Table Of ContentMISSOURI
BOT«Ni-«i.
DEC
1^99
2 9
Fern
Journal 89(4):221-231
(1999)
1
Comparison
Interpopulational Dose-Mediated
of
Antheridiogen Response Onoclea
in sensibilis
—
Abstract.
Intraspecific variation in antheridiogen response has been reported in several fern
we
species. Here, characterized and compared dose-mediated antheridiogen response among
six
populations spanning the range of Onoclea sensibilis, a species widely used for antheridiogen
A
A
assays. crude aqueous filtrate (CAF) of antheridiogen was obtained from cultures of Pteri-
PT
dium aquilinum and introduced into mineral agar at six concentrations ranging from (control)
10"
to 1 Response was initiall] aaracterized for one population using three criteria: percentage
.
i
of male gametophytes, number of antheridia per gametophyte, and mean surface area of gameto-
phytes. Percent male was consistently the best descriptor of the response profile and was used to
compare among
responses populations. In all populations, gametophytes exhibited or no
little
CAF
response concentrations 10 s and lower; a saturated response, almost gametophytes
at
all
being male at 10 3 and higher; and an intermediate response, usually between 50% and 80%
males, at 10 4 With few exceptions, response levels and profiles were very consistent among
.
populations. Data and methodology provided herein indicate that use of the percent male criterion
can provide a reliable quantitative assay using Onoclea that can facilitate tuture comparative
research antheridiogens.
in
Gametophytes of numerous and diverse taxa within the Polypodiaceae sensu
A
lato share response to antheridiogen (sensu Voeller 1964; in this paper an-
A
theridiogen refers to the general class of antheridiogens produced by and
biologically active in taxa belonging to Polypodiaceae sensu whereas an-
lato,
A A
Naf
theridogen sensu [19791 refers specifically to antheridiogen pro-
PT
, ,
duced by Pteridium aquilinum Kuhn). However, antheridiogen response
(L.)
among
neither pervasive nor equivalent polypodiaceous ferns. Species rep-
is
may minimum
by magnitude
resenting different genera differ orders of for the
A
and
dose required response (Naf 1975; Naf, 1979), congeneric
for et
al.,
PT
may shown Bommeria
species also differ in antheridiogen response, as in
and and and
(Haufler Gastony, 1978), Cystopteris (Haufler Ranker, 1985), po-
may
lystichoid ferns (Yatskievych, 1993). Variation in response also occur
among among
within populations or even individuals within
species, either
Hemionitis palmata lower response was detected pop-
populations. In in
L.,
ulations with lower genetic load (Ranker, 1987), suggesting that reduced re-
may
sponse to antheridiogen facilitate a higher selfing rate, thereby eliminating
A
recessive lethals within populations (Schneller et 1990). genetic basis
al.,
was
determined
for variation in antheridiogen sensitivity in Ceratopteris. Ge-
notypes with lowered antheridiogen response were obtained experimentally
Brongn. through application of mutagens, and a simple two-
in richardii
C.
was
gene basis of inheritance for antheridiogen sensitivity determined (Scott
Warne
and
Hickock, 1987; et 1988).
al.,
combine
These observations to indicate that antheridiogen response sub-
is
AMERICAN FERN JOURNAL: VOLUME NUMBER
222 89
4 (1999)
may
evolutionary change. Because antheridiogen be importance
ject to of in
mode
determining or facilitating the of sexual reproduction by gametophytes,
may
evolutionary changes in antheridiogen response represent variation in an
important
history
life attribute.
