Citation
Hybridization and species boundaries in carex sections ceratocysis and bicolores (cyperaceae)

Material Information

Title:
Hybridization and species boundaries in carex sections ceratocysis and bicolores (cyperaceae)
Creator:
Weil, Sarah Jane
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
ix, 56 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Carex -- Hybridization ( lcsh )
Carex -- Speciation ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 53-56).
Thesis:
Biology
Statement of Responsibility:
by Sarah Jane Weil.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
227988011 ( OCLC )
ocn227988011
Classification:
LD1193.L45 2007m W44 ( lcc )

Downloads

This item has the following downloads:


Full Text
HYBRIDIZATION AND SPECIES BOUNDARIES IN CAREX SECTIONS
CERA TOCYSIS AND BICOLORES (CYPERACEAE)
by
Sarah Jane Weil
B.S., University of Puget Sound, 2004
A thesis submitted to the
University of Colorado at Denver and Health Sciences Center
in partial fulfillment
of the requirements for the degree of
Master of Science
Biology
2007


This thesis for the Masters of Science
degree by
Sarah Jane Weil
has been approved
by
Leo P. Bruederle
Charles A. Ferguson
Date


Weil, Sarah J. (Masters of Science, Biology)
Hybridization and Species Boundaries in Carex
Sections Ceratocystis and Bicolores (Cyperaceae)
Thesis directed by Associate Professor Leo P. Bruederle
ABSTRACT
Speciation, the hallmark event of evolution, has been more frequent in certain
genera than in others. Notable among these is the genus Carex, which includes
approximately 2000 species of sedges worldwide. Many species of Carex hybridize
frequently and, occasionally fertile hybrids mediate gene exchange between otherwise
well-defined species, a process referred to as introgression. Hybridization and
introgression have important evolutionary consequences, including increased genetic
diversity within a species and/or speciation. As such, accurate identification of
hybrids and introgressants can inform hypotheses regarding evolution in Carex.
Historically, Carex species were identified and delimited morphologically, in
concert with other data. However, because of a the highly reduced floral morphology
in this genus and the potential for phenotypic plasticity to obscure morphological
traits, accurate identification of species and/or hybrids generally requires additional
lines of evidence. Previous investigators have used geography, allozymes, and
cytology, among other data to differentiate species and identify hybrids. While these


sources of data are valuable, they have limitations and may be inconclusive, time
consuming, or otherwise expensive.
Herein, three populations of Carex, including species from sections
Ceratocystis and Bicolores, were used to test a novel method, based on cleaved
amplified polymorphisms (CAPs) of ribosomal DNA (rDNA) spacer regions, to
distinguish between species and their hybrids in Carex. It is demonstrated that CAPs
of rDNA can differentiate species and hybrids in Carex section Ceratocystis but not
in section Bicolores. Furthermore, in Carex section Ceratocystis CAPs of a region of
chloroplast DNA can determine maternity for hybrid individuals. The use of CAPs to
screen two mixed populations of Carex section Ceratocystis, indicated a lack of
introgression, which is consistent with the observation that hybrids in these
populations were completely sterile. As such, this research could not confirm the
ability of CAPs to positively identify introgressants. A discussion of the possible
limitations of CAPs and other useful molecular techniques for detecting hybrids and
introgressants is included.
The abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signed
Leo P. Bruederle


ACKNOWLEDGEMENT
I would like to thank Leo Bruederle for the opportunity to conduct this research and
his continued guidance and encouragement throughout. My thanks also to Michele
Engel for help with restriction digest protocols and to Lisa Johansen for frequent use
of her lab space. Finally, my acknowledgments to Nathan Derieg for providing
sequence data and continuous advice and support.


CONTENTS
Figures......................................................................viii
Tables.........................................................................ix
Chapter
1. Introduction.................................................................1
1.1 Species Boundaries and Hybridization.......................................1
1.2 Characterization of Car ex Hybrids.........................................3
1.3 Detecting Introgression....................................................6
1.4 Ribosomal DNA as a Genetic Marker in Car ex...............................8
1.5 Cleaved Amplified Polymorphisms...........................................11
1.6 Model System: Carex Sections Ceratocystis and Bicolores...................11
2. Hypotheses.................................................................13
3. Materials and Methods......................................................14
3.1 Sampling..................................................................14
3.2 DNA Extraction............................................................15
3.3 Amplification of ITS, ETS, and Chloroplast Sequences......................16
3.4 Sequencing and Restriction Digests........................................17
4. Results....................................................................20
4.1 Carex cryptolepis x C. viridistellata.....................................20
4.2 Carex cryptolepis x C. viridula...........................................24
vi


[
4.3 Carex bicolor x C. aurea..................................................30
5. Discussion..................................................................31
5.1 Carex cryptolepis x C. viridistellata.....................................31
5.2 Carex cryptolepis x C. viridula...........................................33
5.3 Carex bicolor x C. aurea..................................................36
5.4 Detecting Introgression with Cleaved Amplified
Polymorphisms of Ribosomal DNA.............................................37
6. Future Research............................................................40
Appendix
A. Genomic DNA Extraction Solutions...........................................43
B. Loading Dye and Sodium Borate Gel Buffer...................................45
C. Genomic DNA Extraction Protocol............................................46
D. PCR Protocol...............................................................48
E. Restriction Digest Protocol................................................49
F. Pouring and Running an Agarose Gel.........................................50
References.....................................................................53
Vll


LIST OF FIGURES
Figure
1. Angiosperm rDNA Gene Region..................................................10
2. CAPs of ETS and ITS Sequences for Representative
Springville Marsh Species and Hybrids........................................21
3. Chloroplast CAPs of Springville Marsh Species and
Hybrids......................................................................22
4. Representative Springville Marsh Individuals Screened
for Introgression............................................................23
5. Representative Amplifications of ITS and ETS Regions
from the Fish Fry Lakes Population...........................................24
6. CAPs of ETS and ITS Sequences for Representative
Fish Fry Lakes Species and Hybrids...........................................26
7. Representative ETS CAPs from Fish Fry Lakes
Individuals..................................................................28
8. Representative ITS CAPs from Fish Fry Lakes
Individuals..................................................................29
9. ITS CAPs for Two Species in Carex Section
Bicolores....................................................................30


LIST OF TABLES
Table
1. Enzymes Used to Generate Cleaved Amplified Polymorphisms
for Carex Species in Three Populations................................19


1. Introduction
1.1 Species Boundaries and Hybridization
What defines and delimits species? A species is commonly defined as a group
of individuals that freely exchange genetic material with each other, but not with
members of other such groups. Thus, members of the same species can reproduce
and their offspring are fertile, while members of different species are genetically
isolated. However, many naturally occurring events confound this definition.
In microorganisms, gene transfer between members of different species can
occur horizontally (between members of the same generation) by plasmid transfer or
virally mediated DNA transfer. Asexual species, which theoretically lack any
exchange of genetic material, seem to defy the aforementioned definition completely
(Strickberger, 2000b). Of particular interest here are situations that commonly arise
in plants. For example, individuals that interbreed and produce fertile offspring in an
experimental setting may not do so in their natural environment because of
geographical (Strickberger, 2000b) or phenological (Cayouette, 1992) barriers to
mating. In these cases pre-zygotic barriers tentatively maintain reproductive
isolation. Alternatively, members of other species may hybridize freely. If hybrids
are sterile, genetic isolation is maintained post-zygotically (Strickberger, 2000a).
However, partially fertile hybrids may preclude complete genetic isolation between
species. The resulting complete or partial amalgamation of gene pools from different
1


species has important evolutionary consequences, several of which have been
documented in Carex.
One possible consequence of hybridization is speciation. If fertile hybrid
individuals become established through selection, they may become reproductively
isolated from the parental species. Within Carex section Cryptocarpae, for example,
C. salina Wahl, and C. recta Boott are thought to be of hybrid origin (Cayouette,
1985). Evidence for the hybrid origin of these species includes morphological
intermediacy with respect to the putative parental species and partial sterility due to
meiotic irregularities. Though C. salina and C. recta behave as stable species, the
persistence of meiotic irregularities suggests that speciation is recent or in progress
and it is possible that reproductive isolation from the parental species may not be
complete. On the other hand, hybridization can lead to speciation with immediate
reproductive isolation. Allopolyploid speciation, which may explain the origin of C.
pediformis C.A. Mey. (Carex section Digitatae) (Tyler, 2003), occurs when a hybrid
undergoes genomic duplication. The polyploid hybrid has twice as many
chromosomes as either parental species. As such, meiosis in backcrossed individuals
will not occur due to the lack of homology between chromosomes of the hybrid and
either parental species (Strickberger, 2000a).
If partially fertile hybrids are not reproductively isolated from the parental
species, they may backcross to one species or the other. Repeated backcrosses
following a hybridization event can result in introgression, whereby genes are
2


transferred between two otherwise well-defined species. The evolutionary
importance of introgression is that it can facilitate increased genetic variability within
a species. This has been reported in subspecies curvula and rosae of Carex curvula
All. (Choler, 2004). Populations of each subspecies were studied in both optimal and
marginal habitats. Though these subspecies can interbreed, they normally do not. For
both subspecies, the frequency of genetic markers of the other subspecies was low in
populations occupying optimal habitats and significantly higher in marginal
populations. In the case of the C. curvula subspecies, it seems that hybridization and
gene transfer between subspecies aids in adapting marginal populations to non-
optimal environments.
1.2 Characterization of Carex Hybrids
Given the potentially important evolutionary role of hybridization in Carex,
we must consider the frequency of this phenomenon in the genus. Though
hybridization is common in vascular plants, it may be frequent in some families while
nearly absent from others. A survey of five major biosystematic floras revealed that
hybrids are observed most often in perennial, outcrossing species that are capable of
vegetative reproduction, though these characteristics are clearly not required for
hybridization to occur (Ellstrand, 1996). This survey also found that members of
Carex were among the most commonly hybridizing species in Cyperaceae. Carex
hybrids were first reported in the mid-1800s and by 1992 a total of 253 had been
reported in North America, with hybrids particularly common in sections
3


Ceratocystis, Phacocystis, and Vesicariae (Kukkonen, 1988; Cayouette, 1992).
However, identification of Carex hybrids and putative parental species has proved to
be challenging given the reduced floral morphology characteristic of this genus and
the potential for phenotypic plasticity to distort morphological characteristics. As
such, many of the reported hybrids, especially those reported early in the 20th century
when scientific techniques were limited, require additional study to verify their hybrid
status.
Identification of hybrids requires unambiguous differentiation of species.
From the 1970s to the 1990s, several investigators considered species limits and/or
hybridization in Carex. This involved analyses of macromorphology (Cayouette,
1985; Cayouette, 1987; Catling, 1989; Waterway, 1990; Cayouette, 1992; Ford, 1992;
McClintock, 1994; Waterway, 1994), micromorphology (Cayouette, 1987; Catling,
1989), geographical distribution (Cayouette, 1985; Cayouette, 1987; Catling, 1989;
Waterway, 1990; Cayouette, 1992; Ford, 1992; McClintock, 1994; Waterway, 1994;
Choler, 2004), allozyme data (Waterway, 1990; Ford, 1992; McClintock, 1994;
Waterway, 1994), cytology (Cayouette, 1985; Cayouette, 1987; Whitkus, 1988;
Waterway, 1990; Cayouette, 1992; Luceno, 1993; McClintock, 1994; Waterway,
1994), sterility (Cayouette, 1985; Cayouette, 1987; Catling, 1989; Waterway, 1990;
Cayouette, 1992; Ford, 1992; Waterway, 1994; Choler, 2004), phenology (Cayouette,
1992), and flavonoid chromatography data (Toivonen, 1974; Catling, 1989;
Cayouette, 1992).
4


These studies found that Carex species tend to be differentiated
morphologically, geographically, ecologically (e.g., microhabitat), genetically (e.g.,
allozymes), cytologically (e.g., karyotype), and chemically (e.g., flavonoids)
(Toivonen, 1974; McClintock, 1994). On the other hand, Carex hybrids have been
shown to be intermediate for parental species morphological and chemical
characteristics (Toivonen, 1974; Cayouette, 1985; Cayouette, 1987; Catling, 1989;
Waterway, 1990; Cayouette, 1992; Ford, 1992; Waterway, 1994). With respect to
geography, natural hybrids are usually found in the vicinity of one or both parents
and, if only one parent is present, it is likely the maternal species. Ecologically,
hybrids often inhabit disturbed or intermediate (relative to parental habitats) sites,
unsuitable for either parental species (Cayouette, 1992). Not surprisingly, the genetic
characters of hybrids are generally a combination of parental species genetic
characters. Where parental species are homozygous for different alleles at an
allozyme locus, Fi hybrids will be heterozygous (Waterway, 1990; Cayouette, 1992;
Ford, 1992; Waterway, 1994). Furthermore, Fi hybrid karyotypes reflect parental
species chromosome numbers in that the hybrid diploid chromosome number is
generally the sum of the parental haploid numbers. In addition to these characteristics,
hybrids can be readily identified by complete or partial sterility (Cayouette, 1985;
Cayouette, 1992; Waterway, 1994).
Experimental and natural hybridizations have revealed that large differences
in chromosome number between two species of Carex do not significantly reduce
5


their ability to hybridize (Cayouette, 1985; Whitkus, 1988; Waterway, 1990;
Cayouette, 1992; Waterway, 1994). However, Fi sterility in Carex hybrids is
variable, with complete sterility generally occurring in hybrids having more distantly
related parent species and/or large differences in chromosome number. Such hybrids
tend to exhibit a high level of meiotic aberrations, including trivalents or tetravalents,
presumably contributing to sterility (Cayouette, 1985; Waterway, 1990; Cayouette,
1992; Luceno, 1993; Waterway, 1994).
1.3 Detecting Introgression
In cases where sterility is incomplete, some hybrids may be able to self
fertilize or backcross to one of the parental species, resulting in introgression
(Strickberger, 2000a). Suppose that a partially fertile hybrid (species A x species B)
backcrosses to one of the parental species {species B). The resulting introgressed
individual would be expected to contain genetic material from both species, as in the
hybrid, but there should be a higher percentage of species B genes relative to
species A genes. If this backcrossed individual again crosses with a species B
individual we expect further diminution of species A genes in the resulting progeny.
Repeated backcrosses to species B result in the transfer of a small amount of genetic
material from species A to species B.
Given that the amount of genetic material transferred is small, verifying that
introgression has occurred is challenging. Theoretically it is expected that
introgressed individuals will overall resemble a single species genetically and
6


morphologically, differing with respect to one or more traits that more closely
resemble another species. Within Carex, hypotheses of introgression have rested on
tenuous morphological and allozyme evidence. It has been suggested that when an
allele is found in two different species, but is relatively rare in one of them,
introgression may be the source of the rare allele. That is, the allele was transferred
from one species, where it is common, to relatively few individuals in another species
(McClintock, 1994; Tyler, 2003). In combination with knowledge that two species
hybridize and have partially fertile offspring, rare alleles may provide evidence of
introgression. However, in closely related species, where hybridization and
production of fertile offspring is more likely, sharing of alleles could easily be the
result of recent divergence from a common ancestor, not interspecific gene flow
(Tyler, 2003). As such, conclusively detecting introgression requires both that the
characters studied are species specific (i.e., an allele occurs in one species but not the
other) and that there are populations of pure species that do not exhibit the putative
introgressed characters (i.e., the rare allele).
Attempts to detect introgression in other plant genera and some animals have
made use of a variety of molecular biology techniques involving various genetic
markers such as RAPDs (random amplified polymorphic DNA), AFLPs (amplified
fragment length polymorphisms), microsatellites, and RFLPs (restriction fragment
length polymorphisms) (Arnold M. L., 1990; Jarvis, 1999; Choler, 2004; Grant,
2004). Again, these markers can only detect introgression if they are species specific.
7


Furthermore, a marker must be present in the introgressed region for the gene transfer
to be detected. RAPDs and AFLPs essentially amplify random regions of the genome
(Williams et al., 1990; Klug, 2005). Subsequent sequencing or visualization of
fragment sizes provides many markers distributed throughout the genome, increasing
the chance of finding sequences (or fragments) specific to a species and detecting
introgression, if it has occurred. AFLPs have been used to detect introgression
between subspecies of C. curvula (Choler, 2004). Microsatellites, which are
tandemly repeated short DNA sequences, have also proved to be effective markers.
The repeated segments are distributed throughout the genome and the number of
repeats in each segment varies among individuals. Microsatellites have been used to
detect introgression in Darwins finches (Grant, 2004). Finally, RFLPs that use
enzymes with cleavage sites occurring in repeated DNA regions can be used to
identify introgression. If cleavage sites within the repeated region vary among
species, and the region is distributed throughout the genome, RFLPs can identify
species-specific genetic material. For example, RFLPs of ribosomal DNA (rDNA),
which occurs in tandem repeats throughout the genome, have been used to identify
introgression in Iris (Arnold M. L., 1990).
1.4 Ribosomal DNA as a Genetic Marker in Carex
Recently, investigators have used ribosomal DNA (rDNA) sequences to
construct phylogenies that hypothesize evolutionary relationships in Carex (Roalson
E. H., 2001; Starr, 2003; Hendrichs, 2004). Two nuclear sequences have been
8


particularly useful, namely internal and external transcribed spacers (ITS and ETS
respectively), which are located between highly conserved ribosomal genes (Fig 1).
Together, the spacers and rDNA genes are found in arrays of tandem repeats at many
loci and it is estimated that these sequences are present in hundreds to thousands of
copies throughout the genome. ITS and ETS sequences can be used to assess
phylogenetic relationships at the species level because, although the rDNA genes are
conserved, the spacer regions evolve relatively quickly, providing enough resolution
to distinguish among species (Starr, 2003).
Given that ITS and ETS rDNA sequences differentiate species and are
distributed throughout the genome, these sequences could be of use in detecting
hybrids and introgressants in Carex. Specifically, Fi hybrids would be expected to
have copies of rDNA sequences characteristic of both parental species along with
other hybrid characteristics (e.g., heterozygosity for parental allozyme alleles,
sterility, and morphological intermediacy). If introgression has occurred and rDNA is
among the introgressed genetic material, some rDNA characteristic of one species
should be present in the genome of individuals of another species and these
individuals should lack other hybrid characteristics.
9


IQS ftogton
ITS Region
26S ETS 2 1 NTS | ETS 1 18S ITS 1 5.BS ITS 2 26S'

ETS T I NTS*
_______I_________
26$-Fa(24 top)
'26S-fb (427 bp)
2SS-FJ28 bp)
26S
IGS-R (c. 386 bp)
IGS-Ra (c. 384 bp)
IQS-Rb (c. 389 bp)
ETS-F (e. 588 bp)
1 ETS 2 ! NTS ETS 1 1 j ETS 1f
(IGSf) ! !
7T
\
5*>TQAQTKQTA<3'
d motif)
18S
18S-fl(12bp)
Figure 1. Angiosperm rDNA Gene Region (Starr, 2003)
The 26S, 18S, 5.8S and 26S ribosomal RNA genes are interspersed with external transcribed spacers
(ETS), internal transcribed spacers (ITS), and non-transcribed spacers (NTS). The ribosomal genes are
highly conserved and do not vary among species. The spacer regions, which are non-coding, vary
considerably and generally differentiate species. Regions of interest for this research are the ITS region
and ETS If.


