No. 2] Proc. Jpn. Acad., Ser. B 94 (2018) 59

Review Exploring peptide hormones in plants: identification of four peptide hormone-receptor pairs and two post-translational modification enzymes

† By Yoshikatsu MATSUBAYASHI*1,

(Communicated by Shigekazu NAGATA, M.J.A.)

Abstract: The identification of hormones and their receptors in multicellular organisms is one of the most exciting research areas and has lead to breakthroughs in understanding how their growth and development are regulated. In particular, peptide hormones offer advantages as cell-to- cell signals in that they can be synthesized rapidly and have the greatest diversity in their structure and function. Peptides often undergo post-translational modifications and proteolytic processing to generate small oligopeptide hormones. In plants, such small post-translationally modified peptides constitute the largest group of peptide hormones. We initially explored this type of peptide hormone using bioassay-guided fractionation and later by in silico gene screening coupled with biochemical peptide detection, which led to the identification of four types of novel peptide hormones in plants. We also identified specific receptors for these peptides and transferases required for their post- translational modification. This review summarizes how we discovered these peptide hormone– receptor pairs and post-translational modification enzymes, and how these molecules function in plant growth, development and environmental adaptation.

Keywords: secreted peptide, cell-to-cell communication, post-translational modification, Arabidopsis, ligand, receptor

During the past 20 years, biochemical, genetic, 1. Introduction and bioinformatic analyses have identified more than Cell-to-cell signaling mediated by hormones and a dozen secreted peptide hormones and their recep- membrane-localized receptors is one of the essential tors in plants.1)–4) These peptide hormone–receptor mechanisms by which the growth and development pairs have proven to be functionally more diverse of multicellular organisms are regulated. Upon bind- than anticipated. Some of these peptides act as ing of such hormones to the extracellular domains local signals during plant growth and development, of receptors, physicochemical interactions are con- whereas others are root-to-shoot long-distance signals verted into physiological outputs activating down- required for environmental adaptation. The number stream signaling, which modulates cellular functions of functionally characterized peptide hormones now and fates through conformational changes in the exceeds the number of classical plant hormones. receptors. Because membrane-localized receptors act Secreted peptide hormones can be divided into as master switches of complex intracellular signaling two major groups based on structural characteristics processes, the identification of hormone–receptor arising from their biogenesis pathways (Fig. 1). One pairs is one of the central issues of current biological major group of peptide hormones are small post- research. translationally modified peptides characterized by the presence of post-translational modifications *1 Division of Biological Science, Graduate School of Science, mediated by specific transferases and by their small Nagoya University, Nagoya, Japan. – † size (approximately 5 20 amino acids) after proteo- Correspondence should be addressed: Y. Matsubayashi, lytic processing. The second group comprises cys- Division of Biological Science, Graduate School of Science, Nagoya University, Chikusa, Nagoya 464-8602, Japan (e-mail: matsu@ teine-rich peptides characterized by the presence of bio.nagoya-u.ac.jp). an even number of cysteine residues (typically 6 or 8) doi: 10.2183/pjab.94.006 ©2018 The Japan Academy 60 Y. MATSUBAYASHI [Vol. 94,

Peptide hormone gene

Prepropeptide Signal peptide

Propeptide CC C C CC

Post-translational modification Disulfide bond formation Proteolytic processing (Processing)

X X C Mature peptide C C C C C

Small post-translationally modified peptide Cysteine-rich peptide

Fig. 1. Two distinct biogenesis pathways of secreted peptide hormones categorized by their structural characteristics. Peptide hormones can be categorized into the following two groups: peptides with complex post-translational modifications followed by proteolytic processing and peptides with multiple intramolecular disulfide bonds. The first group of peptides are called small post-translationally modified peptides and the latter group are defined as cysteine-rich peptides. This scheme is adapted from our previous review articles.1),73) that participate in the formation of intramolecular terminally encoded peptide (CEP), involved in disulfide bonds. In both cases, peptide hormone genes long-distance nitrogen demand signaling,8) root are initially translated as biologically inactive pre- meristem growth factor (RGF), regulating root propeptides, followed by removal of the N-terminal meristem development,9) and Casparian strip integ- signal peptide by a signal peptidase to afford a rity factor (CIF) required for contiguous Casparian propeptides. Propeptides are further structurally strip formation.10) These critical peptide hormones modified by several enzymes to give biologically had long been overlooked, probably due to their gene functional mature peptides. redundancy. We also identified receptors involved in My research career started with the purification the perception of these peptide hormones and two of a chemical factor involved in the density effect of important transferases required for post-translational plant cell proliferation in vitro as a Ph.D. student at modification of the hormones.11),12) Nagoya University in the laboratory of Prof. Youji This review offers a personal overview of how Sakagami. This bioassay-guided approach led to the we discovered these peptide hormone–receptor pairs identification of a peptide phytosulfokine (PSK), and post-translational modification enzymes, and the first small post-translationally modified peptide how these molecules contribute to plant growth hormone found in plants.5) Since then, I have been and development. Information regarding other small fascinated with the question of to what extent post-translationally modified peptides and cysteine- peptide signaling plays a role in plant growth and rich peptides is reviewed elsewhere.1)–4) development. The major challenge in this research is, fi however, how to distinguish bona fide peptide 2. Novel approaches for the identi cation of hormones from the numerous unrelated peptides peptide hormones and receptors in plants and protein fragments present in extracellular spaces. 2.1 In silico screening for peptide hormone Additionally, because no one can predict the activ- candidates. After our identification of PSK and ities of undiscovered hormones, a conventional bio- its family of precursor polypeptides by conventional assay-guided approach is not applicable. bioassay-guided purification (described in sec- To this end, my group employed an in silico tion 3.1), we noticed several structural character- gene screening approach coupled with structural istics of the amino acid sequences within this family, determination of mature peptides6) and receptor as summarized in Fig. 2A. (a) These precursor identification using a receptor expression library.7) polypeptides were approximately 100 amino acids This molecular-oriented strategy led to the identi- in length and had N-terminal secretion signal fication of three peptide hormones, namely, C- sequences that can be detectable using public web- No. 2] Exploring peptide hormones in plants 61 based software. (b) The hormones (mature peptides) 2.2 LC-MS-based structural elucidation of were encoded near the C-terminal region of the mature peptides. Small peptide hormones are precursor. Moreover, amino acid sequences corre- generated from precursor polypeptides by post-trans- sponding to the mature peptide domain were highly lational modification and proteolytic processing. conserved within the family, but other domains Because such specific modification and trimming exhibited low sequence conservation. This observa- actions are critical for the activity of each peptide tion can be interpreted as functional mature peptide hormone, determination of the mature functional regions being under strong selective pressure and structures of the peptide hormones is indispensable tending to exhibit higher sequence conservation for detailed analysis of peptide signaling. than their neutral flanking regions. (c) The mature To detect the mature peptide generated from peptide was post-translationally modified. Because target genes of interest, it is essential to extract post-translational modifications such as sulfation apoplastic peptides lacking contamination with and glycosylation require co-substrates that contain cytoplasmic molecules that hamper LC-MS-based high-energy phosphate bonds, the biosynthesis of structural analysis, even when the target peptide post-translationally modified peptides requires con- is overexpressed. To this end, we established a siderably more energy than the biosynthesis of other submerged culture system, in which Arabidopsis peptides. Nevertheless, post-translationally modified seeds are directly sown and grown in liquid culture peptides have been evolutionarily conserved, suggest- medium.6) In these conditions, plants develop hyper- ing that these modified peptides offer greater hydric true leaves, which are characterized by a lack physiological benefit to plants than the energy cost of cuticular wax formation. Hyperhydric leaves have for their biosynthesis. In this context, post-transla- vacuolated mesophyll cells with large intercellular tional modifications can be indicative of hormones. spaces filled with water instead of air. Therefore, (d) Genes encoding peptide hormones may exist as secreted peptides present in the intercellular apo- a family. The PSK family consists of five members plastic spaces diffuse into the culture medium with- in Arabidopsis and six in rice. These predictions out artificial manipulation. Peptides and proteins in were strengthened by the identification of additional the submerged culture medium can be extracted by peptide hormones such as the CLAVATA3/CLE phenol extraction followed by acetone precipitation, peptide family,13),14) in which mature peptides are which effectively remove secondary metabolites and also encoded in the C-terminal domain, which is polysaccharides. Liquid chromatography–tandem conserved among 32 members. mass spectrometry (LC-MS/MS) analysis of the Based on this empirical rule, we hypothesized resulting samples permitted highly reproducible and that if a family of secreted peptides in Arabidopsis reliable identification of the mature peptide gener- shares a conserved domain near the C-terminus by ated from each target gene expressed under the in silico analysis and the conserved domain is indeed control of a constitutive promoter. confirmed to be a part of the secreted mature pep- 2.3 Receptor identification using receptor tide after post-translational modification by liquid kinase expression library. Identification of the chromatography-mass spectrometry (LC-MS)-based receptors for hormones is a key step to understand structural analysis (described in section 2.2), this the molecular mechanisms underlying signal trans- family may encode functional peptide hormones. To duction. In addition, functional analysis of hormone test this idea, we performed computational screening signaling pathways from the receptor side is often an of Arabidopsis genes encoding secreted polypeptide effective way to overcome genetic redundancy of the families that fulfilled the above criteria.6) We down- peptide ligands that frequently exist as multi-gene loaded all Arabidopsis protein sequences (approx. families in Arabidopsis. Two approaches may be 35,000) from the TAIR website and extracted applicable for receptor identification: biochemical potential families with primary translated products binding assays using receptor proteins or hormone- of approximately 100-amino-acid secreted polypep- insensitive assays using loss-of-function receptor tides that share short, conserved domains near their mutants. We employed the former approach, because C-terminus. This in silico screening enabled us to specific ligand binding activity is the most critical obtain at least three major functionally uncharac- property of the receptors. terized peptide families that we later found to be The Arabidopsis genome contains 625 receptor critical peptide hormones in plant growth and kinases (RKs) that are thought to function development. as membrane-localized receptors for extracellular 62 Y. MATSUBAYASHI [Vol. 94, ligands including peptide hormones.15) From the cell suspensions incubated at low density, and functional point of view, RKs can be divided into identified a small peptide as a growth-promoting two groups, namely ligand-binding receptors, which signal involved in the above-mentioned density effect directly interact with ligands, and co-receptors, in plant cell cultures of Asparagus, carrot and which form heteromers with ligand-binding receptors Zinnia.5) This peptide is composed of only 5 amino to modulate signaling. Cumulative evidence suggests acids including two sulfated tyrosine residues, and that the extracellular domain of ligand-binding we named this factor PSK (Fig. 2A).5) PSK was the receptors is quite large and mostly exceeds 400 amino first example of a small post-translationally modified acid residues. In addition, almost all genes encoding peptide hormone in plants. In vitro experiments ligand-binding RKs contain no introns within the showed that PSK has pleiotropic effects on cell gene regions corresponding to the extracellular growth and differentiation such as stimulating cell domain. Based on these empirical rules, we selected proliferation in Asparagus and rice,5),20) promoting 9100 RKs that fulfill the above criteria and tracheary element differentiation in Zinnia17),21) and individually overexpressed them in conventional enhancing somatic embryogenesis in carrot and tobacco suspension cells to establish an RK expres- Japanese cedar (Cryptomeria japonica).19),22),23) sion library.7) Photoaffinity labeling using photo- Based on cDNA cloning experiments, we dem- activatable derivatives of peptide hormones and onstrated that PSK is produced from an approx- screening of the RK expression library enabled us imately 80 amino acid precursor polypeptide via to identify peptide ligand–receptor pairs with high tyrosine sulfation and proteolytic processing.24),25) accuracy within a short period.8)–10),16) Five paralogous PSK genes are present in the Arabidopsis genome. PSK precursor polypeptides 3. Peptide hormone-receptor pairs share a conserved domain close to the C-terminus, 3.1 Phytosulfokine and PSK receptor. PSK from which mature functional PSK peptides are was identified by conventional bioassay-guided puri- generated (Fig. 2A). This domain structure of the fication and provided an initial clue to small post- PSK precursor polypeptides gave us a clue for an translationally modified peptide hormones in plants. in silico search for novel peptide hormones (described Several structural characteristics of the amino acid later). PSK genes are widely expressed in a variety sequences within PSK precursor polypeptides of tissues in Arabidopsis, and their expression is prompted the idea of searching for still-undiscovered upregulated by wounding or interaction with micro- peptide hormones by in silico screening as described organisms.24),26),27) PSK homologs are present in both in section 2.1. monocots and dicots as small gene families.28) The relative growth rates of plant cells in At the time we identified PSK peptide, little suspension are often greatly affected by the initial information was available on receptor candidates for cell population density. Dilution of isolated meso- peptide hormones in plants. Therefore, we searched phyll cells of Asparagus officinalis in excess culture for the PSK receptor using affinity purification. medium suppressed cell division.5) A similar phe- Initial assessment of PSK binding sites in several nomenon was also reported for transdifferentiation established cell lines suggested that rice Oc and of mesophyll cells of Zinnia elegans into xylem cells carrot NC cells express a relatively high number of termed tracheary elements.17),18) If Zinnia mesophyll proteins binding PSK on their membranes.20),29),30) cells in suspension are cultured below a critical cell Accordingly, we purified the PSK binding protein density, the frequency of tracheary element forma- from solubilized membrane fractions of carrot NC tion is greatly decreased. Somatic embryo formation cells using an affinity column on which PSK was from carrot cells in suspension also depends on cell immobilized. The carrot PSK receptor (PSKR) density.19) Of note, this density effect is alleviated by turned out to be a member of the leucine-rich repeat the addition of culture medium in which cells were receptor kinase (LRR-RK) family,31) which was later previously grown. This circumstantial evidence sug- recognized as a major receptor family involved in gests that a chemical factor secreted from individual the perception of peptide hormones.32) The crystal cells affects the potential for cellular proliferation and structure of the PSK-PSKR complex was later differentiation. resolved by Chai and colleagues.33) We also found We purified this growth-promoting signal from that Arabidopsis has two PSKR orthologs, which we the culture medium using a bioassay system that named AtPSKR1 and AtPSKR2; the former is the assessed cell division activity of Asparagus mesophyll major component required for PSK perception.24) No. 2] Exploring peptide hormones in plants 63

