RNA SILENCING IN ARABIDOPSIS THALIANA: DEVELOPMENTAL
PROCESSES AND RESPONSES TO ABIOTIC STRESS
Robert Alan Powell
B.S., University of Colorado at Denver and Health Sciences Center, 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
This thesis for the Master of Science
Robert Alan Powell
has been approved
Lisa K. Johansen
Leo P. Bruederle
dp* ( Jsr JotH
Powell, Robert A. (Master of Science, Biology)
RNA Silencing in Arabidopsis thaliana: Developmental Processes and
Responses to Abiotic Stress
Thesis directed by Assistant Professor Lisa K. Johansen
RNA silencing in plants comprises diverse mechanisms regulating
endogenous genes, stabilizing the genome, and defending against viral
infection. Over the past ten years, many of the molecular components
involved in RNA silencing pathways have been characterized.
Furthermore, a few silencing pathways have been linked to specific
physiological processes in Arabidopsis. Two groups of proteins play
important roles in the RNA silencing pathways in plants. RNA-dependent
RNA polymerases (RDR) generate and maintain the dsRNA molecules that
trigger the silencing process. Dicer-like (DCL) enzymes cleave these
dsRNAs into small RNAs that in turn target mRNAs for degradation. The
DCL gene family in Arabidopsis comprises four members, while the RDR
gene family comprises seven. Recent studies have describes two distinct
RNA silencing pathways regulating genes associated with developmental
timing and stress response.
The research described here has focused on characterizing the
developmental and molecular phenotypes associated with loss-of-function
mutations in RDR1-6 and DCL2-4. RDR3, RDR4, and RDR5 are of special
interest because they comprise a three-gene family. In the process, we
have found that 20% of plants with a loss-of-function mutation in DCL4
(dcl4-2) show extensive purpling in rosette leaves. This purpling results
from the accumulation of flavonoid pigments in leaf tissue. Flavonoids are
a functionally diverse group of plant compounds; some of which function in
responses to stress. The purpling observed in the dcl4-2 line suggests a
role for small RNAs processed by DCL4 in regulating genes associated with
flavonoid-related responses in Arabidopsis.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Lisa K. Johansen
TABLE OF CONTENTS
1.1 Arabidopsis as a Genetic Model Organism........................1
1.2 A Brief History of Research Developments in RNA Silencing......2
1.3 The Process of RNA Silencing in Plants.........................4
1.4 Approaches to Analyzing RNA Silencing Processes in Plants.....12
1.5 Overall Research Objective....................................13
1.6 Specific Aims.................................................15
2. Materials and Methods.........................................18
3. Results and Discussion........................................21
3.1 Analysis of Developmental Timing for Silencing-Defective
3.2 Leaf Morphology Among Silencing-Defective Mutants.............22
3.3 Analysis of Small RNA Production Among Members of the RDR
and DCLGene Families..........................................24
3.4 UV-B Exposure and Detection of Anthocyanin in Plant Tissue....26
3.5 Northern Blot Detection of Small RNA-Directed Cleavage of
Flavonoid-Related Gene Transcripts............................27
3.6 RT-PCR Detection of Small RNA-Directed Cleavage of
Flavonoid-Related Gene Transcripts............................32
3.7 Phenotypic Analysis of Double and Triple Mutants..............35
LIST OF FIGURES
1. The General RNA Silencing Pathway in Plants......................5
2. RNA Silencing Pathways in Plants.................................6
3. Trans-Acting siRNA Biogenesis in Arabidopsis.....................9
4. Nat-siRNA Biogenesis in Arabidopsis.............................11
5. Phenotypes of DCL4 and RDR6 Mutant Plant Lines..................14
6. Growth Stage Progression for Mutant Plant Lines.................21
7. Leaf Elongation for Mutant Plant Lines..........................23
8. Blot Assays of Small (21 -24 nt) RNAs...........................25
9. Relative Amounts of Anthocyanin Present in DCL4 and RDR6
Mutant Plant Lines.............................................28
10. The Flavonoid Biosynthetic Pathway..............................29
11. Northern Blot Analysis of Flavonoid-Related Genes...............31
12. Schematic Representation of RT-PCR-Based Detection of
Small RNA-Directed Cleavage of mRNA............................33
13. RT-PCR of Duplicate Samples to Detect Transcript Cleavage of
SCL6-III in Col-0 Leaf and Inflorescence.......................34
14. RT-PCR to Detect Transcript Cleavage of Flavonoid-Related Gene
Transcripts in Col-0 and Four Mutant Lines.....................34
LIST OF TABLES
1. Summary of RDR and DCL functions in silencing pathways
2. Genes involved in flavonoid biosynthesis and the transcription
Factors that activate their expression............................16
1.1 Arabidopsis as a Genetic Model Organism
Arabidopsis thaliana is a member of the Brassicaceae (cabbage
family) that grows as a weed in many temperate parts of the world.
Arabidopsis exhibits the characteristics typical for any genetic model
organism. It has a short life cycle of about six weeks. Each plant produces
many flowers yielding thousands of seeds. The large number of progeny
allows for convenient large-scale mutant screens and phenotypic analysis.
Classical approaches to transmission genetics can be applied to
Arabidopsis, as plants may be manually crossed or allowed to self-fertilize.
Futhermore, it has a diploid number of chromosomes (2n = 10), which
eliminates complexities associated with polyploidy.
Most valuable is its relatively small genome size (125 Mb), compared
to other traditional plant models such as maize (2500 Mb) and wheat
(16,000 Mb). This facilitated the complete sequencing of the Arabidopsis
genome in 2000. As a result, Arabidopsis has become the prominent
model for plant molecular genetics. In recent years, enormous strides have
been made in the areas of plant structural and functional genomics, the
genetic basis of primary and secondary metabolism, and plant interaction
with pathogens and other environmental factors. One of the most intensely
investigated areas is gene regulation due to RNA silencing a collection of
processes where small, non-coding RNAs serve as negative regulators of
gene expression at transcriptional and post-transcriptional levels.
1.2 A Brief History of Research Developments in RNA Silencing
The first evidence of gene silencing in plants came in 1990 when
researchers attempted to enhance flower color in petunias by introducing a
transgene encoding chalcone synthase (CHS), a key enzyme in the
biosynthesis of anthocyanins. Anthocyanins are flavonoid compounds that
contribute to the purple coloration of plant tissues in most species.
Unexpectedly, the transgenic introduction of CHS appeared to somehow
down-regulate the expression of transgenic and endogenous CHS, blocking
the biosynthesis of anthocyanin in petunia flowers (Napoli et al, 1990). In
1997, it was revealed that aberrant double-standed RNA (dsRNA)
structures derived from the CHS transgene serve as the trigger for the
silencing process (Metzlaff et al, 1997 and Stam et al, 1997). Other
researchers demonstrated that RNA targets for silencing are not
necessarily endogenous gene transcripts. They demonstrated that
transforming plants with viral gene constructs producing dsRNA initiate
silencing of virally expressed genes, which conferred immunity to viral
infection (Waterhouse et al, 1998).
