Plant Physiol. (1997) 115: 881-889

Diverse Amino Acid Residues Function within the Type 1 Peroxisoma! Targeting Signal’

lmplications for the Role of Accessory Residues Upstream of the Type 1 Peroxisomal Targeting Signal

Robert T. Mullen, Michael S. Lee, C. Robb Flynn, and Richard N. Trelease*

Department of Plant Biology (R.T.M., R.N.T.), and Graduate Program in Molecular and Cellular Biology (M.S.L., C.R.F.), Arizona State University, Tempe, Arizona 85287-1 601

The purpose of this study was to determine whether the plant type 1 peroxisomal targeting signal (PTS1) utilizes amino acid res- idues that do not strictly adhere to the serine-lysine-leucine (SKL) motif (small-basic-hydrophobic residues). Selected residues were appended to the C terminus of chloramphenicol acetyltransferase (CAT) and were tested for their ability to target CAT fusion proteins to glyoxysomes in tobacco (Nicotiana tabacum L.) cv Bright Yellow 2 suspension-cultured cells. CAT was redirected from the cytosol into glyoxysomes by a wide range of residues, i.e. A/C/G/S/T-H/K/ L/N/R-I/L/M/Y. Although 1 and N at the -2 position (-SLL, -ANL) do not conform to the SKL motif, both functioned, but in a temporally less-efficient manner. Other SKL divergent residues, however, did not target CAT to glyoxysomes, i.e. F or P at the -3 position (-FKL, -PKL), S or T at the -2 position (49, STL), or D at the -1 position (-SKD). The targeting inefficiency of CAT-ANL could be ameliorated when K was included at the -4 position (-KANL). In summary, the plant PTSl mostly conforms to the SKL motif. For those PTSls that possess nonconforming residue(s), other residues upstream of the PTSl appear to function as accessory sequences that enhance the temporal efficiency of peroxisomal targeting.

One hallmark of eukaryotic cells is the compartmental- ization of specialized metabolic pathways into separate subcellular organelles. Such regulation is strictly depen- dent upon the proper sorting of protein constituents into each organelle. This process is mediated by the targeting action of organelle-specific topogenic signals that reside within concise amino acid sequences in the protein (Baker et al., 1996). Peroxisomal matrix-destined proteins, for ex- ample, are directed into the organelle by at least two types of targeting signals, the type 1 and the type 2 PTS. The

This research was supported by National Science Foundation (NSF) grant no. MCB-9305395 to R.N.T. and in part by the William N. and Myriam Pennington Foundation. Research assistantships were funded by the Molecular and Cell Biology Program at Ari- zona State University for M.S.L and C.R.F., by the Achievement Rewards for College Scientists Foundation Fellowship for M.S.L., by the William N. and Myriam Pennington Foundation for M.S.L., and by the NSF Graduate Research Training Program grant no. GER-9553456 for C.R.F.

* Corresponding author; e-mail trelease.dick@asu.edu; fax 1- 602-965-6899.

PTSl is a noncleaved tripeptide (SKL) motif, i.e. small- basic-hydrophobic residues, or a variant thereof, which resides at the extreme C terminus of the majority of per- oxisomal matrix proteins (de Hoop and AB, 1992; Subra- mani, 1993; Purdue and Lazarow, 1994; Olsen and Harada, 1995; Baker, 1996; Gietl, 1996).

The PTS2 is a nonapeptide (R-L/I/ Q-X,-H/Q-L) found within the N-terminal presequences of a smaller subset of peroxisomal matrix-destined proteins that are proteolyti- cally processed (de Hoop and AB, 1992; Gietl, 1996). Ex- amples of these include 3-keto-acyl thiolase in rat (Osumi et al., 1991; Swinkels et al., 1991) and Sacckaromyces cereui- siae (Erdmann, 1994; Glover et al., 1994), amine oxidase in Hansenula polymorpha (Faber et al., 1995), malate dehydro- genase in watermelon (Gietl et al., 1994), and citrate syn- thase in pumpkin (Kato et al., 1996). Another possible matrix protein PTS, which is located internally, has been described for catalase A and carnitine acetyltransferase in S. cereuisiae (Kragler et al., 1993; Elgersma et al., 1995) and acyl-COA oxidase in Candida tropicalis (Small et al., 1988).

The PTSl was first identified by Subramani and co- workers when the C-terminal tripeptide of firefly luciferase (-SKL) was demonstrated to be both necessary and suffi- cient for peroxisomal targeting in cultured mammalian cells (Gould et al., 1987, 1989). Subsequent experimental studies revealed that the SKL motif directed targeting to the peroxisomal matrix in evolutionary diverse organisms such as yeast (Aitchison et al., 1992; Diste1 et al., 1992; Veenhuis, 1992), plants (Volokita, 1991; Olsen et al., 1993; Banjoko and Trelease, 1995; Hayashi et al., 1996; Trelease et al., 1996a; Lee et al., 1997), and trypanosomes (Fung and Clayton, 1991; Sommer et al., 1992).

Other studies with various organisms, however, strongly suggest that the PTSl possesses a greater diversity of func- tional residues than those defined by the SKL motif. For example, in the trypanosomatid protozoa, the acceptable divergency of the SKL tripeptide for targeting proteins to glycosomes (peroxisome-like organelles) is far more diver- gent compared with mammals (Blattner et al., 1992). A

Abbreviations: BY-2, tobacco cv Bright Yellow 2 suspension- cultured cells; CAT, chloramphenicol acetyltransferase; IL, isoci- trate lyase; MS, malate synthase; PTS, peroxisomal targeting signal.

881

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

882 Mullen et al. Plant Physiol. Vol. 11 5, 1 9 9 7

number of amino acid replacements, including S, N, D, and Y (all nonbasic) at the -2 position of the C-terminal tripep- tide of phosphoglycerate kinase were functional for glyco- soma1 targeting (Blattner et al., 1992). In yeast -AKI on C. tropicalis trifunctional enzyme (Aitchison et al., 1991) and -NKL on H. polymorpka dihydroxyacetone synthase (Han- sen et al., 1992) reflect divergency because I at the -1 position is nonfunctional in mammals (Aitchison et al., 1992), and N at the -3 position is large, not small. In S. cerevisiae, Elgersma et al. (1996) demonstrated that a wide range of SKL variants including E, F, G, and P (large) at the -3 position and A, F, Y, and S (nonbasic) at the -2 position in the C-terminal tripeptide of malate dehydrogenase were functional for peroxisomal import.