A Phytosulfokine (PSK)

At1g13590 (PSK1) 1 ----MKTKSEVLIFFFTLVLLLSMASSVILREDGFAPPKPSPTTHEKASTKGDRDGVECKNSDSEEEC-LVKKTVAA-HTDYIYTQDLNLSP 86 At2g22860 (PSK2) 1 MANVSALLTIALLLCSTLMCTARPEPAISISITTAADPCNMEKKIEGKLDDMHMVDENC-GAD-DEDC-LMRRTLVA-HTDYIYTQKKKHP- 87 At3g44735 (PSK3) 1 ------MKQSLCLAVLFLILSTSSSAIRRGKEDQEINPLVSATSVEEDSVNKLMGMEYCGEGDEEC-LRRRMMTESHLDYIYTQHHKH-- 81 At3g49780 (PSK4) 1 ------MGKFTTIF-IMALLLCSTLTYA-ARLTPTTTTALSRENSVKEIEGDKVEEESCNGIG-EEEC-LIRRSLVL-HTDYIYTQNHKP-- 79 At5g65870 (PSK5) 1 ------MVKFTTFLCIIALLLCSTLTHASARLNPTSV--YPEENSFKKLEQGEV---ICEGVG-EEECFLIRRTLVA-HTDYIYTQNHNP-- 77

Tyr(SO3H) IleTyr(SO3H) Thr Gln

B C-terminally encoded peptide (CEP) At1g47485 (CEP1) 1 ------MGMSNRSVSTSIFFLALVVLHGIQDTEERHLKTTSLEIEGIYKKTEAEHPSIVVTYTRRGVLQKEV 66 At1g59835 (CEP2) 1 MKLFIITVVTILTISRVFDKTPATTEARKSKKMVGHEHFNEYLDPTFAGHTFGVVKEDFLEVKKLKKIGDENNLKNRFINEFAPTNPEDSLGIGHPRVLN 100 At2g23440 (CEP3) 1 ------MATINVYVFAFIFLLTISVGSIEGRKLTKFTVTTSEEIRAGGSVLSSSPPTEPLESPP 58 At2g35612 (CEP4) 1 ------MVSRGCSITVLFRFLIVLLVIQVHFENTKAARHAPVVSWSPPEPPKDDFVWYHKINRFK 59 At5g66815 (CEP5) 1 ------MESFMGQKKTLYACYFLMLVFFLGFNCVHGRTLKVDDKINGGHYDSKTMMALAKHNDMMVDDKAMQFSPPPPPPPPSQSG 80 At2g23445 (CEP7) 1 ------MAKTRRVIYLFLTIVLLFCELIDEAQGSRFRCHHSEDYSCKKRSSHHHHHHHHHQQQQHHHKDTPPEELQGSIKTRRSKD 80 At5g66816 (CEP9) 1 ------MKLSVYIILSILFISTVFYEIQFTEARQLRKTDDQDHDDHHFTVGYTDDFGPTSPGNSPGIGHKMKENEENA 72 At5g66818 (CEP10) 1 ------MAKCTLTSLILLLIVLVLIQESHIVEGRPLKSSRISNVSKKFAAGNSNLSSKLTTE 56 At5g66819 (CEP11) 1 ------MAKALFFNFCISLLIIAILVSHEIIPTEARHLRTHRKSIKNSTLTVHEGAGGLRTGGGSVKTDISKE 67

At1g47485 (CEP1) 67 IAHPTDFRPTNPGNSPGVGHSNGRH---- 91 At1g59835 (CEP2) 101 NKFTNDFAPTNPGDSPGIRHPGVVNV--- 126 At2g23440 (CEP3) 59 SHGVDTFRPTEPGHSPGIGHSVHN----- 82 At2g35612 (CEP4) 60 NIEQDAFRPTHQGPSQGIGHKNPPGAP-- 86 At5g66815 (CEP5) 81 GKDAEDFRPTTPGHSPGIGHSLSHN---- 105 At2g23445 (CEP7) 81 IYGLNAFRSTEPGHSPGVGHLIKT----- 104 At5g66816 (CEP9) 73 GGYKDDFEPTTPGHSPGVGHAVKNNEPNA 101 At5g66818 (CEP10) 57 DHSLDAFRPTNPGNSPGIGH------76 At5g66819 (CEP11) 68 EHGVDEFRPTTPGNSPGIGH------87