Over the past decade, many of the molecular components involved in
the silencing process have been identified, and their functional roles in
specific pathways have been characterized. One of the most important
approaches to dissecting silencing processes in Arabidopsis involves
sequence-based cloning strategies to profile genome-wide expression of
small, noncoding RNAs (Kasschau et al, 2007) This type of analysis has
allowed predictions to be made about the genomic origins of small,
noncoding RNAs and their regulatory targets. This information is
maintained in public databases such as the Arabidopsis thatiana Small RNA
Project (ASRP) database developed by the Center for Gene Research and
Biotechnology, Oregon State University, USA. The information in this
database is derived from the analysis of silencing-defective mutants and
has provided an important starting point for understanding the small-RNA
regulatory pathways in Arabidopsis. It is important to note that small-RNA
databases are expanding continuously as profiling strategies are being
applied to developmental processes and responses to environmental
factors. Small, non-coding RNAs may regulate genes in a tissue specific
manner during developmental events. Likewise, they may regulate specific
responses to stress. Like most gene regulatory processes, RNA silencing
is very dynamic. As such, the comprehensive analysis of the small-RNA
transcriptome in Arabidopsis an ongoing process.
1.3 The Process of RNA Silencing in Plants
RNA silencing involves the production of small, non-coding RNA
molecules that serve as negative regulators of gene expression; it is an
ancient gene regulatory process, as it has been observed in animals, fungi,
and plants. Figure 1 illustrates the general RNA silencing pathway. Briefly,
double-stranded RNA (dsRNA) is processed by a Dicer-like (DCL)
endonuclease to yield small dsRNAs of 21 to 24 nucleotides in length. In
several pathways, the dsRNA is synthesized by an RNA-dependent RNA
polymerase (RDR). One of the two strands from the dsRNA is incorporated
into the RNA-induced silencing complex (RISC), where it functions as a
guide for either the inhibition of translation or site specific cleavage of target
gene transcripts by an Argonaute (AGO) protein.
The divergence of several gene families has given rise to multiple
RNA silencing pathways in plants (Figure 2). One pathway results in post-
transcriptional silencing of endogenous genes by microRNAs (miRNAs).
miRNAs direct site-specific cleavage of a cognate messenger RNA
(mRNA). Many of the mRNA targets of miRNAs are the transcripts of
genes that play important roles in developmental processes. In another
pathway, short interfering RNAs (siRNAs) induce transcriptional silencing
Small RNAs (21-24 nt)
processed by Dicer-like
Incorporation of small RNAs
into the RNA-induced
silencing complex (RISC).
Small RNA/AGO-directed inhibition of
translation or cleavage of a target gene
Figure 1. The General RNA Silencing Pathway in Plants. All known
silencing pathways are triggered by dsRNA and involve the activities of
DCL endonucleases and Argonaute proteins (as a component of RISC).
by directing the methylation of endogenous genes or their related histone
proteins. One of the outcomes of this type of silencing is maintenance of
genome stability by preventing transposition. A third pathway involves
siRNAs directing the cleavage of endogenous or exogenous gene
transcripts. This pathway has been shown to silence genes that play a role
in developmental timing (Xie et al. 2005 and Falhgren et al. 2006). It has
also been shown to silence the expression of transgenes, as well as target
RNAs derived from viral replication (Baulcombe, 2004). RDRs generate
miRNA gene ncRNA gene
Si RNAs 55=}
of unrelated mRNA
Genomic sources end targets
? ? ?
Currant Opirvon m Plan! Bkrfc
Figure 2. RNA Silencing Pathways in Plants (Gendrel and Colot, 2005).
All known silencing pathways are triggered by dsRNA. The functions of the
common molecular components are described in the text.
the dsRNA precursors processed in these siRNA pathways. The DCL
gene family in Arabidopsis comprises four members, while the RDR gene
family comprises seven members. DCLs and RDRs appear to work in pairs
in distinct silencing pathways (Xie et al. 2004) (Table 1). DCL1 is required
for the biogenesis of most miRNAs, which are derived from self-
complementary precursors transcribed from a single strand at distinct,
endogenous loci. miRNAs play critical roles in regulating genes associated
with development. The remaining DCL proteins process siRNAs from
dsRNA precursors generated by RDRs. DCL3 processes RDR2-dependent
siRNAs that direct chromatin formation (Xie et al. 2004). DNA and histone
methylation directed by small RNAs involves other pathway-specific protein
components associated with chromatin formation and remodeling. The role
of the RDRs in RNA silencing is to produce dsRNAs, perhaps by using
siRNAs as primers. In this case, the synthesis of dsRNA acts as a signal
amplification step, allowing for systemic silencing of viral gene expression.
DCL4 processes a class of RDR6-dependent siRNAs known as trans-acting
siRNAs (ta-siRNAs) that regulate genes associated with developmental
timing (Xie et al. 2005). As for the remaining RDRs, RDR1 as well as
RDR6 are involved in viral defense (Yu et al. 2003). RDR7 is thought to be
nonfunctional, yet it is highly expressed in pollen. The specific functions of
RDR3, RDR4, and RDR5 remain unknown, although RDR5 is highly
expressed in seed.
Table 1. Summary of RDR and DCL functions in silencing pathways in
Pathwav Small RNA Requlatorv function
DCL1 miRNAs (21-nt) Cleavage of gene transcripts associated with developmental processes
RDR2 and DCL3 siRNAs (24-nt) Maintenance of heterochromatin by directing methylation of cytosine and histone H3
RDR1, RDR6, and DCL2 siRNAs (21-nt) Cleavage of RNAs derived from viruses and transgenes
DCL1, RDR6, and DCL4 ta-siRNAs (21-nt) Cleavage of gene transcripts associated with developmental processes
DCL2, RDR6, and DCL1 nat-siRNAs (21 nt and 24nt) Cleavage of gene transcripts associtated with salt tolerance
Most research efforts to unravel the RNA silencing process have
focused on the identification of specific miRNAs, siRNAs, and their target
gene transcripts. They have also involved characterizing small RNA
biogenesis and their regulatory functions in distinct pathways. One recently
described is the trans-acting siRNA (ta-siRNA) pathway (Xie et al. 2005)
(Figure 3). ta-siRNA biogenesis begins with miRNA-directed cleavage of a
pre-ta-SiRNA SGS3 ds pre-ta-sRNA
ta-siRNA _2]____ 21
Inacivated ta-siRNA target
Figure 3. Trans-acting siRNA Biogenesis in Arabidopsis (Xie, et al.,
2005) . In this pathway, the processing of 21-nt siRNA is phased according
to the initial site specific cleavage of pri-ta-siRNA by a miRNA. The
RDR6/DCL4 ta-siRNA pathway regulates the juvenile-to-adult transition of
primary transcript (pri-ta-siRNA) produced from one of four trans-acting
siRNA (TAS) loci. This cleavage product is then converted to a dsRNA by
RDR6. The dsRNA is then cleaved in 21-nt increments by DCL4, with its
starting point determined by the initial miRNA cleavage site. This pathway
has been linked to a specific developmental process. ta-siRNAs derived
from the TAS3 locus target auxin response factors (ARFs) (Fahlgren et al.
