Plant Physiology Preview. Published on July 10, 2015, as DOI:10.1104/pp.15.00367
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Running head: BRI1 control of maize plant architecture
Corresponding Author: Philip W. Becraft Genetics, Development & Cell Biology Dept. 2116 Molecular Biology Bldg. Iowa State University Ames, IA 50011 Email:
[email protected] Phone: 515-294-2903 Research area: Genes, Development and Evolution
1 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
Copyright 2015 by the American Society of Plant Biologists
47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
RNAi knockdown of BRI1 in maize (Zea mays) reveals novel functions for brassinosteroid signaling in controlling plant architecture.
Gokhan Kira,b, Huaxun Yea,b, Hilde Nelissenc,d, Anjanasree K. Neelakandana, Andree S. Kusnandara,b, Anding Luoe, Dirk Inzéc,d, Anne W. Sylvestere, Yanhai Yina,b and Philip W. Becrafta,b,f a. Genetics, Development & Cell Biology Department, Iowa State University, Ames, IA 50011 b. Interdepartmental Genetics & Genomics Program, Iowa State University, Ames, IA 50011 c. Department of Plant Systems Biology, VIB, B-9052 Ghent, Belgium. d. Department of Plant Biotechnology and Bioinformatics, Ghent University, B-9052 Gent, Belgium e. Department of Molecular Biology, University of Wyoming, Laramie, WY 82071-2000 f. Agronomy Department, Iowa State University, Ames, IA 50011
Summary: Brassinosteroid signaling is central to controlling auricle development and establishing the blade-sheath boundary in maize leaf.
82 83 84 85 86 87 88
2 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
89
Footnote:
90
This research was supported by the Iowa State University Plant Sciences Institute and the
91
Genetics, Development & Cell Biology Department. G.K. was supported on an education
92
fellowship from the Ministry of National Education, Republic of Turkey.
93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 3 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
108 109
Abstract Brassinosteroids (BRs) are plant hormones involved in various growth and
110
developmental processes. The BR signaling system is well established in Arabidopsis
111
(Arabidopsis thaliana) and rice (Oryza sativa) but poorly understood in maize (Zea
112
mays). BRI1is a BR receptor and database searches and additional genomic sequencing
113
identified 5 maize homologs including duplicate copies of BRI1 itself. RNA interference
114
(RNAi) using the extracellular coding region of a maize zmbri1 cDNA knocked down
115
expression of all 5 homologs. Decreased response to exogenously applied brassinolide
116
(BL), and altered BR marker gene expression demonstrate that zmbri1-RNAi transgenic
117
lines have compromised BR signaling. zmbri1-RNAi plants showed dwarf stature due to
118
shortened internodes, with upper internodes most strongly affected. Leaves of zmbri1-
119
RNAi plants are dark green, upright and twisted, with decreased auricle formation.
120
Kinematic analysis showed that decreased cell division and cell elongation both
121
contributed to the shortened leaves. A BES1-YFP transgenic line was developed that
122
showed BR inducible BES1-YFP accumulation in the nucleus, which was decreased in
123
zmbri1-RNAi. Expression of the BES1-YFP reporter was strong in the auricle region of
124
developing leaves, suggesting that localized BR signaling is involved in promoting
125
auricle development, consistent with the zmbri1-RNAi phenotype. The blade/sheath
126
boundary disruption, shorter ligule and disrupted auricle morphology of RNAi lines
127
resembles KNOX mutants, consistent with a mechanistic connection between KNOX
128
genes and BR signaling.
129
4 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
130
Introduction
131
Brassinosteroids (BRs) are ubiquitous plant hormones that promote plant growth
132
by regulating cell elongation and division (Clouse, 1996; Clouse et al., 1996). BRs have
133
other diverse roles including enhancing tracheary element differentiation, stimulating
134
ATPase activity, controlling microtubule orientation, controlling flowering time, fertility,
135
and leaf development (Iwasaki and Shibaoka, 1991; Clouse et al., 1996; Li et al., 1996;
136
Schumacher et al., 1999; Catterou et al., 2001; Oh et al., 2011). BRs also function in
137
tolerance to both biotic and abiotic stresses such as extreme temperatures, drought, and
138
pathogens (Krishna, 2003).
