RESEARCH ARTICLE

A M E R I C A N J O U R N A L O F B O TA N Y

B R I E F C O M M U N I C AT I O N

Pollen wall degradation in the Brassicaceae permits cell

emergence after pollination1
Anna F. Edlund2,4, Katrina Olsen3, Christian Mendoza2, Jing Wang3, Trudyann Buckley2, Mai Nguyen2, Brooke Callahan2,
and Heather A. Owen3

PREMISE OF THE STUDY: Despite attempts to degrade the sporopollenin in pollen walls, this material has withstood a hundred years of experimental treat-
ments and thousands of years of environmental attack in insects and soil. We present evidence that sporopollenin, nonetheless, locally degrades only
minutes after pollination in Arabidopsis thaliana flowers, and describe here a two-part pollen germination mechanism in A. thaliana involving both chemi-
cal weakening of the exine wall and swelling of the underlying intine.
METHODS: We explored naturally occurring components from pollen and stigma surfaces and found a tripartite mix of hydrogen peroxide, peroxidase and
catalase enzymes (all at high levels at the pollination interface) to be experimentally sufficient to degrade the sporopollenin of some Brassicaceae family
members.
KEY RESULTS: At pollination, factors carried on the pollen surface may mix with factors on the stigma surface in a reaction that locally oxidizes the exine
pollen wall. Hydrogen peroxide, catalases, and peroxidases are biologically present at the right time and place and, when mixed experimentally, are suf-
ficient to degrade the walls of susceptible pollen.
CONCLUSIONS: Our work on native biochemistry for breaching sporopollenin suggests new research directions in pollen aperture evolution and could aid
efforts to analyze sporopollenin’s composition, needed for application of this corrosion-resistant, but long-intractable material.

KEY WORDS Arabidopsis thaliana; Brassicaceae; exine; pollen germination; oxidation; sporopollenin

Many of the pollen grain walls ever formed on Earth are still pres-
ent in layers of sediment (Edwards et al., 2017). The sporopollenin
from which they are made withstands degradation by time, fungus,
insect gut, and even treatment with concentrated sulfuric or hydro-
fluoric acids (Erdtman, 1960). Sporopollenin’s chemical resistance
makes it both worthy of, and difficult to, study. Hundreds of re-
search articles have been published in the past century seeking to
characterize sporopollenin’s biochemical composition develop-
mentally (Ariizumi and Toriyama, 2011; Dobritsa et al., 2011;
Quilichini et al., 2015; Zhang et al., 2016), or by extraction (Southworth,
1974), tracers, or solid-state methods (Domínguez et al., 1999;
Guilford et al., 1988; Jardine et al., 2015).
Pollen wall morphology, studied since the 17th century (Grew,
1682; Erdtman, 1952, is better described than wall materials and
1
Manuscript received 26 May 2017; revision accepted 26 July 2017
2
Biology Department, Lafayette College, Easton, Pennsylvania 18042 USA; and
3
Department of Biological Sciences, University of Wisconsin–Milwaukee, 3209 North
Maryland Avenue, Milwaukee, Wisconsin 53211 USA
4
Author for correspondence (e-mail: edlunda@lafayette.edu)
https://doi.org/10.3732/ajb.1700201
known for most pollen grains to include a sporopollenin exine out-
side a cellulosic and pectin-rich intine. The exine itself is further
divided into an outer, often-sculpted sexine and inner, continuous
nexine (Fig. 1A, B). To emerge from durable exine walls after pollina-
tion, the delicate cells inside move through openings (apertures) in
the exine of various shape and number (Edlund et al., 2004), aiming
especially for the aperture closest to the stigma (Wolters-Arts et al.,
1998). But apertures are not the only way out; many pollen grains
are inaperturate (Furness, 2007), and in some flowers of the Brassi-
caceae family, including Arabidopsis thaliana, the pollen tube can
breach inter-aperture exine for direct access to the stigma, regardless
of the positions of the grain’s three large apertures (Hoedemaekers
et al., 2015; Edlund et al., 2016).
When pollen is dusted onto a stigma, contact is sometimes at a
grain’s aperture, but often a grain lies so that the shortest path from
its interior to the stigma is across a wall. When this happens in
A. thaliana, tube cells can take that shortest path; they aim directly
through the wall at the pollen–stigma interface, toward the source
of stigma fluid (Edlund et al., 2016). As a desiccated grain hydrates,
the pectin-rich intine underlying the contact point with the stigma

1266 • A M E R I C A N J O U R N A L O F B OTA N Y 104(8): 1266–1273, 2017; http://www.amjbot.org/ © 2017 Botanical Society of America
 AU G U S T 2017 , V O LU M E 104
swells, straining the exine (Hoedemaekers et al., 2015). We previ-
ously used atomic force microscopy to find that the inter-aperture
exine is an order of magnitude stiffer than the material covering the
aperture (Edlund et al., 2016), but because these properties were
measured in pollen growth medium, not on the stigma, it remained
to be seen (1) whether the wall at the break-out site might be weak-
ened in situ by more than simple mechanical strain and (2) if it
were weakened, then by what biochemistry.
After pollination, factors found in the stigma pellicle and pollen
coat mix at their interface or foot (Elleman et al., 1992). We hypoth-
esized that such a mixture (red and yellow to make orange) might
be responsible for degrading adjacent sporopollenin. Pollen and
stigmas are, in fact, often richly pigmented with red and yellow ca-
rotenoids, as in saffron (Negri et al., 2011). Such antioxidants pro-
tect against radiation (He et al., 2007), pathogen response (Baker
and Orlandi, 1995), and especially against oxidative damage from
the hydrogen peroxide that covers mature, receptive stigmas in
many species (McInnis et al., 2006a; Smirnova, et al., 2014). Along
with antioxidants and hydrogen peroxide, a number of peroxidases
are found at very high levels in stigmas (McInnis et al., 2006a, b;
Smirnova et al., 2014). Peroxidase increases have classically been
used as assays for stigma receptivity (Galen and Plowright, 1987).
