EVOLUTION & DEVELOPMENT
10:5, 555 –566 (2008)
Class 1 KNOX genes are not involved in shoot development in the moss
Physcomitrella patens but do function in sporophyte development
Keiko Sakakibara,a,1,2 Tomoaki Nishiyama,b,1 Hironori Deguchi,a and Mitsuyasu Hasebec,d,e,
a
Biological Science, Graduate School of Science, Hiroshima University, Higashi-hiroshima 739-8526, Japan
Advanced Science Research Center, Kanazawa University, Kanazawa 920-0934, Japan
c
National Institute for Basic Biology, Okazaki 444-8585, Japan
d
School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan
e
ERATO, Japan Science and Technology Agency, Okazaki 444-8585, Japan
b
Author for correspondence (email: [email protected])
1
These authors contributed equally.
2
Present address: School of Biological Sciences, Monash University, Clayton Campus, Melbourne, Victoria 3800, Australia.
SUMMARY Although the number and form of metazoan
organs are determined in the embryo, plants continuously
form organs via pluripotent stem cells contained within the
meristem. Flowering plants have an indeterminate meristem in
their diploid generation, whereas the common ancestor of land
plants is inferred to have formed an indeterminate meristem in
its haploid generation, as observed in the extant basal land
plants, bryophytes, including mosses. It is hypothesized that
the underlying gene networks for the diploid meristem were
initially present in the haploid generation of the basal land
plants and were eventually co-opted for expression in the
diploid generation. In flowering plants, the class 1
KNOTTED1-LIKE HOMEOBOX (KNOX) transcription factors
are essential for the function of the indeterminate apical
meristem. Here, we show that the class 1 KNOX orthologs
function in the diploid organ, with determinate growth in the
moss Physcomitrella patens, but do not function in the haploid
indeterminate meristem. We propose that the genetic
networks governing the indeterminate meristem in land
plants are variable, and the networks governing the diploid
indeterminate meristem with the class 1 KNOX genes likely
evolved de novo in the flowering plant lineage.
INTRODUCTION
meristem located at the tip. Mosses also have a
multicellular diploid plant body (sporophyte). Growth in
this latter body terminates after a certain period of
time because the plant lacks an indeterminate meristem
(Fig. 1). Vascular plants, including pteridophytes, gymnosperms, and flowering plants, have a diplontic life
cycle. Pteridophytes have indeterminate meristems in
both the haploid and the diploid generations, whereas
gymnosperms and flowering plants have indeterminate
meristems only in the diploid generation (Gifford and
Foster 1989).
The class 1 KNOTTED1-LIKE HOMEOBOX (KNOX)
genes are essential transcription factors for regulation of the
indeterminate shoot apical meristem in flowering plants. Class
1 KNOX genes control cell division and differentiation in
both the indeterminate and the determinate meristems by
regulating phytohormones (Sakamoto et al. 2001, 2006;
Barley and Waites 2002; Jasinski et al. 2005; Yanai et al.
2005) and metabolism (Mele et al. 2003). Gymnosperm
and pteridophyte class 1 KNOX genes are expressed at the
Plant cells cannot move within the plant body because of
the surrounding cell walls, therefore regulation of cell division
and growth constitutes a critical process during plant
development (Steeves and Sussex 1989). Most of the cells in
a terrestrial plant body are produced from the shoot
apical meristem at the shoot apex, in which a population of
indeterminate, proliferating stem cells is maintained.
However, in the extant green algae most closely related to
land plants, the charophycean green algae (Turmel et al.
2007), an indeterminate meristem is formed in the haploid
generation. Charophycean green algae have a haplontic
life cycle, and the first cell division of the zygote is meiotic.
Bryophytes, which branched from other land plants early
in their evolution, form an indeterminate meristem in
the haploid generation. Mosses are a monophyletic lineage
belonging to the bryophytes. The haploid plant body
(gametophyte) of most mosses forms a gametophore,
which resembles a shoot, with an indeterminate apical
& 2008 The Author(s)
Journal compilation & 2008 Wiley Periodicals, Inc.
555
556
EVOLUTION & DEVELOPMENT
Vol. 10, No. 5, September^October 2008
Fig. 1. Transition of meristem during the life cycle of Physcomitrella patens. P. patens forms multicellular plant bodies with a meristem in
both the haploid (1N) and diploid (2N) generations. The first cell division in a spore (A) is asymmetric, forming an apical cell and a
nonapical cell (B). The apical cell (arrow in B and arrowhead in C) continuously forms a line of nonapical cells. A new apical cell is initiated
as a side branch from a nonapical cell. Most side branch initials differentiate into protonemata (arrows in C) but approximately 5% of side
branch initials differentiate into a gametophore (D and E) with an apical cell (arrows in D and F) with three cutting faces (arrowheads in D).
(F) Longitudinal section of a gametophore tip. An egg cell in an archegonium (arrow in G) and a sperm cell dispersed from the antheridia
(arrowheads in G) come together for fertilization, and a zygote is formed in the archegonium. The first cell division of the zygote is
asymmetric, forming an apical cell (arrow in H) and a basal cell. The apical cell functions as a stem cell (arrows in I, J, and K) until it stops
cell division (L). Later, a seta meristem (brackets in K and L) is formed and the number of cells increases. Before sporogenesis begins, the
seta meristem ceases to divide (M) and spores form and mature (N). Scale bars 5 10 mm (A, B, D, H, I, and J), 50 mm (C, F, G, K, L, M, and
N), or 1 mm (E).
shoot apex. Heterologous overexpression of these genes
in Arabidopsis thaliana mimics overexpressors of native
A. thaliana class 1 KNOX genes (Sundås-Larsson et al.
1998; Harrison et al. 2005; Sano et al. 2005). This suggests
that the function of class 1 KNOX genes is conserved in the
diploid indeterminate meristem of vascular plant shoot apices.
Three class 1 KNOX genes are found in the genome of the
moss Physcomitrella patens. Two of these genes are reported
to be involved in sporophyte development (Singer and Ashton
2007), although their functional characterization during diploid plant body development, especially at the
cellular level, has not been investigated. Furthermore, the
involvement of the class 1 KNOX genes in haploid
indeterminate meristem development is still unknown because
functional characterization of all three class 1 KNOX
genes has not been undertaken. Here, we report triple
deletion of all the class 1 KNOX genes in P. patens and
show that class 1 KNOX genes function by regulating
cell division and growth in the diploid generation but not in
the haploid generation of the moss. Based on the results
obtained from our experiments, we discuss the evolution of
body plans in land plants.
