Jul 11, 1989 - For each fraction, 50 Al was removed to determine TCA-precipitable counts. An additional 50 ..... The minced tissue was ground with a Polytron.
The EMBO Journal vol.8 no.11 pp.3195-3206, 1989
The transit peptide of a chloroplast thylakoid membrane protein is functionally equivalent to a stromal-targeting sequence
J.Mark Hand'2'4, Les J.Szabo3'5, Aurea C.Vasconcelos2 and Anthony R.Cashmore' 'Plant Science Institute, Department of Biology, University of Pennsylvania, Philadelphia, PA 19104, 2Department of Biological Sciences, Nelson Biological Laboratories, Rutgers University, Piscatway, NJ 08854 and 3Laboratory of Cell Biology, Rockefeller University, New York, NY 10021, USA 4Present address: American Cyanamid, Agricultural Research Division, P.O. Box 400, Princeton, NJ 08540, USA 5Present address: USDA, ARS, Cereal Rust Lab, University of Minnesota, St. Paul, MN 55108, USA Communicated by G.Silnatz
The role of transit peptides in intraorganellar targeting has been studied for a chlorophyll a/b binding (CAB) polypeptide of photosystem H (PSII) and the small subunit of ribulose-1,5-bisphosphate carboxylase (RBCS) from Pisum sativum (pea). These studies have involved in vitro import of fusion proteins into isolated pea chloroplasts. Fusion of the CAB transit peptide to RBCS mediates import to the stroma, as evidenced by assembly of RBCS with chloroplast-synthesized large subunit (RBCL) to form holoenzyme. Similarly, fusion of the RBCS transit peptide to the mature CAB polypeptide mediates import and results in integration of the processed CAB protein into the thylakoid membrane. Correct integration was indicated by association with PSII and assembly with chlorophyll to form the lightharvesting chlorophyll a/b protein complex (LHCII). We interpret these results as evidence that the CAB transit peptide is functionally equivalent to a stromal-targeting sequence and that intraorganellar sorting of the CAB protein must be determined by sequences residing within the mature protein. Our results and those of others suggest that import and integration of CAB polypeptides into the thylakoid proceeds via the stroma. Key words: thylakoid membrane/CAB polypeptides/stromal targeting/intraorganellular sorting/organellar targeting
Introduction Many chloroplast proteins are synthesized in the cytosol (Cashmore, 1976; Chua and Gillham, 1977) on free ribosomes as higher mol. wt precursors (Dobberstein et al., 1977; Cashmore et al., 1978; Smith and Ellis, 1979). These precursor proteins contain an amino-terminal extension, referred to as the transit peptide, which mediates chloroplast import (Van den Broeck et al., 1985). Post-translational import is most likely initiated via binding of precursors to a proteinaceous receptor (Cline et al., 1985; Pain et al., 1988) followed by translocation across the membranes in
an energy-dependent fashion (Cline et al., 1985; Grossman et al., 1980; Pain and Blobel, 1987; Flugge and Hinz, 1986). During or subsequent to translocation and suborganellar delivery, the precursor is proteolytically processed to the mature form (Robinson and Ellis, 1984a; Kirwin et al., 1987; Lamppa and Abad, 1987) and may assemble with other components synthesized in either the chloroplast or the cytosol. The predominant chloroplast protein is ribulose1,5-bisphosphate carboxylase/oxygenase (RUBISCO), a bifunctional enzyme localized within the stroma. The large subunit (RBCL) is encoded on organellar DNA while the small subunit (RBCS) is a nucleocytoplasmic product imported post-translationally into the chloroplast as a 20 kd precursor (pre-RBCS) (Chua and Schmidt, 1978; Smith and Ellis, 1979). The 6 kd pre-RBCS transit peptide of Pisum sativum L. (pea) is sufficient to import foreign polypeptides both in vitro and in vivo (Van den Broeck et al., 1985; Lubben and Keegstra, 1986). The precursor is proteolytically cleaved in two distinct steps (Robinson and Ellis, 1984b; Mishkind et al., 1985). Whether this occurs prior to, during or subsequent to assembly into the holoenzyme is unknown. The major protein constituents of chloroplast thylakoid membranes are the chlorophyll a/b-binding (CAB) polypeptides associated with the light-harvesting complex (LHCII) of photosystem II (PSII). The protein/lipid complexes are responsible for the harvesting of radiant energy and subsequent transfer of excitation energy to the PSII reaction center. Genes encoding CAB polypeptides have been isolated from various