Pollen in multiple species, including A. thaliana, are also known to
express high levels of the peroxide-decomposing enzyme catalase
(Mhamdi et al., 2010), and for nearly a century, pollen catalases
have been studied in peroxide-containing honey and nectar (Gillette,
1931; Weston, 2000; Carter and Thornburg, 2004).
Both mechanical and chemical weakening would be evident in
the ultrastructure of A. thaliana’s pollen walls during omni-apertu-
rate germination. And if hydrogen peroxide, catalases, and peroxi-
dases are each present in the flower, at the right time and place,
then perhaps their mixture is sufficient to experimentally degrade
the walls of susceptible pollen. Here we investigated both ultra-
structure and biochemistry during the natural breaching of pollen
walls in A. thaliana; our studies may help with sporopollenin’s
long-sought characterization, at the same time that they raise ques-
tions about aperture evolution and use across taxa.
MATERIALS AND METHODS
Plant growth and field collection—Arabidopsis thaliana seeds were
from the Arabidopsis Biological Resource Center (ABRC; Ohio
State University, Columbus, Ohio, USA) and included Columbia
(Col-4, CS933), Landsberg erecta (Ler, CS20), and Male Sterile-1
(ms-1, CS100). Seeds were sown on soil, stratified at 4°C for 2 d,
and plants were grown under Agrosun fluorescent lights (100 μE;
Hydrofarm, Fairless Hills, Pennsylvania, USA) for 16 h/day at
~50% humidity, watered three times weekly, and fertilized weekly
with Miracle-Gro Plant Food (Scotts Co., Marysville, Ohio, USA).
Brassica rapa (Wisconsin Fast Plant standard, Carolina Biological
Supply, Burlington, North Carolina, USA) was grown from seed in
the laboratory, under 24 h light/day, but otherwise in identical con-
ditions. Flowering Hesperis matronalis, Barbarea orthoceras, and
Barbarea vulgaris plants were collected from the roadside near the
Lafayette College campus in Easton, Pennsylvania (USA) May–July
2015 (40.6986°N, 75.2087°W). Flowering Lunaria annua was col-
lected from East Lake, Atlanta, Georgia (USA) in May 2015 (33.7475°N,
84.3003°W) and replanted in the laboratory. Flowering Arabis
pycnocarpa and Armoracia rusticana were collected from the
• E D LU N D E T A L. — P O L L E N WA L L D E G R A DAT I O N • 1267
roadside in Oxford, Pennsylvania (USA) in April 2016 (39.7854°N,
75.9788°W).
Electron microscopy—Preparation of material by acetolysis for SEM
observation—Gynoecia were removed from mature ms-1 flowers us-
ing a dissecting microscope, and the receptacle was embedded in a
thin layer of 1% w/v agar. At room temperature and humidity, stig-
mas were manually pollinated with wild-type pollen grains of Lands-
berg (WT Ler). After waiting periods of 8–12 min, >12–15 min, or
>15–20 min, stigmas were placed in 5% w/v KOH, and incubated in
an 80° water bath for 10 min, with mixing every 2 min. The tube was
then centrifuged at 9250 RCF for 2 min, and the supernatant was
discarded. The pellet was resuspended in 300 μL concentrated HCl,
mixed well, and water was added to make a total volume of 1 mL. The
tube was then centrifuged at 9250 RCF or 5 min, and the supernatant
was discarded. The pellet was resuspended by adding 1 mL glacial
acetic acid to the tube, and the tube was then centrifuged at 9250 RCF
for 5 min. After discarding the supernatant, 1 mL acetolysis mixture
(9:1 v/v of 99% acetic anhydride and 99.9% concentrated sulfuric
acid; Sigma-Aldrich, St. Louis, Missouri, USA) was slowly added to
the tube. The tube was incubated at 100°C for about 8 min, with mix-
ing every 2 min, and then centrifuged at 9250 RCF for 2 min. At this
time, a dark brown pellet was visible at the bottom of the tube. The
supernatant was removed, and the pellet was resuspended in 1 mL
glacial acetic acid. The tube was centrifuged, the liquid discarded, and
the pellet either stored in the residual liquid at room temperature or
directly processed through the next step. The pellet was washed at
least three times in distilled deionized water and left in 50 μL water at
the last washing step. The tube was vortexed, and about 10 μL of pol-
len suspension was pipetted out and placed onto a poly-l-lysine-
coated 5 × 5 mm silicon chip specimen support (catalog #16008, Ted
Pella, Redding, California, USA). After a wait of about 1 min so that
the pollen debris could precipitate onto the chip, the chip was dehy-
drated with 10% increments of ethanol, and critical-point-dried us-
ing a Balzers Union CPD20 (Oerlikon Balzers, Balzers, Liechtenstein).
After the chips were mounted onto 15-mm stubs with carbon tape,
samples were coated with about 10 nm of iridium using an Emitech
K575× sputter coater (Quorum Emitech, Laughton, UK). A thin strip
of copper tape was used to connect the top of the chip to the base of
the stub for grounding.
Scanning electron microscopy imaging of acetolysed pollen—
Samples prepared by acetolysis were imaged on a Hitachi S-4800
scanning electron microscope (SEM; Tokyo, Japan) at 5 kV acceler-
ating voltage in secondary (SE) detection mode using the mixed
signal from the upper and lower SE detectors.
