Proc. Natl. Acad. Sci. USA 79 (1982)

Correction. In the article "Secretory proteins induced in humanfibroblasts under conditions used for the production of inter-feron 1' by Jean Content, Lucas De Wit, Denis Pierard, RikDerynck, Erik De Clercq, and Walter Fiers, which appearedin number 9, May 1982, of Proc. Nati. Acad. Sci. USA (79,2768-2772), an error occurred in the Proceedings editorial of-fice. The fourth line of the abstract should say "molecularmasses of 22 and 27 kilodaltons. "

Correction. In the article "DNA sequences of the joining re-gions of mouse A light chain immunoglobulin genes" by BonnieBlomberg and Susumu Tonegawa, which appeared in January1982, issue no. 2, of Proc. Natl. Acad. Sci. USA (79, 530-533),the authors request the following correction. Fig. 2 containedan error, reported as the insertion of 1 base pair (thymine) im-mediately after the codon for amino acid position 100, in the J4sequence. The region of the figure showing that portion of thesequence is shown corrected below. In addition, the discussionon p. 531 under the section "The A4 Gene is Probably a Pseu-dogene" should read, beginning with the second sentence: "Itappears that J4 has undergone a 2-base pair deletion in the sig,nal heptamer. In all functional J segments, a dinucleotide G-T,which is an obligatory part of a RNA splicing signal (30) occursat the position corresponding to amino acid residue 110 (Fig.3 and ref. 17). In contrast, this dinucleotide is absent in all read-ing frames of J4 at this position. " The conclusion that A4 is prob-ably a pseudogene remains valid, based on the heptamer changeand donor splice signal change, but there is no insertion of anextra nucleotide in the J4 sequence as compared with the J1sequence. These results are in agreement with those recentlyreported by Miller et al. [Miller, J., Selsing, E. & Storb, U.(1982) Nature (London) 295, 428-430].

Correction. In the article "Role ofpositive charge on the amino-terminal region of the signal peptide in protein secretion acrossthe membrane" by Sumiko Inouye, Xavier Soberon, ThomasFranceschini, Kenzo Nakamura, Keiichi Itakura, and MasayoriInouye, which appeared in June 1982, issue no. 11, ofProc. NatLAcad. Sci. USA (79, 3438-3441), the authors request that thefollowing correction be noted. On page 3439, line 5 should read"at 37TC, 440C, 390C, and 370C for 16-18 hr .

Correction. In the article "5a-Cholest-8(14)-en-3(3-ol-15-one,a potent inhibitor of sterol biosynthesis, lowers serum choles-terol and alters distribution of cholesterol in lipoproteins inbaboons" by George J. Schroepfer, Jr., Edward J. Parish,Alemka Kisic, Evelyn M. Jackson, Cynthia Mersinger Farley,and Glen E. Mott, which appeared in May 1982, issue no. 9,in Proc. Natl. Acad. Sci. USA (79, 3042-3046), an editorial erroroccurred on p. 3043. In the right-hand column, line 23 shouldread "(19%, 28%, and 20%), LDL/VLDL . . ."

98 99 102 106 110

TrpValPheGlyGlyGlyThrLysLeuThrValLeuGZyJ1 Ig25A AAATGCATGC-MGGTTTTTGCATGAGTCTATATCACAGTGCTGGGTGTTCGGTGGAGGAACCAAACTCACTGTCCTAGGTGAGTGACTCCTTCCTCCT

* * * ** *

** * ** * * * ** *TrpValPheGZyGlyGlyThrArgLeuAlrValLeuAsp *J4 Ig1OA1 AGGTACATGCAGAGTTTTTTGCATTAGACTATAT--CAGTGTTGGGTGTTCGGAGGTGGAACCAGATTGACTGTCCTAGATGAGTGACTCCTCCCTCCT

PheIlePheGlySerGlyThrLysValThrValLeuGlyJ3 IgS8.2 TGCTTGCCCCACAGGTTTAGGGTTGGGTTTCAGTCACTGTGGTTTATTTTCGGCAGTGGAACCAAGGTCACTGTCCTAGGTAAGTGGCTTTAATGCTTC