2006) . When the function of either DCL4 or RDR6 is lost, plants undergo
an accelerated transition from the juvenile to adult phase of vegetative
growth. This is characterized phenotypically by premature leaf elongation,
appearance of trichomes on abaxial leaf surfaces, and downward curling of
Another recently described silencing pathway involves a novel class
of siRNA, natural antisense transcript siRNAs (nat-siRNAs), which have
been shown to regulate a response to salt-stress (Borsani et al. 2005). In
this pathway, source of dsRNA is overlapping transcripts from functionally
related genes (Figure 4). DCL2 cleaves the overlapping transcripts into 24-
nt nat-siRNAs, which in turn direct phased processing of 21-nt nat-siRNAs
by DCL1. The 21 -nt nat-siRNAs target mRNAs for degradation.
Other research efforts have revealed a complex functional network for
the DCLs in Arabidopsis. DCL2 and DCL3 have been shown to process
RDR6-dependent siRNAs in the absence of DCL4. These siRNAs are 22-
24 nt in length, instead of 21 nt. It is not clear at this point if these
alternatively processed siRNAs play a functional role in targeting mRNAs.
Furthermore, DCL2 and DCL4 process RDR2-dependent siRNAs in the
absence of DCL3 (Gasciolli et al, 2005). In addition, DCL2 and DCL4 have
been shown to process siRNAs that target virus-derived RNAs (Deleris et
al, 2006). Recently, DCL4 has been found to process two miRNAs, miR822
and miR839 (Rajagopalan et al, 2006). This is intriguing because miRNA
processing is a function performed predominantly by DCL1. These two
miRNAs show tissue specific expression as miR822 was primarily
sequenced from seedlings, and miR839 from flowers. This collected body
of work has resulted in the characterization of distinct silencing pathways,
but it has also indicated functional complexities that will require further
P5CDH | Proline f
Figure 4. Nat-siRNA Biogenesis in Arabidopsis (Borsani, et al., 2005). In
this pathway, the dsRNA trigger for silencing arises from the overlapping
transcripts of functionally related genes. This particular pathway regulates
a response to salt-stress.
1.4 Approaches to Analyzing RNA-Silencing Processes in Plants
Basic molecular biology techniques are useful in studying RNA
silencing in Arabidopsis. Radioactively-labeled probes are easily prepared
from cDNA templates representing genes of interest. These probes are
then hybridized to Northern blots of low-molecular weight RNA to detect the
presence of miRNAs or siRNAs in specific tissues or under specific
environmental conditions. Likewise, hybridization to blots of high-molecular
weight RNA may be used to detect small RNA-directed cleavage of
mRNAs. This approach has proven useful to detect miRNA-directed
cleavage (Llave et al, 2002). Another approach to detect cleavage of
mRNA targets involves rapid amplification of cDNA ends (RACE). An
important advantage to using RACE is that it allows specific target
sequences to be mapped. Another technique is reverse-transcription PCR
(RT-PCR), in which dsDNA (complementary or cDNA) is amplified from
mRNA. The DNA obtained from RT-PCR can be used in a variety of
cloning applications. Quantitative RT-PCR can be used to associate
changes in gene expression with gene regulatory processes such as RNA
silencing. In this research, the potential of general RT-PCR as a method to
detect small RNA-directed cleavage of mRNAs is explored.
1.5 Overall Research Objective
Initially, this project had a relatively broad focus: to characterize
developmental and molecular phenotypes associated with loss-of-function
mutations in all known members of the Arabidopsis RDR and DCL gene
families (RDR1 through 6 and DCL2 through 4). RDR3, RDR4, and RDR5
are particularly interesting because they comprise a linked three gene
family and their exact roles in silencing remain unknown. Although
members of multi-gene families diverge and specialize in function, they may
exhibit some degree of functional redundancy. In order to address this
possibility for the RDR and DCL gene families, we have made general
phenotypic comparisons among silencing-defective mutants. A relatively
convenient phenotype to analyze is developmental timing. For example, we
followed the number of days required for each silencing-defective mutant to
reach different developmental time points. These data were statistically
analyzed to address differences in developmental timing.
In the process of this broad phenotypic analysis of silencing-defective
mutants, we observed that plants with a loss-of-function mutation in DCL4
(dcl4-2) show extensive purpling in leaf tissue from adult plants due to the
accumulation of anthocyanins (Figure 5). Anthocyanins comprise a class of
flavonoids that function to agents of pollen and seed dispersal. Flavonoids
also function in responses to stress by acting as antioxidants, protecting the
plant from radicals generated by UV and metabolic processes (Winkel-
Figure 5. Phenotypes of DCL4 and RDR6 mutant plant lines. 45 day-old
plants of Col-0 (A), dcl4-2 (B), rdr6-15 (C), and the dcl4-2/rdr6-15 double-
mutant (D). dcl4-2, rdr6-15, and the double-mutant show extensive
downward leaf curling as described previously. 20% of dcl4-2 plants show
extensive leaf purpling (left-hand arrow, B), some fail to advance to
reproductive stages of development (right-hand arrow, B). 100% of dcl4-
2/rdr6-15 double-mutant show purpling limited to the petioles of leaves (red
Other stresses that induce flavonoid production include cold, drought,
salinity, and wounding (Zhu and Sunkar, 2004; Li and Strid, 2005).
Importantly, this purpling phenotype has not been observed in any of the
other mutant plant lines studied in this project. Furthermore, a distinct
silencing pathway controlling any flavonoid-related response has yet to be
described. As such, we hypothesize that small RNAs processed by DCL4
negatively regulate genes associated with flavonoid biosynthesis.
1.6 Specific Aims
1. To explore the potential use of anthocyanin accumulation in leaf
tissue as a phenotypic marker for the analysis of stress-related responses
in Arabidopsis. Light in the UV-B range (280-320 nm) is known to induce
flavonoid biosynthesis in plants. If UV-B light can be used to induce
anthocyanin accumulation in leaf tissue, this phenotype would serve as a
controllable marker for analyzing flavonoid biosynthesis as a stress-related
response, and more importantly, the possibility of this response being
regulated by RNA silencing.
