Uncoupling of Co-translational Translocation from Signal Peptidase

Document Sample
Uncoupling of Co-translational Translocation from Signal Peptidase Powered By Docstoc
					THEJOURNALOF BIOLOGICAL      CHEMISTRY                                                             Vol. 264, No. 8 Issue of March 15, pp. 4642-464’7,1989
0 1989 by The American Societyfor Biochemistry and Molecular Biology, Inc.                                                              Prmted in U. A.

Uncoupling of Co-translational Translocation fromSignal Peptidase
Processing in a Mutant Rat Preapolipoprotein-A-IVwith a Deletion
That Includesthe COON-terminal Regionof Its Signal Peptide*
                                                                                                (Received for publication, November 9, 1988)

                Steven F. NothwehrSG, Rodney J. FolzSll, a n d J e f f r e y Gordon$II**.
                From the Departments of $Biological Chemistry and 1) Medicine, Washington University School of Medicine,
                St. Louis, Missouri 631 10

   I n o r d e r to characterize the function of the COOH- (SRP) recognizes the signal peptide as it emerges from the
terminal (c) region of eukaryotic signalpeptides, a 14- ribosome (2), and directs the nascent protein-ribosome com-
amino acid long segmentwas deleted from a secreted plex to theER. SRP subsequently binds to an integral mem-
rat liver and intestinal protein, preapo1ipoprotein-A- brane protein-docking protein (3, 4), and releases the signal
IV. This deletion spanned the c-region all potential peptide which then binds to an integral membrane glycopro-
signal peptidase cleavage sites. The functional conse- tein of the ER known as the signal sequence receptor (5).
quences of this mutation were assessed using a n in Signal peptidase, located onthe luminal side of theER
vitro transcription/translation/microsomal                 membrane membrane (6), removes the signal peptide co-incident with
processing system. Removal of these residues had n o co-translational translocation intothe ERlumen.
effect on interaction of the nascent preprotein with
signal recognition particle as measured by a transla-                    Much work has focused on defining structuralfeatures
tional arrest assay. Although no              signal peptidase cleav- which are importantfor proper signal peptide function. While
age of the mutant was detected, the efficiencyco- comparisons of the primary structures of naturally occurring
                                                             of its
translational translocation          was similar to the wild type eukaryotic signal peptides reveal scant primary sequence iden-
protein. A postinitiation translocation assay was uti- tity, threedomains with distinct physical-chemical properties
lized to compare the translocation capabilities nas- can be found in virtually all signal sequences: a positively
cent wild type and mutant proteins as their chain                      charged NH2 terminus, a central core of7-16 hydrophobic
lengths were progressively increased. No difference residues, and a more polar COOH-terminal domain (termed
was detected between the two species suggesting that the n-, h-, and c-regions, respectively; Refs. 7, 8). Structural
their initial conformations functionally equivalent alterations have been introduced to test the
                                     are                                                                                functional impor-
as measured by the translocation machinery. We con- tance of these regions. A net (+) charge in the n-region
cludefrom these studiesofpreapolipoprotein-A-IV                        appears to be necessary for efficient translocation in prokar-
t h a t (i) the efficiencyof translocation is not dependent yotes (9,lO) while it is not strictly required in eukaryotes (11,
on signal peptidase cleavage (ii)structural features 12). Introduction residues which decrease the hydrophobic-
                                                  and                                     of
present in the c-region of the signal peptide are not ity (13) and/or the ordered secondary structure (14-16) of the
necessary for interaction with signal recognition                 par- h-region result in dramatic reductions in translocation effi-
ticle or f o r s u b s e q u e n t t a r g e t i to, and translocation ciency. Studies analyzing structure/function relationships in
across, microsomal membranes.                                          the c-region have focused on features required for itsfavorable
                                                                       interactions with eukaryotic signal peptidase. Evidence sug-
                                                                       gests that the choice of the site of cleavage depends on at
   Signal peptides are 15-30-amino acid long segments, lo- least two factors: the properties of the -1 residue‘ (small,
cated at theNH, terminus of most eukaryotic secretory pro- neutral amino acids are favored; Ref. 17) and the distance
teins, which direct their co-translational translocationacross between the potential -1 residue and the hjc boundary (4-5
the endoplasmic reticular (ER)’ membrane (1). Signal pep- residues appear optimal) (57). While it is clear that the C-
tides undergo specific interactions with at least 3 components region is important in recognition and cleavageby signal
of the translocation apparatus. Signal recognition particle peptidase, it is not known whether this region is required for
                                                                       efficient translocation.
   * This work was funded by Grant HL-18577 from the National            Two members of the serineproteaseinhibitor(serpin)
Institutes of Health and by the Monsanto Company. The costs of family, chicken ovalbumin (18, 19) and human plasminogen
publication of this article were defrayed in part by the payment of activator inhibitor 2 (Ref. 201, have internal signal peptides
page charges. This article must therefore be hereby marked “aduer- which are not removed during translocation, suggesting that
tzsement” in accordance with 18 U.S.C. Section1734 solely to indicate
this fact.
                                                                       efficient translocation and secretion can occur without struc-
      Supported by a predoctoral fellowship from the JosiahP. Macy, tural determinants for signal peptidase cleavage. Unfortu-
Jr. Foundation.                                                        nately, since the precise locations of the signal peptide equiv-
   T Supported in part by Medical Scientist Training Program Grant alents in the primary translation products of ovalbumin and
GM-07200.                                                              human plasminogen activator inhibitor 2 mRNAs have not
    ** An Established Investigatorof the American Heart Association. been defined, their structures cannotbe directly compared to
To whom correspondence should be addressed: Dept. of Biological those of cleavable signal sequences. However, one explanation
Chemistry, Washington University School of Medicine, 660 S. Euclid
Ave., Box 8094, St. Louis, MO 63110.                                   for the “dissociation” of co-translational translocation and
     Theabbreviations used are: ER, endoplasmic reticulum; SRP,
signal recognition particle;apo, apolipoprotein; SDS, sodium dodecyl       In this numbering system the site of signal peptidase processing
sulfate.                                                               occurs between the -1 and +1 residues.

