changes in targeting efficiencies of proteins to plant microbodies

Plant Cell Physiol. 38(6): 759-768 (1997)JSPP © 1997

Changes in Targeting Efficiencies of Proteins to Plant Microbodies Causedby Amino Acid Substitutions in the Carboxy-terminal Tripeptide

Makoto Hayashi, Masahiro Aoki, Maki Kondo and Mikio Nishimura'Department of Cell Biology, National Institute for Basic Biology, Okazaki, 444 Japan

It has been demonstrated that the carboxyl terminusof microbody enzymes functions as a targeting signal to mi-crobodies in higher plants. We have examined an ability of24 carboxy-terminal amino acid sequences to facilitate thetransport of a cytosolic passenger protein, /^glucuroni-dase, into microbodies in green cotyledonary cells of trans-genic Arabidopsis. Immunoelectron microscopic analysisrevealed that carboxy-terminal tripeptide sequences of theform [C/A/S/P]-[K/R]-[I/L/M] function as a microbody-targeting signal, although tripeptides with proline at thefirst amino acid position and isoleucine at the carboxyl ter-minus show weak targeting efficiencies. All known micro-body enzymes that are synthesized in a form similar in sizeto the mature molecule, except catalase, contain one ofthese tripeptide sequences at their carboxyl terminus.

Key words: Arabidopsis thaliana — Glyoxysome — Leafperoxisome — Microbody — Protein sorting — Targetingsignal.

Three functionally different types of microbodies (gly-oxysomes, leaf peroxisomes and unspecialized micro-bodies) are known to exist in plant tissues (Beevers 1979,Huang et al. 1983). Glyoxysomes contain enzymes foryS-ox-idation of fatty acids and for the glyoxylate cycle. These en-zymes are engaged in the conversion of storage lipids to su-crose in the cells of germinated fatty seeds and senescingtissues (De Bellis et al. 1991). Leaf peroxisomes are foundin photosynthetic tissues, such as green cotyledons andmature leaves. They contain enzymes involved in glycolatemetabolism, such as glycolate oxidase and hydroxypyru-vate reductase, and play a crucial role in photorespirationin combination with chloroplasts and mitochondria. Unspe-cialized microbodies are found in other tissues, but theirfunction is obscure.

Immunocytochemical analysis has revealed that glyox-ysomes can reversibly change into leaf peroxisomes andvice versa during seedling growth (Nishimura et al. 1986,

Abbreviations: MS, Murashige and Skoog; PMSF, phenyl-methanesulphonyl fluoride.1 To whom correspondence should be addressed.

Titus and Becker 1985). Since microbodies do not havetheir own protein synthesizing system, the functional differ-entiation of these microbodies requires the import of spe-cific proteins that are synthesized in the cytosol. Two typesof proteins are known to independently target micro-bodies. One type, which includes most microbody proteins,are proteins that are synthesized in a form similar in size tothe mature molecule, while the other proteins are synthe-sized as precursor molecules with larger molecular masses(Olsen and Harada 1995).

Targeting signals responsible for the transport of pro-teins into microbodies have been extensively studied inmammals and yeasts. Subramani's group first revealed thata unique tripeptide sequence, SKL at the carboxyl terminusof firefly luciferase is necessary and sufficient for the trans-port of this protein into microbodies (Gould et al. 1987,1988). Experiments on amino acid substitutions within thetripeptide showed that tripeptides with S, A or C at the - 3position, K, R or H at the —2 position, and L at the carb-oxyl terminal position can function as targeting signalsfor microbodies in mammalian cells (Gould et al. 1989,Swinkels et al. 1992).

Volokita (1991) demonstrated that the carboxyl ter-minus of a microbody enzyme, glycolate oxidase, functionsas a targeting signal to plant microbodies. He generatedtransgenic tobacco expressing a fusion protein consistingof /?-glucuronidase and spinach glycolate oxidase, andshowed that the carboxy-terminal six amino acids of the gly-colate oxidase is sufficient to target the chimeric protein toleaf peroxisomes. Further analysis using similar in vivo im-port systems revealed that the carboxy-terminal five aminoacids of pumpkin malate synthase and carboxy-terminalfour amino acids of rape seed isocitrate lyase were sufficientfor the transport of chimeric proteins into microbodies(Hayashi et al. 1996a, Olsen et al. 1993). These data sug-gested that peptides containing certain carboxy-terminaltripeptide sequences (ARL, SRL, SRM, SKL and PRL)function as targeting signals to glyoxysomes as well as leafperoxisomes and unspecialized microbodies in roots.Recently,- Banjoko and Trelease (1995) clearly demon-strated that the carboxy-terminal SKL tripeptide is ne-cessary and sufficient for targeting a chimeric protein toplant microbodies, as is the case in mammalian cells. How-ever, the divergence of carboxy-terminal tripeptide se-

759

Downloaded from https://academic.oup.com/pcp/article-abstract/38/6/759/1868883by gueston 07 April 2018

760 Carboxy-terminal targeting signal to microbodies

quences that function as microbody-targeting signals inplant cells is still not known.

In the present study, we generated transgenic Arabi-dopsis that express fusion proteins containing various carb-oxyl termini. We examined the targeting-efficiency of thesefusion proteins to microbodies using immunoelectron mi-croscopy and discuss the sequence specificity of the micro-body-targeting signal in plants.