Of compared by Naf
the species (1979), the greatest sensitivity to antheri-
A
diogen was exhibited by Onoclea which was used
sensibilis therefore
L.,
PT
as a reference for interspecific comparisons and has become the standard fern
species for assaying antheridiogen A. This species broadly distributed, oc-
is
curring over most of eastern North America and disjunct in eastern Asia (Gas-
tony and Ungerer, 1997). Although geographic sources of O. sensibilis have
among
varied studies, the possibility of geographically based variation in an-
theridiogen response has not specifically been addressed. Most workers
utiliz-
and
ing O. sensibilis Haulier Ranker, 1985; Miller, 1968) have reported
(e.g.,
highly sensitive antheridiogen response and lack of spontaneous without
(i.e.,
exposure to antheridiogen) antheridium formation, comparable
to that indi-
cated by Naf (1979). However, unusual responses have been reported from
occasional frond collections example spontaneous antheridium
(Naf, 1979), for
formation (Klekowski and Lloyd, 1968; Miller and Miller, 1970; Rubin and
way
which
Paolillo, 1983). Furthermore, the in response levels in relationship
to antheridiogen doses have been quantified has varied among and
studies
sometimes
has not been fully explicit. The objectives of this study were to
determine
consistent criteria for quantifying antheridiogen response of O. sen-
and
determine whether among
sibilis to there are differences in response pop-
from
ulations
different regions.
Materials and Methods
The
antheridiogen exudate used in this study was obtained from gameto-
A
phytes of Pteridium aquilinum, a standard source of antheridiogen due
to
the high activity of exudate. Spores from aquilinum accession
its P. Sheffield
56 from England were sown under
axenic conditions as follow. Spores were
from by
separated sporangia rinsing through a Nitex screen with 100 u.m open-
ings, using a wetting solution drops of Tween-80 per 100 ml
(2 distilled water).
Spores were made
collected in a cylindrical "basket" from an inverted
plastic
mm
The
30 centrifuge tube. conical bottom of the tube was cut and the
off,
A
was
cap perforated with numerous holes using a hot dissecting needle.
fine-
meshed
Nitex screen u.m openings) was placed between the cap and the
(5
centrifuge tube opening to form a "basket" bottom that would retain spores
but allow passage of
liquids.
Spore samples, wet with Tween were
the solution, rinsed three times with
water and
distilled set in the dark overnight to allow germination of fungal
The
spores. following day, the baskets were placed in a sterilizing solution of
5%
commercial
hypochlorite 90 then placed
for sec, in three rinses of
sterile
distilled water. Spores were then sown by pipetting onto mineral agar in den-
cm
ranging from 5-30
sities spores per 2 in plastic petri dishes.
Mineral agar contained Parker's macronutrients and Thompson's micronu-
& WERTH: ANTHERIDIOGEN RESPONSE ONOCLEA
IN
5
Collection localitu [his study. Collectors a
Population Location
GA
Gray, Georgia: Jones Count}. 129 South of Gray (C. R. Werth)
S.
1
W&OD
VA
Vienna, Vnginia Fnrfax County, along near intersection with Gallows
trail
Tennessee Davidson County, plant under cultivation, collected near Nash-
(D. Whittier).
ville
P.
PA TR
Berwick, Pennsylvania: Luzerne County, Beach Haven, Salem township 438
(J.
Montgomery)
D.
VT
Richmond, Vermont: Chittenden County, 3 miles northeast of Richmond (C. A. Paris
and D. Barrington)
WI
Eagle, Wisconsin: Waukesha County, Section 29, County Highway mile west-
5,
1
(W
southwest of Eagle C. Taylor)
trients, as described in Klekowski (1969). Media were autoclaved and poured
into petri dishes. Petri dishes inoculated with spores were placed under con-
tinuous light of 45 micromole quanta provided by 40 watt cool-white fluores-
cent bulbs. The water in which the spores were introduced evaporated (or was
absorbed by the agar) after one to two days. Petri dishes were then placed in
sandwich minimize
sealable bags to further evaporation.
grown
Gametophytes aquilinum were approximately one month.
of for
P.
The agar was then frozen and thawed, and the aqueous phase was separated
from the agar matrix by filtration, yielding a crude aqueous filtrate (CAF). The
A
CAF summer
present study used 89-1 produced during the of 1989, which
PT
was stored frozen in 100 ml plastic bottles until use in the present series of
experiments conducted over the years 1990-1992. Single bottles of antheri-
diogen extract were used for several experiments and these bottles were kept
thawing and used successive experiments. Antheridiogen
in
refrigerated after
was apparently stable over the duration of the experiments under these
activity
was
conditions, as no appreciable decrease in response observed (see results).
