Báo cáo khoa học: The consensus motif for N-myristoylation of plant proteins in a wheat germ cell-free translation system ppt

To clarify the relationship between the efficiency of protein N-myristoylation and the amino acid sequence of the substrate in plants, we have applied a wheat germ cell-free translation s

in a wheat germ cell-free translation system

Seiji Yamauchi1, Naoki Fusada1,2, Hidenori Hayashi1, Toshihiko Utsumi3, Nobuyuki Uozumi4, Yaeta Endo1,5and Yuzuru Tozawa1

1 Cell-Free Science and Technology Research Center and Venture Business Laboratory, Ehime University, Matsuyama, Japan

2 Department of Applied Biological Science, College of Bioresource Sciences, Nihon University, Fujisawa, Japan

3 Department of Biological Chemistry, Faculty of Agriculture, Yamaguchi University, Japan

4 Department of Biomolecular Engineering, Graduate School of Engineering, Tohoku University, Sendai, Japan

5 Systems and Structural Biology Center, RIKEN, Yokohama, Japan

Introduction

N-myristoylation is a form of lipid modification that

targets a wide variety of eukaryotic proteins and plays

important roles in cell physiology In many instances,

N-myristoylation alters the lipophilicity of the target

protein and facilitates its interaction with membranes,

thereby affecting its subcellular localization [1–3] In

mammals, N-myristoylated proteins include protein

kinases, phosphatases, guanine nucleotide-binding

proteins, and Ca2+-binding proteins, many of which participate in signal transduction pathways [1,4–7] Protein N-myristoylation is catalyzed by the enzyme myristoyl-CoA:protein N-myristoyltransferase (EC 2.3.1.97) (NMT), and involves the covalent attachment of myristic acid, a C14saturated fatty acid,

to the a-amino group of the N-terminal Gly of the target protein This modification usually occurs at the

Keywords

cell-free translation; myristoylation;

N-myristoyltransferase; plant; wheat germ

Correspondence

Y Tozawa, Division of Biomolecular

Engineering, Cell-Free Science and

Technology Research Center, Ehime

University, 3 Bunkyo-cho, Matsuyama,

Ehime 790-8577, Japan

Fax: +81 89 927 8528

Tel: +81 89 927 8274

E-mail: tozaway@ccr.ehime-u.ac.jp

(Received 10 May 2010, revised 4 July

2010, accepted 8 July 2010)

doi:10.1111/j.1742-4658.2010.07768.x

Protein N-myristoylation plays key roles in various cellular functions in eukaryotic organisms To clarify the relationship between the efficiency of protein N-myristoylation and the amino acid sequence of the substrate in plants, we have applied a wheat germ cell-free translation system with high protein productivity to examine the N-myristoylation of various wild-type and mutant forms of Arabidopsis thaliana proteins Evaluation of the rela-tionship between removal of the initiating Met and subsequent N-myristoy-lation revealed that constructs containing Pro at position 3 do not undergo N-myristoylation, primarily because of an inhibitory effect of this amino acid on elimination of the initiating Met by methionyl aminopeptidase Our analysis of the consensus sequence for N-myristoylation in plants focused on the variability of amino acids at positions 3, 6 and 7 of the motif We found that not only Ser at position 6 but also Lys at position 7 affects the selectivity for the amino acid at position 3 The results of our analyses allowed us to identify several A thaliana proteins as substrates for N-myristoylation that had previously been predicted not to be candidates for such modification with a prediction program We have thus shown that

a wheat germ cell-free system is a useful tool for plant N-myristoylome analysis This in vitro approach will facilitate comprehensive determination

of N-myristoylated proteins in plants

Abbreviations

AGG1, Arabidopsis thaliana G protein c subunit 1; ARF1A1c, ADP-ribosylation factor 1; AtNMT1, Arabidopsis thaliana myristoyl-CoA:protein N-myristoyltransferase 1; DHFR, dihydrofolate reductase; GFP, green fluorescent protein; MAP, methionyl aminopeptidase;

NMT, myristoyl-CoA:protein N-myristoyltransferase; RGLG2, RING domain ligase 2.

Gly that is exposed during cotranslational elimination

of the initiating Met by methionyl aminopeptidase

(MAP) [4,8], but the N-terminal Gly revealed after

post-translational proteolysis, such as that mediated by

caspases, can also be myristoylated if the downstream

amino acid sequence matches a consensus motif for

myristoylation [9,10] NMT appears to be ubiquitous

in eukaryotes, and corresponding genes have been

iso-lated and characterized in many organisms [11–16] In

the case of plants, two cDNAs encoding NMT-like

proteins, Arabidopsis thaliana NMT1 (AtNMT1) and

A thalianaNMT2, have been isolated from A thaliana

and characterized, and AtNMT1 has been genetically

confirmed to be required for plant viability [17]

Most N-myristoylated proteins contain a

myristoyla-tion motif, generally defined as

Met1-Gly2-Xaa3-Xaa4-Xaa5-Ser⁄ Thr6-Xaa7-Xaa8 (where Xaa indicates

any amino acid), at their N-terminus [18] This motif

interacts with the substrate-binding pocket of NMT,

and thereby ensures binding of the substrate protein to

the enzyme [19] Several rules for the myristoylation

motif have been proposed, on the basis of biological

information such as the N-terminal sequences of

known myristoylated proteins and the crystal

struc-tures of NMTs [19,20] A biochemical approach

showed that the combination of amino acids at

posi-tions 3, 6 and 7 in the myristoylation motif is a major

determinant of protein N-myristoylation in

mamma-lian systems [21] However, evidence suggests that the

consensus sequence for myristoylation in plants differs

slightly from those of mammals and yeast [15,17,22]