2. To identify genes involved in flavonoid biosynthesis as potential
targets for RNA silencing. Six genes encoding enzymes involved in
flavonoid biosynthesis and their associated transcription factors were
selected as potential targets (Table 3.) (Kranz et al, 1998; Mehrtens et al,
2005; Preston et al, 2004). The most commonly analyzed flavonoid-related
gene encodes chalcone synthase (CHS), which controls the first committed
step in the synthesis of all flavonoid compounds. We selected multiple
genes encoding enzymes throughout the flavonoid pathway as potential
targets for RNA silencing. We also included transcription factors that are
known to positively regulate expression of these flavonoid-related genes.
Table 2. Genes involved in flavonoid biosynthesis and the transcription
factors that activate their expression.
Gene products Transcription factor
chalcone synthase (CHS) MYB12
flavonone 3-hydroxylase (F3H)
flavonol synthase (FLS)
dihydroflavonol 4-reductase (DFR) MYB75 (PAP1)
glutathione S-transferase (GST)
anthocyanidin synthase (ANS) MYB32
3. To determine the extent of purpling in dcl2-1, dcl3-1, and dcl4-2
double and triple mutants. If DCL2, DCL3, and DCL4 have redundant
functions associated with the purpling phenotype, the additional loss-of-
function in double and triple mutants is expected to show an additive effect
on the phenotype. Double mutant plants (dcl2-1/dcl3-1, dcl2-1/dcl4-2, dcl3-
1/dcl4-2) and triple mutant plants (dcl2-1/dcl3-Vdcl4-2) were grown to
determine whether or not double and triple mutants show a frequency of
purpling greater than 20%, that observed in the single mutant.
This project involved whole-organism and molecular analysis of
silencing defective mutants in order to further characterize functions for the
RDR and DCL gene families in Arabidopsis. RDR3, RDR4, and RDR5 are
of particular interest because their roles in silencing have yet to be
identified. The purpling phenotype observed in dcl4-2 plants suggests a
role for DCL4-processed small RNAs in regulating flavonoid biosynthesis.
2. Materials and Methods
All silencing-defective mutants are T-DNA insertion lines (in a Col-0
genetic background). Homozygousity for the null alleles was confirmed by
PCR-based genotyping using primers specific for RDR and DCL genes and
the T-DNAs they contain.
Growth Conditions and Developmental Timing Analysis
All seeds were cold-synchronized at 4C for 48 hours prior to planting.
Plants for all experiments were grown on soil in 3-inch pots (4 plants per
pot), under conditions conducive to robust growth: 22-25C and 14-hour
photoperiod. Watering was carried out twice each week by lowering flats
into water (with fertilizer) for 10 to 20 minutes. Plant positions in each flat
were rotated on watering days to provide plants with equal light exposure.
Col-0 and silencing-defective mutant plants (n=80 90) were examined
every other day for growth stage progression. Statistical analysis was
performed using the data analysis feature of Excel (Microsoft).
Plants were grown under the conditions described above. Leaf
measurements for length/width ratios were taken on plants (n=40) between
days 21 and 29 of growth. All measurements were taken using a digital
caliper (Vernier), to 0.01 mm.
Northern Blot Analysis
Total RNA was extracted from leaf and inflorescence tissue in Trizol
reagent. Low molecular weight RNA was normalized to 12.5 pg/sample,
separated on 17% acrylamide-TBE-urea gels, and transferred to nylon
membranes using the semi-dry transfer technique. High molecular weight
RNA was normalized to 5 pg/sample, separated on 1.5% agarose gels, and
transferred to nylon membranes using the conventional transfer apparatus.
Two exceptions to this were RNA samples from dcl4-2 (purple leaves) and
inflorescence, which were normalized to 4 pg and 4.5 pg, respectively. Blot
hybridization was done in PerfectHyb buffer (Sigma) with 32P-random-
primed probes. Hybridization was detected using a Cyclone phosphoimager
Col-0 seedlings (15d) and adult plants (35d) were exposed to light in
the UV-B range (280-320nm) (USHIO G40T10E bulbs) for 15 minutes over
a seven day period, giving a range of exposure from 15 to 105 minutes.
Anthocyanin Extraction and Detection
Anthcyanin was extracted from whole rosettes with roots and
inflorescence stems removed (about 0.750 mg of tissue) using 1% HCL in
methanol. Anthocyanin content was determined by measuring absorbance
at 530nm and 657nm and using the equation: A530- (0.25)A657 (Mancinelli,
All reactions were performed using Superscript III One-Step RT-PCR
System (Invitrogen). Template RNA for RT-PCR was the same as that
used for Northern blots. All samples were normalized to 100 ng/ pi, 1pl of
template RNA was used in each reaction. Amplification products were run
on 1.0% agarose gels and visualized with ethidium bromide.
3. Results and Discussion
3.1 Analysis of Developmental Timing for Silencing-Defective
Plants from each mutant line were grown and the developmental
stages studied include the emergence of four rosette leaves (1.04), ten
rosette leaves (1.10), reproductive bud (5.10), and first flower (6.0) (Figure
6) (Boyes et al. 2001). 30% of rdr4-3 plants failed to progress to stages
0 10 20 30 40
Figure 6. Growth stage progression for mutant plant lines. Growth stages
comprise the appearance of: four rosette leaves (1.04), ten rosette leaves
(1.10), reproductive bud (5.10), and first flower (6.00). (n=80-90).
or 5.10. This suggests that the loss-of-function of RDR4 may influence
viability. rdr3-1 and rdr5-2 showed essentially no difference in growth stage
progression compared to Col-0. Mutant lines dcl2-1, dcl3-1, and dcl4-2
appear to take longer to reach the reproductive stage in development
(stage 5.1) and rdr6-15 appears to take longer to flower (stage 6.0)
compared to Col-0.
This developmental timing analysis was repeated to compare Col-0,
dcl4-2, rdr6-15, and the double mutant line, dcl4-2/rdr6-15. These mutant
lines were analyzed separately from the other DCL and RDR lines because
DCL4 and RDR6 are known to function together in the ta-siRNA silencing
pathway. The differences in developmental timing were tested for
significance using two-sample t-tests. The most striking differences in
timing took place at flowering. dcl4-2 takes approximately 1.7 days longer
to flower compared to Col-0 (p = 0.022). rdr6-15 and rdr6-15/dcl4-2 take
3.1 and 3.4 days longer, respectively (p < 0.0001).
3.2 Leaf Morphology Among Silencing-Defective Mutants
A specific developmental timing phenotype has been linked to a
silencing pathway. Loss-of-function mutations in DCL4 (dcl4-2) and RDR6
(rdr6-15) result in the loss of siRNAs that normally down-regulate the
expression of auxin response factors. This loss of negative control results
in an accelerated transition to the adult stage of vegetative development
and is characterized by early leaf elongation and extreme downward
curling. This specific phenotype can be used as a marker to determine
Figure 7. Leaf elongation for mutant plant lines. Leaf length/width ratios
for wild type compared to RDR mutant plant lines (A) and DCL mutant plant
lines (B). Measurements taken between days 21 and 29 (n=40).
whether or not other RDRs or DCLs play a role in this developmental
process. As expected, rdr6-15 and dcl4-2 plants demonstrated greater leaf
length/width ratios in rosette leaves 1 to 6, compared to Col-0 (Figure 7).