* * *

* * * * * *TyrVaZPheGlyGlyGZyThrLysValThrValLeuGly * * * *J2 IglOAl TGCTGGCCCCATAGGTTTTGGGTTGGGTTTTAGTCATTGTGTTATGTTTTCGGCGGTGGAACCAAGGTCACTGTCCTAGGTAAGTAGTTTCAAAGC

FIG. 2. Comparison of nucleotide sequences of germ-line A J segments and surrounding regions. *, Nonidentical base pairs in comparisons ofthe J1-J4 and J3-J2 sequences. Signal nonamer and heptamer sequences 5' to the J regions are underlined. Amino acids encoded by the nucleotidesequences are shown in italics. The J1 DNA sequence is taken from ref. 28. (The remainder of this legend is shown on p. 532 of this article.)

4828 Corrections

Dow

nloa

ded

by g

uest

on

June

17,

202

1 D

ownl

oade

d by

gue

st o

n Ju

ne 1

7, 2

021

Dow

nloa

ded

by g

uest

on

June

17,

202

1 D

ownl

oade

d by

gue

st o

n Ju

ne 1

7, 2

021

Dow

nloa

ded

by g

uest

on

June

17,

202

1 D

ownl

oade

d by

gue

st o

n Ju

ne 1

7, 2

021

Dow

nloa

ded

by g

uest

on

June

17,

202

1 D

ownl

oade

d by

gue

st o

n Ju

ne 1

7, 2

021

Proc. Natl. Acad. Sci. USA 79 (1982) 6107

Correction. In the article "Role of positive charge on theamino-terminal region ofthe signal peptide in protein secretionacross the membrane" by Sumiko Inouye, Xavier Soberon,Thomas Franceschini, Kenzo Nakamura, Keiichi Itakura, andMasayori Inouye, which appeared in number 11, June 1982, ofProc. NatL Acad. Sci. USA (79, 3438-3441), an incorrect versionof Fig. 1 was printed by mistake. The correct figure is printedbelow.

WILD TYPE

1 5F-METLYSALATHRLYSLEU-

:-MTAATGAAAGCTACTAAACTG-

CHARGE

:+2

OLIGONUCLEOTIDE 1 (16-MER):

OLIGONUCLEOTIDE 2 (19-MER):

OLIGONUCLEOTIDE 3 (16-MER):

OLIGONUCLEOTIDE 4 (15-MER):

51ATGAAAGATACTAAACOH3F-METLYSTgHRLYS-

5tTTAATAATG-GCTACTAMCOH"F-METEIALATHRLYS-

51ATAATG -GATACTAAACOH3'F-METE:i THRLYS-

51ATAATG6AAGATACTOH3'F-METGEM THRLYS-

FIG. 1. Nucleotide sequences of synthetic oligonucleotides. Bases that are altered from the wild-type sequences (6) are marked by a dot. Aminoacid residues changed as a result of mutations are boxed.

Correction. In the article "Amino acid sequence of mousesubmaxillary gland renin" by Kunio S. Misono, Jin-Jyi Chang,and Tadashi Inagami, which appeared in number 16, August1982, of Proc. Natl. Acad. Sci. USA (79, 4858-4862), theauthors request that the following correction be noted. TheMr of the heavy chain in the reduced form was given in-correctly as 31,036 in line 4 of the Abstract (p. 4858) andin line 4 of the Discussion (p. 4859). The correct Mr of theheavy chain in the reduced form is 31,043.

Correction. In the article "Clonal analysis of human cytotoxicT lymphocytes: T4' and T8' effector T cells recognize productsofdifferent major histocompatibility complex regions" by StefanC. Meuer, Stuart F. Schlossman, and Ellis L. Reinherz, whichappeared in number 14, July 1982, ofProc. Natl. Acad. Sci. USA(79, 4395-4399), a typographic error was not caught. On p.4395, in the fourth line from the bottom in the left-hand column,'T4/8+" should have been "T5/8'."

Correction. In the article "Mitochondrial DNA polymorphismin a maternal lineage of Holstein cows" by William W. Haus-wirth and Philip J. Laipis, which appeared in number 15, Au-gust 1982, of Proc. Nati Acad. Sci. USA (79, 4686-4690), theauthors request that the following correction be noted. In thelegend to Fig. 3 on p. 4688, lines 9 and 10 should read ". . .cloned fragment of H949B."