NMT activity has been detected in eukaryotic

cell-free translation systems, including rabbit reticulocyte

lysate [23], insect [24] and wheat germ extract [25]

sys-tems Some of these systems have been applied to the

characterization of protein N-myristoylation, and such

studies have resulted in the identification of 18 novel

human N-myristoylated proteins [26] Wheat germ

extracts have proved useful for analysis of

N-myristoy-lation of plant proteins, with N-myristoyN-myristoy-lation of

sev-eral A thaliana proteins, such as a Ca2+-dependent

protein kinase, a GTPase, and RING finger-type

ubiquitin ligases, having been demonstrated in vitro

[3,27–29] Two groups have described the

‘N-myristoy-lome’ of A thaliana [22,30] However, the number of

N-myristoylated plant proteins confirmed

experimen-tally has remained far smaller than that predicted

in silico, and the consensus motif for plant protein

myristoylation is still imprecise

We have now clarified the relationship between the

efficiency of protein N-myristoylation and the identity

of the amino acids at positions 3, 6 and 7 of the

con-sensus motif in plant target proteins with the use of a

wheat germ cell-free translation system We identified several new N-myristoylated proteins from A thaliana that were predicted to be noncandidates for N-myri-stoylation by an existing prediction program Our results thus update the consensus sequence motif for N-myristoylation in plants

Results

In vitro protein N-myristoylation with a wheat germ cell-free translation system

Wheat germ extracts have previously been shown to contain protein N-myristoylation activity [25] To inves-tigate the mode of protein N-myristoylation in plant cells in more detail, we established a cell-free protein N-myristoylation system as an application of an advanced wheat germ cell-free translation system that allows protein synthesis on a preparative scale [31–33]

We first analyzed the N-myristoylation of two proteins

of A thaliana as model substrates: RING domain ligase 2 (RGLG2) and ADP-ribosylation factor 1 (ARF1A1c) N-myristoylation of these proteins was described previously, and that of RGLG2 was demon-strated in a wheat germ extract [29] We also generated G2A mutants of these proteins as negative controls (nonmyristoylated proteins) by substitution of the codon for Gly2 with a codon for Ala (Fig 1A) The wild-type and mutant proteins were produced with the wheat germ cell-free translation system in the presence

of [14C]Leu or [14C]myristic acid, and the translation products were analyzed by SDS⁄ PAGE and autoradiog-raphy In the presence of [14C]Leu, the translation reac-tions for wild-type and G2A mutants of RGLG2 and ARF1A1c gave rise predominantly to labeled proteins with the expected molecular masses of 52 and 21 kDa, respectively (Fig 1B) In the presence of [14C]myristic acid, however, the wild-type proteins were labeled with

14C, whereas the G2A mutants were not (Fig 1C)

We next analyzed the in vitro N-myristoylation of proteins fused to an N-terminal myristoylation motif For this analysis, we selected green fluorescent protein (GFP) from jellyfish and dihydrofolate reductase (DHFR) from Escherichia coli as model proteins DNA encoding the myristoylation motif Met-Gly-Ala-Ala-Ala-Ser-Ala-Ala-Ala-Ala was fused to the ORF of GFP or DHFR with the use of PCR to construct genes encoding the chimeric proteins Myr–GFP and Myr–DHFR, respectively (Fig 2A) Furthermore, DNA encoding the N-terminal nonmyristoylation motif Met-Ala-Ala-Ala-Ala-Ser-Ala-Ala-Ala-Ala was also fused to each ORF to construct genes for the G2A mutants Myr–GFP-G2A and Myr–DHFR-G2A

(Fig 2A) Translation of mRNAs encoding Myr–GFP

or Myr–DHFR in the presence of [14C]Leu gave rise

to a single labeled protein with the expected molecular

mass of 29 or 25 kDa, respectively (Fig 2B) Such

translation in the presence of [14C]myristic acid also

showed that Myr–GFP and Myr–DHFR were

modi-fied as expected (Fig 2C) By contrast,

Myr–GFP-G2A and Myr–DHFR-Myr–GFP-G2A were not labeled with

[14C]myristic acid These results thus confirmed that

N-myristoylation in the advanced wheat germ cell-free

translation system occurs in a manner dependent on

the canonical consensus sequence [18]

Analysis of in vitro-synthesized N-myristoylated

proteins

To confirm the N-myristoylation of proteins

synthe-sized with the cell-free system, we next performed

MALDI-TOF MS analysis Wild-type forms of

ARF1A1c and Myr–GFP were synthesized in vitro and

purified by Ni2+-affinity column chromatography on

the basis of their His6 tags Tryptic peptides derived

from the purified proteins were then analyzed by MALDI-TOF MS Reaction mixtures supplemented with myristic acid yielded a specific peak with m⁄ z ratios of 818.4 for ARF1A1c and 1184.4 for Myr–GFP (Table 1; Fig S1), values that are in good

A

Fig 2 In vitro N-myristoylation of GFP and DHFR fused to a myri-stoylation consensus motif (A) Structures and N-terminal amino acid sequences of wild-type (wt) and G2A mutant forms of Myr– GFP and Myr–DHFR (B, C) The wild-type and G2A mutant proteins were translated in vitro with the use of a wheat germ cell-free translation system in the presence of [ 14 C]Leu (B) or [ 14 C]myristic acid (C) Portions of the reaction products were analyzed by SDS ⁄ PAGE on a 15% gel and autoradiography Arrowheads indi-cate proteins of the expected size.