                                   UncouplingTranslocation                              from Cleavage                                  4643
cleavage in ovalbumin and human plasminogen activator in-              cationlProcessing Experiments”pSP65 plasmids containing wild
hibitor 2 is that structural features important for transloca-         type or mutant preapoA-IV cDNA inserts were digested with BomHI.
                                                                       In vitro run-off transcription of the linear DNA was achieved using
tion in the c-region of uncleavable signal peptides could be           the protocol described by Folz and Gordon (32) except that 10 pCi of
“retained” in their signal peptides even though structural             [5,6-3H]UTP (41.6 Ci/mmol) was substituted for the [CY-~’P]UTP.
features necessary for cleavage are not. In thisview, structures       Co-translational processing and translocation assays were performed
important for translocation and cleavage may normally co-              with 30-pl reactions containing [36S]methionine(37 pCi), RNasin (40
exist in the c-region but remain functionally distinct.                units), 3 pl of canine pancreatic microsomal membranes (1 pl = 1 eq
   There are a previous studies which demonstrate uncou-
                few                                                    = 0.05 Am unit; Ref. 26), nuclease-treated rabbit reticulocyte lysate
pling of translocation from cleavage after alterationsof secre-        (10 pl), and mRNA (60 ng). After incubating this mixture for 90 min
                                                                       a t 23 “C, trypsin (250 nglpl) and a-chymotrypsin (250 ng/pl) were
tory proteins. Haguenauer-Tsapis and Hinnen (21) have de-              added with or without 1% octyl glucoside. Protease digestion was
scribed a yeast alkaline phosphatase mutant in     which residues      allowed to proceed for 2 h at 0 “C and then stopped by adding a 5-
-3 to +10 were deleted. When the protein was expressed in              fold excess (by weight) of soybean trypsin inhibitor and aprotinin
yeast, some translocation occurred but no signal peptidase             (0.145 trypsin inhibitor units). Radiolabeled apoA-IV polypeptides
cleavage wasdetected. However, a directcomparison of trans-            were recovered with a monospecific rabbit anti-rat apoA-IV sera (30)
location efficiencies of the wild type andmutant proteins was          pre-conjugated to protein A-Sepharose (33). Antigen-antibody-pro-
                                                                       tein A-Sepharose complexes were washed extensively (34), disso-
not reported. Moreover, 30% of the mature domain of this               ciated, and subjected to electrophoresis through 10%polyacrylamide
protein canbe translocated in viuo even when its entiresignal          gel containing 0.1% SDS (35). Autoradiographs of SDS-polyacryl-
peptide is deleted, suggesting the presence of aninternal              amide gels werescanned using an LKBUltroscan XL laser densitom-
targeting signal (22). Substitution of P-hydroxynorvaline for          eter to quantitate the extent of co-translational translocation. The
threonine throughout preprolactin inhibited its cleavage but           loss of the initiator methionine was taken into account when calcu-
not its entry into the ER in viuo (23). Also, an Escherichia           lating the percentage of preapoA-IV translocated into microsomal
coli ompF-lpp fusion protein containing amino acid substitu-              Identification of the Site of Co-translational Proteolytic Cleavage-
tions at the -1, +1, and +2 positions of the ompF domain               Translation reactions were carried out as described above except L-
can undergo efficient post-translational translocation in an                                                            (25
                                                                        [3,4-3H]valine (25 pCi) or ~-[4,6-~H]leucine pCi) was included
E. coli in vitro system, without signal peptidase cleavage (24).       instead of [36S]methionine.Mutant and wild type rat apoA-IV poly-
   We have constructed a 14-amino acid deletion mutant of              peptides synthesized in the presence of canine pancreatic microsomal
rat preapoA-IV, a component of plasma high density lipopro-            membranes were purified by immunoprecipitation and denaturing
                                                                       SDS-polyacrylamide gel electrophoresis. Eluted proteins were sub-
teins and chylomicrons (25). This preprotein has a 20-amino            jected to automated sequential Edman degradation using a 0.33 M
acid long, cleavable NH2-terminal signal peptide with a 3-             Quadrol program (36) and a Beckman 890C spinning cup sequanator
domain organization typical of eukaryotic prepeptides (26).            (30).
This deletion removes 1 residue of the h-region, all of the c-            Assay of Inhibition of Translation by SRP-SRP was prepared from
region, and all potential signal peptidase cleavage sites. When        salt-washed canine pancreatic microsomal membranes according to
translocation and processing of the wild type and mutant               Walter and Blobel (37). SRP (0.5-3 pl of the DEAE-Sepharose CL-
                                                                       6B eluate; Ref. 35) wasadded to wheat germ translation mixtures (25
preproteins were analyzed in uitro, we found that thedeletion          pl) containing RNasin (24 units), [35S]methionine(37 pCi), and 35
did not disrupt SRP interaction or the efficiency of translo-          ng of globin, human preproapoA-I, wild type rat preapoA-IV, or
cation even though no cleavage of the mutant by signal                 mutant rat preapoA-IV mRNAs. Equal volumes of SRP buffer were