MATERIALS AND METHODS
Plant materials and culture conditions
The Gransden wild-type strain of P. patens Bruch and Schimp.
subsp. patens (Ashton and Cove 1977) was cultured on BCDATG
or BCD medium at 251C under continuous light (Nishiyama et al.
2000) for observation of protonemata and gametophores. For the
observation of gametangia and sporophytes, protonemata were
transplanted onto sterile peat pellets (Jiffy-7; Jiffy Products International AS, Kristansand, Norway) and cultured for a month at
251C under continuous light. The pellets were then moved to 151C
under 8-h light and 16-h dark conditions to induce gametangia and
sporophytes.
Sequence determination of class 1 KNOX genes
Genomic sequences flanking the MKN2, MKN4, and MKN5
sequences were isolated by thermal asymmetric interlaced polymerase chain reaction (PCR) (Liu et al. 1995), and the region
containing each gene was amplified, cloned, and sequenced. The
cDNA sequences were obtained by 30 - and 5 0 -rapid amplification
of cDNA ends and reverse transcription polymerase chain reaction
(RT-PCR) using gene-specific primers.
Sakakibara et al.
Evolution of class 1 KNOX genes and shoot development
Phylogenetic analysis
Sequences for phylogenetic analyses were collected from public
databases. Sequences of P. patens and Selaginella moellendorffii were
obtained from raw whole genome shotgun sequences using the program TBLASTN and were assembled into contigs (http://moss.nib
b.ac.jp/cgi-bin/blast-assemble). The contig list (Table S1) includes
the TI numbers used for the phylogenetic analyses. The coding
regions of P. patens and S. moellendorffii sequences were inferred
using GenomeScan (http://genes.mit.edu/genomescan.html). The
putative oxidase genes from gymnosperms were obtained through
a PlantGDB BLAST search (http://www.plantgdb.org/PlantGDBcgi/blast/PlantGDBblast) for assembled EST clones of Pinus taeda.
Other deduced amino acid sequences were obtained from the nonredundant dataset at NCBI (http://www.ncbi.nlm.nih.gov/blast/)
using the program BLASTP ver. 2.0.13. The deduced amino acid
sequences were aligned using MAFFT ver. 5.734 (Katoh et al. 2005)
and then revised manually. Alignments are available on request.
After removing ambiguous regions using MacClade 4.0 (Maddison
and Maddison 2000), pair-wise maximum likelihood (ML) distances
under the JTT model (Jones et al. 1992) were calculated using
ProtML in MOLPHY ver. 2.3b3 (Adachi and Hasegawa 1996).
Starting with the neighbor-joining (NJ) tree obtained using the distances with NJDist, the ML tree under the JTT model (Jones et al.
1992) was searched by the nearest neighbor interchange algorithm
implemented in ProtML (local rearrangement search). Bootstrap
analyses were performed by preparing 1000 bootstrap datasets with
SEQBOOT in PHYLIP ver. 3.65 (Felsenstein 2005) and using the
CONSENSE. The local bootstrap probabilities were estimated with
the resampling-of-estimated-log-likelihood method (Kishino et al.
1990; Hasegawa and Kishino 1994).
Gene targeting
G418 (Nishiyama et al. 2000), hygromycin, and zeocin (Kasahara
et al. 2004) resistance cassettes on pTN182 (AB267706), pTN86
(AB267705), and p35S-zeo (EF451822) were used to replace MKN2,
MKN4, and MKN5, respectively. To obtain MKN-GUS fusion
transformants, a genomic fragment of each MKN gene, from the
middle to the last codon of the coding sequence, was inserted inframe 50 to the coding region of the b-glucuronidase (GUS) gene in
the pGUS-NPTII-2 plasmid (Sakakibara et al. 2003) or into pTN85
(AB267707), thereby creating an in-frame fusion of the MKN and
uidA genes. The genomic fragment containing the 30 -flanking region
of each MKN gene was inserted into the 30 -region of the NPTII
expression cassette of this plasmid. The primers used for plasmid
construction are shown in Table S2. Polyethylene glycol-mediated
transformation was performed as described previously (Nishiyama
et al. 2000) using 3 10 mg of a linearized plasmid. Stable transformants were screened by PCR for those having the construct integrated with homologous recombination in both 50 and 30 ends.
The candidates were further analyzed using Southern hybridization
to exclude transformants with nonhomologous or tandem-integrated transgenes using a probe on the homologous region of the
transgene (Fig. S1).
Determination of the germination rate of spores
Spores were spread on solidified BCDAT medium containing
10 mM CaCl2 (Sakakibara et al. 2001) and cultured at 251C under
557
continuous light. Five days after sowing, the numbers of
germinated and ungerminated spores were counted under a
stereomicroscope.
Transgenic A. thaliana
The MKN2 cDNA fragment was amplified using the primers
5 0 -CAACTTCTTCAACACCACGGCATA-30 and 50 -TAGCGG
CCGCGTCGCTCTGTTTGCTGTATCTCC-30 . The product
was cloned into the SmaI site of pBS SKII1(Stratagene) and the
1.45-kb SphI-SalI fragment from this plasmid was used to replace
uidA in the pBI121Hm vector (Akama et al. 1992). MKN2 was
flanked by the cauliflower mosaic virus (CaMV) 35S promoter and
a transcriptional terminator. This construct was introduced into
Agrobacterium tumefaciens C58C1Rifr harboring the plasmid
pGV2260 with electroporation, and then into A. thaliana Columbia with floral dipping. The transformed A. thaliana was selected
on 1.5% agar plates containing 2.2 g/l Murashige and Skoog
culture medium basal salts (Wako, Osaka, Japan), 1 ml/l
Gamborg’s vitamin solution 1000 (Sigma, St. Louis, MO,
USA), 10 g/l sucrose, and 50 mg/l hygromycin. The seedlings were
transplanted into soil. Mature plants were observed for morphological alterations under long-day conditions (16-h light and 8-h
dark) at 221C. Total RNA was extracted from leaves of one wild
type-like line and three lines with serrated leaves using the RNeasy
Plant Mini kit (Qiagen, Hilden, Germany) and the expression of
MKN2 was detected by quantitative RT-PCR using One Step
SYBR PrimeScript RT-PCR Kit II (Perfect Real Time) (Takara
Bio, Shiga, Japan). Five nanograms of RNA were used as a template. Standard curves were obtained using 20, 5.0, 1.0, and 0.20 ng
total RNA extracted from one of the high MKN2 expression lines.
Expression levels were normalized with the values obtained for
the ORNITHINE TRANSCARBAMILASE (OTC) gene, which
was used as an internal reference gene (Jasinski et al. 2005).