Sporophylls of O. sensibilis were obtained between January and February
before spores were released) of 1990 and 1991 from various locations
(i.e.,
by
Spores were harvested from loose chaff or inducing sporangial
(Table
1).
was found dehiscence could be induced by thoroughly
dehiscence. that
It
them on smooth and
soaking sporophylls in water, placing paper, allowing
them dry overnight. This procedure resulted in release of large quantities
to
of spores with minimal amounts of sporangia (note: sporophylls collected in
the did not release spores after soaking, implicating a role for vernalization
fall
and wetting timely spore release in nature.) Spore collections represented
in
from 5-20 separate sporophylls each population
pooled spores released for
which was
TN, from sporophyll
except the Nashville, site, a single obtained.
Spores were and sown using the same technique as described above
sterilized
aquilinum.
for
P.
Antheridiogen enriched media were prepared by adding various amounts
-1 -10~ CAF Unenriched media provided
(10 6 of prior to autoclaving. a control
)
224 AMERICAN FERN VOLUME NUMBER
JOURNAL:
89
4 (1999)
for antheridiogen dose experiments. Cultures were scored 21 days sow-
at after
Although
ing. axenic conditions were sought, fungal contamination was
ex-
A
perienced number
in a sizable of replicates at all concentrations, appar-
PT
ently resulting from use age-weakened
of hypochlorite To
solution. evaluate
we
the influence of fungal contamination, performed an
analysis of covariance
to determine differences between contaminated and uncontaminated
repli-
contamination -)
cates, state (+, being the covariate. There were no
significant
male—
differences in response (percent see below) between contaminated and
uncontaminated CAF =
same
replicates of the concentration
(P 0.248). There-
contaminated were
fore, dishes included
in the data except a very few
set, for
where
had overgrown
fungi
the gametophytes.
For
experiments, gametophytes
initial thirty per treatment were harvested
in an unbiased fashion, mounted on microscope
slides in a solution of part
1
medium
Hoyers mounting
and
1 part acetocarmine, and examined one day
under low power compound
later using
a microscope. For each gametophyte,
length and width were measured
using an ocular micrometer, and number
the
of each kind gametangium
of antheridia or archegonia) was
(i.e., recorded.
Later experiments were
evaluated only percentage male
for of gametophytes
was
per found
culture.
It that the addition of acetocarmine
directly to cultures
would
both fix the gametophytes and stain gametangia, and gametophytes
that
could be scored
reliably for sex expression under a dissecting microscope
without removing them
from
the
petri dish. In these latter experiments, 50
gametophytes
per treatment were
scored
for sexual expression. All
experi-
ments were
repeated
at least once.
Results
Characterization
of Response to Antheridiogen.—
Consistent with most
previous
observations
(Naf, 1979), O. sensibilis gametophytes grown on media
CAF
lacking
(control cultures) failed to produce antheridia, with
rare excep-
tions (a single male individual was observed
in a control culture of the Gray,
GA,
population). Although
not individuals matured
all sexually during
the
week
three experimental
period, nearly gametophytes
all in control
cultures
allowed grow
to greater lengths of time ultimately became
female,
pro-
i.e.,
duced
archegonia
in the region of their meristematic Male
notch. gameto-
phytes, possessing away
antheridia scattered from the meristematic
notch,
CAF
consistently appeared
in cultures with 10 4 and Also
higher. consis
with
previous
reports (Dopp, 1950; Naf, morphology
1979), differences in
v
observed between male and
female
O. sensibilis gametophytes. Females
\
invariably cordate and than which
larger males, tended develop numei
to
A
marginal and
lobes often were
ameristic at higher concentrations (10 and
!
PT
10"
2 CAF). Frequently, in cultures that possessed 100% males
initially (10 3
and
CAF,
10 2 see below) but were allowed
continue growing
to past the sam-
number
pling date, a small gametophytes became
of and
large cordate
in the
week
fourth or later after sowing. Microscopic examination
revealed that these
were
hermaphrodites had
that apparently switched from being male and had
STEVENS WERTH: ANTHERIDIOGEN
&
GA
Gray
gametophytes from GA,
the Gray, varied doses
rilis site to
(i.e.,
A
by mean number
>en as characterized of antheridia per
PT
,
x
gametophyte surface area, estimated as length width
(right).
developed archegonia in the vicinity of their meristematic notches. The ten-
CAF
dency for these hermaphrodites to appear seemed to be greater in 10 3
10~
cultures than in 2 CAF. Similar observations were reported by Naf
(1979).