:+1

:+1

: o

: -1

Corrections

Proc. Natd Acad. Sci. USAVol. 79, pp. 3438-3441, June 1982Biochemistry

Role of positive charge on the amino-terminal region of the signalpeptide in protein secretion across the membrane

(loop model/site-specific mutagenesis/synthetic oligonucleotide/lipoprotein/outer membrane)

SUMIKO INOUYE*, XAVIER SOBERONtf, THOMAS FRANCESCHINI*, KENZO NAKAMURA*§, KEIICHI ITAKURAt,AND MASAYORI INOUYE**Department of Biochemistry, State University of New York at Stonv Brook, Stony Brook, New York 11794; and tDivision of Biology, City of Hope ResearchInstitute, 1450 East Duarte Road, Duarte, California 91010

Communicated by Cyrus Levinthal, February 23, 1982

ABSTRACT The positively charged amino-terminal region ofthe signal peptide has been proposed to have an important roleat an initial step of protein secretion across the membrane (loopmodel). To test this hypothesis, the charge on the amino-terminalregion ofthe signal peptide ofthe prolipoprotein ofthe Escherichiacoli outer membrane was altered by using synthetic oligonucleo-tides from +2 to + 1, 0, and -1 by guided site-specific mutagenesisof a plasmid DNA carrying an inducible lipoprotein gene. Thewild-type sequence of this section, Met-Lys-Ala-Thr-Lys (+2), wasthus changed to Met-Lys-Asp-Thr-Lys (I-1; + 1), Met-Ala-Thr-Lys(I-2; + 1), Met-Asp-Thr-Lys (I-3; 0), and Met-Glu-Asp-Thr-Lys(I4; -1). After induction of lipoprotein production, cells werepulse labeled with [35S]methionine for 10 sec. The lipoprotein ofI-1, 1-2, and I-3 was assembled in the membrane, although therates of lipoprotein production progressively decreased as thecharge on the signal peptide became more negative. Conversely,in the case of I4, only a small amount of lipoprotein assembledin the membrane while a large amount of glycerol-unmodifiedprolipoprotein accumulated in the cytoplasm. This soluble proli-poprotein was gradually and posttranslationally secreted acrossthe membrane to be modified and assembled in the membrane.These results indicate that the positively charged amino-terminalregion of the signal peptide plays an important role in efficientprotein secretion across the membrane.

The signal peptide, an extension of the amino-terminal end ofa secretory protein, is, required for secretion of the proteinacross the membrane (for review, see ref. 1). Amino acid se-quences of signal peptides from various proteins have been de-termined for both prokaryotes and eukaryotes. In contrast toeukaryotic signal peptides, the structures of prokaryotic signalpeptides have many common features (1). These common fea-tures substantiate a model (loop model) originally proposed toexplain the functions of the signal peptide (2). In this model,the positively charged amino-terminal region allows the attach-ment initially of the signal peptide and consequently of thepolysome to the negatively charged inner surface of the cyto-plasmic membrane by ionic interaction (1). Subsequently, thenext hydrophobic section of the protein is inserted into the cy-toplasmic membrane, thus forming a loop at the signal peptide(1, 3). This model contradicts the linear model in which the sig-nal peptides are secreted linearly across the membrane (4).We have examined the role of the positive charge on the

amino-terminal region of the signal peptide during protein se-cretion. For this purpose, we changed the charge on the amino-terminal region of the prolipoprotein of the Escherichia coliouter membrane from +2 to + 1, 0, and -1. We found that,

when the charge was -1, a large amount of the glycerol-un-modified prolipoprotein accumulated in the cytoplasm. Thissoluble prolipoprotein was gradually secreted across and assem-bled into the membrane. These results support the loop model.