Asp Phe Arg Hyp Thr Asn Pro Gly Asn Ser Hyp Gly Val Gly His

C Root meristem growth factor (RGF) At5g60810 (RGF1) 1 ------MVSIRVICYLLVFSVLQVHAKVSNANFNSQAPQMKNSEGLGASNGTQIAKKHAEDVIENRKTLKHVNVKVEAN 73 At1g13620 (RGF2) 1 ------MTNITSSFLCLLILLLFCLSFGYSLHGDKDEVLSVDVGSNAKVMKHLDGDDAMKKAQ--VRGRSGQEFSKETTK 72 At2g04025 (RGF3) 1 ------MTTL-SKILCVLIILLLCFSFRYSLHEDGNQQSSRDFVSTAKAIKYGDVMKKMIRGRK-LMMASGEK-EEAETK 71 At3g30350 (RGF4) 1 ---MRFTIIVIAFLLIIQSLEEEQILVYARKGREACHKSLDYQGDQDSSTLHPKELYDAPRKVRFGRATRAEKEQVTAMNNDSWSFKISGASKHLIVERK 97 At5g51451 (RGF5) 1 ------MSSIHVASM-ILLLFLFLH-HSDSRHLDNVHITAS-RFSLVK-DQNVVS-S 46 At4g16515 (RGF6) 1 ------MSCSLRSGLVIVFCFIL-LLLSSNVGCASAARRLRSHKHHH 40 At3g02240 (RGF7) 1 ------MEMKKWSYANLITLALLFLFFIILLLAFQGGSRDDDHQHVHVAIRTKDISMGRKLKSLK 59 At2g03830 (RGF8) 1 ------MKLIRVTLFLCALAILLLVTPTS-SLQL-KHPYSSPSQGLSKKIVTKMATRKLMIISSEYVMTSTSHEGSSEQLR 73 At5g64770 (RGF9) 1 ------MAIRVSHKSFLVALLLILFISSPTQARSLR-EVVRN 35 At3g60650 (RGF10) 1 MDMLRSACFYFLLIVFVILSWSLLCDSRHLGHMEKKLSVNLDLLNKDNEEITKLEAPSTNKTNTLLSQSHAVVNHGDNGQINGKKTKEIHRVKRASDKKV 100

At5g60810 (RGF1) 74 EKNGLEIESKEMVKKRKNKKRLTKTESLTADYSNPGHHPPRHN------116 At1g13620 (RGF2) 73 MMM--K-KTTKKETNVE-EED-DLVAY-TADYWKPRHHPPKNN------109 At2g04025 (RGF3) 72 MKR-GN-RETERNSSKSVEED-GLVAY-TADYWRAKHHPPKNN------110 At3g30350 (RGF4) 98 LGFHKRSKSSSFKWKPKKKKSSGPFVAFYDDYRGPARHPPRHNL------141 At5g51451 (RGF5) 47 STSKEPVKVSRFVPGPLKHHHRRPPLL-FADYPKPSTRPPRHN------88 At4g16515 (RGF6) 41 HKVASLDVFNGGERRRALGGVETGEEVVVMDYPQPHRKPPIHNEKS---- 86 At3g02240 (RGF7) 60 PINPTKKNGFEYPDQGSHDVQEREVYVELRDYGQRKYKPPVHN------102 At2g03830 (RGF8) 74 VTSSGKSKDEEKKLSEEEEEKKALAKYLSMDYRTFRRRRPVHNKALPLDP 123 At5g64770 (RGF9) 36 RTLLVVEKSQESRKIRHEGGGSDVDGLMDMDYNSANKKRPIHNR------79 At3g60650 (RGF10) 101 SSKRVSRTWKIPKYPKKQPKSDQEHPGFNLDYMQPTTHPPHHN 143

Asp Tyr(SO3H) Ser Asn Pro Gly His His Pro Hyp Arg His Asn

D Casparian strip integrity factor (CIF)

At2g16385 (CIF1) 1 ------MGMSPLTVKKLGFIFMIVSASALSVSFAGRPSIFVHKKINLREEMVERSMHEHERLLRMNTKDYGNNSPSPRLERPPFKLIPN 83 At4g34600 (CIF2) 1 ------MGLLPLVKKLGFIIFLLVSASAFALCSAGRSSILIYSQEDDHPEVVERRIHEHERILRMNSRDYGHSSPKPKLVRPPFKLIPN 83 MTR_8g037800 1 ------MDFMFLKKFTLLFLLISGSLLTTSFAGRASNFIRISNEDVNAVHEVTTKMAMNEEEVRSIHERLLRANTKDYGRYDPSPTFSKPPFKLIPN 91 POP09s12270 1 ---MGLMLLKKISLLFLLISASFLSTSFAGRRSKSVNKLAEEVEVSAATYEEISSKPSHNNEATTIHERLLKANTKDYGNYKPAPALVRPPFKLIPN 94 Os01g0260800 1 MGAVKRSSFLLVVVVFALLLLTSMAAGGRKMLINKHQVQSMETSDDESMHQGQEDDEMLAMVHER---ILRQVKTNDYGTYDPTPTMAKPHAKEIPN 94 Sb03g010100 1 ------MEPRKSNYLVALVLASLLLSAMAGGHRKRLLNKDDASESMETSESMQQLQEDDEMAVVVHERILRQVKMNDYGRYDPSPTMAKPHFKDIPI 91 Sb0010s013050 1 ------MVLVLVLLLIVLPALSIGGRELADEKDHNKQHSTAASEKGATASEDMVKTNDYGRYDPSPAFSKPRFKLIPN 72 Bradi2g09500 1 MEINKSSCNYLVAFFFAALILSSMVAGAQRKLLDQDQVGSMETSESSTEQLQQEDEEVLVVALHGGRILRQVKNTNDYGTYDPSPTMAKPHFKDIPN 97 Bradi3g11170 1 -----MHCVSSPEMRMQRRSTALFALVFTLLLSTSLAGRQRSLLTDQESLGQQAEGTEAGQDEAVHARMLKAVATSDYGSYDPSPSMEKPHFKLIPN 92

Asp Tyr(SO3H) Gly Asn Asn Ser Hyp Ser Hyp Arg Leu Glu Arg Pro Pro Phe Lys Leu Ile Pro Asn

Fig. 2. Structural characteristics of the primary amino acid sequences of precursor polypeptides that generate small post-translationally modified peptide hormones. Shown are the deduced amino acid sequences of the (A) PSK; (B) CEP; (C) RGF; and (D) CIF families. Domains encoding the mature peptides are underlined, and experimentally elucidated mature peptide structures are shown below. Identical amino acid residues are highlighted in black, and similar amino acid residues are highlighted in gray. 64 Y. MATSUBAYASHI [Vol. 94,

The pskr1 mutant exhibits premature senescence and together with loss-of-function phenotypes similar to gradually loses the potential to form callus as tissue N starvation responses, suggested that the CEP– matures. This mutant was later shown to be defective CEPR system may be involved in long-distance in innate plant immune responses triggered upon the signaling regulating N acquisition. Related to this, perception of elicitors released by pathogens, in root it was first reported in the 1970s that N starvation and hypocotyl elongation, and in Agrobacterium- of a portion of the root system in a heterogeneous induced tumor growth (reviewed in 27)). Thus, soil N environment can upregulate nitrate uptake in a available evidence indicates that PSK signaling distant part of the roots exposed to a N-rich medium, affects cellular longevity and potential for growth, compensating for N deficiency.36) This long-distance and thereby exerts a pleiotropic effect on growth systemic response was postulated to be controlled and development in response to environmental by a N-demand signal emitted from the N-starved conditions. roots,37)–39) and ultimately led to the idea that CEP 3.2 C-terminally encoded peptide and CEP family peptides act as the long-sought N-demand receptor. The above-mentioned in silico gene signal (Fig. 3A, magenta arrow). screening identified at least three families of peptide Based on this hypothesis, we analyzed whether hormone candidates that contain family-specific the CEP–CEPR system is involved in long-distance conserved domains near the C-terminus (Fig. 2B– N-demand signaling by studying CEP expression D). We determined that one of these candidate genes patterns in a N-starved environment, together with encodes a polypeptide that generates a 15 amino cepr mutant phenotypes in split-root conditions, in acid mature peptide with two hydroxyproline which the root system is separated into two parts (Hyp) residues, and named it and its homologs the exposed to different nutrient conditions (Fig. 3B). CEP family after their structural characteristics We also examined CEP translocation from roots (Fig. 2B).6) A total of 15 CEP family genes have to shoots, and finally concluded that this ligand– been found in the Arabidopsis genome.6),34),35) CEP receptor pair is a critical component for the first genes are present in both dicots and monocots but are half of the systemic N-demand signaling pathway absent in lower land plants such as moss and green (Fig. 3A). CEP family peptides are induced in the algae.34) Expression of CEP genes is mainly found portion of the root system directly experiencing N in vascular tissues of the lateral roots, suggesting a starvation and act as root-derived ascending N- role in roots, but severe genetic redundancy obscured demand signals transported through the xylem to the functions of the CEPs at that time. the leaves (Fig. 3A, magenta arrow), where they are We overcame genetic redundancy on the ligand recognized by CEPR localized on the phloem side in side by identifying receptors for CEP by exhaustive the vasculature of the leaves.8) Perception of CEP by photoaffinity labeling using the RK expression CEPR triggers the production of a shoot-derived library.8) In contrast to multiple redundancy on the descending signal (described later) that upregulates ligand side, only two RKs were found for the CEP nitrate transporter genes in the distant part of the receptor, which we named CEPR1 and CEPR2. roots to compensate for local N deficiency. These two RKs belong to LRR-RK subfamily XI. Identification of the ascending CEP family Expression of the major receptor CEPR1 was not peptides raised the question of the molecular identity limited to roots but rather was predominantly of the descending shoot-derived secondary signal. detected in the vascular veins of cotyledons and We searched for this signal by mechanically isolating mature leaves. The receptor double mutant ulti- vascular tissues, in which CEPR is expressed, from mately revealed loss-of-function phenotypes of CEP wild-type, CEP-treated, and cepr mutant plants, signaling, which were characterized by pale-green followed by transcriptome analysis using a micro- leaves, enhanced lateral root elongation, and shorter array system. This approach identified several CEP- stems accompanied by anthocyanin accumulation. inducible genes that encode small phloem-specific These phenotypes are reminiscent of typical re- polypeptides. We found that, when overexpressed, sponses to nitrogen (N) starvation, and indeed were two of them led to upregulation of the nitrate accompanied by a considerable reduction in the transporter gene NRT2.1 in roots and accordingly expression of high-affinity nitrate transporter gene named them CEP downstream (CEPD) polypep- NRT2.1 in roots. tides.40) CEPDs are approximately 100 amino acid These spatially distant expression patterns of polypeptides that belong to a plant-specific, non- CEP family peptide ligands and the receptor CEPR1, secreted peptide family.41) Of note, although CEPD No. 2] Exploring peptide hormones in plants 65