Specimen preparation for ultrathin sectioning—Stigmas of recep-
tive ms1 flowers were manually pollinated with pollen grains of
wild-type Ler. After 8–12 min, >12–15 min, >15–20 min or >20–30
min, pollinated flowers were placed in fixative overnight at room
temperature. After two rinses in 0.1 M HEPES pH 7.2, tissues were
postfixed in 1% w/v OsO4 in HEPES overnight. Samples were rinsed
in water twice, followed by overnight fixation in 1% w/v uranyl ac-
etate in HEPES, dehydration with 10% increments of acetone, and
infiltration with modified Spurr’s resin (Sigma-Aldrich).
Ultrathin sectioning for transmission electron microscopy—Ultrathin
sections were cut with a diamond knife on a RMC MT 7000
ultramicrotome (Boeckeler, Tucson, Arizona, USA) and picked up
1268 • A M E R I C A N J O U R N A L O F B OTA N Y
on 200-mesh copper grids. Sections were stained with 1% w/v
aqueous uranyl acetate and Reynold’s lead citrate, then imaged
with a Hitachi H-600 transmission electron microscope operating
at 75 kv.
Detection of reactive oxygen species—Carpels were removed from
newly opened flowers of each species and soaked in 50 μM 2,7-
dichlorodihydrofluorescein diacetate (DCFH2; Sigma-Aldrich) in
KCl-MES buffer (5 μm KCl, 10 mM MES, 50 μM CaCl2, pH 6.15).
After 20 min, carpels were mounted in KCl-MES buffer and im-
mediately viewed with the microscope; all images were captured
within 5 min of mounting. Some carpels were pretreated for 20 min
in 100 mM Na-pyruvate (a hydrogen peroxide scavenger) in dis-
tilled water, before they were soaked for 20 min in 50 μM DCFH2
and mounted in KCl-MES buffer.
Experimental sporopollenin degradation—Freshly dehisced an-
thers were collected into half-strength Karnovsky’s fixative (2.5%
v/v gluteraldehyde, 2.7% v/v formaldehyde in 0.15 M HEPES buf-
fer, pH 7.4) from each of the six Brassicaceae species. The Columbia
accession was used for A. thaliana flowers, with anthers from 40
flowers (~240 anthers total) fixed in each microcentrifuge tube.
Tubes were vortexed at low speed for 15 s to separate pollen from
anthers. Anther and filament tissues were removed using forceps.
Tubes were centrifuged at 9250 RCF for 3 min, the supernatant re-
moved, and pellets resuspended in subsets of the following: 500 μL
50% H2O2, 500 μL KH2PO4 buffer, 2360 U/mL horseradish peroxi-
dase (Thermo Scientific, Nazareth, Pennsylvania, USA), and 23 U/
mL beef liver catalase (Sigma-Aldrich). KH2PO4 buffer was added
to dilute the H2O2 to 25% for each treatment. We also treated some
grains with 3% H2O2 only (Fig. 4). We repeated one-, two-, and
three-component treatments of pollen (H2O2 alone, or H2O2 + per-
oxidase, or H2O2 + peroxidase + catalase), using catalase and per-
oxidase enzymes from three sources each (Fig. 3) (catalases: human
erythrocytes, Corynebacterium glutamicum bacteria, and bovine
liver; peroxidases: Phanerochaete chrysosporium white rot fungus,
fungal lignin peroxidase, and horseradish; all from Sigma-Aldrich).
All treated samples were incubated for 1 h at room temperature, in
KH2PO4 buffer alone or in one-, two-, or three-component mix-
tures before spinning at 9250 RCF for 3 min and resuspending in 1 mL
of KH2PO4 buffer, then centrifuging one final time. For visualiza-
tion and scoring of sample degradation, pollen pellets were resus-
pended in one drop of 0.01% w/v auramine O in 0.15 M HEPES
buffer and viewed at 400× magnification with fluorescence micros-
copy. For each treatment, at least 500 grains were scored, across five
replicates. The extent of degradation varied for any single pollen
grain; cracks, small or large holes, or missing wall sectors were all
simply scored as “degraded,” and the percentage degraded grains
was calculated for each control and treatment replicate. All statisti-
cal analyses were conducted using the language R for statistical
computing (R Core Team, 2013, R Foundation for Statistical Com-
puting, Vienna, Austria).
RESULTS
We hand-pollinated and fixed A. thaliana stigmas in a time series,
using concentrated sulfuric acid and acetolysis to remove every-
thing except the sporopollenin ghosts of the germinating pollen
grains, which were viewed with scanning electron microscopy
(SEM). Eight minutes postpollination, these ghosts were flattened
at their interface with the stigma (Fig. 1C, arrow). Several minutes
later, the nexine layer became fenestrated at the floors of the lacu-
nae (Fig. 1D). At >12–15 min after pollination, larger rips were vis-
ible in the nexine, and the baculae tipped sideways, as they lost their
rooting in the degraded nexine (Fig. 1E). By >15–20 min postpol-
lination, a single, large hole had formed where the tube was cross-
ing (Fig. 1F).
