ORIGINAL PAPER

RNase T2 genes from rice and the evolution of secretoryribonucleases in plants

Gustavo C. MacIntosh Melissa S. Hillwig

Alexander Meyer Lex Flagel

Received: 16 November 2009 / Accepted: 2 February 2010

Springer-Verlag 2010

Abstract The plant RNase T2 family is divided into two

different subfamilies. S-RNases are involved in rejection of

self-pollen during the establishment of self-incompatibility

in three plant families. S-like RNases, on the other hand,

are not involved in self-incompatibility, and although gene

expression studies point to a role in plant defense and

phosphate recycling, their biological roles are less well

understood. Although S-RNases have been subjects of

many phylogenetic studies, few have included an extensive

analysis of S-like RNases, and genome-wide analyses to

determine the number of S-like RNases in fully sequenced

plant genomes are missing. We characterized the eight

RNase T2 genes present in the Oryza sativa genome; and

we also identified the full complement of RNase T2 genes

present in other fully sequenced plant genomes. Phyloge-

netics and gene expression analyses identified two classes

among the S-like RNase subfamily. Class I genes show

tissue specificity and stress regulation. Inactivation of

RNase activity has occurred repeatedly throughout evolu-

tion. On the other hand, Class II seems to have conserved

more ancestral characteristics; and, unlike other S-like

RNases, genes in this class are conserved in all plant spe-

cies analyzed and most are constitutively expressed. Our

results suggest that gene duplication resulted in high

diversification of Class I genes. Many of these genes are

differentially expressed in response to stress, and we pro-

pose that protein characteristics, such as the increase in

basic residues can have a defense role independent of

RNase activity. On the other hand, constitutive expression

and phylogenetic conservation suggest that Class II S-like

RNases may have a housekeeping role.

Keywords S-like RNases Evolution Rice Housekeeping Inactive RNase

Introduction

Ribonucleases (RNases) belonging to the RNase T2 family

are acidic endoribonucleases without base specificity that

are targeted to the secretory pathway. These RNases have

molecular masses of the protein moiety (some enzymes of

this family are glycosylated) between 20 and 25 kDa, and

are typified by RNase T2, an extracellular RNase from

Aspergillus oryzae (Sato and Egami 1957). The discovery

that some RNase T2 family members (called S-RNases) are

involved in gametophytic self-incompatibility (McClure

et al. 1989) lead to detailed analyses of this type of

enzymes in plants (Bariola and Green 1997). In addition to

S-RNases, present in the Solanaceae, Scrophulariaceae,

and Rosaceae families, plants possess other RNase T2

Communicated by P. Westhoff.

Electronic supplementary material The online version of thisarticle (doi:10.1007/s00438-010-0524-9) contains supplementarymaterial, which is available to authorized users.

G. C. MacIntosh (&) M. S. Hillwig A. MeyerInterdepartmental Genetics Program, Iowa State University,

2214 Molecular Biology Bldg., Ames, IA 50011, USA

e-mail: gustavo@iastate.edu

G. C. MacIntosh M. S. Hillwig A. MeyerDepartment of Biochemistry, Biophysics and Molecular

Biology, Iowa State University, Ames, IA 50011, USA

G. C. MacIntosh

Plant Sciences Institute, Iowa State University, Ames,

IA 50011, USA

L. Flagel

Department of Ecology, Evolution, and Organismal Biology,

Iowa State University, Ames, IA 50011, USA

123

Mol Genet Genomics

DOI 10.1007/s00438-010-0524-9

http://dx.doi.org/10.1007/s00438-010-0524-9

enzymes, known as S-like RNases, which are not involved

in self-incompatibility (Green 1994). S-like RNases have

been found in all plant species, but their evolution and

biological role are less well understood.

Members of the RNase T2 family are in fact present in

the genome of almost all organisms, including bacteria,

viruses, fungi, animals, and plants (Irie 1999; Deshpande

and Shankar 2002; Hillwig et al. 2009); with the exception

of some bacteria (Condon and Putzer 2002) and trypano-

somes (G. MacIntosh, unpublished). This almost absolute

conservation during the evolution of eukaryotes suggests

an important function for this family of RNases (Hillwig

et al. 2009).

Extensive analysis of gene expression in plants suggests

that S-like RNases are involved in phosphate recycling

during senescence and are upregulated during periods of

inorganic phosphate (Pi)-starvation. The expression of two

tomato RNases, RNase LE, and RNase LX, is induced when

cultivated tomato cells or seedlings are grown in Pi-deficient

media (Kock et al. 1998). Two RNase T2 genes from Ara-

bidopsis, RNS1, and RNS2, are also induced by Pi starvation

(Taylor et al. 1993; Bariola et al. 1994). In addition, the

expression of RNS2 and RNase LX increases during senes-

cence (Taylor et al. 1993; Lers et al. 1998), and ZRNaseI

from Zinnia is expressed in the late stage of in vitro tracheary

element differentiation (Ye and Droste 1996); indicating

that the corresponding proteins could be involved in recy-

cling of phosphate during processes involving cell death.

Moreover, antisense suppression of RNase LX expression

results in delayed senescence and leaf abscission, suggesting

that RNase LX may not only recycle phosphate during

senescence and abscission, but could also participate in the

control of these processes (Lers et al. 2006).

S-like RNases also participate in defense responses.

Tobacco RNase NE expression is induced in response to

Phytophthora parasitica, and purified RNase NE can

inhibit hyphal elongation of this pathogen (Hugot et al.

2002). Other Nicotiana genes, RNase NW, and RNase Rk1,

are induced in response to tobacco mosaic virus and

cucumber mosaic virus infections, respectively (Kurata

et al. 2002; Ohno and Ehara 2005). Mechanical wounding

also induces expression of several S-like RNases, including

RNS1, RNase NW, ZnRNaseII, RNase LE, and RNase Nk1

(Ye and Droste 1996; Kariu et al. 1998; LeBrasseur et al.

2002; Gross et al. 2004). Although the role of S-like

RNases during wounding is not clear, it has been proposed

that they could be antimicrobial enzymes, or participate in

phosphate remobilization during the healing process

(LeBrasseur et al. 2002; Ohno and Ehara 2005).

S-like RNases are also found in monocotyledonous

plants, although they have been studied to a lesser extent.

Two genes belonging to the RNase T2 family have been

described in rice. The first, RNase Os, encodes an active

enzyme that was purified from rice bran (Iwama et al.

2001). A barley enzyme, GAR-RNase, also accumulates in

the aleurone layer, and GAR-RNase expression is induced

by gibberellins and repressed by abscisic acid (Rogers and

Rogers 1999). It was proposed that this enzyme may con-

tribute to digestion of RNA in the dead starchy endosperm

cells and thereby may function to help mobilize Pi for the

developing embryo (Rogers and Rogers 1999). A second

rice gene belonging to this family was identified by a

proteomic analysis of rice leaves undergoing drought stress

(Salekdeh et al. 2002). The most highly induced protein in

that experiment, RNase DIS, has homology to RNase T2,

but lacks the conserved residues in the active site of the

enzyme. The same gene, although renamed OsRRP (Wei

et al. 2006), was reported to be expressed preferentially in

stems and to be down-regulated in an increased tillering

dwarf mutant. The biological role of this non-functional

RNase is unknown (Wei et al. 2006). Similar non-func-

tional S-like RNase genes have been found in wheat

(Chang et al. 2003) and barley (Gausing 2000).

Despite extensive phylogenetic analyses of S-RNases in

several plant families (Igic and Kohn 2001; Steinbachs and

Holsinger 2002; Vieira et al. 2008), a genome-wide survey

of the entire RNase T2 enzyme family has not been done

for any plant species other than Arabidopsis (Igic and Kohn

2001). Because S-like RNases from monocots have been

sparsely studied, and since the full sequence of the

rice genome is available (International Rice Genome

Sequencing Project 2005), we set out to characterize the

RNase T2 family in this species. We identified eight RNase

T2 genes in the rice genome. Using phylogenetic analysis

and gene expression studies, we characterized two different

classes of RNase T2 genes, all part of the S-like RNase

subfamily. Class I genes show tissue specificity and are

regulated by stress, as expected for S-like RNases. Among

this class, inactivation of RNase activity has occurred

repeatedly throughout evolution, including a highly con-

served monocot-specific class of RNase T2 genes lacking

RNase activity. On the other hand, Class II seems to have

conserved more ancestral characteristics; and, unlike other

S-like RNases, genes in this class are constitutively

expressed and conserved in all plant species analyzed.

These characteristics suggest that Class II S-like RNases

may have a housekeeping role.

Materials and methods

Database searches, sequences identification, and protein

sequence analyses

Identification of rice RNase T2 genes was done by BLAST

searches (Altschul et al. 1990) using the Arabidopsis RNS1

Mol Genet Genomics

123

and RNS2 protein sequences as query in the Oryza sativa

genome available at Phytozome (http://www.phytozome.

net/rice.php), which corresponds to TIGR Release 5 of the

annotation of the genome of the japonica subspecies of

O. sativa, and the annotated genome assemblies at the

MSU Rice Genome Annotation Database (Ouyang et al.

2007), and the Rice Annotation Project Database [RAP-DB

(Rice Annotation Project 2008)]. Then, the rice sequences

obtained were used to search again the rice genome.

Sequences from soybean (Glycine max), sorghum (Sor-

ghum bicolor), black cottonwood (Populus trichocarpa),

spikemoss (Selaginella moellendorffii), the moss Physc-

omitrella patens, and the green alga Chlamydomonas

reinhardtii were obtained by BLAST searches of the cur-

rent version of their respective genomes, as of June 2009,

accessed through Phytozome. Genome searches of the

green algae Volvox carteri and Ostreococcus tauri were

performed using the assemblies deposited at the Depart-

ment of Energy Joint Genome Institute (http://www.jgi.

doe.gov/); and the Arabidopsis thaliana genome was

searched using the TAIR9 genome release from the

Arabidopsis Information Resource (Swarbreck et al. 2008).

Additional sequences were obtained by BLASTP searches

of the non-redundant protein database and TBLASTN

searches of the non-human/non-mouse EST database at the

National Center for Biotechnology Information (http://

www.ncbi.nlm.nih.gov/).