1.5 Cleaved Amplified Polymorphisms
Rather than sequencing rDNA regions for many individuals, which is time
consuming and costly, sequences can be distinguished by cleaved amplified
polymorphisms (CAPs). This method only requires sequence information for a few
individuals from each species in order to determine that the sequences vary among
species but not within them. Sequences are used to choose a restriction enzyme that
cleaves the sequences of two species differently. Restriction endonucleases are
derived from bacteria, where they degrade foreign DNA by recognizing and cleaving
specific sequences (Weaver, 2005). If restriction enzyme recognition sites differ
between sequences of equal size, then digestion with that enzyme will yield different
sized fragments. Following amplification of the rDNA regions, PCR-amplified DNA
is digested with the appropriate enzyme and fragments are visualized on an ethidium
bromide agarose gel. Different species will yield different banding patterns,
providing a fast and inexpensive method to distinguish between species-specific
rDNA sequences.
1.6 Model System: Carex Sections Ceratocystis and Bicolores
Two sections of Carex, namely sections Ceratocystis and Bicolores, were
used to test several hypotheses using CAPs of rDNA spacer regions and a chloroplast
region. Within Carex section Ceratocystis the rDNA sequences ETS If (referred to
as simply ETS hereafter) and ITS (Fig. 1) from all North American members have
been used to construct a phylogeny for this section. The species include Carex
11


cryptolepis Mack., C. viridula Michx., and the recently described C. viridistellata,
among others. For these three species there is little to no variability in ETS and ITS
sequences within species, but consistent differences among them (N. J. Derieg, pers.
comm., 2007). Additionally, many hybrids involving species in Carex section
Ceratocystis have been reported, including: C. cryptolepis x C. viridula, C. flava L. x
C. hostiana DC., C. flava x C. viridula, and C. hostiana x C. viridula (Cayouette,
1992). Mixed populations that contain either C. cryptolepis and C. viridistellata and
hybrids (Springville Marsh population) or C. cryptolepis and C. viridula and hybrids
(Fish Fry Lakes population) have been collected. Hybrids in each population were
identified by sterility, morphology, and allozyme data. Since these hybrids are
completely sterile, it is expected that introgression is not occurring among the species
in these populations.
Within Carex section Bicolores ITS sequences for C. bicolor All. and C.
aurea Nutt, were obtained from Genbank. Again the sequences are different for each
species. Furthermore, hybrids within this section are often reported as partially fertile
(Ball, 2002). A population presumably comprising C. bicolor and purported C.
bicolor x C. aurea hybrids and/or introgressants has been collected (Donnelly
Campground population) with a population of C. aurea observed nearby (Bruederle,
pers. comm., 2007). Given the reported partial fertility of hybrids within Carex
section Bicolores, this population is expected to contain introgressants.
12


2. Hypotheses
Analysis of cleaved amplified polymorphisms (CAPs) was used to test several
hypotheses.
Individuals from two populations exhibiting F i hybrid characters, (that
is, heterozygosity for parental alleles at allozyme loci, sterility, and
intermediate morphology) are expected to be exhibit CAPs of ITS and
ETS sequences that are characteristic of both parental species.
In hybrids demonstrating ITS and ETS CAPs characteristic of both
parental species, the maternal species can be identified using CAPs of
a chloroplast DNA (cpDNA) sequence, which is expected to be
identical to the CAP of the maternal species.
Mixed populations of Carex section Ceratocystis should reveal a lack
of introgression in this section, given that the known hybrids in these
populations are completely sterile.
A mixed population of Carex section Bicolores should reveal
introgression, given that hybrids have been shown to be partially
fertile in this section (Ball, 2002).
13


3. Materials and Methods
3.1 Sampling
Populations to be screened for introgression were sampled as part of related
research describing genetic diversity and structure within Carex sections Ceratocystis
and Bicolores. The individuals studied from Carex section Ceratocystis include C.
cryptolepis, C. viridula, and the recently described C. viridistellata, with various
hybrid combinations, and, from Carex section Bicolores, putative C. bicolor and C.
bicolor x C. aurea hybrids and/or introgressants (Bruederle, pers. comm., 2007). For
each individual, tissue was dried and preserved in silica gel and a pressed, dried
voucher of the specimen was deposited at the University of Colorado in Denver.
Three populations were chosen for screening based on the presence of known hybrids
or suspected introgressants: 50 individuals from Springville Marsh in Ohio
(405922.38 N, 832357.67 W), 36 individuals from Fish Fry Lakes in
Minnesota (473851.44 N, 9r2634.21 W), and 59 individuals from Donnelly
Campground in Alaska (634025.49 N, 145 5259.00 W). The Springville
Marsh population reportedly comprises a mixture of C. cryptolepis and C.
viridistellata (N. J. Derieg, pers. comm., 2007), though the collection consists of 47
C. cryptolepis and three sterile hybrids (C. cryptolepis x C. viridistellata). Likewise,
the Fish Fry Lakes population comprises C. cryptolepis and C. viridula (Bruederle,
pers. comm., 2007), while the collection consists of 34 C. cryptolepis and two sterile
14


hybrids (C. cryptolepis x C. viridula). The composition of the Donnelly Campground
population is questionable, likely comprising C. bicolor, C. bicolor x C. aurea
hybrids and introgressants (Bruederle, pers. comm., 2007).
3.2 DNA Extraction
Two methods were used to extract DNA from the specimens. Initial
extractions were performed with DNeasy Plant Mini Kit (Qiagen). Briefly, 10-60 mg
of dried tissue was ground in liquid nitrogen and homogenized in lysis buffer
containing RNase to release the contents of cells. Following centrifugation the
supernatant was spun through a QIAshredder mini-column to remove cell debris.
DNA from the flow through was precipitated with 95% ethanol and selectively bound
to the membrane in a DNeasy mini-column. Bound DNA was washed and then
eluted.
The majority of extractions followed a protocol using isopropanol to
precipitate the DNA without the use of spin columns. 10-60mg of dried tissue was
ground in liquid nitrogen and combined with 750 pi of buffer containing 50 mM Tris
at pH 8, 10 mM EDTA at pH 8, 100 mM NaCl, 1% SDS, and 10 mM B-
Mercaptoethanol to lyse cells and denature proteins. After 10 minutes at 65C, 150 pi
of a solution of 3 M potassium acetate and 11.5% glacial acetic acid was added
followed by a 20-minute incubation on ice to precipitate proteins. Samples were then
centrifuged and 750 pi of the supernatant was added to 750 pi of isopropanol to
precipitate DNA. After centrifugation, the supernatant was removed and DNA
15


precipitate washed with 80% ethanol. Following removal of the ethanol, DNA pellets
were air-dried overnight and resuspended in 50-200 pi sterile water.
Tomato plant DNA was extracted simultaneously with the Fish Fry Lakes and
Donnelly Campground populations to control for cross contamination during the
extraction process. All extractions were quantified using a biophotometer
(Eppendorf) and stored at -20C.
See APPENDICES A and C for detailed instructions on extractions and
preparation of required solutions.
3.3 Amplification of ITS, ETS, and Chloroplast Sequences
PCR was performed using ITS5i and ITS4i primers specific for the ITS region
(Roalson E. H., 2001), and ETS-F and 18S-R primers specific for ETS If (Starr,
2003) (Fig 1). Reagents for the amplification were purchased from New England
Biolabs. Reaction mixtures contained 5 pi 1 OX PCR buffer, 1.5 mM MgCfy 200 pM
of each dNTP, .5 pM of each primer, 1-10 pg/ml template DNA, .02 units/pl Taq
DNA polymerase, and de-ionized water to a final volume of 50 pi. Amplifications
were performed in an Eppendorf Mastercycler with an initial denaturation for 2
minutes at 95C followed by 31 cycles of denaturation at 95C for 1 minute,
annealing at 55C for 1 minute, and elongation at 72C for 1 minute, with a 10 minute
final extension at 72C and 4C holding temperature.
TabC and TabF primers were used to amplify a roughly lkb chloroplast
region of hybrids (Shaw, 2005, 2007). Reaction conditions were identical to those for
16


ITS and ETS except for a concentration of 3 mM MgCh and .4 pM of each primer.
Amplifications were performed in an Eppendorf Mastercycler with an initial
denaturation for 2 minutes at 94C followed by 31 cycles of denaturation at 94C for
1 minute, annealing at 55C for 30 seconds, and elongation at 72C for 2 minutes,
with a 10 minute final extension at 72C and 4C holding temperature.
Amplification products were visualized in a 1% agarose gel containing .33
pg/ml ethidium bromide and quantified by comparison to a PCR ladder (New
England Biolabs). All PCR reactions were run with a negative control lacking DNA
to ensure that the PCR reagents were not contaminated.
See APPENDIX D for a general PCR protocol. See APPENDICES B and F
for detailed instructions on making gel buffer and pouring and running gels.
3.4 Sequencing and Restriction Digests
The ITS region (Fig. 1) was sequenced for five C. cryptolepis individuals
representing five populations, two C. viridistellata individuals representing two
populations, and one C. viridula individual. The ETS region (Fig. 1) was sequenced
for six C. cryptolepis individuals representing six populations, three C. viridistellata
individuals representing three populations, and two C. viridula individuals
representing two populations. The chloroplast region was sequenced for one
individual each of C. cryptolepis and C. viridistellata and two C. viridula individuals.
All sequencing was performed by Nathan Derieg, with the use of the sequencer (CEQ
8000, Beckman Coulter) at the Rocky Mountain Center for Conservation Genetics
17


and Systematics at the University of Denver. ITS sequences for C. bicolor and C.
aurea were obtained from Genbank with accession numbers AY278283 and
AF285062, respectively.
NEB cutter V2.0 (New England Biolabs) was used to determine restriction
enzyme cut sites for the species specific ITS, ETS, and chloroplast sequences.
Enzymes were chosen based on their ability to differentially cut the sequences of two
species known to hybridize (Table 1).
Digestions were performed at a final volume of 20 pi with 50-100 ng/pl PCR-
amplified DNA, 2 pi of 10X buffer (contents vary with enzyme), .1 pg/pl BSA (when
required), and .5-5 units enzyme, brought to volume with de-ionized water.
Appropriate buffer and BSA were supplied with the enzyme from New England
Biolabs. Digestions were incubated at 37 C for 2-6 hours and visualized on a 1.5%
agarose gel containing .33 pg/ml ethidium bromide. All digests included an uncut
control, lacking enzyme. For the Fish Fry Lakes and Donnelly campground
populations DNA extracted from tomato was also amplified and digested to check for
cross-contamination during extractions.
See APPENDIX D for a detailed restriction digest protocol. See
APPENDICES B and F for instructions on making gel buffer and pouring and
running gels.
18


Table 1. Enzymes Used to Generate Cleaved Amplified Polymorphisms for Carex Species in Three
Populations
Enzymes and predicted fragment sizes of ITS, ETS, and chloroplast DNA sequences following digestion.
Fragment sizes were predicted with New England Biolabs Cutter (V 2.0). Enzymes were chosen based on
their ability to differentially cut the sequences of two species known or suspected to hybridize.
Population Screened Sequence Enzyme Species Fragments (bp) Uncut Size (bp)
Spnngv> le ETS Hmfl C. cryptotepis 533,144 677
Marsh ETS Hihfl C. vindistellata 144,217,319 677
Sp'ngvi le ITS ECO01091 C. cryptotepis 386,291 677
Marsh ITS ECO01091 C. vindistcllata 386.200,91 677
Spmgvi le Cp taoc-tab' Alul C. cryptotepis 225,751,72 1048
Marsh Co taoc-tab' Alul C. vmdtstetiata 225.823 1048
ETS Msll C. cryptotepis 240,204,233 677
Fish Fry Lake ETS MslI C. vindula 444.233 677
ITS Bccl C. cryptolepis 357,75,245 677
Fish Fry Lake ITS BccI C. vindufa 135.222.75.245 677
Donne i-y ITS BsaHI C. Dicolor 160,27,429,61 677
CanrDQ'ound ITS BsaHI C. aurea 160.27.490 677


I
i
I
I
4. Results
i
4.1 Carex cryptolepis x C. viridistellata
Amplifications of ETS and ITS sequences yielded DNA about 680 base pairs
in length and amplification of the chloroplast region DNA about 1050 base pairs
(Table 1). In general, the ETS and chloroplast regions amplified well, but
amplification of the ITS region was relatively inconsistent (data not shown).
Restriction digests of ETS and ITS DNA sequences for the three hybrid
individuals from this population yielded a banding pattern comprising the combined
patterns of both parental species (Fig 2). For example, Hinfl cleaves C. cryptolepis
ETS sequence into fragments 533 and 144 base pairs in size and C. viridistellata ETS
sequence into fragments 319, 217, and 144 base pairs in size (Table 1 and Fig 2a).
The C. cryptolepis x C. viridistellata ETS CAP yields fragments 533, 319, 217, and
144 base pairs in size. Results from the ITS CAPs were similar, demonstrating
combined parental banding patterns for the three hybrids (Fig. 2b). Flowever, the
CAP of C. viridistellata ITS sequence should generate three bands, as predicted by
NEB cutter (Table 1) but the smallest band of 91 base pairs is visible neither in the
ITS CAP of C. viridistellata nor the three hybrids. Additionally, the 200 base pair
band specific to the ITS CAP of C. viridistellata is relatively faint in the hybrids.
20


s
I
M V) to 5 * 5 to =5
3. 3. 5. g 3.2 Q. £
£ 4 £ 4 a 0) to
Q o Q. To Q 11 11
& & C > ^d
d O O O d x d x
i t £ S
$ o 5 to
*
S' S' S' 1
d b b d
Figure 2. CAPs of ETS and ITS Sequences for Representative Springville Marsh
Species and Hybrids
Uncut C. cryptolepis ETS (a) and ITS (b) DNA is approximately 680 base pairs (bp).
Digestion of ETS DNA (a) with Hinfl and ITS DNA (b) with Eco0109I yields
distinct banding patterns for C. cryptolepis and C. viridistellata. Hybrids have bands
present in both parental species for ETS (a) and ITS (b), though the 200 bp ITS band
is weak in hybrids and the 91 bp ITS band is not visible for C. viridistellata and
hybrids (see Table 1).
21


The CAP banding pattern for the C. cryptolepis x C. viridistellata TabF-TabC
chloroplast region (cpDNA) contained only the bands characteristic of C. cryptolepis
(Fig 3), lacking the additive pattern seen in the nuclear sequences (Fig. 2).
Additionally, a 1048 base pair band, corresponding to the uncut sequence size,
occurred in all of the digestions. This band was diminished but not eliminated after
increased enzyme concentration and digestion time (data not shown).
CO CO Q. (O 3 3 3 3 co 5 3 3 CO % 3 3 co %
Ql Ql 3 3 5.2 5.2 5.2 o
2 3 3 4> yt .CO .co c
o O O <0 6 CO 1 O S o S O S o
k. Q. Q. | o
0) b £ .c b b * +* 3
<0 O O o > o o O o c
2 b b d b b d x ox b x D
bp
12 3456789 10
Figure 3. Chloroplast CAPs of Springville Marsh Species and Hybrids
Digestion of TabC-TabF cpDNA with Alul yields distinct banding
patterns for C. cryptolepis and C. viridistellata. Hybrids exhibit the
banding pattern characteristic of C. cryptolepis. The uncut sequence is
approximately 1050 bp and a small amount of uncut DNA is present in
all of the digestions.
22


Finally, each of the remaining 47 individuals in this population exhibited ETS
and ITS CAPs characteristic of C. cryptolepis (Fig. 4). ITS bands (Fig. 4b) were
often faint compared to ETS bands (Fig. 4a).
a
00 s o> o r- 61 (*) fO CM CO CO CO s ID CO 8 r- CO 00 CO o> CO o O' CM CO 3 lf> <0
> > > > > > > > > > > > > > > > > >
CO (O CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO
Figure 4. Representative Springville Marsh Individuals Screened for Introgression
(a) Digestions with Hinfl of non-hybrid ETS sequence from Springville Marsh
individuals 28-30 and 34-46 have a banding pattern characteristic of C.
cryptolepis. Hybrid individuals 31-33 exhibit the combined banding pattern of C.
cryptolepis and C. viridistellata. (b) Digestions with Eco0109I of non-hybrid ITS
sequence from Springville Marsh individuals 9-20 have a banding pattern
characteristic of C. cryptolepis. ITS bands were fainter than ETS bands.
23


4.2 Carex cryptolepis x C. viridula
As with the Springville Marsh population, amplification of ETS and ITS
regions yielded DNA about 680 base pairs in length, with ETS yield generally higher
and more consistent than ITS (Fig. 5a and b). Additionally, for one hybrid individual
(FFL02), ETS amplification yielded a strong 680 base pair band and a weak 350 base
pair band (Fig. 5b and c, arrows). Since the TabF-TabC chloroplast regions of C.
cryptolepis and C. viridula have identical sequences (unpublished data) an enzyme
that would generate different CAPs in each species could not be chosen, as such, this
region was not amplified nor digested.
1 23456789 10
b
123 4 56789 10
C
bp
500
1 2
bp
677
350
Figure 5. Representative Amplifications of ITS and ETS Regions from the Fish
Fry Lakes Population
Amplification of the ITS (a) but not ETS (b) region varied among individuals. The
same regions in tomato, a positive control, amplified well. The negative controls
lacked DNA template. A 350 bp amplification product was also seen only in the
ETS amplification product of individual 2 from this population (arrows in (b) and
(c))-
24