peptidase could be detected. These results suggest that thec-          added to control reactions so that the                      of
                                                                                                              final concentration potassium
region contains structural features important cleavage but
                                                  for                                                    M
                                                                       acetate was maintained at 86 m irrespective of SRP or SRP buffer
not necessarily for translocation, i.e. efficient co-translational     additions. After incubation at 23 “C for 60 min, aliquots were re-
                                                                       moved, denatured, and subjected to SDS-polyacrylamide gel electro-
translocation can be uncoupled from co-translational cleav-            phoresis. The extent of SRP-inducedtranslation arrest (38) was
age.                                                                   quantitiated by densitometric scanning of autoradiographs.
                                                                           Synchronized Postinitiation Translocation Assay--Two min after
                   EXPERIMENTAL PROCEDURES                             addition of wildtype or mutant preapoA-IV mRNA to the        reticulocyte
   Materials and Radiochemicals”SP6 polymerase, RNasin, and nu- cell free system, further translation initiation was blocked byadding
clease-treated reticulocyte lysate were from Promega Biotec. Wheat edeine to a final concentration of 3 p~ (39-41). Three p1 of canine
germ lysate was prepared using the method of Anderson et al. (27). pancreatic microsomal membranes were then added to 27 pl of the
Nuclease-treated canine pancreaticmicrosomal membranes were iso- translation mixture a t various times (1-60 min) afteredeine synchro-
lated according to Walter and Blobel (28). TqDNA ligase and rabbit nization. Following introduction of microsomes, each reaction mix-
globin mRNA were     from Bethesda Research Laboratories. The trans- ture was incubated at 23 “C for an additional 60 min. The extent of
lation inhibitoredeine was a gift from Robert Thach and   Glen Lawson translocation of the [35S]methionine-labeled,mascent apoA-IV poly-
(Washington University). ~-[~‘S]Methionine       (1150 Ci/mmol) was peptides into the lumen of the microsomal vesicles was assayed by
purchased from Du Pont-New England Nuclear while ~ - [ 3 , 4 - ~ H ]addition of trypsin and a-chymotrypsin inthe presence or absence of
valine (45 Ci/mmol) and ~[4,6-~H]leucine Ci/mmol) were from octyl glucoside. Following a 2-h incubation a t 0 “C, the protease
Amersham Corp.                                                         activity was stopped by addition of soybean trypsin inhibitor and
   Construction of aRatPreapoA-IV Deletion Mutant-We have aprotinin. The radiolabeled proteins were then denatured and applied
previously described the production of a recombinant 4329-base pair to 10%polyacrylamide gels containing 0.1% SDS.
plasmid containing a 1330-bp rat preapoA-IV cDNA cloned into the
SmaI site of pSP65 (29, 30). Digestion of this plasmid with BalI                            RESULTSANDDISCUSSION
liberated a &?-base fragment encoding residues 15-28 of preapoA-
IV (Fig. 1).The remaining 4287-base pair fragment was re-ligated          Design and Construction of a PreapoA-IV Deletion Mutant
using T, DNAligase and used to transform E. coli strain RR1 (PreapoA-IVAIU-The NHz-terminal sequence                                     of rat
according to theprocedure of Hanahan (31). Since removal of the 42- preapoA-IV including its 20-residue signal peptide is shown
base pair fragment resulted in the loss of NciI and HpaII restriction in Fig. 1. Positions 5-15 of the rat preapoA-IV signal peptide
sites in the recombinant plasmid (Fig. l), digestion with these en- contain a block of hydrophobic residues which are predicted
zymes was used to confirm that the deletion had been successfully by the rules of Chou and Fasman (42) to have a high proba-
accomplished (data not shown). Digestion with BalI was also per-
formed to check that this site had  been “regenerated” thus conserving bility for assuming a P-sheet structure (Fig. 2, panels A and
the two codons which flank the deletion (i.e. GCC and AAT in Fig. B ) . This area represents a typical eukaryotic signal peptide
 )                                                                     h-region (8) as defined by both hydrophobicity and ordered
   In Vitro Transcription, Translation, and Co-translational Translo- secondary structure. This h-region is interrupted by a Thr
4644                               Uncoupling Co-translational Translocation from Cleavage
                                                                   Hpa I1
                                                           Bat I    Nci I                    -
                                                                                             Bal I
                     MelPheleulysAlaValVaILeuThrValAlaleuVslAI                                   snV~lMetTrpAspTyrPheThrGlnLeu...
                      1           5             10                                                                35
                                                                         cleavage site
               FIG. 1. Nucleotide and amino acid sequences of preapoA-IV and the preapoA-IVA14 mutant. The
            first 114 nucleotides of the preapoA-IV cDNA coding regionare shown including Ball and NciI restriction enzyme
            sites used in screening and construction of the preapoA-IVA14 mutant “Experimental Procedures”). Sequences
            removed toproduce the preapoA-IVAl4 deletion mutant shown in black background. This deletionencompasses
            residues 15-28 of preapoA-IV protein includingits signal peptidasecleavage site.