Greater divergence from the SKL motif has been demon- strated recently for mammalian proteins. Motley et al. (1995) reported that -KKL, -SQL, -NKL, or -SSL at the C terminus of human Alaglyoxylate aminotransferase all were functional for import into human fibroblast peroxi- somes. Purdue and Lazarow (1996) demonstrated that N (nonbasic) at the -2 position of the C-terminal tripeptide of human liver catalase was functional for import into human fibroblast and yeast peroxisomes.

Severa1 examples of divergence within the SKL motif have been reported for plants. Trelease et al. (1996b) dem- onstrated that the C-terminal tripeptide of rat liver catalase with a nonbasic N at the -2 position was sufficient to re-route CAT from the cytosol to tobacco (Nicotiana taba- cum) BY-2 glyoxysomes. Hayashi et al. (1996) showed that P (large) at the -3 position of a 10-amino acid fragment (-PRL) appended to GUS was sufficient to target this GUS construct to glyoxysomes and leaf peroxisomes in trans- genic Arabidopsis. Hayashi et al. (1996) concluded that this observation was significant because the C-terminal tripep- tide of glycolate oxidase from pumpkin is -PRL and there- fore this tripeptide likely represents a functional variant of the SKL motif in plants. More recently, Mullen et al. (1997) reported SKL divergency by the observation that the C- terminal tripeptide of cottonseed catalase with P at the -3 position and S (nonbasic) at the -2 position was the PTSl for this constitutive matrix enzyme.

The purpose of this study was to ascertain the range of amino acid residues, both conserved and divergent from the SKL motif, that were functional within the PTSl in plants. To that end, we generated various CAT constructs that coded for permutations of a SKL tripeptide appended to the C terminus of bacterial passenger protein CAT. The subcellular localization of these transiently expressed CAT fusion proteins was analyzed using tobacco BY-2 suspen- sion cells serving as an in vivo glyoxysome import system (Banjoko and Trelease, 1995; Trelease et al., 1996a).

Experimental results revealed that a wide range of amino acid residues, most of which conform to the SKL motif, were functional within the PTSl in BY-2 cells. This range is considerably greater than those reported for the targeting of proteins to mammalian and certain yeast peroxisomes, but perhaps not for trypanosomes. On the other hand, certain SKL-nonconforming residues were more efficient within the PTSl only when in the proper context conveyed

by upstream amino acid residues within the C terminus of certain native peroxisomal proteins.

MATERIALS A N D METHODS

Plasmid Constructions

AI1 DNA manipulations were performed using standard procedures (Sambrook et al., 1986). Restriction enzymes and other DNA-modifying enzymes were purchased from Promega or New England Biolabs and were used as rec- ommended by the suppliers. Custom synthetic oligonucle- otides were synthesized at the Arizona State University (ASU) Bioresources Facility (Tempe, AZ).

pRTL2 / CAT, which is a plant expression vector that contains the entire CAT open reading frame, was created as described previously (Trelease et al., 199610). CAT chimeric constructs were generated via PCR-based site-directed mu- tagenesis with essentially the same reaction conditions as those described by Trelease et al. (1996b). The various PCR primers used in this study are listed in Table I. PCR mix- tures included a forward primer corresponding to an 18-bp region upstream of a unique NcoI site in CAT and a reverse primer that introduced sequences in the 3’ untranslated region of CAT encoded for a specific appended polypep- tide, a stop codon, and an XbaI site. The resulting PCR products were ethanol-precipitated overnight, resus- pended in water, and then digested with NcoI and XbaI and ligated into NcoI-XbaI-digested pRTL2 / CAT. Construction of pRTLZ/CAT-ANL (Trelease et al., 199613) and pRTL2/ CAT-SKL (Lee et al., 1997) has been described elsewhere.

pRTLZ/IL-ANL was created by modifying the DNA se- quences encoding the C-terminal -RM of cIL to -NL via PCR-based mutagenesis. The PCR reaction included a for- ward primer upstream of unique NheI site within cotton- seed IL (Lee et al., 1997) and a reverse mutagenic primer

AGCC-3’) that introduced the C-terminal substitutions (i.e. -RM to -NL) as well as an XbaI site in the 3’ untranslated region of cIL. The resulting PCR fragment was gel purified, ligated into pCR 2.1 (TA cloning vector; Invitrogen, San Diego, CA), then subcloned into NkeI-XbaI-digested pRTL2/IL (Lee et al., 1997).

AI1 DNA constructs were verified by sequencing at the ASU Bioresources Facility employing TaqCycle auto- mated sequencing with DyeDeoxy terminators (Applied Biosystems).

(5‘-CTGACGTCTAGAGCTTGTCTTAGATGCTAGGCTT-

Cell Culture and Microprojectile Bombardment

Tobacco (Nicotiana tabacum L. cv Bright Yellow 2) cell- suspension cultures were grown in darkness at 25°C as described previously (Banjoko and Trelease, 1995). BY-2 cells were harvested 4 d after subculture by centrifugation at 400g, resuspended in an equal volume of 1 X transfor- mation buffer ( lx growth media without 2,4-D, and with 250 mM sorbitol and 250 mM mannitol) (Banjoko and Tre- lease, 1995), and spread on filter papers premoistened with 1 X transformation buffer. After 1 h of equilibration, BY-2 cells were transiently transformed by microprojectile bom-

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

Characterization of the Type 1 Peroxisomal Targeting Signal 883

bardment with 10 pg of plasmid DNA and a Biolistic Particle Delivery System according to the manufacturer’s recommendations (Bio-Rad). Bombarded BY-2 cells were then left in unwrapped Petri dishes for 5 h, 20 to 22 h, or 40 to 45 h to allow transient gene expression and import into glyoxysomes.

lmmunofluorescence Microscopy

Transiently transformed BY-2 cells were fixed in 4% (v/ v) formaldehyde in 1 X growth media minus 2,4-D (Ban- joko and Trelease, 1995) for 1 h at room temperature. Cells were then washed several times with l x PBS (4.3 mM Na,HPO,, 1.4 mM KH,PO,, 2.7 mM KC1, and 137 mM NaC1, pH 7.4), and incubated with 0.1% (w/v) pectolyase Y-23 (Seishin Pharmaceutical Co., Tokyo, Japan) in I X PBS for 2 h at 30°C. After washings in l x PBS, plasma, and or- ganellar membranes were permeabilized by incubating cells in 0.3% (v/v) Triton X-100 (Sigma) for 15 min at room temperature (Lee et al., 1997).

For experiments designed to demonstrate the actual im- port of transiently expressed proteins into the glyoxysomal matrix, BY-2 cells were harvested in 1 X growth medium and then fixed in 2% (v /v) formaldehyde in 50 mM KHJ’O,, pH 7.2, as described previously (Lee et al., 1997). Following pectolase treatment, cells were differentially permeabilized with digitonin (25 pg / mL) (Sigma) rather than with Triton X-100 for 15 min at room temperature to selectively permeablilize the plasma membrane (Lee et al., 1997).