Digests of ETS sequences for the two hybrid individuals in this population
yielded banding patterns comprising a combination of the parental banding patterns
(Fig. 6a and b). The 350 base pair band seen following amplification of ETS of
hybrid FFL02 is still seen following digestion. Digests of ITS sequences for one of
the two hybrid individuals in this population yielded a banding pattern comprising a
combination of the parental banding patterns (Fig. 6c). The other hybrid individual
did not yield discemable bands following digestion due to an insufficient amount of
ITS DNA despite repeated amplification attempts and the use of DMSO and betaine
to increase PCR yield (data not shown). Additionally, a 75 base pair band is
predicted for both C. cryptolepis and C. viridula ITS CAPs (Table 1). While this
band is visible in C. viridula and hybrids, it is not seen in the C. cryptolepis
individual in Fig. 6c, likely due to low amplification yield of this individual.
However, this band was seen in all digestions of individuals from the Fish Fry Lakes
population after DNA concentration was increased (Fig. 8b).
25


Figure 6. CAPs of ETS and ITS Sequences for Representative Fish Fry
Lakes Species and Hybrids
In (a) and (b) digestion of ETS DNA with MslI yields distinct banding
patterns for C. cryptolepis and C. viridula. Hybrid individuals exhibit both
parental banding patterns. Additionally, the 350 bp amplification product is
seen in individual 2 ((a), lane 2, Fig. 5b and c). (c) Digestion of ITS DNA
with BccI also yields distinct banding patterns for C. cryptolepis and C.
viridula. One hybrid has bands present in both parental species.
26


ETS CAPs of the remaining 34 individuals generally yielded banding patterns
characteristic of C. cryptolepis, however a faint band specific to C. viridula at 444
base pairs was often visible. Doubling the enzyme concentration and increasing
digestion time to six hours diminished most of these bands (compare Fig. 7a and b).
ITS CAPs of the remaining 34 individuals also resembled C. cryptolepis, though
some smearing or irregular bands were seen. Reamplification followed by digestion
diminished most irregularities though smearing and some unpredicted bands persisted
(compare Fig. 8a and b). The ITS and ETS CAPs of tomato were different from
either of the Carex species in this population and contamination from tomato to
Car ex or vice versa was not detected (data not shown).
27


a
3
u.
LL
bp
2
7
b
500
200
CM
CO
3
8
-i
sg
c
o
3 o
bp
1 2 3 4 5 6 7
Figure 7. Representative ETS CAPs from Fish Fry Lakes Individuals
ETS fragments for the same individuals after 3 hours digestion and 3.75
untis MslI (a) and 6 hours digestion and 5 units MslI (b). In both digests,
three bands characteristic of C. cryptolepis at 240,233, and 204 bp can be
seen. A 444 bp band characteristic of C. viridula is diminished in (b).
28


a
bp
CO
O
LL
o>
3
U_
LL
$ 3
C. a
<0 g
O S
*- o
3 *3
O C
C O
D O
bp
12 3 4
6 7
10 11
b
bp
a>
a
S
3
U-
o s
bp
123 45678 9 10
Figure 8. Representative ITS CAPs from Fish Fry Lakes Individuals
(a) ITS CAPs after 1 hour digestion with 5 untis of Bccl. FFL02 and FFL24 are
the two hybrids in this population. C. viridula from another population is present
for comparison, (b) The results of digestions for 2 hours of the same individuals
as in (a) with the same reaction conditions using freshly amplified DNA. C.
cryptolepis and C. viridula have different banding patterns and the hybrid FFL24
has bands characteristic of both parental species, but all individuals exhibit some
smeared bands. FFL02, the other hybrid, does not yield a clear banding pattern.
FFL08 has a distinct unpredicted band at about 320 bp.
29


4.3 Carex bicolor x C. aurea
ITS was amplified from all individuals in the Donnelly Campground
population and from four C. aurea individuals collected from a site in Vermont.
Again, ITS bands were approximately 680 base pairs and amplification yield was
inconsistent. An enzyme was chosen to cut C. bicolor and C. aurea Genbank
sequences differently. However, the CAPs of both species were identical following
digestion with the same enzyme (Fig. 9). As such, screening using that enzyme
would be uninformative for identifying hybrids and/or introgressants.
1 2 3 4 5 6
Figure 9. ITS CAPs for Two Species in Carex Section Bicolores
Following digestion with BsaHI, C. bicolor and C. aurea yield identical
banding patterns, though lanes are slightly staggered.
30


5. Discussion
5.1 Car ex cryptolepis x C. viridistellata
CAPs of ITS and ETS sequences for C. cryptolepis and C. viridistellata were
species specific, and the three hybrid individuals in the Springville Marsh population
had CAP banding patterns characteristic of both parental species, as expected (Fig. 2).
This indicates not only that hybrids have copies of both parental species rDNA, but
also that both species rDNA sequences are amplified when they are combined in a
single individual. Inability to see the smallest fragment (91 base pairs) in the ITS
CAP of C. viridistellata and hybrids (Table 1 and Fig. 2b) is likely due to low ITS
amplification yield. Additionally, ITS CAPs of the hybrids exhibited a weak 200
base pair band specific to C. viridistellata. This could indicate that both species ITS
regions are not amplified with equal efficiency when present in the same reaction.
That is, the C. cryptolepis ITS region may outcompete the C. viridistellata ITS region
for primers and thus be amplified at a higher rate. In this case, relatively more C.
cryptolepis ITS DNA would be present in the digestions and the C. cryptolepis bands
would appear brighter. Alternatively, a longer digestion time may be required to
cleave the C. viridistellatas 291 base pair band into two bands of 200 and 91 base
pairs (see discussion of ETS CAPs for the Fish Fry Lakes population).
CAPs of the cpDNA indicate that C. cryptolepis is maternal in the hybrids.
Since pollen contributes only nuclear DNA, all chloroplastic (and mitochondrial)
31


DNA is inherited maternally. The hybrid cpDNA CAPs were identical to those
specific for C. cryptolepis (Fig.3). This is congruent with the morphology of the two
species. Since C. viridistellata has a longer perigynia beak than C. cryptolepis
(unpublished observation) it is more likely that C. viridistellata pollen tube growth
will be sufficient for sperm to reach a C. cryptolepis egg than vice versa (A. A.
Reznicek, pers. comm., 2007).
Interestingly, there were persistent uncut bands in all of the digests of cpDNA.
Extended digestions with increased enzyme concentration diminished but did not
abolish these bands. The enzyme used in these digestions expired in 2001 and,
though previously unused, may have lost some of its activity. This hypothesis is
consistent with the fact that the cpDNA region amplified well and thus the
concentration of DNA in the digestions was high. Alternatively, other sequences,
which lack the same enzyme recognition sites, could be present. This seems unlikely
given that sequences for the cpDNA were clean, and did not have ambiguous
nucleotides. Sequencing was performed using DNA amplified by the same protocol
used to generate the CAPs and so it is expected that the same sequences were present
in the sequencing reactions as in the digestions. Furthermore, any sequences not cut
by the enzyme would have to differ exactly in the cut recognition sites, which is
improbable.
Finally, CAPs of ITS and ETS sequences for all other Springville Marsh
individuals exhibited bands specific to C. cryptolepis (Fig. 4), indicating that rDNA
32


has not introgressed from C. viridistellata to C. cryptolepis. Since ITS amplification
yields were low, and some hybrid bands were faint, ITS DNA concentration was
increased (1.5 times) in digestions for all other individuals in this population. Again,
it is possible that C. cryptolepis ITS DNA is preferentially amplified and therefore C.
viridistellata-spec'ific bands were not visible. However, given the strength of both
species ETS bands, and that ITS and ETS regions occur in tandem, it is unlikely that
one region has introgressed without the other. Furthermore, given that hybrids were
completely sterile, it was expected that introgression has not occurred in this
population.
5.2 Carex cryptolepis x C. viridula
In the Fish Fry Lakes population, CAPs of ITS and ETS sequences also
differentiated the two species, identified hybrids (Fig. 5), and indicated a lack of
introgression of rDNA between the species. However, there were a few unexpected
results. Interestingly, one individual from this population (FFL02) yielded the
expected 680 base pair sequence along with an unpredicted 350 base pair sequence
following amplification of the ETS region (Fig. 5b and c). This band was seen when
DNA obtained from two different extraction procedures was amplified, suggesting
that the DNA was not contaminated. Contamination during PCR is doubtful given
that contamination was not detected from the positive control tomato DNA, or from
the reagents, as indicated by the negative controls. It is also improbable that the 350
base pair band is due to non-specific amplification given that it occurred only in a
33


single individual. Thus the most likely explanation for this small band is the presence
of a truncated 350 base pair ETS sequence where regions of primer annealing have
not been affected. Since this smaller ETS sequence did not amplify as well, it is
either present in fewer copies than the larger ETS sequence, or has a decreased
amplification efficiency, possibly due to some changes in the primer regions. Though
one primer is anchored in the highly conserved 18S rDNA gene, the other primer is
located in a conserved region within ETS1 (Starr, 2003) (Fig. 1). It is conceivable
that mutations have occurred in this primer region, since it is non-coding and changes
are neutral with respect to selection.
Curiously, ITS amplification was extremely weak in this same individual and
discrete bands were not discemable following digestion (Fig. 8). Several repetitions
of amplification and digestion yielded some bands specific to C. cryptolepis (data not
shown). On the other hand, ETS CAPs for this individual contained both species
bands, verifying that the individual is a hybrid (Fig. 6a). If the 350 base pair ETS
region in this individual is evidence of an evolving ETS region, it is possible that the
ITS region has also sustained mutations. If changes have occurred in the primer
regions, which is possible if the primers are not anchored in the rDNA genes,
amplification efficiency could be affected. Since ITS was difficult to amplify in
many individuals, there may be a general weak annealing of primers or formation of
secondary structure in the single-stranded DNA during PCR. Though addition of
DMSO (dimethylsulfoxide) and betaine seemed to increase ITS amplification yield
34


for the troublesome hybrid, digestions of the resulting PCR products did not yield
discemable bands. New England Biolabs has reported that organic solvents,
including DMSO, can contribute to an enzymes star activity (non-specific cleavage).
Overall, digestions of ITS and ETS regions for this population were less
consistent than for the Springville Marsh population. ETS CAPs of several
individuals appeared to have a band characteristic of C. viridula (at 444 base pairs),
but these bands diminished after increased enzyme concentration and digestion time
(Fig. 7). This suggests that the enzyme cleaves the ETS sequence into 444 and 233
base pairs readily, but the second cut, which cleaves the 444 base pair fragment into
240 and 204 base pair fragments, requires more time. Preferential cleaving by
enzymes has been reported by New England Biolabs. On the other hand, the C.
viridula bands could represent introgression of C. viridula rDNA into C. cryptolepis.
This possibility can only be excluded if longer digestions completely eliminate the C.
viridula bands.
The fact that ITS CAPs resembled C. cryptolepis supports the conclusion that
introgression of rDNA has not occurred. Again, this is consistent with complete
sterility of hybrids in this population, but it is possible that amplification efficiencies
differ for different species ITS or ETS regions and introgression may not be detected
by the CAPs. Additionally, ITS CAPs for the Fish Fry Lakes population were
irregular and bands were often indistinct and smeared. This could be idiosyncratic to
the enzyme, though starring activity has not been reported for it. Alternatively,
35


degradation of the DNA due to exonucleases or repeated freezing and thawing could
be responsible. Reamplification did yield discemable bands following digestion, but
some smearing was still visible as well as unpredicted bands (Fig. 8). Specifically, a
320 base pair band was seen in several individuals (FFL08 in Fig. 8b). This may
again be the result of preferential cleaving whereby the 320 band has not been
cleaved into the predicted 75 and 245 base pair bands.
Finally, as mentioned, cpDNA could not be used to determine the maternal
species for the hybrids in this population. However, based on the hypothesis that the
species with the shorter perigynia beak is maternal (A. A. Reznicek, pers. comm.,
2007), it is suggested that C. viridula is the maternal species, given that the range of
beak lengths for C. cryptolepis and C. viridula is 1.4-2.5mm and .3-1.3mm,
respectively (Crins, 2002).
5.3 Carex bicolor x C. aurea
The enzyme chosen to generate different CAPs for these species did not cut
the sequences as expected. It may be that the difference between the CAPs of these
two species is too small to detect (Table 1). On the other hand, the sequences used to
choose the enzyme represent a single individual from each species and these
sequences differed at only one nucleotide position (Hendrichs, 2004). As such, this
difference may not be consistent and some populations of either species may have
ITS sequences identical to some populations of the other species. ETS sequences,
36


which tend to vary more among species (Starr, 2003), were not available for these
species.
5.5 Detecting Introgression with Cleaved Amplified
Polymorphisms of Ribosomal DNA
Since ITS and ETS occur in tandem, they represent two different markers for
the same genomic region. Thus, even though ITS amplification was variably weak in
the Springville Marsh and Fish Fry Lakes populations (Fig. 5a), the efficient
amplification of ETS (Fig. 5b) and relatively strong ETS hybrid banding patterns
(Fig. 2a and 6a) confirm the results of the ITS CAPs in these two populations. The
results are consistent with the observation that hybrids within Carex section
Ceratocystis tend to be sterile (Bruederle, unpublished observation), which would
prevent introgression.
Given that there are many copies of rDNA sequences in the genome, it is
likely that if introgression has occurred, it may involve ITS and ETS regions.
However, the distribution of rDNA repeats within the genome of the species studied
here is currently unknown. Thus, it is still possible that other genomic regions have
introgressed. Alternatively, rDNA may be a suitable marker for detecting
introgression but the method of CAPs used here is not sensitive enough to identify
small amounts of genetic exchange.
Though there is a lack of knowledge regarding the distribution of rDNA
repeats throughout the genomes of the species in Carex sections Ceratocystis and
37


Bicolores, it is likely that rDNA regions are adequate markers for introgressing genes.
Vanzela et al. (1998) used in situ hybridization to determine the location and
distribution of the 18S-5.8S-26S rDNA repeats in eight species of Rhynchospora
(Cyperaceae). This study found that in all of the species rDNA repeats were
localized at telomeres, though sites of localization varied from 4-8 for species with 2n
= 10-30 up to 30 for species with 2n = 50. This variation makes it difficult to predict
the number and distribution of rDNA repeats in Carex. However, if at least one half
of the chromosomes contain rDNA repeat regions, it seems likely that these would
serve as a suitable marker for recent introgression.
rDNA has been used to detect introgression among species of Iris using
RFLPs (restriction length fragment polymorphisms). Arnold et al. (1990) digested
genomic DNA with restriction enzymes, separated the fragments by gel
electrophoresis, transferred the separated fragments to a filter, and used a 32P-labeled
rDNA probe to detect fragment size differences. This method is based on species
specific polymorphisms of rDNA, but also allows for quantification of fragments via
the strength of the radioactive probe, providing an estimate of how recently the
original hybridization event occurred. F i hybrids would have equally strong RFLP
signals from each parent. Repeated backcrossing to one parental species will result in
a weaker RFLP signal from the other parental species. A weaker signal would
indicate that more crosses have occurred since the original hybridization event. A
similar method has also been used to detect introgression in cottonwoods using 35
38


different labeled probes for loci known to be scattered throughout the genome
(Martinsen, 2001). Visualization of DNA fragments with radioactive probes is more
sensitive then by ethidium bromide and UV light, which is a weakness of the method
used in this research.
39


6. Future Research
Further research is required to resolve several open questions in the Fish Fry
Lakes population. First, the maternal species of the hybrids has not been determined
conclusively since in the TabC-TabF chloroplast region was identical for C.
cryptolepis and C. viridula. Amplification of an adjacent chloroplast region in these
species yielded many different sized fragments, which could not be sequenced
(unpublished data). The amplification product could be cloned into a vector thereby
separating the fragments and allowing them to be sequenced. If the sequences vary
between species, a restriction digest could be designed to test the hypothesis that C.
viridula is the maternal species for these hybrids.
Exploration of the anomalies from the Fish Fry Lakes population may be
informative as well. Sequencing of ETS and ITS sequences for the FFL02 hybrid
may elucidate the origin of the 350 base pair ETS band or the lack of a discemable
ITS CAP. Sequencing of several other individuals in the population could reveal if
ETS digestions were incomplete, or if introgression has occurred and why so many
ITS CAPs resulted in smeared bands.
Sequencing of the ITS region of several known C. bicolor and C. aurea
individuals would verify the presence of variation between these species sequences.
Additional sequencing of ETS regions may provide more variable sequences (Starr,
40


2003) that could be used to choose an enzyme that can cleave the sequences
differently and be used to test for introgression.
Because introgression was not expected in the populations of Carex section
Ceratocystis, and CAPs could not be performed in the population of Carex section
Bicolores, it is yet to be determined if the CAPs are sensitive enough to detect gene
flow between species. Further studies of populations known to contain introgressants
would indicate the efficacy of this method. The best way to confirm test this method
on known introgressants would be to create introgressants from crosses with
greenhouse plants. Though experimental crosses of Carex are reportedly difficult
(Cayouette, 1992), few have been tried and the effort would be worthwhile. The use
of known introgressants to compare CAPs of rDNA to other methods used to detect
introgression would also be extremely informative.
Another, perhaps more efficient way to detect species specific ITS or ETS
sequences is to design species-specific primers that anneal in regions where species
vary. This would negate the need for restriction digests because the presence of a
species sequence would be detected if it could be amplified from genomic DNA.
However, in order to detect two species sequences in a single individual
amplification of different sequences in the same reaction must occur with equal
efficiency, as in the CAPs method.
Finally, given that DNA can be extracted from herbarium specimens or
vouchers, future investigators can potentially use CAPs to identify hybrids and
41


species and/or screen populations that have already been collected. This data will
continue to increase understanding of evolution in the large and systematically
complex genus Carex.
42