139
Deficiencies in BR biosynthesis or signaling produce characteristic dwarf plant
140
phenotypes (Clouse et al., 1996; Szekeres et al., 1996; Fujioka et al., 1997; Fujioka et al.,
141
1997). Plant height is an important agricultural trait as seen in the Green Revolution,
142
where semi-dwarf mutants contributed to increased yields in small grain crops (Salas
143
Fernandez et al., 2009). BR deficient dwarf rice produced increased grain and biomass
144
yields because the erect leaf habit allowed higher planting densities under field conditions
145
(Sakamoto et al., 2006). In fact, Green Revolution ‘uzu’ barley (Hordeum vulgare) is
146
based on a mutation of the uzu1 gene, which encodes a homolog of
147
BRASSINOSTEROID INSENSITIVE1 (BRI1), a BR receptor (Chono et al., 2003).
148
Genes functioning in BR pathways have been identified by analysis of dwarf
149
mutants in several species, including Arabidopsis (Arabidopsis thaliana) and rice (Oryza
150
sativa). Arabidopsis bri1 mutants are shortened, have reduced apical dominance and are
151
male-sterile (Clouse et al., 1996). BRI1 encodes a leucine-rich repeat receptor-like kinase
5 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
152
(LRR-RLK) that is located in the plasma membrane and contains an extracellular domain
153
responsible for BR binding, a transmembrane sequence, and a cytoplasmic protein kinase
154
domain (Li and Chory, 1997; Vert et al., 2005; Belkhadir and Chory, 2006). The Island
155
Domain (ID) and subsequent LRR22 are critical for BR binding (Kinoshita et al., 2005;
156
Hothorn et al., 2011; She et al., 2011). Phosphorylation of conserved residues Ser-1044
157
and Thr-1049 in the kinase activation loop activates the BRI1 kinase (Wang et al., 2005),
158
while dephosphorylation of BRI1 by protein phosphatase 2A (PP2A) inhibits its function
159
(Wu et al., 2011).
160
BRI1 is partially redundant in BR signaling with related BRI1-LIKE RECEPTOR
161
KINASE (BRL) paralogs, both in Arabidopsis and rice. In Arabidopsis, even though null
162
alleles of brl1 or brl3 did not show obvious phenotypic defects in shoots, they enhanced
163
the developmental defects of a weak bri1-5 mutant. In contrast to ubiquitously expressed
164
BRI1, BRL1, 2 and 3 are tissue specific, mostly expressed in vascular tissues, while BRL1
165
and 3 are also expressed in root apices (Cano-Delgado et al., 2004; Zhou et al., 2004;
166
Fabregas et al., 2013). Both BRL1 and BRL3 can bind BL (Cano-Delgado et al., 2004).
167
In rice, OsBRI1 is similar to the Arabidopsis BRI1 gene and phenotypes of OsBRI1 rice
168
mutants include dwarf plants with shortened internodes, erect leaves that are twisted and
169
dark green, and photomorphogenesis in the dark (Yamamuro et al., 2000). There are three
170
BR receptors in rice as well, and while OsBRI1 is universally expressed in all organs,
171
OsBRL1 and OsBRL3 are expressed mostly in roots (Nakamura et al., 2006).
172
To date, two mutant genes of the BR biosynthetic pathway have been reported in
173
maize (Zea. mays). A classic dwarf mutant, nana plant1 (na1), has a mutation in a DET2
174
homologous gene, which encodes a 5α-reductase enzyme in the BR biosynthesis pathway 6 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
175
(Hartwig et al., 2011), while the brd1 gene encodes brC-6 oxidase (Makarevitch et al.,
176
2012). The maize BR deficient mutants have shortened internodes, twisted, dark green,
177
erect leaves and feminized male flowers (Hartwig et al., 2011; Makarevitch et al., 2012).
