202 (1998) 122 – 133.
Mineral nutrition in White Spruce (Picea glauca [Moench] Voss) seeds
and somatic embryos; II. EDX analysis of globoids and Fe-rich
particles
Daryl A. Reid a, John N.A. Lott a,*, Stephen M. Attree b,1, Larry C. Fowke b
b
a
Department of Biology, McMaster Uni6ersity, Hamilton, ON, L8S 4K1, Canada
Department of Biology, Uni6ersity of Saskatchewan, Saskatoon, SK, S7N 5E2, Canada
Received 13 July 1998; received in revised form 26 October 1998; accepted 28 October 1998
Abstract
The elemental compositions of globoids and Fe-rich particles were investigated in white spruce (Picea glauca [Moench] Voss)
somatic embryos, zygotic embryos and female gametophytes using energy dispersive X-ray analysis. Globoids, phytate deposits in
seed protein bodies, were found throughout the female gametophyte of seeds and in the ground meristem and procambium from
all regions of both somatic and zygotic embryos. Iron-rich particles, believed to be Fe-associated phytate deposits in seed
proplastids, were found throughout the female gametophyte of seeds and in the protoderm, ground meristem, and procambium
from all regions of both somatic and zygotic embryos. Globoids in somatic and zygotic embryos ranged from 0.5– 3.0 mm in
diameter, but globoids typically ranged from 1.5– 2.0 mm in diameter in somatic embryos and from 2.0– 3.0 mm in diameter in
zygotic embryos. Globoids in female gametophyte tissue ranged from 0.5– 6.0 mm in diameter. All Fe-rich particles studied from
somatic embryos and from seeds ranged from 0.14– 0.25 mm in diameter. Globoids in somatic embryos and seeds contained high
P, moderate K and Mg with occasional traces of Fe and little if any Ca and Zn. Globoids in the zygotic embryo cotyledon
procambium tissue also contained moderate levels of Fe and had significantly higher Fe:P ratios, which were not found in any
other regions in seeds or in somatic embryos. Iron-rich particles in somatic embryos and seeds contained high P and Fe, moderate
K and Mg, and little if any Ca and Zn. Typically, spectra of Fe-rich particles in somatic embryos had P peaks higher than Fe
peaks and spectra of Fe-rich particles in seeds had P peaks lower than Fe peaks. Overall, the composition of globoids and Fe-rich
particles in somatic embryos and zygotic embryos were very similar with only minimal differences found. © 1999 Elsevier Science
Ireland Ltd. All rights reserved.
Keywords: Somatic embryos; Seeds; White spruce; EDX analysis; Globoids; Fe-rich particles
1. Introduction
In this study, energy dispersive X-ray (EDX)
analysis was used to investigate the mineral nutrient reserves of white spruce seeds and somatic
embryos. This paper will focus on those elements
* Corresponding author. Tel.: + 1-905-5259140 ext. 24589; fax:
+ 1-905-5226066; e-mail: lott@mcmaster.ca.
1
Present address: Pacific Regeneration Technologies Inc., Victoria,
BC, V8T 2W1, Canada.
measured on a whole tissue basis in paper I of this
two paper series. Energy dispersive X-ray analysis
is a convenient tool for studying the elemental
compositions of specific spots or larger areas
within the sample and allows the simultaneous
detection of elements with a high detection sensitivity [1,2]. We used EDX analysis to determine if
phytate reserves in mature somatic embryos of
white spruce have similar element compositions to
those found in mature white spruce seeds. This
research will provide information on mineral nu-
0168-9452/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved.
PII: S 0 1 6 8 - 9 4 5 2 ( 9 8 ) 0 0 2 1 2 - X
20
D.A. Reid et al. / Plant Science 141 (1999) 19–27
trient storage which may be useful to those developing artificial seeds.
Globoids (or globoid crystals) are naturally electron-dense deposits within seed protein bodies and
are composed of phytate. Phytate is a mixed
cation salt of myo-inositol hexa-kisphosphoric
acid or phytic acid [3,4]. Phytate is the typical
mineral nutrient storage compound in seeds. Due
to their natural electron density and relatively
large size, these inclusions are convenient particles
for spectral EDX analysis. Energy dispersive Xray analysis of globoids from different tissues and
species have shown the presence of P, Mg and K
with occasional storage of Ca, Mn, Na, Zn, Ba,
and Fe [5].
Studying mature seeds from eleven Pinus species, small ( 5 0.33 mm in diameter) naturally electron-dense
inclusions
were
found
in
membrane-bound structures resembling proplastids and were termed Fe-rich particles based on
EDX analysis [6]. These particles contained high P
and Fe, moderate levels of K and Mg, and traces
of Ca, Mn, and Zn [6]. Based on the similarities of
EDX analysis spectra of Fe-rich particles, globoids
and precipitated ferric phytate, it was proposed
that these particles could represent storage deposits of Fe-rich phytate [6,7]. The high levels of P
also suggested that these deposits were not phytoferritin inclusions, which is a common iron storage
compound found in chloroplasts [8].