From the exterior, by SEM, we could not know whether rips and
fenestrations visible in the nexine (Fig. 1D) were due to thinning or
rupture by mechanical strain alone, or whether additional bio-
chemical degradation was involved. We next used transmission
electron microscopy (TEM) to follow A. thaliana pollen tube emer-
gence in transect. At 8 min postpollination, the vegetative cell was
already polarized toward the stigma surface, with visible changes to
the plasma membrane and cytoplasm adjacent to the stigma, but
without detectable changes to the intine or exine anywhere around
the grain (Fig. 1G). Minutes later, the intine was swollen into a lami-
nated lentil shape, but most strikingly, the local, overlying nexine
layer was absent, while the local, overlying sexine architecture was
disrupted but still intact (Fig. 1H, I). Transects showing the initial
absence of the nexine layer, but not the sexine, suggested bio-
chemical degradation of the nexine and distinguished this germi-
nation mechanism from simple strain or rupture. In other words,
material appeared to be missing, not only torn. The pollen germina-
tion mechanism of A. thaliana can thus be dissected into two
parts—a swelling intine pushing against a weakened exine.
We looked for factors at the pollen–stigma contact sites that
could be responsible for local weakening of exine sporopollenin
(Fig. 2). The three reagents—hydrogen peroxide, peroxidase, and
catalase—are a famous triplet united in the abdomen of the bom-
bardier beetle, where their exothermic, reduction–oxidation reac-
tion is well described (Aneshansley et al., 1969). We suspected that
a similar reaction, at the site of pollination, could oxidize the sporo-
pollenin wall locally, helping to release the escaping pollen tube. To
experimentally mimic the redox reactions of pollination, A. thali-
ana pollen were bathed in hydrogen peroxide mixed with peroxi-
dase and catalase. Fluorescent staining of the treated sporopollenin
with auramine O showed that one quarter of the pollen was frag-
mented and/or globally degraded (Fig. 3A). The three-component
mixture of hydrogen peroxide, peroxidase and catalase degraded
the most grains; the two-part mixture of peroxidase and hydrogen
peroxide degraded a smaller percentage, and hydrogen peroxide,
alone, an even smaller percentage. Peroxidase enzyme by itself did
not degrade pollen walls (Fig. 3A).
The sporopollenin was strikingly sensitive to treatment with
even 3% hydrogen peroxide (Fig. 4), despite its evident complete
resistance to acetolysis (Fig. 1A, B), but we cannot yet relate the ex-
perimental concentrations used for degradation to naturally occur-
ring concentrations of any reagents. In particular, hydrogen
peroxide concentrations of 25% are so high that enzyme denatur-
ation becomes a concern (McInnis et al., 2006b). We were reas-
sured to see lengthy, vigorous bubbling (oxygen production from
catalase activity) and reproducible increases in degradation follow-
ing enzyme addition. Although naturally occurring catalase and
peroxidase enzymes have not yet been isolated from the A. thaliana
flower, RNA levels for endogenous enzymes are high in mature
pollen and receptive stigmas at the time of pollination (Fig. 2). Ad-
ditionally, all six commercially available catalases and peroxidases
 AU G U S T 2017 , V O LU M E 104 • E D LU N D E T A L. — P O L L E N WA L L D E G R A DAT I O N • 1269

FIGURE 1 Scanning (A–F) and transmission electron micrographs (G–I) of Arabidopsis thaliana pollen grains. (A, B) Broken pollen grain after acetolysis,
critical point drying, and sputter coating. Acetolysis has removed the pollen coat, intine, and cellular contents of the grain, leaving only the sporopol-
lenin structures. Nexine (n), baculae (b) and tectum (t). The nexine is degraded during germination. Time series of in situ germination in (C–F) SEM and
(G–I) TEM images. For SEM, stigmas were fixed at (C) 8–12, (D, E) >12–15, and (F) >15–20 min after hand-pollination. Arrow in 1C points to indentation
at the former contact point between stigma cell and pollen grain. For TEM, stigmas were fixed at (G) 8–12 or (H, I) >12–15 min after hand pollination.
Arrow in Fig. 1I points to the edge of the degraded nexine layer. Scale bars: A, I = 1 μm; B = 5 μm; C–G = 2 μm; H = 4 μm.

(regardless of source) were able to increase degradation above that
seen for H2O2 alone (Fig. 3B). A Wilcoxon rank–sum test was used
to compare every peroxidase type with H2O2 alone and the different
catalase types with the horseradish peroxidase treatments. All val-
ues proved to be significant (P < 0.01).
To determine whether pollen grains of other species would be
similarly sensitive to reactive oxygen species (ROS), we treated pol-
len grains from three obligate aperture users (Brassica rapa, Bar-
barea vulgaris, and Barbarea orthoceras), and three “break-out”
species able to germinate through inter-aperture walls (A. thaliana,
Lunaria annua, and Hesperis matronalis; Edlund et al., 2016). By
fluorescence microscopy, only species with natural inter-aperture
germination were detectably sensitive to experimental redox wall
degradation (Fig. 3A). The three species that germinate only through
apertures (Brassica rapa, Barbarea orthoceras, Barbarea vulgaris)
had low sensitivity to ROS degradation, and there was no signifi-
cant variation between treatments, with species considered as an
error term (F2,4 = 3.78, P = 0.12 by ANOVA). Degradation for the
three species that germinate by break-out (Arabidopsis thaliana,
Hesperis matronalis, Lunaria annua), however, differed between
the treatments, with species considered as an error term (F2,4 = 47.1,
P = 0.0016 by ANOVA). For A. thaliana, all pairwise differences
were significant (below P = 0.01) except the pair control and per-
oxidase (P = 0.99, Tukey honestly significant difference [HSD]).
Treatments differed significantly between break-out and nonbreak-
out species, but only slightly for the control (respectively, P < 0.001,
P < 0.001 and P < 0.160, Wilcoxon rank–sum test with Bonferroni
correction).
