with nontransgenic plants, the T 3 transgenic
plants showed a greater number of spikelets
per spike (0.9 spikelets; fig. S10A), longer spike
length (4.4 cm; fig. S10B) and more spikes
per plant (0.29 spikes; fig. S10C), again with no
compensatory loss in thousand grain weight
(fig. S10E). No significant difference was ob-
served, however, in grain number per spike
(fig. S10D). The net effect of these spike and
grain traits was that grain yield increased
between 7.8% and 19.8% among the four
transgenic Yangmai18 lines (fig. S10F) in stan-
dard yield plots, averaging an 11.9% increase
over nontransgenic Yangmai18 (table S3). We
concluded that the constitutive overexpression
ofTaCol-B5in T 1 ,T 2 , and T 3 transgenic plants
modified spike architecture and increased the
numbers of spikelets and spikes, but its posi-
tive effect on the number of grains in the T 3
population was suppressed by plant density
or environmental effects.TaCol-B5as a sin-
gle gene increased grain yield in the indig-
enous cultivar Yangmai18.
TaCOL-B5was primarily expressed in the
shoot apex and tiller bud, consistent with its
potential role in promoting tillering and thus
more spikes, but it was also expressed in leaves
and roots of juvenile plants at the five- to six-
leaf stage (fig. S11). However, there was no
significant difference in the spatial or tem-
poral expression ofTaCOL-B5between the
two alleles (fig. S11), excluding the possibility
that the traits described above were regu-
lated byTaCOL-B5at the transcript level
and leading to the alternative hypothesis
that phenotypic differences were probably
determined by differences inTaCOL-B5 at
the protein level.
Three amino acid substitutions, Phe^243 /
Leu^243 , Ser^269 /Gly^269 , and Ala^338 /Thr^338 , were
found betweenTaCol-B5 andTacol-B5 pro-
teins (fig. S4). We next investigated whether
any of these amino acid substitutions affected
the interaction ofTaCol-B5 orTacol-B5 with
other proteins. From a wheat yeast two-hybrid
(Y2H) library, we identified a clone encodingTraesCS4D02G196100, orTaK4, which is an
ortholog of riceOsK4encoding a serine/
threonine protein kinase (GenBank Q852Q1)
( 9 ). Indeed,TaCol-B5 andTacol-B5 showed
differential interactions withTaK4 in the Y2H
system (Fig. 3A and fig. S12) and in a transient
expression system in tobacco leaves (fig. S13).
Protein sequence analysis suggests phosphor-
ylation sites in the amino acid substitutions
(Fig. 3B). Furthermore, comparative in vitro
phosphorylation interaction studies showed
that the Ser^269 /Gly^269 substitution inTaCol-B5
andTacol-B5 resulted in potential differential
protein phosphorylation byTaK4 (Fig. 3C).
This study provides an example that protein
phosphorylation may be involved in spike ar-
chitecture and grain yield in plants.
Constitutive overexpression ofTaCol-B5also
was found to regulate heading date (earlier)
in the greenhouse and plant height (taller) in
thegreenhouseandfield(fig.S14).Wetested
whether the CCT (CONSTANS, CO-like, and
TOC1) domain ofTaCOL-B5 is manifestedSCIENCEscience.org 8 APRIL 2022¥VOL 376 ISSUE 6589 181
Fig. 1. Mapping and positional cloning of
QSns.osu-7B.(A) Mapping ofQSns.osu-7B.
Physical locations of the GBS markers are
provided in table S1. The horizontal dashed line
represents a threshold LOD value of 3.0.
(B) Physical map of crossovers detected in
four critical recombinant plants. The populations
derived from these plants were mapped with
black dots representing PCR markers (fig. S3),
red dots representingTraesCS7B02G400600, and
yellow dots representingTraesCS7B02G400700.
The red“X”indicates a crossover between
markers.“A”represents the CItr 17600 allele;
“B,”the Yangmai18 allele; and“H,”heterozygotes.
CS, Chinese Spring. (CtoF) Average SNS in
the four F 6 populations. Populations P11-58 (C)
and P19-236 (D) each show significant segregation
for SNS. Populations P19-1121 (E) and P11-63
(F) show no association between SNS and
two candidate genes,TraesCS7B02G400600
andTraesCS7B02G400700. More detailed
phenotypic analyses of these populations are
provided in fig. S2.RESEARCH | REPORTS