species, including pea (Cashmore, 1984) and numerous in vitro studies have analyzed the import characteristics of the precursors from a variety of species. Cumulative results indicate that imported precursors are processed to the mature form (Schmidt et al., 1981; Cline et al., 1985; Pichersky et al., 1987), assembled into photosystem II (Chitnis et al., 1986; Pichersky et al., 1987) and bind chlorophyll (Schmidt et al., 1981). Several steps of the import pathway of thylakoid membrane proteins are not well understood. For example, it is not known how translocation from the envelope to the thylakoid membrane occurs. It has been proposed by Douce et al. (1984) that import of thylakoid membrane proteins may occur through the stroma, en route to the thylakoids, via vesicles derived from the inner envelope membrane. However, recent evidence (Cline, 1986, 1988; Smeekens et al., 1986; Hageman et al., 1986; Lamppa and Abad, 1987; Lamppa, 1988) suggests translocation of thylakoid or lumen directed proteins may occur through the stroma as soluble precursors. Similarly, it is not known whether the precursor is processed to the mature form before or after thylakoid membrane insertion. Chitnis et al. (1986) demonstrated that pre-CAB can integrate into PSII, although it is possible that translocation of a Lemna gibba protein into barley etiochloroplasts may prevent optimal processing of pre-CAB. Similarly, while pre-CAB may integrate into the
3195
J.M.Hand et al.
A
X'~~~~
_l~~~~~~~~_X
l.~~ . ~ ~ ~ YxX
~~~~~~~~~
Fig. 1. Structure of pre-CAB, pre-RBCS, CAB-SS and SS-CAB. The structure of the various polypeptides are shown and the translation products are displayed by fluorography of SDS-polyacrylamide gels. In CAB-SS, the CAB transit peptide (CAB TP) ends at amino acid +3 within the mature CAB protein, this is followed by the amino acids P-A-G derived from polylinker sequences, followed by the entire mature RBCS sequence. In SSCAB, the entire RBCS transit peptide (RBCS TP) is fused to the CAB coding region, minus amino acid 1, with a D-P-L-E amino acid sequence in the polylinker fusion junction. The mature CAB protein (Mat-CAB) lacks a transit peptide.
thylakoid membrane prior to processing in vitro (Cline, 1986), this may not be the case in vivo. Membrane integration of a pre-CAB protein has been demonstrated in a reconstituted system and requires thylakoid membranes (and not envelope membranes), ATP and a soluble proteinaceous factor residing in the stroma (Cline, 1986, 1988; Fulson and Cline, 1988; Chitnis et al., 1988). This ability of pre-CAB to integrate into the thylakoid membrane in the absence of envelope membranes argues favorably for a translocation model whereby pre-CAB passes through the stroma. An important question concerning RBCS and CAB import
addresses the role that the respective transit peptides play in intraorganellar sorting. If translocation of CAB polypeptides occurs via the stroma then one might expect the CAB transit peptide to mediate import of a soluble protein to the stroma. Recent data addressing this point is somewhat equivocal. A CAB transit peptide was shown to mediate transport of neomycin phosphotransferase II to a soluble fraction of the chloroplast (Van den Broeck et al., 1988). It seems likely, and was concluded, that this corresponded to translocation to the stroma. However, evidence excluding localization between the membranes of the chloroplast envelope was not provided. In contrast, a CAB transit peptide containing four or 16 amino-terminal amino acids from the
3196
mature CAB protein did not mediate chloroplast import when fused to 3-glucuronidase (Kavanagh et al., 1988). In related experiments it was recently demonstrated that translocation of a CAB polypeptide to the thylakoid can be mediated by an RBCS transit peptide (Lamppa, 1988). Whether, in this case, the CAB polypeptide integrated correctly and assembled to form an LHCII complex, was not examined. In experiments reported here we rigorously explore the functional relatedness of the RBCS and CAB transit peptides. We demonstrate that the CAB transit peptide has the potential to translocate the RBCS polypeptide to the stroma, as evidenced by assembly of RBCS into holoenzyme. Furthermore, we show that the RBCS transit peptide can translocate the CAB polypeptide to the thylakoid, where the latter integrates and assembles into an LHCII complex. We conclude that the two transit peptides are functionally
equivalent.