178
However, no genes in BR signaling have yet been reported in maize. Understanding BR
179
signaling in maize might help improve this important crop for production of biofuels,
180
biomass and grain yield. Here, we took a transgenic RNAi approach to generate maize
181
plants partially deficient for BRI1. These knockdown lines demonstrate that BRI1
182
functions are generally conserved in maize compared to other plant species, but they also
183
exhibit unique phenotypes suggesting either that maize possesses novel BR-regulated
184
developmental processes, or that aspects of maize morphology reveal processes not
185
evident in other plants.
186 187
7 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
188
Results
189
BRI1 homologs in maize
190
Database searches were conducted to determine whether maize possessed the potential to
191
encode all the core components of the canonical BR signal transduction pathway. As
192
reported in the supplemental data, high confidence homologs were identified for BRI1,
193
BAK1 (BRI1 ASSOCIATED KINASE1), BKI (BRI1 KINASE INHIBITOR1), BSK1
194
(BRASSINOSTEROID-SIGNALING KINASE1), CDG1 (CONSTITUTIVE
195
DIFFERENTIAL GROWTH1), BSU1 (BRI1 SUPPRESSOR1), BIN2
196
(BRASSINOSTEROID-INSENSITIVE2), and BES1/BZR1 (BRI1-EMS-
197
SUPPRESSOR1 / BRASSINAZOLE-RESISTANT2). As such, it appears highly likely
198
that the core BR signaling system is conserved in maize. Interestingly, an unequivocal
199
homolog of rice DLT (DWARF AND LOW-TILLERING) could not be identified.
200
Whether this reflects a gap in the genome assembly or a biological difference between
201
rice and maize remains to be determined.
202
BRI1 belongs to a family of LRR receptor-like kinases and at least 2 BRI1-like
203
genes are involved in BR signaling in Arabidopsis (Cano-Delgado et al., 2004; Zhou et
204
al., 2004). Maize BRI1 homologs were identified by using the Arabidopsis and rice BRI1
205
amino acid sequences to search the maize genome. A phylogenetic analysis was carried
206
out to identify relationships among maize BRI1 homologs. Protein products were aligned
207
(Supplemental Fig. S1) and a phylogenetic tree obtained by the neighbor-joining method
208
(Fig. 1). Five maize genes encode proteins belonging to the BRI1 family (Table 1).
209
Figure 1B shows that two maize gene products are most closely related to Arabidopsis
8 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
210
and rice BRI1 while three additional genes encode proteins similar to BRL1, BRL2 and
211
BRL3. The top hit, herein designated zmbri1a, corresponds to GRMZM2G048294 on
212
chromosome 8. The ZmBRI1a protein product shows 54 % amino acid identity and 69 %
213
similarity to Arabidopsis BRI1 (AtBRI1) and 79 % amino acid identity and 88 % 9 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
214
similarity to rice BRI1 (OsBRI1). A second gene, GRMZM2G449830, located on
215
chromosome 5 contained only a partial sequence in the genome assembly (B73 v.3) and
216
was annotated as a transposable element. However, it was found through cloning and
217
sequencing to contain a complete gene, designated zmbri1b (Genbank accession number
218
KP099562). The predicted ZmBRI1b protein shows 93 % amino acid identity and 95 %
219
similarity to maize ZmBRI1a, 79 % amino acid identity and 88% similarity with OsBRI1,
220
and 54 % amino acid identity and 69 % similarity with AtBRI1. The island domain or
221
insertion domain between LRRs 21 and 22 is a distinguishing characteristic of the BRI1
222
family of receptor kinases and the site of BR binding in BRI1 (Kinoshita et al., 2005;
223
Hothorn et al., 2011; She et al., 2011). All 5 of maize BRI1/BRL homologs contained a
224
conserved island domain and LRR 22 sequences involved in BR binding (Fig. 1C).