Construction of chimeric genes—The DNA fragmentencoding /?-glucuronidase was amplified by the polymerasechain reaction (PCR) using 10 ng of pBI221 (CLONTECH,Palo Alto, CA) as a template. The reaction mixture contain-ed 2.5 units of AmpliTaq DNA polymerase (Perkin ElmerJapan, Chiba, Japan), a 5'primer (CCGGATCCGTCGAC-CATGAGTCCGTCCTGAT), a 3' primer (CCGAATTC-TCAGATCTCTTGTTTGCCTCCCTGCTGC), and an ap-propriate buffer in a total volume of 50//I. Each cycle ofPCR consisted of 95°C for 45 s, 60°C for 45 s and 72°C for45 s. The PCR product obtained after 25 cycles ofamplification was then incubated for 2 h at 72°C to in-troduce deoxyadenosine at the 3' ends. The DNA fragmentwas then subcloned into a T-vector prepared using Blue-script KS+ (Stratagene, La Jolla, CA) as described in a pre-vious report (Marchuk et al. 1990). The DNA fragment in-serted in the vector contained a BarriHl site at the 5' flank-ing region and Bglll and £coRI sites on either side of thestop codon of /?-glucuronidase.

In order to make chimeric transgenes containing an ad-ditional amino acid sequence at the carboxyl terminus of /?-glucuronidase, the BgM-EcoRl fragment of the DNA frag-ment was replaced with various double-stranded DNAsproduced by the annealing of complementary oligonucleo-tides. The complementary oligonucleotides were designedto produce protruding ends that can ligate to the Bgfll and£coRI site of the /?-glucuronidase gene. The nucleotide se-quences of the coding strands of the synthetic oligonucleo-tides used to construct chimeric genes are shown in Table 1.

The BamHl-EcoRl fragments that contain /?-glucu-ronidase as well as various chimeric genes were insertedinto the Bgfll-EcoRl site of a Ti-plasmid, pMAT037 (Mat-suoka and Nakamura 1991). Standard procedures wereused for all DNA manipulations (Sambrook et al. 1989).The Ti-plasmids produced were then transformed intoAgrobacterium tumefaciens (strain C58ClRifR) by elec-troporation (Nagel et al. 1990).

Plant transformation and selection of transformants—Transformation of Arabidopsis thaliana (ecotype C24)using Agrobacterium tumefaciens was performed accord-ing to the method of Valvekens et al. (1988). Primary trans-formants were designated TO plants. Tl seeds collectedfrom TO plants were surface sterilized in 2% NaCIO,0.02% Triton X-100, and grown on germination media(2.3 m g m P 1 MS salt (Wako, Osaka, Japan), \% sucrose,100ng ml"1 myo-inositol, l / zgmF 1 thiamine-HCl, 0.5 fig

ml ' pyridoxine, 0.5 fig ml"1 nicotinic acid, 0.5mgml 'MES-KOH (pH5.7), 0.2% gellan gum (Wako, Osaka,Japan)) containing lOO^gml"1 of kanamycin. T2 seedswere collected from approximately 10 independent Tlplants. T2 plants that accumulated the highest amount oftransgene product were selected on the basis of immuno-blotting using a /?-glucuronidase-specific antibody. Homo-zygous T3 plants obtained from progenies of the T2 plantwere scored by kanamycin resistance.

Immunoblotting—Transgenic Arabidopsis were grownon germination media containing lOO^gml"1 kanamycinfor 7 d under continuous illumination. Seedlings were ho-mogenized in 100 fi\ of buffer containing 50 raM Tris-HCl(pH 8.3), 1 mM EDTA and 1 mM PMSF. The homogenatewas centrifuged for 15 min at 15,000 *g at 4°C to removecell debris and the supernatant was used as the crude ex-tract. The amount of total protein in the crude extract wasmeasured using the Bio-Rad Protein Assay (Bio-Rad, Her-cules, CA).

Nine fig of total protein were loaded onto lanes ofan 10% SDS-polyacrylamide gel. After electrophoresis,proteins were transferred to a nitrocellulose membrane(Schleicher & Schuell, Dassel, Germany) in a semi-dry elec-troblotting system. The membrane was blocked with 3%nonfat dry milk in Tris-buffered saline, pH 7.4 and immu-noblotted with a 1 : 1,000 dilution of /?-glucuronidase-spe-cific antibody (Molecular Probe, Eugene, OR, U.S.A.).Bands were visualized with an ECL Western blotting de-tection kit (Amersham Japan, Tokyo, Japan) using a 1 :5,000 dilution of peroxidase-conjugated goat antibodiesagainst rabbit IgG following the instructions of the manu-facturer.

Immunoelectron microscopy—Immunoelectron mi-croscopy was performed as described previously (Nishi-mura et al. 1993). Green cotyledons were obtained fromhomozygous T3 plants grown on germination media con-taining 100figm\~l kanamycin for 7d under continu-ous illumination. They were vacuum-infiltrated for 1 hwith a fixative that consisted of 4% paraformaldehyde, 1%glutaraldehyde and 0.06 M sucrose in 0.05 M cacodylatebuffer (pH 7.4). The fixed samples were then cut into slicesof less than 1 mm in thickness and treated for another2 h with freshly prepared fixative. After washing with thesame buffer, the samples were dehydrated in a gradeddimethylformamide series at — 20° C and embedded in LR-White resin (London Resin Co. Ltd., Basingstoke, Hamp-shire, U.K.). Blocks were polymerized under a UV lamp at— 20°C for 24 h. Ultrathin sections were mounted on un-coated nickel grids. The sections were treated with blockingsolution (\% bovine serum albumin in phosphate-bufferedsaline) for 1 h at room temperature, and were then incubat-ed overnight at 4°C in a solution of yS-glucuronidase-spe-cific antibodies that had been diluted 1 : 500 in the blockingsolution at 4°C. After washing with phosphate-buffered


Carboxy-terminal targeting signal to microbodies 761

saline, sections were incubated for 30 min at room tempera-ture in a solution of protein A-gold (15 nm; AmershamJapan, Tokyo, Japan) that had been diluted 1 :20 in theblocking solution. The sections were washed with distilledwater and then stained with 4% uranyl acetate and leadcitrate. After staining, all sections were examined under atransmission electron microscope (1200EX; JEOL, Tokyo,Japan) operated at 80 kV.