Results Construction of plasmids for chloroplast import studies A DNA template encoding pre-CAB (Cashmore, 1984) was inserted downstream from the SP6 promoter within the in vitro transcription plasmid, pSP65 (Figure 1). This construct
CAB transit peptide equivalence to stromal targeting sequence
c.
1i
-.
'Al-
;.:..".
'.
.. X.
...
--.;W A".4-,
f.:
Pin-1
.:, :-3e.
Fig. 2. Import of pre-CAB and CAB. Import assays were performed and chloroplasts reisolated as detailed in Materials and methods. Mg-ATP and thermolysin were utilized at 10 mM and 100 lg/mnl, respectively. Apyrase (1 U/pA) was utilized at a final concentration of 0.02 of import mix. Radioactive bands were cut out of 12.5% polyacrylamide gels previously subjected to fluorography and counted directly in units/1l the presence of ACSII, a xylene/2-ethoxyethanol based scintillant. T = total chloroplast lysate; S = soluble and I = insoluble.
contains a non-translated leader sequence of 63 nucleotides and was utilized in the analysis of import of a PSH CAB protein. A plasmid utilizing the methionine codon 37 of the CAB transit peptide as an initiating codon was constructed for the production of a 'mature' CAB polypeptide lacking the transit peptide. DNA sequences encoding the transit peptide from preCAB fused to the mature RBCS (CAB-SS) and the transit peptide from pre-RBCS fused to the mature CAB (SS-CAB) were ligated downstream from the SP6 promoter, in the modified in vitro transcription plasmid pGEM4 (Figure 1). A prime concern in the construction of both chimeric fusion constructs was to include the putative processing region of the respective transit peptide, while deleting the homologous region contributed from the mature coding region. To insure that the processing region would be included, cab-ss was constructed such that the entire transit peptide of pre-CAB was encoded in addition to the initial four amino acids of the mature coding region. The polylinker region contributes a P-A-G amino acid sequence, which is followed by amino acid 1 (Met) of mature RBCS (Figure 1). The addition of four amino acids from the pre-CAB mature protein as well as the lack of the Cys-Met cleavage site for pre-RBCS, was expected to result in processing proceeding in a similar manner to that found for pre-CAB.
For construction of a plasmid encoding SS-CAB, a sequence encoding the transit peptide of RBCS was fused to a sequence encoding the mature coding region of preCAB. The fusion junction of the polypeptide consists of four amino acids (D-P-L-E) contributed by the oligonucleotides inserted from pGSSTneol (Van den Broeck et al., 1985) and the polylinker region of the plasmid encoding the mature CAB polypeptide (Figure 1). The presence of the putative RBCS processing enzyme binding site and the Cys -Met
cleavage site was expected to ensure processing properties similar to those described for pre-RBCS (Mishkind et al., 1985). In vitro chloroplast import of pre-CAB, pre-RBCS and the fusion proteins We first determined whether the in vitro synthesized polypeptides were translocated across the chloroplast
envelope membrane, as indicated by protection against subsequent protease treatment. Furthermore, we determined
whether the precursor was proteolytically processed to the mature form. During a 60 min import assay, pre-CAB is imported and processed to the mature form (Figure 2, lane 1). A significant amount of precursor remains bound to the envelope membrane and is sensitive to treatment with thermolysin (Figure 2, compare lane 1 with lane 2). 3197
,.1
J.M.Hand et al.
i.