Portions of pectolyase-treated, detergent-permeabilized, and PBS-washed cells were processed for immunofluores- cence microscopy. Applications of primary and fluorescent dye-conjugated secondary antibodies were performed as de- scribed by Trelease et al. (1996a). After each incubation cells were washed with l x PBS. Antibody sources and concen- trations used were as follows: mouse anti-CAT monoclonal antibody (undiluted hybridoma medium; gift from S. Sub- ramani, San Diego, CA), rabbit anti-cottonseed catalase IgGs

(prepared with protein A affinity columns; Kunce et al., 1988) at 1:500 (v/v), rabbit anti-cottonseed IL IgGs (prepared with protein A affinity columns; Lee et al., 1997) at 1:2000 (v/ v), mouse anti-a-tubulin monoclonal antibody at 1:500 (v / v) (Amersham), goat anti-rabbit rhodamine at 1:lOOO (v / v) (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA), goat anti-mouse BODIPY 1 : l O O O (v /v) (MO- lecular Probes Inc., Eugene, OR), and goat-anti mouse Cy3 at 1:2000 (v /v ) (Jackson ImmunoResearch Laborato- ries, Inc.).

Immunostained cells were mounted on glass slides in 90% (v/v) glycerol with n-propyl gallate (Sigma) to pre- vent bleaching of fluorescence and viewed using a fluores- cent microscope (Axiovert 100, Zeiss). Photographs were taken with T-Max 400 ASA black-and-white print film (Kodak). Images shown in the figures are representative of data obtained from viewing several thousands of BY-2 cells (1-2% transient transformation) under 40-mm coverslips from at least two separate experiments.

RESULTS

lmport of CAT-SKL into BY-2 Clyoxysomes

In our initial studies related to in-vivo targeting of fusion proteins to peroxisomes in BY-2 cells, we biolistically in- troduced either pCAMVCN, which contained sequences encoding CAT, or pCAMVC-SKL-N, which contained se- quences encoding -SKL at the C terminus of CAT (Banjoko and Trelease, 1995). In a11 of our subsequent studies, we biolistically introduced CAT constructs within the pRTL2 vector (e.g. Trelease et al., 1996a, 1996b). The same applies to the current study, i.e. wild-type CAT and modified versions of CAT with appended C-terminal polypeptides were transiently expressed from pRTL2 in BY-2 cells. As- sessments of the functionality and / or temporal efficiency of the appended amino acid residues were made relative to results obtained with transiently expressed CAT possess- ing the prototypical -SKL tripeptide (pRTL2 / CAT-SKL).

Table 1. Oligonucleotides used for PCR-based generation of CA J fusion construcb Each of the synthetic oligonucleotides listed introduced substitutions within the 3’ untranslated

region of CAT that encoded for a specific C-terminal polypeptide, a stop codon, and an Xbal site via PCR-based mutarienesis.

Mutant Oligonucleotide Sequences (5’ to 3’)

pRTL21CAT-CKL pRTL2fCAT-GKL pRTL21CAT-FKL pRTLZ/CAT-PKL pRTL21CAT-TKL pRLTZ/CAT-SHL pRTL21CAT-SLL pRTL21CAT-S R L pRTL21CAT-STL pRLT21CAT-SKD pRTL21CAT-SKF pRTL21CAT-SKI pRLT21CAT-SKY pRLT21CAT-SSI DRLT~ICAT-KANL

CGCCATCTAGAATAACTGCCTTAAAGTTTACACGCCCCGCCC CGCCATCTAGAATAACTGCCTTACAATTTTCCCGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAAGTTTGAACGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAAATTTTGGCGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAAGTTTAGTCGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAAGATGTGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAAGAAGTGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTACAAACGAGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTACAAAGTAGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTAGTCTTTTGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAAATTTTGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAATTTTTGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTAATATTTTGACGCCCCGCCC CGCCATCTAGAATAACTGCCTTAAATAGAAGACGCCCCGCCC CGCCATCTAGAATAACTTCATAATTAAGCTTTCGCCCCGCCC

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

884 Mullen et al. Plant Physiol. Vol. 115, 1997

Figure 1, A and B, includes representative photomicro-graphs illustrating the subcellular localizations of CAT andCAT-SKL, respectively, in biolistically transformed BY-2cells. During the 20-h postbombardment period allowedfor expression and sorting of these introduced CAT fusionproteins, CAT accumulated and remained throughout thecytosol in transiently transformed cells (Fig. 1A). In con-trast, CAT-SKL was redirected during the same period oftime from the cytosol to organelles, presumably to glyoxy-somes (the punctate pattern shown in Fig. IB). Compellingevidence for glyoxysomes being the targeted site of CAT-SKL was the co-localization of CAT-SKL with endogenousglyoxysomal catalase (Banjoko and Trelease, 1995). In con-trol experiments in which BY-2 cells were transformedtransiently with the gene encoding cottonseed IL (pRTL2/IL), immunostaining was not observed following applica-tion of anti-CAT monoclonal antibodies (data not shown).

It is important to note, however, that the immunofluo-rescence co-localization of transiently expressed CAT-SKLwith endogenous catalase does not necessarily demon-strate actual import into the glyoxysomal matrix. Rather,

Figure 1. Immunofluorescence localization of CAT-SKL in BY-2glyoxysomes. Twenty to 22 h following biolistic bombardment, BY-2cells were fixed in formaldehyde, treated with pectolyase Y-23, incu-bated either in 0.3% (v/v) Triton X-100 or 25 jAg/rnL digitonin todifferentially permeabilize the plasma and glyoxysomal boundarymembranes, then incubated in primary and secondary antibodies. A,BODIPY immunofluorescence attributable to staining of CAT through-out the cytosol of a transiently transformed cell treated with TritonX-100 and incubated in mouse anti-CAT and anti-mouse BODIPY-conjugated IgGs. B, BODIPY immunofluorescence attributable tostaining of CAT-SKL in glyoxysomes in cells fixed in formaldehyde,treated with Triton X-100, and incubated in mouse anti-CAT IgGs. C,Representative lack of BODIPY immunofluorescence in the samebatch of cells as B but treated with digitonin and incubated in mouseanti-CAT IgGs. D, Cy3 immunofluorescence of a-tubulin in the cytosolof the same digitonin-treated cells shown in C. Bar = 10 urn.

observed punctate immunofluorescence patterns such asthose shown in Figure IB could simply reflect the bindingof the CAT-SKL to the cytosolic face of the glyoxysomalboundary membrane. Recently, Lee et al. (1997) demon-strated by differential permeabilization of the plasma andglyoxysomal boundary membranes in BY-2 cells that intro-duced cottonseed IL was targeted and imported into theglyoxysomes. Cells were incubated in 0.3% (v/v) TritonX-100 to permeabilize both the plasma membrane and theperoxisomal boundary membrane, or in 25 /xg/mL digito-nin to permeabilize only the plasma membrane. Punctateimmunofluorescence of introduced IL correlated with flu-orescence imaging of matrix-localized catalase, rather thanwith the imaging of cytosolic tubulin, indicating that ILwas contained within the glyoxysomes following translo-cation through the boundary membrane (Lee et al., 1997).