A

Fig 1 In vitro N-myristoylation of two A thaliana substrates (A)

Structures and N-terminal amino acid sequences of wild-type (wt)

and G2A mutant forms of RGLG2 and ARF1A1c (B, C) The wild-type

and G2A mutant proteins were translated in vitro with the use of a

wheat germ cell-free translation system in the presence of [ 14 C]Leu

(B) or [ 14 C]myristic acid (C) Portions of the reaction products were

analyzed by SDS ⁄ PAGE on a 15% gel and autoradiography Control

reaction mixtures without mRNA were similarly analyzed

Arrow-heads indicate proteins of the expected size The positions of

pro-tein size standards are shown on the left of the gel.

Table 1 Observed and calculated mass values for the tryptic N-terminal peptides of ARF1A1c and Myr–GFP ND, not detected.

Sequence of tryptic N-terminal peptide

Calculated mass (Da)

Observed mass (Da)

Myristic acid (+)

Myristic acid ( )) ARF1A1c

Myr–GFP

Myristate–GAAASAAAAVSKb 1184.5 1184.4 ND a

N-terminal peptide lacking the initiating Met. bMyristoylated N-terminal peptide.

agreement with the calculated mass for the

correspond-ing myristoylated N-terminal fragments (m⁄ z ratios of

818.1 for ARF1A1c and 1184.5 for Myr–GFP)

(Table 1) In the absence of myristic acid, a peak

attributable to the N-terminal fragment lacking the

ini-tiating Met was detected for both ARF1A1c and Myr–

GFP (m⁄ z ratios of 608.4 and 974.1, respectively)

(Table 1; Fig S1) We did not observe a peak

corre-sponding to the N-terminal fragment of either protein

that retained the initiating Met in the presence or

absence of myristic acid These results thus confirmed

removal of the initiating Met and subsequent

N-myri-stoylation of ARF1A1c and Myr–GFP synthesized in

the wheat germ cell-free translation system

Effect of the combination of amino acids at

positions 3 and 6 on protein N-myristoylation in

the wheat germ cell-free translation system

Ser6 in the myristoylation consensus motif has been

shown to affect the selectivity for the amino acid at

position 3 in N-myristoylation catalyzed by rabbit

retic-ulocyte lysate [22,34] To clarify further the sequence

specificity of the myristoylation motif in plants, we first

examined the relationship between amino acids at

positions 3 and 6 for protein N-myristoylation in the

wheat germ cell-free translation system We generated cDNAs encoding A thaliana G protein c subunit 1 (AGG1) fused at its N-terminus to the myristoylation motifs Met-Gly-Xaa-Ala-Ala-Ala-Ala-Ala-Ala-Ala (Myr–AGG1-3X6A) or Met-Gly-Xaa-Ala-Ala-Ser-Ala-Ala-Ala-Ala (Myr–AGG1-3X6S), with position 3 of each motif separately occupied by each of the 20 amino acids (Fig 3A) Translation of the corresponding mRNAs in the presence of [14C]Leu gave rise to main products with the expected molecular mass of 13 kDa, indicating that all proteins were effectively translated in the wheat germ system (Fig 3B,C) Translation in the presence of [14C]myristic acid revealed that the require-ment of protein N-myristoylation for the amino acid at position 3 differed between Myr–AGG1-3X6A and Myr–AGG1-3X6S Only two amino acids (Asn and Gln) at position 3 allowed efficient N-myristoylation of Myr–AGG1-3X6A (Fig 3B) In contrast, 12 amino acids (Gly, Ala, Ser, Cys, Thr, Val, Asn, Leu, Ile, Gln, His, and Met) at position 3 supported efficient N-myristoylation of Myr–AGG1-3X6S (Fig 3C) These results thus indicated that Ser6 in the plant myri-stoylation motif has a marked effect on selectivity for the amino acid at position 3 in protein N-myristoyla-tion, as previously observed in rabbit reticulocyte lysate [22,34]

A

B

C

Fig 3 Effect of Ser6 in the myristoylation

consensus motif on selectivity for the amino

acid at position 3 in initiator Met elimination

and N-myristoylation (A) Structure of

mature AGG1 fused at its N-terminus to the

myristoylation motifs

Met-Gly-Xaa-Ala-Ala-Ala-Ala-Ala-Ala-Ala (Myr–AGG1-3X6A) or

Met-Gly-Xaa-Ala-Ala-Ser-Ala-Ala-Ala-Ala

(Myr–AGG1-3X6S) (B, C) Each of the 20

mRNAs corresponding to Myr–AGG1-3X6A

(B) or Myr–AGG1-3X6S (C) was translated

with a wheat germ cell-free translation

system in the presence of [ 14 C]Leu (upper

panels), [ 14 C]myristic acid (middle panels),

or [35S]Met (lower panels) Portions of the

reaction products were analyzed by

SDS ⁄ PAGE on a 15% gel and

autoradiogra-phy The molecular masses of labeled

translation products are shown on the right.