In this study the elemental composition of
globoids and Fe-rich particles in various regions of
white spruce somatic embryos, zygotic embryos
and female gametophytes was investigated using
EDX analysis. This paper will concentrate on the
levels of P, K, Mg, Ca, Fe, and Zn within these
deposits since these nutrients were of focus in part
one of this two paper series. We believe this to be
the first study of globoid and Fe-rich particle
composition in white spruce and the first comparison of their composition in somatic embryos and
seeds of any species.
2. Materials and methods
2.1. Tissue preparation for EDX analysis
White spruce (Picea glauca [Moench] Voss)
seeds were collected from the Big River and
Christopher Lake regions north of Saskatoon,
Saskatchewan. Mature zygotic embryos were excised from artificially dried seeds using fine-tipped
forceps. White spruce somatic embryos were prepared and desiccated as previously described
[9,10]. To prevent the extraction of water-soluble
phytate, while allowing enough tissue expansion
that epoxy resin can penetrate the tissue, a lowwater-content preparation procedure was used
[11]. Tissue samples were initially placed into 80%
ethanol overnight. These samples were then placed
into 100% ethanol for at least 8 h and again in
100% ethanol overnight. To complete the dehydration, the tissue samples were treated with propylene oxide for at least 7 h.
Using a standard hardness Spurr’s epoxy resin,
the samples were infiltrated on a tissue rotator
using a propylene oxide:Spurr’s epoxy resin series
(2:1, 1:2, 0:1, and 0:1) for 24 h in each solution.
Following infiltration, each embryo was sectioned
into cotyledons, hypocotyl and radicle regions and
each region was placed into a separate rubber
mold. Additionally, each female gametophyte was
divided into cotyledon-associated, hypocotyl-associated and radicle-associated regions prior to dehydration and placed into separate molds
following infiltration with Spurr’s resin. All molds
were filled with Spurr’s resin and allowed to
harden in a 70°C oven overnight.
A Reichert OM U2 ultramicrotome (Reichert,
Austria) was used to cut thick sections (1–1.5 mm)
using dry glass knives. Each section was flattened
using eyelashes and cactus spines mounted on
wooden sticks and picked up using 100 mesh
Formvar-carbon coated copper grids. Drops of
100% ethanol were used to help the sections adhere to the Formvar.
2.2. EDX analysis of globoids and Fe-rich
particles
EDX analysis was carried out at 80 kV and a
magnification of 25 000 using the scanning mode
of a JEOL-1200 EX-II TEMSCAN microscope
(JEOL, Tokyo) and a PGT model IMIX-II microanalysis system (Princeton Gamma-Tech., Princeton, NJ). All spectra were collected for 60 s. with
aperture, spot size, tilt and detector distance kept
the same for all analyses. Only those spectra collected with greater than 1000 counts s − 1. and less
than 5% dead time were used.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
For each of five somatic embryos and five
zygotic embryos, five spectra of globoids from
different cells and of different sizes were collected
from both the ground meristem and procambium
in each of the cotyledon, hypocotyl and radicle
regions. Additionally, for each of these same five
embryos, five spectra of Fe-rich particles were
collected from each of the protoderm, ground
meristem and procambium in each of the cotyledon, hypocotyl and radicle regions. Five spectra of
globoids and five spectra of Fe-rich particles were
collected from each of the cotyledon-associated,
hypocotyl-associated and radicle-associated regions from each of five female gametophytes.
Therefore in total, 525 spectra of Fe-rich particles
and 375 spectra of globoids were collected.
X-ray counts for all the spectra were collected
by integrating the peaks at the selected window
widths: Mg, 1153.7–1354.3 eV; P, 1905.3–2120.7
eV; K, 3193.7 –3432.3 eV; Ca, 3568.6–3813.4 eV;
Mn, 5758.5 –6037.5 eV; Fe, 6259.9–6546.1 eV; Cu,
7891.7 –8200.3 eV; and Zn, 8478.9–8795.1 eV [12].
In order to produce the best fit background line
for all elements, points were connected at the
following eV values: 510, 660, 814, 1453, 1730,
2500, 2800, 3020, 4200, 5400, 8400, 9400, and
11 000 eV. Background subtraction was performed
for each element in each spectrum by subtracting
the number of counts in the background from the
total number of counts in the element. A useful
format to express peak values is in the form of
peak-to-background (P/B) ratios, which have been
defined as the number of counts above the background for a peak divided by the number of
background counts [13]. Peak-to-background ratios were calculated for each element in each spectrum by using the total number of counts before
and after subtraction [14]. Ratios of elements to P
were also calculated since P is the main anion to
which the elements are bound in phytate.