MATERIALS AND METHODSBacterial Strains and Plasmids. E. coli K-12 strain JA221

(hsdM+hsdRileuB6 lacY A trpE5) was used. JA221 lpp- wasisolated from JA221 as a globomycin-resistant mutant on an L-broth plate. JA221 Ipp-/F'laciq lacZ' lacY+ proA+ proB+ wasobtained by transferring F prime factor from X90 (ara lac-pronalA argEam rtfr thi)/F' laCiq lacZ+ proA+ proB+ (obtainedfrom J. Beckwith, Harvard University) into JA221 lpp-.

Cells were grown in M9 medium supplemented with glucose(4 mg/ml), leucine (20 ,g/ml), tryptophan (20 ,ug/ml), methi-onine (2 ,ug/ml), thiamin (2 ,ug/ml), and MgSO4-7H20 (200yg/ml). Ampicillin (50 ,g/ml) was added for growth of theplasmid-harboring bacteria.

Plasmid pKEN125 (3.9 kilobases) was used for guided site-specific mutagenesis. This plasmid was derived from pBR322and carried an inducible lpp gene. This gene was constructedby inserting a DNA fragment carrying the lacUV5 promoter-operator region as a transcriptional control switch into the 5'-untranslated region of the lpp gene (5). JA221 Ipp-/F'laciq wasused as a recipient and, in the absence of a lac inducer, theproduction of the lipoprotein was negligible. Addition of iso-propyl-,B3D-thiogalactoside (iPr-S-Gal; Sigma) induces lipopro-tein production to =4 times that of the Lpp+ wild-type cells.

Guided Site-Specific Mutagenesis. Mutagenesis was carriedout by using synthetic oligonucleotides (Fig. 1) according to themethod of Wallace et al. (7) with the following modification.pKEN125 was partially digested with EcoRI (Bethesda Re-search Laboratories) in the presence of ethidium bromide, andthen the ethidium bromide was removed by regular phenolextraction and the DNA was precipitated with ethanol withoutusing Sephadex G-25. The DNA was treated with exonucleaseIII (New England BioLabs) and digested with Hinf I (New En-gland BioLabs), and this was followed by phenol extraction andethanol precipitation. Then, the DNA was treated with bacterialalkaline phosphatase (Worthington). The final single-strandedclosed circular DNA was extracted with phenol and precipitatedwith ethanol without using Sephadex G-25.

The oligonucleotides were synthesized by a solid-phase tries-ter method (ref. 8; H. Ito, unpublished data). Plasmid pKEN125

Abbreviation: iPr-S-Gal, isopropyl-3-D-thiogalactoside.t Present address: Instituto de Investigaciones Biomedicas, ApartadoPostal 70228, Ciudad Universitaria, 20 D. F., Mexico.

§ Present address: Mitsubishi-Kasei Institute of Life Sciences, 11 Min-amiooya, Machida-shi, Tokyo 194, Japan.

3438

The publication costs ofthis article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertise-ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 79 (1982) 3439

1 5F-METLYSALATHRLYSLEU-

-AATAATGAAAGCTACTAAACTG-

OLIGONUCLEOTIDE 1 (16-MER):

OLIGONUCLEOTIDE 2 (19-MER):

OLIGONUCLEOTIDE 3 (16-MER):

OLIGONUCLEOTIDE 4 (15-MER):

51ATGAAAGATACTAAACOH 3F-METLYS9IHRLYS-

5AATAATG-GCTACTAMCOHH3F-METEJALATHRLYS-

51ATAATG -GATACTAAACOH3'F-METEli~THRLYS-

51ATAATG4MGATACTOH3'F-MET!D9THRLYS-

FIG. 1. Nucleotide sequences of synthetic oligonucleotides. Bases that are altered from the wild-type sequences,(6) are marked by a dot. Aminoacid residues changed as a result of mutations are boxed.

DNA was used for mutagenesis with oligonucleotides 1, 2, and3. Plasmid DNA from the mutant isolated with use of oligo-nucleotide 1 was used for mutagenesis with oligonucleotide 4.Colony hybridization;to detect mutant colonies was carried outat 370C, 440C, 490C, and 370C for 16-18 hr for oligonucleotides1, 2, 3, and 4, respectively.