Parameters for the predicted protein sequences were

obtained using the ProtParam tool at ExPASy (Gasteiger

and Hoogland 2005). Prediction of signal peptides and

subcellular localization was carried out using PSORT

(Nakai and Horton 1999), WoLF PSORT (Horton et al.

2007), and SignalP and TargetP (Emanuelsson et al. 2007).

Phosphate starvation microarray data (Cheng et al. 2007)

were obtained from Dr. Huixia Shou (Zhejiang University);

other microarray data were obtained using Genevestigator

(Zimmermann et al. 2005). Massively parallel signature

sequencing data (MPSS) were obtained using the Rice

MPSS database (Nobuta et al. 2007).

Plant material and analysis of gene expression

Rice plants from the subspecies japonica c.v. Nipponbare

were grown in growth chamber with 12 h photoperiod at

30C at day time and 28C at night. Tissue samples werecollected at day 75 after sowing. RNA extraction was

performed using TRI RNA extraction buffer (Ambion)

following manufacturers instructions. All rice material

was provided by Dr. Bing Yang (Iowa State University).

Petunia hybrida plants were obtained from a local market

and maintained in the greenhouse, tissue samples were

collected from adult plants about 3-month old. Tomato

plants (Solanum lycopersicum c.v. Early Girl Hybrid) were

grown in growth chamber with 16 h photoperiod at 21Cuntil they reached maturity. Petunia and tomato RNA was

extracted using the RNeasy Plant Mini Kit (Qiagen), RNA

was then treated with DNase to remove genomic DNA

contamination, and cDNA was synthesized using the

i-Script Select Kit (Bio-Rad) following manufacturers

protocols. PCR was performed using GoTaq Green

MasterMix (Promega), and the number of cycles was

optimized for each primer pair. Primers used are shown in

Supplementary Table 1 (ESM1). Primers were designed to

hybridize to regions of each gene that were not shared by

other members of the family; and specificity of amplifica-

tion was confirmed by sequencing of PCR products. PCR

products were analyzed in 1% agarose gels. For cloning,

PCR products were purified using the QIAquick PCR

Purification Kit (Qiagen), and the purified fragments were

cloned into the pGEM-T Easy Vector (Promega). Clones

were sequenced at the DNA Sequencing Facility at Iowa

State University. Partial cDNA sequences for OsRNS7,

OsRNS8 and RNase Phy2 were deposited in GenBank

under the accession numbers GQ507488, GQ507489, and

GQ507487, respectively. All experiments were repeated at

least three times, with two independent biological repli-

cates. Representative samples were chosen for each figure.

Phylogenetic analyses

Protein sequences were aligned using the CLC bio software

package, followed by manual adjustments. Only the region

between the first conserved region after the signal peptide

and the last conserved C residue was used in phylogenetic

analyses. PAUP 4.0 software (Swofford 2002) was used

for neighbor-joining (1,000 bootstrap replications) and

parsimony analyses, using default parameters. Mr Bayes

(Huelsenbeck and Ronquist 2001) was used for Bayesian

tree estimation, using a 1,000,000 generation run length

and allowing the program to optimize between nine dif-

ferent amino acid substitution matrices (using the prset

aamodelpr = mixed command), while leaving all other

parameters at their default settings. Due to the good fit of

the Wag substitution matrix (Whelan and Goldman 2001)

to the substitution patterns found in the data, this matrix

was used 99.96% of the time, although all nine matrices

were available throughout the analysis.

Results

The rice genome contains eight RNase T2 genes

A survey of the two available annotation databases for rice,

O. sativa, revealed that several RNase T2 genes are present

in its genome. Searches in the Michigan State University

Mol Genet Genomics

123

http://www.phytozome.net/rice.phphttp://www.phytozome.net/rice.phphttp://www.jgi.doe.gov/http://www.jgi.doe.gov/http://www.ncbi.nlm.nih.gov/http://www.ncbi.nlm.nih.gov/

(MSU) Rice Genome Annotation Database (Ouyang et al.

2007), which contains sequence information for the sub-

species japonica c.v. Nipponbare, returned eight gene

models with homology to RNase T2 proteins. In contrast,

searches in the most current version of the Rice Annotation

Project Database [RAP-DB (Rice Annotation Project

2008)], also based on the sequence of Nipponbare, pro-

duced only six gene models belonging to this family. These

six genes corresponded to six of the eight genes identified

in the MSU database; a search in previous versions of

RAP-DB using the sequences of the two genes only present

in the MSU database showed that these genes were also

present in older RAP-DB assemblies.

Additional searches using BLASTP and BLASTN,

either on gene models or the complete genome assembly

deposited in Phytozome (www.phytozome.net), failed to

identify any other sequences with homology to the RNase

T2 family. Based on analysis of EST sequences (not

shown) and our own gene expression experiments (see

below) we are confident that the eight genes correspond to

expressed proteins belonging to this ribonuclease family.

We named these genes OsRNS1-8, following the nomen-

clature used for the RNase T2 genes found in Arabidopsis

thaliana (Taylor and Green 1991). The corresponding IDs

from the two rice annotation databases are given in

Table 1.

We performed an analysis of the intron structure of the

genes in the rice RNase T2 family (Fig. 1). Four genes,

OsRNS1, OsRNS3, OsRNS7 and OsRNS8, contain only one

intron within the coding region, although the OsRNS7 and

OsRNS8 gene models do not have EST support. OsRNS3

has an additional intron in the 30 UTR. On the other hand,OsRNS2, OsRNS4, OsRNS5, and OsRNS6 have multiple

introns in the coding region. OsRNS4, OsRNS5 contain

three and four introns, respectively; while OsRNS2 and

OsRNS6 have seven and nine.

The proteins encoded by these genes (Figs. 1, 2) have

predicted molecular weights between 24 and 32 kDa,

similar to other RNases in the family (Deshpande and

Shankar 2002). All are predicted to have signal peptides

that would target these proteins to the secretory pathway

(Fig. 2), as is the norm among RNase T2 proteins. The

predicted subcellular localization for the eight rice proteins

is shown in Fig. 1. Several rice RNases have acidic pIs, as

described for most S-like enzymes; however, OsRNS2 has

a near neutral pI, and OsRNS5 and OsRNS6 have basic pIs.

Most S-like RNases have been found to have acid pIs while

basic pIs are characteristic of S-RNases, although some

basic S-like RNases have been found (Yamane et al. 2003

and references therein).

The eight rice genes encode proteins with homology to

RNase T2 proteins, and are more similar to S-like RNases

(not shown, but see below); however, the level of conser-

vation of amino acid residues that are important for ribo-

nuclease activity and structure (Irie 1999) varies among

them (Fig. 2). Five proteins (OsRNS2, OsRNS6, OsRNS8,

OsRNS1, OsRNS3) contain the two histidines that are

present in the conserved active site (CAS) I and II motifs,

and are essential for ribonuclease activity (Irie 1999).

OsRNS3 corresponds to the previously described RNase

Os, an active RNase purified from rice bran (Iwama et al.

2001). On the other hand, OsRNS4 and OsRNS5 have lost

most conserved residues in the two CAS. These changes

are bona fide mutations, not the result of sequencing errors,

as the same sequences can be found in the genome of

japonica and indica, and in many EST collections for both

subspecies (not shown). As a consequence, these two

proteins have clearly lost their ribonuclease activity, as

previously described for RNase DIS (Salekdeh et al. 2002)

(a.k.a. OsRPP: Wei et al. 2006), which corresponds to

OsRNS4.

OsRNS7 also has a mutation in the first active site

His. In this protein the amino acid in this position has

changed to Arg. This could be seen as a conservative

change; however CalsepRRP, the major protein of rest-

ing rhizomes of Calystegia sepium and also a member of

the RNase T2 family, has a substitution to Lys in the

same position and is devoid of any ribonuclease activity

(Van Damme et al. 2000). Thus, it is likely that OsRNS7

is also an inactive protein. No EST were found for this

gene; thus, it would be possible that OsRNS7 represents

a pseudogene. To test this idea, we obtained rice mRNA

and amplified a cDNA corresponding to OsRNS7. This is

the first proof of expression of OsRNS7, and it indicates

that this gene is not a pseudogene. Sequencing of this

cDNA also confirmed the change in CAS I. RT-PCR

analysis was also used to confirm expression of OsRNS8,

the other rice gene for which proof of expression

was missing. Both sequences have been deposited in

GenBank under accession numbers GQ507488 and

GQ507489, respectively.

Table 1 Equivalence of our nomenclature and gene identificationnumbers from the Japanese and US annotations of the rice genome

Gene MSU rice genome

annotation ID

Rice annotation project

database (RAP-DB) ID

OsRNS1 LOC_Os07g43670 Os07g0630400

OsRNS2 LOC_Os01g67180 Os01g0897200

OsRNS3 LOC_Os08g33710 Os08g0434100

OsRNS4 LOC_Os09g36680 Os09g0537700

OsRNS5 LOC_Os09g36700 Os09g0538000

OsRNS6 LOC_Os01g67190 Os01g0897300

OsRNS7 LOC_Os07g43600 Os07g0629300

OsRNS8 LOC_Os07g43640 Os07g0629900

Mol Genet Genomics

123

http://www.phytozome.net

Finally, although OsRNS6 has both active site His res-

idues conserved, this protein has a substitution in the first

Lys of CAS II. This is also a bona fide change since all

japonica ESTs corresponding to this gene have the same

mutation, and the same sequence can be found in the indica

genome (GenBank accession number EEC71956). Bio-

chemical characterization of Rhizopus niveous mutant

RNases indicates that this residue is important for activa-

tion of the enzyme (Irie 1999). Changes in this position can

have different degrees of effect on RNase activity; a

K108T substitution resulted in a protein with 20-fold less

activity than WT RNase Rh (Ohgi et al. 1996). Thus,

OsRNS6 could be an attenuated RNase.

Structural residues, such as Cys residues involved in

disulfide bridges that are important for proper folding (Irie

1999), also show different levels of conservation. Inter-

estingly, the Cys residue in CAS II is absolutely conserved,

even though the residues involved in activity are not.

