The Lotus japonicus Genome

(SAM) maintenance via cell–cell communication
(Clark et al. 1997 ).CLV1is specifically expressed
in the centre of SAM, whereasHAR1and legume
orthologs are expressed in various organs, such as
leaves, stems and roots, but significantly sup-
pressed in the shoot apex (Krusell et al. 2002 ;
Nishimura et al. 2002 ; Schnabel et al. 2005 ; Searle
et al. 2003 ). Further expression analysis using a
GUS reporter gene driven by theHAR1native
promoter has elucidated thatHAR1is expressed
predominantly in the phloem tissues of leaves and
stems (Nontachaiyapoom et al. 2007 ). This
phloem-specific expression ofHAR1 seems to
make sense because phloem tissues function as the
conduit for long-distance communication
between distantly located organs. As the HAR1
ligand is presumed to act as a root-derived long-
distance mobile signal, the next important ques-
tion is‘What is a ligand for HAR1?’

7.2 Root-Derived Signal and

Arabinosylated CLE Peptide

The existence of the root-derived signal, also

known as‘Q’, wasfirst proposed by grafting and

split root experiments using soybeanntsmutants

(Caetano-Anollés and Gresshoff 1990 ). It is

thought that the root-derived signal is generated

in roots via early symbiotic signalling activated

by Nod factor secreted from rhizobia and then

translocated to the shoot. However, since the

early 1990s, the chemical nature of the root-

derived signal had remained unknown. In 2009,

Okamoto et al. ( 2009 )first reported a candidate

through an in silicosearch of theL. japonicus

genome database. They found thatL. japonicus

CLEgenes (CLE-RS1andRS2) are specifically

and rapidly induced in the roots in response to its

symbiotic bacteriaMesorhizobium lotiand that

the overexpression ofCLE-RS1/2genes drasti-

cally reduced or abolished nodulation in a HAR1-

dependent manner. Of particular importance is

that this inhibitory effect travels systemically

from transformed roots to untransformed roots

(Okamoto et al. 2009 ). SimilarCLEpeptide genes

showing local and systemic suppression of nod-

ulation have been reported in Medicagoand

Glycine(Mortier et al. 2010 , 2011 ; Reid et al.

2011 ). However, application of synthesized CLE

peptides deduced fromCLE-RS1/2gene struc-

tures did not suppress nodulation even at a

micromolar concentration, implying that some

specific posttranslational modification is required

for the biological activity of CLE-RS peptides.

Recently, Okamoto et al. determined the

mature structure of CLE-RS2 peptide by the

overexpression ofCLE-RS2 genes using Ara-

bidopsis submerged culture and L. japonicus

hairy root culture systems (Okamoto et al. 2013 ).

Through nano-liquid chromatography–mass

spectrometry (nano-LC-MS) and nano-liquid

chromatography–tandem mass spectrometry

(nano-LC-MS/MS) analyses of peptides diffused

into culture media, they identified a 13-amino-

acid CLE-RS2 peptide modified with three or

more residues of arabinose. Chemically synthe-

sized arabinosylated CLE-RS peptides bind

TML

Arabinosylated

CLE-RS peptides

Root-Derived

Signal Shoot-Derived Inhibitor

Nod factor

Mesorhizobium loti

Fig. 1 Schematic illustration of a model for HAR1-,
KLV, CLV2 and TML-mediated autoregulation of nod-
ulation (AON). 1 Perception of the rhizobial Nod factor
initiates the production of a long-distance inhibitor termed
the root-derived signal. 2 Arabinosylated CLE-RS1 and
CLE-RS2 peptides are transported to the shoot, and 3
activate the production of the shoot-derived signal.
HAR1, KLV and CLV2 mediate this process. 4 The
shoot-derived signal(s) is translocated to the root and
negatively regulates nodulation via TML. TML F-box
protein is a root factor acting at thefinal stage of AON

74 M. Kawaguchi