Relationship between initiator Met elimination

and N-myristoylation

We next examined the relationship between initiator

Met elimination and N-myristoylation with the Myr–

AGG1-3X6A and Myr–AGG1-3X6S constructs Given

that the mature AGG1 polypeptide does not contain a

Met, it was possible to investigate the efficiency of

initia-tor Met elimination by metabolic labeling with [35S]Met

In the case of translation products containing Pro at

position 3 (Myr–AGG1-3P6A and Myr–AGG1-3P6S),

for which the initiating Met is the only Met in the entire

encoded amino acid sequence, the extent of [35S]Met

incorporation was similar to that observed with Myr–

AGG1-3M6A or Myr–AGG1-3M6S (Fig 3B,C), for

which Met residues are present at both positions 1 and

3, indicating that the initiating Met is retained in Myr–

AGG1-3P6A and Myr–AGG1-3P6S This result is in

good agreement with our previous finding of an

inhibi-tory effect of Pro at position 3 of full-length translation

products on MAP function in the wheat germ cell-free

translation system [35] We also detected low levels of

[35S]Met incorporation in the translation products with

Gly, Thr, Asp or Glu at position 3 of the myristoylation

motif in Myr–AGG1-3X6A or Myr–AGG1-3X6S

(Fig 3B,C) Given that the AGG1 mutants containing

Pro3 were not modified by N-myristoylation

(Fig 3B,C), our results indicate that Pro at position 3

prevents substrate recognition not by NMT but rather

by MAP, which functions upstream of NMT

Effect of Lys7 in the myristoylation motif on

selectivity for the amino acid at position 3 in

protein N-myristoylation

Podell and Gribskov recently developed a program

(plantsp) for the prediction of N-myristoylation sites

in plant proteins [30] The construction of this

pro-gram relied on 80 plant proteins selected on the basis

of direct evidence for their N-myristoylation,

subcellu-lar localization, and N-terminal sequence conservation

We examined the N-terminal sequences of these 80

proteins, and categorized them into four groups: 18

proteins (22.5%) without Ser6 and Lys7, designated

the Met-Gly-Xaa-Xaa-Xaa-[^Ser]-[^Lys] group (where

[^Ser] and [^Lys] mean that Ser and Lys are excluded);

26 proteins (32.5%) with Ser6 but without Lys7,

desig-nated the Met-Gly-Xaa-Xaa-Xaa-Ser-[^Lys] group; 14

proteins (17.5%) without Ser6 but with Lys7,

desig-nated the Met-Gly-Xaa-Xaa-Xaa-[^Ser]-Lys group;

and 22 proteins (27.5%) with both Ser6 and Lys7,

designated the Met-Gly-Xaa-Xaa-Xaa-Ser-Lys group

The numbers of individual amino acids located at

position 3 in each group were then counted (Fig 4) In the Met-Gly-Xaa-Xaa-Xaa-[^Ser]-[^Lys] group, most proteins (17 of 18) had Asn3 (Fig 4A), whereas vari-ous amino acids were present at this position in the Xaa-Xaa-Ser-[^Lys] and Met-Gly-Xaa-Xaa-Xaa-Ser-Lys groups (Fig 4B,D) These results are consistent with the amino acid requirements at posi-tion 3 for protein N-myristoylaposi-tion shown in Fig 3

On the other hand, although proteins in the Met-Gly-Xaa-Xaa-Xaa-[^Ser]-Lys group do not have Ser6, five amino acids – Cys, Thr, Asn, Leu, and Gln – were present at position 3 (Fig 4C) Therefore, to investi-gate whether Lys7 also affects the selectivity for amino acids at position 3 in the plant myristoylation motif,

we constructed cDNAs encoding Myr–AGG1-3X6A7K and Myr–AGG1-3X6S7K, corresponding to Myr–AGG1-3X6A and Myr–AGG1-3X6S, respec-tively, with Ala7 changed to Lys (Fig 5A) We then examined these constructs for initiator Met elimination and N-myristoylation by metabolic labeling All of the constructs were efficiently translated as determined from the incorporation of [14C]Leu (Fig 5B,C), and the pattern of [35S]Met incorporation was the same as that for the corresponding Myr–AGG1-3X6A and Myr–AGG1-3X6S constructs (data not shown) Label-ing with [14C]myristic acid revealed that the selectivity for amino acids at position 3 for N-myristoylation in the Myr–AGG1-3X6A7K constructs was the same as that observed with Myr–AGG1-3X6S; that is, 12 amino acids – Gly, Ala, Ser, Cys, Thr, Val, Asn, Leu, Ile, Gln, His, and Met – were permitted at position 3 for N-myristoylation (Fig 5B) In the case of Myr– AGG1-3X6S7K, [14C]myristic acid incorporation was detected in all constructs, with the exception of those with Pro3 or Asn3 (Fig 5C), indicating that selectivity for the amino acid at position 3 for N-myristoylation

in Myr–AGG1-3X6S7K was extended relative to that

in Myr–AGG1-3X6S or Myr–AGG1-3X6A7K Together, these results indicated that not only Ser6 but also Lys7 contributes to selectivity for the amino acid

at position 3 for protein N-myristoylation in plants The combination of Ser6 and Lys7 in the motif thus allows more varieties of acceptable amino acids at position 3 than that of Ser6 and [^Lys]7 This result updated the consensus sequence for N-myristoylation