1270 • A M E R I C A N J O U R N A L O F B OTA N Y

FIGURE 2 Catalases, peroxidases, and reactive oxygen species (ROS) are present at pollination in Arabidopsis thaliana and other Brassicaceae. (A) Heat
maps display microarray RNA abundance data for three catalase and four peroxidase genes (http://arabidopsis-heat-tree.org). Gene expression values
are shown for stages of pollen development (bicellular, mature, and dry pollen; Honys and Twell, 2004), pollen tubes grown in vitro (0.5 h pollen, 4 h
pollen; Qin et al., 2009), pollen tubes that grew through pistil tissue (semi in vitro; Qin et al., 2009), the stigma and ovary (Swanson et al., 2005), the
unpollinated pistil, and pollinated pistils (0.5, 3.5, and 8 h after pollination; Boavida et al., 2011). (B) Carpels stained for ROS. (1) A. thaliana carpel
treated with 100 mM Na-pyruvate (a hydrogen peroxide scavenger) then stained with 50 mM 2,7-dichlorodihydrofluorescein diacetate (DCFH2);
(2) A. thaliana, (3) Arabis pycnocarpa, (4) Armoracia rusticana, (5) Brassica rapa carpels, respectively, stained with DCFH2 to reveal the presence of ROS
on each stigma. Scale bars = 0.5 mm.

DISCUSSION
With so much variety in pollen grain and stigma structures, a single
aperture-based model of pollen germination will not apply for tube
escape across all taxa. An alternative, functionally omni-aperturate
model is found in A. thaliana, and ~25% of other surveyed Brassi-
caceae (Edlund et al., 2016), by which cells escape the pollen grains
through holes in the walls between apertures. We report here the
likely involvement of redox reactions in making these holes and a
simple reagent mixture that might be exploited further for material
science application.
At pollination, pollen grains contact hydrogen peroxide-covered
stigmas. For sensitive species, hydrogen peroxide alone is sufficient
to alter the pollen wall, with degradation increasing if peroxidase
and catalase enzymes are added. One interpretation is that when
catalase mixes with stigma hydrogen peroxide, the hydrogen peroxide
decomposes, producing not only heat, but short-lived, highly reac-
tive oxygen radicals, which together with stigma peroxidases oxi-
dize nearby sporopollenin. Other ROS-associated factors that
localize to stigma and pollen at the time of pollination and could
further affect biochemistry at the interface include superoxide
dismutase (Speranza et al., 2012), nitric oxide (McInnis et al.,
2006a; Hiscock et al., 2007; Sharma and Bhatla, 2013), glutathione
(Zechmann et al., 2011), NADPH respiratory burst oxidase homo-
logs (RBohs; Jiménez-Quesada et al., 2016), and dietary antioxidants
(Alicic et al., 2014). In future studies, pollen germination changes
would be expected in A. thaliana flowers, if ROS-associated factors
were suppressed experimentally. Reminiscent of the bombardier
beetle’s exothermic, redox reactions with hydroquinone, a pollen
hydroquinone alternative called tasselin has also been identified
(Wille and Berhow, 2011). In preliminary studies with a bomb calo-
rimeter, we detected small increases in temperature (~0.0005°)
 AU G U S T 2017 , V O LU M E 104 • E D LU N D E T A L. — P O L L E N WA L L D E G R A DAT I O N • 1271

FIGURE 3 Percentages of degraded Brassicaceae pollen grains following treatment in different mixtures of buffer, H2O2, peroxidase, and catalase. In box
graphs, the line indicates the median, and whiskers are highest and lowest points within 1.5 times the interquartile range. (A) The three species that
germinate only through apertures (reds; Brassica rapa, Barbarea orthoceras, Barbarea vulgaris) were less sensitive to ROS degradation, and there was
no significant variation among treatments with H2O2 and enzymes (F2,4 = 3.78, P = 0.12), whereas the three species that germinate by break-out (blues;
Arabidopsis thaliana, Hesperis matronalis, Lunaria annua) differed among treatments (F2,4 = 47.1, p = 0.0016). (B) Percentages of degraded A. thaliana
pollen grains following treatments with H2O2 and six enzymes (three peroxidases and three catalases). When every peroxidase type was compared
with H2O2 alone and the different catalase types with the two-part H2O2 and horseradish peroxidase treatments, all values differed significantly (P <
0.01). By Wilcoxon rank–sum test, the following differences were all significant: H2O2 alone to H2O2 + horseradish peroxidase (W = 71, P = 0.0012); H2O2
alone to H2O2 + lignin peroxidase (W = 30, P = 0.0043); H2O2 alone to H2O2 + manganese peroxidase (W = 24, P = 0.0095); H2O2 + horseradish peroxidase
to H2O2 + horseradish peroxidase + liver catalase (W = 69, P = 0.0023); H2O2 + horseradish peroxidase to H2O2 + horseradish peroxidase + erythrocyte
catalase (W = 60, P = 0.0018), and H2O2 + horseradish peroxidase to H2O2 + horseradish peroxidase + Corynebacterium catalase (W = 60,
P = 0.0018).

when multiple stigmas are simultaneously dusted with pollen
(data not shown), consistent with catalase’s exothermic decompo-
sition of hydrogen peroxide. Thermogenic flowers, such as the lo-
tus, use a distinct mechanism to make heat for pollinator reward,
although their rise in flower temperature is synchronized with
a rise in stigma peroxidase expression (Seymour and Blaylock,
2000).