To ascertain whether CAT-SKL had actually been im-ported into the glyoxysomal interior, transformed cells weretreated with digitonin as described by Lee et al. (1997).When the same batch of BY-2 cells shown in Figure IB wasselectively permeabilized with digitonin, CAT-SKL was notimmunostained (Fig. 1C), whereas cytosolic tubulin wasreadily immunostained (Fig. ID). These results indicate thatCAT-SKL was not adhered to the outside of the glyoxy-somes, but was imported into the organelles. The samedifferential permeabilization results were obtained for allother CAT constructs examined in this study whether theywere targeted efficiently or inefficiently to glyoxysomes.Figures illustrating these results are not presented.

Conservation of the SKL Motif

We demonstrated previously that the C-terminal tripep-tides of rat-liver catalase (-ANL) and cottonseed catalase(-PSI) were necessary for targeting these enzymes to BY-2glyoxysomes (Trelease et al., 1996b; Mullen et al., 1997).These observations were somewhat surprising becauseboth tripeptides were conspicuously divergent from theconsensus SKL motif (de Hoop and AB, 1992; Subramani,1993); i.e. P at the -3 position was large not small, and S andN at the -2 position were nonbasic residues. It became ofinterest, therefore, to determine the range of amino acidresidues that could function within the PTS1. Table II sum-marizes the results of these experiments. Figure 2 showsrepresentative immunofluorescence photomicrographs il-lustrating the subcellular localizations of selected CAT con-structs listed in Table II.

Figure 2A shows a punctate BODIPY immunofluores-cence pattern attributable to the transient expression of aCAT fusion protein containing a C, a small residue at the -3position, i.e. CAT-CKL. When the same field of cells shownin Figure 2A, which were double-labeled with anti-cottonseed catalase IgGs, was viewed with the rhodaminefilter, an identical punctate immunofluorescence patternattributable to endogenous catalase was observed in thetransformed cell and in all neighboring cells (results notshown). Therefore, -CKL, which conforms to the SKL motif,was sufficient for redirecting CAT from the cytosol toglyoxysomes.

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

Characterization of the Type 1 Peroxisomal Targeting Signal 885

Table II. Subcellular localizations of CAT fusion constructs 20 hpostbombardment

Twenty hours postbiolistic bombardment, BY-2 cells were fixed informaldehyde, treated with pectoylase Y-23 to partially digest thecell walls, permeabilized with 0.3% (v/v) Triton X-100, and incu-bated in mouse anti-CAT and anti-mouse BODIPY-conjugated IgGs.Results are given for immunofluorescence microscopic analyses ofthe subcellular localizations of transiently expressed CAT fusionproteins with permutations of amino acid residues either at the -3position (-XKL), the -2 position (-SXL), the -1 position (-SKX), or otherpositions within the C-terminal tripeptide. Targeting efficiency ofC-terminal tripeptides are shown either as + (efficient, glyoxysomallocalization), ± (inefficient, glyoxysomal and cytosolic localization),or - (not targeted, cytosolic localization).

-XKL

-CKL +-GKL +-TKL +-FKL --PKL -

-SXL

-SHL +-SRL +-SLL ±-STL -

-SKX

-SKI +-SKF +-SKY +-SKD -

Other Positions

-ANL ±-SSI -

Similar results were observed when other small aminoacid residues such as G (-GKL) or T (-TKL) were substi-tuted for S at the -3 position (Table II). However, targetingof fusion proteins to glyoxysomes was abolished when alarge amino acid residue such as F (Fig. 2B) or P (Fig. 2C)replaced S at the -3 position. Figure 2, B and C, shows thatCAT-FKL and CAT-PKL, respectively, accumulatedthroughout the cytosol of transformed cells.

To test the requirement of a basic amino acid residue atthe -2 position, K was substituted either with a conservedor a divergent residue (Table II). As shown in Figure 2D,the transient expression of CAT-SHL resulted in a punctateBODIPY immunofluorescence pattern in transformed BY-2cells. CAT-SHL immunofluorescence co-localized with en-dogenous catalase staining (results not shown), indicatingthat basic H was functional at the -2 position for glyoxy-somal targeting. Similarly, substitution of K with basic R(-SRL) also preserved targeting of CAT to glyoxysomes(not illustrated; Table II). However, if K was substitutedwith a nonbasic amino acid residue such as L or T, target-ing of CAT to glyoxysomes either was markedly reducedor abolished, respectively. For example, Figure 2E showsthat at 20 h postbombardment, L at the -2 position (CAT-SLL) results in punctate and cytosolic immunofluorescencewithin the same cell. These repeatable data indicate thatthis fusion protein is not completely targeted from thecytosol to glyoxysomes within the 20-h period. Targetingalso was inefficient when nonbasic N was at the -2 positionand A was at the -3 position (CAT-ANL) (Table II). How-ever, Figure 2F shows that when nonbasic T was substi-tuted at the -2 position (CAT-STL), glyoxysomal targetingwas not apparent during the 20-h expression period. After40 h some CAT-STL was targeted to glyoxysomes (resultsnot shown).

The requirement of a hydrophobic amino acid residue atthe -1 position was also tested. Figure 2G shows that sub-stitution of L at the -1 position with hydrophilic D (CAT-SKD) abolishes glyoxysomal targeting. However, substitu-tions at the -1 position with various hydrophobic amino

acid residues preserved glyoxysomal targeting (CAT-SKI,Fig. 2H; CAT-SKF, Fig. 21; CAT-SKY, Table II). On the otherhand, when I was at the -1 position with a nonbasic S at the-2 position (CAT-SSI), glyoxysomal targeting was not evi-dent (Table II).

Temporal Targeting Efficiency of CAT Constructs

Because CAT-ANL and CAT-SLL were inefficiently tar-geted to BY-2 glyoxysomes (Table II), we investigated thekinetics of expression and sorting of these and other fusionproteins following biolistic bombardment. Figure 3 is agroup of representative immunofluorescence photomicro-graphs illustrating the temporal efficiency of glyoxysomaltargeting of CAT-SKL, CAT-ANL, CAT-SLL, and CAT-SSI5 h, 20 to 22 h, and 40 to 45 h after bombardment.