The primers used are as follows: MKN2qF, TGCGGCCACTAC
GTTGACAC; MKN2qR, GGTCAACAATGACGCTTTCGTC
TA; OTCF, TGAAGGGACAAAGGTTGTGTATGTT; and
OTCR, CGCAGACAAAGTGGAATGGA. The measurements
were performed on a Light Cycler 3501SP (Roche, Basel,
Switzerland). The cycle involved reverse transcription at 451C for
5 min, denaturation at 951C for 10 sec, and 40 cycles of amplification, at 951C for 5 sec and 601C for 20 sec. Fluorescence was
measured at the end of each cycle. Cycle threshold (Ct) values were
determined by the second derivative maximum method.
Microscopy
The histochemical detection of GUS activity was performed as
described previously (Nishiyama et al. 2000). Spores were directly
observed using an environmental scanning electron microscope
(XL30 ESEM; Philips, Hillsboro, OR, USA). For semithin sectioning of resin-embedded tissues, sporophytes were fixed in 1%
glutaraldehyde, 3% paraformaldehyde, and 0.01% Triton X-100 in
0.1 M phosphate buffer (pH 7.0) for 2 h and in 1% osmium tetroxide for 1 h on ice. Sporophyte tissues were then dehydrated in a
graded acetone series and embedded in Quetol 651 epoxy resin
(Nisshin EM, Tokyo, Japan). Semithin sections (1 mm) were examined after staining with 5 g/l toluidine blue in 0.1 M sodium
phosphate buffer (pH 7.0).
558
EVOLUTION & DEVELOPMENT
Vol. 10, No. 5, September^October 2008
Expression analysis of MKN2, MKN4, and MKN5 by
semiquantitative RT-PCR
Protonemata were grown on solidified BCDAT medium for
1 week (Nishiyama et al. 2000). Gametophores were grown on
solidified BCD medium for 4 weeks (Sakakibara et al. 2001).
Various developmental stages of sporophytes were collected from
mosses grown on peat pellets for 4 weeks after the transition to
151C (Tanahashi et al. 2005). Total RNA was extracted using the
RNeasy Plant Mini kit (Qiagen). The cDNA was synthesized
from 1.0 mg of total RNA with Superscript II reverse transcriptase
RNaseH- (Invitrogen, Carlsbad, CA, USA) using the dT20
primer. PCR was performed using TaKaRa Ex Taq DNA polymerase Hot Start Version (Takara Bio). The primer pairs are
shown in Table S3.
Expression analysis of downstream candidates by
quantitative RT-PCR
P. patens was grown on peat pellets for 4 weeks after the transition
to 151C, and sporophytes were then collected. Total RNA was
extracted from 100 young sporophytes with longitudinal lengths of
350 450 mm (Fig. S2) using the RNeasy Micro kit (Qiagen) and
subjected to on-column DNase treatment according to the manufacturer’s instructions (Qiagen). The cDNA was synthesized from
0.5 mg of total RNA using the dT20 primer and Ready-To-Go YouPrime First-Strand Beads (GE Healthcare, Buckinghamshire,
England). Three cDNA preparations were synthesized independently. We searched the latest release of the P. patens genome sequence (Rensing et al. 2008). Seven loci for isopentenyl transferase
(IPT) gene homologs were found in the P. patens genome, of which
two genes have a single nucleotide difference. cDNA clones of five
genes were obtained by cloning RT-PCR products (PpIPT2,
AB294239, corresponding to the two loci; PpIPT3, AB294241;
PpIPT4, AB294242; PpIPT5, AB294243) or from the EST library
(PpIPT1, AB294240). The cDNA for PpIPT6 (represented by
ti|993727240 and other WGS sequences; Table S1) was not detected
by the primer sets that successfully amplified genomic DNA (Table
S4 and Fig. 7). Twenty gibberellic acid (GA) oxidase gene homologs
were found in the P. patens genome (Table S1), and the phylogenetic relationship of the P. patens genes and the representative land
plant GA oxidase genes was inferred using the ML method (Fig.
S3). Four P. patens genes orthologous to seed plant GA2 oxidase
genes (PpGA2ox1, AB294413; PpGA2ox2, AB294414; PpGA2ox3,
AB294415; PpGA2ox4, AB294416) and six genes orthologous to
GA20 oxidase genes (PpGA20ox1, AB294417; PpGA20ox2,
AB294418; PpGA20ox3, AB294419; PpGA20ox4, AB294420;
PpGA20ox5, AB294421; PpGA20ox6, AB294422) were detected.
Their expression patterns were first analyzed by RT-PCR (Fig. 7).
Genes detected by RT-PCR in the RNA extracted from young
sporophytes were further analyzed by real-time PCR analyses using
the 7500 Real-Time PCR System (Applied Biosystems, Foster City,
CA, USA) with SYBR Premix Ex Taq (Perfect Real Time; Takara
Bio) according to the manufacturer’s instructions. The PCR of each
cDNA sample was performed in triplicate using gene-specific primers (Table S4) and the following cycling conditions: 951C for 10 sec,
and 40 cycles of 951C for 5 sec and 601C for 34 sec. The amount of
cDNA for each gene was quantified using a log-linear regression
curve of the threshold cycle and the amount of standard template
prepared from a cDNA clone. The expression levels of the IPT and
GA oxidase genes relative to the P. patens PpGapC1 gene (X72381)
were calculated by dividing the amount of the target gene cDNA by
the amount of PpGapC1 cDNA in each preparation.
RESULTS AND DISCUSSION
Class 1 KNOX genes do not function in
gametophyte development
We searched for P. patens class 1 KNOX genes in the
P. patens v1.1 repeat masked main genome sequence (http://
genome.jgi-psf.org//Phypa1_1/Phypa1_1.home.html) and raw
whole genome shotgun sequence reads with TBLASTN using
the A. thaliana class 1 KNOX gene, SHOOT MERISTEMLESS as a query. Five genomic loci were found, four of which
corresponded to previously reported MKN1/MKN3, MKN2,
MKN4, and MKN5 genes (Champagne and Ashton 2001).
The other locus was designated MKN6. We performed phylogenetic analyses of KNOX genes incorporating these
Fig. 2. A maximum likelihood phylogenetic tree of the KNOX
homeobox genes in representative green plants. Bootstrap probabilities of more than 50% are shown on the branches. The horizontal branch length is proportional to the estimated evolutionary
distance. The bar is 0.1 substitutions per site. Symbols after gene
names indicate the higher classification: dicots (solid circles),
monocots (open circles), gymnosperms (double circles), ferns (open
squares), a lycophyte (open triangles), mosses (solid triangles), and
a green alga (solid squares). The brackets to the right indicate the
subfamilies of the KNOX gene family. The identification of
SmKN1 to 4 and MKN6 is described in Table S1.