Sporopollenin is a dynamic, changeable substance inside flow-
ers, even if it is unchangeable elsewhere (Mascarenhas, 1975). As it
forms in the anther, and first becomes resistant to acetolysis, it may
be cross-linked by peroxide radicals (Scott, 1994). If it is later weak-
ened by oxidation on the stigma, sporopollenin may regain its sen-
sitivity to acetolysis (although we have yet to try treating A. thaliana
pollen sequentially with hydrogen peroxide and then acetolysis so-
lution). Our report of localized exine degradation in the Brassicaceae
joins several past reports, for other species, of more global changes
to exine at the time of pollination, such as spines melting and walls
becoming amorphous or altered in their staining (Gherardini and
Healey, 1969; Dickinson and Lewis, 1974; Rowley and Rowley,
1983). Wall disintegration is also described during spore germi-
nation in the protist Chlamydomonas monoica (Malmberg and
VanWinkle-Swift, 2001). Unlike some other cell walls, however,
the A. thaliana pollen wall is defined (by virtue of its resistance to
acetolysis) as true sporopollenin. One future line of research would
be to explore fertile, inaperturate pollen (found in over 50 families),
looking for additional mechanisms of sporopollenin degradation—
given that even cells in inaperturate grains must have a means of
escape.
Our research accentuates the need for systematic surveys of pol-
len germination mechanisms, like those of pollen morphology and
ontogeny (Erdtman, 1952; Taylor and Osborn, 2006). Other pollen
tubes will be found that fail to emerge at apertures, and other spo-
ropollenin that fails to resist degradation (Hesse and Waha, 1989),
but with what systematic patterns? There is an evolutionary trend
in angiosperms for an increase in aperture number; inaperturate
and monosulcate pollen are common among basal angiosperms
and monocots, whereas eudicots most frequently have three aper-
tures or more. Increasing numbers of apertures, and thus germina-
tion sites, would seem to offer a selective advantage (Furness and
Rudall, 2004), as would the ability to germinate anywhere, directly
through exine walls. It is possible that the true apertures retained in
functionally omni-aperturate pollen, as in A. thaliana, are used
when grains fall aperture-side down and to allow volume change
and fluid and signal movement (Rehman et al., 2004; Vieira and
Feijó, 2016). In other species, pollen tubes may escape omni-aperturate
walls by brute force and/or reduction–oxidation reactions, or these
two steps we describe may be specific to germination in the Brassi-
caceae, or to species having dry stigmas, thin exine, tricellular pollen,
and rapid germination. They do seem to have evolved more than
once in this family (Edlund et al., 2016).
1272 • A M E R I C A N J O U R N A L O F B OTA N Y
FIGURE 4 Reactive oxygen species (ROS) degradation of Arabidopsis thali-
ana sporopollenin. Confocal laser scanning micrographs of pollen grains
after they were fixed in half-strength Karnovsky’s, treated for 1 h in either
(A) KH2PO4 buffer or (B) 3% hydrogen peroxide, washed, and then fluo-
rescently stained with 0.1% w/v auramine O. Scale bars = 10 μm.
Tons of pollen cell walls are available for harvest annually from
fields and woodlands. Uniquely nontoxic, long-lasting, and chemi-
cally unreactive sporopollenin could be developed for lining nuclear
or chemical waste containers and pipes, or for other industrial, trans-
portation, food science, or biomedical coatings. The material behaves
as a strong and species-specific adhesive on the stigma surface (Zinkl
et al., 1999) and even self assembles, forming ornaments and con-
tours by physical, rather than purely genetic, means (Blackmore et al.,
2007). To date, however, sporopollenin’s intractability has slowed
progress on determining its composition and on developing applica-
tions. The redox reactions we describe here provide an early step to-
ward characterizing and exploiting this organic material that persists
thousands of years, withstands concentrated sulfuric acid, and yet
degrades with the simple tripartite combination of hydrogen perox-
ide, peroxidase, and catalase at the moment of pollination.
ACKNOWLEDGEMENTS
This research was supported by National Science Foundation Grant
RUI#0744943 to A.F.E. The authors thank the reviewers for helpful
comments, A. Antia for statistics and graphing, and E. Kasumi,
J. Redes, and C. Yerdon for technical help.
CONTRIBUTIONS
A.F.E. conceived the study and wrote the manuscript. Transmission
electron microscopy was done by K.O. and H.A.O., scanning
electron microscopy by J.W. and H.A.O. B.C. stained Brassicaceae
stigmas for hydrogen peroxide. C.M., T.B., and M.N. treated pollen
from six Brassicaceae species using hydrogen peroxide, peroxidase,
and catalase mixtures.
AUTHOR INFORMATION
The authors declare no competing financial interests. A.F.E. and
Lafayette College have filed a provisional patent application for the
sporopollenin degradation protocol described in this manuscript
and for sporopollenin application in coatings and adhesives.
LITERATURE CITED
Alicic, D., Subaric, D., Jasic, M., Pasalic, H., and D. Ackar. 2014. Antioxidant
properties of pollen. Hrana u Zdravlju i Bolesti, Znanstveno-Strucni Casopis
za Nutricionizam i dijetetiku 3: 6–12.
Aneshansley, D. J., T. Eisner, J. M. Widom, and B. Widom. 1969. Biochemistry
at 100°C: Explosive secretory discharge of bombardier beetles (Brachinus).
Science 165: 61–63.
Ariizumi, T., and K. Toriyama. 2011. Genetic regulation of sporopollenin syn-
thesis and pollen exine development. Annual Review of Plant Biology 62:
437–460.
Baker, C. J., and E. W. Orlandi. 1995. Active oxygen in plant pathogenesis.
Annual Review of Phytopathology 33: 299–321.
Blackmore, S., A. H. Wortley, J. J. Skvarla, and J. R. Rowley. 2007. Pollen wall
development in flowering plants. New Phytologist 174: 483–498.