In vitro N-myristoylation of A thaliana proteins predicted not to be candidates for the modification

To perform an N-myristoylation assay based on the results shown in Figs 3 and 5, we selected eight genes from a search of potential N-myristoylated proteins in

the A thaliana database All of these proteins were

predicted not to be candidates for N-myristoylation

with the plantsp program, but they possess an amino

acid sequence consistent with the new variation of the

myristoylation consensus motif identified in the present

study and based on the combination of amino acids at

positions 3, 6 and 7 (Table 2) We analyzed the eight

proteins for potential N-myristoylation with the wheat

germ cell-free translation system In the presence of

[14C]Leu, the translation products of At1G64850,

At4G00305, At3G55450, At5G03200 and At5G03870

were detected at positions corresponding to their

expected molecular masses (18, 14, 43, 38 and 43 kDa,

respectively), whereas those of At1G66480, At3G18430

and At5G64690 were detected at positions

correspond-ing to molecular masses larger than those calculated

(25, 20 and 38 kDa, respectively) (Fig 6A) In all

instances, proteins labeled with [14C]myristic acid were

detected at positions similar to those of the proteins

labeled with [14C]Leu (Fig 6B), indicating that all

eight proteins were N-myristoylated With the use of cell-free analysis, we were thus able to identify novel substrates for N-myristoylation that had not previ-ously been shown to undergo such modification by biochemical analysis and were not predicted to do so with the plantsp program

Discussion

Protein N-myristoylation promotes the membrane association that is essential for appropriate protein localization, and N-myristoylated proteins play key roles in various cellular functions [1,4,8] We have applied an advanced wheat germ cell-free translation system as a tool to characterize protein N-myristoyla-tion in plants We first evaluated the utility of this system for analysis of N-myristoylated proteins N-myristoylation of the tested proteins was detected

by labeling with [14C]Leu and [14C]myristic acid, and was confirmed by MS analysis

Fig 4 Combination of amino acids at positions 3, 6 and 7 in 80 plant proteins with a myristoylation consensus motif The numbers of the different amino acids located at position 3 in the 80 plant proteins with a myristoylation consensus motif listed in a recent study [30] were counted Results are shown for proteins in the Gly-Xaa-Xaa-Xaa-[^Ser]-[^Lys] group (A), Gly-Xaa-Xaa-Xaa-Ser-[^Lys] group (B), Met-Gly-Xaa-Xaa-Xaa-[^Ser]-Lys group (C), or Met-Gly-Xaa-Xaa-Xaa-Ser-Lys group (D).

Given that myristic acid is attached to a Gly at the

N-terminus of a protein, the initiating Met must be

cleaved by MAP prior to N-myristoylation We have

previously shown that a Pro at position 3 markedly

inhibits cleavage of the initiator Met by MAP, even if

the penultimate amino acid is Ala, Cys, Gly, Pro, Ser,

or Thr, all of which generally allow efficient Met

removal from a translated polypeptide [35] Before this

finding, the antepenultimate amino acid had not been

thought to affect the substrate selectivity of MAP

Rather, a Pro at position 3 had been considered to

affect substrate recognition by NMT [34] In the

pres-ent study, by taking advantage of the fact that the

mature AGG1 polypeptide does not contain Met, we

also examined the efficiency of elimination of the

initiator Met in a series of mutants by labeling with [35S]Met We found that the translation products of AGG1 mutants containing Pro3 retained [35S]Met, indicating that cleavage of the initiating Met by MAP did not occur This result is thus consistent with our previous data obtained from analysis of the sequence specificity and efficiency of endogenous MAP activity

in the wheat germ cell-free translation system [35] The AGG1 mutant proteins containing Pro3 were not mod-ified by N-myristoylation We thus demonstrated that constructs containing the amino acid sequence Met-Gly-Pro at their N-terminus are not myristoylated, primarily because of the substrate specificity of MAP With respect to the consensus motif for N-myristoy-lation, although Gly2 is absolutely required for protein

Table 2 Selected A thaliana proteins for analysis of N-myristoylation in vitro N-myristoylation prediction was performed with the PLANTSP N-myristoylation prediction program [30].

A thaliana

database

entry

GenBank

accession

number

N-terminal sequence

N-myristoylation prediction

Prediction

A

B

C

Fig 5 Effects of Ser6 and Lys7 in the myristoylation consensus motif on selectiv-ity for the amino acid at position 3 in N-myri-stoylation (A) Structure of mature AGG1 fused at its N-terminus to the myristoylation motifs Ala-Ala-Lys-Ala-Ala-Ala (Myr–AGG1-3X6A7K) or Met-Gly-Xaa-Ala-Ala-Lys-Ala-Ala-Ala- Met-Gly-Xaa-Ala-Ala-Ser-Lys-Ala-Ala-Ala (Myr–AGG1-3X6S7K) (B, C) Each of the 20 mRNAs corresponding

to 3X6A7K or Myr–AGG1-3X6S7K was translated with the use of a wheat germ cell-free translation system in the presence of [14C]Leu (upper panels) or [ 14 C]myristic acid (lower panels) Portions of the reaction products were analyzed by SDS ⁄ PAGE on a 15% gel and autoradiogra-phy The molecular masses of labeled trans-lation products are shown on the right.