residue at position 16 which denotes the beginning of the                     at the first measurable step of translocation, i.e. recognition
polar c-region. The c-region continues to thesignal peptidase                 by SRP.
cleavage site at Ala’’1.                                                         These results arecompatible with the notion that themost
   A deletion mutation was designed to test whether the c-                    critical features for SRP recognition are contained in the h-
region of rat preapoA-IV contains structural features neces-                  region of signal peptides (46). Mutations designed to alter the
sary for its efficient translocation intothe ER. Such a deletion              length and/or net charge of the signal peptide n-region also
should ideally eliminate the c-region and simultaneously re-                  do not appearto have any effect on interaction with SRP (12,
move all potential signal peptidase cleavage sites while at the               47), although laterstepscan       be affected (47). When the
same time producing minimal perturbations in the physical-                    hydrophobicity of the h-region of bovine preprolactin is re-
chemical properties of mature domains placed just COOH-                       duced by incorporation of a polar leucine analog, @-hydroxy-
terminal to theretained portion of the prepeptide. Addition-                  leucine (48), the translation-arrest effect of SRP is signifi-
ally, the NH2-terminal domain of the mature protein which                     cantly reduced (49). In addition, a Leu to Pro substitution in
is linked to the h-region as a result of mutagenesis, should                  the h-region of the E. coli. EltB signal peptide abolishes SRP-
not have structural featuressimilar to thedeleted c-region. A                 mediated effects (50).
deletion mutant, named preapoA-IVA14, which lacks 1 resi-                        Assay of Translocation and Processing of Mutant and Wild
due of the h-region (Ile15),all of the c-region (Thr“ to Ala2’),              Type PreapoA-IV-To assess whether preapoA-IVAl4 retains
and the first 8 residues of the mature domain (G1u21to Ala%),                 other functions necessary for translocation and processing,
was produced by BalI digestion and subsequent religation of                   reticulocyte lysate cell-free translation systems containing
a rat preapoA-IV cDNA cloned into the plasmid SP65 (see                       [35S]methioninewere programmed with wild type or mutant
Fig. 1).Fig. 2 provides a comparison of the hydrophobicity,                   preapoA-IV mRNAs and translation was carried out in the
secondary structure probabilities, and potential signal pepti-                presence or absence of canine pancreatic microsomal mem-
dase cleavage sites of the NH2-terminal sequences of the wild                 branes. Fig. 4 shows that when translation of wild type apoA-
type and mutant proteins. Their hydrophobic “profiles” are                    IV mRNA occurs in the presence of membranes, a product
                                                                              appears with slightly greater mobility than the   primary trans-
very similar from residues 1 to 23 (Fig. 2, panel A). @-Sheet
                                                                              lation product (compare lanes 3 and 4 ) . This lower molecular
probability was increased in the mutant from positions 15 to
                                                                              weight polypeptide is protected from post-translational pro-
24, thereby extending the length of its predicted @-sheet
                                                                              tease digestion but becomes protease-sensitive if detergent is
structure (Fig. 2, panel B ) . The two regions of high /3-turn
                                                                              included, suggesting that it is translocated and sequestered
probability in themutant (positions 15 and 22) are only
                                                                              within the closed microsomal vesicles (see lanes 5 and 6).
slightly shifted compared to the wild type (Fig. 2, panel C ) .               Translation of mRNA encoding the mutant rat preapoA-
The mutation accomplished its other goal, namely to remove                    IVA14 results in a primary translation product with slightly
potentialsites     for co-translational proteolytic processing                greater mobility than thewild type primary translation prod-
(Thr”, Ala2’, and Ser25; Fig. 2, panel D). The potential for
                           see                                                uct (compare lanes 3 and 10) consistent with its 14-residue
signal peptidase cleavage between positions 13 and30 is very                  deletion. When microsomal membranes are present during
low (Fig. 2, panel D )as predicted by the computer program,                   translation of the mutant preprotein, no lower molecular
SIGSEQ2 (44). This program identifies signal peptidase proc-                  weight band is detected suggesting that no processing of the
essing sites based on a statistical weight matrix derived from                precursor occurs (compare lanes 9 and 10). Lanes 7 and 8
a database of 161 naturally occurring eukaryotic preproteins                  show that the preapoA-IVAl4 mutant is translocated across
(45).                                                                         the microsomal membrane. A comparison of the amount of
   Arrest of Translation by SRP-SRP interacts with signal                     wild type or mutant protein protected from protease digestion
peptides in the wheat germ translation system thereby arrest-                 (lanes 5 and 8) to the amount of protein synthesized in the
ing further elongation of the nascent proteins (38).  Therefore,              presence of membranes (lanes 4 and 9) reveals that the
inhibition of protein synthesis was used to compare the effi-                 efficiency of their translocationis similar (the average percent
ciency of SRP interaction with mutant and     wild type preapoA-              translocation determined in two independent experiments
IV. Fig. 3 shows that the synthesis of mutant and wild type                   was 23and 21% for the wild type and mutantspecies, respec-
preapoA-IV are inhibited by SRP to a very similar degree.                     tively).
The extent of SRP-mediated translation arrest was compa-                         Synchronized Postinitiatwn TranslocationlProcessing As-
rable to that produced with another wild type preprotein,                     say-One interpretation for the data presented in Fig. 4 is
human preproapoA-I. Synthesis of two cytoplasmically tar-                     that the deleted 14-amino acid segment does notcontain
geted proteins, a and /3 rabbit globin, were not affected by the              structural features necessary for (co-translational) translo-
presence of SRP showing that the inhibition is specific for                   cation. However, another possible explanation for the efficient
secretory proteins which have signal peptides. These results                  translocation of the mutant is that the deletion may have
indicate that thepreapoA-IVA14 mutant retains full function                   altered the folding of the nascent protein so that it remains
                                               Uncoupling Co-translational Translocation from Cleavage                                                           4645
                                                                                                                 -0”-    .I..._....._.