ABReceptor CEPR CEP CEPD

– 3 NO CEP induction

Compensatory N uptake Local N starvation

Fig. 3. Schematic representation of long-distance N-demand signaling mediated by the CEP-CEPR-CEPD system. (A) CEP family peptides are induced by N starvation and transmit to the shoots that roots require nitrogen. Recognition of CEP by the specific receptor CEPR in shoots leads to the generation of a shoot-to-root secondary signal CEPD that promotes N-uptake in distant parts of the roots to compensate for local N starvation. (B) Split-root culture systems where the root system of a plant is separated into two parts (left and right in image), which are then exposed to different nutrient conditions. Image is adapted from our previous report.8) genes are expressed exclusively in leaf phloem, CEPD mobile ascending peptides as “hunger” signals and polypeptides are detected in the root vascular region, phloem mobile descending peptides as “nutrient indicating that CEPDs function as shoot-to-root intake-facilitating” signals (Fig. 3A).42) These funda- mobile signals (Fig. 3A, blue arrow). mental mechanistic insights should provide a con- Our final question was how CEPD polypeptides ceptual framework for understanding systemic long- induce NRT2.1 expression, specifically on the side of distance root-to-shoot-to-root signaling in plants. the root system exposed to the N-rich medium under 3.3 Root meristem growth factor and RGF heterogeneous N conditions. One possibility was that receptor. The initial clue that led to the discovery shoot-derived CEPD polypeptides are selectively of the RGF peptide family came from the identi- translocated to the root system of the N-rich side. fication of Arabidopsis tyrosylprotein sulfotransfer- This possibility was, however, ruled out by the ase (TPST) (described in section 4.1,11)), a post- observation that there was no apparent difference in translational modification enzyme required for tyro- the abundance of CEPDs in the roots even when sine sulfation. Because TPST is a single-copy gene, one side of the root system was starved of N. The the phenotype of a loss-of-function mutant of other possibility was that CEPDs are distributed to Arabidopsis TPST (tpst-1)reflects a deficiency in both sides of the root system but activate NRT2.1 the biosynthesis of all functional tyrosine-sulfated expression only in roots where nitrate is available. peptides. Of note, tpst-1 shows an extremely short- We tested this possibility using CEPD1-overexpress- root phenotype accompanied by a considerable ing plants and found that NRT2.1 expression was decrease in proximal meristem activity in the root specifically induced in roots exposed to the N-rich apical meristem,11),43) suggesting that at least one medium but not in N-starved roots of the split-root tyrosine-sulfated peptide regulates root meristem system. Thus, shoot-derived CEPD polypeptides development. Because proximal meristem activity of upregulate NRT2.1 expression in roots specifically tpst-1 was not recovered by treatment with known when nitrate is present in the soil, thereby compen- sulfated peptides such as PSK, we predicted that an sating for local N deficiency at the whole-plant as yet undiscovered sulfated peptide was indispen- level.40) sable for root meristem development. Plants, as sessile organisms, continuously face We searched for relevant tyrosine-sulfated pep- a complex array of environmental fluctuations. Our tides by phenotypic rescue experiments using syn- findings demonstrated that although plants do not thetic sulfated peptides nominated by in silico gene have a circulation system, like the heart and screening coupled with LC-MS-based structural circulatory system in animals, they have evolved determination. As stated in the previous section, sophisticated long-distance signaling mechanisms to in silico gene screening identified at least three respond to fluctuating environments using xylem- families of peptide hormone candidates that contain 66 Y. MATSUBAYASHI [Vol. 94, family-specific conserved domains near the C- We also uncovered that PLETHORA1 (PLT1) terminus. Notably, two of them encoded tyrosine- and PLT2 transcription factors are very proximal sulfated peptides and thus represent strong candi- molecular targets of RGF signaling. PLT genes, dates for novel hormones. which are specifically expressed in the stem cell area We found that one of these sulfated peptide in the root meristem, encode AP2-domain tran- families indeed rescued proximal meristem activity of scription factors that mediate patterning of the root tpst-1 at nanomolar concentrations and accordingly stem cell niche.49) PLT proteins display a gradient named it the RGF family9) (Fig. 2C). Mature RGF1 distribution, with maximal distribution in the stem is a 13 amino acid sulfated peptide derived from cell area, and this gradient is essential for main- the C-terminal conserved domain of the precursor tenance of the root stem cell niche and transit- polypeptides; in Arabidopsis, 9 members (later found amplifying cell proliferation.50) However, the funda- to be 1144)) belong to this group. RGF1 is specifically mental molecule that defines this gradient has been expressed in quiescent center cells and columellar elusive. stem cells in the root tip.9) An additional 4 RGF In wild-type seedlings, PLT proteins show family members are also expressed in the root stem gradients that extend into the region of transit- cell area, suggesting redundant roles in the root amplifying cells or the elongation zone. However, apical meristem. Indeed, rgf1 rgf2 rgf3 triple mutants in the rgf1 rgf2 rgf3 ligand triple mutant, or rgfr1 show a short-root phenotype characterized by a rgfr2 rgfr3 receptor triple mutant, PLT gradient decrease in the number of meristematic cells.9) dimensions were considerably reduced. Conversely, Because the C-terminal sequence of CLE18 is similar exogenous application of RGF to the rgf1 rgf2 rgf3 to RGF1, some researchers later named the RGF mutant causes drastic enlargement of the PLT family the “CLE-like” family.45) Additionally, because expression domain in the mutant to levels even overexpression of RGF family peptides often causes higher than in wild-type plants. Importantly, because irregular root waving (“golven” in Dutch), some externally added RGF1 restores PLT protein ex- researchers refer to this as the GOLVEN family.46) pression patterns without major changes in PLT gene RGF genes are also found in the rice and poplar expression, this response is not at the transcriptional genomes, suggesting evolutionarily conserved roles of level but rather at the protein level, possibly through RGF across the plant kingdom.46) the stabilization of PLT proteins. We identified receptors for RGF using a custom- Collectively, RGF peptides that are secreted made RK expression library. Exhaustive photoaffin- from the stem cell region create a diffusion-based ity labeling revealed that 3 LRR-RKs in subfamily XI concentration gradient extending shootward from the specifically interact with RGF family peptides.7) We stem cell area in roots9) (Fig. 4). This RGF peptide observed a considerable decrease in meristematic cell gradient is reflected as a PLT protein gradient by number in the triple mutant to a level similar to that the membrane receptors, RGFRs, and allows correct of the tpst-1 mutant, accompanied by insensitivity pattern formation in the proximal meristem. Despite to RGF, and named these receptor proteins RGFR1, its emerging importance, RGF signaling has been RGFR2, and RGFR3. RGFR1 and RGFR2 are overlooked due to the high degree of genetic predominantly expressed in the proximal meristem redundancy both in ligands and receptors. The including the elongation zone and gradually decrease gradient of RGF is defined by the diffusion coefficient in expression towards the differentiation zone. In of the peptides, which is a physical parameter solely contrast, RGFR3 promoter activity is detected in proportional to the molecular weight, independent of the more basal region of the elongation zone and environmental fluctuations. Regulation of PLT ex- the differentiation zone. Unexpectedly, RGFR2 is pression patterns by a simple diffusion-based RGF also known as ROOT CLAVATA-HOMOLOG1 peptide gradient is a sophisticated system to ensure (RCH1); its promoter is often used to express robust root growth and development in fluctuating transgenes specifically in the root meristem.47) No natural environments. one, however, had noticed the fundamental functions 3.4 Casparian strip integrity factor and CIF of this receptor family due to functional redundancy receptor. Because nutrients often accumulate until our analysis focused on direct ligand binding against a concentration gradient in the root xylem properties. The crystal structure of the RGF1- vessels, vascular plants have evolved a physical RGFR1 complex was later resolved by Chai and barrier that prevents passive apoplastic diffusion of colleagues.48) ions and water across the endodermal cells surround- No. 2] Exploring peptide hormones in plants 67

Differentiation A zone BCD

Elongation zone

Proximal meristem

Quiescent center Columella stem cell

Fig. 4. RGF family peptides regulate root meristem development through the PLT pathway. (A) Schematic representation of the Arabidopsis root. (B) Whole-mount in situ hybridization of RGF1 mRNA. RGF genes are specifically expressed in the stem cell region in the root tip. (C) Whole-mount immunostaining of wild-type roots using an anti-RGF1 antibody. RGF peptides create a diffusion- based concentration gradient extending shootward from the stem cell area in roots. (D) Root meristem of wild-type seedling expressing PLT2-GFP. The RGF peptide gradient is reflected as a PLT transcription factor protein gradient and allows correct pattern formation in the proximal root meristem. Images B and C are adapted from our previous report.9)

Arabidopsis has two peptides that belong to this peptide family (Fig. 2D). We determined that their Extracellular space mature structures are 21 amino acid long and contain Water and ions one residue of sulfated tyrosine and two residues Endodermal cells of Hyp.10) One peptide is expressed in the stele, Casparian strip especially at the phloem pole, of the mature region of the primary roots, and the other one is in the root ff Stele stele in the elongation and di erentiation zones of both primary and lateral roots. We also found by Xylem exhaustive photoaffinity labeling that this peptide family directly binds to GSO1/SGN3 and its closest Fig. 5. Schematic cross-section of roots showing the Casparian homolog GSO2. Importantly, it has been reported strip. The hydrophobic barrier that seals the extracellular / apoplastic space between neighboring endodermal cells is called that GSO1 SGN3 is expressed in root endodermal the Casparian strip (shown in red). The Casparian strip prevents cells and its loss-of-function mutations result in the passive apoplastic diffusion of ions and water across the formation of a repeatedly interrupted, discontinuous endodermal cells (solid blue arrows). Essential mineral ions Casparian strip. These findings indicated that these move across the endodermis by symplastic transport (dashed two sulfated peptides act as ligands for GSO1/ blue arrows). CIF peptides expressed in the stele are required for contiguous Casparian strip formation. This illustration is SGN3 and GSO2 receptors to regulate contiguous modified from a figure in a previous report.74) Casparian strip formation in roots. We confirmed that the double mutant devoid ing the vascular bundles. This hydrophobic barrier of these two peptide genes was a phenocopy of the that seals the extracellular spaces between neighbor- receptor mutant, and external application of the ing endodermal cells is called the Casparian strip51) synthetic peptides restored contiguous Casparian (Fig. 5). The third peptide hormone candidate family strip formation in the mutant roots.10) Therefore, that we identified by in silico screening was found we called this family of peptides CIF. Because CIF is to act in Casparian strip formation,10) because this a tyrosine-sulfated peptide, the tpst-1 mutant, which peptide family binds directly to an LRR-RK, lacks tyrosine sulfation enzymes, is also defective in GASSHO1 (GSO1)/SCHENGEN3 (SGN3),52) which Casparian strip formation. Independently, Geldner regulates the integrity of the Casparian strip. and colleagues identified tpst-1 in the course of 68 Y. MATSUBAYASHI [Vol. 94, screening for Casparian strip defective mutants.53) 4.1 Tyrosine sulfation. Tyrosine sulfation is His group further identified CIF peptides by pheno- a post-translational modification occasionally found typic rescue experiments using synthetic sulfated in peptides and proteins synthesized through the peptides selected by in silico gene screening.53) secretory pathway both in plants and animals.56) A The Casparian strip plays important roles in specific enzyme involved in this modification was first environmental adaptation by acting as a physical identified in mice and humans and named tyrosyl- barrier that prevents unfavorable leakage of ions protein sulfotransferase (TPST).57),58) Animal TPST between the xylem and the soil.54) In nutrient- is a type II membrane protein that has a large limiting conditions such as low potassium, the cif1-1 C-terminal catalytic domain oriented in the lumen of cif2-1 double mutant exhibits potassium deficiency the Golgi and a single transmembrane domain near symptoms that are far more severe than in wild-type the N-terminus. After public release of the mouse plants. Our xylem sap analysis revealed that the TPST sequence, many groups, including ours, potassium level in the xylem sap of the mutant was searched for its plant counterpart by BLAST analy- lower than that of the wild-type, due to concen- sis, but no ortholog was identified in the Arabidopsis tration-dependent outward leakage.10) We also found protein database. This suggested that plants evolved that the cif1-1 cif2-1 mutant is highly sensitive to plant-specific TPSTs with a primary structure excess iron. At normal iron concentrations, the iron distinct from that of animals. content in the xylem sap of the cif1-1 cif2-1 mutant In general, TPST catalyzes the transfer of a was virtually the same as that of the wild-type. In sulfate from a donor 3B-phosphoadenosine 5B-phos- contrast, the iron level in the xylem sap of the cif1-1 phosulfate to the phenolic group of tyrosine within cif2-1 mutant cultured with excess iron was consid- the acceptor peptide by forming an enzyme-substrate erably higher than the wild-type, probably due to ternary complex (Fig. 6A). Therefore, we immobi- inward leakage. Thus, Casparian strip mutants are lized the acceptor peptide on a column in an attempt defective in ion homeostasis in the xylem because of to purify plant TPST from solubilized Arabidopsis inward or outward leakage of ions depending on the membrane fractions by affinity chromatography. ionic concentration gradient across the endodermis However, the initial trial using PSK precursor cell layer, leading to pleiotropic phenotypes in peptide was unsuccessful, probably due to insufficient unfavorable mineral environments. affinity for purification. Another chance for affinity purification of TPST fi 4. Post-translational modi cation enzymes arose from the identification of PSY1, the second Secreted peptides move from cell to cell through tyrosine-sulfated peptide in plants.59) This peptide the extracellular apoplastic space solely by simple was found by peptidomics analysis in a highly acidic diffusion, with the diffusion coefficient dependent on fraction obtained by ion-exchange chromatography molecular size. Because smaller molecular size leads of the culture medium of Arabidopsis cells. PSY1 is to higher diffusion rates, shorter oligopeptides are an 18 amino acid peptide that exhibits cell division theoretically advantageous in cell-to-cell signaling stimulatory activity similar to that of PSK. We over larger proteins. Fewer residues in the peptide again immobilized the PSY1 precursor peptide on a chain, however, results in lower structural diversity column and attempted to purify Arabidopsis TPST. and specificity. To solve this paradox, multicellular Fortunately, this PSY1 column specifically adsorbed organisms likely evolved a post-translational mod- TPST activity, allowing enrichment of TPST protein ification system that alters the physicochemical to a level detectable by SDS-PAGE. It has been properties of peptides by changing their net charge reported that multiple acidic amino acids near the or conformation. Indeed, post-translational modifica- tyrosine residue, as is the case for PSY1, significantly tion is often critical for the physiological function and enhance sulfation,60) and thereby likely increases specific receptor binding activity of small peptide affinity for TPST. hormones. To date, three types of post-translational Arabidopsis TPST (AtTPST) was identified to modifications have been found in peptide hormones be a 62-kD transmembrane protein localized in the in plants: tyrosine sulfation, proline hydroxylation, cis-Golgi.11) As we hypothesized, AtTPST showed and hydroxyproline arabinosylation (reviewed in no sequence similarity with animal TPST and, more 55)). We identified key enzymes for tyrosine sulfa- surprisingly, was a type I membrane protein, which tion and hydroxyproline arabinosylation by affinity has a transmembrane domain near the C-terminus purification.11),12) and a large luminal catalytic domain toward the No. 2] Exploring peptide hormones in plants 69

SO3H A Tyrosylprotein OH O sulfotransferase (TPST)

R R R R 1 H 2 1 H 2 N N O O PAPS PAP Tyr Tyr(SO3H)

B Hyp O-arabinosyl OH transferase O R 1 R 1 N (HPAT) N HO HO O R 2 OH R 2

O O

Hyp OH UDP Hyp O-arabinoside O

HO O UDP OH

Fig. 6. Reaction scheme of post-translational modification enzymes for peptide hormones. (A) Tyrosylprotein sulfotransferase (TPST) catalyzes the transfer of a sulfate group from the donor 3B-phosphoadenosine 5B-phosphosulfate to the phenolic group of tyrosine in peptides. (B) Hydroxyproline arabinosyltransferase (HPAT) catalyzes the transfer of L-arabinofuranose (L-Araf ) from the sugar donor UDP-O-L-Araf to the hydroxyl group of Hyp residues in peptides.

N-terminus, the opposite topological orientation that the tpst-1 mutant showed hypersensitivity to compared with animal TPST. Plants and animals copper deficiency61) and defective formation of the are likely to have independently acquired enzymes Casparian strip in roots.53) A recent discovery of for tyrosine sulfation through convergent evolution. a sulfated CIF peptide that is indispensable for Nonetheless, AtTPST has a subtle footprint of a Casparian strip formation explains this phenotype sulfotransferase near the C-terminus sharing similar- well.10),53) ity with heparan sulfate 6-O-sulfotransferase 2 4.2 Hyp arabinosylation. Proline residues in (HS6ST2). It should be noted that AtTPST had secreted peptides and proteins are often hydroxylated been registered in the Arabidopsis protein database to Hyp, a post-translational modification found both before our discovery as a shorter protein lacking the in plants and animals. In plants, Hyp residues are HS6ST2 domain due to misassignment of the splicing occasionally modified further with three residues of site. This is probably why AtTPST escaped gene L-arabinose, giving rise to Hyp O-triarabinoside. This hunting by bioinformaticians for such a longer Hyp O-arabinosylation was originally discovered in period. extracellular structural proteins such as extensins,62) AtTPST is expressed throughout the plant, and and later found in several peptide hormones.55) the highest expression is in the root apical meristem. The occurrence of pentose residues in peptide A knockout mutant of AtTPST (tpst-1) displayed hormones was initially suggested in the course of pale green leaves and early senescence in the above- mass spectroscopic analysis of a defense peptide, ground tissues and had an extremely short root TobHypSys,63) but the molecular nature of the sugar phenotype accompanied by a considerable decrease moiety remained elusive. We detected pentose sugars in proximal meristem activity.11),43) Because AtTPST also in growth-promoting peptide PSY1 and deter- is a single-copy gene, the phenotype of the tpst-1 mined that the sugar moiety consisted of three mutant should reflect the deficiency in the biosyn- residues of L-arabinose.59) thesis of all functional tyrosine-sulfated peptides. The first structurally characterized Hyp O- Indeed, root defects associated with tpst-1 were arabinosylated peptide hormone was CLV3, which explained later as being caused by loss of sulfation has been shown to play a definitive role in regulating on RGF peptides that are required for maintenance stem cell populations in the shoot and floral of the root stem cell niche.9) It has also been reported meristems of Arabidopsis.64),65) We detected Hyp 70 Y. MATSUBAYASHI [Vol. 94,

O-arabinosylated CLV3 in the medium derived from A recent computational analysis of these sequences a submerged culture of Arabidopsis plants over- suggested that they may be conformationally similar expressing CLV3 during an exhaustive peptidomics to the glycosyltransferase GT8 family.69) search, and determined by gas chromatography– Our identification of Arabidopsis HPAT family mass spectrometry (GC-MS)-based linkage analysis arabinosyltransferases shed light on the function of that one of the Hyp residues is modified with O-1,2- their orthologs in leguminous plants such as Pisum linked triarabinoside. We also chemically synthesized sativum NOD3 and Medicago truncatula RDN1, arabinosylated CLV3 glycopeptide by intramolecular whose loss-of-function mutants show hypernodula- aglycone delivery and found that Hyp O-arabinosy- tion phenotypes.70) Although their enzymatic activ- lation enhanced both biological activity and receptor ities have yet to be confirmed, the fact that NOD3 binding activity by means of conformational alter- and RDN1 belong to the HPAT family strongly ations within the peptide backbone.66) supports our previous findings that Hyp O-arabino- The physiological importance of Hyp O-arabi- sylated CLE-RS2 peptide induced by rhizobia acts nosylation in peptide hormones was further demon- as an autoregulation signal to suppress excess strated by the discovery of the CLE-RS2 gene, which nodulation.68) controls the number of root nodules to achieve Similarly, tomato FIN, whose loss-of-function balance in the symbiotic relationship between legu- mutants show fasciated flowers associated with minous plants and rhizobia.67) This symbiosis enables enlarged shoot apical meristems, has also been nitrogen fixation in the nodules and is beneficial to proved to be an HPAT family protein.71) A collab- the host plants; however, excessive nodule formation orative team including our group found that this is deleterious to the plants because the energy cost phenotype was fully rescued by the in vitro treat- outweighs the need for fixed nitrogen. The CLE-RS2 ment of fin mutant apices with a synthetic Hyp O- gene is upregulated in roots upon rhizobial infection arabinosylated tomato CLV3 peptide.71) Thus, it is and suppresses excessive nodule formation via a likely that loss of FIN caused under-arabinosylation negative feedback loop. We determined that mature of tomato CLV3 and led to downregulation of the CLE-RS2 is a 13 amino acid peptide, in which one of CLV signaling pathway, resulting in overaccumula- the Hyp residues is modified with triarabinoside.68) tion of stem cells in the shoot apical meristems. The physiological importance of arabinosylation of However, this meristem phenotype caused by loss of CLE-RS2 is more pronounced than that of CLV3, HPAT has not been reported in plant species other as the arabinosyl chain is indispensable for both than tomato, suggesting that the extent of contribu- biological activity and receptor binding activity. tion of Hyp O-arabinosylation to CLV3 function Biosynthesis of Hyp triarabinoside involves differs between species.72) two distinct steps, namely initial attachment of Because Hyp O-arabinosylation is also required L-arabinose to Hyp and successive elongation of for cell wall structural proteins, loss-of-function the L-arabinose chain. The first enzyme, designated mutations in HPAT genes in Arabidopsis cause hydroxyproline arabinosyltransferase (HPAT) pleiotropic phenotypes that include enhanced hypo- (Fig. 6B), has been predicted to exist in plants since cotyl elongation, defects in cell wall thickening, the discovery of extensins in the 1960s, but its early flowering, early senescence, and defective molecular identity remained elusive for many years. growth of pollen tubes, causing a transmission defect Making use of our experience in the purification through the male gametophyte.12) Detailed pheno- of TPST, we purified HPAT from solubilized typic analyses of hpat mutants will provide a more Arabidopsis membrane fractions by affinity chroma- complete picture of how Hyp O-arabinosylated tography and identified 42 kD Golgi-localized type II glycoproteins and glycopeptides contribute to plant transmembrane proteins that contained a single growth and development. transmembrane domain near the N-terminus.12) Notably, purified Arabidopsis HPAT1 and HPAT2 5. Future directions showed no sequence similarity to any known func- Numerous research efforts including our own tionally characterized proteins, which is probably have identified more than a dozen secreted peptide why HPAT escaped similarity-based searches for hormones in plants. Over the past decade, the glycosyltransferases. BLAST analysis showed that number of functionally characterized peptide signals Arabidopsis has one additional HPAT family protein has increased several-fold and now exceeds the (HPAT3), resulting in a total of three members. number of classical plant hormones such as auxin No. 2] Exploring peptide hormones in plants 71 and gibberellin. Nonetheless, many genes encoding References possible secreted peptides remain in the Arabidopsis genome. In the TAIR10 protein data set, as many 1) Matsubayashi, Y. (2014) Posttranslationally modi- fi as 1,086 genes encode potential secreted peptides ed small-peptide signals in plants. Annu. Rev. Plant Biol. 65, 385–413. (SignalP score > 0.75) of between 50 and 150 amino 2) Grienenberger, E. and Fletcher, J.C. (2015) Polypep- acid residues. Continued efforts should be made to tide signaling molecules in plant development. further identify peptide hormones and explore how Curr. Opin. Plant Biol. 23,8–14. they function during plant growth and development. 3) Endo, S., Betsuyaku, S. and Fukuda, H. (2014) Previously, peptide hormones were identified Endogenous peptide ligand-receptor systems for fi diverse signaling networks in plants. Curr. Opin. using classical bioassay-guided puri cation. The Plant Biol. 21, 140–146. feasibility of this approach, however, largely depends 4) Marshall, E., Costa, L.M. and Gutierrez-Marcos, J. on the quality and sensitivity of the bioassay system (2011) Cysteine-rich peptides (CRPs) mediate and the abundance of the peptides in the samples. In diverse aspects of cell-cell communication in plant addition, one paradoxical problem is that, because no reproduction and development. J. Exp. Bot. 62, 1677–1686. one knows the activities of undiscovered molecules, 5) Matsubayashi, Y. and Sakagami, Y. (1996) Phyto- it is not possible to establish a bioassay system aimed sulfokine, sulfated peptides that induce the pro- at the detection of novel hormones. Indeed, no liferation of single mesophyll cells of Asparagus additional peptide hormones have been identified officinalis L. Proc. Natl. Acad. Sci. U.S.A. 93, – by bioassay-guided purification since 2006. Similarly, 7623 7627. fi 6) Ohyama, K., Ogawa, M. and Matsubayashi, Y. no peptide hormones have been identi ed using a (2008) Identification of a biologically active, small, forward genetics approach since 2003, indicating that secreted peptide in Arabidopsis by in silico gene the non-redundant peptide genes that produce a screening, followed by LC-MS-based structure visible phenotype upon mutation have already been analysis. Plant J. 55, 152–160. fully characterized. Thus, alternative strategies, such 7) Shinohara, H., Mori, A., Yasue, N., Sumida, K. and Matsubayashi, Y. (2016) Identification of three as genomics- or transcriptomics-based in silico ap- LRR-RKs involved in perception of root meristem proaches, preferably coupled with biochemical and growth factor in Arabidopsis. Proc. Natl. Acad. molecular biology techniques, offer promise for Sci. U.S.A. 113, 3897–3902. further research into peptide hormones in plants. 8) Tabata, R., Sumida, K., Yoshii, T., Ohyama, K., In plants, peptide hormones have long been Shinohara, H. and Matsubayashi, Y. (2014) Perception of root-derived peptides by shoot thought to exclusively mediate short-distance cell-to- LRR-RKs mediates systemic N-demand signaling. cell signaling involved in growth and development. Science 346, 343–346. Cumulative examples including our CEP-CEPR 9) Matsuzaki, Y., Ogawa-Ohnishi, M., Mori, A. and system, however, have provided a new functional Matsubayashi, Y. (2010) Secreted peptide signals model in which peptides respond to environmental required for maintenance of root stem cell niche in Arabidopsis. Science 329, 1065–1067. stimuli and mediate long-distance communication. 10) Nakayama, T., Shinohara, H., Tanaka, M., Baba, K., Plants may employ various types of peptide signals Ogawa-Ohnishi, M. and Matsubayashi, Y. (2017) for organ-to-organ communication and adapt to A peptide hormone required for Casparian strip diverse and complex environmental stresses more dy- diffusion barrier formation in Arabidopsis roots. – namically and ingeniously than previously assumed. Science 355, 284 286. 11) Komori, R., Amano, Y., Ogawa-Ohnishi, M. and Acknowledgements Matsubayashi, Y. (2009) Identification of tyrosyl- protein sulfotransferase in Arabidopsis. Proc. Natl. The series of studies described here were Acad. Sci. U.S.A. 106, 15067–15072. performed in collaboration with many students and 12) Ogawa-Ohnishi, M., Matsushita, W. and fi researchers whose names are recorded as authors Matsubayashi, Y. (2013) Identi cation of three hydroxyproline O-arabinosyltransferases in Arabi- in each referenced paper. I am grateful to all these dopsis thaliana. Nat. Chem. Biol. 9, 726–730. collaborators. This publication was supported in 13) Ito, Y., Nakanomyo, I., Motose, H., Iwamoto, K., part by a Grant-in-Aid for Scientific Research (S) Sawa, S., Dohmae, N. et al. (2006) Dodeca-CLE (number 25221105) from the Japan Society for peptides as suppressors of plant stem cell differ- – Promotion of Science (JSPS) and a Grant-in-Aid entiation. Science 313, 842 845. fi 14) Kondo, T., Sawa, S., Kinoshita, A., Mizuno, S., for Scienti c Research on Innovative Areas (number Kakimoto, T., Fukuda, H. et al. (2006) A plant 15H05957) from the Ministry of Education, Culture, peptide encoded by CLV3 identified by in situ Sports, Science, and Technology (MEXT) of Japan. MALDI-TOF MS analysis. Science 313, 845–848. 72 Y. MATSUBAYASHI [Vol. 94,