APPENDIX A
Genomic DNA Extraction Solutions
Extraction Buffer (100 ml)
Final concentration
Ingredients for 100 ml
50 Mm Tris, pH 8
10 mM EDTA, pH 8
100 mM NaCl
1% SDS
5 ml 1 M stock
2 ml .5 M stock
2 ml 5 M stock
10 ml 10% stock
10 mM B-Mercaptoethanol 70 pi of pure reagent
Mix ingredients and add H2O to bring to final volume.
Sterile filter the solution into a labeled, clean, autoclaved glass container before
addition of B-Mercaptoethanol. (see below for sterile filtration instructions)
B-Mercaptoethanol should be added just before use. If more than one month has
passed since it was added, additional B-Mercaptoethanol should be added to a final
concentration of 10 mM. Store extraction buffer at room temperature.
.5 M EDTA pH 8 (250 ml)
EDTA powder is difficult to dissolve and raising the pH to 8 takes some time.
1. Measure approximately 150 ml of water into a beaker.
2. Weigh 46.53g of powdered EDTA (MW = 372.24 g/mole).
3. Dissolve the EDTA in the water using a magnetic stir bar and a heat setting of
2 or 3. Set the pH meter to read the pH as the EDTA is dissolving.
4. Adjust the pH with 6 N NaOH and a transfer pipette. Initially a full pipette of
NaOH can be added but as the pH approaches 8.0 add the NaOH drop wise.
5. Once the pH is about 8.0, allow any remaining EDTA powder to dissolve.
Add a little more water if the EDTA doesnt dissolve, but dont exceed a total
volume of 250 ml.
6. Double check the pH. (it doesnt have to be 8.00 but 8.OX is ideal)
7. Add more NaOH if necessary.
8. Transfer the solution to a graduated cylinder and add water to reach a total
volume of 250 ml.
9. Store in a labeled glass container at room temperature.
43


BIO 101 Solution III 1500 ml)
Final concentration Ingredients for 500 ml
3 M KOAc 147.225 g 5 M KOAc
11.5% Glacial Acetic acid 57.5 ml glacial acetic acid
Mix 150 ml H2O with glacial acetic acid.
Dissolve KOAc and bring to final volume with H2O.
Sterile filter the solution into a labeled, clean, autoclaved glass container before use.
(see below for sterile filtration instructions)
Sterile Filtration
Extraction Buffer and BIO 101 solutions should be sterile filtered before use.
1. Label an autoclaved glass bottle (large enough to hold the desired amount of
solution).
2. Remove lid and screw on the vacuum filter (funnel shaped).
3. Attach hose to filter apparatus and vacuum source.
4. Turn on vacuum and pour liquid into the filter and cover opening with the
filter lid.
5. Allow liquid to be pulled through the filter.
6. Remove filter and place cap on the bottle.
44


APPENDIX B
Loading Dye and Sodium Borate Gel Buffer
Bromophenol Blue tBPB) Loading Dve (6X) (100ml)
Final concentration
50% Glycerol
50 mM Tris, pH 7.7
5 mM EDTA
.03% BPB
Ingredients for 100 ml
50 ml of 100% Glycerol
5 ml of 1.0 M stock
1 ml of 0.5 M stock (see APPENDIX A)
.03 g BPB
Mix BPB in glycerol and then add EDTA and Tris. Add H2O to 100ml.
When using loading dye use lA the volume of sample (e.g. use 2.5ul for a lOul
sample)
50mM Sodium Borate Buffer. pH 8.5 (1 L of 10X concentration)
Dissolve 19.07 g disodium borate decahydrate (MW = 381.4 g/mole) in 900 ml.
Monitor pH while adding boric acid powder (about 8 grams) to bring to a pH of 8.5.
Boric acid dissolves slowly so allow some time for the pH readings to change. Store
in a labeled glass container at room temperature.
45


APPENDIX C
Genomic DNA Extraction Protocol
Genomic DNA Extraction from Plant Material
1. Autoclave mortars, pestles, spatulas, and scissors. Let cool before using.
2. Set a water bath to 65C and label two sets of tubes.
3. Cut (dried or fresh) leaf or flower tissue (about. 1 g but amount may vary with
age or quality of the tissue) into a ceramic mortar using a scissor.
4. Grind in liquid nitrogen until evaporated, covering the top of the mortar as
you grind.
5. Before tissue thaws add 750 pi of DNA extraction buffer with fresh 13-
Mercaptoethanol (see APPENDIX A).
6. Grind well until the mixture is homogenous and no large fragments of tissue
remain.
7. Transfer to a 1.5 ml eppendorf tube using a spatula, close the tube and vortex
for several seconds.
8. Place the tube 65C water bath for at least 10 minutes. (Tubes can remain in
the bath longer while remaining samples are to be ground)
9. Remove tubes from water bath and add 150 pi BIO 101 solution III to each
tube.
10. Invert to mix.
11. Place tubes on ice for 20 minutes.
12. While waiting, add 750 pi of 70% isopropanol to the second set of labeled
tubes.
13. Remove tubes from ice and spin in the centrifuge for 5 minutes at max speed.
14. Transfer as much supernatant (but not more than 750 pi) from each sample to
isopropanol tubes.
15. Invert to mix.
16. Spin for 2 minutes at max speed. Align tube hinges towards the outside of the
centrifuge (DNA pellets may be barely visible and aligning the tubes ensures
that the pellet is always in the same place).
17. Carefully pour off liquid making sure not to disturb the pellet
18. Carefully add 200 pi of 80% ethanol, swirl, remove liquid with pipette and
leave the pellet (a second wash with 80% EtOH may increase purity).
19. Let pellet air dry for several hours or overnight.
20. Resuspend DNA in 50-200 pi sterile H2O (volume depends on desired
concentration of DNA).
46


21. Measure DNA concentrations with a biophotometer.
47


APPENDIX D
PCR Protocol
The components of a PCR reaction are: buffer, MgCh, deoxynucleotides (dNTPs),
forward and reverse primers, Taq enzyme, DNA template, and water. Buffer and
MgCh are generally supplied with the Taq. Concentrations of reagents and DNA will
depend on the Taq and the region being amplified. Generally, DNA will have to be
diluted 1:10 or 1:100 following extraction.
If restriction digests will be performed on the PCR product, it is a good idea to make
50 pi reactions so that several digests can be performed if needed. Always make a
master mix which comprises enough of all ingredients (except DNA) for all reactions
to be performed. Dont forget to count control reactions and add one or two for
pipetting error. For example if you want to set up 10 PCR reactions, plus one
negative control, make a master mix for 12 reactions. Master mix can be pipetted
into .2 or .5 pi labeled tubes and followed by addition of DNA sample.
A negative control, which is a reaction that does not have DNA template, should
always be run. There should be no amplification in this reaction, which ensures that
your reagents are not contaminated with DNA.
A positive control, using DNA that you know should amplify can be used to ensure
reagents are good.
PCR is run on a thermocycler, which automatically adjusts the denaturation,
annealing, and elongation temperatures cyclically. The temperature and time of each
stage varies with the region to be amplified and the Taq enzyme.
All PCR products should be visualized on an ethidium bromide agarose gel. See
APPENDIX F for instructions pouring and running a gel.
48


APPENDIX E
Restriction Digest Protocol
Restriction Digests (20 pi reaction)
Final concentration
IX buffer
IX BSA (if required)
1 unit Enzyme per 60 ng/pl DNA1
30-500 ng/pl DNA2
H20
Combine reagents in a ,2pl or .5pl tube and incubate at 37C3 for 1-3 hours. A master
mix should be made if many reactions are to be prepared.
The results should be visualized on an ethidium bromide agarose gel.. See
APPENDIX F for instructions on pouring and running a gel.
Ingredients for 20pl
2 pi 1 OX buffer
.2pl 100X BSA
.5-5 units enzyme
10-15pl DNA
1 The amount of enzyme used depends both on how much DNA is being digested and
the concentration of enzyme. Enzymes come in a buffer in concentrations of units/pl
where one unit is the amount of enzyme required to digest DNA at a concentration of
20ng/pl in one hour at 37C. If the digestion time is increased, less enzyme can be
used.
2 Generally lOpl of DNA is enough but if you are using amplified DNA and PCR
yield is low, more DNA may be needed, especially to visualize bands less than 100
base pairs.
3 Most restriction enzymes are active at 37C but some require different temperatures.
Enzyme requirements and buffers/BSA are provided with the enzyme.
49


APPENDIX F
Pouring and Running an Agarose Gel
Pouring a 1% agarose gel (60 ml)
1. Measure 60 ml of IX Sodium Borate (SB) buffer1 2 3, pH 8.5 (or other gel buffer)
in a 500 ml Erlenmeyer flask.
2. Add 0.6 g of agarose..
3. Swirl to mix the agarose.
4. Put a weigh boat on top of the flask.
5. Heat in microwave until boiling. Watch carefully so that the boiling mixture
does not overflow. Use a mitt to retrieve the flask.
6. Swirl rapidly and look at the solution. You want the solution to be clear and
homogenous. You do not want to see floating particles. If there is
undissolved agarose, reheat for 30 seconds (with weigh boat lid on). Continue
until agarose is completely dissolved.
7. Let solution cool so that glass is very warm to the touch but not burning hot.
Alternatively, the flask can be placed in a 55 C water bath to cool and keep
the agarose melted. Do not let the agarose cool enough to start solidifying.
8. Set up a gel apparatus. Be sure gel tray is firmly put into the gel box and that
the gaskets are in place. Use the appropriate size and number of wells. Be
sure you have enough wells for all your samples plus markers and controls.
9. Add 10 uL of 2 mg/ml Ethidium Bromide to flask and swirl.
EtBr IS A CARCINOGEN WEAR GLOVES CLEAN UP ANY SPILLS
IMMEDIATELY FROM NOW ON TREAT THE GEL AS A
BIOHAZARD AND WEAR GLOVES
10. Pour the melted, but cooled agarose into the gel rig with well combs in place.
Let the gel solidify. It will become opaque looking (cloudy).
1 To separate smaller bands up to 2% agarose can be used.
2 For a standard gel rig 60 ml gel should be made. Larger rigs require more for the
wells to be deep enough. The amoung of agarose and ethidum bromide must be
adjusted accordingly.
3 See APPENDIX B for 10X Sodium Borate gel buffer recipe.
50


Running an Agarose Gel
1. Once the gel has solidified, remove the combs
and turn the gel so that the wells are near the
black electrode (the negative end).
2. Pour enough IX buffer (SB) to fill the side
reservoirs to about one cm above the gel. Too
much buffer over the top will result in a
longer running time.
3. Load the gel with your samples, (see loading
instructions below) +
4. Record the order of your samples in your
notebook (left to right).
5. Run the gel at 150-250 Volts1. The current should flow negative to positive
(black on top, red on bottom). DNA is negatively charged so make sure it is
set to run towards the positive electrode.
6. Be sure your samples are running the correct direction down the length of the
gel. Run the gel until the dark blue dye is about 1/2 to 2/3 down the gel from
the wells.
Loading a Gel
(You must add loading dye to samples before loading them into the gel. Loading
dye2 contains glycerol that helps the DNA sample settle into the wells so it does not
float away.)
1. Cut a small piece of parafilm and rest it on a flat area near your gel.
2. Pipette loading dye onto the parafilm, one dot per sample.
The loading dye is made up at a 6X concentration, but a IX concentration
may be exceeded. To load PCR products use 5 pi sample and 2 pi loading
dye. To load 20 pi of restriction digest reactions use 5 pi loading dye.
3. Pipette your samples into the dye spots, keeping track of the sample order.
4. Set your pipette to the total volume. Pipette up and down once to mix the dye
and the sample. Load the entire volume into the well.
1 TBE gels should not exceed 150V but SB gels can be run at higher voltage.
2 See APPENDIX B for loading dye recipe.
51


Visualizing a Gel
1. Turn off the power source, remove the gel, and place it on a tray.
2. Carry the gel on the tray to the UV light box.
3. Put on your UV glasses or face shield. Be aware of others around you, this is
a powerful UV light.
4. Turn on the UV light and view your gel.
5. If using a digital camera, take a picture and save the picture. Record the file
name of the picture in your notebook.
52


REFERENCES
Arnold M. L., B. D. B., E. A. Zimmer. 1990. Natural Hybridization between Iris
fulva and Iris hexagona: Pattern of Ribosomal DNA Variation. Evolution 44:
1512-1221.
Ball, P. W. 2002. Carex Linnaeus sect. Bicolores. In F. o. N. A. E. Committee [ed.],
Flora of North America 424-426. Oxford Univeristy Press, New York.
Catling, P. M., A. A. Reznicek, and K. Denford. 1989. Carex lacustris X C.
trichocarpa (Cyperaceae), a new natural hybrid. Canadian Journal of Botany
67: 790-795.
Cayouette, J. 1987. Carex lyngbyei excluded from the flora of eastern North
America, and taxonomic notes on related species and hybrids. Canadian
Journal of Botany 65: 1187-1198.
Cayouette, J. a. P. M. 1985. Chromosome studies on natural hybrids between
maritime species of Carex (sections Phacocystis and Cryptocarpae) in
northeastern North America and their taxonomic implications. Canadian
Journal of Botany 63: 1957-1982.
Cayouette, J. a. P. M. C. 1992. Hybridization in the Genus Carex with Special
Reference to North America. The Botanical Review 58: 351-438.
Choler, P., B. Erschbamer, A. Tribsch, L. Gielly, andP. Taberlet. 2004.
Genetic introgression as a potential to widen a species niche: Insights from
alpine Carex curvula. Proc. Natl. Acad. Sci. USA 101: 171-176.
Crins, W. J. 2002. Carex Linnaeus sect. Ceratocystis. In F. o. N. A. E. Committee
[ed.], Flora of North America 523-527. Oxford Univeristy Press, New York.
Ellstrand, N. C., R. Whitkus, andL. H. Rieseberg. 1996. Distribution of
spontaneous plant hybrids. Proc. Natl. Acad. Sci. USA 93: 5090-5093.
Ford, B. A., P. W. Ball, and K. Ritland. 1992. Genetic and macromorphologic
evidence bearing on the evolution of members of Carex section Vesicariae
(Cyperaceae) and their natural hybrids. Canadian Journal of Botany 71: 486-
500.
53


Grant, P. R., B. R. Grant, J. A. Markert, L. F. Keller, and K. Petren. 2004.
Convergent Evolution of Darwin's Finches Caused by Introgressive
Hybridization and Selection. Evolution 58: 1588-1599.
Hendrichs, M., F. Oberwinkler, D. Begerow, and R. Bauer. 2004. Carex and
subgenus Carex (Cyperaceae) A phylogenetic approach using ITS
sequences. Plant Systematics and Evolution 246: 89-107.
Jarvis, D. I. A. T. H. 1999. Wild relatives and crop cultivars: detecting natural
introgression and farmer selection of new genetic combinations in
agroecosystems. Molecular Ecology 8: S159-S173.
Klug, W. S. a. M. R. C. 2005. Conservation Genetics, Essentials of Genetics, 552-
568. Pearson Prentice Hall, Upper Saddle River.
Kukkonen, I. A. H. T. 1988. Taxonomy of Wetland Carices. Aquatic Botany 30: 5-
22.
Luceno, M. 1993. Chromosome Studies on Carex L. Section Mitratae Kukenth.
(Cyperaceae) in the Iberian Peninsula. Cytologia 58: 321-330.
Martinsen, G. D., T. G. Whitham, R. J. Turek, and P, Keim. 2001. Hybrid
Populations Selectively Filter Gene Introgression Between Species. Evolution
55: 1325-1335.
McClintock, K. A. a. M. J. W. 1994. Genetic Differentiation between Carex
lasiocarpa and C. pellita (Cyperaceae) in North America. American Journal
of Botany 81: 224-231.
Roalson E. H., J. T. C., and E. A. Friar. 2001. Phylogenetic Relationships in
Cariceae (Cyperaceae) Based on ITS (nrDNA) and tmT-L-F (cpDNA) Region
Sequences: Assessment of Subgeneric and Sectional Relationships in Carex
with Emphasis on Section Arocystis. Systematic Botany 26: 318-341.
Shaw, J., E. B. Lickey, E. E. Schilling, and R. L. Small. 2007. Comparison of
Whole Chloroplast Genome Sequences to Choose Noncoding Regions for
Phylogenetic Studies in Angiosperms: The Tortoise and the Hare III.
American Journal of Botany 94: 275-288.
Shaw, J., E. B. Lickey, J. T. Beck, S. B. Farmer, W. Liu, J. Miller, K. C. Sirpun,
C. T. Winder, E. E. Schilling, and R. L. Small. 2005. The Tortoise and the
54


Hare II: Relative Utility of 21 Noncoding Chloroplast DNA Sequences for
Phylogenetic Analysis. American Journal of Botany 92: 142-166.
Starr, J. R., S. A. Harris, and D. A. Simpson. 2003. Potential of the 5 and 3 Ends
of the Intergenic Spacer (IGS) of rDNA in the Cyperaceae: New Sequences
for Lower-Level Phylogenies in Sedges with an Example from Uncia Pers.
Int. J. Plant Sci 164: 213-227.
Strickberger, M. W. 2000a. From Races to Species, Evolution, 580-605. Jones and
Bartlett Publishers, Sudbury, Massachusetts.
_______. 2000b. Systematics and Classification, Evolution, 236-255. Jones and
Bartlett Publishers, Sudbury, Massachusetts.
Toivonen, H. 1974. Chromatographic comparison of the species of Carex section
Heleonastes and some Carex canescens hybrids in Eastern Fennoscandia.
Annales Botanci Fennici 11: 225-230.
Tyler, T. 2003. Allozyme variation in Carex sect. Digitatae Evidence of
introgression, genetic distinctiveness and evolution of taxa. Plant Systematics
and Evolution 237: 219-231.
Vanzela, A. L., A. Cuadrado, N. Jouve, M. Luceno, AND M. Guerra. 1998.
Multiple locations of the rDNA sites in holocentric chromosomes of
Rhynchospora (Cyperaceae). Chromosome Res 6: 345-349.
Waterway, M. J. 1990. Genetic Differentiation and Hybridization between Carex
gynodynama and C. mendocinensis (Cyperaceae) in California. American
Journal of Botany 77: 826-838.
_______. 1994. Evidence for the hybrid origin of Carex kinieskernii with comments on
hybridization n the genus Carex (Cyperaceae). Canadian Journal of Botany
72:260-871.
Weaver, R. F. 2005. Molecular Biology, 59-89. McGraw Hill, Boston.
Whitkus, R. 1988. Experimental Hybridizations Among Chromosome Races of
Carex pachystachya and the Related Species C. macloviana and C. preslii
(Cyperaceae). Systematic Botany 13: 146-153.
55


Williams, J. G., A. R. Kubelik, K. J. Livak, J. A. Rafalski, AND S. V. Tingey.
1990. DNA polymorphisms amplified by arbitrary primers are useful as
genetic markers. Nucleic Acids Res 18: 6531-6535.
56