N-myristoylation, not all proteins with Gly2 are

N-myristoylated Previous studies have described

pref-erences for certain amino acids at distinct positions

downstream of the N-terminal Gly Mammalian

cell-free translation systems have shown that Ser6 greatly

affects the selectivity for amino acids at position 3 in

protein N-myristoylation [21,34] In addition, a set of

empirical rules for a myristoylation motif has been

proposed on the basis of the structure of

Saccharomy-ces cerevisiae NMT and of in vitro kinetic analysis of

the purified protein and synthetic peptide substrates

[19,36] These rules do not allow Pro, Asp, Glu, His,

Phe, Lys, Tyr, Trp or Arg at position 3; an arbitrary

amino acid is permitted at positions 4 and 5; only Ser,

Thr, Ala, Gly, Cys and Asn are permitted at

posi-tion 6; and only Pro among the 20 amino acids is not

allowed at position 7 By evaluating N-myristoylation

efficiency for a series of AGG1 mutants with a fused

myristoylation motif, we have now shown that not

only Ser6 but also Lys7 influences selectivity of

N-myr-istoylation for the amino acid at position 3

Further-more, in the case of the constructs containing Ser6 and

Lys7, only Pro and Asn were not allowed at position 3 for N-myristoylation These results indicate that selec-tivity for the amino acid at position 3 in the plant myristoylation motif is more complex than that in

S cerevisiae

Regarding other amino acid positions, such as posi-tions 4 and 5, we have not yet investigated the precise selectivity of amino acids in the plant N-myristoylation system Investigations of these positions could further update the consensus sequence

A prediction program for N-myristoylation of plant proteins, plantsp, was recently developed [30] How-ever, given that only a small number of

N-myristoylat-ed proteins have been experimentally confirmN-myristoylat-ed in plants, the accuracy of N-myristoylation prediction has not yet achieved sufficient reliability Indeed, our

in vitro analysis of AGG1 mutants revealed that the prediction program was not sufficiently effective for evaluation of N-terminal amino acid sequences as potential N-myristoylation motifs (Table S1) For example, although effective N-myristoylation was observed in Myr–AGG1-3X6S constructs with Gly, Ala, Ser, Thr, Val, Leu, Ile, Gln, His or Met at posi-tion 3, the predicposi-tion program gave low scores for the possibility of N-myristoylation of these constructs (Table S1) We further selected eight A thaliana pro-teins on the basis of combinations of amino acids at positions 3, 6 and 7 that our results suggested would

be compatible with N-myristoylation, and we exam-ined whether this was the case with the cell-free sys-tem Although these proteins were predicted to be negative candidates by the plantsp program, we detected their myristoylation in vitro (Fig 6) To date,

319 proteins of A thaliana, representing 1.1% of the total proteome, have been predicted to be N-myristoy-lation candidates [30] According to the results obtained from our experiments, we further surveyed

A thalianaproteins that are potentially

N-myristoylat-ed Using the pattern matching program in The Arabidopsis Information Resource, we listed 103

A thaliana proteins (Table S2) Among these, 79 were included in the above-mentioned 319 proteins, but the other 24 were newly found candidates These results suggest that the actual number of N-myristoylated proteins in A thaliana may be substantially larger than the number previously predicted

Our investigation is based on wheat germ extracts

We therefore need to note that the substrate specificity

of the wheat NMT could be technically different from that of the A thaliana enzyme On the other hand, a single-copy gene encoding an AtNMT1 homolog (73% identity) has been reported previously as a candidate gene for wheat NMT [16] We suppose that the wheat

A

B

Fig 6 In vitro N-myristoylation of A thaliana proteins predicted

not to be substrates for the modification Eight proteins, all of

which were predicted to be negative candidates for

N-myristoyla-tion by the PLANTSP program, were translated in vitro with a wheat

germ cell-free translation system in the presence of [ 14 C]Leu (A) or

[14C]myristic acid (B) Portions of the reaction products were

ana-lyzed by SDS ⁄ PAGE and autoradiography Open and closed

arrow-heads indicate the positions of full-length proteins and products of

aborted translation, respectively.