15) Shiu, S.H. and Bleecker, A.B. (2003) Expansion of 28) Lorbiecke, R. and Sauter, M. (2002) Comparative the receptor-like kinase/Pelle gene family and analysis of PSK peptide growth factor precursor receptor-like proteins in Arabidopsis. Plant homologs. Plant Sci. 163, 321–332. Physiol. 132, 530–543. 29) Matsubayashi, Y. and Sakagami, Y. (1999) Charac- 16) Ogawa, M., Shinohara, H., Sakagami, Y. and terization of specific binding sites for a mitogenic Matsubayashi, Y. (2008) Arabidopsis CLV3 pep- sulfated peptide, phytosulfokine-,, in the plasma- tide directly binds CLV1 ectodomain. Science 319, membrane fraction derived from Oryza sativa L. 294. Eur. J. Biochem. 262, 666–671. 17) Matsubayashi, Y., Takagi, L., Omura, N., Morita, A. 30) Matsubayashi, Y. and Sakagami, Y. (2000) 120- and Sakagami, Y. (1999) The endogenous sulfated and 160-kDa receptors for endogenous mitogenic pentapeptide phytosulfokine-, stimulates trache- peptide, phytosulfokine-alpha, in rice plasma ary element differentiation of isolated mesophyll membranes. J. Biol. Chem. 275, 15520–15525. cells of zinnia. Plant Physiol. 120, 1043–1048. 31) Matsubayashi, Y., Ogawa, M., Morita, A. and 18) Fukuda, H. and Komamine, A. (1980) Establishment Sakagami, Y. (2002) An LRR receptor kinase of an experimental system for the study of involved in perception of a peptide plant hormone, tracheary element differentiation from single cells phytosulfokine. Science 296, 1470–1472. isolated from the mesophyll of Zinnia elegans. 32) Torii, K.U. (2004) Leucine-rich repeat receptor Plant Physiol. 65,57–60. kinases in plants: structure, function, and signal 19) Kobayashi, T., Eun, C., Hanai, H., Matsubayashi, transduction pathways. Int. Rev. Cytol. 234,1– Y., Sakagami, Y. and Kamada, H. (1999) Phyto- 46. sulphokine-,, a peptidyl plant growth factor, 33) Wang, J., Li, H., Han, Z., Zhang, H., Wang, T., Lin, stimulates somatic embryogenesis in carrot. J. G. et al. (2015) Allosteric receptor activation by Exp. Bot. 50, 1123–1128. the plant peptide hormone phytosulfokine. Nature 20) Matsubayashi, Y., Takagi, L. and Sakagami, Y. 525, 265–268. (1997) Phytosulfokine-,, a sulfated pentapeptide, 34) Roberts, I., Smith, S., De Rybel, B., Van Den Broeke, stimulates the proliferation of rice cells by means of J., Smet, W., De Cokere, S. et al. (2013) The specific high- and low-affinity binding sites. Proc. CEP family in land plants: evolutionary analyses, Natl. Acad. Sci. U.S.A. 94, 13357–13362. expression studies, and role in Arabidopsis shoot 21) Motose, H., Iwamoto, K., Endo, S., Demura, T., development. J. Exp. Bot. 64, 5371–5381. Sakagami, Y., Matsubayashi, Y. et al. (2009) 35) Delay, C., Imin, N. and Djordjevic, M.A. (2013) CEP Involvement of phytosulfokine in the attenuation genes regulate root and shoot development in of stress response during the transdifferentiation response to environmental cues and are specificto of zinnia mesophyll cells into tracheary elements. seed plants. J. Exp. Bot. 64, 5383–5394. Plant Physiol. 150, 437–447. 36) Drew, M.C. and Saker, L.R. (1975) Nutrient supply 22) Hanai, H., Matsuno, T., Yamamoto, M., and growth of seminal root system in barley: II. Matsubayashi, Y., Kobayashi, T., Kamada, H. Localized, compensatory increases in lateral root et al. (2000) A secreted peptide growth factor, growth and rates of nitrate uptake when nitrate phytosulfokine, acting as a stimulatory factor supply is restricted to only part of root system. of carrot somatic embryo formation. Plant Cell J. Exp. Bot. 26,79–90. Physiol. 41,27–32. 37) Gansel, X., Munos, S., Tillard, P. and Gojon, A. ! 23) Igasaki, T., Akashi, N., Ujino-Ihara, T., (2001) Differential regulation of the NO3 and D Matsubayashi, Y., Sakagami, Y. and Shinohara, NH4 transporter genes AtNrt2.1 and AtAmt1.1 K. (2003) Phytosulfokine stimulates somatic em- in Arabidopsis: relation with long-distance and bryogenesis in Cryptomeria japonica. Plant Cell local controls by N status of the plant. Plant J. 26, Physiol. 44, 1412–1416. 143–155. 24) Matsubayashi, Y., Ogawa, M., Kihara, H., Niwa, M. 38) Ruffel, S., Freixes, S., Balzergue, S., Tillard, P., and Sakagami, Y. (2006) Disruption and over- Jeudy, C., Martin-Magniette, M.L. et al. (2008) expression of Arabidopsis phytosulfokine receptor Systemic signaling of the plant nitrogen status gene affects cellular longevity and potential for triggers specific transcriptome responses depending growth. Plant Physiol. 142,45–53. on the nitrogen source in Medicago truncatula. 25) Yang, H., Matsubayashi, Y., Nakamura, K. and Plant Physiol. 146, 2020–2035. Sakagami, Y. (1999) Oryza sativa PSK gene 39) Ruffel, S., Krouk, G., Ristova, D., Shasha, D., encodes a precursor of phytosulfokine-,, a sulfated Birnbaum, K.D. and Coruzzi, G.M. (2011) Nitro- peptide growth factor found in plants. Proc. Natl. gen economics of root foraging: transitive closure of Acad. Sci. U.S.A. 96, 13560–13565. the nitrate-cytokinin relay and distinct systemic 26) Loivamaki, M., Stuhrwohldt, N., Deeken, R., signaling for N supply vs. demand. Proc. Natl. Steffens, B., Roitsch, T., Hedrich, R. et al. (2010) Acad. Sci. U.S.A. 108, 18524–18529. A role for PSK signaling in wounding and micro- 40) Ohkubo, Y., Tanaka, M., Tabata, R., Ogawa- bial interactions in Arabidopsis. Physiol. Plant. Ohnishi, M. and Matsubayashi, Y. (2017) Shoot- 139, 348–357. to-root mobile polypeptides involved in systemic 27) Sauter, M. (2015) Phytosulfokine peptide signalling. regulation of nitrogen acquisition. Nat. Plants 3, J. Exp. Bot. 66, 5161–5169. 17029. No. 2] Exploring peptide hormones in plants 73