Full Text

PAGE 1

HYBRIDIZATION AND SPECIES BOUNDARIES IN CAREXSECTIONS CERATOCYSIS AND BICOLORES (CYPERACEAE) by Sarah Jane Weil B.S., University of Puget Sound, 2004 A thesis submitted to the University of Colorado at Denver and Health Sciences Center in partial fulfillment of the requirements for the degree of Master of Science Biology 2007

PAGE 2

This thesis for the Master s of Science degree by Sarah Jane Weil has been approved by Leo P. Bruederle Charles A. Ferguson ( l Date

PAGE 3

Weil, Sarah J. (Master's of Science, Biology) Hybridization and Species Boundaries in Carex Sections Ceratocystis and Bicolores (Cyperaceae) Thesis directed by Associate Professor Leo P. Bruederle ABSTRACT Speciation the hallmark event of evolution, has been more frequent in certain genera than in others. Notable among these is the genus Carex which includes approximately 2000 species of sedges worldwide. Many species of Carex hybridize frequently and occasionally fertile hybrids mediate gene exchange between otherwise well-defined species, a process referred to as introgression. Hybridization and introgression have important evolutionary consequences, including increased genetic diversity within a species and/or speciation. As such accurate identification of hybrids and introgressants can inform hypotheses regarding evolution in Carex Historically, Carex species were identified and delimited morphologically in concert with other data. However because of a the highly reduced floral morphology in this genus and the potential for phenotypic plasticity to obscure morphological traits, accurate identification of species and/or hybrids generally requires additional lines of evidence. Previous investigators have used geography allozymes and cytology, among other data to differentiate species and identify hybrids. While these

PAGE 4

sources of data are valuable, they have limitations and may be inconclusive, time consuming or otherwise expensive. Herein, three populations of Carex, including species from sections Ceratocystis and Bicolores, were used to test a novel method, based on cleaved amplified polymorphisms (CAPs) of ribosomal DNA (rDNA) spacer regions, to distinguish between species and their hybrids in Carex. It is demonstrated that CAPs of rDNA can differentiate species and hybrids in Carex section Ceratocystis but not in section Bicolores. Furthermore, in Carex section Ceratocystis CAPs of a region of chloroplast DNA can determine maternity for hybrid individuals. The use of CAPs to screen two mixed populations of Carex section Ceratocystis indicated a lack of introgression which is consistent with the observation that hybrids in these populations were completely sterile. As such this research could not confirm the ability of CAPs to positively identify introgressants. A discussion of the possible limitations of CAPs and other useful molecular techniques for detecting hybrids and introgressants is included. The abstract accurately represents the content of the candidate's thesis. I recommend its publication. :.t Signed ..: Leo P. Bruederle

PAGE 5

ACKNOWLEDGEMENT I would like to thank Leo Bruederle for the opportunity to conduct this research and his continued guidance and encouragement throughout. My thanks also to Michele Engel for help with restriction digest protocols and to Lisa Johansen for frequent use of her lab space. Finally my acknowledgments to Nathan Derieg for providing sequence data and continuous advice and support.

PAGE 6

CONTENTS Figures .......................................................................................... viii Tables ............................................................................................ .ix Chapter 1. Introduction .................................................................................. 1 1.1 Species Boundaries and Hybridization ................................................... 1 1.2 Characterization of Carex Hybrids ...................................................... .3 1.3 Detecting Introgression ..................................................................... 6 1.4 Ribosomal DNA as a Genetic Marker in Carex ........................................ 8 1.5 Cleaved Amplified Polymorphisms ..................................................... 11 1.6 Model System: Carex Sections Ceratocystis and Bicolores ......................... 11 2. Hypotheses ................................................................................. 13 3. Materials and Methods .................................................................... 14 3.1 Sampling .................................................................................... 14 3.2 DNA Extraction ........................................................................... 15 3.3 Amplification ofiTS, ETS, and Chloroplast Sequences ............................. 16 3.4 Sequencing and Restriction Digests ..................................................... 17 4. Results ....................................................................................... 20 4.1 Carex cryptolepis x C. viridistellata .................................................... 20 4.2 Carex cryptolepis x C. viridula .......................................................... 24 Vl

PAGE 7

4.3 Carex bicolor x C aurea ................................................................. 30 5. Discussion .................................................................................. 31 5.1 Carex crypto/epis x C. viridistellata ................................................... .31 5.2 Carex cryptolepis x C. viridula ......................................................... .33 5.3 Carex bicolor x C. aurea ................................................................. 36 5.4 Detecting Introgression with Cleaved Amplified Polymorphisms of Ribosomal DNA .................................................... 37 6. Future Research ........................................................................... 40 Appendix A. Genomic DNA Extraction Solutions .................................................... .43 B. Loading Dye and Sodium Borate Gel Buffer ........................................... .45 C. Genomic DNA Extraction Protocol ..................................................... .46 D. PCR Protocol ............................................................................... 48 E. Restriction Digest Protocol ............................................................... .49 F. Pouring and Running an Agarose Gel ................................................... 50 References ....................................................................................... 53 Vll

PAGE 8

LIST OF FIGURES Figure 1. Angiosperm rDNA Gene Region ......................................................... 10 2. CAPs of ETS and ITS Sequences for Representative Springville Marsh Species and Hybrids .................................................. 21 3. Chloroplast CAPs of Springville Marsh Species and Hybrids ....................................................................................... 22 4. Representative Springville Marsh Individuals Screened for Introgression ............................................................................. 23 5. Representative Amplifications of ITS and ETS Regions from the Fish Fry Lakes Population ....................................................... 24 6. CAPs of ETS and ITS Sequences for Representative Fish Fry Lakes Species and Hybrids ...................................................... 26 7. Representative ETS CAPs from Fish Fry Lakes Individuals .................................................................................... 28 8. Representative ITS CAPs from Fish Fry Lakes Individuals .................................................................................... 29 9. ITS CAPs for Two Species in Carex Section Bicolores ..................................................................................... 30 Vlll

PAGE 9

LIST OF TABLES Table 1. Enzymes Used to Generate Cleaved Amplified Polymorphisms for Carex Species in Three Populations ................................................ 19 lX

PAGE 10

1. Introduction 1.1 Species Boundaries and Hybridization What defmes and delimits species? A species is commonly defmed as a group of individuals that freely exchange genetic material with each other but not with members of other such groups. Thus members of the same species can reproduce and their offspring are fertile while members of different species are genetically isolated. However many naturally occurring events confound this definition. In microorganisms gene transfer between members of different species can occur horizontally (between members of the same generation) by plasmid transfer or virally mediated DNA transfer. Asexual species which theoretically lack any exchange of genetic material seem to defy the aforementioned defmition completely (Strickberger 2000b). Of particular interest here are situations that commonly arise in plants. For example individuals that interbreed and produce fertile offspring in an experimental setting may not do so in their natural environment because of geographical (Strickberger, 2000b) or phenological (Cayouette 1992) barriers to mating. In these cases pre-zygotic barriers tentatively maintain reproductive isolation. Alternatively members of other species may hybridize freely. If hybrids are sterile genetic isolation is maintained post-zygotically (Strickberger 2000a). However partially fertile hybrids may preclude complete genetic isolation between species. The resulting complete or partial amalgamation of gene pools from different 1

PAGE 11

species has important evolutionary consequences, several of which have been documented in Carex. One possible consequence of hybridization is speciation. If fertile hybrid individuals become established through selection they may become reproductively isolated from the parental species. Within Carex section Cryptocarpae for example C. salina Wahl. and C. recta Boott are thought to be of hybrid origin (Cayouette 1985). Evidence for the hybrid origin of these species includes morphological intermediacy with respect to the putative parental species and partial sterility due to meiotic irregularities. Though C. salina and C. recta behave as stable species, the persistence of meiotic irregularities suggests that speciation is recent or in progress and it is possible that reproductive isolation from the parental species may not be complete. On the other hand hybridization can lead to speciation with immediate reproductive isolation. Allopolyploid speciation which may explain the origin of C. pediformis C.A. Mey. (Carex section Digitatae) (Tyler 2003), occurs when a hybrid undergoes genomic duplication. The polyploid hybrid has twice as many chromosomes as either parental species As such meiosis in backcrossed individuals will not occur due to the lack of homology between chromosomes of the hybrid and either parental species (Strickberger 2000a) If partially fertile hybrids are not reproductively isolated from the parental species they may backcross to one species or the other. Repeated backcrosses following a hybridization event can result in introgression whereby genes are 2

PAGE 12

transferred between two otherwise well-defined species. The evolutionary importance of introgression is that it can facilitate increased genetic variability within a species. This has been reported in subspecies curvula and rosae of Carex curvula All. (Choler, 2004). Populations of each subspecies were studied in both optimal and marginal habitats. Though these subspecies can interbreed, they normally do not. For both subspecies, the frequency of genetic markers of the other subspecies was low in populations occupying optimal habitats and significantly higher in marginal populations. In the case of the C. curvula subspecies, it seems that hybridization and gene transfer between subspecies aids in adapting marginal populations to non optimal environments. 1.2 Characterization of Carex Hybrids Given the potentially important evolutionary role of hybridization in Carex, we must consider the frequency of this phenomenon in the genus. Though hybridization is common in vascular plants, it may be frequent in some families while nearly absent from others. A survey of five major biosystematic floras revealed that hybrids are observed most often in perennial, outcrossing species that are capable of vegetative reproduction, though these characteristics are clearly not required for hybridization to occur (Ellstrand, 1996). This survey also found that members of Carex were among the most commonly hybridizing species in Cyperaceae. Carex hybrids were first reported in the rnid-1800s and by 1992 a total of253 had been reported in North America, with hybrids particularly common in sections 3

PAGE 13

Ceratocystis, Phacocystis, and Vesicariae (Kukkonen 1988; Cayouette, 1992). However, identification of Carex hybrids and putative parental species has proved to be challenging given the reduced floral morphology characteristic of this genus and the potential for phenotypic plasticity to distort morphological characteristics. As such many of the reported hybrids especially those reported early in the 20th century when scientific techniques were limited require additional study to verify their hybrid status. Identification of hybrids requires unambiguous differentiation of species. From the 1970s to the 1990s, several investigators considered species limits and/or hybridization in Carex. This involved analyses of macromorphology (Cayouette, 1985; Cayouette 1987; Catling 1989; Waterway 1990; Cayouette 1992 ; Ford 1992 ; McClintock, 1994 ; Waterway 1994) micromorphology (Cayouette 1987; Catling, 1989) geographical distribution (Cayouette, 1985 ; Cayouette, 1987; Catling 1989; Waterway 1990; Cayouette 1992; Ford 1992 ; McClintock, 1994; Waterway 1994 ; Choler, 2004) allozyme data (Waterway, 1990; Ford 1992; McClintock 1994; Waterway 1994) cytology (Cayouette 1985 ; Cayouette 1987 ; Whitkus 1988 ; Waterway 1990 ; Cayouette 1992; Lucefio 1993; McClintock, 1994; Waterway 1994), sterility (Cayouette 1985; Cayouette 1987; Catling 1989; Waterway, 1990; Cayouette, 1992; Ford 1992 ; Waterway 1994 ; Choler 2004), phenology (Cayouette 1992), and flavonoid chromatography data (Toivonen, 1974 ; Catling 1989; Cayouette 1992) 4

PAGE 14

These studies found that Carex species tend to be differentiated morphologically geographically ecologically (e.g. microhabitat) genetically (e.g. allozymes) cytologically (e.g. karyotype) and chemically (e .g., flavonoids) (Toivonen 1974 ; McClintock, 1994). On the other hand Carex hybrids have been shown to be intermediate for parental species morphological and chemical characteristics (Toivonen 1974 ; Cayouette 1985 ; Cayouette 1987 ; Catling 1989 ; Waterway 1990 ; Cayouette 1992 ; Ford 1992 ; Waterway 1994). With respect to geograph y, natural hybrids are usually found in the vicinity of one or both parents and if only one parent is present it is likely the maternal species. Ecologically hybrids often inhabit disturbed or intermediate (relative to parental habitats) sites unsuitable for either parental species (Cayouette 1992). Not surprisingly the genetic characters o f hybrids are generally a combination of parental species genetic characters. Where parental species are homozygous for different alleles at an allozyme locus F 1 hybrids will be heterozygous (Waterway 1990 ; Cayouette 1992 ; Ford 1992; Waterway 1994). Furthermore F1 hybrid karyotypes reflect parental species chromosome numbers in that the hybrid diploid chromosome number is generally the sum of the parental haploid numbers. In addition to these characteristics hybrids can be readily identified by complete or partial sterility (Cayouette 1985 ; Cayouette 1992 ; Waterway 1994). E x perimental and natural hybridizations have revealed that large differences in chromosome number between two species of C arex do not significantly reduce 5

PAGE 15

their ability to hybridize (Cayouette, 1985; Whitkus, 1988; Waterway, 1990; Cayouette, 1992; Waterway, 1994). However, F1 sterility in Carex hybrids is variable, with complete sterility generally occurring in hybrids having more distantly related parent species and/or large differences in chromosome number. Such hybrids tend to exhibit a high level of meiotic aberrations, including trivalents or tetravalents, presumably contributing to sterility (Cayouette, 1985; Waterway, 1990; Cayouette, 1992; Luceiio, 1993; Waterway, 1994). 1.3 Detecting Introgression In cases where sterility is incomplete, some hybrids may be able to self fertilize or backcross to one of the parental species, resulting in introgression (Strickberger, 2000a). Suppose that a partially fertile hybrid (species A x species B) backcrosses to one of the parental species (species B). The resulting introgressed individual would be expected to contain genetic material from both species, as in the hybrid, but there should be a higher percentage of species' B genes relative to species A genes. If this backcrossed individual again crosses with a species B individual we expect further diminution of species A genes in the resulting progeny. Repeated backcrosses to species B result in the transfer of a small amount of genetic material from species A to species B. Given that the amount of genetic material transferred is small, verifying that introgression has occurred is challenging. Theoretically it is expected that introgressed individuals will overall resemble a single species genetically and 6

PAGE 16

morphologically, differing with respect to one or more traits that more closely resemble another species. Within Carex, hypotheses of introgression have rested on tenuous morphological and allozyme evidence. It has been suggested that when an allele is found in two different species, but is relatively rare in one of them, introgression may be the source of the rare allele. That is, the allele was transferred from one species, where it is common, to relatively few individuals in another species (McClintock, 1994; Tyler, 2003). In combination with knowledge that two species hybridize and have partially fertile offspring, rare alleles may provide evidence of introgression. However, in closely related species, where hybridization and production of fertile offspring is more likely sharing of alleles could easily be the result of recent divergence from a common ancestor, not interspecific gene flow (Tyler, 2003). As such, conclusively detecting introgression requires both that the characters studied are species specific (i.e., an allele occurs in one species but not the other) and that there are populations of pure species that do not exhibit the putative introgressed characters (i.e., the rare allele). Attempts to detect introgression in other plant gener!l and some animals have made use of a variety of molecular biology techniques involving various genetic markers such as RAPDs (random amplified polymorphic DNA), AFLPs (amplified fragment length polymorphisms), microsatellites, and RFLPs (restriction fragment length polymorphisms) (Arnold M. L. 1990; Jarvis 1999; Choler, 2004; Grant, 2004). Again, these markers can only detect introgression if they are species specific. 7

PAGE 17

Furthermore, a marker must be present in the introgressed region for the gene transfer to be detected. RAPDs and AFLPs essentially amplify random regions of the genome (Williams et al. 1990 ; Klug 2005). Subsequent sequencing or visualization of fragment sizes provides many markers distributed throughout the genome increasing the chance offmding sequences (or fragments) specific to a species and detecting introgression if it has occurred. AFLPs have been used to detect introgression between subspecies of C. curvula (Choler 2004). Microsatellites which are tandemly repeated short DNA sequences have also proved to be effective markers. The repeated segments are distributed throughout the genome and the number of repeats in each segment varies among individuals. Microsatellites have been used to detect introgression in Darwin s finches (Grant, 2004). Finally RFLPs that use enzymes with cleavage sites occurring in repeated DNA regions can be used to identify introgression. If cleavage sites within the repeated region vary among species and the region is distributed throughout the genome RFLPs can identify species-specific genetic material. For example RFLPs of ribosomal DNA (rDNA) which occurs in tandem repeats throughout the genome have been used to identify introgression in Iri s (Arnold M. L., 1990). 1.4 Ribosomal DNA as a Genetic Marker in Carex Recently investigators have used ribosomal DNA (rDNA) sequences to construct phylogenies that hypothesize evolutionary relationships in C arex (Roalson E. H ., 2001 ; Starr 2003 ; Hendrichs, 2004). Two nuclear sequences have been 8

PAGE 18

particularly useful, namely internal and external transcribed spacers (ITS and ETS respectively), which are located between highly conserved ribosomal genes (Fig 1). Together, the spacers and rDNA genes are found in arrays of tandem repeats at many loci and it is estimated that these sequences are present in hundreds to thousands of copies throughout the genome. ITS and ETS sequences can be used to assess phylogenetic relationships at the species level because, although the rDNA genes are conserved, the spacer regions evolve relatively quickly, providing enough resolution to distinguish among species (Starr, 2003). Given that ITS and ETS rDNA sequences differentiate species and are distributed throughout the genome, these sequences could be of use in detecting hybrids and introgressants in Carex. Specifically, F 1 hybrids would be expected to have copies of rDNA sequences characteristic of both parental species along with other hybrid characteristics (e.g., heterozygosity for parental allozyme alleles, sterility, and morphological intermediacy). lfintrogression has occurred and rDNA is among the introgressed genetic material some rDNA characteristic of one species should be present in the genome of individuals of another species and these individuals should lack other hybrid characteristics. 9

PAGE 19

0 IGS Region ITS Region ................................................................................ ................................. I I I ) ', ,' ................... ,' ............ ', ', ', ,, ,' ............ ', ,' 21S.Fa(24bp) '',, 'zes-Fb(427bp) bp) f:TS.F 588 bp) I I NTS I ), 26S ETS 1 ETS 1f 18S IGS-R (c. 388 bp) IGS.Ra (c 384 bp) IGS.Rb (c 319 bp) Figure 1. Angiosperm rDNA Gene Region (Starr, 2003) ; 5'> TGAGTKGTA<3' (-mollf) 115-R(12bp) The 26S, 18S 5.8S and 26S' ribosomal RNA genes are interspersed with external transcribed spacers (ETS), internal transcribed spacers (ITS) and non-transcribed spacers (NTS). The ribosomal genes are highly conserved and do not vary among species The spacer regions which are non-coding vary considerably and generally differentiate species Regions of interest for this research are the ITS region and ETS If.