                                                                                                                                             g bn
                                                                                                           :                             0

                                 5    10 15 20 25 30               35
                                     Amino AcM Posltion

                                                                              FIG.3. Effects of SRP on nascent protein synthesis. Parallel
                                                                           wheat germ cell-free translation reactions were programmed with a-
                                                                           and @-globinmRNAs (0);       preproapoA-I mRNA (A); preapoA-IV
                                                                           mRNA (0); preapoA-IVA14 mRNA (0).              Translation mixtures
                                                                           containing [35S]methionineand varying amounts of SRP as eluted
                                 5 1510             20   25   30   35      from a DEAE-Sepharose CL-GB column (37) were incubated a t 23 “C
                                                                           for 60 min. Control reactionscontained SRP buffer alone. The
                                El                                         amount of protein synthesis was determined by electrophoretic frac-
                                                                           tionation of the reaction products on SDS-polyacrylamide gels, au-
                                                                           toradiography, and laser densitometry. The datashown for preapoA-
                                                                           IV, preapoA-IVAl4, and preproapoA-I represent the mean of two
                                                                           independent experiments.

                                                                                        Wlld Type   Am             Mutant A E A 1 4
                                                                                ”           - +          + + I1 + +    + “ -                     I
                                                                                           -    -+ + + +
                                                                                            - “ + + - ” -
                                                                                                                       -        - + +                Proteose
                                 5        10   15   20   25   30   35                                                                        +       Oclyl Glucoslde(l%)

                                                                                                        -         -   “

                                                                            1       2      3 4 5 6 7 8 9 1 0 1 1 1 2
                                                                              FIG. 4. In vitro co-translational translocation and cleavage
                            I                                 1            assays ofpreapoA-IV and preapoA-IVA14. Reticulocyte lysates
                                     15     20     25         30           containing [nsS]methionine were programmed with mutant or wild
                                     Amino Acid Position                   type preapoA-IV mRNA in the presence or absence of canine pan-
   FIG. 2. Comparative structural predictions of preapoA-IV                creatic microsomal membranes. After a 90-min incubation a t room
and preapoA-IVA14. Hydrophobicity (panel A ) , @-sheetprobabil-            temperature, apoA-IV was recovered by immunoprecipitation with
ity (panel B ) , @-turnprobability (panel C),and prediction of the site    monospecific antisera and subjected to electrophoresis through a10%
of signal peptide cleavage (panel D )are plotted for mutant (- -) and      polyacrylamide gel containing 0.1% SDS. Where indicated immuno-
wild type (-) preapoA-IV. Hydrophobicity plots were calculated by          precipitation was preceded by incubation with trypsin and a-chymo-
averaging residue-specific values across a windowof four residues,         trypsin (250 ng/pl each) and/or octyl glucoside (1%). Equal volumes
according to the algorithm of Kyte and Doolittle (43). Positive and        from each reaction were subjected to the indicated treatments. The
negative values indicate hydrophobic and hydrophilic sequences, re-        band corresponding to processed wild type apoA-IV is denoted by the
spectively. @-Sheet  and @-turn probabilities were calculated according    arrow.
to the rules of Chou and Fasman (42) using windows of 5 and 4
residues, respectively. As the curve for @-sheetprobability rises above
                                                                anism that is distinct from the route taken by other secretory
1.05, one criteria for propagation of a @-sheetis satisfied. The com-
                                                                proteins in vitro.
puter program SIGSEQ2 (44) was used to calculate signal peptidase
processing values (S-values) at the indicated amino acid positions.To address this issue, we compared the ability of nascent
                                                                wild type and mutant proteins of increasing chain length to
This program, based on rules from von Heijne (45), predicts process-
                                                                be translocated across microsomal membranes. Translation
ing to occur where the S-value is the highest. Typical S-values for
cleavage sites of eukaryotic preproteins range from +3 to +15. The
                                                                ofwild type and mutant mRNAs in the reticulocyte lysate
position of signal peptidase cleavage ofwild type preapoA-IV is system was “synchronized” by adding edeine, a translation
denoted by the arrow.                                           initiation inhibitor (39, 40, 53). Edeine was added to parallel
                                                                reaction mixtures two minutes after they were programmed
in a translocation-competent state for a longer period than with either preapoA-IV or preapoA-IVAl4 mRNAs. Under
the wild type preprotein. There is a precedent for sucha view: these conditions full length preapoA-IV and preapoA-IVA14
correlation of a lack of folding of the E. coli maltose-binding first appear between 10 and 20 min of translation and con-
protein with export competence has been observed previously tinue to appear until the 30-min time point, after which very
(51, 52). Moreover, since the mature domain of apoA-IV is little additional chain completion occurs. No translation was
largely composedof tandemly arrayed, docosapeptide repeats detected if edeine was added at 0 min, indicating that the
withpredicted amphipathic a-helical structure (29), it is inhibition of initiation by edeine was complete (see panel A ,
possible that this polypeptide may be able to insert itself Fig. 5). Fig. 5B, presents the results of an experiment where
through (or interact with) the membrane bilayer via a mech- aliquots of “synchronized” translation mixtures containing
4646                            Uncoupling
                     Co-translational                                    Translocation from Cleavage
                                                                         6, 7, 10, and 13 and leucine residues at positions 3, 8, and 12.
                                         40   I                     After removal of the 20-amino acid signal peptide, valine
                                                                    appears at positions 2, 7, and 10 while leucine appears at 18.
                                                                    Panels A and C, Fig. 6, show the results of sequencing im-
                                                                    munopurified wild type apoA-IV synthesized in the presence
                                                                    of microsomes and [3H]valine and [3H]leucine, respectively.
                                                                    Dominant peaks of [3H]valine were noted at cycles 2, 7, and
 gcl                                                                10. A dominant peak at cycle 18 was observed when [3H]
    e t
        Id&/ I   , , , ,       ~           1, \.
                                           !   U      , -, , ,
                                                      " "
                                                     " "
                                                                    leucine wasused. These peaks areconsistent with signal
                                                                    peptide removal. Peaks of radioactivity at cycles 6 and 13 in
                                                                    panel A as well as at cycles 3,8, and12 of panel C correspond
          0     20     40   60            0       20     40      60 to uncleaved preapoA-IV. The primary translation product of
             Time Time                        After Edeine Addition
                (minutes)                        (minutes)          the preapoA-IVAl4 mutant hasvaline residues at positions 6,
                                                                    7, 10, 13, and 16 and leucine residues 3, 8, 12, and 24. Panels
   FIG.5. Edeine-synchronizedpostinitiationtransloca-               B and D show that when the mutant is synthesized in the
tion assay. Panel A shows the rate of synthesis of [36S]methionine-
labeled preapoA-IV ( 7 and mutant preapoA-IVA14 (0)in reticulo-
cyte lysates where the translation inhibitor edeine was added 2 min
after the initiation of translation. The result of adding edeine just
before addition of preapoA-IV mRNA (time = 0 min) is also shown
(A).The primary translation products were quantitated by densito-
metric scanning of autoradiographs of SDS-polyacrylamide gels. The
data at each time point for the mutant and wild type proteins have
been arbitrarily connected. The results of assaying the translocation
competency of wild type and mutantnascent chainsat various stages                                                          preapo AIV (UtlClMVed)
of their synthesis are shown in panel B . As in panel A , edeine was                                                       apo AIV ( c b m d )
introduced 2 min after the addition of mRNAs encoding preapoA-IV
or preapoA-IVAl4. At the indicated times, aliquots of the edeine-
synchronized translation mixtures were added to canine microsomes
and incubated for an additional 60 min. Reaction products were
separated by electrophoresis through 10% polyacrylamide gels con-
taining 0.1% SDS and quantitated by densitometric scanning of
autoradiographs. The percent translocation was calculated by divid-
ing the amount of mutant or wild type apoA-IV protein protected
from digestion from exogenous protease by the total amount present
(translocated plus nontranslocated) after incubation with the micro-
somes. The mutant and    wild type proteins protected from proteolysis
in this assay were completely digested in the presence of protease
plus 1%octyl glucoside (data not shown).