225
The expression patterns of these five BRI1/BRL homologs were examined by
226
querying publicly available gene expression datasets. According to the Maize Gene
227
Expression Atlas (Sekhon et al., 2011), the zmbri1a and zmbri1b transcripts are widely
228
expressed in most tissues sampled, with zmbri1a showing somewhat higher expression
229
levels and a more ubiquitous pattern than zmbri1b (Supplemental Fig. S2). Notably, both
230
are expressed in shoot apices, developing leaves, internodes and roots. The zmbrl genes
231
all show generally lower steady state transcript levels than the zmbri1 genes except in a
232
couple instances such as roots and germinating seeds where zmbrl transcripts are present
233
at higher levels than zmbri1b. The zmbrl1 and zmbrl3 transcripts are more highly
234
expressed than zmbrl2 except in roots where all 3 genes are expressed at comparable
235
levels. In most tissues, the zmbrl2 transcript was below reliable detection levels.
10 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
236
Expression along the developmental gradient of a growing seedling leaf was also
237
examined (Wang et al., 2014). As shown in Supplemental Fig. S3, zmbri1a and zmbri1b
238
expression was high at the base of the leaf in the zones of active cell division and cell
239
elongation, gradually declining as cells differentiate and mature. Expression of the zmbrl
240
genes was generally very low except for a brief increase in zmbrl1 transcript levels near
241
the region of the leaf where tissues change from acting as carbon sinks and become
242
photosynthetically active as source tissue.
243 244
zmbri1-RNAi lines show compromised BR signaling
245
To generate loss-of-function RNAi suppression of maize BRI1-related genes, a 498bp
246
fragment encoding LRRs 6 to 12 of the extracellular domain in the protein product was
247
cloned in forward and reverse orientations into a pMCG1005 vector, controlled by the
248
maize ubiquitin1 promoter and transformed into maize. Among 17 independent bri1-
249
RNAi T0 transgenic events, 13 showed clear dwarf phenotypes similar to those reported
250
in rice d61 or maize na1 seedlings (Yamamuro et al., 2000; Hartwig et al., 2011). Two
251
events were selected for further analysis, one producing strong phenotypes and one
252
intermediate. Each was backcrossed three or more generations to a B73 inbred
253
background.
254
Because BRI1 belongs to a gene family, it is likely that multiple members
255
participate in BR signaling, perhaps redundantly, and that multiple members were
256
targeted by the RNAi construct. To determine which genes in the maize genome were
257
likely to be targeted by the zmbri1-RNAi construct, nucleotide database searches were
258
conducted using the zmbri1a cDNA fragment contained in the RNAi construct. The 11 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
259
results of this search were consistent with the phylogeny shown in Fig.1B. The two top
260
hits included the original zmbri1a, and zmbri1b with 93% nucleotide identity. The zmbril
261
homologs showed somewhat lower homology ranging from 43.9% to 50.4% nucleotide
262
identity, although localized stretches of nucleotides showed higher levels of homology
263
(alignments shown in Supplemental Fig. S4). The expression of these 5 genes was
264
examined using RNA obtained from young developing shoot tissue, which included shoot
265
apical meristems as well as nodes, internodes and growing regions of leaves of
266
approximately the seven youngest plastochrons, from the upper, most strongly affected
267
nodes of the plant. Both zmbri1-RNAi lines showed decreased expression of all five
268
BRI1 homologs compared to non-transgenic siblings (Fig.1D).
269
Expression of BR marker genes brd1 and cpd, encoding enzymes involved in BR
270
biosynthesis, was examined by qRT-PCR. These genes are negatively feedback regulated
271
by BR signaling in Arabidopsis and rice, causing increased expression in mutants with
272
decreased BR signaling (Mathur et al., 1998; Hong et al., 2002; Bai et al., 2007). As seen
273
in Fig. 2, both shoot and stalk tissues from strongly affected regions of the plants have
274
increased expression of cpd and brd1, consistent with decreased BR signaling activity in
275
strong zmbri1-RNAi lines compared to non-transgenic siblings.