PAGE 20

1.5 Cleaved Amplified Polymorphisms Rather than sequencing rDNA regions for many individuals which is time consuming and costly sequences can be distinguished by cleaved amplified polymorphisms (CAPs). This method only requires sequence information for a few individuals from each species in order to determine that the sequences vary among species but not within them. Sequences are used to choose a restriction enzyme that cleaves the sequences of two species differently. Restriction endonucleases are derived from bacteria, where they degrade foreign DNA by recognizing and cleaving specific sequences (Weaver, 2005). If restriction enzyme recognition sites differ between sequences of equal size then digestion with that enzyme will yield different sized fragments. Following amplification of the rDNA regions PCR-amplified DNA is digested with the appropriate enzyme and fragments are visualized on an ethidium bromide agarose gel. Different species will yield different banding patterns providing a fast and inexpensive method to distinguish between species-specific rDNA sequences. 1.6 Model System: Carex Sections Ceratocystis and Bicolores Two sections of Carex namely sections Ceratocystis and Bicolores, were used to test several hypotheses using CAPs of rDNA spacer regions and a chloroplast region. Within Carex section Ceratocystis the rDNA sequences ETS lf(referred to as simply ETS hereafter) and ITS (Fig. 1) from all North American members have been used to construct a phylogeny for this section. The species include Carex 11

PAGE 21

cryptolepis Mack., C. viridula Michx., and the recently described C. viridistellata, among others. For these three species there is little to no variability in ETS and ITS sequences within species, but consistent differences among them (N.J. Derieg, pers. comm., 2007). Additionally, many hybrids involving species in Carex section Ceratocystis have been reported, including: C. cryptolepis x C. viridula, C. jlava L. x C. hostiana DC., C. jlava x C. viridula, and C. hostiana x C. viridula (Cayouette, 1992). Mixed populations that contain either C. cryptolepis and C. viridistellata and hybrids (Springville Marsh population) or C. cryptolepis and C. viridula and hybrids (Fish Fry Lakes population) have been collected. Hybrids in each population were identified by sterility, morphology, and allozyme data. Since these hybrids are completely sterile, it is expected that introgression is not occurring among the species in these populations. Within Carex section Bicolores ITS sequences for C. bicolor All. and C. aurea Nutt. were obtained from Genbank. Again the sequences are different for each species. Furthermore, hybrids within this section are often reported as partially fertile (Ball, 2002). A population presumably comprising C. bicolor and purported C. bicolor x C. aurea hybrids and/or introgressants has been collected (Donnelly Campground population) with a population of C. aurea observed nearby (Bruederle, pers. comm., 2007). Given the reported partial fertility of hybrids within Carex section Bicolores, this population is expected to contain introgressants. 12

PAGE 22

2. Hypotheses Analysis of cleaved amplified polymorphisms (CAPs) was used to test several hypotheses. Individuals from two populations exhibiting F 1 hybrid characters, (that is heterozygosity for parental alleles at allozyme loci sterility, and intermediate morphology) are expected to be exhibit CAPs of ITS and ETS sequences that are characteristic of both parental species. In hybrids demonstrating ITS and ETS CAPs characteristic of both parental species the maternal species can be identified using CAPs of a chloroplast DNA ( cpDNA) sequence, which is expected to be identical to the CAP of the maternal species. Mixed populations of Carex section Ceratocystis should reveal a lack of introgression in this section given that the known hybrids in these populations are completely sterile. A mixed population of Carex section Bico/ores should reveal introgression given that hybrids have been shown to be partially fertile in this section (Ball, 2002). 13

PAGE 23

3 M a terials and Methods 3.1 Sampling Populations to be screened for introgression were sampled as part of related research describing genetic diversity and structure within Carex sections Ceratocystis and Bicolores. The individuals studied from Carex section Ceratocystis include C. cryptolepis C. viridula, and the recently described C. viridistellata with various hybrid combinations, and from Carex section Bicolores, putative C. bicolor and C. bicolor x C. aurea hybrids and/or introgressants (Bruederle, pers. comm. 2007). For each individual tissue was dried and preserved in silica gel and a pressed dried voucher of the specimen was deposited at the University of Colorado in Denver. Three populations were chosen for screening based on the presence of known hybrids or suspected introgressants: 50 individuals from Springville Marsh in Ohio (40 59'22.38" N, 8323' 57.67" W) 36 individuals from Fish Fry Lakes in Minnesota (4T38'51.44" N 91 26'34.21" W) and 59 individuals from Donnelly Campground in Alaska (63 40'25.49 N 145" 52 '59.00" W). The Springville Marsh population reportedly comprises a mixture of C. cryptolepis and C. viridistellata (N. J. Derieg pers. comm. 2007), though the collection consists of 4 7 C. cryptolepis and three sterile hybrids (C. cryptolepis x C. viridistellata). Likewise the Fish Fry Lakes population comprises C. cryptolepis and C. viridula (Bruederle, pers. comm. 2007) while the collection consists of 34 C. cryptolepis and two sterile 14

PAGE 24

hybrids (C. cryptolepis x C. viridula). The composition of the Donnelly Campground population is questionable likely comprising C. bicolor C. bicolor x C. aurea hybrids and introgressants (Bruederle, pers. comm. 2007). 3.2 DNA Extraction Two methods were used to extract DNA from the specimens. Initial extractions were performed with DNeasy Plant Mini Kit (Qiagen). Briefly 10-60 mg of dried tissue was ground in liquid nitrogen and homogenized in lysis buffer containing RNase to release the contents of cells. Following centrifugation the supernatant was spun through a QIAshredder mini-column to remove cell debris. DNA from the flow through was precipitated with 95% ethanol and selectively bound to the membrane in a DNeasy mini-column. Bound DNA was washed and then eluted The majority of extractions followed a protocol using isopropanol to precipitate the DNA without the use of spin columns. 1 0-60mg of dried tissue was ground in liquid nitrogen and combined with 750 f..ll of buffer containing 50 mM Tris at pH 8 10 mM EDTA at pH 8 100 mM NaCl 1% SDS and 10 mM BMercaptoethanol to lyse cells and denature proteins. After 10 minutes at 65 C 150 f..ll of a solution of 3 M potassium acetate and 11.5% glacial acetic acid was added followed by a 20-minute incubation on ice to precipitate proteins. Samples were then centrifuged and 750 f..ll of the supernatant was added to 750 f..ll of isopropanol to precipitate DNA. After centrifugation the supernatant was removed and DNA 15

PAGE 25

precipitate washed with 80% ethanol. Following removal of the ethanol DNA pellets were air-dried overnight and resuspended in 50-200 Ill sterile water. Tomato plant DNA was extracted simultaneously with the Fish Fry Lakes and Donnelly Campground populations to control for cross contamination during the extraction process. All extractions were quantified using a biophotometer (Eppendorf) and stored at -20 C. See APPENDICES A and C for detailed instructions on extractions and preparation of required solutions. 3.3 Amplification of ITS, ETS, and Chloroplast Sequences PCR was performed using ITS5i and ITS4i primers specific for the ITS region (Roalson E H., 2 001) and ETS-F and 18S-R primers specific for ETS lf(Starr, 2003) (Fig 1 ) Reagents for the amplification were purchased from New England Biolabs. Reaction mixtures contained 5 J..ll10X PCR buffer 1.5 mM MgCh, 200 11M of each dNTP 5 11M of each primer 1-10 J..lg/ ml template DNA, .02 units /Ill Taq DNA polymerase and de-ionized water to a final volume of 50 111. Amplifications were performed in an Eppendorf Mastercycler with an initial denaturation for 2 minutes at 95 C followed by 31 cycles of denaturation at 9YC for 1 minute annealing at 5YC for 1 minute and elongation at 72 C for 1 minute with a 10 minute final extension at 72 C and 4 C holding temperature. TabC and TabF primers were used to amplify a roughly 1kb chloroplast region of hybrids (Shaw 2005 2007). Reaction conditions were identical to those for 16

PAGE 26

ITS and ETS except for a concentration of 3 mM MgCb and .4 J..LM of each primer. Amplifications were performed in an Eppendorf Mastercycler with an initial denaturation for 2 minutes at 94 C followed by 31 cycles of denaturation at 94 C for 1 minute, annealing at 55"C for 30 seconds, and elongation at 72 C for 2 minutes, with a 10 minute final extension at 72 C and 4 C holding temperature. Amplification products were visualized in a 1% agarose gel containing .33 J..Lg/ml ethidium bromide and quantified by comparison to a PCR ladder (New England Biolabs). All PCR reactions were run with a negative control lacking DNA to ensure that the PCR reagents were not contaminated. See APPENDIX D for a general PCR protocol. See APPENDICES B and F for detailed instructions on making gel buffer and pouring and running gels. 3.4 Sequencing and Restriction Digests The ITS region (Fig. 1) was sequenced for five C. cryptolepis individuals representing five populations, two C. viridistellata individuals representing two populations, and one C. viridula individual. The ETS region (Fig. 1) was sequenced for six C. cryptolepis individuals representing six populations, three C. viridistellata individuals representing three populations and two C. viridula individuals representing two populations. The chloroplast region was sequenced for one individual each of C. cryptolepis and C. viridistellata and two C. viridula individuals. All sequencing was performed by Nathan Derieg with the use of the sequencer (CEQ 8000, Beckman Coulter) at the Rocky Mountain Center for Conservation Genetics 17

PAGE 27

and Systematics at the University of Denver. ITS sequences for C. bicolor and C. aurea were obtained from Genbank with accession numbers AY278283 and AF285062, respectively. NEB cutter V2.0 (New England Biolabs) was used to determine restriction enzyme cut sites for the species specific ITS, ETS, and chloroplast sequences. Enzymes were chosen based on their ability to differentially cut the sequences of two species known to hybridize (Table 1). Digestions were performed at a final volume of20 111 with 50-100 ng/J.ll PeRamplified DNA 2 111 of 1 OX buffer (contents vary with enzyme), .1 J.lg/J.ll BSA (when required) and .5-5 units enzyme brought to volume with de-ionized water. Appropriate buffer and BSA were supplied with the enzyme from New England Biolabs. Digestions were incubated at 3TC for 2-6 hours and visualized on a 1.5% agarose gel containing .33 J.lg/ml ethidium bromide. All digests included an uncut control lacking enzyme. For the Fish Fry Lakes and Donnelly campground populations DNA extracted from tomato was also amplified and digested to check for cross-contamination during extractions. See APPENDIX D for a detailed restriction digest protocol. See APPENDICES Band F for instructions on making gel buffer and pouring and running gels. 18

PAGE 28

....... \0 Table 1. Enzymes Used to Generate Cleaved Amplified Polymorphisms for Carex Species in Three Populations Enzymes and predicted fragment sizes ofiTS, ETS, and chloroplast DNA sequences following digestion. Fragment sizes were predicted with New England Biolabs Cutter (V 2.0). Enzymes were chosen based on their ability to differentially cut the sequences of two species known or suspected to hybridize. Bccl Bee I 6saH 8saH C. aurea Uncut Size 5

PAGE 29

4. Results 4.1 Carex cryptolepis x C. viridistellata Amplifications of ETS and ITS sequences yielded DNA about 680 base pairs in length and amplification of the chloroplast region DNA about 1050 base pairs (Table 1 ). In general the ETS and chloroplast regions amplified well but amplification of the ITS region was relatively inconsistent (data not shown). Restriction digests of ETS and ITS DNA sequences for the three hybrid individuals from this population yielded a banding pattern comprising the combined patterns of both parental species (Fig 2). For example Hinfl cleaves C. cryptolepis ETS sequence into fragments 533 and 144 base pairs in size and C. viridistellata ETS sequence into fragments 319,217, and 144 base pairs in size (Table 1 and Fig 2a). The C. cryptolepis x C. viridistellata ETS CAP yields fragments 533 319 217, and 144 base pairs in size. Results from the ITS CAPs were similar demonstrating combined parental banding patterns for the three hybrids (Fig. 2b). However, the CAP of C. viridistellata ITS sequence should generate three bands as predicted by NEB cutter (Table 1) but the smallest band of91 base pairs is visible neither in the ITS CAP of C. viridistellata nor the three hybrids. Additionally the 200 base pair band specific to the ITS CAP of C. viridistellata is relatively faint in the hybrids. 20

PAGE 30

a .! .! .! 0 !(! !(! !(! .! -II> a.! l j j Cl.-Q.c .! .!! 0 2 2 !(! 4; 0 1!: 0 0 O(j "' c bp ::;; :::> (j (j (j (j 0>< 0>< ux bp ........ ..... --._.. ---II II ---II -----II 2 3 4 5 6 7 8 9 b .! g -a !(! a. .! .,, .,, .!!! -II> -II> j-Q.-Q.-Q.c .9! .9!-!!1 -!! 8 2 0 : 4; 1!: s loi 0 "" O(j "' c bp bp ::;; :::> (j (j (j (j 0>< 0>< ux ----.. ..... -------.. ------II II .--II 2 3 4 5 6 7 8 9 Figure 2. CAPs ofETS and ITS Sequences for Representative Springville Marsh Species and Hybrids Uncut C. cryptolepis ETS (a) and ITS (b) DNA is approximately 680 base pairs (bp). Digestion ofETS DNA (a) with Hinfl and ITS DNA (b) with Eco01091 yields distinct banding patterns for C. cryptolepis and C. viridistellata. Hybrids have bands present in both parental species for ETS (a) and ITS (b), though the 200 bp ITS band is weak in hybrids and the 91 bp ITS band is not visible for C. viridistellata and hybrids (see Table 1). 21

PAGE 31

The CAP banding pattern for the C. cryptolepis x C. viridistellata TabF-TabC chloroplast region ( cpDNA) contained only the bands characteristic of C. cryptolepis (Fig 3), lacking the additive pattern seen in the nuclear sequences (Fig. 2). Additionally a 1 048 base pair band, corresponding to the uncut sequence size, occurred in all of the digestions. This band was diminished but not eliminated after increased enzyme concentration and digestion time (data not shown). e -CD -CD Q. Q. Q.-Q.-.92 .92 .92 c .2 .2 .2 : o:O .2:g 0 :s! Q. -0 s .!:;; 0 0 0 > O(j bp bp ox 1048 751 ---._ .----.... -_, -.. --.. -------' .. 2 3 4 5 6 7 8 9 10 Figure 3. Chloroplast CAPs of Springville Marsh Species and Hybrids Digestion of TabCTabF cpDNA with Alul yields distinct banding patterns for C. cryptolepis and C. viridistellata. Hybrids exhibit the banding pattern characteristic of C. cryptolepis. The uncut sequence is approximately 1050 bp and a small amount of uncut DNA is present in all of the digestions. 22

PAGE 32

Finally, each of the remaining 4 7 individuals in this population exhibited ETS and ITS CAPs characteristic of C. cryptolepis (Fig. 4). ITS bands (Fig. 4b) were often faint compared to ETS bands (Fig. 4a). b 2 3 4 5 6 7 8 9 10 11 12 13 14 Figure 4. Representative Springville Marsh Individuals Screened for Introgression (a) Digestions with Hinfl of non-hybrid ETS sequence from Springville Marsh individuals 28-30 and 34-46 have a banding pattern characteristic of C. cryptolepis. Hybrid individuals 31-33 exhibit the combined banding pattern of C. cryptolepis and C. viridistellata. (b) Digestions with EcoO 1091 of non-hybrid ITS sequence from Springville Marsh individuals 9-20 have a banding pattern characteristic of C. cryptolepis. ITS bands were fainter than ETS bands. 23

PAGE 33

4.2 Carex cryptolepis x C. viridula As with the Springville Marsh population, amplification of ETS and ITS regions yielded DNA about 680 base pairs in length, with ETS yield generally higher and more consistent than ITS (Fig. 5a and b). Additionally, for one hybrid individual (FFL02), ETS amplification yielded a strong 680 base pair band and a weak 350 base pair band (Fig. 5b and c, arrows). Since the TabF-TabC chloroplast regions of C. cryptolepis and C. viridula have identical sequences (unpublished data) an enzyme that would generate different CAPs in each species could not be chosen, as such, this region was not amplified nor digested. a Cll .'3 0 N <"l .,. "' 8 ,.._ :go ..>< "' 0 0 0 0 0 iii E ..J ..J ..J ..J ..J ..J ..J 0 u. u. u. u. u. u. u. ..,c bp :::;; 1-u. u. u. u. u. u. u. z8 bp 500 677 2 3 4 5 6 7 8 9 10 b c a; N 0 ., ...J .9 .. u. Q; M ... "' 8 .... bp ::; u. "' 0 0 0 0 0 E -' -' -' -' -' -' -' bp "' 0 u. u. u. u. u. u. u. bp ::;: Iu. u. u. u. u. u. u. 500 500 677 350 2 2 3 4 5 6 7 8 9 10 Figure 5. Representative Amplifications of ITS and ETS Regions from the Fish Fry Lakes Population bp 677 350 Amplification of the ITS (a) but not ETS (b) region varied among individuals. The same regions in tomato, a positive control, amplified well. The negative controls lacked DNA template. A 350 bp amplification product was also seen only in the ETS amplification product of individual 2 from this population (arrows in (b) and (c)). 24

PAGE 34

Digests of ETS sequences for the two hybrid individuals in this population yielded banding patterns comprising a combination of the parental banding patterns (Fig. 6a and b). The 350 base pair band seen following amplification ofETS of hybrid FFL02 is still seen following digestion. Digests of ITS sequences for one of the two hybrid individuals in this population yielded a banding pattern comprising a combination of the parental banding patterns (Fig. 6c). The other hybrid individual did not yield discernable bands following digestion due to an insufficient amount of ITS DNA despite repeated amplification attempts and the use ofDMSO and betaine to increase PCR yield (data not shown). Additionally, a 75 base pair band is predicted for both C. cryptolepis and C. viridula ITS CAPs (Table 1 ). While this band is visible in C. viridula and hybrids, it is not seen in the C. cryptolepis individual in Fig. 6c, likely due to low amplification yield of this individual. However, this band was seen in all digestions of individuals from the Fish Fry Lakes population after DNA concentration was increased (Fig. 8b). 25