nascent mutantor wild type apoA-IV sequences were removed
at the indicated times and presented to canine pancreatic
microsomal membranes. The extent of translocationasa
function of increasing chain length was measured by compar-
ing the amount of full length [35S]methionine-labeled,wild
type or mutant proteins resistant to protease degradation to
the total amount (translocatedplus nontranslocated) present
                                                                                     301     pl        1

after incubation with the microsomal membranes (as in Fig.
4). After synchronized translation for 20 min, the transloca-
tion competency of both the wild type and mutant nascent
chains decreases -10-fold from 30 to 2-3%. Since only slightly
over one-half of the polypeptide chains are completed after                             0.
                                                                                        M F ~ A V ~ T V A K N A N V M W O Y F mutant proapo AIVAU
20 min (panel A ) , we conclude that translocation of both                                           x)            20
preapoA-IV and the preapoA-IVA14 mutant is essentially a                                          Cycle Number
co-translational event in uitro. The decrease in translocation
competency as afunction of increasing chain length is similar               FIG.6. NHz-terminal sequence analysis of co-translation-
                                                                         ally processed preapoA-IV and preapoA-IVAl4. I n vitro trans-
for the wild type and mutant proteins, indicating that pres-             lation of mRNAs encoding mutant and wild type preapoA-IV was
ervation of efficient translocation after deletion of the 14-            carried out in parallel reticulocyte lysate cell-free systems containing
amino acid segment is due to recognition of "appropriate"                translocation-competent canine pancreatic microsomal membranes
features by the translocation apparatus rather than prolon-              and f3H]valine or [3H]leucine. The radiolabeled products were im-
gation of translocation competency due to formation of less              munoprecipitated using a monospecific antisera then separated by
                                                                         electrophoresis through 10% polyacrylamide tube gels containing
stable secondary or tertiary structures.                                 0.1% SDS. ApoA-IVpolypeptides were passively eluted from gel slices
   Analysis of SignalPeptidaseProcessing-The         data pre-           and subjected to automated sequential Edman degradation. Panels A
sented in Figs. 4 and 5 suggest that deletion of the 14-residue          and B shows the results obtained when 73,000 and 780,000 dpm of
domain uncouples co-translational    translocationand       co-          [3H]valine-labeledpreapoA-IV and preapoA-IVA14 polypeptides were
translational proteolytic processing. To establish definitively          sequenced, respectively. Panels C and D show the results when
                                                                         210,000 and 180,000 dpm of [3H]leucine-labeled preapoA-IV and
that the preapoA-IVAl4 mutant is not cleaved by signal                   preapoA-IVA14 weresequenced, respectively. The solid arrows denote
peptidase, [3H]valine-, or [3H]leucine-labeledproducts of i n            peaks arising from the co-translationally cleaved protein, while
vitro co-translational translocation were subjected to Edman             dashed arrows indicate peaks which correspond to theintact primary
degradation. Rat preapoA-IV has valine residues at positions             translation product.
                                      Uncoupling Co-translational Translocation from Cleavage                                                            4647
                                                                  4. Gilmore, R., Blobel, G., and Walter, P. (1982) J. Cell. BioL 96,463-469
presence of microsomes, immunopurified, and subjected to          5. Weidmann, M., Kurzchalia, T. V., Hartmann, E., and Rapoport, T. A.
Edman degradation, only peaks of radioactivity corresponding            (1987) Nature 328,830-833
                                                                  6. Jackson, R.C., and Blobel, G . (1977) Proc. NatL Acad. Sci. U. S. A. 7 4 ,
to the uncleaved, primary translation product are detected.             559b5602
This datais consistent with the observation that theelectro-      7. Perlman, D., and Halvorson, H. 0.        (1983) J. Mol. Bwl. 167,391-409
                                                                  8. von Heijne, G. (1985) J. Mol. Biol. 184,99-105
phoretic mobility of the preapoA-IVAl4 mutant does not            9. Vlasuk. G. P.. Inouve. S.. Ito. H.. Itakura. K... and Inouve. M. (1983) J.
                                                                                                                      .             - .     .   .
change co-incident with translocation into canine   microsomes          Biol.’Chem.’268,”’7141-7148                   ’