homolog of AtNMT1 is most likely the enzyme

responsible for the protein N-myristoylation occurring

in the wheat germ extracts

The wheat germ cell-free N-myristoylation system is

thus useful for the detection and characterization of

N-myristoylated proteins in plants The substrate

spec-ificity for N-myristoylation of plant proteins revealed

by the wheat germ cell-free myristoylation system will

facilitate the preparation of N-myristoylated proteins

at a preparative scale, as well as the high-throughput

and comprehensive proteomic analysis of

N-myristoy-lated plant proteins

Experimental procedures

Plasmid construction

All restriction endonucleases and DNA-modifying enzymes

for plasmid construction were obtained from Takara Shuzo

(Kyoto, Japan) We modified the pEU3S cell-free

expres-sion vector [37] by introducing a nucleotide sequence into

the 3¢-end of the multiple cloning site, so that the

C-termi-nus of the encoded polypeptide would be conjugated to a

His6 tag To prepare a DNA fragment encompassing the

modified multiple cloning site, we performed PCR with

KOD plus DNA polymerase (Toyobo, Osaka, Japan),

primers pEU3S FW and pEU3S RV (Table S3), and the

pEU3S vector as a template The amplified DNA fragment

was digested with SpeI and NcoI, and then cloned into the

corresponding SpeI–NcoI sites of pEU3S The resulting

plasmid was designated pEU3SH

The coding regions of RGLG2 (GenBank accession

no NM_203051) and ARF1A1c (NM_130285) of A

thali-anawere obtained by RT-PCR with Superscript III reverse

transcriptase (Invitrogen, Carlsbad, CA, USA), KOD plus

DNA polymerase (Toyobo), and corresponding PCR

prim-ers (Table S3) The PCR product for RGLG2 was digested

with SpeI and SmaI, and that for ARF1A1c was digested

with XhoI and SmaI The resultant DNA fragments were

then cloned into the corresponding sites of pEU3SH For

expression of N-myristoylated GFP and E coli DHFR,

each coding region fused with a DNA sequence encoding

the myristoylation motif

Met-Gly-Ala-Ala-Ala-Ser-Ala-Ala-Ala-Ala was amplified by PCR with appropriate primers

(Table S3), digested with SpeI and SmaI, and cloned into

the corresponding sites of pEU3SH Each G2A construct

was prepared by PCR with corresponding G2A primers

(Table S3) and the corresponding pEU3SH-based vector as

the template, and was then cloned into pEU3SH as

described above

For further analysis of N-myristoylation of A thaliana

proteins, the coding regions of At1G64850 (NM_105159),

At1G66480 (NM_105319), At3G18430 (NM_112728),

At3G55450 (NM_115403), At4G00305 (NM_116252),

At5G03200 (NM_120398), At5G03870 (NM_120468) and At5G64690 (NM_125865) were obtained by RT-PCR as described above and with the primers listed in Table S3 The PCR products of At1G66480, At3G18430 and At5G03870 were digested with SpeI and BamHI, and the other PCR products were digested with SpeI and BglII The resultant DNA fragments were then cloned into the SpeI and BglII sites of the pEU3b vector [37]

For analysis of the sequence specificity of the myristoyla-tion motif, the coding region of A thaliana AGG1 (NM_116207) was obtained by RT-PCR as described above and with appropriate primers (Table S3) The PCR product was cloned into the pTA2 vector with the use of TArget Clone (Toyobo), yielding pTA2–AGG1 The coding region for mature AGG1 fused with a DNA fragment encoding the myristoylation motif Met-Gly-Ala-Ala-Ala-Ala-Ala-Ala-Ala-Ala or Met-Gly-Met-Gly-Ala-Ala-Ala-Ala-Ala-Ala-Ala-Ala-Ala-Ser-Met-Gly-Ala-Ala-Ala-Ala-Ala-Ala-Ala-Ala-Met-Gly-Ala-Ala-Ala-Ala-Ala-Ala-Ala-Ala was amplified by PCR with the primers AGG1 RV and 3A6(A⁄ S) FW (Table S3) and with pTA2–AGG1 as the template The resultant DNA fragments were cloned into the SpeI and SmaI sites of pEU3SH for production of Myr– AGG1-3A6A or Myr–AGG1-3A6S, respectively The cDNAs encoding 3X6A and 3X6S, corresponding to 3A6A or Myr–AGG1-3A6S, respectively, with Ala3 replaced with each of the other

19 amino acids, were constructed by PCR with the primer pEU3SH RV and the corresponding 3X6(A⁄ S) FW primer (Table S3) and with the Myr–AGG1-3A6A and Myr– AGG1-3A6S vectors as templates The resultant DNA frag-ments were digested with SpeI and BglII, and then cloned into the corresponding sites of pEU3b For further substitu-tion of Lys for Ala at posisubstitu-tion 7 of Myr–AGG1-3X6A, the cDNA encoding Myr-AGG1-3A6A7K was constructed by PCR with the primers pEU3SH RV and 3A6A7K FW (Table S3) and with the Myr–AGG1-3A6A vector as tem-plate The PCR product was cloned into pEU3b as described above The cDNAs encoding Myr–AGG1-3X6A7K were constructed by an approach similar to that used for con-struction of Myr–AGG1-3X6A cDNAs, with the primers listed in Table S3 and with the Myr–AGG1-3A6A7K vector

as the template The cDNAs encoding Myr–AGG1-3X6S7K, corresponding to Myr–AGG1-3X6A7K with Ser substituted for Ala6, were constructed as described above with appropriate primers (Table S3) and with the corre-sponding Myr–AGG1-3X6A7K vector as template The sequences of all constructs were confirmed by DNA sequenc-ing with the use of an ABI PRISM 310 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA)

In vitro N-myristoylation assay For detection of N-myristoylation of A thaliana proteins, the corresponding plasmid (10 lg) purified with the use of

a Qiagen Midi Kit (Qiagen, Chatsworth, CA, USA) was used as a template for in vitro transcription with SP6 RNA

polymerase (Promega, Madison, WI, USA) In the case of

analysis of the sequence specificity of the myristoylation

motif, template DNAs for in vitro transcription were

pre-pared from the Myr–AGG1-3X6A, Myr–AGG1-3X6S,

Myr–AGG1-3X6A7K and Myr–AGG1-3X6S7K vectors by

PCR Proteins were synthesized by the batch method, with

the use of a wheat germ cell-free translation system (Wepro;