PAGE 35

a bp 500 300 200 c bp 500 300 200 100 )( )( .!/? .!/? .!/? a <>. <>. j-.S!., .S!., .S! o:; o:; .9 l 0 :;; :; "' o:> 0,. 0 (j(j (j(j (j (j c: :::; :::> bp b .!!1 G; ;;; bp ::!; (j 2 3 4 5 6 500 300 )( .!2 a. 200 <>. .S! .!!! 2 .9 : -::. 8 :;; :; 0 s 0 ,. 0 "' (j (j c bp :::!: (j(j :::> 2 3 4 5 Figure 6. CAPs ofETS and ITS Sequences for Representative Fish Fry Lakes Species and Hybrids In (a) and (b) digestion ofETS DNA with Msll yields distinct banding patterns for C. cryptolepis and C. viridula. Hybrid individuals exhibit both parental banding patterns. Additionally, the 350 bp amplification product is seen in individual2 ((a), lane 2, Fig. 5b and c). (c) Digestion ofiTS DNA with Bccl also yie lds distinct banding patterns for C. cryptolepis and C. viridula. One hybrid has bands present in both parental species. 26 bp 444 233

PAGE 36

ETS CAPs of the remaining 34 individuals generally yielded banding patterns characteristic of C. cryptolepis, however a faint band specific to C. viridula at 444 base pairs was often visible. Doubling the enzyme concentration and increasing digestion time to six hours diminished most of these bands (compare Fig. 7a and b). ITS CAPs of the remaining 34 individuals also resembled C. cryptolepis, though some smearing or irregular bands were seen. Reamplification followed by digestion diminished most irregularities though smearing and some unpredicted bands persisted (compare Fig. 8a and b). The ITS and ETS CAPs of tomato were different from either of the Carex species in this population and contamination from tomato to Carex or vice versa was not detected (data not shown). 27

PAGE 37

b bp 500 200 2 N ("') -l lL lL 2 ("') ("') -l lL lL 3 3 "' ("') -l lL lL 4 4 I() ("') -l lL lL 5 5 lL lL 6 6 7 -o oc:c: ::>8 7 bp Figure 7. Representative ETS CAPs from Fish Fry Lakes Individuals ETS fragments for the same individuals after 3 hours digestion and 3.75 untis Msll (a) and 6 hours digestion and 5 units Msll (b). In both digests three bands characteristic of C. cryptolepis at 240 233 and 204 bp can be seen. A 444 bp band characteristic of C. viridula is diminished in (b). 28

PAGE 38

a Q; N co 0 ("') ..,. co Q) ..,. -e 0 0 :i :i :i :i :i :;:,_ ...J ...J 0 c: bp (U u. u. u. u. u. u. u. u. <3 c: 0 bp u. u. u. u. u. u. u. u. ::>(.) 677 500 357 300 245 100 2 3 4 5 6 7 8 9 10 11 b 4i N a) 0 "' .., a) 0> .., ><.!!! 0 0 ::i .... ::i ::i .... ...J ...J -' -' bp .. u. u. u. u. u. u. u. u. bp :::!! u. u. u. u. u. u. u. u. u s 500 200 100 2 3 4 5 6 7 8 9 10 Figure 8. Representative ITS CAPs from Fish Fry Lakes Individuals (a) ITS CAPs after 1 hour digestion with 5 untis ofBcci. FFL02 and FFL24 are the two hybrids in this population. C. viridula from another population is present for comparison. (b) The results of digestions for 2 hours of the same individuals as in (a) with the same reaction conditions using freshly amplified DNA. C. cryptolepis and C. viridula have different banding patterns and the hybrid FFL24 has bands characteristic of both parental species, but all individuals exhibit some smeared bands. FFL02, the other hybrid does not yield a clear banding pattern. FFL08 has a distinct unpredicted band at about 320 bp. 29

PAGE 39

4.3 Carex bicolor x C. aurea ITS was amplified from all individuals in the Donnelly Campground population and from four C. aurea individuals collected from a site in Vermont. Again, ITS bands were approximately 680 base pairs and amplification yield was inconsistent. An enzyme was chosen to cut C. bicolor and C. aurea Genbank sequences differently. However the CAPs of both species were identical following digestion with the same enzyme (Fig. 9). As such, screening using that enzyme would be uninformative for identifying hybrids and/or introgressants. 500 300 200 2 3 4 5 6 Figure 9. ITS CAPs for Two Species in Carex Section Bicolores bp 677 429 160 Following digestion with BsaHI, C. bicolor and C. aurea yield identical banding patterns, though lanes are slightly staggered. 30

PAGE 40

5. Discussion 5.1 Carex cryptolepis x C. viridistellata CAPs of ITS and ETS sequences for C. cryptolepis and C. viridistellata were species specific and the three hybrid individuals in the Springville Marsh population had CAP banding patterns characteristic of both parental species as expected (Fig. 2). This indicates not only that hybrids have copies of both parental species rDNA, but also that both species rDNA sequences are amplified when they are combined in a single individual. Inability to see the smallest fragment (91 base pairs) in the ITS CAP of C. viridistellata and hybrids (Table 1 and Fig. 2b) is likely due to low ITS amplification yield. Additionally ITS CAPs of the hybrids exhibited a weak 200 base pair band specific to C. viridistellata. This could indicate that both species ITS regions are not amplified with equal efficiency when present in the same reaction. That is the C. cryptolepis ITS region may outcompete the C. viridistellata ITS region for primers and thus be amplified at a higher rate. In this case, relatively more C. cryptolepis ITS DNA would be present in the digestions and the C. cryptolepis bands would appear brighter. Alternatively, a longer digestion time may be required to cleave the C. viridistellata s 291 base pair band into two bands of200 and 91 base pairs (see discussion ofETS CAPs for the Fish Fry Lakes population). CAPs of the cpDNA indicate that C. cryptolepis is maternal in the hybrids. Since pollen contributes only nuclear DNA, all chloroplastic (and mitochondrial) 31

PAGE 41

DNA is inherited maternally. The hybrid cpDNA CAPs were identical to those specific for C. cryptolepis (Fig.3). This is congruent with the morphology of the two species. Since C. viridistellata has a longer perigynia beak than C. cryptolepis (unpublished observation) it is more likely that C. viridistellata pollen tube growth will be sufficient for sperm to reach a C. cryptolepis egg than vice versa (A. A. Reznicek pers. comm., 2007). Interestingly, there were persistent uncut bands in all of the digests of cpDNA. Extended digestions with increased enzyme concentration diminished but did not abolish these bands. The enzyme used in these digestions expired in 2001 and, though previously unused, may have lost some of its activity. This hypothesis is consistent with the fact that the cpDNA region amplified well and thus the concentration of DNA in the digestions was high. Alternatively, other sequences, which lack the same enzyme recognition sites, could be present. This seems unlikely given that sequences for the cpDNA were clean, and did not have ambiguous nucleotides. Sequencing was performed using DNA amplified by the same protocol used to generate the CAPs and so it is expected that the same sequences were present in the sequencing reactions as in the digestions. Furthermore, any sequences not cut by the enzyme would have to differ exactly in the cut recognition sites, which is improbable. Finally CAPs ofiTS and ETS sequences for all other Springville Marsh individuals exhibited bands specific to C. cryptolepis (Fig. 4) indicating that rDNA 32

PAGE 42

has not introgressed from C. viridistellata to C. cryptolepis Since ITS amplification yields were low and some hybrid bands were faint, ITS DNA concentration was increased (1.5 times) in digestions for all other individuals in this population. Again it is possible that C. cryptolepis ITS DNA is preferentially amplified and therefore C. viridistellata-specific bands were not visible. However, given the strength of both species ETS bands, and that ITS and ETS regions occur in tandem it is unlikely that one region has introgressed without the other. Furthermore, given that hybrids were completely sterile it was expected that introgression has not occurred in this population. 5.2 Carex cryptolepis x C. viridula In the Fish Fry Lakes population CAPs of ITS and ETS sequences also differentiated the two species identified hybrids (Fig. 5) and indicated a lack of introgression of rDNA between the species. However, there were a few unexpected results. Interestingly one individual from this population (FFL02) yielded the expected 680 base pair sequence along with an unpredicted 350 base pair sequence following amplification of the ETS region (Fig. 5b and c). This band was seen when DNA obtained from two different extraction procedures was amplified, suggesting that the DNA was not contaminated. Contamination during PCR is doubtful given that contamination was not detected from the positive control tomato DNA, or from the reagents as indicated by the negative controls It is also improbable that the 350 base pair band is due to non-specific amplification given that it occurred only in a 33

PAGE 43

single individual. Thus the most likely explanation for this small band is the presence of a truncated 350 base pair ETS sequence where regions of primer annealing have not been affected. Since this smaller ETS sequence did not amplify as well, it is either present in fewer copies than the larger ETS sequence, or has a decreased amplification efficiency, possibly due to some changes in the primer regions. Though one primer is anchored in the highly conserved 18S rDNA gene, the other primer is located in a conserved region within ETSl (Starr, 2003) (Fig. 1). It is conceivable that mutations have occurred in this primer region, since it is non-coding and changes are neutral with respect to selection. Curiously, ITS amplification was extremely weak in this same individual and discrete bands were not discemable following digestion (Fig. 8). Several repetitions of amplification and digestion yielded some bands specific to C. cryptolepis (data not shown). On the other hand, ETS CAPs for this individual contained both species' bands, verifying that the individual is a hybrid (Fig. 6a). If the 350 base pair ETS region in this individual is evidence of an evolving ETS region, it is possible that the ITS region has also sustained mutations. If changes have occurred in the primer regions, which is possible if the primers are not anchored in the rDNA genes, amplification efficiency could be affected. Since ITS was difficult to amplify in many individuals, there may be a general weak annealing of primers or formation of secondary structure in the single-stranded DNA during PCR. Though addition of DMSO (dimethylsulfoxide) and betaine seemed to increase ITS amplification yield 34

PAGE 44

for the troublesome hybrid digestions of the resulting PCR products did not yield discemable bands. New England Biolabs has reported that organic solvents including DMSO can contribute to an enzyme's star activity (non-specific cleavage). Overall digestions of ITS and ETS regions for this population were less consistent than for the Springville Marsh population. ETS CAPs of several individuals appeared to have a band characteristic of C. viridula (at 444 base pairs) but these bands diminished after increased enzyme concentration and digestion time (Fig. 7). This suggests that the enzyme cleaves the ETS sequence into 444 and 233 base pairs readily but the second cut which cleaves the 444 base pair fragment into 240 and 204 base pair fragments requires more time. Preferential cleaving by enzymes has been reported by New England Biolabs. On the other hand the C. viridula bands could represent introgression of C. viridula rDNA into C. cryptolepis. This possibility can only be excluded if longer digestions completely eliminate the C. viridula bands. The fact that ITS CAPs resembled C. cryptolepis supports the conclusion that introgression of rDNA has not occurred. Again this is consistent with complete sterility of hybrids in this population but it is possible that amplification efficiencies differ for different species ITS or ETS regions and introgression may not be detected by the CAPs. Additionally ITS CAPs for the Fish Fry Lakes population were irregular and bands were often indistinct and smeared. This could be idiosyncratic to the enzyme though starring activity has not been reported for it. Alternatively 35

PAGE 45

degradation of the DNA due to exonucleases or repeated freezing and thawing could be responsible. Reamplification did yield discernable bands following digestion, but some smearing was still visible as well as unpredicted bands (Fig. 8). Specifically, a 320 base pair band was seen in several individuals (FFL08 in Fig. 8b ). This may again be the result of preferential cleaving whereby the 320 band has not been cleaved into the predicted 75 and 245 base pair bands. Finally, as mentioned, cpDNA could not be used to determine the maternal species for the hybrids in this population. However, based on the hypothesis that the species with the shorter perigynia beak is maternal (A. A. Reznicek, pers. comm., 2007), it is suggested that C. viridula is the maternal species, given that the range of beak lengths for C. cryptolepis and C. viridula is 1.4-2.5mm and .3-1.3mm, respectively (Crins, 2002). 5.3 Carex hicolor x C. aurea The enzyme chosen to generate different CAPs for these species did not cut the sequences as expected. It may be that the difference between the CAPs of these two species is too small to detect (Table 1). On the other hand, the sequences used to choose the enzyme represent a single individual from each species and these sequences differed at only one nucleotide position (Hendrichs, 2004). As such, this difference may not be consistent and some populations of either species may have ITS sequences identical to some populations of the other species. ETS sequences, 36

PAGE 46

which tend to vary more among species (Starr, 2003), were not available for these species. 5.5 Detecting Introgression with Cleaved Amplified Polymorphisms of Ribosomal DNA Since ITS and ETS occur in tandem, they represent two different markers for the same genomic region. Thus, even though ITS amplification was variably weak in the Springville Marsh and Fish Fry Lakes populations (Fig. 5a), the efficient amplification of ETS (Fig. 5b) and relatively strong ETS hybrid banding patterns (Fig. 2a and 6a) confirm the results of the ITS CAPs in these two populations. The results are consistent with the observation that hybrids within Carex section Ceratocystis tend to be sterile (Bruederle, unpublished observation), which would prevent introgression. Given that there are many copies of rDNA sequences in the genome, it is likely that if introgression has occurred, it may involve ITS and ETS regions. However, the distribution ofrDNA repeats within the genome ofthe species studied here is currently unknown. Thus, it is still possible that other genomic regions have introgressed. Alternatively, rDNA may be a suitable marker for detecting introgression but the method of CAPs used here is not sensitive enough to identify small amounts of genetic exchange. Though there is a lack of knowledge regarding the distribution of rDNA repeats throughout the genomes of the species in Carex sections Ceratocystis and 37

PAGE 47

Bicolores it is likely that rDNA regions are adequate markers for introgressing genes. V anzela et al. (1998) used in situ hybridization to determine the location and distribution of the 18S-5.8S-26S rDNA repeats in eight species of Rhynchospora ( C yperaceae ). This study found that in all of the species rDNA repeats were localized at telomeres though sites of localization varied from 4-8 for species with 2n = 10-30 up to 30 for species with 2n =50. This variation makes it difficult to predict the number and distribution of rDNA repeats in Carex However if at least one half of the chromosomes contain rDNA repeat regions it seems likely that these would serve as a suitable marker for recent introgression. rDNA has been used to detect introgression among species of Iris using RFLPs (restriction length fragment polymorphisms). Arnold eta/. (1990) digested genomic DNA with restriction enzymes separated the fragments by gel electrophoresis transferred the separated fragments to a filter and used a 3 2P-labeled rDNA probe to detect fragment size differences. This method is based on species specific polymorphisms of rDNA but also allows for quantification of fragments via the strength of the radioactive probe providing an estimate of how recently the original hybridization event occurred. F 1 hybrids would have equally strong RFLP signals from each parent. Repeated backcrossing to one parental species will result in a weaker RFLP signal from the other parental species. A weaker signal would indicate that more crosses have occurred since the original hybridization event. A similar method has also been used to detect introgression in cottonwoods using 35 38

PAGE 48

different labeled probes for loci known to be scattered throughout the genome (Martinsen 2001). Visualization of DNA fragments with radioactive probes is more sensitive then by ethidium bromide and UV light, which is a weakness of the method used in this research. 39

PAGE 49

6. Future Research Further research is required to resolve several open questions in the Fish Fry Lakes population. First, the maternal species of the hybrids has not been determined conclusively since in the TabCTabF chloroplast region was identical for C. cryptolepis and C. viridula. Amplification of an adjacent chloroplast region in these species yielded many different sized fragments, which could not be sequenced (unpublished data). The amplification product could be cloned into a vector thereby separating the fragments and allowing them to be sequenced. If the sequences vary between species a restriction digest could be designed to test the hypothesis that C. viridula is the maternal species for these hybrids. Exploration of the anomalies from the Fish Fry Lakes population may be informative as well. Sequencing of ETS and ITS sequences for the FFL02 hybrid may elucidate the origin of the 350 base pair ETS band or the lack of a discernable ITS CAP. Sequencing of several other individuals in the population could reveal if ETS digestions were incomplete or if introgression has occurred and why so many ITS CAPs resulted in smeared bands. Sequencing of the ITS region of several known C. bicolor and C. aurea individuals would verify the presence of variation between these species sequences. Additional sequencing ofETS regions may provide more variable sequences (Starr 40

PAGE 50

2003) that could be used to choose an enzyme that can cleave the sequences differently and be used to test for introgression. Because introgression was not expected in the populations of Carex section Ceratocystis, and CAPs could not be performed in the population of Carex section Bicolores it is yet to be determined if the CAPs are sensitive enough to detect gene flow between species. Further studies of populations known to contain introgressants would indicate the efficacy of this method. The best way to confirm test this method on known introgressants would be to create introgressants from crosses with greenhouse plants. Though experimental crosses of Carex are reportedly difficult (Cayouette 1992) few have been tried and the effort would be worthwhile. The use of known introgressants to compare CAPs of rDNA to other methods used to detect introgression would also be extremely informative. Another perhaps more efficient way to detect species specific ITS or ETS sequences is to design species-specific primers that anneal in regions where species vary. This would negate the need for restriction digests because the presence of a species' sequence would be detected if it could be amplified from genomic DNA. However in order to detect two species sequences in a single individual amplification of different sequences in the same reaction must occur with equal efficiency as in the CAPs method. Finally given that DNA can be extracted from herbarium specimens or vouchers future investigators can potentially use CAPs to identify hybrids and 41

PAGE 51

species and/or screen populations that have already been collected. This data will continue to increase understanding of evolution in the large and systematically complex genus Carex. 42