Sakakibara et al.
Evolution of class 1 KNOX genes and shoot development
sequences (Figs. 2 and S4). MKN2, MKN4, and MKN5 genes
formed a clade with vascular plant class 1 KNOX genes and
MKN5 is sister to MKN2. MKN1 and MKN6 formed a clade
with vascular plant class 2 KNOX genes. We cloned and
sequenced cDNAs of the three class 1 KNOX genes, which
are longer than previously reported (Champagne and Ashton
2001), and genomic DNA with flanking regions (AB266745,
AB266746, and AB266747). We did not find any additional
class 1 KNOX genes in the latest version of the P. patens
genome sequence (Rensing et al. 2008).
To investigate the functions of the MKN2, MKN4, and
MKN5, we generated deletion lines in which all the three class
1 KNOX loci were replaced with selection marker genes (Fig.
S1). We generated single deletions for three genes, including
three lines of MKN2 (MKN2dis-1 to -3), three lines of MKN4
(MKN4dis-1 to -3), and five lines of MKN5 (MKN5dis-1 to 5) (Fig. S5). The MKN5dis-5 line was used to form three
MKN4 MKN5 double deletion lines (MKN4-5dis-1 to -3; Fig.
S5), and then the MKN4-5dis-3 line was used to generate
MKN2 MKN4 MKN5 triple deletion lines (MKN2-4-5dis-1
to -3). Three independent triple deletion lines were examined,
and the phenotypes of these lines were not distinguishable
from one another. The indeterminate meristem of the P. patens gametophore comprises an apical cell functionally equivalent to an animal stem cell, surrounded by proliferating cells.
An apical cell that is morphologically distinct from the apical
cell of the gametophore is formed at the tip of the protonema
filament, a hypha-like structure in the haploid generation. The
morphology of the gametophore and protonemata in the triple deletion lines was not distinguishable from that of the
wild-type plant (Fig. 3, A F).
Function of class 1 KNOX genes in sporophytes
The moss sporophyte lacks an indeterminate meristem. The
first zygotic cell division forms an apical and a basal cell. The
apical cell functions as a stem cell with oblique divisions, but
stops dividing after approximately 12 cell divisions. The cell
division plane of the apical cell was malformed in all three
triple deletion lines (Figs. 3, G and H). The angle between the
two oblique cell walls of the apical cell at the 10-cell stage (Fig.
3, G and H) was 94.8 16.51 (n 5 11) in the wild type and
109.2 17.41 (n 5 7) in the triple deletion line MKN2-4-5dis-1.
After the serial oblique divisions, the cells continued dividing in
both the wild type (Fig. 3, I, J, and M) and the deletion lines
(Fig. 3, K, L, and N), although both cell division and the
direction and amount of cell expansion were abnormal in the
deletion lines. A new multicellular intercalary meristem, a seta
meristem (Goebel 1930; French and Paolillo 1975), is formed in
the middle of a sporophyte, in which defects in cell division and
growth are conspicuous (Fig. 3, M and N). A seta locating
between a sporangium and a foot is mostly formed from the
seta meristem. The boundary between a sporangium and a seta
559
is conspicuous but that between a seta and a foot is unclear.
Therefore, the number of cells in a medial longitudinal section
of the lower half of the mature sporophyte including the seta
and foot was counted. The number of cells was 12.4 1.5
(n 5 5) in the wild type and 7.4 1.5 (n 5 5) in the triple
deletion line MKN2-4-5dis-1 (Fig. 3, P and R). As the effects of
decreased cell number and aberrant cell shape accumulated
during development, the mature sporophyte became aberrant.
However, every organ in the deletion lines could be recognized
(Fig. 3, O R) except for a lack of stomata (Fig. 3, S and T).
This indicates that MKN2, MKN4, and MKN5 do not regulate
tissue or organ identity. The surface ornamentation of spores
did not develop well in the triple deletion lines (Fig. 3, U and
V). The malformed cells in the sporophyte including columella
and spore sac cells, which are involved in tapetum cell formation (Crum 2001), are a possible reason for the aberrant spore
ornamentation. Except for spore ornamentation, no other
irregularities in spores were observed. The number of spores
was severely decreased in the triple deletion lines (wild type,
8.9 2.4 103; MKN2-4-5dis-1, 1.0 0.7 103) and the
rates of spore germination were similar in both the wild type
(34.2 13.0%) and MKN2-4-5dis-1 (31.2 23.2%), indicating that meiosis occurs normally in triple deletion lines and that
class 1 KNOX genes are not involved in meiotic cell divisions.
The decrease in spore number should be caused by a decrease
in mitotic cell divisions in sporogenous cells, which later form
spore mother cells.
Putative expression patterns of P. patens class 1
KNOX genes
Spatio-temporal expression patterns were estimated by making
an in-frame insertion of the GUS reporter gene just before the
stop codon of each MKN gene via homologous recombination
(Fig. S1). This enabled the estimation of MKN expression via
the histochemical detection of GUS. GUS activities were not
detected in the haploid indeterminate meristem in either the
protonemata (Fig. 4, A C) or the gametophores (Fig. 4,
D F). In contrast, GUS activity was detected during
sporophyte development. GUS signals were detected in proliferating and growing cells during early sporophyte development and the patterns of GUS activity varied among the three
class 1 KNOX genes (Fig. 4, G Z). The earliest signal was
detected in presumptive egg cells in MKN2-GUS lines (Fig.
4G), whereas the GUS signal in MKN4-GUS lines started at
the fertilized egg stage (Fig. 4, N–P). After several cell divisions, the GUS signal in MKN2-GUS (Fig. 4, H–J) and
MKN4-GUS (Fig. 4Q) lines was confined to the apical part of
the young sporophyte and was reduced in the basal part where
cells differentiate as foot cells (Fig. 4, J and Q). The GUS
signal in MKN5-GUS lines was not detected at this stage (Fig.
4, T W). After the oblique cell divisions of the apical cell
ceased, the GUS signal in MKN2-GUS and MKN4-GUS
560
EVOLUTION & DEVELOPMENT
Vol. 10, No. 5, September^October 2008
Fig. 3. Disruption of Physcomitrella patens class 1 KNOX genes affects sporophyte development. (AF) Gametophytes of the wild type (A,
C, and E) and triple deletion line MKN2-4-5dis-1 (B, D, and F). Protonemata (AD) and gametophores (E and F). (GT) Development
of the sporophytes in the wild type (G, I, J, M, O, P, S, and U) and the MKN2-4-5dis-1 line (H, K, L, N, Q, R, T, and V). Sporophytes
removed from the archegonia (O and Q). Sporophyte viewed under a Nomarski microscope (G and H) or a stereomicroscope (O and Q).
Toluidine blue-stained semithin sections observed under a microscope (I, K, M, N, P, and R). The apical cell (arrows in G and H)
undergoing cell division. The lines indicate oblique division planes (insets in G and H). The apical cell ceased to divide, although the
daughter cells continued to do so (I–L). The lines indicate the division planes in the young sporophyte (J and L). Cells actively dividing and
differentiating (M and N). A zone corresponding to the seta meristem is indicated as a bracket. A mature sporophyte (O–R) composed of an
operculum (op), a sporangium with spores (sp), a columella (c), a seta (se), and a foot (f) in both the wild type (O and P), and the MKN2-45dis-1 deletion line (Q and R). (S and T) Stomata are visible in the epidermis of wild type (S) but not in the MKN2-4-5dis-1(T). (U and V)
Surface of wild-type (U) and MKN2-4-5dis-1 (V) spores. Scale bar 5 50 mm (A, B, S, and T), 1 mm (C F), 100 mm (G R), and10 mm (U
and V).
Sakakibara et al.
Evolution of class 1 KNOX genes and shoot development
561
Fig. 4. Expression of Physcomitrella patens class 1 KNOX and b-glucuronidase (GUS) fusion proteins. (AZ) GUS activity was detected
in MKN2-GUS-1 (A, D, and G M), MKN4-GUS-1 (B, E, and NS), and MKN5-GUS-1 (C, F, and TZ) lines. GUS activity was not
detected in protonemata or gametophores in the three deletion lines (A F). GUS activity was detected in an egg primordial cell (arrow in
G), an unfertilized egg cell (arrow in H), a fertilized egg cell (arrows in I and P), an apical cell and surrounding cells (J and Q), and a
developing sporophyte (K, L, R, S, X, and Y). Expression diminished during maturation (M and Z). Sporophytes removed from archegonia
in J–M, Q–S, and W–Z. Scale bar 5 100 mm (AC), 1 mm (DF), and 50 mm (GZ).
562
EVOLUTION & DEVELOPMENT
Vol. 10, No. 5, September^October 2008
lines disappeared from the apex and were confined to a presumptive seta meristem (Fig. 4, K and R). The GUS signal in
MKN5-GUS lines was also detected in the seta meristem
(Fig. 4, X and Y). MKN2-GUS was further confined to the
center of a presumptive sporangium (Fig. 4L). The signal of
all genes ceased when the sporangium began to swell (Fig. 4,
M, S, and Z). Although in situ hybridization using riboprobes and antibodies was not successful in the gametophyte
shoot or the sporophyte tissue, the mRNA expression patterns detected by RT-PCR were concordant with the pattern
of GUS activity of the fusion protein (Fig. 5). Partial cDNA
fragments containing ELK and homeobox regions were detected in cDNAs collected from both haploid and diploid
plants, but cDNAs of MEIKNOX regions were detected
only in the diploid generation. Because we could not distinguish the morphology of protonemata and gametophores in
the triple deletion lines or a series of single and double deletion lines (data not shown) from the wild type, the partial
cDNAs are unlikely to represent functional mRNA in the
haploid plant body.
Comparison of class 1 KNOX gene functions
between moss and flowering plants
The morphological observations suggest that the P. patens
class 1 KNOX genes either directly or indirectly regulate the
Fig. 5. Semiquantitative reverse transcription polymerase chain reaction (RT-PCR) of MKN2, MKN4, and MKN5. Two primer sets,
one to amplify the MEIKNOX domain (KNOX) and the other to
amplify the ELK and homeo domains (ELK1HD), were used.
PCR cycle numbers are indicated on the right. The cytosolic
glyceraldehyde-3-phosphate dehydrogenase gene (PpGapC1) was
used as a loading control. Lane 1, protonemata; lane 2, gametophores; lane 3, sporophytes within archegonial tissues.
frequency of cell divisions and cell growth in the diploid
determinate meristem. When BREVIPEDICELLUS/KNAT1
(BP) (Chuck et al. 1996) or KNAT2 (Hamant et al. 2002)
is expressed under the control of the CaMV 35S promoter,
A. thaliana forms serrated leaves. When MKN2 was
overexpressed in A. thaliana under the control of the
CaMV 35S promoter, the leaves became similarly serrated
in 25 of 53 independent lines (Fig. 6). The level of MKN2
mRNA accumulation was consistent with the phenotype
(Fig. 6). In P. patens, the MKN2 single deletion caused
Sakakibara et al.
Evolution of class 1 KNOX genes and shoot development
a similarly severe phenotype to that of the triple deletion
lines (Fig. S5). This suggests that MKN2 may regulate
similar downstream genes in A. thaliana to those regulated by
BP and KNAT2.
In flowering plants, the class 1 KNOX genes regulate
transcript levels of the genes involved in metabolism of the
phytohormones cytokinin and gibberellin (GA), and thus
facilitate correct shoot meristem development (Barley and
Waites 2002; Jasinski et al. 2005; Yanai et al. 2005;
Sakamoto et al. 2006). These phytohormones regulate cell
division and expansion in both flowering plants and mosses
(Mitra and Allsopp 1959; Vaarama and Taren 1963;
Schumaker and Dietrich 1998). However, quantitative
RT-PCR did not detect significant differences in the mRNA
expression of homologs of isopentenyl transferase (PpIPT1–
PpIPT6) between the wild type and triple deletion lines in
the moss (Fig. 7). Similarly, there was no change in the
expression of the homologs of GA2 oxidase and GA20
oxidase genes, which are required in A. thaliana for GA
catabolism and biosynthesis, respectively (PpGA2ox1–
PpGA2ox4 and PpGA20ox1–PpGA20ox6; Table S1, Figs. 6
and S5). Thus, P. patens class 1 KNOX genes are unlikely to
regulate metabolism of cytokinin and GA. The GA response
mediated by GID and DELLA proteins is unlikely to be
conserved between mosses and vascular plants and no GA
response has been reported in P. patens (Hirano et al. 2007;
Yasumura et al. 2007), which suggests that meristematic
activity in this moss is not regulated by GA. Metabolism of
cytokinin is also different in the moss from that in angiosperms (Yevdakova and von Schwartzenberg 2007). These
data imply that regulation of the meristem is not via cytokinin
and GA downstream of the class 1 KNOX genes. The next
challenge is to resolve the class 1 KNOX gene-dependent
regulatory mechanism of the meristem in P. patens.