Subcellular fractionation—Homozygous T3 seedswere grown on germination medium at 22°C for 14 d undercontinuous illumination. The green plants (0.4 g in freshweight), excluding roots, were chopped with a razor bladein a Petri dish with 2.0 ml of chopping buffer (150 raMTricine-KOH; pH 7.5, 1 mM EDTA, 0.5 M sucrose). Theextract was then nitrated with cell strainer (Becton Dickin-son, Franklin Lakes, NJ). Two ml of the homogenate waslayered directly on top of a 16-ml linear sucrose density gra-dient (30-60%, [w/w]) that contained 1 mM EDTA. Cen- .trifugation was performed in an SW 28.1 rotor (Beckman,Palo Alto, CA) at 25,000 rpm for 2.5 h at 4°C. One ml frac-tions were collected using a gradient fractionator (model185; ISCO, Lincoln, NE). Twenty//I of each fraction was

o c o c s

1 3 7 -

6 3 -

Fig. 1 Expression of GUS and various chimeric genes in trans-genic Arabidopsis. Crude extracts of the green cotyledons wereprepared from wild type and transgenic/lratorfo/wiy. The sampleswere analyzed by immunoblotting using a specific antibodyagainst /S-glucuronidase after SDS-polyacrylamide gel electropho-resis. Nine /*g of total protein were applied in each lane. Label onthe top of each lane represents the chimeric gene used totransform Arabidopsis, with the exception of WT, which repre-sents wild-type Arabidopsis (ecotype C24). Numbers at the left in-dicate the sizes and positions of the molecular mass markers inkilodaltons.

subjected to immunoblotting to analyze distribution of/?-glucuronidase and chimeric proteins using /?-glucuroni-

Table 1 Oligonucleotides used for generation of GUS fusion proteins

Sequence of

ATAGAGATA

ATAATAATAATAATAATAATA

ATAATAATAATAATA

ATAATAATAATAATAATAATAATAATA

CATCTGCAT

CATCATCATCATCATCATCAT

CATCATCATCATCAT

CATCATCATCATCATCATCATCATCAT

synthetic

CATTCCCAT

CATCATCATCATCATCATCATCATCATCATCATCATCATCATCATCATCATCATCATCATCAT

oligonucleotides

CCCAGGCCC

CCCCCCCCCCCCCCCCCCCCC

CCCCCCCCCCCCCCC

CCCCCCCCCCCCCCCCCCCCCCCCCCC

AGGCTCAGG

AGGAGGAGGAGGAGGAGGAGG

AGGAGGAGGAGGAGG

AGGAGGAGGAGGAGGAGGAGGAGGAGG

used in

GAGTGAGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAGGAG

this study

CTGGCTG

CTGCTGCTGCTGCTGCTGCTG

CTGCTGCTGCTGCTG

CTGCTGCTGCTGCTGCTGCTGCTGCTG

TCC

TGA

TCCTCCTCCTCCTCCTCCtecTCCTCCTCCTCCTCC

GCCTGCGAATTCGGCAAACTCCCCTAC

AGG

G

AGGAGGAGGAGGAGGAGGAGG

GGGCACATTAAATCC

AGGAGGAGGAGGAGGAGGAGGAGGAGG

CTC

GAATTTATTAAAATGTCCGTT

CTCCTCCTCCTCCTC

CTCCTCCTCCTCCTCCTCCTCCTCCTC

TGA

TGATGATGATGATGATGATGA

TGATGATGATGATGA

TGATGATGATGATGATGATGATGATGA

G

GGGGGGG

GGGGG

GGGGGGGGG

Abbreviation offusion proteins

SRLMS5ASRL

SRESRFSRISRKSRMSRSSRVSGLSHLSILSKLSSLARLCRLERLFRLGRLKRLLRLPRLYRL



dase-specific antibody, and catalase using catalase-specificantibody (Yamaguchi and Nishimura 1984).

Expression of chimeric genes in transgenic Arabi-dopsis—To investigate the role of carboxy-terminalsequences in targeting proteins to microbodies, we con-structed chimeric genes that encoded /3-glucuronidase fu-sion proteins having various carboxy-terminal sequences.The chimeric genes were produced by replacing the ter-minal codon of the )S-glucuronidase gene with varioussynthetic complementary oligonucleotides. Nucleotide se-quences of the coding strands of the synthetic oligonucleo-tides as well as the names of the chimeric genes used in thisstudy are shown in Table 1. The original carboxy-terminalamino acid sequence of /?-glucuronidase (GUS) is QG-GKQ. SRL encoded a chimeric protein that combined theten carboxy-terminal amino acids of pumpkin malate syn-thase (Mori et al. 1991) at the carboxyl terminus of y3-glucu-

ronidase. MS5 contained the gene for GUS with an addi-tional five carboxy-terminal amino acids (ELSRL) thatwere the same as those of pumpkin malate synthase,whereas JSRL encoded a protein that were the same asthat encoded by SRL except that the three carboxy-ter-minal amino acids were deleted. Other chimeric gene con-structs encoded proteins that were the same as those encod-ed by SRL except that they contained single amino acidsubstitutions in the three carboxy-terminal amino acids.The names of these chimeric genes represent the sequencesof their three carboxy-terminal amino acids.

GUS and the chimeric genes were inserted downstreamof the cauliflower mosaic virus 35S promoter in the binaryplant expression vector pMAT037 (Matsuoka and Naka-mura 1991). The constructs were transferred into Arabi-dopsis by the Agrobacterium tumefaciens-media.ted genetransfer procedure (Valvekens et al. 1988). Primary trans-

H

Fig. 2 Ettects of substitutions at the carboxy-terminal amino acid on subcellular localization of chimeric proteins. Subcellular localiza-tion of chimeric proteins in the cells of transgenic Arabidopsis was analyzed by immunoelectron-microscopy. Sections of 7-day old greencotyledons prepared from Arabidopsis transformed using various chimeric genes were incubated with /J-glucuronidase-specific antibody,and then visualized with protein A-gold. Chimeric genes transformed into each transgenic Arabidopsis are; SRL (A), JSRL (B), SRE(C), SRF (D), SRI (E), SRK (F), SRM (G), SRS, (H) and SRV (I). Bar on (I) indicates 1 urn.



formants were designated TO plants. We selected one T2plant of each construct that accumulated the highestamount of the transgene product based on the results of im-munoblotting using a /?-glucuronidase-specific antibody.