                                                                 10. Iino, T., Takahashi, M., and Sako, T. (1987) J. BWL Chem. 262, 7412-
(Fig. 4) and with the prediction of the SIGSEQ2 computer                7417
program (Fig. 20) that the mutantprotein has no favorable 11. Lipp, J., and Dohberstein, B. (1986) Cell 46,1103-1112
cleavage sites. We can therefore conclude that preapoA-IVA14 12. Garcia. P. D.. Ghraveb. J.. Inouve.’ M..’ and Walter.. P. (1987) J. Biol. Chem.
                                                                        262; 946319468- ‘                    ’
                                                                                                                                .     .
is efficiently translocated with an uncleavable signal se- 13. Rapoport, T. A. (1986) CRC Crit. Rev. B k h e m . 20,73-137
                                                                 14. Emr, S. D.,and Silhavy, T. J. (1983) Proc. Natl. Acad. Sci. U.S. A. 80,
quence.                                                                 4599-460.1
                                                                            . ..
                                                                         .. ..
   The demonstration that in vitroco-translational cleavage 15. Bankaitis, V.A,, Rasmussen, B.A., and Bassford, P. J. (1984) Cell 37,
of preapoA-IVAt4 was prohibited is compatible with the 16. Briggs, M. S., and Gierasch, L. M. (1984) Biochemistry 23,3111-3114
results of recent studies that examined structural character- 17. Folz, R. J., Nothwehr, S. F., and Gordon, J. I. (1987) J. BwL Chem. 263,
istics important for recognition and cleavage by signal pepti- 18. Palmiter, R. D., Gagnon, J., and Walsh, K. A. (1978) Pmc. NatL Acad. Sci.
dase, As noted inthe Introduction, the site of cleavage chosen          U. S. A. 76,94-98
by eukaryotic signal peptidase is most likely to occur 4-5 19. Meek, R.L., Walsh, K. A., and Palmiter, R. D. (1982) J. Bwl. Chem. 267,
residues COOH-terminal to theboundary between the h- and 20. Ye, R. D., Wun, T.-C., and Sadler, J. E. (1988) J. BWL Chem. 263,4869-
c-regions (57)and after small, neutral residues such as Ala 21. Haguenauer-Tsapis, R., and Hinnen, A. (1984) MOL Cell. Biol. 4 , 2668-
and Gly (17). In addition, a direct correlation between p-turn          2675
                                                                 22. Silve, S., Monod, M., Hinnen, A., and Haguenauer-Tsapis. R. (1987) Mol.
probability in the region of the cleavage site and signal pep-          Cell. Biol. 7 , 3306-3314
tidase cleavage has been observed in both prokaryotic (54,55) 23. Hortin, G., and Boime, I. (1981) J. Biol. Chem. 286,1491-1494
                                                                 24. Yamane, K., Matsuyama, S., and Mizushima, S. (1988) J. Bwl. Chem. 263,
and eukaryotic signal peptides (57). While analysis of the              5368-5372
preapoA-IVAl4 sequence shows only a slight decrease in the 25. Swaney, J. B., Braithwaite, F., and Eder, H. A. (1977) Biochemistry 16,
length of its hyrophobic domain (Fig. 2, panel A ) , the lack of 26. Gordon, J. I., Smith, D. P., Alpers, D. H., and Straws,A. W. (1982) J. Biol.
                                                                        Chem. 267,8418-8423
disruption of predicted @-sheetstructure from positions 16- 27. Anderson, C. W., Straus, J. W., and Dudock, B. S. (1983) MethodsEnzymol.
22 (Fig. 2, panel B ) may prevent access of signal peptidase to         10 1. -- - - - .
                                                                        - - -, fiR.5444
any cryptic cleavage sites. Moreover, the absence of favorable 29. Walter, P., and Blobel, G. (1983) Methods J. M., and Gordon, J. I. (1984)
                                                                     Boguski, M. S., Elshourba           N., Taylor,
                                                                                                                     EnzymoL 96,84-93
-1 amino acids located at anoptimal distance downstream of              P m . Natl. Acad. Sei. U . F A . 81,5021-5025
                                                  ~ ~ - h e ~ 30. Fob, R. J., and Gordon, J. I. (1986) J. Biol. Chem. 261, 14752-14759
the h/c-boundary ( T ~ p ~ * - A s p ~ ~ - T y rwouldP also~be) ~31. Hanahan, D. (1985) DNA Cloning Technology, Cold Spring Harbor Labo-
expected to prohibit cleavage.                                          ratory, Cold Spring Harbor, NY
                                                                                   I., Smith, D. P., Andy, R., Alpers, D. H.,
   Finally, it remains to be determined if the n- plus h-domain 32. Gordon, J. A. W. (1982) J. Biol Chem. 267,971-978 Schonfeld, G., and
of the rat apoA-IV signal peptide is sufficient to promote co- 33. Gordon J. I., Budelier, K. A,, Sims, H. F., Edelstein, C., Scanu, A. M., and
translational translocation if it is displaced to other internal 34. Gordon, J. A. W. (1983b) J.Lentz,Chem. 268,14054-14059
                                                                                   I., Sims, H. F.,
                                                                                                             S. R., Edelstein, C., Scanu, A.M., and
regions of the mature plasma protein segment or is “ligated”            Strauss, A. W. (1983a) J. Bioi. Chem. 268,4037-4044
                                                                 35. Laemmli, U.K. (1970) Nature 227,680-685
to the NH, terminus of proteins which arenot normally 36. Thomas, K. A., Silverman, R. E., Jeng I., Baglan, N. C., and Bradshaw, R.
secreted (see Ref. 56). Moreover, while our results strongly            A. (1981) J. BioL Chem. 256,9147-’9155
suggest that the c-region is primarily involved in signal pep- 38. Walter P., and Blobel G. (1981) J. Cell. Biol. 91 557-561
                                                                 37. Walter,
                                                                                   and Blobel,
                                                                                                    (1983) Methods Enzymol. 96,682-691