When we incorporate the result of MKN2 overexpression
in A. thaliana, part of the regulatory mechanism of the class 1
KNOX genes is conserved between these distantly related
taxa, mosses and angiosperms. Future studies on the detailed
regulatory mechanisms will provide insights into their
evolution. This result may also suggest that even though
heterologous gene can mimic a function of a native gene,
Fig. 6. Overexpression of MKN2 in Arabidopsis thaliana. (A, G,
and I) wild type and (B–E, H, and J) CaMV35S-MKN2 lines.
Rosette leaves of wild type (A) and CaMV35S-MKN2-1 to -4 lines
(B to E, respectively). The first cauline leaf of the wild type (G) and
the CaMV35S-MKN2-2 line (H). A flower of the wild type (I) and
the CaMV35S-MKN2-2 line (J). (F) Expression level of MKN2 in
the wild type (wt) and the CaMV35S-MKN2 transformants.
CaMV35S-MKN2-1 to -4 lines correspond to 1 to 4 on the x-axis,
respectively. The expression level was normalized against OTC.
Standard curves for measurement were made with the RNA of the
CaMV35S-MKN2-4 line. No amplification of MKN2 was detected
in the wild type. a.u., arbitrary unit. Scale bar 5 1 cm.
563
gene networks may not be conserved between these
distant taxa.
Evolution of the body plan in land plants
Based on phylogenetic, morphological, and paleobotanical
studies, it is hypothesized that developmental networks in the
sporophyte were co-opted from the preceding gametophyte
(Kenrick and Crane 1997a, b; Friedman et al. 2004). Comparisons between the gametophytic transcriptome of P. patens and
the A. thaliana genome show strong similarities, supporting this
hypothesis (Nishiyama et al. 2003). The development of cells
with a rooting function is controlled by a common mechanism
in the moss gametophyte and the A. thaliana sporophyte,
suggesting that the mechanism was recruited from the gametophyte to the sporophyte (Menand et al. 2007). In contrast,
the lack of full-length class 1 KNOX gene expression in the
haploid determinate meristem of the fern Ceratopteris richardii
suggests that there are different molecular mechanisms at work
in the haploid and diploid meristems (Sano et al. 2005).
Previous studies (Champagne and Ashton 2001; Singer
and Ashton 2007) suggested that P. patens class 1 KNOX
genes may function in gametophytes, based on the amplification of cDNA fragments by RT-PCR using gametophyte
RNA (Champagne and Ashton 2001). This previous study
used PCR primers that amplify a part of the cDNA of class 1
KNOX genes. Our RT-PCR results indicate that only partial
fragments of mRNAs are expressed in gametophytes and that
full-length cDNAs are detected only in sporophytes. Furthermore, we could not detect any GUS activity in gametophytes
of the GUS knock-in lines. Absence of a detectable gametophytic phenotype in the triple deletion lines confirms that
class 1 KNOX genes are not required for normal gametophyte development in P. patens.
The lack of class 1 KNOX gene function in the haploid
indeterminate meristem of P. patens indicates that there are
alternative mechanisms regulating the indeterminate meristem
in the haploid and diploid generations of land plants. In addition, there may also be differences in the molecular mechanisms of leaf formation. Leaf formation in flowering plants is
partly controlled by the genes ASYMMETRIC LEAVES1,
ROUGH SHEATH2, PHANTASTICA (ARP), and YABBY, which interact with class 1 KNOX genes. However,
homologs of these genes have not been found in the genome
of P. patens (Floyd and Bowman 2007). These findings indicate that the molecular mechanisms governing the indeterminate meristem in the sporophyte of land plants were not coopted in their entirety from the pre-existing gene networks in
the haploid indeterminate meristem. Such mechanisms at least
partly evolved de novo during land plant evolution.
Metazoans usually form a multicellular body in the diploid
generation. Core developmental genes such as transcription
factors are well conserved among the bilaterians, whose
564
EVOLUTION & DEVELOPMENT
Vol. 10, No. 5, September^October 2008
Fig. 7. Expression analyses of Physcomitrella patens IPT, GA2 oxidase, and GA20 oxidase gene homologs (PpIPT1–PpIPT6, PpGA2ox1–
PpGA2ox4, and PpGA20ox1–PpGA20ox6) in sporophytes. (A) RT-PCR for the PpIPT, PpGA2ox, and PpGA20ox genes with cDNA
synthesized using young sporophyte RNA extracted from the wild-type plant (lane 2), MKN2-4-5dis-1 triple deletion line (lane 3), and
MKN2-4-5dis-2 triple deletion line (lane 4). PCR fragments amplified using a plasmid containing the cDNA or genomic region of each gene
(lane 1) were used as a control. Genes with detectable expression in the young sporophyte were used for real-time PCR analyses (). (B)
Real-time RT-PCR analyses of PpIPT1, PpGA2ox3, PpGA20ox1, 3, 4, and 6. Expression levels of the PpIPT, PpGA2ox, and PpGA20ox
genes were normalized against the PpGapC1 gene. Black, hatched, and white bars represent the signals in the wild type (wt), MKN2-4-5dis-1
(dis1), and MKN2-4-5dis-2 (dis2), respectively.
common ancestor arose approximately 600 Ma (Peterson et
al. 2004). The regulatory change of core genes to downstream
genes likely caused the diversification of the body plan
(Carroll et al. 2001). Our study indicates that some of the core
developmental genes required for leafy shoot formation
changed during the evolution of land plants. This change
should be studied in greater detail, based on global analyses of
other developmental genes in land plants.
Acknowledgments
We wish to thank K. Nakamura, H. Tsukaya, Y. Hiwatashi, and
R. Kofuji for providing plasmids, the NIBB Center for Analytical
Instruments for DNA sequencing, Futamura Chemical Industries
Co. Ltd. for providing the cellophane, Kyowa Hakko Kogyo Co.
Ltd. for Driselase, M. Shimamura, T. Murata, K. Miyawaki,
T. Kurata, M. Kubo, S. Miyazaki, T. Tanahashi, Y. Hiwatashi, and
R. Sano for technical suggestions and help, and J. L. Bowman,
T. Kurata, and M. Kubo for comments on the manuscript. The
Sakakibara et al.
Evolution of class 1 KNOX genes and shoot development
computations were partly done in the Computer Lab of NIBB. This
research was partly supported by grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to K. S.,
T. N., H. D., and M. H.) and Japan Society for the Promotion of
Science (to T. N., H. D., and M. H.). K. S. is a research fellow of the
Japan Society for the Promotion of Science.
REFERENCES
Adachi, J., and Hasegawa, M. 1996. MOLPHY version 2.3: programs for
molecular phylogenetics based on maximum likelihood. Comput. Sci.
Monogr. 28: 1–150.
Akama, K., Shiraishi, H., Ohta, S., Nakamura, K., Okada, K., and Shimura, Y. 1992. Efficient transformation of Arabidopsis thaliana comparison of the efficiencies with various organs, plant ecotypes and
Agrobacterium strains. Plant Cell Rep. 12: 7–11.