Fig. 1 Structure of rice RNase T2 genes and properties of theencoded proteins. Open reading frames are shown by open boxes, anduntranslated regions are shown by gray boxes. Introns are shown aslines. Predicted molecular weight (MW) and isoelectric point (pI) of

the pre-proteins (P) and mature proteins (M) are indicated. Signalpeptides were predicted using Signal P. Predicted subcellular

localizations of the mature peptides are also indicated. V vacuole, Ssecreted (apoplast), ER endoplasmic reticulum, N nucleus

Fig. 2 Protein sequence alignment of the rice S-like RNases.Predicted signal peptides are underlined. Residues conserved in otherRNase T2 enzymes (Irie 1999) are shaded. The two conserved active

site (CAS) regions are indicated, and the two catalytic histidines are

indicated by asterisks

Mol Genet Genomics

123

Tissue specificity and stress regulation of OsRNSs

expression

As a first step toward characterizing the RNase T2 family

in rice, we analyzed the expression of the eight genes in

different organs, and took advantage of public data from

expression databases to identify biotic and abiotic stresses

that regulate the expression of rice S-like RNase genes. We

obtained root, leaf, stem, and inflorescence tissue from

adult rice plants, prepared RNA and tested for the presence

of RNase T2 transcripts by RT-PCR (Fig. 3). The expres-

sion of OsRNS2, OsRNS3, OsRNS4, and OsRNS8 was

detected in all organs. Although our RT-PCR assay is only

semi-quantitative, it is evident that even though OsRNS4 is

present in all tissues tested, its expression is stronger in

stems. OsRNS1, OsRNS5, OsRNS6, and OsRNS7 show

stronger organ specificity: OsRNS5 is only expressed in

stem and inflorescences, OsRNS6 is expressed in all tissues

except leaves, and OsRNS1and OsRNS7 are specifically

expressed in roots.

Analysis of public expression data (microarrays and

MPSS) confirmed the tissue specificity observed in our

experiments (not shown). Moreover, this analysis indicated

that many of the OsRNS genes are regulated by biotic and

abiotic stress (Table 2). We were particularly interested in

regulation by Pi starvation, since one of the functions

proposed for this family of enzymes is phosphate remobi-

lization in response to limitations in this nutrients

availability. Three genes showed elevated expression in

response to low Pi. OsRNS8 was induced in leaf and

root tissues. Interestingly, OsRNS7 and OsRNS5 are also

induced in roots in response to Pi starvation, even though

these proteins lacks the amino acid residues necessary for

ribonuclease activity.

Another proposed function for this family of RNases is a

role in plant defense against pests. Consistent with this

idea, OsRNS4 and OsRNS5 showed increased expression in

response to insect attacks (beet armyworm, Spodoptera

exigua, and water weevil, Lissorhoptrus oryzophilus) and

wounding. The expression of OsRNS4 and OsRNS5 was

also differentially increased in response to the bacterium

Xanthomonas oryzae and to the fungus Magnaporthe gri-

sea (Table 2). Remarkably, neither of these proteins seems

to be an active RNase, suggesting that the defense role for

these proteins should be independent of the enzymatic

activity. OsRNS3 was also differentially expressed in

response to M. grisea.

Several OsRNS genes are also induced by abiotic

stresses, but in this case genes are both induced and

repressed (Table 2). Salt stress results in strong induction

of OsRNS3, OsRNS4, and OsRNS5 expression, but

repression of OsRNS1. Similarly drought stress results in

induction of OsRNS3 but repression of OsRNS1; while

arsenate toxicity causes increase in the levels of OsRNS4

and decrease of OsRNS1. OsRNS3 expression is also

upregulated by cold, and OsRNS4 is upregulated in

response to root impedance changes. Finally, several RNase

genes (OSRNS1, OsRNS3, and OsRNS4) are repressed when

plants encounter oxygen deprivation.

We also found some information on hormonal regula-

tion of these genes (Table 2). OsRNS1 is upregulated by

gibberellins; while cytokinins regulate expression of

OsRNS3 (up) and OsRNS4 (down). Downregulation of

OsRNS3 expression in plants overexpressing OsRR6, a

negative regulator of CK signaling (Hirose et al. 2007),

confirmed that OsRNS3 is under cytokinin control.

Phylogenetic analysis of plant RNase T2 proteins

Phylogenetic analyses of plant RNase T2 proteins have

been carried out before. However, none of the previous

analyses incorporates an exhaustive analysis of all fully

sequenced plant genomes. This is important because

analysis of fully sequenced genomes can provide infor-

mation to verify the reasonability of the estimates obtained

for the gene duplications, and the fixation or loss of

homologs after each duplication event. To gain insights

into the evolution of the S-like RNase family in plants, we

performed a phylogenetic analysis including all the pro-

teins belonging to this family present in five fully

sequenced plant genomes. We carried out an exhaustive

Fig. 3 Expression patterns of the S-like RNase genes from rice. TotalRNA was obtained from stem (S), inflorescence (F), root (R) and leaf(L) tissues of 75-day-old plants, and expression of the eight RNase T2genes was evaluated by RT-PCR. Amplification of 18S cDNA was

used as loading control

Mol Genet Genomics

123

analysis of the genomes from rice, Arabidopsis (Arabid-

opsis thaliana) (Swarbreck et al. 2008), black cottonwood

(Populus trichocarpa) (Tuskan et al. 2006), sorghum

(S. bicolor) (Paterson et al. 2009), and soybean (G. max).

Searches in the fully sequenced genomes were per-

formed using the genome assemblies deposited in

Phytozome (www.phytozome.net), using TBLASTN. Gene

models were confirmed by EST analyses, using TBLASTN

to search the NCBI EST database. After these analyses, we

found that the Arabidopsis genome only contains the five

already known RNase T2 genes (Taylor and Green 1991;

Igic and Kohn 2001). The cottonwood genome also has five

genes belonging to the RNase T2 family (PtrRNS1-5; but

see below), while the soybean genome has twelve genes

(GmaRNS1-12), and the sorghum genome contains six

genes (SbRNS1-6). We also identified a partial RNase T2

sequence in the cottonwood chromosome LG X with high

degree of similarity to PtrRNS1, a partial sequence similar

to PtrRNS4 in a scaffold not included in the assembled

chromosomes, and a sequence with similarity to RNase T2

proteins in chromosome LG XII that has several stop

codons in the putative coding region, suggesting that it is a

pseudogene. These sequences were not included in our

analysis due to the lack of gene models or ESTs supporting

their expression and the possibility that they represent

pseudogene remnants.

We also collected published protein sequences for plant

S-like RNases from eudicots, monocots, moss, and algae.

In addition, we included other known protein sequences or

sequences derived from EST collections or cDNA collec-

tions for phylogenetic groups (like mosses and conifers)

that were not represented in our dataset. Finally, after a

Table 2 OsRNSs expression data extracted from public databases

Gene Overall

expressionaPi-

starvationb,cOther abiotic stressb Biotic stressb Hormoneb

OsRNS1 Variable Anoxia Gibberellin

Arsenate toxicity

Salt

Drought

OsRNS2 High

OsRNS3 High Anoxia Magnaporthe grisea OsRR6-ox

Arsenate toxicity Cytokinin

Drought

Salt

Cold

OsRNS4 Variable Anoxia Magnaporthe grisea Cytokinin

Salt Xanthomonas oryzae

Root impedance Beet armyworm

Arsenate toxicity water weevil

Wounding

OsRNS5 Variable Root Wounding Magnaporthe grisea

Salt Xanthomonas oryzae

Beet armyworm

OsRNS6 Low

OsRNS7 Low Root

OsRNS8 Low Root

Leaf

a Overall expression was estimated by the average intensity from microarray data obtained from Genevestigator, the number of sequenced tags

from the MPSS database, and EST abundanceb Specific responses to biotic and abiotic stresses and hormones were obtained from Genevestigator and MPSS databases; blue indicates

repression of transcript accumulation, yellow indicates induction, and gray indicates no change; light colors are used for 25-fold differences, and

dark colors for expression changes of more than fivefoldc Gene expression in response to Pi starvation was extracted from microarray data (Cheng et al. 2007); the tissue where expression was detected

is specified

Mol Genet Genomics

123

http://www.phytozome.net

preliminary phylogenetic analysis determined that a

monocot-specific clade exists, other monocot sequences

derived from ESTs were incorporated in our analysis.

In total, 78 protein sequences were analyzed using a

Bayesian phylogenetic approach to create a tree of plant

S-like RNases (Fig. 4). A full list of the proteins with the

corresponding accession number and species are shown

in Supplementary Table 2 (ESM2). The Bayesian tree

presented a well define clade for only a small portion of

the sequences, but the rest of the tree had less definition.

A Neighbor-Joining tree (Fig. 5; Supplementary Fig. 1,

ESM3) was also created to test whether a distance

analysis could resolve the topology of this portion of the

tree. However, the results from both approaches were

similar. The trees allowed us to make several inferences

on the evolution of the RNase T2 family in plants. First,

our analysis agrees with previous phylogenetic studies

that sorted plant RNase T2 proteins into three classes

(Igic and Kohn 2001; Steinbachs and Holsinger 2002;

Roalson and McCubbin 2003); Classes I and II corre-

spond to S-like RNases, while Class III corresponds to

S-RNases. Though our analysis did not include canonical

S-RNases, we did identify two main clusters for S-like

RNases corresponding to Classes I and II [we use the

class names as defined by Igic and Kohn (2001)], and a

small cluster corresponding to either proto-S-RNases or

relic S-RNases, that could correspond to Class III. The

position of the potential S-RNases corresponding to Class

III in a clade with Class I RNases could be an artifact of

the small number of Class III RNases in our analysis, or

could mean that Classes III and I share a common

ancestor.