Leaf elongation among the remaining mutant lines was no different from
1 and 2 3and4 5 and 6
Rosette leaf position
up-regulated dramatically in purple leaf tissue from dcl4-2*; indicating that
flavonoid biosynthesis is active in these plants.
3.7 Phenotypic Analysis of Double and Triple Mutants
Functional redundancy among the DCL proteins may account for the
lack of leaf purpling observed in 80% of the dcl4-2 plants grown under
optimal conditions. If this is the case, DCL double and triple knockouts are
expected to show an additive phenotypic effect. In other words, we
expected to observe an increase in the frequency of leaf purpling
compared to the single mutant. To address this, double mutant plants
(dcl2-1/dcl3-1, dcl2-1/dcl4-2, dcl3-1/dcl4-2) and triple mutant plants
(dcl2-1/dcl3-Mdcl4-2) were grown under optimal conditions in order to make
phenotypic comparisons. The only plant showing a frequency of purpling
similar to that observed in the single mutant (20%) was dcl3-1/dcl4-2 (data
not shown). The purpling phenotype observed in these plants is assumed
to result from the DCL4 loss-of function. Unexpectedly, the other double
mutants containing dcl4-2, as well as the triple mutant, did not show leaf
purpling. This suggests that the DCLs may not play compensatory roles
with respect to regulating flavonoid biosynthesis.
Since the first evidence of RNA silencing in plants in 1990, many of the
molecular components involved in the silencing process have been
identified, and their functional roles in specific pathways have been
characterized. The sequencing of its relatively small genome has made
Arabidopsis the prominent model for plant molecular genetics. Arabidopsis
has more silencing components than any other organism; its six RDR and
four DCL genes have demonstrated complex functional relationships and
are likely to be involved in a wide range of processes.
Growth stage-based phenotypic analysis of Arabidopsis is a
convenient starting point to analyzing gene function (Boyes et al, 2001).
30% of rdr4-3 plants failed to progress to stages 1.10 or 5.10. This finding
is important because there are no published data on the functions of RDR4.
Future work should address this apparent influence of RDR4 loss-of-
function on plant viability. Both DCL4 and RDR6 appear the play a role in
the emergence of reproductive buds (stage 5.1) and first flower (stage 6.0).
The most striking differences in timing take place at flowering. dcl4-2 takes
approximately 1.7 days longer to flower compared to Col-0 and rdr6-15
and rdr6-15/dcl4-2 take 3.1 and 3.4 days longer, respectively. This is
interesting considering that DCL4 and RDR6 function in the same silencing
pathway. This also raises an interesting question: how can loss of function
mutations in either DCL4 or RDR6 accelerate one developmental event (the
transition to the adult phase), yet inhibit progression to reproductive stages
of development? One possibility is that dcl4-2 (and perhaps rdr6-15) plants
are up-regulating flavonoid biosynthesis in order to counteract the activities
of auxin in these plants. Flavonoids have been shown to be endogenous
negative regulators of auxin transport. The precise mechanism of inhibition
is not known, but it has been suggested that flavonoids bind to proteins that
control auxin efflux. Auxin transport inhibition has distinct growth
phenotypes. Seedlings grown on NPA (a known inhibitor of auxin transport)
and naringenin (a precursor to flavonoid biosynthesis) both show a
decrease in root growth and gravitropism (Brown et al, 2001). It would be
interesting to see whether or not dcl4-2, rdr6-15, or dcl4-2/rdr6-15 show a
similar phenotype in seedlings. Functional characterizations of multi-gene
families should include a known phenotype associated with the genes of
interest. In this work, leaf elongation (length/width ratio) proved to be a
useful phenotypic marker to rule-out the involvement of DCL2, DCL3, and
RDR1 through RDR5 in the juvenile-to-adult transition in development.
Phenotypic observations must be validated by molecular analysis. In this
work, Northern blots of low molecular weight RNA were used to
successfully screen for shared roles in known small RNA pathways.
The purpling phenotype observed in the DCL4 knockout is intriguing.
This phenotype shows incomplete penetrance (20%) and variable
expressivity (Figure 4). As such, the genetic basis for the favonoid
accumulation in these plants is likely to be quite complex. Future work
must address how this trait is inherited. It is possible that the dcl4-2 line
used in our research is heterozygous for a recessive allele at an additional
locus, producing the purpling phenotype when the plants self-fertilize. To
rule-out this possibility, dcl4-2 plants should be back-crossed to Col-0 and
progeny observed for the purpling phenotype over several generations.
The disappearance of purpling over subsequent generations would indicate
that a functionally unrelated locus is responsible for the phenotype.
Ideally, a phenotype of interest is inducible, controllable under
experimental conductions, and/or easy to analyze at the molecular level.
Purpling could serve as an ideal phenotypic maker for studying potential
roles for flavonoids in developmental and stress-related processes in
Arabidopsis. Specifically, anthocyanin accumulation can allow for visual
comparisons of tissue involvement. In addition, relative amounts of
anthocyanin in plant tissues is easily determined by spectrophotometry or
HPLC. In this research, light in the UV-B range was used to assess the
inducibility and controllability of the purpling phenotype. Although these
experiments did not reveal leaf purpling as either inducible or controllable
there may be detectable changes in flavonoid-related gene expression
taking place at the RNA level. This should be addressed by performing
quantitative RT-PCR methods on flavonoid-related gene transcripts. CHS
is an excellent candidate as the activity of this enzyme is the first committed
step in flavonoid biosynthesis.
If flavonoid biosynthesis is regulated by an RNA silencing mechanism,
mRNA transcripts of flavoniod-related genes should show detectable
cleavage. We have attempted to detect small RNA-directed cleavage for
nine genes in the flavonoid biosynthetic pathway by Northern blot of high
molecular weight RNA. Based on our results, it is not clear if the multiple
bands present represent small RNA-directed cleavage. The most
reasonable explanation for these results is that the probes are hybridizing to
various mRNA degradation products, and not exclusively to products of
small RNA-directed cleavage. In addition, siRNAs, especially ta-siRNAs,
may target mRNAs at multiple sites. This is expected to yield multiple,
unstable cleavage fragments that are not easily detectable by Northern
Based on our hypothesis that flavonoid-related gene expression is
down-regulated via RNA silencing, dcl4-2 plants showing leaf purpling were
expected to show a relatively large amount of intact mRNA compared to
cleavage product in Northern blots. This was not the case, as dcl4-2 plants
did not show banding patterns different from those detected in the other
plant lines for the nine flavonoid-related genes analyzed. As such, these
results fail to support our hypothesis that these flavonoid-related genes are
RNA silencing targets. Accurate detection of cleavage products can be
achieved using 5'-RACE, an RT-PCR-based method. The products of 5'-
RACE can be further cloned as sequenced. This method is routinely used
to validate mRNA targets predicted by large-scale small RNA expression
profiles; in the process cleavage sites are mapped to specific nucleotides.
We explored the potential use of general RT-PCR to detect small
RNA-directed cleavage. Using this approach, we were able to detect
cleavage of SCL6-III mRNA, a transcript known to undergo miRNA-directed
cleavage. When compared to RT-PCR of SCL6-III mRNA, the transcripts of
CHS, PAP1, and DFR do not appear to be cleaved (Figures 13 and 14).
However, these differences in amplification may be due to priming of the
PCR reactions. For example, we found that cleavage of the SCLIII-6
transcript was easier to detect with primer pairs annealing in the coding
sequence rather than the 5'- and 3'-UTRs. As such, re-amplification of
these sections with different primer pairs might confirm that differences in
amplification are in fact the result of small RNA-directed cleavage.
Furthermore, these differences in amplification need to be analyzed
quantitatively. This would allow the differentiation between regulatory
control at the transcriptional and post-transcriptional levels.
Recent research by others has produced some very exciting results.
High-throughput techniques used to identify miRNAs, siRNA, and there
targets have identified MYB90, MYB75, and MYB12 as silencing targets
(Rajagopalan et al, 2006 and Falgren et al, 2007). These transcription
factors have been shown to be up-regulated under various stress
conditions, including exposure to UV light (Kranz et al, 1998). In addition, a
loss-of-function mutation in a gene encoding the ds-RNA binding protein,
DRB4, results in extensive anthocyanin accumulation in adult plants
(Nakazawa et al, 2007). Furthermore, these researchers have shown that
DRB4 interacts with DCL4 in the ta-siRNA pathway. In light of these recent
developments, its reasonable to assume that RNA silencing regulates
flavonoid-related functions in Arabidopsis. In time, research efforts will very
likely reveal a specific pathway involving RNA silencing.
Napoli, Carolyn, Lemieux, Christine, and Richard Jorgensen. Introduction
of a Chimeric Chalcone Synthase Gene into Petunia Results in
Reversible Co-Supression of Hologous Gene in trans.
The Plant Cell 2 (1990): 279-289.
Metzlaff, M., ODell, M., Cluster, P.D., and Flavell, R.B. RNA-Mediated
RNA Degradation and Chalcone Synthase A Silencing in Petunia.
Cell 88 (1997): 845-854.
Stam, Maike, de Bruin, Rob, Kenter, Susan, van der Hoorn, Renier A.L.,
van Blokland, Rik, Mol, Joseph N.M., Kooter, Jan M. Post-
transcriptional silencing of chalcone synthase in Petunia by inverted
transgene repeats. The Plant Journal 12 (1997): 63-82.
Waterhouse, Peter M., Graham, Michael W., and Ming-Bo-Wang. Virus
resistance and gene silencing in plants can be induced by
simultaneous expression of sense and antisense RNA. Proceedings
of the National Academy of Sciences 95 (1998): 13959-13964.
Kasschau, K.D., Fahlgren, N., Chapman, E.J., Sullivan, C.M., Cumbie, J.S.,
Givan, S.A., and Carrington, J.C. Genome-Wide Profiling and
Analysis of Arabidopsis siRNAs. PLoS Biology 5(3) (2007): e57.
Xie, Z., Allen, E., Wilken, A., and Carrington, J.C. DICER-LIKE 4 functions
in trans-acting small interfering RNA biogenesis and vegetative phase
change in Arabidopsis thaliana." Proceedings of the
National Academy of Sciences 102 (2005): 2164-2175.
Fahlgren, Noah, Montgomery Taiowa A., Howell, Miya D., Allen, Edwards,
Dvorak, Sarah K., Alexander, Amanda L., and James C. Carrington.
Regulation of AUXIN RESPONSE FACTOR3 by TAS3 ta-siRNA
Affects Developmental Timing and Patterning in Arabidopsis.
Current Biology 16 (2006): 939-944
Baulcombe, David. RNA silencing in plants. Nature 431 (2004): 356-363.
Gendrel, Anne-Valerie, and Vincent Colot. Arabidopsis epigenetics: when
RNA meets chromatin. Current Opinion in Plant Biology 8 (2005):
Xie, Z., Johansen, L.K., Gustafson, A.M., Kasschau, K.D., Lellis, A.D.,
Zilberman, D., Jacobsen, S.E., and Carrington, J.C. Genetic and
functional diversification of small RNA pathways in plants. PLoS
Biology 2 (2004): 642-652.
Yu, Diqiu, Fan, Baofang, MacFarlane, Stuart A., and Zhixiang Chen.
Analysis of the Involvement of an Inducible Arabidopsis RNA-
Dependent RNA Polymerase in Antiviral Defense. Molecular Plant-
Microbe Interactions 16 (2003): 206-216.
Borsani, Omar, Zhu, Jianhua, Verslues, Paul E., Sunkar, Ramanjulu, and
Jian-Kang Zhu. Endogenous siRNAs Derived from a Pair of Natural
cis-Antisense Transcripts Regulate Salt Tolerance in Arabidopsis."
Cell 123 (2005): 1279-1291.
Gasciolli, V., Mallory, A.C., Bartel, D.P., and Vaucheret, H. Partially
redundant functions of the Arabidopsis DICER-like enzymes and a role
for DCL4 in producing trans-acting siRNAs. Current Biology
15 (2005): 1494-1500.
Deleris, Angelique, Gallego-Bartolome, Javier, Bao, Jinsong, Kasschau,
Kristiin D., Carrington, James C., and Olivier Voinnet. Hierarchical
Action and Inhibition of Plant Dicer-Like Proteins in Antiviral Defense.
Science 313 (2006): 68-71.
Rajagopalan, Ramya, Vaucheret, Herve, Trejo, Jerry, and David P. Bartel.
A diverse and evolutionarily fluid set of microRNAs in Arabidipsis
thaliana. Genes and Development 20 (2006): 3407-3425.
Llave, Cesar, Xie, Zhixin, Kasschau, Kristin D., and James C. Carrington.
Cleavage of Scarecrow-like mRNA Targets Directed by a Class
of Arabidopsis miRNA. Science 297 (2002): 2053-2056.
Winkel-Shirley, Brenda. Biosynthesis of flavonoids and effects of stress.
Current Opinion in Plant Biology 5 (2002): 218-223.