Boavida, L., F. Borges, J. D. Becker, and J. A. Feijo. 2011. Whole genome
analysis of gene expression reveals coordinated activation of signaling and
metabolic pathways during pollenpistil interactions in Arabidopsis. Plant
Physiology 155: 2066–2080.
Carter, C., and R. W. Thornburg. 2004. Is the nectar redox cycle a floral defense
against microbial attack? Trends in Plant Science 9: 320–324.
Dickinson, H. G., and D. Lewis. 1974. Changes in the pollen grain wall of
Linum grandiflorum following compatible and incompatible intraspecific
pollinations. Annals of Botany 38: 23–29.
Dobritsa, A. A., A. Geanconteri, J. Shrestha, A. Carlson, N. Kooyers, D.
Coerper, E. Urbanczyk-Wocniak, et al. 2011. A large-scale genetic screen
in Arabidopsis to identify genes involved in pollen exine production. Plant
Physiology 157: 947–970.
Domínguez, E., J. A. Mercado, M. A. Quesada, and A. Heredia. 1999. Pollen
sporopollenin: Degradation and structural elucidation. Sexual Plant Repro-
duction 12: 171–178.
Edlund, A. F., R. Swanson, and D. Preuss. 2004. Pollen and stigma structure and
function: The role of diversity in pollination. Plant Cell 16: S84–S97.
Edlund, A. F., Q. Zheng, N. Lowe, S. Kuseryk, K. L. Ainsworth, R. H. Lyles, S. J.
Sibener, and D. Preuss. 2016. Pollen from Arabidopsis thaliana and other
Brassicaceae are functionally omniaperturate. American Journal of Botany
103: 1006–1019.
Edwards, K. J., R. M. Fyfe, and S. T. Jackson. 2017. The first 100 years of pollen
analysis. Nature Plants 3: 17001.
 AU G U S T 2017 , V O LU M E 104
Elleman, C. J., V. Franklin-Tong, and H. G. Dickinson. 1992. Pollination in
species with dry stigmas: The nature of the early stigmatic response and the
pathway taken by pollen tubes. New Phytologist 121: 413–424.
Erdtman, G. 1952. Pollen morphology and plant taxonomy: Angiosperms.
Almqvist and Wiksell, Stockholm, Sweden.
Erdtman, G. 1960. The acetolysis method. Svensk Botanisk Tidskrift 54:
561–564.
Furness, C. A. 2007. Why does some pollen lack apertures? A review of in-
aperturate pollen in eudicots. Botanical Journal of the Linnean Society 155:
29–48.
Furness, C. A., and P. J. Rudall. 2004. Pollen aperture evolution: A crucial fac-
tor for eudicot success? Trends in Plant Science 9: 154–158.
Galen, C., and R. C. Plowright. 1987. Testing the accuracy of using peroxidase
activity to indicate stigma receptivity. Canadian Journal of Botany 65: 107–111.
Gherardini, G. L., and P. L. Healey. 1969. Dissolution of outer wall of pollen
grain during pollination. Nature 224: 718–719.
Gillette, C. C. 1931. Honey catalase. Journal of Economic Entomology 24:
605–606.
Grew, N. 1682. The anatomy of plants. Digital copy available at http://www.
botanicus.org/item/31753000008869. Botanicus [digital library], Missouri
Botanical Garden, St. Louis, Missouri, USA.
Guilford, W. J., D. M. Schneider, J. Labowitz, and S. J. Opella. 1988. High reso-
lution solid state 13C NMR spectroscopy of sporopollenins from different
plant taxa. Plant Physiology 86: 134–136.
He, J.-M., X.-L. Bai, R.-B. Wang, B. Cao, and X.-P. She. 2007. The involvement of
nitric oxide in ultraviolet-B-inhibited pollen germination and tube growth of
Paulownia tomentosa in vitro. Physiologia Plantarum 131: 273–282.
Hesse, M., and M. Waha. 1989. A new look at the acetolysis method. Plant
Systematics and Evolution 163: 147–152.
Hiscock, S. J., J. Bright, S. M. McInnis, R. Desikan, and J. T. Hancock. 2007.
Signaling on the stigma: Potential new roles for ROS and NO in plant signal-
ing. Plant Signaling & Behavior 2: 23–24.
Hoedemaekers, K., J. Derksen, S. W. Hoogstrate, M. Wolters-Arts, S.-A. Oh, D.
Twell, C. Mariani, and I. Rieu. 2015. Bursting pollen is required to organize
the pollen germination plaque and pollen tube tip in Arabidopsis thaliana.
New Phytologist 206: 255–267.
Honys, D., and D. Twell. 2004. Transcriptome analysis of haploid male game-
tophyte development in Arabidopsis. Genome Biology 5: R85.
Jardine, P. E., W. T. Fraser, B. H. Lomax, and W. D. Gosling. 2015. The impact
of oxidation on spore and pollen chemistry. Journal of Micropalaeontology 34:
139–149.
Jiménez-Quesada, M. J., J. A. Traverso, and J. de Dios Alche. 2016. NADPH
oxidase-dependent superoxide production in plant reproductive tissues.
Frontiers in Plant Science 7: 359.
Malmberg, A. E., and K. P. VanWinkle-Swift. 2001. Zygospore germination
in Chlamydomonas monoica (Chlorophyta): Timing and pattern of second-
ary zygospore wall degradation in relation to cytoplasmic events. Journal of
Phycology 37: 86–94.
Mascarenhas, J. P. 1975. The biochemistry of angiosperm pollen development.
Botanical Review 41: 259–314.