The presence of only one RNase T2 gene in fully

sequenced genomes of green algae (Derelle et al. 2006;

Merchant et al. 2007) suggests that the gene duplication

that gave rise to the two classes (Class II and ClassI/III) of

RNases occurred after the split of Chlorophyta from the

main Viridiplantae stem. Only RNases belonging to Class I

could be found in the moss Physcomitrella patens and in

the spikemoss Selaginella moellendorffii, both of which

have fully sequenced genomes (Rensing et al. 2008; Banks

2009). The earliest plant in which both classes of RNases

were found is a conifer (Picea glauca). This could mean

that the duplication that originated Classes I and II hap-

pened after the separation of Euphyllophyta and Ly-

copodiophyta, and before the separation of Gymnosperms

and Angiosperms. However, it seems that Class II con-

served more ancestral characteristics because some algal

RNases and the only sequence obtained from a March-

antiophyta (Marchantia polymorpha) are included in the

Class II clade. Thus, and alternative hypothesis could be

that the gene duplication occurred early in the Embryo-

phyte branch, but Bryophytes and Lycopodiophyta could

have lost the gene corresponding to Class II after the split

from the main land plant branch.

As mentioned, part of the Bayesian tree is well resolved.

The resulting strongly supported clade corresponds to the

Class II S-like RNases. This clade includes at least one

gene from each of the fully sequenced higher plant gen-

omes, and in the cases in which more than one gene for this

class was identified (rice and soybean) the gene duplica-

tions have likely occurred recently, as they form mono-

phyletic groups. Moreover, in general, the associations

among Class II RNases follow established taxonomic

relationships.

Another clade with strong support contains several

proteins from soybean, a protein from cottonwood, and

RNase LC1, RNase LC2 and RNase MC, three proteins

that are expressed in seeds and are proposed to be either

storage proteins or have a defense role. The three seed

proteins and one of the soybean proteins (GmaRNS5) have

been proposed to be S-RNases, and part of Class III (Igic

and Kohn 2001; Vieira et al. 2008). The same is true for

GmaRNS5, which was derived from an EST sequence

(accession number BM520744) by Vieira et al. (2008), and

included among S-RNases based on amino acid patterns.

These genes could represent rare changes in function from

a shared ancestral S-RNase (Igic and Kohn 2001), or

S-RNase genes that acquired a new function after GSI loss

on several independent occasions (Vieira et al. 2008).

In contrast to the well-structured Classes II and III,

Class I S-like RNases form a clade with little structure,

with several proteins from each genome that do not form

species-specific clades, indicating that the gene duplication

events that originated these RNases predate speciation.

Then, diversification of Class I RNases resulted from dif-

ferential retention and expansion in different lineages by

gene sorting, a process characterized by rapid gene dupli-

cation and deactivation occurring differentially among

lineages (Zhang et al. 2000). For example, Arabidopsis

RNS4 and soybean GmaRNS9 form a small clade with

strong support. This indicates that the gene duplication that

gave rise to these genes occurred before the separation

between Eurosids I (which includes soybean) and Eurosids

II (which includes Arabidopsis). In further support of gene

sorting, no ortholog to the aforementioned clade could be

found in cottonwood, which also belongs to the Eurosids I,

suggesting that in this species this gene was lost. Similarly,

gene duplication and loss events can be evoked to explain

the clade containing Arabidopsis RNS3 and RNS5 among

other proteins.

One major source of incongruence in phylogenetic

reconstructions is homoplasy (OhUigin et al. 2002). To

establish its effect on our analysis of plant RNS genes, we

performed a parsimony analysis (not shown) of the same

sequences, which produced highly similar tree topologies

Mol Genet Genomics

123

to the Bayesian and neighbor-joining methods, and used

this tree to calculate the homoplasy index (HI). The HI for

the whole tree was unusually high (0.5460). However,

when we partitioned the data set into only Classes I and II

groups the sequences in the Class II clade had an HI of only

0.1588, whereas the Class I clade had an HI of 0.3807,

indicating that Class I is primarily responsible for the high

HI index of the tree containing the whole set of sequences.

Among Class I S-like RNases, a cluster with strong

support can be identified. This clade is monocot specific

and includes proteins that have lost their ribonuclease

activity due to mutations in CAS I and/or CAS II (Fig. 6).

This clade includes two of the three inactive RNSs from

rice, OsRNS4 and OsRNS5. Barley and sorghum also have

the two homologs of OSRNS4 and OsRNS5; HvuRNS4

corresponds to the previously described protein RSH1 that

is exclusively expressed in young leaf tissue and induced

by light (Gausing 2000). Thus, in this case, the gene

duplication that originated the clade may have occurred in

an ancestral grass genome, before the whole genome

duplication that gave rise to the modern genomes (Salse

et al. 2008). Although these genes have mutations that

Fig. 4 Phylogenetic relationships of plant S-like RNases. Tree wasestimated by the Bayesian method, using only conserved regions.

Bootstrap percentages greater than 50 are shown on interior branches.

The tree was rooted using algae sequences. Classes I, II and III clades

are indicated, as well as algal proteins. The monocot-specific clade

that groups inactive S-like RNases (inside Class I) is also labeled

Mol Genet Genomics

123

likely result in no RNase activity, they are clearly not

pseudogenes, because high levels of expression were

detected for OsRNS4 and OsRNS5 in different tissues

(Fig. 3), and high levels of OsRNS4 protein accumulate in

response to drought (Salekdeh et al. 2002).

Interestingly, inactivation of the CAS I and CAS II

RNase domains happened more than once during the

evolution of the S-like RNases. Another rice protein, Os-

RNS7, has a mutation in CAS I that likely inhibits RNase

activity. This protein does not cluster with the monocot-

specific clade. Inactive RNases are also present in eudicot

genomes (Fig. 6). CalsepRRP, the major protein of resting

rhizomes of Calystegia sepium, is an inactive RNase (Van

Damme et al. 2000) that belongs to the Class II S-like

RNases; in addition, we identified three soybean genes

(GmaRNS3, GmaRNS4, and GmaRNS5) that encode for

inactive RNases belonging to Class III.

Class II RNases are highly conserved and expressed

constitutively

The Class II clade contains one homolog from each fully

sequenced seed plant genome, with the exception of soy-

bean and rice, which have two proteins in the clade. Both

grasses and plants of the genus Glycine have undergone

ancestral whole genome duplication events (Shoemaker

et al. 2006; Salse et al. 2008), which may explain the

second gene belonging to Class II.

The analysis of gene expression indicates that OsRNS2

is expressed constitutively in all tissues and all develop-

mental stages (Fig. 3; Table 2, and data not shown). The

conservation of these genes could extend also to their

expression since Arabidopsis RNS2, the other gene for

which expression analyses are available, is also expressed

constitutively in all tissues and developmental stages

(Taylor et al. 1993; A. Meyer and G.C. MacIntosh

unpublished). This could indicate that the Class II S-like

RNases have an important function that has preserved these

enzymes throughout evolution. If this is the case, we would

expect that other members of this clade should have similar

expression patterns, and that plants for which a Class II

S-like RNases has not been found are likely to have a

homolog in their genome.

To test these hypotheses, we prepared RNA from dif-

ferent tissues from tomato, and analyzed expression of

RNaseLER, the tomato Class II RNase (Fig. 7a). We

observed that, as expected, RNaseLER is expressed con-

stitutively in root, stem, flower, and leaf tissues. We also

used primers designed to amplify Solanaceae Class II

RNases to search for a Class II sequence in another plant of

the family, Petunia hybrida, for which no Class II S-like

RNase is known. RT-PCR analysis resulted in the ampli-

fication of a cDNA fragment corresponding to a Class II

S-like RNase as expected (Fig. 7a, b). Gene expression

analysis indicated that this gene, which we called RNase

Liverwort

AlgaepI= 5.20.4

pI= 5.9

Class IIpI= 6.10.9

Class III, putative proto-S-RNases

pI= 8.21.2

Class IpI= 5.10.7 *

Monocotinactive clade

pI= 7.31.3

0.2

Fig. 5 Phylogenetic tree of plant S-like RNases obtained by theneighbor-joining method. Unrooted tree generated by the NJ method

with the same sequences used in Fig. 4. Class I (orange), Class II(blue), Class III (yellow), and algae (green) clusters are shaded. Theliverwort (Marchantia polymorpha) sequence is also indicated.Average isoelectric point (standard deviation) for predicted mature

proteins in each class is indicated (asterisk indicates that the Class Iaverage was calculated excluding proteins in the monocot inactive

clade). A detailed, rooted version of this tree, including bootstrap

values is shown in Supplementary Fig 1 (ESM3)

CAS I CAS II

* *RNS1 FGIHGLWP FWEHEWEKHGTCOsRNS3 FGIHGLWP FWAHEWEKHGTC

CalsepRRP FTIKGLWP DLAYEWAKHGTC

OsRNS7 FTIRGLWP LWSHEWSKHGTC

GmaRNS5 FTISYLHP LWRDQWRKFGSCGmaRNS4 FTISYLHP LWRDQWRKFGSCGmaRNS3 FTISYLHP LWRDQWRMFGSC

TaeRNS4 FFVQSFTT TWKSEWRSYGVCHvuRNS3 FFVEGFMT TWKSEWRSYGVCOsRNS4 FFVKSFMT SWKSEWNSYGVC MonocotSbRNS1 FFVEFFQT AWKSAWDNYGVC inactiveZmaRNS1 FYITGFTV SWKNAWKKAGAC clusterSbRNS3 FYITGLTV SWKNAWKKAGACOsRNS5 FYVAGFTV GWKNAWETSGVCHvuRNS4 FYVSGFTV SWKSAWKTSGVC

Fig. 6 Mutations in conserved active site residues in plant S-likeRNases. The alignment shows the conserved CAS I and CAS II

regions characteristic of RNase T2 enzymes. The catalytic histidines

are marked with asterisks. Either one or the two histidines are lost inthese proteins indicating that they are inactive RNases. The active

sites of OsRNS3 and RNS1, two active RNases, are shown for

comparison. The proteins belonging to the monocot-specific cluster

are indicated

Mol Genet Genomics

123

Phy2, is also expressed in root, stem, flower, and leaf tis-

sues (Fig. 7a). We sequenced this cDNA fragment; BLAST

analyses showed that the encoded protein is highly similar

to tobacco and tomato RNases from Class II (Fig. 7b). The

RNase Phy2 sequence has been deposited in GenBank

under the accession number GQ507487. These results

suggest that the absence of sequences from other species

belonging to this class likely reflects lack of detailed

studies in those species, and that in fact Class II RNases are

conserved and that they are expressed constitutively in all

tissues of the plant.