Figure 3A shows that transient expression for 5 h ofCAT-SKL results in cytosolic and punctate immunofluores-cence in the same cell, indicating that some of the CAT-SKLis targeted to BY-2 glyoxysomes during this period. WhenCAT-SKL was expressed transiently for 20 to 22 h (Fig. 3B)or 40 to 45 h (Fig. 3C), only punctate immunofluorescencewas observed, indicating that CAT-SKL was completelytargeted to glyoxysomes during these periods. The lack ofany detectable free cytosolic fluorescence after 20 h indi-cates that a reverse equilibrium of CAT-SKL targeting/import was not occurring. These data (Fig. 3, B and C) are

Figure 2. Representative immunofluorescence photomicrographs ofthe subcellular localizations of selected CAT fusion proteins. Twentyto 22 h following biolistic bombardment, BY-2 cells were fixed informaldehyde, treated with pectolyase Y-23, permeabilized (mem-branes) with 0.3% (v/v) Triton X-100, and incubated in mouse anti-CAT and anti-mouse BODIPY-conjugated IgCs. Fusion proteins shownare selected permutations of amino acid residues within the C terminusof CAT-SKL: at the -3 position, A, CAT-CKL; B, CAT-FKL; C, CAT-PKL;at the -2 position, D, CAT-SHL; E, CAT-SLL; F, CAT-STL; and at the 1position, G, CAT-SKD; H, CAT-SKI; I, CAT-SKF. Bar = 10 (j.m.

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

886 Mullen et al. Plant Physiol. Vol. 115, 1997

Figure 3. Representative immunofluorescence photomicrographs il-lustrating the temporal targeting efficiencies of selected CAT fusionproteins in BY-2 cells. Transiently transformed cells were held onsterile Petri plates for 5 h, 20 to 22 h, or 40 to 45 h following biolisticbombardment, then fixed in formaldehyde, treated with pectolyaseY-23 and 0.3% (v/v) Triton X-100, and incubated in mouse anti-CATand anti-mouse BODIPY-conjugated IgCs. Immunofluorescence local-izations shown are: CAT-SKL at 5 h (A), 22 h (B), or 42 h (C); CAT-ANLat 5 h (D), 22 h (E), or 45 h (F); CAT-SLL at 5 h (G), 21 h (H), or 45 h(I); CAT-SSI at 5 h (]), 20 h (K), or 45 h (L). Bar = 10 /LUTI.

consistent with our results for co-localization of CAT-SKLwith endogenous catalase presented in Figure 1, B and C.

CAT-ANL transiently expressed for 5 h remained in thecytosol (Fig. 3D), whereas transient expression for 20 to22 h was sufficient time for import of some of the CAT-ANL as illustrated in Figure 3E. At 40 to 45 h CAT-ANLstill was not completely localized within glyoxysomes, i.e.cytosolic staining of CAT-ANL often was observed (Fig.3F). Comparison of the temporal targeting efficiency ofCAT-ANL with CAT-SLL reveals similar results. At 5 hpostbombardment, essentially all of the detectable CAT-SLL was in the cytosol (Fig. 3G), whereas after 20 to 22 h(Fig. 3H) or 40 to 45 h (Fig. 31), fluorescence in transformedcells was both cytosolic and glyoxysomal (Fig. 3H).

We examined the temporal targeting efficiency of a CATfusion protein, CAT-SSI, which possesses a putative non-functional PTS1 (Table II), to determine whether the pro-longed transient expression (i.e. 40-45 h) of introducedCAT fusion proteins artificially resulted in glyoxysomallocalization. Figure 3, J through L, shows that accumulated

CAT-SSI remained in the cytosol for the entire 40- to 45-hpostbombardment period. Similar results were obtainedwhen CAT without any appended amino acids was tran-siently expressed for 40 to 45 h (results not shown).

Role of K at the -4 Position in Glyoxysomal Targeting

Results shown in Figure 4 demonstrate that the function-ality of a PTS1 can be affected by the context conveyed byat least one upstream amino acid residue. CAT-ANL waspartially targeted to glyoxysomes after 21 h (Fig. 4A), butwhen -KANL was appended to CAT (CAT-KANL), essen-tially all of this fusion protein was redirected from thecytosol to glyoxysomes after 20 to 22 h (Fig. 4B). Theseobservations suggest that the addition of K at the -4 posi-tion enhanced the efficiency of an -ANL to target CAT toBY-2 glyoxysomes. To test the role of K already at the -4position on a peroxisomal protein, we substituted the C-terminal -RM of cottonseed IL with -NL, resulting in IL-KANL because cottonseed IL possesses a KA at the -4 and-3 positions, respectively (Turley et al., 1990). Figure 4Cshows that 20 h postbombardment, IL-KANL accumulatesin the cytosol and glyoxysomes within the same cell. After40 h IL-KANL still was only partially targeted to glyoxy-somes, which is similar to the results with CAT-ANL.However, transiently expressed wild-type IL (IL-KARM)was completely targeted to glyoxysomes during the 20-hexpression period (results not shown), indicating that K atthe -4 position in IL was not sufficient to compensate forthe inefficient targeting activity of -ANL.

DISCUSSION

Cultured BY-2 suspension cells constitute a model in-vivo system for elucidating the targeting signals responsi-ble for sorting peroxisomal-destined proteins (Banjoko andTrelease, 1995; Trelease et al., 1996b). In this study immu-nofluorescence microscopic analyses of the subcellular lo-calizations of various transiently expressed CAT fusionproteins allowed us to characterize the PTS1 in plants. Awide range of amino acid residues, i.e. A/C/G/S/T-H/K/L / N / R-I / L/ M/ Y, within appended C-terminal tripeptides

Figure 4. Comparison of the immunofluorescence localizations ofCAT-ANL, CAT-KANL, and IL-ANL 21 h following biolistic bombard-ment. A, Representative immunfluorescence image depicting por-tions of CAT-ANL in the cytosol and imported into glyoxysomes of atransiently transformed cell. B, Representative immunofluorescenceimage illustrating that essentially all of the CAT-KANL imported intoglyoxysomes of a transiently transformed cell. C, Representativeimmunofluorescence image showing that portions of cottonseed IL-ANL in the cytosol and imported into glyoxysomes of a transientlytransformed cell. Bar = 10 jam.

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

Characterization of the Type 1 Peroxisomal Targeting Signal 887

were functional for redirecting CAT from the cytosol to BY-2 glyoxysomes. The majority of these residues conform to the SKL motif, i.e. small-basic-hydrophobic residues, and seem to be within PTSls that are the most efficient for targeting.