McInnis, S. M., R. Desikan, J. T. Hancock, and S. J. Hiscock. 2006a. Production
of reactive oxygen species and reactive nitrogen species by angiosperm
stigmas and pollen: Potential signaling crosstalk. New Phytologist 172:
221–228.
McInnis, S. M., D. C. Emery, R. Porter, R. Desikan, J. T. Hancock, and S. J. Hiscock.
2006b. The role of stigma peroxidases in flowering plants: Insights from fur-
ther characterization of a stigma-specific peroxidase (SSP) from Senecio squal-
idus (Asteraceae). Journal of Experimental Botany 57: 1835–1846.
Mhamdi, A., G. Queval, S. Chaouch, S. Vanderauwera, F. van Breusegem, and G.
Noctor. 2010. Catalase function in plants: A focus on Arabidopsis mutants
as stress-mimic models. Journal of Experimental Botany 61: 4197–4220.
• E D LU N D E T A L. — P O L L E N WA L L D E G R A DAT I O N • 1273
Negri, G., E. W. Teixeira, M. L. T. M. F. Alves, A. C. Moreti, I. P. Otsuk, R. G.
Borguini, and A. Salatino. 2011. Hydroxycinnamic acid amide derivatives,
phenolic compounds and antioxidant activities of extracts of pollen sam-
ples from Southeast Brazil. Journal of Agricultural and Food Chemistry 59:
5516–5522.
Qin, Y., A. R. Leydon, A. Manziello, R. Pandey, D. Mount, S. Denic, B. Vasic,
et al. 2009. Penetration of the stigma and style elicits a novel transcrip-
tome in pollen tubes, pointing to genes critical for growth in a pistil. PLOS
Genetics 5: e1000621.
Quilichini, T. D., E. Grienenberger, and C. J. Douglas. 2015. The biosynthesis,
composition and assembly of the outer pollen wall: A tough case to crack.
Phytochemistry 113: 170–182.
R Core Team. 2013. R: A language and environment for statistical computing.
R Foundation for Statistical Computing: Vienna, Austria. URL http://www.
R-project.org/
Rehman, S., E. S. Rha, M. Ashraf, K. J. Lee, S. J. Yun, Y. G. Kwak, N. H. Yoo, and
J. K. Kim. 2004. Does barley (Hordeum vulgare L.) pollen swell in fractions
of a second? Plant Science 167: 137–142.
Rowley, J. R., and J. S. Rowley. 1983. Modification of the exine during germi-
nation of pollen grains of Ulmus. In D. L. Mulcahy and E. Ottaviano [eds.],
Pollen: Biology and implications for plant breeding. 173–182. Elsevier,
Amsterdam, Netherlands.
Scott, R. J. 1994. Pollen exine: The sporopollenin enigma and the physics of pat-
tern. In R. J. Scott and A. D. Stead [eds.], Molecular and cellular aspects of plant
reproduction, 49–80. Cambridge University Press, New York, New York, USA.
Seymour, R. S., and A. J. Blaylock. 2000. Stigma peroxidase activity in associa-
tion with thermogenesis in Nelumbo nucifera. Aquatic Botany 67: 155–159.
Sharma, B., and S. C. Bhatla. 2013. Accumulation and scavenging of reactive
oxygen species and nitric oxide correlate with stigma maturation and pollen–
stigma interaction in sunflower. Acta Physiologiae Plantarum 35: 2777–2787.
Smirnova, A. V., N. P. Matveyeva, and I. P. Yermakov. 2014. Reactive oxy-
gen species are involved in regulation of pollen wall cytomechanics. Plant
Biology 16: 252–257.
Speranza, A., R. Crinelli, V. Scoccianti, and A. Geitmann. 2012. Reactive oxy-
gen species are involved in pollen tube initiation in kiwifruit. Plant Biology
14: 64–76.
Southworth, D. 1974. Solubility of pollen exines. American Journal of Botany
61: 36–44.
Swanson, R., T. Clark, and D. Preuss. 2005. Expression profiling of Arabidopsis
stigma tissue identifies stigma-specific genes. Sexual Plant Reproduction 18:
163–171.
Taylor, M. L., and J. M. Osborn. 2006. Pollen ontogeny in Brasenia
(Cabombaceae, Nymphaeales). American Journal of Botany 93: 344–356.
Vieira, A. M., and J. A. Feijó. 2016. Hydrogel control of water uptake by pectins
during in vitro pollen hydration of Eucalyptus globulus. American Journal of
Botany 103: 437–451.
Weston, R. J. 2000. The contribution of catalase and other natural prod-
ucts to the antibacterial activity of honey: A review. Food Chemistry 71:
235–239.
Wille, J. J., and M. A. Berhow. 2011. Bioactives derived from ripe corn tassels:
A possible new natural skin whitener, 4-hydroxy-1-oxindole-3-acetic acid.
Current Bioactive Compounds 7: 126–134.
Wolters-Arts, M., W. M. Lush, and C. Mariani. 1998. Lipids are required for
directional pollen-tube growth. Nature 392: 818–821.
Zechmann, B., B. E. Koffler, and S. D. Russell. 2011. Glutathione synthesis is
essential for pollen germination in vitro. BMC Plant Biology 11: 54.
Zhang, D., J. Shi, and X. Yang. 2016. Role of lipid metabolism in plant pollen
exine development. Sub-Cellular Biochemistry 86: 315–337.
Zinkl, G. M., B. I. Zwiebel, D. G. Grier, and D. Preuss. 1999. Pollen–stigma ad-
hesion in Arabidopsis: A species-specific interaction mediated by lipophilic
molecules in the pollen exine. Development 126: 5431–5440.