Protein properties may define different biological roles

for plant S-like RNases

Protein properties of predicted mature forms of all the

RNases included in our phylogenetic analyses showed that

the different classes of plant T2 RNases have a marked

difference in isoelectric point and size (Fig. 5; Supple-

mentary Table 3, ESM4). Most RNase T2 proteins are acid

RNases (Irie 1999). However, S-RNases are basic proteins

(Deshpande and Shankar 2002; Cruz-Garcia et al. 2003),

and the Class III clade identified in our analysis is made of

proteins with a average pI of 8.16. Thus, the protein

properties of these enzymes are similar to those of

canonical S-RNases, strengthening the idea that these

RNases are either precursors of S-RNases or S-RNases that

lost their original function.

Comparisons between Classes I and II (Supplementary

Table 3, ESM4) showed that Class II proteins are larger

(average of 247 aa in the mature peptide) than Class I

proteins (average of 210 aa). Class II proteins are also less

acidic, with a pI average of 6.13; while canonical Class I

RNases (excluding the monocot-specific clade) are acidic

proteins with a pI average of 4.99. Moreover, a detailed

analysis of Class I RNases suggests that there are two types

of proteins in this class (Supplementary Fig. 1, ESM3;

Supplementary Table 3, ESM4): proteins with a size

smaller than 210 aa and an average pI of 4.78; and proteins

larger than 210 aa with a pI average of 5.58. These two

types of proteins seem to form somewhat separated groups

in the Class I clade (Supplementary Fig. 1, ESM3),

although the support for the topology of that portion of the

tree is weak. These differences in pI and size could indicate

different functions for Class II and the two types of Class I

proteins.

Among Class I proteins, the monocot-specific clade that

contains inactive RNases is the most different. These

proteins are significantly (P \ 0.005) more basic than otherClass I proteins, with a pI average of 7.29. Analyses of

sequence alignments failed to recognize any amino acid

pattern specific for this class of RNases, although many

seem to have a C-terminal extension with charged amino

acids (not shown). Other changes in charged amino acids,

responsible for the elevated pI, are spread throughout the

proteins.

Amino acid pattern analysis did however identify a

Class II-specific pattern present in all Class II RNases and

in the liverwort RNase (Fig. 8). This region, immediately

before the second CAS, is a segment of nine amino acids

either polar or positively charged that most likely is

exposed and in contact with the substrate, since it is close

proximity with a region implicated in substrate binding in

other RNase T2 enzymes (Tanaka et al. 2000). The

increase in basic residues could increase binding of the

acidic substrate. An exception to this pattern is found in

trichomaglin and CalsepRRP, two proteins that have

acquire a different function (Van Damme et al. 2000; Gan

et al. 2004). In these two proteins acidic residues interrupt

the amino acid pattern.

Discussion

Ribonucleases belonging to the RNase T2 family are highly

conserved among all organisms. Although their enzymatic

properties are well understood, their biological roles are still

not clear. S-like RNases, a subfamily of the RNase T2

family, are present in the genomes of all plants studied thus

far; however, there has not been a genome-wide analysis of

this gene family prior to our study. In this work, we ana-

lyzed the rice genome and found eight genes belonging to

the S-like RNase family. We detected expression of every

one of these genes in at least one rice tissue, indicating that

they are all functional genes. We also did an exhaustive

search for this type of proteins in other fully sequenced

plant genomes, and found that the number of S-like RNases

in each genome is variable, from five genes in Arabidopsis

to 12 in soybean. This finding suggests a great degree of

diversification for the S-like RNase family in plants.

Fig. 7 Conservation and constitutive expression of Class II S-likeRNases. a RT-PCR analysis of the expression of RNase LER fromtomato, and a newly identified Class II S-like RNase, RNase Phy2,from petunia. Tomato and petunia RNA from stems (S), flowers (F),roots (R) and leaves (L) was analyzed. b Sequence alignment of thepredicted peptide encoded by the RNase Phy2 fragment amplified in aand RNase NGR2, a Class II S-like RNase

Mol Genet Genomics

123

Rice S-like RNases also show diverse expression pat-

terns. OsRNS2 is highly expressed in all tissues (Fig. 3). In

addition, it does not appear that this gene is regulated by

stress conditions, either biotic or abiotic (Table 2). The

other genes are expressed at different intensities, have

tissue specificity, and are regulated by abiotic and/or biotic

stress. Phylogenetic analyses indicate that S-like RNases

can be clustered in two classes, Classes I and II (Figs. 5, 6:

Igic and Kohn 2001; Steinbachs and Holsinger 2002;

Roalson and McCubbin 2003). OsRNS2 clusters with

Class II. This finding is also supported by the intronexon

structure of the genes (Fig. 1). Class II RNases have seven

or more introns (Igic and Kohn 2001), as is the case for

OsRNS2 and OsRNS6, the two Class II genes from rice. On

the other hand, Class I genes have normally between one

and three introns, as was found for OsRNS1, OsRNS3,

OsRNS4, OsRNS7, and OsRNS8. The last gene, OsRNS5,

has an extra intron (four), but phylogenetic analysis clearly

positions this RNase in the Class I category.

RNases in Class I form a cluster with little structure. The

diversification of Class I RNases seems be the result of

gene sorting (Zhang et al. 2000). This idea is also sup-

ported by the high homoplasy observed for these RNases.

Similar high levels of homoplasy have been shown to be

the result of either positive selection or purifying selection

(Rokas and Carroll 2008). Gene sorting is a common fea-

ture of several host-defense gene families, including the

vertebrates-specific RNase A family (Cho and Zhang

2007). Accordingly, the expression of several Class I rice

genes is regulated by pathogens and pests, which suggests

that the encoded proteins can have a role in defense. Other

S-like RNases have also been implicated in defense (Sup-

plementary Table 4, ESM5). For example, the expression

of RNase NE, a Class I S-like RNase from Nicotiana

tabacum, is induced in response to Phytophthora parasitica

(Galiana et al. 1997). Furthermore, purified RNase NE has

antimicrobial activity, and can inhibit hyphal elongation of

P. parasitica (Hugot et al. 2002). It was also reported that a

RNS1 ISDLLTSMKK---SWPTLAC---------PSG-SGEAFWEHEWEKHGTCS RNaseNE VSDLISRMQQ---NWPTLAC---------PSG-TGSAFWSHEWEKHGTCS ZRNaseII ISDLTSHMQS---EWPTLLC---------PSN-NGLAFWGHEWDKHGTCS GmaRNS7 ISDLISNMEK---DWPSLSC---------PSS-NGMRFWSHEWEKHGTCA GmaRNS11 ISDLISNMEK---DWPSLSC---------PSS-NGMRFWSHEWEKHGTCA RNasePD1 ISELTSNLEK---NWPSLTC---------PSS-NGFRFWSHEWEKHGTCS PtrRNS1 ISDLTSSLQK---DWPSLSC---------PSS-TGFRFWSHEWEKHGTCA RNS5 ISDLVSSLKK---NWPTLSC---------PSN-EGFNFWEHEWEKHGTCS RNS3 VSDLMSDLQR---EWPTLSC---------PSN-DGMKFWTHEWEKHGTCA FE1 ISDLVTRMQS---EWPTLAC---------PSG-DGTAFWTHEWEKHGTCS RNaseLX FSDLISTMEK---NWPSLAC---------PSS-DGLKFWSHEWLKHGTCS NGR3 ISDLISTMEK---NWPSLAC---------PSS-DGVRFWSHEWLKHGTCS PglRNS3 ISDLRKEMNE---KWPSLAC---------PSS-NSMNFWEHEWEKHGTCS ZRNaseI VSNLESELQV---HWPTLAC---------PSG-DGLKFWRHEWEKHGTCA RNaseDAI IQDLLSQLQT---QWPSLTC---------PSS-DGTSFWTHEWNKHGTCS HvuRNS1 VSDLLSSLRA---EWPTLAC---------PAS-DGLQFWAHEWEKHGTCA TaeRNS1 VSDLLSSLRR---EWPTLAC---------PTS-DGLQFWAHEWEKHGTCA SbRNS6 VSDLLSSLRA---KWPTLAC---------PSN-DGLRFWGHEWEKHGTCA OsRNS3 VSDLLGSMRS---EWPTLAC---------PSN-DGIRFWAHEWEKHGTCA PglRNS1 ILDLKEELNK---YWGSLSC---------PSS-DGHQFWGHEWEKHGTCP SmRNS1 IGGLRGDMDA---LWSSLSC---------PSS-NSEKFWAHEWEKHGTCS PpRNS1 LADVVDQMDD---DWGSLAC---------PAS-DSHSFWTHEWTKHGTCS RNS4 ISDLVCQMEKKWTEWGVWAC---------PSNETNL--WEHEWNKHGTCV RNS2 ISTLMDGLEKY---WPSLSCGSPSSCNGGKGS-----FWGHEWEKHGTCS PtrRNS2 ISTLHDALEKY---WPSLSCGSPSSCHGTKGS-----FWAHEWEKHGTCS RNaseLER ISTLLEPMRKY---WPSLSCSSPRSCHHKKGP-----FWGHEWEKHGTCA NGR2 ISTLLEPLRKY---WPSLSCGSPRSCHHRKGP-----FWAHEWEKHGTCA GmaRNS6 ILTLTNSLEQY---WPSLSCSKPSLCHGGKGT-----FWAHEWEKHGTCS GmaRNS2 ISTLTNALEQY---WPSLSCSKPSLCHGGKGT-----FWAHE--KHGTCS AhSL28 ISTLLGDLNKY---WPSLSCGSPSNCHGGKGL-----FWEHEWEKHGTCS PglRNS2 IVSLLGELERD---WPSLSCSSPSNCHGKRGS-----FWEHEWEKHGTCA SbRNS2 ILPLMEVLNKY---WPSLYCSKSGTCFSGKGL-----FWAHEWEKHGTCS ZmaRNS2 ILPLKEVLDKY---WPSLYCSKSGTCFSGKGL-----FWAHEWEKHGTCS OsRNS2 ILPLKPTLEKY---WPSLYCSSSSTCFSGKGP-----FWAHESEKHGTCS OsRNS6 ISRLKTILEEY---WPSLYCGSFSTCFGGKRP-----FWVHEWETHGTCG trichomaglin IMPFFGKLVEY---WPTYRCALEQSCNNRKEI-----LWGQQYEKHGTCA CalsepRRP ISILSNDLSKY---WPSYSCMSSSACGSFDAS-----DLAYEWAKHGTCS MpRNS VAELYDYLVEY---WPTLSCNSPKQCKGRKSG--PYGFWQHEWEKHGTCS CrRNS VSDLEEALD---LEWPSFM-------------GENADFWDHEWSKHGTCA VRN1 LEDLMDDLE---VEWPSVY-------------DSDETFWEHEWSKHGTCA RNaseBm2 -KDLLEELSS---EWPSYY-------------GSNYGFWKHEWEKHGTCA

CAS II

Fig. 8 Conserved domain inClass II S-like RNases.