Homozygous T3 plants, which were obtained from theprogenies of the T2 plants, were used for further analyses.As shown in Figure 1, homozygous T3 plant yielded singleimmunoreactive band with a molecular mass similar to thatof GUS, while non-transformed wild-type Arabidopsis didnot. This result indicated that fusion proteins expressedfrom the chimeric genes accumulated at similar level ingreen cotyledons of all T3 plants, and excluded the posibili-ty that levels of expression of these transgenes were marked-ly different among the transgenic plants analyzed in thepresent study.

Effects of amino acid substitutions at the carboxyl ter-minus—To quantify targeting efficiencies of proteins con-taining various carboxy-terminal amino acid sequences tomicrobodies, thin sections of green cotyledons of trans-genic Arabidopsis expressing various chimeric genes wereanalyzed by immunoelectron microscopy after stainingwith antibodies against /?-glucuronidase and then with pro-tein A-gold. We have previously shown that the carboxy-terminal amino acids of pumpkin malate synthase functionas targeting signals for glyoxysomes, leaf peroxisomes andunspecialized microbodies in roots (Hayashi et al. 1996a).In agreement with this, gold particles were observed ex-clusively within microbodies in green cotyledons of trans-

genic Arabidopsis expressing the fusion proteins encodedby SRL and MS5 (black dots in Fig. 2A, 3A). In contrast,no signal was detected in thin sections of green cotyledonsexpressing the fusion protein encoded by .dSRL (Fig.2B)or the unmodified /?-glucuronidase gene (data not shown).

Figure 2 shows the effects of single amino acid sub-stitutions at the carboxyl terminus of the fusion proteins onmicrobody-targeting efficiencies. We found high densitiesof gold particles within microbodies of Arabidopsis trans-formed with SRL (Fig. 2A) and SRM (Fig. 2G). A lowerdensity of gold particles within microbodies was observedin Arabidopsis transformed with SRI (Fig. 2E), in spite ofthe fact that similar amount of fusion proteins accumulat-ed in transgenic Arabidopsis transformed with SRI, SRLand SRM (see Fig. 1). In contrast, no signal was detected re-producibly within microbodies of Arabidopsis transform-ed with SRE (Fig. 2C), SRF (Fig. 2D), SRK (Fig. 2F), SRS(Fig. 2H) or SRV (Fig. 21). These fusion proteins were nottransported into microbodies, and remained in the cytosolwhere their concentrations are too low to be detected byimmunocytochemical technique as discussed in previouspaper (Hayashi et al. 1996a).

Effects of substitutions at the second amino acid fromthe carboxyl terminus—Figure 3 shows the effects of sub-stitutions of the second amino acid from the carboxylterminus of the fusion proteins on microbody-targetingefficiencies. A high density of gold particles was observedwithin microbodies when the carboxy-terminal tripeptide

Fig. 3 Effects of substitutions at the second amino acid in the carboxy-terminal tripeptide on subcellular localization of chimeric pro-teins. Subcellular localization of chimeric proteins in the cells of transgenic Arabidopsis was analyzed by immunoelectron-microscopy.Chimeric genes transformed into each transgenic Arabidopsis are; MS5 (A), SGL (B), SHL (C), SIL (D), SKL (E)and SSL (F). Bar in (F)indicates 1 fim.



sequence was changed from SRL to SKL (Fig. 3E). How-ever, no signal was detected reproducibly within micro-bodies of Arabidopsis transformed with SGL (Fig. 3B),SHL (Fig. 3C), SIL (Fig. 3D) or SSL (Fig. 3F).

Effects of substitutions at the third amino acid fromthe carboxyl terminus—Figure 4 shows the effects of singlesubstitutions at the third amino acid from the carboxylterminus of the fusion proteins on microbody-targetingefficiencies. High densities of gold particles were observedwithin microbodies when the carboxy-terminal tripeptidesequence was changed from SRL to ARL (Fig.4A) andfrom SRL to CRL (Fig.4B). As is the case with SRI(Fig. 2E), a lower density of gold particles was observedwithin microbodies of Arabidopsis transformed with PRL(Fig.4H). At least some of the fusion protein containingPRL at the carboxyl terminus seemed to target the micro-bodies, since we have recently demonstrated with immuno-fluorescent microscopy that the fusion protein was local-

ized within glyoxysomes as well as leaf peroxisomes(Hayashi et al. 1996a). In contrast, no signal was detectedreproducibly within microbodies of Arabidopsis transform-ed with ERL (Fig.4C), FRL (Fig.4D), GRL (Fig.4E),KRL (Fig.4F), LRL (Fig.4D), GRL (Fig.4E), KRL(Fig.4F), LRL (Fig.4G) or YRL (Fig. 41).

Subcellular distribution ofGUS and chimeric proteinsin transgenic Arabidopsis plants—To confirm the data ob-tained by the immunoelectron microscopic observations,we performed subcellular fractionation of transgenic Arabi-dopsis plants transformed with GUS, SRL and 4SRL us-ing sucrose density gradient centrifugation. The distribu-tion of the proteins expressed from GUS, SRL and ASKLin the gradients was analyzed by immunoblotting using spe-cific antibody against /?-glucuronidase. As shown in Figure5, the fusion protein encoded by SRL (SRL in Fig. 5) wasdetected in the fractions 13 and 14 in addition to the firstthree fractions (fractions 1 to 3). Immunoblot analysis us-

Fig. 4 Effects of substitutions at the first amino acid in the carboxy-terminal tripeptide on subcellular localization of chimeric proteins.Subcellular localization of chimeric proteins in the cells of transgenic Arabidopsis was analyzed by immunoelectron-microscopy.Chimeric genes transformed into each transgenic Arabidopsis are; ARL (A), CRL (B), ERL (C), FRL (D), GRL (E), KRL (F), LRL (G),PRL (H) and YRL (I). Bar in (I) indicates 1 ftm.