Mutational analyses of amino acid residues examined with in vivo (Gould et al., 1989; Swinkels et al., 1992) and in vitro (Miura et,,al., 1992) mammalian systems revealed that conserved substitutions conforming to the SKL motif preserved peroxisomal targeting. For instance, replace- ments of small S with A or C at the -3 position, basic K with R or H at the -2, or hydrophobic L with M at the -1 position permitted targeting of transiently expressed firefly lucif- erase or CAT fusion proteins into peroxisomes of cultured mammalian cells. Each of the amino acid residues that were shown to function in mammalian cells was found in our study to be sufficient for targeting CAT fusion proteins to BY-2 glyoxysomes (Table 11). Although we did not di- rectly test M at the -1 position, results from other studies with plant cells indicated that C-terminal polypeptides containing M, e.g. -ARM (Lee et al., 1997), -KSRM (Olsen et al., 1993; Trelease et al., 1996a), and -SRM (Hayashi et al., 1996), were sufficient for targeting CAT (or GUS) to per- oxisomes (glyoxysomes).

Severa1 amino acid residues other than those shown to function in mammalian cells were found to be sufficient for redirecting CAT from the cytosol to BY-2 glyoxysomes. For example, conserved substitutions of small S with G or T at the -3 position or hydrophobic L at the -1 position with either I, F, or Y a11 preserved glyoxysomal targeting (Table 11). These latter results contrast with findings for peroxiso- mal targeting in mammalian cells where I or F (Y not tested) were not functional at the -1 position (Subramani, 1993). However, I, F, and Y at the -1 position were func- tional for targeting to glycosomes and peroxisomes in try- panosomes (Blattner et al., 1992; Sommer et al., 1992) and in yeast (Aitchison et al., 1991), respectively. These appar- ently conflicting observations for residues that are func- tional within the PTSl of various organisms have been explained previously by species-specific divergency (Pur- due and Lazarow, 1994).

The nontargeting of CAT-PKL to glyoxysomes provides additional evidence that the plant PTSl mostly conforms to the SKL motif (Fig. 2C). This result is consistent with our previous results in which the C-terminal -PSI of cottonseed, catalase (large P at the -3 position and nonbasic S at the -2 position) was not sufficient for directing CAT to BY-2 glyoxysomes (Mullen et al., 1997). Substitution of large res- idues at the -3 position were not examined in mammalian peroxisomal import studies. In other studies of plants and yeast, however, P at the -3 position was capable of directing peroxisomal targeting. Hayashi et al. (1996) demonstrated that a substitution of S with P (-PRL) on a 10-amino acid C-terminal fragment identical to residues in pumpkin MS appended to GUS was sufficient for targeting this fusion protein to glyoxysomes and leaf-type peroxisomes in trans- genic Arabidopsis. Elgersma et al. (1996) reported that a substitution of S at the -3 position with P (-PKL) or F (-FKL) in epitope-tagged S. cerevisiae malate dehydrogenase still permitted targeting in vivo to peroxisomes. This latter ob-

servation contrasts with our findings that -FKL could not direct CAT to BY-2 glyoxysomes (Fig. 28).

These apparent discrepancies seem to be due not to spe- cies diversity, but to a consequence of the context conveyed by residues upstream of the PTS1. Elgersma et al. (1996) attributed the inability of some PTSls with divergent SKL residues to direct a reporter protein to peroxisomes to be due to the lack of a homologous context that included ”ac- cessory sequences.” As an example, they mentioned the insufficiency of -SSL (found on glycosomal phosphoglyceral kinase) to target various reporter proteins to peroxisomes (Gould et al., 1989; Blattner et al., 1992). Substitution of -SSL for the PTSl on native A1a:glyoxylate aminotransferase or malate dehydrogenase, however, resulted in these proteins being directed to mammalian (Motley et al., 1995) and S. cerevisiae (Elgersma et al., 1996) peroxisomes, respectively.

Elgersma et al. (1996) concluded that the ability of -SSL to function as a PTSl only at the end of a native peroxiso- mal protein likely was due to the role of ”accessory se- quences” within the C terminus of the protein. Conse- quently, the authors proposed that recognition of peroxisomal-destined proteins bearing a divergent PTSl could be mediated via interactions between the PTSl re- ceptor (PEX5; Diste1 et al., 1996) and other ”accessory se- quences” aside from the PTS1. Elgersma et al. (1996) noted that the SKL motif likely constitutes the most favorable PTSl signal because the majority of peroxisomal proteins possess the C-terminal consensus sequence (de Hoop and AB, 1992; Subramani, 1993). Therefore, for those peroxiso- mal proteins that do not possess an “ideal” PTS1, evolu- tionary selection pressure for efficient peroxisomal target- ing may have worked on nearby upstream residues to function as accessory amino acids (Elgersma et al., 1996).

Evidence for this model comes from targeting studies of human and cottonseed catalase in which -ANL and -PSI, respectively, were insufficient for redirecting CAT from the cytosol to peroxisomes (Purdue and Lazarow, 1996; Mullen et al., 1997). However, both catalase tripeptides were suf- ficient for targeting CAT to peroxisomes (glyoxysomes) when a specific additional residue was included at the -4 position in an appended polypeptide, i.e. -KANL and -RPSI. These findings, as pointed out by the authors, sug- gest that the adjacent residue ”compensates” for residues within the PTSl that do not conform to the SKL motif. This interpretation also could explain our results for the inabil- ity of -PKL or -FKL to direct CAT to BY-2 glyoxysomes (Fig. 2, B and C). That is, P appeared to be nonfunctional within the tripeptide appended to CAT, but was functional when in the proper context conveyed by residues within a native peroxisomal-destined protein. For example, CAT was targeted to BY-2 glyoxysomes when R was included at the -4 position of an appended polypeptide (-RPSI) (Mullen et al., 1997), and GUS was targeted to Arabidopsis peroxi- somes when -PRL was at the end of a 10-amino-acid frag- ment identical to the C terminus of pumpkin MS (Hayashi et al., 1996).

Other examples for the importance of putative accessory residues in peroxisomal targeting have been reported and may provide additional evidence for our interpretations of the role of K at the -4 position in CAT-KANL. Mullen et al.