To determine the suitability of various criteria that could characterize and
quantify antheridiogen response, initial experiments using a single accession
GA)
(Gray, evaluated three observable attributes of response by O. sensibilis
A
^-10
10
gametophytes over concentrations of 5 CAF. These attributes were:
PT
mean number mean
of antheridia per gametophyte, surface area per
1) 2) ga-
metophyte, and percentage of male gametophytes.
3)
mean number
Antheridiogen response characterized as of antheridia per
ga-
Number
metophyte illustrated in Fig. 1A. of antheridia per gametophyte was
is
Maximum
very both the control and 10 5 CAF. response was
close to in at
CAF
Two
observed concentrations of 10 3 or higher. of the three replicates
at
CAF ^-10
showed
very similar response levels across concentrations 10 \ sug-
mean number
gesting that at this concentration of antheridia exhibits a satu-
showed
However,
rated response. the replicate a substantially higher
first re-
showed
sponse 10 2 than either 10 3 or 10 \ All three replicates an
at at in-
mean number
termediate response 10 4 CAF. In replicate the of antheridia
at
3,
per gametophyte was substantially lower than in the two This
first trials. is
believed be due differences in the spore sowing density.
to to
The second gametophyte was evaluated because known
criterion, size,
it is
and num-
to be affected by antheridiogen extracts (Naf, 1979) to influence the
gametophyte, demonstrated and
ber of antheridia per as in Cystopteris (Haufler
Ranker, 1985). To determine the most effective means to estimate gametophyte
surface gametophyte outlines were traced onto paper using a camera-
area,
lucida. These were cut out, their length and width measured, and their surface
Area was
area determined using a leaf area meter. regressed against the follow-
Squan
Of Square of
5
St Length Width X
width Length width
t
2 0.78 0.70 064
r 0.7 0.93
65.55 43.56
68.4 34.01
T
<0.001
<0.001
<0.001
001
N
20 lo
20
ing independent
variables: length, width, square
1) 2) 3) of length, 4) square
of width, and length x width. The product of two dimensions X
5) (length
width) provided
a substantially better approximation of area than any
of the
other variables (Table and product was
this therefore used. Variation
2), in
gametophyte
surface area (as estimated by the product of length and width)
in
CAF
response
to concentration is illustrated in Fig. IB. The area response was
CAF
not as consistently predicted by concentration was mean number
as either
of antheridia or percent male (see next paragraph), and furthermore was
in-
consistent across experiments. However, tendency
a for area to decrease with
CAF
increased was
concentration
clearly indicated, consistent with previous
reports
(Naf, 1979).
Antheridiogen
response
characterized as percent male
is illustrated for the
GA, CAF
Gray, population
Response
in A.
Fig. 2 concentration 10 was
at 5
at
or near zero and indistinguishable from CAF
the Response
control. concen-
at
trations 10 \ 10 and 10 3 was at or near 100% males (96.5-100
2, %), indi-
cating a saturated response CAF
for this criterion. At 10 4 an intermediate
response was
observed
(68.9-79.3%
males), indicating
that the threshold
for
substantial response lies between 10 5 and 10 4 CAF. This concentration seems
highly comparable Naf
to s (1956) determination of 3.2 X 10 5 as a response
A
threshold concentration
for exudate. Values for percent male each con-
at
PT
were
centration highly much
consistent across the three more
replicates, so
than number
for either of antheridia
or area.
The
relationship between gametophyte
surface area and antheridium num-
was
ber explored. To determine whether number
the
of antheridia per game-
tophyte increased with
area, regression analysis was performed
within each
Number
treatment
(Table was
of antheridia
3). positively associated with
ga-
metophyte
surface area
for all three replicates at 10 CAF, for 10 2 two
" all at
,
of the three at 10 and two at 10 4 In other treatments, was no
all there
'
.
significant relationship between and
area antheridial number. Thus,
significant
regressions tended to be observed in cultures where most
or gametophytes
all
possessed
antheridia, indicating that male gametophytes, number
for
of an-
theridia increased with
size of the gametophyte. However,
the strength of
this
was
association not
great, the highest value of 2 being Nor was
r 0.722. the
etophytes from six populat
STEVENS & WERTH: ANTHERIDIOGEN RESPONSE ONOCLEA
IN
GrayGA
Eagle V
AMERICAN FERN VOLUME NUMBER
JOURNAL:
89
4
Table
Relationships betwe
3.
phytes for each of the three repl
degrees of freedom, and
F
>
gametophyte
unit of even approximately
i
;een by comparison between 1A and
Figs.
Clearly, percent male gametophytes
not only provided more
a consistent
response
criterion, but also could be more rapidly scored than number
of an-
theridia. Therefore, subsequent experiments were
evaluated using only the
percent male
criterion.
Comparison
of Populations.—
Five
additional populations were compared
to
each and
other GA,
to the Gray, population
antheridiogen
for response using
the percent male
To
criterion (Fig. 2). allow for the possibility that some pop-
ulations might A
exhibit greater sensitivity, the range of concentration was
PT
CAF
extended
to include 10 The was
<> level of response similar in popu-
all
Response
approached
lations. saturation 10 3 CAF, with means
at across rep-
licates for populations varying from 96.6-100% male
(the lowest value an
for
individual replicate was 85%). most
In populations, response 10 5 and 10 «
at
CAF
was
similar to that of the control, few to no males. The exception
i.e.,
was
the Eagle, WI, population, which numbers
in substantial males were
of
observed some
in replicates at both 10 5 (21% males) and 10"8 (16%
males);
control cultures for this population contained no males.
All cultures exhibited
intermediate Mean
levels of response 10 4 CAF. response
at values dose
at this
among
differed substantially populations, ranging from 18% TN)
(Nashville,
78%
to (Gray, GA).
Relative to the amount of variation among populations, was
there a substan-
STEVENS & WERTH: ANTHERIDIOGEN RESPONSE ONOCLEA
IN 229
amount
tial of variation in response at 10~ 4 between replicates for some pop-
among
ulations. Variation in response replicates was greatest in the Richmond,
VT, population (26-84% males at 10 4 CAF) and least in the Gray, GA, pop-
ulation (68.9-79.3% males 10 4 CAF).
at
Discussion
we
Herein have components by
quantified three of response O. sensibilis to
A
varied concentration of antheridiogen Percent males and mean number of
PT
.
antheridia per gametophyte increased and gametophyte surface area decreased
A
with increasing doses of Of these, percent male was the most predictable,
PT
.
and provides a consistent and convenient means characterize antheridiogen
to
A
response. Using this criterion, threshold and saturation doses of concen-
PT
were found by approximately two magnitude
tration to differ orders of
(ca.
10 5 and 1G- 3 CAF, respectively), with the intermediate concentration (10 4
CAF)
an intermediate response. This response by and
eliciting profile large
is
among
consistent populations sampled across the species range. However,
some among
there is evidence of at least variation populations at the inter-
mediate dose 4 CAF). Notably, the Nashville, TN, population exhibited
(10 a
CAF
18%
(mean
substantially lower response of males) at 10* 4 than other
all
mean 50%
which
populations, exhibited response levels in excess of males
at
10~ 4 CAF. There were also some populations in which males appeared con-
at
WI
below 10" which
centrations 4 notably the Eagle, population, exhibited
,
numbers males 10 5 and 10 6 CAF.
substantial of
at
among
The
fundamentally similar results populations could reflect a lack of
determining antheridiogen-response phenotype
genetic variation in O. sensi-
may
Alternatively, substantial genetic variation for response exist, but be
bilis.