Due to peak overlaps, correction factors were
used to calculate the actual Ca, Fe, and Zn counts
in each spectrum [7]. Correction factors were used
prior to background subtraction. The Kb peak of
K overlaps the Ka peak of Ca and therefore a
correction factor of 8.80% of the total X-ray
counts for the K Ka peak was subtracted from the
total counts in the Ca window to give the actual
counts for Ca. Similarly, the Kb peak of Mn
overlaps the Fe Ka peak and therefore a correction
factor of 11.60% of the total X-ray counts for Mn
21
Ka peak was subtracted from the total counts in
the Fe window to give a corrected Fe value.
Finally, the Kb peak of Cu overlaps the Ka peak
of Zn and therefore 2.00% of the net Cu Ka counts
was subtracted from the total Zn counts to give a
corrected Zn value.
2.3. Statistical analysis
The statistical significance between means was
determined by using the MINITAB analysis of
variance test. When a difference was found,
Tukey’s test was used to determine which of the
means were statistically different at P \0.05 [15].
Due to the use of manually predetermined background points and relatively short sampling times,
several negative counts were obtained for Ca, Fe
and Zn P:B values as well as for the Ca:P, Fe:P,
and Zn:P ratios. For each P/B value and ratio of
P/B values, all the negative values from all the
spectra collected were averaged and any means
equal to or less than the absolute value of this
mean were considered to be not significantly different from the background line and were assigned
a value of zero. Peak-to-background values and
ratios of P/B values for individual embryos were
analysed to see if there were any significant variations from embryo-to-embryo.
The MINITAB two-sample t-test for the difference between two means at P \0.05 was used to
compare EDX analysis values for somatic embryos to values from zygotic embryos. In this
comparison, for each P/B value and ratio of P/B
values all the globoids in somatic embryos were
pooled together and compared to all the globoids
pooled together for zygotic embryos. Each element
P/B value and each ratio of an element to P in
somatic embryo globoids were compared to the
corresponding value in zygotic embryo globoids.
This was then repeated for all the EDX analysis
values for Fe-rich particles.
3. Results
3.1. General obser6ations
Naturally electron-dense globoids were found
throughout the female gametophyte tissue and in
the ground meristem and procambium in the
cotyledons, hypocotyl and radicle regions of so-
22
D.A. Reid et al. / Plant Science 141 (1999) 19–27
Figs. 1 – 4. Scanning transmission electron micrographs of
thick sections (1– 1.5 mm) from white spruce somatic embryos,
zygotic embryos and female gametophytes. Fig. 1. Naturally
electron-dense globoids (G) variable in size and a smaller
Fe-rich particle (F) in ground meristem tissue from the radicle
of a somatic embryo. Scale bar, 2 mm.
matic and zygotic embryos. Globoids were not
found in the protoderm of either somatic or
zygotic embryos. Typically, globoids in somatic
embryos were found to be 0.5 –2.0 mm in diameter
(Figs. 1 and 2), but were occasionally observed to
be 3.0 mm in diameter. Globoids in zygotic embryos were also found to be between 0.5 and 3.0
mm in diameter (Fig. 3), but typically they were
found to be between 2.0 and 3.0 mm in diameter.
Female gametophyte tissue was found to contain
globoids that ranged in size from 0.5 –6.0 mm in
Fig. 2. Hypocotyl ground meristem tissue of a somatic embryo showing a small Fe-rich particle (F) and protein bodies
(P), which contain globoids (G). Scale bar, 2 mm.
Fig. 3. Iron-rich particles (F) and protein bodies (P), which
contain globoids (G), are shown in the hypocotyl ground
meristem tissue of a zygotic embryo. Scale bar, 2 mm.
diameter (Fig. 4). Globoids in the ground meristem of somatic and zygotic embryos tended to be
larger than those found in the procambium of the
same embryo.
Generally only one globoid was observed per
section of a single protein body in somatic and
zygotic embryos, but occasionally as many as
three were visible. Protein bodies from female
gametophyte tissue varied in the number of
globoids per section, with some protein bodies
containing one globoid and another protein body
within the same cell containing four or more
globoids. Globoids were most frequent per unit
area in female gametophyte tissue and least frequent per unit area in somatic embryos.