tidase interactions, it will be important to expand these  types 39. Hunt, +. (1974) Ann. k. Y. Acnd. Sci. 241,223-;31
                                                                                                      J. (1978) J. BWL Chum. 263,6568-6577
of studies to other preproteins where the consequences of 40. Kozak, M.,M., Shatkin, A.E., and Thach,R. E. (1982) J. Bwl. Chem. 257,
                                                                 41. Detjen, B.
                                                                                       Walden, W.
deleting their c-regions or substituting naturally occurring,           amr;-onfin  I”    ““Y

                                                                     Chou P. Y.,                                  Annu. Rev. Biochem. 47,251-276
“heterologous,” or synthetic c-regions with “epitomized” 42. Kyte,’J., andand Fasman G. D. (1978)MOLBiol. 157, 105-132
                                                                 43.                  Doolittle, R: F. (1982) J.
structures, are assessed both in vitro in vivo.
                                        and                      44. Folz, R. J., and Gordon, J. I. (1987a) Biochern. Biophys. Res. Commun.
                                                                              45. von Heijne, G. (1986) Nucleic Acids Res. 14,4683-4690
  Acknowledgments-We are grateful to Glen Hortin for critical                 46. Finkelstein, A. V., Bendzko, P., and Rapoport, T. A. (1983) Febs. Lett.
reading of the manuscript and insightful comments, to Mark Frazier            47. Szczesna-Skorupa, E., Mead, D. A., and Kemper, B. (1987) J. Biol. Chem.
for maintaining the spinning cup protein sequencer, as well as Nick                 262,8896-8900
Davidson (University of Chicago) for generously providing rabbit              48. Hortin, G., and Boime, I. (1980) Pmc. Natl. Acad. Sci. U.S. A. 77, 1356-
anti-rat apoA-IV sera.                                                               1360
                                                                              49. Walter, P.,Ibrahami, I. and Blobel G. (1981) J. Cell Biol. 91,545-550
                                                                              50. Ibrahimi I., and Gentz,’R. (1987) Biol. Chem. 262,10189-10194
                              REFERENCES                                      51. Randal1,’L. L., and Hardy, S. J. S. (1986) Cell 46,921-928
                                                                              52. Collier, D. N., Bankaitis, V. A., Weiss, J. B., and Bassford, P. J. (1988) Cell
 1. Walter P., Gilmore, R., and Blobel, G. (1984) Cell 38,5-8                       63,273-283
 2. Walter, P., and Blobel, G . (1981) J. Cell Biol. 91,557-561               53. Folz, R. J., and Gordon, J. I. (1987) J. Bwl. Chem. 262,17221-17230
 3. Meyer, D. I., Krause, E., and Dobberstein, B. (1982) Nature 297,647-650   54. Inouye, S., Duffaud, G., and Inouye, M. (1986) J. Biol. Chem. 261,10970-
                                                                              55. Duffaud, G., and Inouye, M. (1988) J. BWL Chem. 263,10224-10228
   The numbering system used here for the deletion mutant is the              56. Perara, E., and Lingappa, V. (1985) J. Cell Bwl. 101,2292-2301
same as is shown for the wild type in Fig. 1.                                 57. Nothwehr, S. F., and Gordon, J. I. (1989) J. Biol. Chem. 264,3979-3987

Shared By: