directly to an LRR receptor domain of HAR1 and
significantly suppress nodulation at nano-molar
concentrations when applied from the cut surface
of the cotyledon. Furthermore, the arabinosylated
CLE-RS2 peptide can be detected in xylem sap
collected from the soybean shoots when CLE-
RS2 is specifically expressed in the soybean
hairy roots (Okamoto et al. 2013 ). These results
suggest that arabinosylated CLE-RS2 peptide is
the long sought after root-derived signal for the
onset of AON. On the other hand, the mature
structure of CLE-RS1 peptide remains unknown.
7.3 KLAVIER and CLV2 as Shoot
Factors
L. japonicus mutant har1 and other legume
mutants carrying mutations in theHAR1orthologs
do not exhibit anyclv1-like shoot phenotypes such
as fasciation by enlargement of SAM. Thisfinding
indicates that in legumes,CLV1orthologs play a
specific role in the systemic regulation of nodu-
lation but not in the feedback control of SAM
development. Although the genes responsible for
the regulation of SAM size were not identified in
legumes until relatively recently, several hyper-
nodulating mutants were known to exhibitclv1-
like phenotypes. InL. japonicus,klavier (klv)
exhibits not only a typical hypernodulation phe-
notype, but alsoclv-like phenotypes such as fas-
ciated stems, an increased number offlowers per
peduncle and bifurcated pistils (Oka-Kira et al.
2005 ; Miyazawa et al. 2010 ). Grafting experi-
ments usingklvshoots and wild-type roots have
demonstrated that KLV functions in the shoots to
control nodule numbers, as is the case for HAR1
(Oka-kira et al. 2005 ). Similarly, thepea sym28
mutant also exhibits shoot-regulated hypernodu-
lation, fasciated stems and an increased number of
flowers (Sagan and Duc 1996 ). These pleiotropic
phenotypes suggest a potential evolutionary link
between SAM maintenance and AON in legumes.
Positional cloning identified KLV, which
encodes an LRR receptor-like kinase (Miyazawa
et al. 2010 ). More importantly,KLV is most
closely related to ArabidopsisRPK2, which is
essential for SAM maintenance (Kinoshita et al.
2010 ). A double-mutant analysis indicates that
KLVandHAR1act in the same genetic pathway
that governs the long-distance control of nodu-
lation.KLVis predominantly expressed in the
vascular tissues of mature leaves, as isHAR1.
The biochemical analyses actually demonstrated
that KLV physically interacts with HAR1 in
Nicotiana benthamiana(Miyazawa et al. 2010 ),
suggesting that the potential KLV–HAR1
receptor complex systemically regulates nodula-
tion by receiving the root-derived arabinosylated
CLE peptides.
In the reproductive phase, peasym28shoots
develop additionalflowers, fascinated stems and
abnormal phyllotaxis as well as hypernodulation
(Sagan and Duc 1996 ). Recently, molecular
genetic approaches identified the causal gene
Sym28(Krusell et al. 2011 ).Sym28encodes an
LRR receptor-like protein and is closely related
toArabidopsis CLAVATA2(CLV2) that lacks a
kinase domain and forms a complex with CO-
RYNE/SOR2, a membrane-associated kinase
that regulates the SAM maintenance in Arabid-
opsis (Bleckmann et al. 2010 ; Guo et al. 2010 ;
Zhu et al. 2010 ). On the other hand, downregu-
lation of theL. japonicus Clv2gene by RNAi
resulted in enhanced nodulation (Krusell et al.
2011 ). Thus, as with KLV, legume CLV2
receptor-like proteins appear to be involved in
AON via long-distance signalling as well as the
maintenance of SAM development.7.4 TOO MUCH LOVE as a Root
FactorIn contrast to shoot factors,Pisum sativum nod3,
M. truncatula sickleandrdn1mutants belong to
the category of the root genotype-determined
hypernodulating mutants (Postoma et al. 1998 ;
Penmetsa and Cook 1997 ; Schnabel et al. 2011 ).
Furthermore,L. japonicus rdh1,too much love
(tml), andplenty,have been also isolated as root-
determined hypernodulating mutants (Ishikawa
et al. 2008 ; Magori et al. 2009 ; Yoshida et al.
2010 ). Asrdh1turned out to be allelic totmlit has
been designated tml- 4 (Takahara et al. 2013 ).
Some of these mutants are thought to be deficient7 Genes for Autoregulation of Nodulation 75