GUS

in(3

1 2 3 4SRL

5 6 7 8 9 10 11 12 13 14 15 16

o

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16ASRL

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16fraction number

Fig. 5 Subcellular fractionation of transgenic Arabidopsisplants expressing GUS, SRL and JSRL. Transgenic Arabidopsisplants harboring GUS, SRL and JSRL (see Table 1) were grownfor 14 d under continuous illumination at 22°C. The whole plantswithout roots were homogenized and the extracts were then sub-jected to sucrose density gradient centrifugation (30-60% [w/w]).After the centrifugation, fractions were collected from the top ofthe gradient. Twenty /A of each fraction was analyzed by immuno-bloting using specific antibodies against yS-glucuronidase (a-GUS)and catalase (a-CAT).

ing antibody against catalase clearly indicated that micro-bodies were located in the fractions 13 and 14. The immu-noreactive polypeptides were also found in the fractions 1to 3, since certain amount of proteins was released frombroken microbodies during homogenization and sub-sequent centrifugation. In contrast, GUS (GUS in Fig. 5)and the fusion protein encoded by ASRL (ASRL in Fig. 5)were detected in the first three fractions that include cyto-solic proteins, and no band was detected in microbodies(fractions 13 and 14). These results indicated that the pro-tein encoded by SRL was transported to microbodieswhereas the proteins encoded by GUS and zlSRL were lo-calized in cytosol.

Microbody-targeting efficiencies of carboxy-terminaltripeptides—Based on the density of gold particles ob-served by immunocytochemical analysis, the targetingefficiencies of various tripeptide sequences at the carboxylterminus of the fusion proteins could be grouped into threeclasses (Table 2). The first class (+ + ) , which showed highdensities of gold particles within microbodies, includesSRL, CRL, ARL, SKL and SRM. The results indicate thatthese tripeptide sequences function as efficient targetingsignals to microbodies. SRI and PRL belonged to the sec-ond class ( + ), which showed detectable but weak stainingby immunoelectron microscopy. The density of gold par-ticles within microbodies was less than one tenth of the firstclass ( + + ) , although the similar amounts of the fusionproteins were accumulated. This result suggests that the mi-crobody-targeting activities of the latter tripeptides arelower than those of the first class. The remaining tripeptidesequences that we tested belong to the third class (—). Wecould not detect any signal within microbodies of thesetransgenic plants, although we performed the immunoelec-tron microscopic observations using several sections pre-pared from independent experiments. The result obtainedfrom subcellular fractionation experiments supports theidea that these fusion proteins are not transported into mi-crobody, and remain in cytosol where their concentrationsare too low to be detected by immunoelectron microscopy.

Immunoelectron microscopic analysis of transgenicArabidopsis expressing fusion proteins with various carb-oxyl termini has allowed the role of these terminal se-quences in targeting microbodies to be determined. Analy-sis by electron microscopy indicates that addition of thepentapeptide ELSRL at the carboxyl terminus is sufficientto target /S-glucuronidase into microbodies in the cells ofgreen cotyledons. Since the minimum requirement for aplant microbody-targeting signal has been reported to becarboxy-terminal SKL (Banjoko and Trelease 1995), wewere especially interested in the function of the carboxy-ter-minal tripeptide. Our results strongly suggest that certaincarboxy-terminal tripeptide sequences function as micro-body-targeting signals in plant cells. The results of thisstudy indicate that the consensus amino acid sequence ofthe microbody-targeting signal is [C/A/S/P]-[K/R]-[I/L/M]. The highest microbody-targeting efficiencies were ob-served when the carboxy-terminal tripeptides were [C/A/S]-[K/R]-[L/M], while sequences with proline at the firstposition and isoleucine at the last position function asweak targeting signals.

Most proteins localized in microbodies are known tobe synthesized in a form similar in size to the mature mole-cule, i.e., they are not first synthesized as larger precursormolecules. As summarized in Table 3, the tetrafunctionalenzyme for fatty acid ^-oxidation (Preisig-Muller et al.1994), isocitrate lyase (Beeching and Northcote 1987,Mano et al. 1996, Turley et al. 1990b, Zhang et al. 1993),



Table 2 Targeting efficiencies of carboxy-terminal tripeptides

Substitutions offirst amino acid

Substitutions ofsecond amino acid

Substitutions ofC-terminal amino acid

LRLFRLCRLARLGRLSRLYRLPRLERLKRL

SILSGLSSLSHLSKLSRL

SRISRVSRLSRFSRMSRSSRESRK

Amino acid sequences of carboxy-terminal tripeptides are shown by single letter codes. Targeting efficiencies of proteins containing thesetripeptides at the carboxy terminus were presented by + + (efficient), + (detectable) and — (inefficient).

malate synthase (Comai et al. 1989, Graham et al. 1989,Mori et al. 1991, Rodriguez et al. 1990, Turley et al.1990a), glycolate oxidase (Tsugeki et al. 1993, Volokita andSomerville 1987), hydroxypyruvate reductase (Hayashi etal. 1996b) and uricase (Nguyen et al. 1985) contain carb-oxy-terminal tripeptides that are consistent with the abovemicrobody-targeting consensus sequence. Although wehave not directly tested the targeting efficiencies of the carb-oxy-terminal tripeptides ARM and PRM that are found insome of these enzymes, these tripeptides obey the rule formicrobody-targeting signals denned in this study. Thesedata suggest that all these enzymes are translocated into mi-crobodies by the recognition of these carboxy-terminal

tripeptide sequences. Although our results indicated thatCRL and SRI function as microbody-targeting signals, nomicrobody enzymes have yet been found to contain thesetripeptides at the carboxyl terminus.