*i
ml
i
1
. ..
C
,~ ~
4M BE 5 0Nl.j1 4 ,t oi tU ..i ij'L L J 4 15 r
BT NO
..........
AL)t0'"I
-iLi I:iP
-.1 2...6
i Z 1 r v- |~~~~
K;
0
F~~~ -
I
Fig. 4. Import and assembly of pre-RBCS. Reisolated chloroplasts from a large-scale import and assembly assay of pre-RBCS were separated into soluble and insoluble components. The soluble fraction was subjected to ultracentrifugation at 115 000 g for 16 h. Fractions of - 800 y1 were collected, of which 50 IA was added to 20 itl of concentrated SDS-PAGE gel loading buffer, heated at 90°C for 5 min and electrophoresed through a 12.5% SDS-polyacrylamide gel. A Coomassie blue-stained SDS-PAGE of respective fractions is shown (a) and an autoradiograph of the gel is shown in (b).
1A 1 1 l-o, _ Si
a
0
AS5
Fig. 3. Time course and fractionation of import assays. Individual small-scale import assays were performed for pre-CAB, CAB-SS and SS-CAB. Subsequent to termination of and treatment with the protease thermolysin, intact chloroplasts were reisolated through 40% Percoll cushions. The chloroplast pellet was washed, repelleted and resuspended in ddH2O and separated into soluble (S) and insoluble (I) fractions as described in Materials and methods. Each sample was analyzed as described in Figure 2.
However, the mature form is protected from protease treatment, demonstrating inclusion within the chloroplast. Further analysis reveals that pre-CAB is localized within the membrane fraction (insoluble fraction; see Figure 2, lane 3) and no detectable amount of precursor or the mature protein is localized within the stroma (soluble fraction; see Figure 2, lane 4). Import of pre-CAB was optimal in the presence of light (Figure 2, lane 5). A requirement for ATP is demonstrated by an almost complete inhibition of import when assays are incubated in the dark (Figure 2, lane 7). Inclusion of 10 mM Mg-ATP in the assay mix during incubation in the dark results in a significant increase toward control levels (Figure 2, lane 6, compare to lane 5). The residual amount of import recorded during dark import assays minus exogenous Mg-ATP is most likely due to ATP present in either isolated chloroplasts or the wheat germ translation mix. Pre-incubation of [ S]methionine-labeled polypeptides with apyrase (an adenosine 5'-triphosphatase and adenosine 5'-diphosphatase) results in a complete inhibition of import
3198
assays conducted in the dark (Figure 2, lane 9). However, over a 60 min import assay, light-incubated samples pretreated with apyrase show slightly lower levels than control assays (Figure 2, lane 8). Import assays conducted in the presence of apyrase heated to 95°C before incubation with wheat germ translation products demonstrated control levels of import (Figure 2, lane 10). In import experiments with the mature CAB protein a substantial amount of the CAB protein binds to the chloroplast outer envelope membrane but is sensitive to treatment with the protease thermolysin (compare Figure 2, lane 11 to lanes 12 and 13). Therefore, no detectable amount of mature CAB is imported and translocated to either the thylakoid membrane or stroma of the chloroplast (Figure 2, lanes 12-13). Both of the fusion proteins, CAB-SS and SS-CAB, were imported into chloroplasts, processed to a lower mol. wt form and protected from treatment of intact chloroplasts with protease (data not shown). The time course of suborganellar localization of pre-CAB and the fusion proteins was determined (Figure 3). For pre-CAB, a steady increase in the amount of precursor imported, processed and localized within the thylakoid membrane occurs during the initial 20 min of the import assay (Figure 3a). CAB-SS import increases linearly toward a plateau level at -40-60 min (Figure 3b). The precursor is processed to a smaller mol. wt form and is localized within the soluble fraction (i.e. either the stroma or intraenvelope space). After the initial 10-20 min interval, import and processing of CAB-SS is at a similar maximal level to that measured for pre-CAB. In contrast, even at saturation levels of CAB-SS import (after -40-60 min) the level of detectable radioactivity in the soluble fraction is CAB-SS > pre-CAB).