Ashton, N. W., and Cove, D. J. 1977. The isolation and preliminary characterisation of auxotrophic and analogue resistant mutants in the moss
Physcomitrella patens. Mol. Gen. Genet. 154: 87–95.
Barley, R., and Waites, R. 2002. Plant meristems: the interplay of KNOX
and gibberellins. Curr. Biol. 12: R696–R698.
Carroll, S. B., Grenier, J. K., and Weatherbee, S. D. 2001. From DNA
to Diversity: Molecular Genetics and the Evolution of Animal Design.
Blackwell science, Malden, MA.
Champagne, C. E. M., and Ashton, N. W. 2001. Ancestry of KNOX genes
revealed by bryophyte (Physcomitrella patens) homologs. New Phytol.
150: 23–36.
Chuck, G., Lincoln, C., and Hake, S. 1996. KNAT1 induces lobed leaves
with ectopic meristems when overexpressed in Arabidopsis. Plant Cell 8:
1277–1289.
Crum, H. 2001. Structural Diversity of Bryophytes. The University of
Michigan Herbarium, Ann Arbor.
Felsenstein, J. 2005. PHYLIP (Phylogeny Inference Package). Version 3.65
Distributed by the Author. Department of Genetics, University of Washington, Seattle.
Floyd, S. K., and Bowman, J. L. 2007. The ancestral developmental tool kit
of land plants. Int. J. Plant Sci. 168: 1–35.
French, J. C., and Paolillo, D. J. 1975. Intercalary meristematic activity in
the sporophyte of Funaria (Musci). Am. J. Bot. 62: 86–96.
Friedman, W. E., Moore, R. C., and Purugganan, M. D. 2004. The evolution of plant development. Am. J. Bot. 91: 1726–1741.
Gifford, E. M., and Foster, A. S. 1989. Morphology and Evolution of Vascular Plants. 3rd ed. Freeman and Co., New York.
Goebel, K. 1930. Organographie der Pflanzen: Inbesondere der Archegoniaten und Samenpflanzen. Gustav Fischer, Jena.
Hamant, O., et al. 2002. The KNAT2 homeodomain protein interacts with
ethylene and cytokinin signaling. Plant Physiol. 130: 657–665.
Harrison, C. J., Corley, S. B., Moylan, E. C., Alexander, D. L., Scotland,
R. W., and Langdale, J. A. 2005. Independent recruitment of a conserved
developmental mechanism during leaf evolution. Nature 434: 509–514.
Hasegawa, M., and Kishino, H. 1994. Accuracies of the simple methods for
estimating the bootstrap probability of a maximum-likelihood tree. Mol.
Biol. Evol. 11: 142–145.
Hirano, K., et al. 2007. The GID1-mediated gibberellin perception mechanism is conserved in the lycophyte Selaginella moellendorffii but not in
the bryophyte Physcomitrella patens. Plant Cell 19: 3058–3079.
Jasinski, S., et al. 2005. KNOX action in Arabidopsis is mediated by
coordinate regulation of cytokinin and gibberellin activities. Curr. Biol.
15: 1560–1565.
Jones, D. T., Taylor, W. R., and Thornton, J. M. 1992. The rapid
generation of mutation data matrices from protein sequences. Comput.
Appl. Biosci. 8: 275–282.
Kasahara, M., Kagawa, T., Sato, Y., Kiyosue, T., and Wada, M. 2004.
Phototropins mediate blue and red light-induced chloroplast movements
in Physcomitrella patens. Plant Physiol. 135: 1388–1397.
565
Katoh, K., Kuma, K., Toh, H., and Miyata, T. 2005. MAFFT version 5:
improvement in accuracy of multiple sequence alignment. Nucleic Acids
Res. 33: 511–518.
Kenrick, P., and Crane, P. R. 1997a. The Origin and Early Diversification of
Land Plants: A Cladistic Study. Smithsonian Institution Press, Washington, DC.
Kenrick, P., and Crane, P. R. 1997b. The origin and early evolution of
plants on land. Nature 389: 33–39.
Kishino, H., Miyata, T., and Hasegawa, M. 1990. Maximum likelihood
inference of protein phylogeny and the origin of chloroplasts. J. Mol.
Evol. 30: 151–160.
Liu, Y.-G., Mitsukawa, N., Oosumi, T., and Whittier, R. F. 1995. Efficient
isolation and mapping of Arabidopsis thaliana T-DNA insert junctions
by thermal asymmetric interlaced PCR. Plant J. 8: 457–463.
Maddison, D., and Maddison, W. 2000. MacClade 4: Analysis of Phylogeny
and Character Evolution. Sinauer Associates, Sunderland, MA.
Mele, G., Ori, N., Sato, Y., and Hake, S. 2003. The knotted1-like homeobox
gene BREVIPEDICELLUS regulates cell differentiation by modulating
metabolic pathways. Genes Dev. 17: 2088–2093.
Menand, B., et al. 2007. An ancient mechanism controls the development of
cells with a rooting function in land plants. Science 316: 1477–1480.
Mitra, G. C., and Allsopp, A. 1959. The effects of various physiologically
active substances on the development of the protonemata and bud
formation in Pohlia nutans (Hedw.) Linb. Phytomorphology 9: 64–71.
Nishiyama, T., Hiwatashi, Y., Sakakibara, K., Kato, M., and Hasebe, M.
2000. Tagged mutagenesis and gene-trap in the moss, Physcomitrella
patens by shuttle mutagenesis. DNA Res. 7: 9–17.
Nishiyama, T., et al. 2003. Comparative genomics of Physcomitrella
patens gametophytic transcriptome and Arabidopsis thaliana:
implication for land plant evolution. Proc. Natl. Acad. Sci. USA 100:
8007–8012.
Peterson, K. J., Lyons, J. B., Nowak, K. S., Takacs, C. M., Wargo, M. J.,
and McPeek, M. A. 2004. Estimating metazoan divergence times with a
molecular clock. Proc. Natl. Acad. Sci. USA 101: 6536–6541.
Rensing, S. A., et al. 2008. The Physcomitrella genome reveals evolutionary
insights into the conquest of land by plants. Science 319: 64–69.
Sakakibara, K., Nishiyama, T., Kato, M., and Hasebe, M. 2001. Isolation
of homeodomain-leucine zipper genes from the moss Physcomitrella
patens and the evolution of homeodomain-leucine zipper genes in land
plants. Mol. Biol. Evol. 18: 491–502.