Sunkar, Ramanjulu, and Jian-Kang Zhu. Novel and Stress-Regulated
MicroRNAs and Other Small RNAs from Arabidopsis. The Plant Cell
16 (2004): 2001-2019.
Li, Shaoshan, and Ake Strid. Anthocyanin accumulation and changes in
CHS and PR-5 gene expression in Arabidopsis thaliana after removal
of inflorescence stem (decapitation). Plant Physiology and
Biochemistry 43 (2005): 521-525.
Kranz Harald D., Denekamp, Marten, Greco, Raffaella, Jin, Hailing, Leyva,
Antonio, Meissner Ruth C., Petroni, Katia, Urzainqui, Ana, Bevan,
Michael, Martin, Cathie, Smeekens, Sjef, Tonelli, Chiara, Paz-Ares,
Javier, and Bernd Weisshaar. Towards functional characterization of
the members of the R2R3-MYB gene family from Arabidopsis
thaliana. The Plant Journal 16 (1998): 263-276.
Mehrtens, Frank, Kranz, Harold, Bednarek, Pawel, and Bernd Weisshaar.
The Arabidopsis Transcription Factor MYB12 Is a Flavonol-Specific
Regulator of Phenylpropanoid Biosynthesis. Plant Physiology
138 (2005): 1083-1096.
Preston, Jeremy, Wheeler, Janet, Heazlewood, Joshua, Li, Song Feng and
Roger W. Parish. AtMYB32 is required for normal pollen development
in Arabidopsis thaliana. The Plant Journal 40 (2004): 979-995
Mancinelli A.L. Interaction between light quality and light quantity in the
photoregulation of anthocyanin production. Plant Physiology
92 (1990): 1191-1195.
Brown, Dana E., Rashotte, Aaron M., Murphy, Angus S., Normanly,
Jennifer, Tague, Brian W., Peer, Wendy A., Taiz, Lincoln, and Gloria
K. Muday. Flavonoids Act as Negative Regulators of Auxin Transport
in Vivo in Arabidopsis. Plant Physiology 126 (2001): 524-535.
Boyes, Douglas C., Sayed, Adel M., Ascenzi, Robert, McCaskill, Amy J.,
Hoffman, Neil E., and Keith R. Davis. Growth Stage-Based
Phenotypic Analysis of Arabidopsis: A Model for High Throughput
Functional Genomics in Plants. The Plant Cell 13 (2001): 1499-1510.
Fahlgren, N., Howell, M.D., Kasschau, K.D., Chapman, E.J., Sullivan, C.M.,
Cumbie, J.S., Givan, S.A., Law, T.F., Grant, S.R., Dangl, J.L., and
James C. Carrington. High-Throughput Sequencing of Arabidopsis
microRNAs: Evidence for Frequent Birth and Death of MIRNA Genes.
PLoS ONE 2(2) (2007): e219.doi:10.1371/journal.pone.0000219
Nakazawa, Yukihiro, Hiraguri, Akihiro, Moriyama, Hiromitsu, and Toshiyuki
Fukuhara. The dsRNA-binding protein DRB4 interacts with the Dicer-
like protein DCL4 in vivo and functions in the trans-acting siRNA
pathway. Plant Molecular Biology 63 (2006): 777-785.
that observed in Col-0. As such, RDR3, RDR4, and RDR5 do not appear to
play a role in this particular developmental event.
3.3 Analysis of Small RNA Production Among Members of the RDR
and DCL Gene Families
Northern blot assays were performed to compare production of
miRNAs, siRNAs, and ta-siRNAs, among the RDR and DCL knockout lines
(Figure 8). miRNAs, miR163 and miR173, were detected in all of the
knockout lines. This indicates that neither DCL2 through 4 nor RDR1
through 6 are involved in the processing of these specific miRNAs.
This is expected because these specific miRNAs are known to be
processed by DCL1. These blots confirm results described previously on
the processing of some endogenous siRNAs. dcl4-2 and rdr6-15 fail to
produce the ta-siRNAs, siR255 and siR1511 (Xie et al, 2005) (Figure 8,
blue squares). ta-siRNA, siR1511, was detected as a 22-nt small RNA
thought to be processed by either DCL2 or DCL3 in the absence of DCL4
(Gasciolli et al, 2005) (Figure 8, blue arrow). Production of an siRNA
arising from an inverted duplication, siRNA02, is reduced in rdr2-1 (Figure
8, red square). rdr2-1 and dcl3-1 fail to produce siRNA1003, which arises
from 5S rRNA loci (Xie et al, 2004) (Figure 8, green squares). rdr3-1, rdr4-
3, and rdr5-2 did not show differences in miRNA, siRNA or ta-siRNA
production, as compared to wild type, indicating that these RDRs are not
involved in the processing of these specific endogenous small RNAs.
Figure 8. Blot assays of small (21-24 nt) RNAs. dcl4-2 and rdr6-15 fail to
produce the ta-siRNAs, siR255 and siR1511 (blue squares). siR1511 was
detected as a 22-nt small RNA thought to be processed by either DCL2 or
DCL3 in the absence of DCL4 (blue arrow). Production of an siRNA arising
from an inverted duplication, siRNA02, is reduced in rdr2-1 (red square).
rdr2-1 and dcl3-1 fail to produce siRNA1003, which arises from 5S rRNA
loci (green squares). As expected, these blots show that neither DCL2
through 4 nor RDR1 through 6 are involved in the processing of these
specific miRNAs. This is expected because most miRNAs have been found
to be processed by DCL1.
3.4 UV-B Exposure and Detection of Anthocyanin in Plant Tissue
If leaf purpling can be induced experimentally by exposure to UV-B
light, it could serve as a phenotypic marker in the analysis of flavonoid-
related responses. For example, if flavonoid biosynthesis is controlled at
the posttranscriptional level by small RNAs, we expect to dectect a
decrease in the production of these small RNAs with an associated
increase in production of flavonoids. Anthocyanins are an end product of
flavonoid biosynthesis, easy to detect in plant tissue, and thus could
potentially serve as a marker.
Light in the UV-B range (280-320 nm) was used as an experimental
treatment to induce flavonoid production because it is an environmental
stressor that is very easy to control. Four groups of Col-0, dcl4-2 and rdr6-
15 seedlings and adult plants were exposed to UV-B light for 15 minutes,
30 minutes, 45 minutes, and 1 hour. Exposure periods 30 minutes and
longer proved lethal to both seedlings and adult plants. The non-lethal
exposure period (15 min.) resulted in variable leaf darkening, but failed to
induce purpling in these three lines. We hypothesized that leaf purpling is
inducible with cumulative exposure to the non-lethal dose of UV-B light. To
test this, the experiment was repeated by exposing six groups of Col-0 for
15 minutes, everyday, over a seven day period. This range of exposure, 15
minutes minimum to 105 minutes maximum, failed to induce leaf purpling.