Sequence alignment of Class I

(red fonts), Class II (blue fonts)and green algae and liverwort

(green fonts) showing the regionadjacent to the second CAS

(CAS II). The Class II-specific

domain is shaded (green polaramino acids, blue basic groups,red acidic groups)

Mol Genet Genomics

123

highly similar gene from N. tabacum, RNase Nk1, is

induced by cucumber mosaic virus infections (Ohno and

Ehara 2005); and expression of RNase NW is induced in

Nicotiana glutinosa plants infected with tobacco mosaic

virus (Kurata et al. 2002). Many S-like RNases are also

induced by wounding and insect infestations (Ye and

Droste 1996; Kariu et al. 1998; LeBrasseur et al. 2002;

Gross et al. 2004; Bodenhausen and Reymond 2007). Thus,

the evolutionary pattern observed for Class I RNases could

reflect a high rate of gene turnover that is consistent with

the role of the gene family in defending against the dif-

ferent pathogens and pests encountered by each plant

species. A similar explanation has been proposed for the

evolution of the RNase A family in animals (Zhang et al.

2000; Cho and Zhang 2007).

It is important to note that most rice S-like RNases

associated with defense responses are proteins that have

lost their RNase activity due to changes in the conserved

active sites (OsRNS4, OsRNS5 and OsRNS7). This

seemingly unusual finding also has a parallel in the RNase

A family. Several members of the RNase A family have

antimicrobial properties. Eosinophil associated RNases

have antiviral (RNase 2 and RNase 3 in humans) and

antibacterial (RNase 3) functions, and angiogenin and

RNase 7 have antibacterial and antifungal activities

(reviewed in Boix and Nogues 2007). However, their

ribonuclease activity is not necessary for this antimicrobial

role (Rosenberg 1995; Torrent et al. 2009). It has been

proposed that their antimicrobial activity is due to their

membrane destabilizing properties. A similar mechanism

could be envisioned for inactive S-like RNases, like

OsRNS4, OsRNS5, and OsRNS7, which are produced in

response to pathogen attacks. It has been suggested and

experimentally verified in several antibacterial RNases that

positively charged amino acid residues are important for

the disruption of negatively charged bacterial cell

membranes and thus are key to the bactericidal activity

(Carreras et al. 2003; Zhang et al. 2003; Huang et al. 2007).

Thus, the increase in pI observed for the members of the

monocot inactive clade could represent a gain of antibac-

terial activity.

Similarly, the Class III RNases described here present a

elevated pI, comparable to canonical S-RNases. S-RNases

participate in gametophytic self-incompatibility in at least

three plant families (Hua et al. 2008). S-RNases are

secreted into style mucilage, where they abort the growth

of pollen bearing the same S-allele (Clarke and Newbigin

1993). This cytotoxic activity and their expression in

flowers lead to the hypothesis that gametophytic self-

incompatibility may have evolved through the recruitment

of an ancient flower ribonuclease involved in defense

mechanisms against pathogens for the use in defense

against invasion by self-pollen tubes (Hiscock et al.

1996; Nasrallah 2005). Although it has been shown that

RNase activity is essential for the self-incompatibility

function of S-RNases, a high pI could contribute to their

cytotoxic effect. Then, it would be possible that the Class

III RNases included in our analysis are either precursors of

S-RNases (hence the denomination proto-S-RNases), or

are relic S-RNases no longer involved in self-incompati-

bility, but in either case they could be involved in defense.

In addition, the identification of a Class III RNase,

PtrRNS5, in a family in which S-RNases have not been

previously reported (Salicaceae) supports the hypothesis

that S-RNases are a monophyletic group that appeared only

once during evolution of angiosperms (Igic and Kohn

2001; Steinbachs and Holsinger 2002; Vieira et al. 2008).

Several rice Class I S-like RNases are also induced by

abiotic stresses that limit water availability, like drought

and high salinity. Comparable regulation has been

observed for Arabidopsis RNS1. This RNase is regulated

by abscisic acid (Hillwig et al. 2008), the main hormone

involved in the regulation of water stress in plants

(Verslues and Zhu 2007), and is induced in response to

high salinity (M. Hillwig and G.C. MacIntosh, unpub-

lished). Again, it is not clear whether RNase activity is

important for a role in this abiotic stress response, since

OsRNS4 and OsRNS5, two inactive RNases, are highly

induced in response to water stress (Table 2; Salekdeh

et al. 2002).

The presence of inactive RNases could be the result of

subfunctionalization or neofunctionalization of the genes

resulting from a gene duplication event (Lawton-Rauh

2003; Zhang 2003). Although more research is needed to

determine which case applies to these S-like RNases, some

evidence from yeast and humans points to subfunctional-

ization. Only one RNase T2 gene is present in the yeast

genome (MacIntosh et al. 2001). Thompson and Parker

(2009) demonstrated that this protein is responsible for the

cleavage of tRNAs in response to oxidative stress, and that

it can modulate cell survival in response to this stress

condition. While tRNA cleavage is dependent on ribonu-

clease activity, the modulation of cell fate is independent of

RNA catalysis. Similarly, human RNASET2, an active

RNase, is a tumor suppressor that functions in a manner

independent of its ribonuclease activity (Acquati et al.

2005; Smirnoff et al. 2006). Thus, in these cases, it is

evident that two different functions reside in the same

polypeptide. After gene duplication in plants, it is possible

that the two functions present in the original gene were

individually conserved in each of the duplicates.

Class II S-like RNases, on the other hand, seem to be

constitutively expressed; and they are conserved in all the

plant genomes analyzed. These RNases seem to have

conserved more ancestral characteristics, since they cluster

with lower plant T2 RNases. Evolution of this clade is

Mol Genet Genomics

123

similar to that observed for the RNase T2 family in animals

(Hillwig et al. 2009). Characterization of zebrafish RNase

T2 homologs indicated that animal RNase T2 genes are

also expressed constitutively; and based on the gene

expression and phylogenetic analyses, we proposed that

animal RNaste T2 proteins have a housekeeping role

(Hillwig et al. 2009). The conservation of Class II S-like

RNases, expression in all tissues analyzed, and ancestral

placement in the phylogenetic trees suggest that these

enzymes could also have a housekeeping role in plants.

This hypothesis is further supported by similarities in

subcellular localization between Class II S-like RNases and

animal RNase T2 enzymes, such as RNASET2, which has

been shown to localize in the lyzosome in human cells

(Campomenosi et al. 2006). Similarly, Arabidopsis RNS2

is an intracellular enzyme that may be located in the vac-

uole, the ER, or both (Bariola et al. 1999; Carter et al.

2004; A. Meyers and G.C. MacIntosh, unpublished); and

the two Class II RNases in rice are predicted to have an

intracellular localization (Fig. 1).

In addition to this putative housekeeping role, it is evi-

dent that some Class II RNases have acquired novel

functions as well. Proteins such as trichomaglin and Cals-

epRRP are probably storage proteins, and show differences

in expression patterns and conservation of Class II-specific

domains. Others, such as RNS2 and AhSL28 are expressed

at even higher levels during senescence (Taylor et al. 1993;

Liang et al. 2002), suggesting that in addition to a house-

keeping role these enzymes could have a role during

development.

In conclusion, the rice genome contains eight genes

belonging to the S-like RNase family. These genes belong

to two different phylogenetic classes. Class I S-like RNases

seem to have undergone diversification through differential

fixation and loss of genes after gene duplication events.

Many genes in this class are regulated by biotic and abiotic

stresses, and are tissue- or organ-specific (Table 2; Sup-

plementary Table 4, EMS 5). They seem to be part of a

defense response against a variety of pests, although this

role may not be dependent on catalytic activity, since

several of the genes induced by pests have mutations that

should result in proteins with no RNase activity.

On the other hand, Class II S-like RNases are conserved

in all plant genomes analyzed, and the genes are expressed

constitutively. Similarities with animal RNase T2 proteins,

in terms of evolution, gene expression and protein locali-

zation, suggest that Class II S-like RNases could be per-

forming a housekeeping function similar to that proposed

for animal RNase T2 proteins.

Acknowledgments We thank Dr. Bing Yang for providing the riceRNA and sharing unpublished microarray results, Dr. Huixia Shou for

providing Pi starvation microarray data, and Matthew Studham for

mining the soybean microarray data. This work was supported in part

by Grants from Iowa State University and the Roy J. Carver Chari-

table Trust to GCM.

References

Acquati F, Possati L et al (2005) Tumor and metastasis suppres-

sion by the human RNASET2 gene. Int J Oncol 26(5):1159

1168

Altschul SF, Gish W et al (1990) Basic local alignment search tool.

J Mol Biol 215(3):403410

Banks JA (2009) Selaginella and 400 million years of separation. Ann

Rev Plant Biol 60(1):223238. doi:10.1146/annurev.arplant.59.

032607.092851

Bariola P, Green P (1997) Plant ribonucleases. In: DAlessio G,

Riordan J (eds) Ribonucleases: structures and functions. Aca-

demic Press, New York, pp 163190

Bariola PA, Howard CJ et al (1994) The Arabidopsis ribonuclease

gene Rns1 is tightly controlled in response to phosphate

limitation. Plant J 6(5):673685

Bariola PA, MacIntosh GC et al (1999) Regulation of S-like

ribonuclease levels in arabidopsis: antisense inhibition of

RNS1 or RNS2 elevates anthocyanin accumulation. Plant

Physiol 119(1):331342

Bodenhausen N, Reymond P (2007) Signaling pathways controlling

induced resistance to insect herbivores in Arabidopsis. Mol Plant

Microbe Interact 20(11):14061420. doi:10.1094/Mpmi-20-

11-1406

Boix E, Nogues MV (2007) Mammalian antimicrobial proteins and

peptides: overview on the RNase A superfamily members

involved in innate host defence. Mol Biosyst 3(5):317335. doi:

10.1039/B617527a

Campomenosi P, Salis S et al (2006) Characterization of RNASET2,

the first human member of the Rh/T2/S family of glycoproteins.