Labeling Experiment. Cultures were grown at 370C to 2X 10' cells/ml, and then iPr-S-Gal was added to a final con-centration of 2 mM. Twenty minutes after induction, 1 ml ofthe culture was pulse labeled with 50 ,Ci of [3S]methionine(Amersham; 1,390 Ci/mmol; 1 Ci = 3.7 x 10'° becquerels) for10 sec at 37°C. Incorporation was stopped by adding 1 ml of 40mM sodium phosphate buffer, pH 7.1/0.8% formaldehyde con-taining methionine (2 mg/ml) (stopping solution). Chase ex-periments were carried out by adding 40 A1l of methionine (5mg/ml) at 10 sec and stopped at 2 min by adding 1 ml of stop-ping solution. The membrane fraction was prepared as de-scribed (9). Immunoprecipitation was carried out by using rab-bit antilipoprotein serum as described (10).

NaDodSO4/Polyacrylamide Gel Electrophoresis. To sep-arate the unmodified prolipoprotein and the mature lipopro-tein, the following NaDodSO4/polyacrylamide gel system wasused. Separation gels were prepared with 17.5% acrylamide/0.074% bisacrylamide in 0.335 M sodium phosphate, pH 7.1/0.1% NaDodSO4. Stacking gels were prepared with 5% acryl-amide/0. 13% bisacrylamide in 0.063 M sodium phosphate, pH7.1/0.1% NaDodSO4. Polymerization was started by adding0.05% ammonium persulfate/0. 1% N,N,N',N'-tetramethyl-ethylenediamine. Electrophoresis was carried out with 0.1 Msodium phosphate, pH 7.1/0.5% NaDodSO4 at 40 mA per gel(75 x 130 X 1.5 mm) for 20 hr. Amounts of protein were mea-sured by densitometric tracing of autoradiograms or extractingthe radioactivity from the bands. The prolipoprotein containsthree methionine residues as opposed to two in the lipoprotein;thus, an appropriate correction was made to estimate the finalyield.

RESULTSIsolation and Identification of the Mutants. The amino acid

1

sequence of the signal peptide of the prolipoprotein is Met-Lys-

5 10 15Ala-Thr-Lys-Leu -Val-Leu -Gly-Ala-Val- Ile -Leu-Gly-Ser-Thr-

20Leu-Leu-Ala-Gly (2). According to the loop model for the mech-anism ofprotein secretion across the membrane, the +2 chargefrom the two lysine residues in the amino-terminal region playsan important role at an initial step of protein secretion. To ex-amine the effects of the charge of this section, guided site-spe-cific mutagenesis was carried out using the synthetic oligonu-cleotides shown in Fig. 1. Oligonucleotide 1, a 16-mer, has amismatch in the codon for alanine-3. This mismatch alters ala-nine to aspartic acid, resulting in a change in the net charge ofthis section from +2 to + 1. Oligonucleotide 2, a 19-mer, hasa three-base deletion at the codon for lysine-2, thus changingthe net charge from +2 to + 1. Oligonucleotide 3, a 16-mer, hasthe mutations carried by both oligonucleotides 1 and 2, result-ing in a change in the net charge from +2 to 0. Oligonucleotide4, a 15-mer, has mismatches in the codons for lysine-2 and ala-nine-3. These mismatches cause alterations of lysine to glutamicacid and alanine to aspartic acid, resulting in a change in thenet charge from +2 to -1.

These oligonucleotides were used for guided site-specificmutagenesis as described in Materials and Methods. Detectionof the desired mutants was carried out by colony hybridizationin which individual 32P-labeled oligonucleotides were used asprobes. Yields of positive colonies were 1.5%, 1.0%, 0.3%, and0.4% of the total number of transformants for experiments witholigonucleotides 1, 2, 3, and 4, respectively. Positive colonieswere isolated from a second transformation in which DNA ex-tracted from the initial positive colonies was used to removewild-type segregants. All of the secondary transformants werepositive with the exception of that from oligonucleotide 1, inwhich half of the transformants were negative.

Final identification of the mutants was carried out by DNAsequence analysis (Fig. 2). Sequences ofthe mutant DNAs wereidentical to those predicted from the mutagenesis.