. ^- I
CAB transit peptide equivalence to stromal targeting sequence
a
.
"N- 4j
/-
I.
0
pre-RBC
-
x
a)
S-:2
L1
L2
.-
v
....
3
4
F ACTION NUMBFR
F.
7
x-
83
11
1.
4
15
10'
(
*,r
60'
......
Fig. 5. Time course analysis of import and assembly of pre-RBCS. Soluble components from large-scale import assays of pre-RBCS were fractionated and collected as described in Figure 4. For each fraction, 50 Al was removed to determine TCA-precipitable counts. An additional 50 Id was removed, mixed with 20 yl of concentrated SDS-PAGE loading buffer, heated at 90°C for 5 min and electrophoresed through a 12.5% SDS-polyacrylamide gel. Each time course assay is represented via graphic analysis as well as a fluorograph. In vitro import and assembly of pre-RBCS and CABSS into the holoenzyme of RUBISCO In order to determine the exact suborganellar targeting of each precursor, we studied their respective assembly characteristics. In the case of pre-RBCS, we first confirmed earlier in vitro import data (Chua and Schmidt, 1978; Smith and Ellis, 1979) that RBCS assembles into the RUBISCO holoenzyme. Figure 4a illustrates that the majority of soluble proteins sediment near the top of a 0.1-0.7 M sucrose gradient. However, the RUBISCO holoenzyme sediments within the middle fractions and upon dissociation with SDS and heating, the RBCL and RBCS subunits migrate independently in SDS-PAGE. The [35S]methionine imported RBCS protein is identical in size to that of the native RBCS protein (Figure 4b). Therefore, pre-RBCS is imported, processed and assembled into the stromal holoenzyme. Import of RBCS reaches a plateau level at 10 min (data not shown). Figure 5 illustrates that this level of import correlates with RUBISCO assembly. Separate import assays of pre-RBCS were terminated at 10 and 60 min and the soluble fraction was sedimented through sucrose gradients and fractionated as shown in Figure 4. Identical aliquots from each fraction were removed for determination of trichloroacetic acid (TCA)-precipitable counts and analysis via 12.5% SDS -PAGE/fluorography. We next determined the import characteristics of the fusion protein CAB-SS. Similarly to pre-RBCS, this protein was
proteolytically processed and targeted to a soluble fraction of the chloroplast. We used the chloroplast import/RUBISCO assembly assay to determine if CAB-SS is assembled into the RUBISCO holoenzyme. Figure 6 illustrates that peak radioactive fractions of [35S]methionine-labeled processed CAB-SS sediment with the RUBISCO holoenzyme as depicted by both TCA-precipitable counts and SDS -PAGE fluorographs. By SDS -PAGE the processed form of CABSS appears to be slightly larger (- 1-3 amino acids) than the mature form of RBCS (data not shown). From these results it is evident that the CAB transit peptide can substitute for the transit peptide of RBCS, albeit in a less efficient manner. That is, the CAB transit peptide will deliver the 'mature' form of RBCS to its final destination (the chloroplast stroma) in the final form (the assembled holoenzyme of RUBISCO). In vitro import and thylakoid integration of SS-CAB The demonstration that the CAB transit peptide translocates a stromal protein to its correct suborganellar location prompted us to ask if a transit peptide known to be processed within the stroma (i.e. the RBCS transit peptide) could substitute for the CAB transit peptide in mediating translocation to the thylakoid membrane. The data presented in Figure 3 demonstrate that this is the case; import experiments with SS-CAB show that it is inserted into the thylakoid membrane in a processed form. The maximum level of imported SS-CAB occurs within the initial 10 min of the
3199
J.M.Hand et al.