CellFree Sciences, Matsuyama, Japan) in the presence of

[14C]Leu (Perkin Elmer, Waltham, MA, USA), [14C]myristic

acid (American Radiolabeled Chemicals, St Louis, MO,

USA), or [35S]Met (American Radiolabeled Chemicals),

under the conditions recommended by the manufacturer

[38] The batchwise reaction mixture (25 lL) contained

wheat germ extract (final concentration, 60 A260 nm units),

4.7 lL of reaction buffer, creatine kinase (0.4 mgÆmL)1),

5 lL of mRNA, and either 0.5 lL of [14C]Leu (316

CiÆmol)1, 100 lCiÆmL)1), 0.5 lL of [14C]myristic acid (55

CiÆmol)1, 100 lCiÆmL)1), or 0.25 lL of [35S]Met (810.3

CiÆmmol)1, 10.89 mCiÆmL)1), and was incubated at 26C

for 3 h Samples were denatured by boiling for 3 min in

SDS sample buffer, and were then analyzed by

SDS⁄ PAGE on a 15% gel The gel was dried under

vacuum and then subjected to autoradiography

Protein purification

Myristoylated or nonmyristoylated proteins were synthesized

with the cell-free translation system with or without the

addi-tion of myristic acid (Nacalai tesque, Kyoto, Japan) at a

final concentration of 75 lm The reaction mixture was

cen-trifuged at 20 400 g for 20 min at 4C, and the resulting

supernatant was mixed with 10 volumes of a solution

con-taining 50 mm Tris⁄ HCl (pH 7.5), 500 mm NaCl, and

20 mm imidazole (buffer A) and then applied to a 1 mL

HiTrap chelating column (GE Healthcare, Little Chalfont,

UK) that had been equilibrated with buffer A The column

was washed with 10 mL of buffer A containing 40 mm

instead of 20 mm imidazole, and then with 10 mL of

buf-fer B (50 mm Tris⁄ HCl, pH 7.5, 50 mm NaCl, 40 mm

imid-azole) Elution was then performed with buffer B containing

400 mm instead of 40 mm imidazole, as well as 10% glycerol,

and the eluate was stored at)80 C until use

MALDI-TOF MS analysis

Purified N-myristoylated or nonmyristoylated proteins were

separated by SDS⁄ PAGE on a 12.5% gel and stained

with Coomassie Brilliant Blue R250 The stained protein

bands were excised from the gel, incubated in 100 lL of

30% acetonitrile containing 25 mm ammonium bicarbonate

for 10 min to remove the stain, dehydrated with 100 lL

of 100% acetonitrile for 5 min, and dried for 15 min

under vacuum The protein bands were then reduced and

alkylated before digestion and extraction with the use of an

XL-trypKit (Promega) The extracted peptides were

desalt-ed and concentratdesalt-ed with the use of a ZipTip C18 device (Millipore, Billerica, MA, USA), and samples eluted from the device with 1 lL of saturated a-cyano-4-hydroxycin-namic acid (Sigma-Aldrich, St Louis, MO, USA) in 90% acetonitrile containing 0.1% trifluoroacetic acid were spot-ted onto a MALDI target plate (Applied Biosystems) MALDI-TOF mass spectra were acquired with a Voyager

DE Biospectrometry Workstation (Applied Biosystems) as described previously [35]

Bioinformatics analysis Proteins containing objective sequences in the myristoylation motif were searched for in The Arabidopsis Information Resource (http://arabidopsis.org), with the use of the pattern matchingprogram (http://arabidopsis.org/cgi-bin/ patmatch/nph-patmatch.pl) The prediction of protein N-myristoylation was performed with the plantsp program (http://plantsp.genomics.purdue.edu/html/myrist.html) [30]

Acknowledgement

This work was supported by a grant-in-aid for Scien-tific Research on Priority Areas (Plant Membrane Transport) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (no

20053016 to Y Tozawa)

References

1 Resh MD (1999) Fatty acylation of proteins: new insights into membrane targeting of myristoylated and palmitoylated proteins Biochim Biophys Acta 1451, 1–16

2 Batisticˇ O, Sorek N, Schu¨ltke S, Yalovsky S & Kudla J (2008) Dual fatty acyl modification determines the localization and plasma membrane targeting of CBL⁄ CIPK Ca2+signaling complexes in Arabidopsis Plant Cell 20, 1346–1362

3 Benetka W, Mehlmer N, Maurer-Stroh S, Sammer M, Koranda M, Neumu¨ller R, Betschinger J, Knoblich JA, Teige M & Eisenhaber F (2008) Experimental testing of predicted myristoylation targets involved in asymmetric cell division and calcium-dependent signaling

Cell Cycle 7, 3709–3719

4 Boutin JA (1997) Myristoylation Cell Signal 9, 15–35

5 Chen CA & Manning DR (2001) Regulation of G proteins

by covalent modification Oncogene 20, 1643–1652

6 O’Callaghan DW & Burgoyne RD (2003) Role of myristoylation in the intracellular targeting of neuronal calcium sensor (NCS) proteins Biochem Soc Trans 31, 963–965

7 Utsumi T, Ohta H, Kayano Y, Sakurai N & Ozoe Y (2005) The N-terminus of B96Bom, a Bombyx mori