PAGE 52

APPENDIX A Genomic DNA Extraction Solutions Extraction Buffer (100 ml) Final concentration 50 Mm Tris, pH 8 10 mM EDTA, pH 8 100 mMNaCl 1%SDS 10 mM B-Mercaptoethanol Ingredients for 1 00 ml 5 ml1 M stock 2 ml.5 M stock 2 ml5 M stock 10 ml10% stock 70 f.ll of pure reagent Mix ingredients and add H20 to bring to final volume. Sterile filter the solution into a labeled, clean autoclaved glass container before addition of B-Mercaptoethanol. (see below for sterile filtration instructions) B-Mercaptoethanol should be added just before use. If more than one month has passed since it was added, additional B-Mercaptoethanol should be added to a final concentration of 1 0 mM. Store extraction buffer at room temperature 5 M EDT A pH 8 (250 ml) EDT A powder is difficult to dissolve and raising the pH to 8 takes some time. 1. Measure approximately 150 ml of water into a beaker. 2. Weigh 46.53g of powdered EDTA (MW = 372.24 g/mole). 3. Dissolve the EDT A in the water using a magnetic stir bar and a heat setting of 2 or 3. Set the pH meter to read the pH as the EDTA is dissolving. 4. Adjust the pH with 6 N NaOH and a transfer pipette. Initially a full pipette of NaOH can be added but as the pH approaches 8.0 add the NaOH drop wise. 5. Once the pH is about 8.0, allow any remaining EDTA powder to dissolve. Add a little more water if the EDTA doesn't dissolve, but don't exceed a total volume of 250 ml. 6. Double check the pH. (it doesn't have to be 8.00but 8.0X is ideal) 7. Add more NaOH if necessary. 8. Transfer the solution to a graduated cylinder and add water to reach a total volume of 250 ml. 9. Store in a labeled glass container at room temperature. 43

PAGE 53

BIO 101 Solution III (500 ml) Final concentration 3 MKOAc 11.5% Glacial Acetic acid Ingredients for 500 ml 147.225 g 5 M KOAc 57.5 ml glacial acetic acid Mix 150 ml H20 with glacial acetic acid. Dissolve KOAc and bring to final volume with HzO. Sterile filter the solution into a labeled, clean, autoclaved glass container before use. (see below for sterile filtration instructions) Sterile Filtration Extraction Buffer and BIO 101 solutions should be sterile filtered before use. 1. Label an autoclaved glass bottle (large enough to hold the desired amount of solution). 2. Remove lid and screw on the vacuum filter (funnel shaped). 3. Attach hose to filter apparatus and vacuum source. 4. Turn on vacuum and pour liquid into the filter and cover opening with the filter lid. 5. Allow liquid to be pulled through the filter. 6. Remove filter and place cap on the bottle. 44

PAGE 54

APPENDIXB Loading Dye and Sodium Borate Gel Buffer Bromophenol Blue (BPB) Loading Dye (6X) (lOOml) Final concentration 50% Glycerol 50 mM Tris, pH 7.7 5 mMEDTA .03%BPB Ingredients for 1 00 rnl 50 ml of 100% Glycerol 5 rnl of 1. 0 M stock 1 rnl of0.5 M stock (see APPENDIX A) .03 gBPB Mix BPB in glycerol and then add EDT A and Tris. Add H20 to 1 OOrnl. When using loading dye use V.. the volume of sample (e.g. use 2.5ul for a 1 Oul sample) 50mM Sodium Borate Buffer. pH 8.5 (1 L of lOX concentration) Dissolve 19.07 g disodium borate decahydrate (MW = 381.4 g/mole) in 900 rnl. Monitor pH while adding boric acid powder (about 8 grams) to bring to a pH of 8.5. Boric acid dissolves slowly so allow some time for the pH readings to change. Store in a labeled glass container at room temperature. 45

PAGE 55

APPENDIX C Genomic DNA Extraction Protocol Genomic DNA Extraction from Plant Material 1. Autoclave mortars, pestles spatulas, and scissors. Let cool before using. 2. Set a water bath to 65 C and label two sets of tubes. 3. Cut (dried or fresh) leaf or flower tissue (about .1 g but amount may vary with age or quality of the tissue) into a ceramic mortar using a scissor. 4. Grind in liquid nitrogen until evaporated, covering the top of the mortar as you grind. 5. Before tissue thaws add 750 Jll of DNA extraction buffer with fresh BMercaptoethanol (see APPENDIX A). 6. Grind well until the mixture is homogenous and no large fragments of tissue remam. 7. Transfer to a 1.5 ml eppendorf tube using a spatula, close the tube and vortex for several seconds. 8. Place the tube 65 C water bath for at least 10 minutes. (Tubes can remain in the bath longer while remaining samples are to be ground) 9. Remove tubes from water bath and add 150 Jll BIO 101 solution III to each tube. 10. Invert to mix. 11. Place tubes on ice for 20 minutes. 12. While waiting, add 750 Jll of70% isopropanol to the second set of labeled tubes. 13. Remove tubes from ice and spin in the centrifuge for 5 minutes at max speed. 14. Transfer as much supernatant (but not more than 750 Jll) from each sample to isopropanol tubes. 15. Invert to mix. 16. Spin for 2 minutes at max speed. Align tube hinges towards the outside of the centrifuge (DNA pellets may be barely visible and aligning the tubes ensures that the pellet is always in the same place). 17. Carefully pour off liquid making sure not to disturb the pellet 18. Carefully add 200 Jll of 80% ethanol, swirl, remove liquid with pipette and leave the pellet (a second wash with 80% EtOH may increase purity). 19. Let pellet air dry for several hours or overnight. 20. Resuspend DNA in 50-200 Jll sterile H20 (volume depends on desired concentration of DNA). 46

PAGE 56

21. Measure DNA concentrations with a biophotometer. 47

PAGE 57

APPENDIX D PCR Protocol The components of a PCR reaction are: buffer, MgClz, deoxynucleotides (dNTPs) forward and reverse primers, Taq enzyme, DNA template, and water. Buffer and MgCh are generally supplied with the Taq. Concentrations of reagents and DNA will depend on the Taq and the region being amplified. Generally, DNA will have to be diluted 1: 10 or 1: 100 following extraction. If restriction digests will be performed on the PCR product, it is a good idea to make 50 !!1 reactions so that several digests can be performed if needed. Always make a master mix which comprises enough of all ingredients (except DNA) for all reactions to be performed. Don't forget to count control reactions and add one or two for pipetting error. For example if you want to set up 10 PCR reactions plus one negative control, make a master mix for 12 reactions. Master mix can be pipetted into .2 or .5 Ill labeled tubes and followed by addition of DNA sample. A negative control, which is a reaction that does not have DNA template, should always be run. There should be no amplification in this reaction, which ensures that your reagents are not contaminated with DNA. A positive control, using DNA that you know should amplify can be used to ensure reagents are good. PCR is run on a thermocycler, which automatically adjusts the denaturation, annealing, and elongation temperatures cyclically. The temperature and time of each stage varies with the region to be amplified and the Taq enzyme. All PCR products should be visualized on an ethidium bromide agarose gel. See APPENDIX F for instructions pouring and running a gel. 48

PAGE 58

APPENDIX E Restriction Digest Protocol Restriction Digests (20 f.tl reaction) Final concentration IX buffer 1 X BSA (if required) 1 unit Enzyme per 60 ng/f.tl DNA 1 30-500 ng/f.tl DNA2 H20 Ingredients for 20t-tl 2f.tl 1 OX buffer .2f.tl 1 OOX BSA .5-5 units enzyme 10-15!-tl DNA Combine reagents in a .2f.tl or .5f.tl tube and incubate at 3TC3 for 1-3 hours. A master mix should be made if many reactions are to be prepared. The results should be visualized on an ethidium bromide agarose gel. See APPENDIX F for instructions on pouring and running a gel. 1 The amount of enzyme used depends both on how much DNA is being digested and the concentration of enzyme. Enzymes come in a buffer in concentrations of units/f.tl where one unit is the amount of enzyme required to digest DNA at a concentration of 20ng/f.tl in one hour at 3TC. If the digestion time is increased less enzyme can be used. 2 Generally lOf.tl ofDNA is enough but if you are using amplified DNA and PCR yield is low, more DNA may be needed, especially to visualize bands less than 100 base pairs. 3 Most restriction enzymes are active at 3TC but some require different temperatures. Enzyme requirements and buffers/BSA are provided with the enzyme. 49

PAGE 59

APPENDIX F Pouring and Running an Agarose Gel Pouring a 1%1 agarose gel ( 60 mil 1. Measure 60 ml of IX Sodium Borate (SB) buffer3 pH 8.5 (or other gel buffer) in a 500 ml Erlenmeyer flask. 2. Add 0.6 g of agarose .. 3. Swirl to mix the agarose. 4. Put a weigh boat on top of the flask. 5. Heat in microwave until boiling. Watch carefully so that the boiling mixture does not overflow. Use a mitt to retrieve the flask. 6. Swirl rapidly and look at the solution. You want the solution to be clear and homogenous. You do not want to see floating particles. If there is undissolved agarose, reheat for 30 seconds (with weigh boat lid on). Continue until agarose is completely dissolved. 7. Let solution cool so that glass is very warm to the touch but not burning hot. Alternatively, the flask can be placed in a 55 C water bath to cool and keep the agarose melted. Do not let the agarose cool enough to start solidifying. 8. Set up a gel apparatus. Be sure gel tray is fmnly put into the gel box and that the gaskets are in place. Use the appropriate size and number of wells. Be sure you have enough wells for all your samples plus markers and controls. 9. Add 10 uL of 2 mg/ml Ethidiurn Bromide to flask and swirl. EtBr IS A CARCINOGENWEAR GLOVES -CLEAN UP ANY SPILLS IMMEDIATELYFROM NOW ON TREAT THE GEL AS A BIOHAZARD AND WEAR GLOVES 10. Pour the melted, but cooled agarose into the gel rig with well combs in place. Let the gel solidify. It will become opaque looking (cloudy). 1 To separate smaller bands up to 2% agarose can be used. 2 For a standard gel rig 60 ml gel should be made. Larger rigs require more for the wells to be deep enough. The amoung of agarose and ethidum bromide must be adjusted accordingly. 3 See APPENDIX B for 1 OX Sodium Borate gel buffer recipe. 50

PAGE 60

Running an Agarose Gel 1. 2. 3. 4. 5. 6. Once the gel has solidified, remove the combs and turn the gel so that the wells are near the black electrode (the negative end). Pour enough lX buffer (SB) to fill the side reservoirs to about one em above the gel. Too much buffer over the top will result in a longer running time. Load the gel with your samples. (see loading instructions below) Record the order of your samples in your notebook (left to right). -----... ... .. 5 -.... "' "' Q. Q. 5 5 .. "' "' "' + Run the gel at 150-250 Volts1 The current should flow negative to positive (black on top, red on bottom). DNA is negatively charged so make sure it is set to run towards the positive electrode. Be sure your samples are running the correct direction down the length of the gel. Run the gel until the dark blue dye is about 1/2 to 2/3 down the gel from the wells. Loading a Gel (You must add loading dye to samples before loading them into the gel. Loading dye2 contains glycerol that helps the DNA sample settle into the wells so it does not float away.) 1. Cut a small piece of parafilm and rest it on a flat area near your gel. 2. Pipette loading dye onto the parafilm, one dot per sample. The loading dye is made up at a 6X concentration, but a 1 X concentration may be exceeded. To load PCR products use 5 !J.l sample and 2 !J.lloading dye. To load 20 !J.l of restriction digest reactions use 5 !J.lloading dye. 3. Pipette your samples into the dye spots, keeping track of the sample order. 4. Set your pipette to the total volume. Pipette up and down once to mix the dye and the sample. Load the entire volume into the well. 1 TBE gels should not exceed 150V but SB gels can be run at higher voltage. 2 See APPENDIX B for loading dye recipe. 51

PAGE 61

Visualizing a Gel 1. Turn off the power source remove the gel, and place it on a tray. 2 Carry the gel on the tray to the UV light box. 3. Put on your UV glasses or face shield. Be aware of others around you, this is a powerful UV light. 4. Turn on the UV light and view your gel. 5. If using a digital camera, take a picture and save the picture. Record the file name of the picture in your notebook. 52

PAGE 62

REFERENCES ARNOLD M. L. B. D. B. E. A. ZIMMER. 1990. Natural Hybridization between Iris fulva and Iris hexagona: Pattern of Ribosomal DNA Variation. Evolution 44: 1512-1221. BALL, P. W 2002. Carex Linnaeus sect. Bico/ores. In F. o. N. A. E. Committee [ed.] Flora of North America 424-426. Oxford Univeristy Press New York. CATLING, P.M., A. A. REZNICE K AND K. D ENFORD. 1989. Carex /acustris X C. trichocarpa (Cyperaceae ) a new natural hybrid. Canadian Journal of Botan y 67: 790-795 CAYOUETTE, J. 1987. Carex lyngbyei excluded from the flora of eastern North America and taxonomic notes on related species and hybrids. Canadian Journal of Botany 65 : 1187-1198. CAYOUETTE J. A. P.M. 1985. Chromosome studies on natural hybrids between maritime species of Carex (sections Phacocystis and Cryptocarpae) in northeastern North America and their taxonomic implications. Canadian Journal of Botany 63: 1957-1982. CAYOUETTE J. A P M. C 1992. Hybridization in the Genus Carex with Special Reference to North America. The Botanical Review 58: 351-438. CHOLE R P. B. ERSCHBAMER, A. TRIBSCH, L. GIELLY, ANDP. TABERLET. 2004. Genetic introgression as a potential to widen a species niche: Insights from alpine Carex curvu/a. Proc Nat/ A cad. Sci. USA 101: 171-176. CRINS, W. J. 2002. Carex Linnaeus sect. Ceratocystis. In F. o. N. A. E. Committee [ed.] Flora ofNorth America 523-527. Oxford Univeristy Press New York. ELLSTRAND, N.C., R. WHITKU S ANDL. H. RIE S E B ERG. 1996 Distribution of spontaneous plant hybrids. Proc. Nat/. Acad. Sci. USA 93: 5090-5093 FoRD, B. A., P. W. BALL, AND K. RITLAND. 1992. Genetic and macromorphologic evidence bearing on the evolution of members of Carex section V esicariae (Cyperaceae) and their natural hybrids. Canadian Journal of Botany 71: 486500. 53

PAGE 63

GRANT, P. R., B. R. GRANT, J. A. MARKERT, L. F. KELLER, AND K. PETREN. 2004. Convergent Evolution of Darwin's Finches Caused by Introgressive Hybridization and Selection. Evolution 58: 1588-1599. HENDRICHS, M., F. 0BERWINKLER, D. BEGEROW, AND R. BAUER. 2004. Carex and subgenus Carex (Cyperaceae)-A phylogenetic approach using ITS sequences. Plant Systematics and Evolution 246: 89-107. JARVIS, D. I. A. T. H. 1999. Wild relatives and crop cultivars: detecting natural introgression and farmer selection of new genetic combinations in agroecosystems. Molecular Ecology 8: S 159-S 173. KLuG, W. S. A.M. R. C. 2005. Conservation Genetics, Essentials of Genetics, 552568. Pearson Prentice Hall, Upper Saddle River. KUKKONEN, I. A. H. T. 1988. Taxonomy of Wetland Carices. Aquatic Botany 30: 522. LUCENO, M. 1993. Chromosome Studies on Carex L. Section Mitratae Kiikenth. (Cyperaceae) in the Iberian Peninsula. Cytologia 58: 321-330. MARTINSEN, G. D., T. G. WHITHAM, R. J. TUREK, AND P, KEIM. 2001. Hybrid Populations Selectively Filter Gene Introgression Between Species. Evolution 55: 1325-1335. McCLINTOCK, K. A. A. M. J. W. 1994. Genetic Differentiation between Carex lasiocarpa and C. pellita (Cyperaceae) in North America. American Journal of Botany 81: 224-231. ROALSON E. H., J. T. C., AND E. A. FRIAR. 2001. Phylogenetic Relationships in Cariceae (Cyperaceae) Based on ITS (nrDNA) and tmT-L-F (cpDNA) Region Sequences: Assessment of Subgeneric and Sectional Relationships in Carex with Emphasis on Section Arocystis. Systematic Botany 26: 318-341. SHAw, J., E. B. LICKEY, E. E. SCHILLING, AND R. L. SMALL. 2007. Comparison of Whole Chloroplast Genome Sequences to Choose Noncoding Regions for Phylogenetic Studies in Angiosperms: The Tortoise and the Hare III. American Journal of Botany 94: 275-288. SHAW, J., E. B. LICKEY, J. T. BECK, S. B. FARMER, W. LIU, J. MILLER, K. C. SIRPUN, C. T. WINDER, E. E. SCHILLING, AND R. L. SMALL. 2005. The Tortoise and the 54

PAGE 64

Hare II: Relative Utility of21 Noncoding Chloroplast DNA Sequences for Phylogenetic Analysis. American Journal of Botany 92: 142-166. STARR, J. R. S. A. HARRJs, AND D. A. SIMPSON. 2003. Potential of the 5' and 3' Ends of the Intergenic Spacer (IGS) ofrDNA in the Cyperaceae: New Sequences for Lower-Level Phylogenies in Sedges with an Example from Uncia Pers. Int J Plant Sci 164: 213-227. STRICKBERGER, M. W. 2000a. From Races to Species, Evolution, 580-605 Jones and Bartlett Publishers, Sudbury Massachusetts. ___ 2000b. Systematics and Classification, Evolution 236-255. Jones and Bartlett Publishers, Sudbury, Massachusetts. TOIVONEN, H. 1974. Chromatographic comparison of the species of Carex section Heleonastes and some Carex canescens hybrids in Eastern Fennoscandia. Annales Botanci Fennici 11: 225-230. TYLER, T. 2003. Allozyme variation in Carex sect. Digitatae --Evidence of introgression, genetic distinctiveness and evolution of taxa. Plant Systematics and Evolution 237: 219-231. VANZELA, A. L. A. CUADRADO, N. JOUVE, M. LUCENO, AND M. GUERRA. 1998. Multiple locations of the rDNA sites in holocentric chromosomes of Rhynchospora (Cyperaceae). Chromosome Res 6: 345-349. WATERWAY, M. J. 1990. Genetic Differentiation and Hybridization between Carex gynodynama and C. mendocinensis (Cyperaceae) in California. American Journal of Botany 77: 826-838 ---. 1994. Evidence for the hybrid origin of Carex kinieskernii with comments on hybridization n the genus Carex (Cyperaceae). Canadian Journal of Botany 72: 260-871. WEAVER, R. F. 2005. Molecular Biology, 59-89 McGraw Hill, Boston. WHITKUS, R. 1988. Experimental Hybridizations Among Chromosome Races of Carex pachystachya and the Related Species C. macloviana and C. pres Iii (Cyperaceae). Systematic Botany 13: 146-153. 55

PAGE 65

WILLIAMS, J. G., A. R. KUBELIK, K. J. LIVAK, J. A. RAFALSKI, AND S. V. TINGEY. 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res 18: 6531-6535. 56