Fig. 4. Female gametophyte tissue containing globoids (G)
variable in size within protein bodies (P). Scale bar, 6 mm.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
23
Table 1
Mean peak-to-background ratios and ratios of Fe:P in globoids and Fe-rich particles from resin embedded thick sections for five
white spruce female gametophytesa
Region
Particle
P
K
Mg
Fe
Fe:P
Cotyledon-associated
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Globoid
3.39 1.0a
11.4 9 1.9b
3.3 9 1.1a
12.09 1.9b
4.2 91.7a
11.7 92.2b
2.3 9 0.8a
8.7 92.4b
2.5 9 0.9a
10.39 1.4b
2.79 1.4a
9.99 2.2b
0.9 90.3a
3.9 91.1b
0.9 90.2a
4.5 9 0.7b
0.99 0.3a
4.79 1.1b
9.8 93.3a
0.5 9 0.3b
10.99 4.6a
0.3 9 0.2b
13.79 5.9a
0.29 0.2b
3.09 0.6a
0.1 90.0b
3.2 90.6a
0.0 9 0.0b
3.49 0.8a
0.09 0.0b
Hypocotyl-associated
Radicle-associated
a
Each mean ( 9S.D.) was calculated using 25 values; each value in a single column followed by the same letter is not
significantly different at P\0.05.
Iron-rich particles were found in the protoderm,
ground meristem and procambium from the
cotyledons, hypocotyl and radicle regions of somatic and zygotic embryos (Figs. 1–3). Iron-rich
particles were also found throughout the female
gametophyte tissue, but they tended to be very
difficult to locate. Typically Fe-rich particles were
0.14 –0.25 mm in diameter and were also naturally
electron-dense. Occasionally Fe-rich particles were
found in clusters in embryo tissues, but never in
female gametophyte tissue.
3.2. EDX analysis of globoids and Fe-rich
particles
Typical EDX analysis spectra (based on the
mean P/B values given in Tables 1–3) of globoids
from somatic embryos, zygotic embryos and female gametophytes were similar in composition
and a representative spectrum is shown in Fig. 5.
Typical globoids contained high levels of P, moderate levels of Mg and K, occasional traces of Fe
and little if any Ca and Zn. One exception to this
typical composition was that globoids from the
procambium of zygotic embryo cotyledons contained moderate levels of Fe and slightly reduced
K and Mg (Fig. 6). Spectra of globoids from the
same tissue region in somatic embryos showed no
significant levels of Fe and more typical K and Mg
levels.
Iron-rich particles in somatic embryos were
found to be similar in composition to those found
in zygotic embryos and female gametophytes. All
Fe-rich particles contained high P and Fe with
moderate levels of K and Mg, and little if any Ca
and Zn. Iron-rich particles sampled from somatic
embryos typically had P peaks higher than Fe
peaks (Fig. 7); whereas, Fe-rich particles from
seeds typically had P peaks lower than Fe peaks
(Fig. 8). Although Fe-rich particles in seeds and
somatic embryos had means that differed in the
proportions of P to Fe, particles of similar P and
Fe proportions were found in both seeds and
somatic embryos. Comparison of a typical globoid
spectrum (Fig. 5) with Fe-rich particle spectra
(Figs. 7 and 8), show that K and Mg peaks were
slightly lower in Fe-rich particles.
Mean P/B values and the ratio of Fe:P for
female gametophytes, zygotic embryos, and somatic embryos are given in Tables 1–3, respectively. Since little if any Ca or Zn were detected in
globoids and Fe-rich particles, these values are not
given in these tables. No significant differences
were found in P/B values from different female
gametophytes or from embryo-to-embryo for either somatic or zygotic embryos. The majority of
the variation was found to be from particle-toparticle.
Comparison of P/B ratios and ratios of Fe:P in
globoids and Fe-rich particles tested in various
regions of female gametophyte tissue showed no
significant variation from region-to-region (Table
1). The P P/B values were between 2.7 and 3.6
times higher in globoids than in Fe-rich particles.
Globoids also had between 3.2 and 4.5 times
higher K and 4.3 to 5.2 times higher Mg than
Fe-rich particles. Iron P/B values were between
19.6 and 68.5 times higher in Fe-rich particles than
in globoids. Globoids in the female gametophyte
tissue had trace amounts of Fe and therefore very
low Fe:P ratios. The Fe:P ratio was 3.0 and higher
in Fe-rich particles.