Although we have demonstrated that the tripeptide se-quence PRL functions as a microbody-targeting signal notonly for leaf peroxisomes in green cotyledons but also forglyoxysomes in etiolated cotyledons and for leaf peroxi-somes in leaves by using immunofluorescent microscopy(Hayashi et al. 1996a), the targeting efficiency for PRL israther low in Arabidopsis cells. This result is curious be-cause glycolate oxidase containing a carboxy-terminal PRLand the tetrafunctional enzyme containing a carboxy-ter-

Table 3 Targeting signals for various microbody proteins

Protein

CatalaseTetrafunctional enzymeThiolaseIsocitrate lyase

Malate synthase

Citrate synthaseMalate dehydrogenaseGlycolate oxidase

Localization

all kinds of microbodiesglyoxysomeglyoxysomeglyoxysome

glyoxysome

glyoxysomeglyoxysome and leaf-peroxisomeleaf-peroxisome

N-terminalextra peptide

nonoyesno

no

yesyesno

C-terminaltripeptide0

PRM

SRMARMSRLSKL

PRL

Hydroxypyruvate reductaseUricaseAconitase

leaf-peroxisomemicrobody in root nodulecytosol

nonono

ARLSKLSKL

" The carboxy-terminal tripeptide sequences coinciding with the consensus amino acids for microbody-targeting signal are shown. Twokinds of tripeptides are shown in some cases, since these enzymes are known to contain different carboxy-terminal tripeptides in differentspecies.



minal PRM seemed to be efficiently translocated into micro-bodies in pumpkin and cucumber cells, respectively. Onepossibility is that targeting efficiency is species specific, andthat PRL and PRM might function efficiently as micro-body-targeting signals in pumpkin and cucumber but not inArabidopsis.

Among other enzymes localized in microbodies, thio-lase (Kato et al. 1996b, Preisig-Muller and Kindl 1993),which is involved in fatty acid yS-oxidation, and citrate syn-thase (Kato et al. 1995) and malate dehydrogenase (Gietl1990), which are involved in the glyoxylate cycle, areknown to be synthesized as precursor molecules with anextra peptide at their amino terminus (Table 3). We havedemonstrated that these enzymes contain a microbody-targeting signal within their amino-terminal extra peptides(Kato et al. 1996a). In contrast, the targeting signal forplant catalase is still unknown, while aconitase, a memberof the glyoxylate cycle, and which was formerly believed tobe localized in microbodies, has been revealed to be a cyto-solic enzyme (Hayashi et al. 1995).

In the present study, we examined the microbody-targeting efficiency of the tripeptides only for green cotyle-donary cells that contain leaf peroxisomes. We recentlydemonstrated that fusion proteins containing SRL, SRM,SKL, ARL and PRL at their carboxyl termini are importedin vivo into glyoxysomes in etiolated cotyledonary cells aswell as into leaf peroxisomes in mature leaf cells (Hayashiet al. 1996a). All of these tripeptides coincide with the con-sensus sequence of the microbody-targeting signal dennedabove. In addition, results of several studies obtained by invivo and in vitro import systems suggested that there existsconserved machinery for recognition of targeting signalsin glyoxysomes, leaf peroxisomes and unspecialized mi-crobodies (Mori and Nishimura 1989, Olsen et al. 1993,Onyeocha et al. 1993). We are currently trying to identifywhether all consensus tripeptide sequences of the form [C/A/S/P]-[K/R]-[I/L/M] function as targeting signals for alltypes of plant microbodies including glyoxysomes, leaf per-oxisomes in cotyledonary cells, leaf peroxisomes in matureleaves and unspecialized microbodies in roots.

The microbody-targeting signals of mammals, yeastsand trypanosomes have also been analyzed, and the diver-gence in permissible tripeptide sequences among these or-ganisms has been discussed (Purdue and Lazarow 1994).The consensus tripeptide sequence for the mammalian mi-crobody-targeting signal has been reported as [S/A/C]-[K/R/H]-[L] (Gould et al. 1989, Swinkels et al. 1992). Com-parison of this consensus sequences with that for plantsrevealed that there are some features that are specific forplant microbody-targeting signals. A significant characteris-tic that differs from other organisms is that histidine at thesecond position of the carboxy-terminal tripeptide is notrecognized as a microbody-targeting signal in plants, but isin mammals and trypanosomes. In contrast, methionine at

the carboxyl terminus acts as a microbody-targeting signalin plants and trypanosomes, but does not function efficient-ly in mammals (Swinkels et al. 1992). Proline at the firstposition of the carboxy-terminal tripeptide acts as a micro-body-targeting signal in plants and trypanosomes, but pro-line in this position has not been examined experimentallyin mammals and yeast.

Recently, Trelease et al. (1994) demonstrated thatcottonseed isocitrate lyase retains an ability to be import-ed in vivo into microbodies of mammalian cells and discus-sed the evolutional conservation of microbody importmachinery between plants and mammals. Although thismachinery appears to be essentially conserved, our observa-tions suggest that there is some evolutional divergence inthe carboxy-terminal microbody-targeting signal and in themechanism for its recognition among organisms. The recep-tors for carboxy-terminal microbody-targeting signals havebeen characterized in yeast (Terlecky et al. 1995, Vanderleijet al. 1993) and human (Fransen et al. 1995). Further analy-sis of the receptor in higher plants and its comparison withthose from other organisms will provide information onthe evolutional divergence of the mechanisms for the trans-location of proteins into microbodies at molecular level.