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

888 Mullen et al. Plant Physiol. Vol. 11 5, 1997

(1997) demonstrated that mutations in residues within the C terminus of cottonseed catalase abolished glyoxysomal targeting in BY-2 cells even when the PTSl was left intact. Wolins and Donaldson (1997) studied the interactions be- tween an acyl-CoA oxidase synthetic peptide and putative PTSl receptors integrated into boundary membranes iso- lated from castor bean glyoxysomes. They found that the PTSl sequence (-SKL) of the acyl-COA oxidase peptide bound to a low-affinity site and that adjacent amino acid residues within the peptide bound to a high-affinity site. These authors suggested that the high-affinity site consti- tutes the initial binding interaction event between the PTSl receptor and the peroxisomal matrix-destined protein, whereas binding to the low-affinity site provides the selec- tion for translocation through the boundary membrane via the import machinery. It is conceivable that in our study interactions with a putative high-affinity site were not involved when CAT was used to analyze targeting results. Nevertheless, our results are meaningful, given that a ma- jority of the CAT fusion proteins tested were imported into glyoxysomes.

CAT-ANL (Fig. 3E) was inefficiently targeted to BY-2 glyoxysomes 20 h after biolistic bombardment compared with the targeting of CAT-SKL (Fig. 3B). However, CAT- ANL targeting inefficiency was ameliorated when K was included at the -4 position (CAT-KANL; Fig. 48). Based on the interpretations published by severa1 groups (Elgersma et al., 1996; Purdue and Lazarow, 1996; Mullen et al., 1997; Wolins and Donaldson, 1997), it is possible that addition of K at the -4 position allowed for a more efficient interaction between the divergent -ANL and the putative PTSl recep- tor in BY-2 cells. The complete targeting of CAT-SKL to glyoxysomes during the 20-h period (Fig. 3B) presumably did not require additional accessory residues. These inter- pretations may also explain why IL-KANL was ineffi- ciently targeted to glyoxysomes after 20 h (Fig. 4C). Cot- tonseed IL is an essential enzyme of the glyoxylate cycle and recently has been demonstrated to be targeted to BY-2 glyoxysomes by its C-terminal SKL-conforming -ARM (Lee et al., 1997). It is possible that the cottonseed IL does not possess accessory residues required for efficient glyoxyso- mal targeting. Consequently, accessory residues would not be present within the IL-KANL construct to compensate for a PTSl possessing a less-efficient N at the -2 position.

Puzzling questions related to these interpretations are as follows. Why was IL-KANL less efficiently targeted to BY-2 glyoxysomes than a CAT-fusion protein containing the similar C-terminal tetrapeptide (CAT-KANL)? Similarly, why was CAT-SSI not directed to BY-2 glyoxysomes (Fig. 3, J-L) even though it was similar to CAT-ANL, i.e. a nonbasic residue at the -2 position? One possible explanation of these results is that residues within CAT, but not IL, inher- ently resemble accessory residues involved in peroxisomal targeting. Furthermore, these ”cryptic” accessory residues may be capable of only compensating for certain divergent residues within the PTSl, i.e. N but not S at the -2 position. Wolins and Donaldson (1997) identified a cluster of basic residues upstream of the PTSl in the acyl-COA oxidase fragment that were essential for binding to the putative PTSl receptor in isolated membranes. However, a compar-

ison of the sequences within the C termini of cottonseed IL, CAT, and cottonseed catalase did not reveal a similar basic cluster of residues.

In summary, this and other related targeting studies clearly indicate that residues involved in the PTSl pathway may not be restricted to the C-terminal tripeptide. Accu- mulating evidence suggests that accessory residues within peroxisomal-destined proteins that do not possess a con- served SKL motif play an essential role during interaction with the PTSl receptor. Nevertheless, our results predict that a characterization of a putative plant peroxisomal- destined protein based on a C-terminal tripeptide com- posed of small-basic-hydrophobic residues is applicable.

ACKNOWLEDCMENT

We thank Houston Thompson for maintaining the plant cell cultures.

Received April 17, 1997; accepted July 15, 1997. Copyright Clearance Center: 0032-0889/97/115/0881/09.

LITERATURE ClTED

Aitchison JD, Murrary WW, Rachubinski R (1991) The carboxy- terminal tripeptide ala-lys-ile is essential for targeting Candida tropicalis trifunctional enzyme to yeast peroxisomes. J Biol Chem

Aitchison JD, Szilard RK, Nuttley WM, Rachubinski R (1992) Antibodies directed against a yeast carboxyl-terminal peroxiso- mal targeting signal specifically recognize peroxisomal proteins from various yeasts. Yeast 8: 721-734

Baker A (1996) Biogenesis of plant peroxisomes. In M Smallwood, JP Knox, DJ Bowles, eds, Membranes: Specialized Functions in Plants. BioScientific, Oxford, UK, pp 421440

Baker A, Kaplan CP, Pool MR (1996) Protein targeting and trans- location; a comparative survey. Biol Rev 71: 637-702

Banjoko A, Trelease RN (1995) Development and application of an in vivo plant peroxisome import system. Plant Physiol 107: 1201-1208

Blattner J, Swinkels B, Dorsam H, Prospero T, Subramani S, Clayton C (1992) Glycosome assembly in trypanosomes: varia- tions in the acceptable degeneracy of a COOH-terminal micro- body targeting signal. J Cell Biol 119: 1129-1136

de Hoop MJ, AB G (1992) Import of proteins into peroxisomes and other microbodies. J Biochem 286: 657-669

Distel B, Erdmann R, Gould SJ, Blobel G, Crane DI, Cregg JM, Dodt G, Fujiki Y, Goodman JM, Just WW, and others (1996) A unified nomenclature for peroxisome biogenesis factors. J Cell

. Biol 135: 1-3 Distel B, Gould SJ, Voorn Brouwer T, van der Berg M, Tabak HF,

Subramani S (1992) The carboxy-terminal tripeptide serine- lysine-leucine of firefly luciferase is necessary but not sufficient for peroxisomal import in yeast. New Biol 4: 157-165

Elgersma Y, Van Roermund CWT, Wanders RJA, Tabak HF (1995) Peroxisomal and mitochondrial carnitine acetyltrans- ferase of Sacchuromyces cereuisiae are encoded by a single gene.