A higher frequency of leaf darkening with longer exposure periods was
expected, but this was not the case; all of the plant groups showed
variability in leaf darkening across the range of exposure.
Anthocyanin was extracted from whole rosettes of 8-week old plants
and measured by photospectrometry. The relative amounts of anthocyanin
present in Col-0, dcl4-2, rdr6-15, and the dcl4-2/rdr6-15 double mutant are
shown in Figure 9. The dcl4-2 plants analyzed in this experiment showed
extensive purpling in the petioles and leaves, whereas the dcl4-2/rdr6-15
double mutant plants show purpling, limited to the petioles (Figure 5).
3.5 Northern Blot Detection of Small RNA-Directed Cleavage of
Flavonoid-Related Gene Transcripts
Northern blots of total RNA from replicate tissue samples of leaf from
Col-0, dcl2-1, dcl3-1, dcl4-2, and rdr6-15 were used to detect flavonoid-
related gene transcripts and RNA silencing products. If a particular gene
transcript is a target for RNA silencing, we expect to detect both the intact
mRNA and the product of cleavage, which is detectable as a stable 3'-end
of the transcript (Llave et al, 2002).
We selected multiple genes encoding enzymes throughout the
flavonoid pathway as potential targets for silencing (Table 3). mRNAs
encoding transcription factors are common silencing targets (Rajagopalan
et al, 2006).
Figure 9. Relative amounts of anthocyanin present in DCL4 and RDR6
mutant plant lines. Col-0 and rdr6-15 accumulate comparable levels of
anthocyanin. The dcl4-2 mutant shows extensive purpling in rosette leaves,
whereas the dcl4-2/rdr6-15 double mutant shows purpling limited to the
petioles (see Figure 5).
As such, we included transcription factors that are known to positively
regulate expression of these flavonoid-related genes (Figure 10).
Transcription factors and their associated structural genes are expected to
show co-expression, although transcription factors may show a lesser
degree of expression compared to the genes they regulate. Leaves from
dcl4-2 plants showing purpling provided enough RNA to analyze a single
sample normalized to 4 pg while all other samples were normalized to 5 pg.
Pigmentation for recruitment of
pollinators ana seed dispersers
Figure 10. The flavonoid biosynthetic pathway (modified from Winkel-
Shirley, 2002). Flavonoids comprise a functionally diverse group of
secondary metabolites. Three MYB transcription factors (in red) and the
structural genes they regulate (boxed) were analyzed as potential targets of
For some reason, it is difficult to obtain sufficient amounts of RNA from this
mutant. Likewise, inflorescence from dcl4-2 plants provided a relatively low
concentration of RNA, these samples were normalized to 4.5 pg. As a
transcript cleavage control, probes for SCL6-III were hybridized to RNA
from Col-0, dcl4-2, and rdr6-15 inflorescence. The SCL6-III transcript is
known to be cleaved in Col-0 inflorescence tissue yielding a detectable 3'-
end. The control blot of SCL6-III showed results consistent with transcript
cleavage; a single dark band representing intact transcript is present in all
of the leaf samples (Figure 11, bottom panel, lanes 1 through 11). In
inflorescence, two dark bands representing intact transcript and 3'-
cleavage product are present (Figure 11, bottom panel, lanes 12 through
17). As expected, CHS and F3H follow the same general pattern of
expression as MYB12, as they are expressed together in both leaf and
inflorescence. FLS does not appear to be expressed in inflorescence
despite the expression of MYB12 (Figure 11, top panel). PAP1 and DFR
remained undetectable, even after a long exposure period. PAP1 may be
difficult to detect due to a low level of expression. The lack of signal from
the DFR blot may be the result of a defective probe. GST showed a
difference in expression in leaf and inflorescence, with multiple bands
appearing for leaf samples, and single bands appearing for inflorescence.
Figure 11. Northern blot analysis of flavonoid-related genes. SCL6-III,
which is not in the pathway, served as a positive control for the detection of
transcript cleavage (red arrow). Mutiple bands for some flavonoid genes
may represent transcript cleavage (black arrows).
ANS is detectable as a single band in leaf tissue (Figure 11, bottom panel,
lanes 1 through 11), and shows down-regulated expression in inflorescence
(Figure 11, bottom panel, lanes 12 through 17). Furthermore, the pattern of
expression for ANS appears to parallel that of MYB32, the transcription
factor that positively regulates ANS.
Based on these Northern blots, it is not clear if the multiple bands
detected represent small RNA-directed cleavage. It is possible that the
gene-specific probes are hybridizing to various mRNA degradation
products, and not exclusively to products of small RNA-directed cleavage.
As such, we explored RT-PCR as a method for detecting mRNA cleavage
associated with RNA silencing.
3.6 RT-PCR Detection of Small RNA-Directed Cleavage of Flavonoid-
Related Gene Transcripts
In order to detect transcript cleavage using RT-PCR, three of the nine
flavonoid related genes were analyzed. The transcripts for these genes
were amplified in three sections, designated A, B, and C (Figure 12).
If transcript cleavage takes place within a particular section, this section
should show decreased amplification. Any amplification from this section is
assumed to reflect intact transcript present. Again, SCL6-III was used as a
positive control (Figure 13). SCL6-III is expressed in leaf and stem tissue,
but most abundantly in inflorescence. Cleavage of the SCL6-III transcript is
Figure 12. Schematic representation of RT-PCR-based detection of
small RNA-directed cleavage of mRNA. mRNAs are amplified in sections
(A, B, and C) of approximately 500 base pairs in length. Here, section B
fails to amplify as a result of cleavage of the mRNA at the indicated site.
directed by miRNA 171 in wild type inflorescence tissue. The cleavage site
for miRNA 171 has been mapped to a position between 804 and 824
nucleotides into the coding sequence of SCL6-III. This site lies within
section B of the RT-PCR-amplified transcript in this experiment. Cleavage
of the SCL6-III transcript is clearly detectable by RT-PCR.
Differences in amplification were seen from within transcripts of the
flavonoid-related genes, CHS, PAP1, and DFR (Figure 14). Sections A and
C of the CHS and DFR transcripts amplify to relatively low levels compared
to region B. The A and B sections of PAP1 appear to amplify to roughly
equal amounts. As expected, these flavonoid-related gene transcripts are
Figure 13. RT-PCR of duplicate samples to detect transcript cleavage of
SCL6-III in Col-0 inflorescence and leaf. SCL6-III served as a control for
the RT-PCR experiments, as its transcript is known to be cleaved in Col-0
CHS PAPI DPR
Figure 14. RT-PCR to detect transcript cleavage of flavonoid-related
gene transcripts in Col-0 and four mutant lines. dcl4-2* plants show
extensive leaf purpling. Gene transcripts were amplified in sections (A, B,
and C) of equal length. All RNA samples were taken from leaf tissue.