276
To further analyze BR signaling, we generated a BES1-YFP reporter for BR
277
activity. BES1 is a transcription factor that is post-translationally stabilized and
278
translocated into the nucleus to regulate gene expression in response to BR signaling (Yin
279
et al., 2002). As such, BR signaling is expected to increase nuclear fluorescence due to
280
BES1-YPF accumulation (Ryu et al., 2010). The maize genome was searched for BES1
281
homologs and GRMZM2G102514 was clearly the top candidate. The predicted protein 12 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
282
encoded by this gene model contains 313 amino acids and shows 50% overall identity,
283
62% similarity to Arabidopsis BES1 and 78% identity and 82% similarity to OsBZR1. To
284
create a stable BES1-YFP maize line, the YFP coding sequence was translationally fused
285
to the 3’ end of the maize BES1 coding region within the native genomic context of the 13 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
286
gene, including 5’ and 3’ regions as well as introns (Mohanty et al., 2009). As expected,
287
nuclear fluorescence in leaf sheath cells is BR inducible in transgenic BES1-YFP lines,
288
indicating the fusion protein responds to BRs and can serve as a BR reporter (Fig. 2C).
289
Since BES1-YFP responds to BRs, a decrease in BR signaling should cause less
290
BES1-YFP accumulation in the nucleus. To test the hypothesis that zmbri1-RNAi
291
disrupts BR signaling, we examined several tissues in zmbri1-RNAi; BES1-YFP lines.
292
As seen in Fig. 2C, there is decreased nuclear fluorescence in zmbri1-RNAi leaf sheath
293
cells, compared to the WT cells, consistent with decreased levels of BR signaling and
294
BES1-YFP accumulation. To see if these transgenic cells were able to respond to BR, we
295
exposed both transgenic and non-transgenic leaf sheath cells to 1μM BL. Fig. 2C shows
296
that WT cells are more responsive to BL than RNAi cells confirming that BR signaling is
297
impaired in zmbri1-RNAi lines.
298
To further confirm disrupted BR signaling in zmbri1-RNAi lines, we performed a
299
BR root growth inhibition assay, a commonly used bioassay for BR sensitivity (Clouse,
300
1996; Chono et al., 2003; Müssig et al., 2003; Kim et al., 2007; Wang et al., 2007;
301
Hartwig et al., 2012). Germinated seeds were treated with ddH2O (mock), 20 nM BL, or
302
100 nM BL. After 10 days, WT plants showed a two-fold greater inhibition of root
303
growth compared to zmbri1-RNAi seedlings, indicating decreased BL sensitivity in
304
transgenic seedlings (Fig. 2B and Supplemental Fig. S5).
305 306 307 308
zmbri1-RNAi alters maize plant architecture In both mild and strong zmbri1-RNAi lines, plant height is decreased compared to non-transgenic siblings (Fig.3 and Table 2). Internodes were analyzed to determine the 14 Downloaded from www.plantphysiol.org on July 19, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
309
basis of the shorter plant height. While the number of internodes was not significantly
310
changed (Table 2), internode length was dramatically shorter in zmbri1-RNAi lines than
311
their non-transgenic siblings. In the milder lines, internodes are shortened by similar
312
proportions throughout the plant (Supplemental Fig.S6), while in strong lines, internodes
313
between the ear and tassel nodes were most dramatically affected (Fig.3, B-D).
314
Epidermal cell lengths in the strongly affected ninth internodes were shortened an
315
average of 40% in the mild, and ~70% in the strong, zmbri1-RNAi-RNAi lines compared
316
to their non-transgenic counterparts, suggesting decreased cell elongation is a major
317
contributing factor to the decreased internode length (Fig.3, E-G).
318
Leaf blades and sheathes were both shorter in RNAi lines (Fig. 4, Table 2). Adult
319
leaves in zmbri1-RNAi transgenic lines were dark green, upright, thickened, and had a
320
wrinkled surface, compared to non-transgenic lines (Fig.3A, Fig. 4C and Supplemental
321
Fig. S7). These phenotypes were more apparent in adult leaves than in seedlings, and
322
stronger in field-grown plants than in the greenhouse. Like internode epidermal cells, leaf
323
epidermal cells of zmbri1-RNAi lines were also shortened compared to non-transgenic
324
siblings (Fig. 4D,E). In the first leaf above the ear node, epidermal pavement cells
325
sampled halfway along the blade length and midway between the midrib and margin had
326
a mean length of 85.8 ± 3.4 μm in WT, while zmbri1-RNAi cells averaged 46.3 ± 2.7 μm
327
(Student’s t-test p-value