Elgersma Y, Vos A, van den Berg M, van Roermund CWT, van der Sluijs P, Distel B, Tabak HF (1996) Analysis of the carboxyl- terminal peroxisomal targeting signal1 in a homologous context in Saccharomyces cereuisiae. J Biol Chem 271: 26375-26382

Erdman R (1994) The peroxisomal targeting signal of 3-oxoacyl- COA thiolase from Saccharomyces cerevisiae. Yeast 10: 935-944

Faber KN, Keizer-Gunnink I, Pluim D, Harder W, Ab G, Veen- huis M (1995) The N-terminus of amino oxidase of Hansenula polymorpha contains a peroxisomal targeting signal. FEBS Lett

266: 23197-23203

EMBO J 14: 3472-3479

357: 115-120

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022

Characterization of the Type 1 Peroxisomal Targeting Signal 889

Fung K, Clayton C (1991) Recognition of a peroxisomal tripeptide entry signal by the glycosomes of Trypanosoma brucei. Mo1 Bio- chem Parasito1 45: 261-264

Gietl C (1996) Protein targeting and import into plant peroxi- somes. Physiol Plant 97: 599-608

Gietl C, Faber KN, Van der Klei IJ, Veenhuis M (1994) Muta- tional analysis of the N-terminal topogenic signal of watermelon glyoxysomal malate dehydrogenase using heterologous host Hansenula polymorpha. Proc Natl Acad Sci USA 91: 3151-3155

Glover JR, Andrews DW, Subramani S , Rachubinski RA (1994) Mutagenesis of the amino targeting signal of Saccaromyces cer- evisiae 3-ketoacyl-COA thiolase reveals conserved amino acids required for import into peroxisomes in vivo. J Biol Chem 269:

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

Gould SJ, Keller GA, Subramani S (1987) Identification of a peroxisomal targeting signal at the carboxy terminus of firefly luciferase. J Cell Biol 105: 2923-2931

Hansen H, Didion T, Thiemann A, Veenhuis M, Roggenkamp R (1992) Targeting sequences of the two major peroxisomal pro- teins in the methylotrophic yeast Hansenula polymorpha. Mo1 Gen Genet 235: 269-278

Hayashi M, Aoki M, .Kato A, Kondo M, Nishimura M (1996) Transport of chimeric proteins that contain a carboxy-terminal targeting signal into plant microbodies. Plant J 1 0 225-234

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

Kragler F, Langeder A, Raupachova J, Binder M, Hartig A (1993) Two independent peroxisomal targeting signals in catalase A in Snccharomyces cerevisiae. J Cell Biol 120: 665-673

Kunce CM, Trelease RN, Turley RB (1988) Purification and bio- synthesis of cottonseed (Gossypium hirsutum L.) catalase. Bio- chem J 251: 147-155

Lee MS, Mullen RT, Trelease RN (1997) Oilseed isocitrate lyases lacking their essential type 1 peroxisomal targeting signal are piggybacked to glyoxysomes. Plant Cell 9: 185-197

Miura S , Kasuy-Arai I, Mori H, Miyazawa S, Osumi T, Hashi- moto T, Fujiki Y (1992) Carboxyl-terminal consensus Ser-Lys- Leu-related tripeptide of peroxisomal proteins functions in vitro as a minimal peroxisome-targeting signal. J Biol Chem 267

Motley A, Lumb MJ, Oatey PB, Jennings PR, De Zoysa PA, Wanders RJ, Tabak HF, Danpure CJ (1995) Mammalian alanine: glyoxylate aminotransferase 1 is imported into peroxisomes via the PTSl translocation pathway: increased degeneracy and con- text specificity of the mammalian PTSl motif and implications for the peroxisome-to-mitochondrion mistargeting of AGT in primary hyperoxaluria type 1. J Cell Biol 131: 95-109

Mullen RT, Lee MS, Trelease RN (1997) Identification of the peroxisomal targeting signal for cottonseed catalase. Plant J 12:

7558-7563

14405-14411

313-322

Olsen LJ, Ettinger WF, Dansz B, Matsudaira K, Webb MA, Harada JJ (1993) Targeting of glyoxysomal proteins to peroxi- somes in leaves and roots of a higher plant. Plant Cell5: 941-952

Olsen LJ, Harada JJ (1995) Peroxisomes and their assembly in higher plants. Annu Rev Plant Physiol Plant Mo1 Biol46: 123-146

Osumi T, Tsukamoto T, Hata S , Yokota S , Fujiki Y, Miyazawa S , Hashimoto T (1991) Amino-terminal presequence of the precur- sor of peroxisomal 3-ketoacyl-COA thiolase is a cleavable signal peptide for peroxisomal targeting. Biochem Biophys Res Com- mun 181: 947-954

Purdue PE, Lazarow PB (1994) Peroxisomal biogenesis: multiple pathways of protein import. J Biol Chem 269: 30065-30068

Purdue PR, Lazarow PB (1996) Targeting of human catalase to peroxisomes is dependent upon a novel COOH-terminal perox- isomal targeting sequence. J Cell Biol 134 849-862

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory Manual, Ed 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY

Small GM, Szabo LJ, Lazarow PB (1988) Acyl-COA oxidase con- tains two targeting sequences each of which can mediate protein import into peroxisomes. EMBO J 7: 1167-1173

Sommer JM, Cheng Q-L, Keller G-A, Wang CC (1992) In vivo import of firefly luciferase into the glycosomes of Trypanosoma brucei and the mutational analysis of the C-terminal targeting signal. Mo1 Biol Cell 3: 749-759

Subramani S (1993) Protein import into peroxisomes and biogen- esis of the organelle. Annu Rev Cell Biol 9: 445478

Swinkels BW, Gould SJ, Bodnar AG, Rachubinski RA, Subra- mani S (1991) A novel, cleavable peroxisomal targeting signal at the amino-terminus of the rat 3-ketoacyl-COA thiolase. EMBO J

Swinkels BW, Gould SJ, Subramani S (1992) Targeting efficien- cies of various permutations of the consensus C-terminal trip- eptide peroxisomal targeting signal. FEBS Lett 305: 133-136

Trelease RN, Lee MS, Banjoko A, Bunkelmann J (1996a) C- terminal polypeptides are necessary and sufficient for in vivo targeting of transiently-expressed proteins to peroxisomes in suspension-cultured plant cells. Protoplasma 195: 156-167

Trelease RN, Xie W, Lee MS, Mullen RT (199613) Rat liver catalase is sorted to peroxisomes by its C-terminal tripeptide Ala-Asn-Leu, not by the interna1 Ser-Lys-Leu motif. Eur J Cell Biol 71: 248-258

Turley RB, Choe SM, Trelease RN (1990) Characterization of a cDNA clone encoding the complete amino acid sequence of cottonseed isocitrate lyase. Biochem Biophys Acta 1049: 223-226

Veenhuis M (1992) Peroxisome biogenesis and function in Nan- senula polymorpha. Cell Biochem Funct 1 0 175-184

Volikita M (1991) The carboxy-terminal end of glycolate oxidase directs a foreign protein into tobacco leaf peroxisomes. Plant J 1:

Wolins NE, Donaldson RP (1997) Binding of the peroxisomal targeting sequence SKL is specified by a low-affinity site in castor bean glyoxysomal membranes. A domain next to the SKL binds to a high-affinity site. Plant Physiol 113: 943-949

1 0 3255-3262

361-366

Dow

nloaded from https://academ

ic.oup.com/plphys/article/115/3/881/6071271 by guest on 16 February 2022