evenly distributed among populations such that population means appear
among
equivalent. Phenotypic differences genetically different individuals
would be small antheridiogen response variation polygenically
deter-
if is
mined. uncertain whether the differences in response by individual ga
It is
CAF
among
metophytes exhibited 10 4 are a result of genetic differences
at
among
gametophytes or instead represent stochastic response variations
A^
metophytes that are genetically uniform with respect to determinants of
CAF
response. The lower response value at 10 4 in the Nashville, TN, popu
some gametophytes
lation and the occasional response of (e.g., in the Eagle
WI, population) extremely low doses (10 5 and 10~ 6 CAF) suggests that a
at
may
some However,
genetic variation exist. the heritability of antheridi
least
unknown
any
ogen response completely in O. sensibilis or in other poly
is
mutation
podiaceous (sensu lato) fern. In Ceratopteris richardii (Parkeriaceae),
was shown
have simple
induced lack of antheridiogen sensitivity to a (two
gene) mode of inheritance (Scott and Hickok, 1987; Warne et al., 1988). How-
more
remains possible that this species includes subtly differentiated,
ever,
it
continuously varying phenotypes that are determined polygenically.
As most on antheridiogen response, the present experiments
with studies
were under highly conditions that constrain their
carried out artificial rele-
AMERICAN VOLUME NUMBER
FERN
230 JOURNAL:
89 4
(1999)
vance
to the natural life history characteristics of O. sensibilis. Unnatural con-
ditions included use of agar solidifed media instead of heat
soil, sterilization
media and
of antheridiogen via autoclaving temperatures exceeding
at greatly
those encountered continuous and
in nature, illumination, use
of antheridi-
ogen derived from bracken rather than from Onoclea Use
of definable
itself.
make
artificial conditions such experiments more and
feasible, repeatable,
comparable to the broad literature which has used similar conditions. The
and
validity of this other similar experiments and relevance nature could
to
be evaluated by repeating them under more
natural conditions, has been
as
done
but and
rarely (Haufler Ranker,
1985).
whether
Irrespective of the present
results
reflect life-history attributes of O.
they do have on
sensibilis, bearing the use of this species as a standard assay
organism
detecting antheridogen The
for A. similar response observed
profile
across a large portion of the species range further validates the and
reliability
consistency
of O. sensibilis for this purpose. The means by which
antheridi-
ogen
response among
is quantified has varied studies, having included
deter-
minimum
mining
the dose can produce any
that response Naf
Naf, 1956;
(e.g.,
et al., 1975), percentage of males in cultures (Naf, 1965; Klekowski and Lloyd,
1968; Schedlbauer and Klekowski, Rubin and
1972; and
Paolillo, 1983; Scott
Hickok, 1987; Nester-Hudson number
et 1997), or of antheridia per
al., ga-
metophyte
and
(Haufler Ranker, 1985). The response variation among popu-
lations the lower
at concentrations may
indicates
that inconsistencies be
ex-
perienced using the minimal response
same
criterion, unless the spore source
We
Onoclea
of is always used. found that the most easily scored response
attribute, percent male, most A
also the
is reliable for judging concentration.
PT
A
The
strength
of extracts could be standardized by determining
PT the dilution
which 50%
at of the gametophytes
are male, analogous LD50
the
to criterion
used
in toxicology. Moreover,
the discovery
of individuals such one
as the
from
TN,
Nashville, with lower
sensitivity could provide an
additional assay
more
point
for precisely standardizing
antheridogen Such
concentrations.
standardization have
will importance
as research on antheridiogen response
CAF
continues, as different may
extractions vary
in their /
and
diminish
over
time.
We
are grateful to Paul Wolf providing
for bracken
spores, Ralph
to Brooks, Carol Kuhn,
Chris
Haulier, John
Miller, Carl Taylor, James Montgomery, Dean
Whittier, David Barrington, and Cathy
Paris providing
for collections of Onoclea sporophylls, and
to Lisa Wellborn for help in developing
methodology. Helpful comments
by an anonymous
reviewer
resulted
in
i
REU
supported by a supplement NSF
to grant
1
LITERATURE
ClTED
Ein die Antheridienbildung
Farnen
bei fordernde Substanz
.
aquilinum Kuhn.
(L.) Ber. Deutsch.
r? Bot. Ges. 63:139-147.