Peak-to-background ratios and ratios of Fe:P
for globoids and Fe-rich particles in zygotic embryos (Table 2) were very similar to those for
female gametophyte tissue (Table 1). Generally,
D.A. Reid et al. / Plant Science 141 (1999) 19–27
24
Table 2
Mean peak-to-background ratios and ratios of Fe:P in globoids and Fe-rich particles from resin embedded thick sections for five
white spruce zygotic embryosa
Region
Tissue
Particle
P
K
Mg
Fe
Fe:P
Cotyledons
Protoderm
Ground meristem
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
4.99 0.9c
5.1 9 1.3c
11.3 9 2.3a
4.7 9 1.4c
6.9 92.2b
3.8 9 1.1c
3.5 9 1.0c
11.8 9 2.0a
3.5 91.1c
10.2 92.5a
4.0 91.9c
3.5 90.9c
12.0 9 1.3a
3.6 9 1.0c
10.5 91.9a
2.5 9 1.0a
2.79 0.9a
7.69 2.5cd
2.9 9 1.0a
5.69 1.2d
2.2 91.0a
2.3 90.7a
9.7 94.0b
2.7 9 1.0a
7.39 2.2cd
2.79 1.9a
2.19 0.8a
9.39 3.4bc
2.7 9 0.9a
7.39 1.5cd
1.19 0.4a
1.3 9 0.3a
4.29 0.8cd
1.09 0.3a
3.29 2.4b
4.79 3.4c
2.39 2.8a
4.4 90.9c
0.8 90.3a
3.6 9 0.7d
1.3 9 1.2a
0.99 0.3a
4.79 0.7c
0.9 90.2a
4.1 9 0.8cd
17.09 4.8bc
18.69 4.9b
0.59 0.4a
15.29 3.0bcd
7.49 5.7e
13.29 4.3cd
14.5 9 5.1bcd
0.2 90.2a
12.8 93.6cd
0.3 90.1a
11.0 94.1d
13.3 9 3.7cd
0.19 0.1a
13.8 93.9cd
0.2 90.1a
3.59 0.8c
3.8 9 0.9c
0.09 0.0a
3.39 0.7c
1.19 0.8b
3.59 0.7c
4.19 0.8c
0.0 90.0a
3.8 90.9c
0.0 9 0.0a
3.3 9 1.5c
3.99 0.8c
0.09 0.0a
4.0 90.7c
0.0 90.0a
Procambium
Hypocotyl
Protoderm
Ground meristem
Procambium
Radicle
Protoderm
Ground meristem
Procambium
a
Each mean ( 9S.D.) was calculated using 25 values; each value in a single column followed by the same letter is not
significantly different at P\0.05.
Fe-rich particles from zygotic embryos had significantly smaller P/B values for P, K and Mg than
globoids. Globoids sampled in the cotyledon procambium tissue in zygotic embryos had significantly higher Fe values and smaller P, K and Mg
values than other globoids within these embryos.
Globoids from the cotyledon procambium tissue
of zygotic embryos had Fe:P ratios above 1.0.
With the exception of this moderate Fe in globoids
from the cotyledon procambium tissue of zygotic
embryos, the composition of globoids and Fe-rich
particles were very consistent from tissue-to-tissue.
Table 3 illustrates the P/B ratios and ratios of
Fe:P for globoids and Fe-rich particles in somatic
embryos. As found with zygotic embryos, the
composition of these deposits did not differ greatly
from tissue-to-tissue. Globoids had significantly
higher P, K and Mg than Fe-rich particles and
globoids in the cotyledon procambium of somatic
embryos did not have the moderate levels of Fe
found in zygotic embryos.
Comparison of globoids and Fe-rich particles in
zygotic and somatic embryos using the MINITAB
two-sample t-test (Table 4) found that, in general,
globoids in zygotic embryos had slightly higher
P/B values for P, K, Fe and Fe:P than globoids in
somatic embryos. Iron-rich particles from zygotic
embryos had slightly higher Fe:P values and significantly lower P and K values than Fe-rich particles in somatic embryos. Overall, globoid and
Fe-rich particle compositions were very similar in
somatic and zygotic embryos.
4. Discussion
The focus of this study was phytate-containing
globoids that are found inside protein bodies [16].
The low-water-content preparation procedures
used here were designed to retain the phytate in
the tissue. It has been shown that K-phytate and
Na-phytate are very water-soluble, while phytates
containing divalent and trivalent cations tend to
be less soluble in water [17,18]. Aqueous fixation
and thin sectioning on a water-filled microtome
boat can result in major losses of P, Mg and K
[19]. Prolonged fixation and washing can cause the
complete extraction of globoids [20]. Thicker sections cut on dry knives prevent the shattering of
globoids as well as eliminating extraction problems during microtomy. Since globoids are naturally electron dense, they can be readily identified
in these thicker sections and the addition of electron-dense stains can be omitted.