Arch Biochem Biophys 449(12):1726. doi:10.1016/j.abb.

2006.02.022

Carreras E, Boix E et al (2003) Both aromatic and cationic residues

contribute to the membrane-lytic and bactericidal activity of

eosinophil cationic protein. Biochemistry 42(22):66366644.

doi:10.1021/bi0273011

Carter C, Pan S et al (2004) The vegetative vacuole proteome of

Arabidopsis thaliana reveals predicted and unexpected proteins.Plant Cell 16(12):32853303. doi:10.1105/tpc.104.027078

Chang SH, Ying H et al (2003) Expression of a wheat S-like RNase

(WRN1) cDNA during natural- and dark-induced senescence.

Acta Bot Sin 45(9):10711075

Cheng LJ, Wang F et al (2007) Mutation in nicotinamide amino-

transferase stimulated the Fe(II) acquisition system and led to

iron accumulation in rice. Plant Physiol 145(4):16471657. doi:

10.1104/pp.107.107912

Cho S, Zhang JZ (2007) Zebrafish ribonucleases are bactericidal:

Implications for the origin of the vertebrate RNase a superfam-

ily. Mol Biol Evol 24(5):12591268. doi:10.1093/molbev/

msm047

Clarke AE, Newbigin E (1993) Molecular aspects of self-incompat-

ibility in flowering plants. Ann Rev Genet 27(1):257

Condon C, Putzer H (2002) The phylogenetic distribution of bacterial

ribonucleases. Nucleic Acids Res 30(24):53395346

Cruz-Garcia F, Hancock CN et al (2003) S-RNase complexes and

pollen rejection. J Exp Bot 54(380):123130. doi:10.1093/

jxb/erg045

Derelle E, Ferraz C et al (2006) Genome analysis of the smallest free-

living eukaryote Ostreococcus tauri unveils many unique

Mol Genet Genomics

123

http://dx.doi.org/10.1146/annurev.arplant.59.032607.092851http://dx.doi.org/10.1146/annurev.arplant.59.032607.092851http://dx.doi.org/10.1094/Mpmi-20-11-1406http://dx.doi.org/10.1094/Mpmi-20-11-1406http://dx.doi.org/10.1039/B617527ahttp://dx.doi.org/10.1016/j.abb.2006.02.022http://dx.doi.org/10.1016/j.abb.2006.02.022http://dx.doi.org/10.1021/bi0273011http://dx.doi.org/10.1105/tpc.104.027078http://dx.doi.org/10.1104/pp.107.107912http://dx.doi.org/10.1093/molbev/msm047http://dx.doi.org/10.1093/molbev/msm047http://dx.doi.org/10.1093/jxb/erg045http://dx.doi.org/10.1093/jxb/erg045

features. Proc Natl Acad Sci USA 103(31):1164711652. doi:

10.1073/pnas.0604795103

Deshpande RA, Shankar V (2002) Ribonucleases from T2 family.

Crit Rev Microbiol 28(2):79122. doi:10.1080/1040-84029

1046704

Emanuelsson O, Brunak S et al (2007) Locating proteins in the cell

using TargetP, SignalP and related tools. Nat Protoc 2(4):953

971. doi:10.1038/nprot.2007.131

Galiana E, Bonnet P et al (1997) RNase activity prevents the growth

of a fungal pathogen in tobacco leaves and increases upon

induction of systemic acquired resistance with elicitin. Plant

Physiol 115(4):15571567

Gan JR, Yu L et al (2004) The three-dimensional structure and X-ray

sequence reveal that trichomaglin is a novel S-like ribonuclease.

Structure 12(6):10151025. doi:10.1016/j.str.2004.03.023

Gasteiger E, Hoogland C (2005) Protein identification and analysis tools

on the ExPASy server. In: Walker JM et al (eds) The proteomics

protocols handbook. Humana Press, Totowa, pp 571607

Gausing K (2000) A barley gene (rsh1) encoding a ribonuclease

S-like homologue specifically expressed in young light-grown

leaves. Planta 210(4):574579

Green PJ (1994) The ribonucleases of higher plants. Annu Rev Plant

Phys 45:421445

Gross N, Wasternack C et al (2004) Wound-induced RNaseLE

expression is jasmonate and systemin independent and occurs

only locally in tomato (Lycopersicon esculentum cv. Lukullus).Phytochemistry 65(10):13431350. doi:10.1016/j.phytochem.

2004.04.036

Hillwig MS, Lebrasseur ND et al (2008) Impact of transcriptional,

ABA-dependent, and ABA-independent pathways on wounding

regulation of RNS1 expression. Mol Genet Genomics

280(3):249261. doi:10.1007/s00438-008-0360-3

Hillwig MS, Rizhsky L et al (2009) Zebrafish RNase T2 genes and the

evolution of secretory ribonucleases in animals. BMC Evol Biol

9(1):170. doi:10.1186/1471-2148-9-170

Hirose N, Makita N et al (2007) Overexpression of a type-A response

regulator alters rice morphology and cytokinin metabolism. Plant

Cell Physiol 48(3):523539. doi:10.1093/pcp/pcm022

Hiscock SJ, Kues U et al (1996) Molecular mechanisms of self-

incompatibility in flowering plants and fungidifferent means

to the same end. Trends Cell Biol 6(11):421428. doi:

S0962-8924(96)10037-4[pii]

Horton P, Park KJ et al (2007) WoLF PSORT: protein localization

predictor. Nucleic Acids Res 35:W585W587. doi:10.1093/

Nar/Gkm259

Hua ZH, Fields A et al (2008) Biochemical models for S-RNase based

self-incompatibility. Mol Plant 1(4):575585

Huang YC, Lin YM et al (2007) The flexible and clustered lysine

residues of human ribonuclease 7 are critical for membrane

permeability and antimicrobial activity. J Biol Chem 282(7):

46264633. doi:10.1074/jbc.M607321200

Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference

of phylogenetic trees. Bioinformatics 17(8):754755

Hugot K, Ponchet M et al (2002) A tobacco S-like RNase inhibits

hyphal elongation of plant pathogens. Mol Plant Microbe

Interact 15(3):243250

Igic B, Kohn JR (2001) Evolutionary relationships among self-

incompatibility RNases. Proc Natl Acad Sci USA 98(23):13167

13171

International Rice Genome Sequencing Project (2005) The map-based

sequence of the rice genome. Nature 436(7052):793800. doi:

http://www.nature.com/nature/journal/v436/n7052/suppinfo/

nature03895_S1.html

Irie M (1999) Structure-function relationships of acid ribonucleases:

lysosomal, vacuolar, and periplasmic enzymes. Pharmacol Ther

81(2):7789 S0163-7258(98)00035-7[pii]

Iwama M, Ogawa Y et al (2001) Amino acid sequence and

characterization of a rice bran ribonuclease. Biol Pharm Bull

24(7):760766

Kariu T, Sano K et al (1998) Isolation and characterization of a

wound-inducible ribonuclease from Nicotiana glutinosa leaves.Biosci Biotech Biochem 62(6):11441151

Kock M, Theierl K et al (1998) Extracellular administration of

phosphate-sequestering metabolites induces ribonucleases in

cultured tomato cells. Planta 204(3):404407

Kurata N, Kariu T et al (2002) Molecular cloning of cDNAs encoding

ribonuclease-related proteins in Nicotiana glutinosa leaves, asinduced in response to wounding or to TMV-infection. Biosci

Biotech Biochem 66(2):391397

Lawton-Rauh A (2003) Evolutionary dynamics of duplicated genes in

plants. Mol Phylogenet Evol 29(3):396409. doi:10.1016/j.

ympev.2003.07.004

LeBrasseur ND, MacIntosh GC et al (2002) Local and systemic

wound-induction of RNase and nuclease activities in Arabidop-

sis: RNS1 as a marker for a JA-independent systemic signaling

pathway. Plant J 29(4):393403

Lers A, Khalchitski A et al (1998) Senescence-induced RNases in

tomato. Plant Mol Biol 36(3):439449

Lers A, Sonego L et al (2006) Suppression of LX ribonuclease in

tomato results in a delay of leaf senescence and abscission. Plant

Physiol 142(2):710721. doi:10.1104/pp.106.080135

Liang L, Lai Z et al (2002) AhSL28, a senescence and phosphate

starvation-induced S-like RNase gene in Antirrhinum. Biochim

Biophys Acta 1579(1):6471. doi:S0167478102005079[pii]

MacIntosh GC, Bariola PA et al (2001) Characterization of Rny1, the

Saccharomyces cerevisiae member of the T-2 RNase family ofRNases: unexpected functions for ancient enzymes? Proc Natl

Acad Sci USA 98(3):10181023

McClure BA, Haring V et al (1989) Style self-incompatibility gene

products of Nicotiana alata are ribonucleases. Nature342(6252):955957. doi:10.1038/342955a0

Merchant SS, Prochnik SE et al (2007) The Chlamydomonas genomereveals the evolution of key animal and plant functions. Science

318(5848):245250. doi:10.1126/science.1143609

Nakai K, Horton P (1999) PSORT: a program for detecting sorting

signals in proteins and predicting their subcellular localization.

Trends Biochem Sci 24(1):3435

Nasrallah JB (2005) Recognition and rejection of self in plant self-

incompatibility: comparisons to animal histocompatibility.

Trends Immunol 26(8):412418

Nobuta K, Venu RC et al (2007) An expression atlas of rice mRNAs

and small RNAs. Nat Biotechnol 25(4):473477. doi:

10.1038/Nbt1291

OhUigin C, Satta Y et al (2002) Contribution of homoplasy and of

ancestral polymorphism to the evolution of genes in anthropoid

primates. Mol Biol Evol 19(9):15011513

Ohgi K, Iwama M et al (1996) Enzymatic activities of several K108

mutants of ribonuclease (RNase) isolated from Rhizopus niveus.Biol Pharm Bull 19(8):10801082

Ohno H, Ehara Y (2005) Expression of ribonuclease gene in

mechanically injured of virus-inoculated Nicotiana tabacumleaves. Tohoku J Agric Res 55(34):11

Ouyang S, Zhu W et al (2007) The TIGR rice genome annotation

resource: improvements and new features. Nucleic Acids Res

35:D883D887. doi:10.1093/nar/gkl976

Paterson AH, Bowers JE, et al (2009) The Sorghum bicolor genomeand the diversification of grasses. Nature 457(7229):551556.

http://www.nature.com/nature/journal/v457/n7229/suppinfo/

nature07723_S1.html

Rensing SA, Lang D et al (2008) The Physcomitrella genome revealsevolutionary insights into the conquest of land by plants. Science

319(5859):6469. doi:10.1126/science.1150646

Mol Genet Genomics

123

http://dx.doi.org/10.1073/pnas.0604795103http://dx.doi.org/10.1080/1040-840291046704http://dx.doi.org/10.1080/1040-840291046704http://dx.doi.org/10.1038/nprot.2007.131http://dx.doi.org/10.1016/j.str.2004.03.023http://dx.doi.org/10.1016/j.phytochem.2004.04.036http://dx.doi.org/10.1016/j.phytochem.2004.04.036http://dx.doi.org/10.1007/s00438-008-0360-3http://10.1186/1471-2148-9-170http://dx.doi.org/10.1093/pcp/pcm022http://dx.doi.org/S0962-8924(96)10037-4[pii]http://dx.doi.org/10.1093/Nar/Gkm259http://dx.doi.org/10.1093/Nar/Gkm259http://dx.doi.org/10.1074/jbc.M607321200http://www.nature.com/nature/journal/v436/n7052/suppinfo/nature03895_S1.htmlhttp://www.nature.com/nature/journal/v436/n7052/suppinfo/nature03895_S1.htmlhttp://dx.doi.org/10.1016/j.ympev.2003.07.004http://dx.doi.org/10.1016/j.ympev.2003.07.004http://dx.doi.org/10.1104/pp.106.080135http://dx.doi.org/S0167478102005079[pii]http://dx.doi.org/10.1038/342955a0http://dx.doi.org/10.1126/science.1143609http://dx.doi.org/10.1038/Nbt1291http://dx.doi.org/10.1093/nar/gkl976http://www.nature.com/nature/journal/v457/n7229/suppinfo/nature07723_S1.htmlhttp://www.nature.com/nature/journal/v457/n7229/suppinfo/nature07723_S1.htmlhttp://dx.doi.org/10.1126/science.1150646

Rice Annotation Project (2008) The rice annotation project database

(RAP-DB): 2008 update. Nucl Acids Res 36(Suppl 1):D1028

D1033. doi:10.1093/nar/gkm978

Roalson EH, McCubbin AG (2003) S-RNases and sexual incompat-

ibility: structure, functions, and evolutionary perspectives. Mol

Phylogenet Evol 29(3):490506. doi:10.1016/S1055-7903(03)

00195-7

Rogers SW, Rogers JC (1999) Cloning and characterization of a

gibberellin-induced RNase expressed in barley aleurone cells.

Plant Physiol 119(4):14571464

Rokas A, Carroll SB (2008) Frequent and widespread parallel

evolution of protein sequences. Mol Biol Evol 25(9):19431953.

doi:10.1093/molbev/msn143

Rosenberg HF (1995) Recombinant human eosinophil cationic

proteinribonuclease-activity is not essential for cytotoxicity.

J Biol Chem 270(14):78767881

Salekdeh GH, Siopongco J et al (2002) Proteomic analysis of rice

leaves during drought stress and recovery. Proteomics

2(9):11311145

Salse J, Bolot S et al (2008) Identification and characterization of

shared duplications between rice and wheat provide new insight

into grass genome evolution. Plant Cell 20(1):1124. doi:

10.1105/tpc.107.056309

Sato K, Egami F (1957) Studies on ribonucleases in Takadiastase 1. J

Biochem Tokyo 44(11):753767

Shoemaker RC, Schlueter J et al (2006) Paleopolyploidy and gene

duplication in soybean and other legumes. Curr Opin Plant Biol

9(2):104109. doi:10.1016/j.pbi.2006.01.007

Smirnoff P, Roiz L et al (2006) A recombinant human RNASET2

glycoprotein with antitumorigenic and antiangiogenic character-

istics: expression, purification, and characterization. Cancer

107(12):27602769. doi:10.1002/Cncr.22327

Steinbachs JE, Holsinger KE (2002) S-RNase-mediated gametophytic

self-incompatibility is ancestral in eudicots. Mol Biol Evol

19(6):825829

Swarbreck D, Wilks C, et al (2008) The Arabidopsis information

resource (TAIR): gene structure and function annotation. Nucl

Acids Res 36(Suppl 1): D1009D1014. doi:10.1093/nar/gkm965

Swofford DL (2002) PAUP*. Phylogenetic analysis using parsimony

(*and other methods). Version 4.0b10. Sinauer Associates,

Sunderland

Tanaka N, Arai J et al (2000) Crystal structure of a plant ribonuclease,

RNase LE. J Mol Biol 298(5):859873. doi:10.1006/jmbi.

2000.3707

Taylor CB, Green PJ (1991) Genes with homology to fungal and S-

Gene RNases are expressed in Arabidopsis thaliana. PlantPhysiol 96(3):980984. doi:10.1104/pp.96.3.980

Taylor CB, Bariola PA et al (1993) Rns2a senescence-associated

Rnase of Arabidopsis that diverged from the S-Rnases before

speciation. Proc Natl Acad Sci USA 90(11):51185122

Thompson DM, Parker R (2009) The RNase Rny1p cleaves tRNAs

and promotes cell death during oxidative stress in Saccharomy-ces cerevisiae. J Cell Biol 185(1):4350. doi:10.1083/jcb.200811119

Torrent M, de la Torre BG et al (2009) Bactericidal and membrane

disruption activities of the eosinophil cationic protein are largely

retained in an N-terminal fragment. Biochem Jl 421(3):425434.

doi:10.1042/bj20082330

Tuskan GA, DiFazio S, et al (2006) The genome of black cottonwood,

Populus trichocarpa (Torr. & Gray). Science 313(5793):15961604. doi:10.1126/science.1128691

Van Damme EJM, Hao Q et al (2000) Major protein of resting

rhizomes of Calystegia sepium (hedge bindweed) closelyresembles plant RNases but has no enzymatic activity. Plant

Physiol 122(2):433445

Verslues PE, Zhu J-K (2007) New developments in abscisic acid

perception and metabolism. Curr Opin Plant Biol 10(5):447452

Vieira J, Fonseca NA et al (2008) An S-RNase-based gametophytic

self-incompatibility system evolved only once in eudicots. J Mol

Evol 67(2):179190. doi:10.1007/s00239-008-9137-x

Wei JY, Li AM et al (2006) Cloning and characterization of an

RNase-related protein gene preferentially expressed in rice

stems. Biosci Biotech Biochem 70(4):10411045

Whelan S, Goldman N (2001) A general empirical model of protein

evolution derived from multiple protein families using a

maximum-likelihood approach. Mol Biol Evol 18(5):691699

Yamane H, Tao R et al (2003) Identification of a non-S RNase, a

possible ancestral form of S-RNases, in Prunus. Mol Genet

Genomics 269(1):90100. doi:10.1007/s00438-003-0815-5

Ye ZH, Droste DL (1996) Isolation and characterization of cDNAs

encoding xylogenesis-associated and wounding-induced ribonu-

cleases in Zinnia elegans. Plant Mol Biol 30(4):697709Zhang JZ (2003) Evolution by gene duplication: an update. Trends

Ecol Evol 18(6):292298. doi:10.1016/S0169-5347(03)00033-8

Zhang JZ, Dyer KD et al (2000) Evolution of the rodent eosinophil-

associated RNase gene family by rapid gene sorting and positive

selection. Proc Natl Acad Sci USA 97(9):47014706

Zhang J, Dyer KD et al (2003) Human RNase 7: a new cationic

ribonuclease of the RNase A superfamily. Nucleic Acids Res

31(2):602607

Zimmermann P, Hennig L et al (2005) Gene-expression analysis and

network discovery using Genevestigator. Trends Plant Sci

10(9):407409

Mol Genet Genomics

123

http://dx.doi.org/10.1093/nar/gkm978http://dx.doi.org/10.1016/S1055-7903(03)00195-7http://dx.doi.org/10.1016/S1055-7903(03)00195-7http://dx.doi.org/10.1093/molbev/msn143http://dx.doi.org/10.1105/tpc.107.056309http://dx.doi.org/10.1016/j.pbi.2006.01.007http://dx.doi.org/10.1002/Cncr.22327http://dx.doi.org/10.1093/nar/gkm965http://dx.doi.org/10.1006/jmbi.2000.3707http://dx.doi.org/10.1006/jmbi.2000.3707http://dx.doi.org/10.1104/pp.96.3.980http://dx.doi.org/10.1083/jcb.200811119http://dx.doi.org/10.1083/jcb.200811119http://dx.doi.org/10.1042/bj20082330http://dx.doi.org/10.1126/science.1128691http://dx.doi.org/10.1007/s00239-008-9137-xhttp://dx.doi.org/10.1007/s00438-003-0815-5http://dx.doi.org/10.1016/S0169-5347(03)00033-8

RNase T2 genes from rice and the evolution of secretory ribonucleases in plantsAbstractIntroductionMaterials and methodsDatabase searches, sequences identification, and protein sequence analysesPlant material and analysis of gene expressionPhylogenetic analyses

ResultsThe rice genome contains eight RNase T2 genesTissue specificity and stress regulation of OsRNSs expressionPhylogenetic analysis of plant RNase T2 proteinsClass II RNases are highly conserved and expressed constitutivelyProtein properties may define different biological roles for plant S-like RNases

DiscussionAcknowledgmentsReferences

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org?) /PDFXTrapped /False

/Description >>> setdistillerparams> setpagedevice