4Xi p&B".-:;b--m - ;i ;,
:*.4%' ,
--.:--:!.rl4- !i
-A
;.
-.1
..I
11 i
Ow I x .., i.. ;.?W,w. :.,.. SO&
-,' ., l..-I"i
j..
.-x-.
.
9
>: :
....
_
'.
e
..;
.-i
...'.;
>';w.s
i
-
s ;
"
w
I
1,
--
RI.!'i-S:7
2-,
=.:'-:
P.,
:'
-; ;5>--
,
&4 %
s
Fig. 6. Time course analysis of import and assembly of CAB-SS. Soluble components from large-scale import assays and collected as described in Figure 4. Analysis of each fraction was as described in Figure 5.
assay (similar to pre-RBCS), while the amount of precursor translocated, processed and associated with the insoluble fraction was consistently greater than the amount imported for pre-CAB over a series of import assays. While the rate of import of SS-CAB was greater than pre-CAB, the total amount of precursor imported during a 60 min assay was usually less than CAB-SS and always less than pre-RBCS (pre-CAB < SS-CAB ' CAB-SS < pre-RBCS).
In vitro import and association of pre-CAB and SSCAB with the photosystem 11 complex of the chloroplast thylakoid membrane The transit peptide for a known stromal protein (RBCS) was shown to direct mature CAB to a membrane fraction, presumed to be the thylakoids (Figure 3c). We now asked whether the CAB polypeptides derived from both pre-CAB and SS-CAB were associated with the PSII complex subsequent to intraorganellar sorting. Triton-X solubilized thylakoids were sedimented through a 0.1/0.7 M sucrose gradient in 0.02% Triton-X100 with a 2 M sucrose cushion. The LHCII CAB/PSH complex sediments near the top of the gradient while Photosystem I (PSI) and its associated LHCI CAB proteins sediment at the 0.7/2 M sucrose interface. Figure 7 illustrates the results for pre-CAB
(Figure 7a) and SS-CAB (Figure 7b). We repeatedly observed -50% more SS-CAB associated with PSII than we observed for pre-CAB. In Figure 7 the fluorograph for each construct is located under the respective graphical analysis of TCA-precipitable counts and total chlorophyll.
3200
of CAB-SS were fractionated
In each case, -90-95% of TCA-precipitable counts are located within the same fractions corresponding to the PSII peak. The remaining counts were in the peak chlorophyll fraction at the 0.7/2 M sucrose interface (the PSI complex). The [35S]methionine-labeled mature polypeptide sedimenting with the PSI peak is likely to be due to crosscontamination of these chlorophyll-containing complexes as it also occurs with pre-CAB import and PSI complexes recovered and concentrated contain small amounts of the major LHCH CAB polypeptide, as demonstrated by SDS -PAGE Coomassie blue staining (data not shown). Further analysis revealed that radiolabeled CAB polypeptides derived from both imported pre-CAB and SSCAB were selectively precipitated with the LHCII/PS11 complex (Burke et al., 1978). The LHCH precipitates derived from import of pre-CAB and SS-CAB were analyzed by SDS -PAGE followed by Coomassie blue staining (Figure 8; Panel A) and fluorography (Figure 8; Panel B). The radiolabeled proteins from peak radioactive fractions are selectively precipitated along with native LHCII proteins. The processed form of SS-CAB possesses a slightly lower mobility in 12.5% SDS -PAGE than the native CAB polypeptide (Figure 8, Panel B). This lower mobility indicates that processing most likely occurs at the Cys -Met cleavage site in the RBCS transit peptide via the stromal localized protease. Processing at this site would result in an SS-CAB 'mature' form which would be four amino acids larger than the proposed mature form of CAB (Mullet, 1983). The mature form of SS-CAB is not accompanied by
CAB transit peptide equivalence to stromal targeting sequence a.
!?00 0
0 ir
80
1-
cn
3,UC4J
0 0