Energy dispersive X-ray analysis of individual
globoids in white spruce somatic embryos, zygotic
embryos and female gametophyte tissue revealed
the presence of P, K, and Mg with occasional
traces of Fe and little if any Ca or Zn. All
globoids sampled from somatic embryos, zygotic
D.A. Reid et al. / Plant Science 141 (1999) 19–27
25
Table 3
Mean peak-to-background ratios and ratios of Fe:P in globoids and Fe-rich particles from resin embedded thick sections for five
white spruce somatic embryosa
Region
Tissue
Particle
P
K
Mg
Fe
Fe:P
Cotyledons
Protoderm
Ground meristem
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
Fe-rich particle
Fe-rich particle
Globoid
Fe-rich particle
Globoid
4.4 9 1.1d
5.5 9 1.6d
10.3 9 2.1ab
4.8 9 1.4d
10.2 9 1.7ab
4.8 9 1.1d
4.7 9 1.5d
11.1 9 2.1a
4.7 9 1.3d
8.7 9 2.5bc
4.8 9 1.3d
5.2 9 2.0d
10.6 9 1.9a
4.9 9 2.0d
7.4 9 2.2c
2.6 90.8e
3.19 1.2e
6.1 91.4bc
2.79 1.1e
6.49 1.9b
3.0 90.8e
3.69 1.1de
8.89 2.5a
3.0 90.8e
6.3 91.7bc
3.49 1.0e
3.4 91.4e
8.39 2.0a
3.29 1.5e
4.9 91.9cd
0.9 90.3d
1.2 9 0.4d
4.2 9 1.0b
1.1 90.3d
4.49 1.0ab
1.0 90.2d
1.0 90.3d
4.69 0.7ab
1.0 9 0.4d
3.5 91.1c
1.0 9 0.2d
1.1 90.4d
4.89 1.0a
1.0 9 0.4d
3.2 90.7c
12.89 3.6b
14.69 4.4b
0.2 90.2a
13.99 4.9b
0.19 0.1a
14.0 93.9b
13.6 9 4.5b
0.29 0.2a
14.1 95.3b
0.1 90.1a
14.79 4.6b
16.49 6.6b
0.19 0.2a
14.4 96.1b
0.1 90.2a
2.9 90.6bc
2.69 0.4c
0.09 0.0a
2.9 90.5bc
0.0 90.0a
2.9 9 0.5bc
2.9 90.4bc
0.0 90.0a
2.99 0.6bc
0.0 9 0.0a
3.19 0.5b
3.1 90.4b
0.0 90.0a
3.09 0.4b
0.0 9 0.0a
Procambium
Hypocotyl
Protoderm
Ground meristem
Procambium
Radicle
Protoderm
Ground meristem
Procambium
a
Each mean ( 9S.D.) was calculated using 25 values; each value in a single column followed by the same letter is not
significantly different at P\0.05.
embryos and female gametophytes had this typical
composition except for globoids from the cotyledon procambium of zygotic embryos, which had
Fe peaks similar in height to K. Energy dispersive
X-ray analysis of globoids for white spruce, as
investigated in this study, showed similar compositions to globoids from seeds of various Pinus
species [6]. As found here for white spruce, the
protein bodies in Pinus also contained one or more
globoids [6]. In zygotic and somatic embryos of
white spruce, protein bodies often contained one
globoid, but as many as three were observed in a
Figs. 5 – 8. Energy dispersive X-ray analysis spectra of
globoids and Fe-rich particles from thick sections of white
spruce zygotic and somatic embryos. The energy lines for
each element illustrated are as follows: Mg (Ka = 1.2 keV); P
(Ka =2.0 keV); Cl (Ka =2.6 keV); K (Ka = 3.3 keV; Kb = 3.6
keV); and Fe (Ka = 6.4 keV; Kb =7.1 keV). Note the peaks at
8.0 and 8.9 keV are the Cu Ka and Cu Kb peaks, respectively,
from the copper grids used for holding the sections. Fig. 5.
Typical EDX analysis spectrum of a globoid from the
hypocotyl ground meristem tissue of a somatic embryo.
single section.
It has been proposed, based on studies of angiosperm seeds, that the size and frequency of
globoids was related to the ratio of divalent
cations (Mg2 + and Ca2 + ) to monovalent cations
(K + ), as measured per g dry weight [21]. According to this hypothesis, seed tissues with a high
(Mg +Ca):K ratio have larger and more frequent
globoids; whereas, a low ratio is correlated to
smaller and less frequent globoids. A lower ratio
would result in more K-phytate, which is more
water-soluble and therefore would be present
throughout the proteinaceous matrix rather than
in discrete globoids [21,22]. In part I of this two
paper series, atomic absorption spectroscopy was
used to measure the concentrations of K, Mg and
Ca. From these results the (Mg +Ca):K ratios for
Fig. 6. Energy dispersive X-ray analysis spectrum of a globoid
from the cotyledon procambium tissue of a zygotic embryo
showing moderate Fe levels.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
26
Fig. 7. Typical EDX analysis spectrum of an Fe-rich particle
from the radicle ground meristem of a somatic embryo showing the P peak higher than the Fe peak.