Sakakibara, K., Nishiyama, T., Sumikawa, N., Kofuji, R., Murata, T., and
Hasebe, M. 2003. Involvement of auxin and a homeodomain-leucine
zipper I gene in rhizoid development of the moss Physcomitrella patens.
Development 130: 4835–4846.
Sakamoto, T., Kamiya, N., Ueguchi-Tanaka, M., Iwahori, S., and Matsuoka, M. 2001. KNOX homeodomain protein directly suppresses the
expression of a gibberellin biosynthetic gene in the tobacco shoot apical
meristem. Genes Dev. 15: 581–590.
Sakamoto, T., et al. 2006. Ectopic expression of KNOTTED1-like homeobox protein induces expression of cytokinin biosynthesis genes in rice.
Plant Physiol. 142: 54–62.
Sano, R., et al. 2005. KNOX homeobox genes potentially have similar
function in both diploid unicellular and multicellular meristems, but not
in haploid meristems. Evol. Dev. 7: 69–78.
Schumaker, K. S., and Dietrich, M. A. 1998. Hormone-induced signaling
during moss development. Annu. Rev. Plant Physiol. Plant Mol. Biol. 49:
501–523.
Singer, S. D., and Ashton, N. W. 2007. Revelation of ancestral roles of
KNOX genes by a functional analysis of Physcomitrella homologues.
Plant Cell Rep. 26: 2039–2054.
Steeves, T. A., and Sussex, I. M. 1989. Patterns in Plant Development. 2nd
ed. Cambridge University Press, Cambridge.
Sundås-Larsson, A., Svenson, M., Liao, H., and Engström, P. 1998. A
homeobox gene with potential developmental control function in the
meristem of the conifer Picea abies. Proc. Natl. Acad. Sci. USA 95:
15118–15122.
Tanahashi, T., Sumikawa, N., Kato, M., and Hasebe, M. 2005. Diversification of gene function: homologs of the floral regulator FLO/LFY
control the first zygotic cell division in the moss Physcomitrella patens.
Development 132: 1727–1736.
566
EVOLUTION & DEVELOPMENT
Vol. 10, No. 5, September^October 2008
Turmel, M., Pombert, J.-F., Charlebois, P., Otis, C., and Lemieux, C. 2007.
The green algal ancestry of land plants as revealed by the chloroplast
genome. Int. J. Plant Sci. 168: 679–689.
Vaarama, A., and Taren, N. 1963. On the separate and combined effects of
calcium, kinetin and gibberellic acid on the development of moss protonemata. J. Linn. Soc. Lond. Bot. 58: 297–304.
Yanai, O., et al. 2005. Arabidopsis KNOXI proteins activate cytokinin
biosynthesis. Curr. Biol. 15: 1566–1571.
Yasumura, Y., Crumpton-Taylor, M., Fuentes, S., and Harberd, N. P.
2007. Step-by-step acquisition of the gibberellin-DELLA growth-regulatory mechanism during land-plant evolution. Curr. Biol. 17: 1225–1230.
Yevdakova, N. A., and von Schwartzenberg, K. 2007. Characterisation of a
prokaryote-type tRNA-isopentenyltransferase gene from the moss
Physcomitrella patens. Planta 226: 683–695.
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article:
Table S1. List of contigs.
Table S2. Primer sequences for plasmid construction.
Table S3. Primer sequences for semi quantitative RT-PCR
of MKN2, MKN4, and MKN5.
Table S4. Primer sequences for real-time PCR.
Fig. S1. Gene targeting of MKN2, MKN4, and MKN5.
(A) Genomic structures of MKN2, MKN4, and MKN5 in
the wild type, triple MKN deletion lines, and MKN-GUS
transformants. Boxes and lines between the boxes indicate
exons and introns, respectively. The circle indicates a putative stop codon. The blue, red, and green boxes indicate
a MEIKNOX, an ELK, and a homeodomain, respectively. The labels uidA, nos-ter, nptII, hpt, and zeocin
designate the uidA coding region, nopaline synthase
polyadenylation signal, NPTII expression cassette, HPT
expression cassette, and zeocin resistance cassette, respectively. Each bracket above the genome structure indicates
the region contained in the plasmid used for gene targeting. B, BglII; E1, EcoRI; E14, EcoT14I; H, HindIII. Thick
black lines indicate the probes used for Southern hybridization. (B E) Southern blots to confirm gene targeting.
Probes used for Southern hybridization and restriction
enzymes used for the digestion of P. patens genomic DNA
are indicated. (B) MKN2-4-5dis-1 3 (lanes 1–3). (C)
MKN2-GUS-1 and -2 (lanes 1 and 2). (D) MKN4-GUS-1
and -2 (lanes 1 and 2). (E) MKN5-GUS-1-4 (lanes 1–4).
The sizes of the signals are indicated on both sides (kb).
Fig. S2. A sporophyte at the developmental stage used for
quantitative RT-PCR. The arrow indicates longitudinal
length. Scale bar 5 100 mm.
Fig. S3. A maximum-likelihood tree for the GA oxidase
genes. Local bootstrap probabilities supported by more
than 50% are shown on the branches. The horizontal
branch length is proportional to the estimated evolutionary
distance. P. patens genes are shown in red. The putative
orthologs for GA2 oxidase and GA20 oxidase are indicated
by arrows.
Fig. S4. The maximum-likelihood tree for 136 KNOX
genes found using a local rearrangement search. Bootstrap
probabilities supported by more than 50% are shown on the
branches. The horizontal branch length is proportional to the
estimated evolutionary distance. The color of the gene name
corresponds to the higher classification: light blue, eudicots;
blue, monocots; purple, gymnosperms; orange, ferns; green,
lycophytes; red, mosses; black, green algae. Brackets to the
right indicate the subfamilies of the KNOX gene family.
Fig. S5. Sporophytes from the wild type and MKN deletion lines. (A) Wild type. (B) Triple deletion line (MKN2-45dis-1). (C) MKN2 single deletion line (MKN2dis-1). (D)
MKN4 single deletion line (MKN4dis-3). (E) MKN5 single
deletion line (MKN5dis-3). (F) MKN4 MKN5 double deletion line (MKN4-5dis-3). Scale bars are 100 mm. (G I)
Southern blots to confirm gene targeting. The probes shown
in Fig. S1 and the restriction enzymes used for digestion of the
P. patens genomic DNA are listed. (G) MKN2dis-1 to 3
(lanes 1–3). (H) MKN4dis-1 to 3 (lanes 1–3) and MKN4-5
dis-1 to 3 (lanes 4–6). (I) MKN5dis-1-5 (lanes 1–5). The size of
each signal is indicated on either side (kb).
Please note: Wiley Periodicals, Inc. are not responsible for
the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the
article.