The authors thank Dr. John J. Harada (University of Califor-nia, Davis, CA) for technical advice on the generation of trans-genic Arabidopsis. They also thank Dr. Kenzo Nakamura andDr. Ken Matsuoka (Nagoya University, Japan) for providingpMAT037. This work was supported in part by a grant ofResearch for the Future Program from the Japanese Society forthe Promotion of Science (JSPS-RFTF96L00407), and by Grants-in-Aid for Scientific Research from the Ministry of Education,Science and Culture of Japan (nos. 04454017, 06248230,06261239) and by a grant from the Nissan Science Foundation(Tokyo, Japan).

References

Banjoko, A. and Trelease, R.N. (1993) Development and application of anin vivo plant peroxisome import system. Plant Physiol. 107: 1201-1208.

Beeching, J.R. and Northcote, D.H. (1987) Nucleic acid (cDNA) andamino acid sequences of isocitrate lyase from castor bean. Plant. Mol.Biol. 8: 471-475.

Beevers, H. (1979) Microbodies in higher plants. Annu. Rev. PlantPhysiol. 30: 159-193.

Comai, L., Baden, C.S. and Harada, J.J. (1989) Deduced sequence of amalate synthase polypeptide encoded by a subclass of the gene family. / .Biol. Chem. 264: 2778-2782.

De Bellis, L., Tsugeki, R. and Nishimura, M. (1991) Glyoxylate cycle en-zymes in peroxisomes isolated from petals of pumpkin (Cucurbita sp.)during senescence. Plant Cell Physiol. 32: 1227-1235.

Fransen, M., Brees, C , Baumgart, E., Vanhooren, J.C.T., Baes, M., Man-naerts, G.P. and Vanveldhoven, P.P. (1995) Identification and character-ization of the putative human peroxisomal C-terminal targeting signalimport receptor. J. Biol. Chem. 270: 7731-7736.

Gietl, C. (1990) Glyoxysomal malate dehydrogenase from watermelon issynthesized with an amino-terminal transit peptide. Proc. Natl. Acad.Sci. USA 87: 5773-5777.

Gould, S.J., Keller, G.A., Hosken, N., Wilkinson, J. and Subramani, S.(1989) A conserved tripeptide sorts proteins to peroxisomes. J. Cell.Biol. 108: 1657-1664.



Gould, S.J., Keller, G.A. and Subramani, S. (1988) Identification of perox-isomal targeting signals located at the carboxy terminus of four peroxi-somal proteins. J. Cell. Biol. 107: 897-905.

Gould, S.J., Keller, G.-A.A. and Subramani, S. (1987) Identification of aperoxisomal targeting signal at the carboxy terminus of firefly luciferase.J. Cell. Biol. 105: 2923-2931.

Graham, I.A., Smith, L.M., Brown, J.W., Leaver, C.J. and Smith, S.M.(1989) The malate synthase gene of cucumber. Plant Mol. Biol. 13: 673-684.

Hayashi, M., Aoki, M., Kato, A., Kondo, M. and Nishimura, M. (1996a)Transport of chimeric proteins that contain a carboxy-terminal targetingsignal into plant microbodies. Plant J. 10: 225-234.

Hayashi, M., De Bellis, L., Alpi, A. and Nishimura, M. (1995) Cytosolicaconitase participates in the glyoxylate cycle in etiolated pumpkin cotyle-dons. Plant Cell Physiol. 36: 669-680.

Hayashi, M., Tsugeki, R., Kondo, M., Mori, H. and Nishimura, M.(1996b) Pumpkin hydroxypyruvate reductases with and without aputative C-terminal signal for targeting to microbodies may be producedby alternative splicing. Plant Mol. Biol. 30: 183-189.

Huang, A.H.C., Trelease, R.N. and Moore, T.S. (1983) Plant Peroxi-somes. New York: NY: Academic Press.

Kato, A., Hayashi, M., Kondo, M. and Nishimura, M. (1996a) Targetingand processing of a chimeric protein with the amino-terminal prese-quence of the precursor to glyoxysomal citrate synthase. Plant Cell 8:1601-1611.

Kato, A., Hayashi, M., Mori, H. and Nishimura, M. (1995) Molecularcharacterization of a glyoxysomal citrate synthase that is synthesized as aprecursor of higher molecular mass in pumpkin. Plant Mol. Biol. 27:377-390.

Kato, A., Hayashi, M., Takeuchi, Y. and Nishimura, M. (1996b) cDNAcloning and expression of a gene for 3-ketoacyl-CoA thiolase in pump-kin cotyledons. Plant Mol. Biol. 31: 843-852.

Mano, S., Hayashi, M., Kondo, M. and Nishimura, M. (1996) cDNA clon-ing and expression of the isocitrate lyase gene in pumpkin cotyledons.Plant Cell Physiol. 37: 941-948.

Marchuk, D., Drumm, M., Saulino, A. and Collins, F.S. (1990) Construc-tion of T-vectors, a rapid and general system for direct cloning of unmod-ified PCR products. Nucl. Acids Res. 19: 1154.

Matsuoka, K. and Nakamura, K. (1991) Propeptide of a precursor to aplant vacuolar protein required for vacuolar targeting. Proc. Natl.Acad. Sci. USA 88: 834-838.

Mori, H. and Nishimura, M. (1989) Glyoxysomal malate synthase is specifi-cally degraded in microbodies during greening of pumpkin cotyledons.FEBS Lett. 244: 163-166.

Mori, H., Takeda, Y.Y., Hara-Nishimura, I. and Nishimura, M. (1991)Pumpkin malate synthase: cloning and sequencing of the cDNA and Nor-thern blot analysis. Eur. J. Biochem. 197: 331-336.

Nagel, A., Elliot, A., Masel, A., Birch, R.G. and Manners, J.M. (1990)Electroporation of binary Ti plasmid vector into Agrobacterium tumefa-ciens and Agrobacterium rhizogenes. FEMS Microbiol. Lett. 67: 325-328.

Nguyen, T., Zelechowska, M., Foster, V., Bergmann, H. and Verma, D.(1985) Primary structure of the soybean nodulin-35 gene encodinguricase II localized in the peroxisomes of uninfected cell of nodules.Proc. Natl. Acad. Sci. USA 82: 5040-5044.