Fig. 8. Typical EDX analysis spectrum of an Fe-rich particle
from the radicle ground meristem of a zygotic embryo showing the P peak lower than the Fe peak.
white spruce somatic embryos, zygotic embryos
and female gametophytes, were 0.38, 0.49 and
0.62, respectively. White spruce somatic embryos,
with the smallest (Mg + Ca):K ratio, had smaller
globoids and less frequent globoids than in either
zygotic embryos or female gametophytes. White
spruce female gametophyte tissue, with the highest
(Mg + Ca):K ratio, had the largest globoids and
the highest frequency of globoids. These results
support the theory proposed for angiosperm seeds
[21] and this is the first report of this ratio for
conifer seeds or somatic embryos. The relationship
of this ratio to globoid size and frequency therefore appears to be valid in both gymnosperm and
angiosperm seed tissues.
The second type of storage material studied here
was Fe-rich particles. The composition of Fe-rich
particles in white spruce seeds and somatic embryos are consistent with those reported for Pinus
seeds [6]. Recently, seeds from nine genera in the
Pinaceae family were studied for the presence and
composition of Fe-rich particles [23]. It was found
that Fe-rich particles are a common characteristic
of conifers in the family Pinaceae and that they
have similar compositions with the exception of
lower Fe:P ratios in Cedrus and Abies [23]. From
EDX analysis studies and comparisons to prepared phytate and phytoferritin deposits, it was
proposed that Fe-rich particles represent stores of
Fe-rich phytate [6,7,23]. The findings of this current study are consistent with that proposal.
The current study has found that phytate stores
within white spruce somatic embryos are very
similar in location and composition to those found
in white spruce zygotic embryos. It was found that
the (Mg+Ca):K ratio for whole tissue described
for angiosperm seeds holds true for this conifer
species. The fact that somatic embryos are produced using prepared media holds the possibility
of altering the media composition of Mg, Ca, and
K to study their effects on phytate biosynthesis. It
may be possible to influence globoid size and
frequency by optimizing the ratio of these elements within the media. In addition, since a fully
mature white spruce somatic embryo now closely
resembles a mature white spruce zygotic embryo,
this system of culturing white spruce somatic embryos may be very useful in studying the processes
of phytate biosynthesis. It can be very difficult to
obtain large quantities of conifer embryos at vari-
Table 4
MINITAB two-sample t-test comparison of mean peak-to-background values and ratios of Fe:P in globoids and Fe-rich particles
from white spruce somatic and zygotic embryos. For each value, all the globoids in somatic embryos were compared to all the
globoids in zygotic embryos and all the Fe-rich particles in somatic embryos were compared to all the Fe-rich particles in zygotic
embryosa
Particle type
Embryo type
P
K
Mg
Fe
Fe:P
Globoid
Somatic
Zygotic
Somatic
Zygotic
9.7 9 2.4a
10.5 9 2.6b
4.9 91.5c
4.1 9 1.3d
6.89 2.3a
7.89 2.9b
3.19 1.1c
2.59 1.1d
4.19 1.1a
3.9 91.1a
1.0 90.3c
1.09 0.5c
0.1 9 0.2a
1.4 93.5b
14.39 5.0c
14.4 94.7c
0.09 0.0a
0.29 0.5b
2.9 90.5c
3.79 0.9d
Fe-rich particle
a
Each mean ( 9S.D.) for globoids was calculated using 150 values and each mean ( 9S.D.) for Fe-rich particles was
calculated using 225 values; each value in a single column followed by the same letter is not significantly different at P\0.05.
D.A. Reid et al. / Plant Science 141 (1999) 19–27
ous stages of development to perform detailed
biochemical and molecular studies. Somatic embryos offer a virtually unlimited supply of embryos at every stage and these embryos may be
useful in discovering some of the key aspects of
phytate production, which could then be further
explored in zygotic embryos. Somatic embryos
offer the potential to learn about aspects of mineral nutrition that have, for various reasons, been
very difficult to study in natural zygotic embryos.
Acknowledgements
This work was supported by Natural Sciences
and Engineering Research Council of Canada
(NSERC) grants awarded to John N.A. Lott and
Larry C. Fowke and an NSERC/industrial grant
with Pacific Regeneration Technologies Inc. (Victoria, BC) awarded to Larry C. Fowke and
Stephen M. Attree. We would like to thank Marcia West, Klaus Schultes and Tim Dament for
their advice and technical assistance in this study.
References
[1] J.N.A. Lott, E. Spitzer, X-ray analysis studies of elements stored in protein body globoid crystals of Triticum
grains, Plant Physiol. 66 (1980) 494 – 499.