41) Rouhier, N., Couturier, J. and Jacquot, J.P. (2006) 54) Robbins, N.E. 2nd, Trontin, C., Duan, L. and Genome-wide analysis of plant glutaredoxin sys- Dinneny, J.R. (2014) Beyond the barrier: commu- tems. J. Exp. Bot. 57, 1685–1696. nication in the root through the endodermis. Plant 42) Okamoto, S., Tabata, R. and Matsubayashi, Y. Physiol. 166, 551–559. (2016) Long-distance peptide signaling essential 55) Matsubayashi, Y. (2011) Post-translational modifi- for nutrient homeostasis in plants. Curr. Opin. cations in secreted peptide hormones in plants. Plant Biol. 34,35–40. Plant Cell Physiol. 52,5–13. 43) Zhou, W., Wei, L., Xu, J., Zhai, Q., Jiang, H., Chen, 56) Moore, K.L. (2003) The biology and enzymology of R. et al. (2010) Arabidopsis tyrosylprotein sulfo- protein tyrosine O-sulfation. J. Biol. Chem. 278, transferase acts in the auxin/PLETHORA path- 24243–24246. way in regulating postembryonic maintenance of 57) Beisswanger, R., Corbeil, D., Vannier, C., Thiele, C., the root stem cell niche. Plant Cell 22, 3692–3709. Dohrmann, U., Kellner, R. et al. (1998) Existence 44) Fernandez, A., Drozdzecki, A., Hoogewijs, K., of distinct tyrosylprotein sulfotransferase genes: Nguyen, A., Beeckman, T., Madder, A. et al. molecular characterization of tyrosylprotein sulfo- (2013) Transcriptional and functional classifi- transferase-2. Proc. Natl. Acad. Sci. U.S.A. 95, cation of the GOLVEN/ROOT GROWTH 11134–11139. FACTOR/CLE-like signaling peptides reveals 58) Ouyang, Y., Lane, W.S. and Moore, K.L. (1998) their role in lateral root and hair formation. Plant Tyrosylprotein sulfotransferase: purification and Physiol. 161, 954–970. molecular cloning of an enzyme that catalyzes 45) Meng, L., Buchanan, B.B., Feldman, L.J. and Luan, tyrosine O-sulfation, a common posttranslational S. (2012) CLE-like (CLEL) peptides control the modification of eukaryotic proteins. Proc. Natl. pattern of root growth and lateral root develop- Acad. Sci. U.S.A. 95, 2896–2901. ment in Arabidopsis. Proc. Natl. Acad. Sci. U.S.A. 59) Amano, Y., Tsubouchi, H., Shinohara, H., Ogawa, 109, 1760–1765. M. and Matsubayashi, Y. (2007) Tyrosine-sulfated 46) Whitford, R., Fernandez, A., Tejos, R., Perez, A.C., glycopeptide involved in cellular proliferation and Kleine-Vehn, J., Vanneste, S. et al. (2012) expansion in Arabidopsis. Proc. Natl. Acad. Sci. GOLVEN secretory peptides regulate auxin carrier U.S.A. 104, 18333–18338. turnover during plant gravitropic responses. Dev. 60) Hanai, H., Nakayama, D., Yang, H., Matsubayashi, Cell 22, 678–685. Y., Hirota, Y. and Sakagami, Y. (2000) Existence 47) Casamitjana-Martinez, E., Hofhuis, H.F., Xu, J., of a plant tyrosylprotein sulfotransferase: novel Liu, C.M., Heidstra, R. and Scheres, B. (2003) plant enzyme catalyzing tyrosine O-sulfation of Root-specific CLE19 overexpression and the sol1/2 preprophytosulfokine variants in vitro. FEBS Lett. suppressors implicate a CLV-like pathway in the 470,97–101. control of Arabidopsis root meristem maintenance. 61) Wu, T., Kamiya, T., Yumoto, H., Sotta, N., Curr. Biol. 13, 1435–1441. Katsushi, Y., Shigenobu, S. et al. (2015) An 48) Song, W., Liu, L., Wang, J., Wu, Z., Zhang, H., Arabidopsis thaliana copper-sensitive mutant sug- Tang, J. et al. (2016) Signature motif-guided gests a role of phytosulfokine in ethylene produc- identification of receptors for peptide hormones tion. J. Exp. Bot. 66, 3657–3667. essential for root meristem growth. Cell Res. 26, 62) Kieliszewski, M.J. and Lamport, D.T. (1994) 674–685. Extensin: repetitive motifs, functional sites, post- 49) Aida, M., Beis, D., Heidstra, R., Willemsen, V., translational codes, and phylogeny. Plant J. 5, Blilou, I., Galinha, C. et al. (2004) The 157–172. PLETHORA genes mediate patterning of the 63) Pearce, G., Moura, D.S., Stratmann, J. and Ryan, Arabidopsis root stem cell niche. Cell 119, 109– C.A. (2001) Production of multiple plant hor- 120. mones from a single polyprotein precursor. Nature 50) Galinha, C., Hofhuis, H., Luijten, M., Willemsen, V., 411, 817–820. Blilou, I., Heidstra, R. et al. (2007) PLETHORA 64) Ohyama, K., Shinohara, H., Ogawa-Ohnishi, M. proteins as dose-dependent master regulators of and Matsubayashi, Y. (2009) A glycopeptide Arabidopsis root development. Nature 449, 1053– regulating stem cell fate in Arabidopsis thaliana. 1057. Nat. Chem. Biol. 5, 578–580. 51) Geldner, N. (2013) The endodermis. Annu. Rev. 65) Fletcher, J.C., Brand, U., Running, M.P., Simon, R. Plant Biol. 64, 531–558. and Meyerowitz, E.M. (1999) Signaling of cell fate 52) Pfister, A., Barberon, M., Alassimone, J., Kalmbach, decisions by CLAVATA3 in Arabidopsis shoot L., Lee, Y., Vermeer, J.E. et al. (2014) A receptor- meristems. Science 283, 1911–1914. like kinase mutant with absent endodermal dif- 66) Shinohara, H. and Matsubayashi, Y. (2013) Chemi- fusion barrier displays selective nutrient homeo- cal synthesis of Arabidopsis CLV3 glycopeptide stasis defects. eLife 3, e03115. reveals the impact of hydroxyproline arabinosyla- 53) Doblas, V.G., Smakowska-Luzan, E., Fujita, S., tion on peptide conformation and activity. Plant Alassimone, J., Barberon, M., Madalinski, M. Cell Physiol. 54, 369–374. et al. (2017) Root diffusion barrier control by a 67) Okamoto, S., Ohnishi, E., Sato, S., Takahashi, H., vasculature-derived peptide binding to the SGN3 Nakazono, M., Tabata, S. et al. (2009) Nod factor/ receptor. Science 355, 280–284. nitrate-induced CLE genes that drive HAR1- 74 Y. MATSUBAYASHI [Vol. 94,

mediated systemic regulation of nodulation. Plant 71) Xu, C., Liberatore, K.L., MacAlister, C.A., Huang, Cell Physiol. 50,67–77. Z., Chu, Y.H., Jiang, K. et al. (2015) A cascade of 68) Okamoto, S., Shinohara, H., Mori, T., Matsubayashi, arabinosyltransferases controls shoot meristem size Y. and Kawaguchi, M. (2013) Root-derived CLE in tomato. Nat. Genet. 47, 784–792. glycopeptides control nodulation by direct binding 72) MacAlister, C.A., Ortiz-Ramirez, C., Becker, J.D., to HAR1 receptor kinase. Nat. Commun. 4, 2191. Feijo, J.A. and Lippman, Z.B. (2016) Hydroxypro- 69) Nikolovski, N., Rubtsov, D., Segura, M.P., Miles, line O-arabinosyltransferase mutants oppositely G.P., Stevens, T.J., Dunkley, T.P. et al. (2012) alter tip growth in Arabidopsis thaliana and Putative glycosyltransferases and other plant Physcomitrella patens. Plant J. 85, 193–208. Golgi apparatus proteins are revealed by LOPIT 73) Matsubayashi, Y. (2011) Small post-translationally proteomics. Plant Physiol. 160, 1037–1051. modified Peptide signals in Arabidopsis. Arabidop- 70) Schnabel, E.L., Kassaw, T.K., Smith, L.S., Marsh, sis Book 9, e0150. J.F., Oldroyd, G.E., Long, S.R. et al. (2011) The 74) Grebe, M. (2011) Unveiling the Casparian strip. ROOT DETERMINED NODULATION1 gene Nature 473, 294–295. regulates nodule number in roots of Medicago truncatula and defines a highly conserved, unchar- acterized plant gene family. Plant Physiol. 157, (Received Sep. 1, 2017; accepted Oct. 25, 2017) 328–340.

Profile

Yoshikatsu Matsubayashi was born in 1971 in Mie prefecture, Japan. He graduated from the School of Bio-Agricultural Sciences, Nagoya University in 1993 and received his Ph.D. degree from Nagoya University in 1997. He joined the Graduate School of Bio- Agricultural Sciences in Nagoya University as an Assistant Professor in 1999 and became Associate Professor in the same department in 2002. In 2011 he moved to National Institute for Basic Biology (NIBB) where he was appointed a Professor. He then moved to the Graduate School of Science, Nagoya University in 2014. He has been conducting research on peptide signaling in plants and identified various novel peptide hormones that are critical for plant growth, development, and environmental adaptation. For his accomplishments, he received the Japan Society for Bioscience, Biotechnology and Agrochemistry (JSBBA) Award for Young Scientists in 2001, Japanese Society of Plant Physiologists Young Investigator Award in 2008, Molecular Biology Society of Japan Mitsubishi Chemical Award in 2010, and the Japan Society for the Promotion of Science (JSPS) Prize in 2016.