Nishimura, M., Takeuchi, Y., De Bellis, L. and Hara-Nishimura, I. (1993)Leaf peroxisomes are directly transformed to glyoxysomes during senes-cence of pumpkin cotyledons. Protoplasma 175: 131-137.

Nishimura, M., Yamaguchi, J., Mori, H., Akazawa, T. and Yokota, S.(1986) Immunocytochemical analysis shows that glyoxysomes are direct-ly transformed to leaf peroxisomes during greening of pumpkin cotyle-dons. Plant Physiol. 80: 313-316.

Olsen, L.J., Ettinger, W.F., Damsz, B., Matsudaira, K., Webb, M.A. andHarada, J.J. (1993) Targeting of glyoxysomal proteins to peroxisomes inleaves and roots of a higher plant. Plant Cell 5: 941-952.

Olsen, L.J. and Harada, J.J. (1995) Peroxisomes and their assembly inhigher plants. Annu. Rev. Plant Physiol. Plant Mol. Biol. 46: 123-146.

Onyeocha, I., Behari, R., Hill, D. and Baker, A. (1993) Targeting of castorbean glyoxysomal isocitrate lyase to tobacco leaf peroxisomes. PlantMol. Biol. 22: 385-396.

Preisig-Muller, R., Guhnemann-Schafer, K. and Kindl, H. (1994) Domainsof the tetrafunctional protein acting in glyoxysomal fatty acid beta-oxida-tion. Demonstration of epimerase and isomerase activities on a peptidelacking hydratase activity. / . Biol. Chem. 269: 20475-20481.

Preisig-Muller, R. and Kindl, H. (1993) Thiolase mRNA translated in vitroyields a peptide with a putative N-terminal presequence. Plant Mol. Biol.22: 59-66.

Purdue, P.E. and Lazarow, P.B. (1994) Peroxisomal biogenesis: multiplepathways of protein import. / . Biol. Chem. 269: 30065-30068.

Rodriguez, D., Ginger, R.S., Baker, A. and Northcote, D.H. (1990) Nucle-otide sequence analysis of a cDNA clone encoding malate synthase ofcastor bean (Ricinus communis) reveals homology to DAL7, a gene in-volved in allantoin degradation in Saccharomyces cerevisiae. Plant Mol.Biol. 15: 501-504.

Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: aLaboratory Manual, 2nd ed. Cold Spring Harbor, NY: Cold Spring Har-bor Laboratory.

Swinkels, B.W., Gould, S.J. and Subramani, S. (1992) Targeting efficien-cies of various permutations of the consensus C- terminal tripeptide per-oxisomal targeting signal. FEBS Lett. 305: 133-136.

Terlecky, S.R., Nuttley, W.M., McCollum, D., Sock, E. and Subramani,S. (1995) The Pichia pastoris peroxisomal protein PAS8p is the receptorfor the C-terminal tripeptide peroxisomal targeting signal. EMBO J. 14:3627-3634.

Titus, D.E. and Becker, W.M. (1985) Investigation of the glyoxysome-per-oxisome transition in germinating cucumber cotyledons using double-la-bel immunoelectron microscopy. / . Cell. Biol. 101: 1288-1299.

Trelease, R.N., Choe, S.M. and Jacobs, B.L. (1994) Conservative aminoacid substitutions of the C-terminal tripeptide (Ala-Arg-Met) on cotton-seed isocitrate lyase preserve import in vivo into mammalian cell peroxi-somes. Eur. J. Cell Biol. 65: 269-279.

Tsugeki, R., Hara-Nishimura, I., Mori, H. and Nishimura, M. (1993) Clon-ing and sequencing of cDNA for glycolate oxidase from pumpkin cotyle-dons and Northern blot analysis. Plant Cell Physiol. 34: 51-57.

Turley, R.B., Choe, S.M., Ni, W. and Trelease, R.N. (1990a) Nucleotidesequence of cottonseed malate synthase. Nucl. Acids Res. 18: 3643.

Turley, R.B., Choe, S.M. and Trelease, R.N. (1990b) Characterization ofa cDNA clone encoding the complete amino acid sequence of cotton iso-citrate lyase. Biochim. Biophys. Ada 1049: 223-226.

Valvekens, K., Montagu, M.V. and Lijsebettens, M.V. (1988) Agrobacte-rium tumefaciens-medisittd transformation of Arabidopsis thaliana rootexplants by using kanamycin selection. Proc. Natl. Acad. Sci. USA 85:5536-5540.

Vanderleij, I., Franse, M.M., Elgersma, Y., Distel, B. and Tabak, H.F.(1993) PAS10 is a Tetratricopeptide-Repeat Protein That is Essential forthe Import of Most Matrix Proteins into Peroxisomes of Saccharomycescerevisiae. Proc. Natl. Acad. Sci. USA 90: 11782-11786.

Volokita, M. (1991) The carboxy-terminal end of glycolate oxidase directsa foreign protein into tobacco leaf peroxisomes. Plant J. 1: 361-366.

Volokita, M. and Somerville, C.R. (1987) The primary structure of spinachglycolate oxidase deduced from the DNA sequence of a cDNA clone. / .Biol. Chem. 262: 15825-15828.

Yamaguchi, J. and Nishimura, M. (1984) Purification of glyoxysomalcatalase and immunochemical comparison of glyoxysomal and leaf per-oxisomal catalase in germinating pumpkin cotyledons. Plant Physiol.74: 261-267.

Zhang, J.Z., Gomezpedrozo, M., Baden, C.S. and Harada, J.J. (1993) 2Classes of isocitrate lyase genes are expressed during late embryogenyand postgermination in Brassica napus L. Mol. Gen. Genet. 238: 177-184.

(Received April 14, 1997; Accepted May 8, 1997)


changes in targeting efficiencies of proteins to plant microbodies

Documents