[2] J.A. Chandler, X-ray microanalysis in the electron microscope, in: A.M. Glauert, (Ed.), Practical Methods in
Electron Microscopy, Elsevier/North-Holland Biomedical Press, Amsterdam, The Netherlands, 1977, pp. 327 –
518.
[3] D.J. Cosgrove, The chemistry and bio-chemistry of inositol polyphosphates, Rev. Pure Appl. Chem. 16 (1966)
209 – 224.
[4] L.F. Johnson, M.E. Tate, Structure of ‘phytic acids’,
Can. J. Chem. 47 (1969) 63 – 73.
[5] J.N.A. Lott, Accumulation of seed reserves of phosphorus and other minerals, in: D.R. Murray (Ed.), Seed
Physiology, vol. 1 Development, Academic Press, Australia, 1984, pp. 139 – 166.
[6] M.M. West, J.N.A. Lott, Studies of mature seeds of
eleven Pinus species differing in seed weight. II. subcellular structure and localization of elements, Can. J. Bot. 71
(1993) 577 – 585.
[7] J. Pittermann, M. West, J.N.A. Lott, Characterization of
globoids and iron-rich particles in cotyledons of Pinus
banksiana seeds and seedlings, Can. J. For. Res. 26
.
27
(1996) 1697– 1702.
[8] B.B. Hyde, A.J. Hodge, A. Kahn, M.L. Birnstiel, Studies
on phytoferritin. I. identification and localization, J.
Ultrastruct. Res. 9 (1963) 248 – 258.
[9] S.M. Attree, M.K. Pomeroy, L.C. Fowke, Production of
vigorous, desiccation tolerant white spruce (Picea glauca
[Moench.] Voss.) synthetic seeds in a bioreactor, Plant
Cell Rep. 13 (1994) 601 – 606.
[10] S.M. Attree, M.K. Pomeroy, L.C. Fowke, Development
of white spruce (Picea glauca (Moench.) Voss) somatic
embryos during culture with abscisic acid and osmoticum, and their tolerance to drying and frozen storage, J. Exp. Bot. 46 (1995) 433 – 439.
[11] J.N.A. Lott, D.J. Goodchild, S. Craig, Studies of mineral
reserves in pea (Pisum sati6um) cotyledons using low-water-content procedures, Aust. J. Plant Physiol. 11 (1984)
459 – 469.
[12] P. Beecroft, J.N.A. Lott, Changes in the element composition of globoids from Cucurbita maxima and Cucurbita
andreana cotyledons during early seedling growth, Can.
J. Bot. 74 (1996) 838 – 847.
[13] I. Ockenden, J.N.A. Lott, Beam sensitivity of globoid
crystals within seed protein bodies and commercially
prepared phytates during X-ray microanalysis, Scanning
Microsc. 5 (1991) 767 – 778.
[14] A. Stewart, H. Nield, J.N.A. Lott, An investigation of
the mineral content of barley grains and seedlings, Plant
Physiol. 86 (1988) 93 – 97.
[15] J.H. Zar, Biostatistical analysis, Prentice-Hall, Englewood Cliffs, NJ, 1984.
[16] J.N.A. Lott, Protein bodies in seeds, Nord. J. Bot. 1
(1981) 421 – 432.
[17] E.C. Brown, M.L. Heit, D.E. Ryan, Phytic acid: an
analytical investigation, Can. J. Bot. 39 (1961) 1290–
1297.
[18] D.E.C. Crean, D.R. Haisman, The interaction between
phytic acid and divalent cations during the cooking of
dried peas, J. Sci. Food Agric. 14 (1963) 824 – 833.
[19] H.R. Skilnyk, J.N.A. Lott, Mineral analyses of storage
reserves of Cucurbita maxima and Cucurbita andreana
pollen, Can. J. Bot. 70 (1992) 491 – 495.
[20] T. Wada, E. Maeda, An improved method for the retention of globoids in aleurone grains in light microscopy
and electron microscopy, Jpn. J. Crop Sci. 48 (1979)
206 – 213.
[21] J.N.A. Lott, P.J. Randall, D.J. Goodchild, S. Craig,
Occurrence of globoid crystals in cotyledonary protein
bodies of Pisum sati6um as influenced by experimentally
induced changes in Mg, Ca and K contents of seeds,
Aust. J. Plant Physiol. 12 (1985) 341 – 353.
[22] C.A. Prattley, D.W. Stanley, Protein-phytate interactions
in soybeans. I. localization of phytate in protein bodies
and globoids, J. Feed Biochem. 6 (1982) 243 – 245.
[23] D.A. Reid, H.C. Ducharme, M.M. West, J.N.A. Lott,
Iron-rich particles in embryos of seeds from the family
Pinaceae, Protoplasma 202 (1998) 122 – 133.