Production of the Lipoprotein. JA221 lpp-/F'laciq cells car-rying pKEN125 or its mutant derivatives (I-1, 1-2, 1-3, and 1-4for mutants isolated by using oligonucleotides 1, 2, 3, and 4,respectively) do not produce lipoprotein in the absence of iPr-S-Gal, a lac inducer, because a DNA fragment containing thelac promoter-operator region has been inserted in the lpp gene

WILD TYPE

CHARGE

:+2

:+1

:+1

: 0

:-1

Biochemistry: Inouye et aL

3440 Biochemistry: Inouye et aL

b

G A C T G A C T G A C T4FF -'~_f-a

10 AT

5W ' T _ AT1

iweP

d

G A C T

_ok TA

_ ...

_A IfT

4

e

G A C T

,-N

_AO T

^_AlG

CTTC-CA.T

a

iTAACICTACTAAA ATGAAAGATACTAAA ATGGCTACTAAA ATGGATACTAAA ATGGAAGAIACTAAA

FIG. 2. DNA sequence analysis gels corresponding to the region of the translation initiation site of the ipp genes from pKEN125 (a) and mutantsisolated by using oligonucleotides 1 (b), 2 (c), 3 (d), and 4 (e). The plasmid DNAs were digested with Xba I (New England BioLabs). The unique XbaI site in the ribosome binding site of the lpp gene (6) was then labeled with [32P]dNTP by using the Klenow fragment of DNA polymerase I (Boeh-ringer Mannheim). The labeled DNA was then cleaved with HinfI and the -120-base pair Xba I/HinfI fragment was purified and analyzed ac-cording to Maxam and Gilbert (11). The cleavage products were applied to a 20% polyacrylamide gel and subjected to electrophoresis. Only thesequences from the initiation codon to the codon for the lysine residue at position 5 are shown. The sequences at the sides of the gels are compli-mentary to the sense sequences. The sense sequences deduced from the gels are shown below the gels. Small arrows indicate positions of the firstbase of the initiation codon. Small dots indicate bases that are different from the sequence of pKEN125. Large open arrows indicate positions ofa 3-base pair deletion.

(5). Therefore, to examine the effects of mutations in the signalpeptide on the lipoprotein assembly, the Ipp gene was inducedby iPr-S-Gal and, 20 min later, cells were pulse labeled for 10sec with [3S]methionine. Total radioactivities recovered in theimmunoprecipitates with antilipoprotein serum were 76%,75%, 48%, and 45% of that of pKEN125 for I-1, 1-2, 1-3, and1-4, respectively (data not shown). This indicates that the mu-tations affect the overall rate of lipoprotein production (lipo-protein plus prolipoprotein).To further examine the effects of mutation on lipoprotein

assembly, the membrane and the soluble fractions were sepa-rated and treated with antilipoprotein serum. The immunopre-cipitates were analyzed by NaDodSO4/polyacrylamide gel elec-trophoresis. It is known that there are two types of the prolip-oprotein; unmodified and modified by glycerol at the cysteine-21 residue. The modified form can be separated from the ma-ture lipoprotein by a NaDodSO4 gel system using Tris HC1 buf-fer. We found that there was no accumulation of modified pro-lipoprotein in any case (data not shown). However, when ana-lyzed with a NaDodSO4 gel system using phosphate buffer, the

A B1 2 3 4 5 1 2 3 4 5

PLP.-LPP -MbftAho- ..

unmodified prolipoprotein was detected in all cases (Fig. 3).The slower moving band was identified as unmodified proli-poprotein because it formed a dimer in the absence of 2-mer-captoethanol due to the unmodified cysteine-21 residue (datanot shown).

Fig. 3A shows that the amount of mature lipoprotein in themembrane fraction progressively decreased as the net chargeon the amino-terminal region became more negative: Fractionsfrom I-1, 1-2; 1-3, and 1-4 had 77%, 66%, 40%, and 7%, re-spectively, of the amount of mature lipoprotein in pKEN125.The amounts of unmodified prolipoprotein, however, were ataratherhigh level in all cases; 12%, 12%, 7.2%, 5.2%, and 8.0%ofthe amount ofmature lipoprotein ofpKEN125 for pKEN125,I-1, 1-2, I-3, and 1-4, respectively.

In contrast to the membrane fraction, the amount of un-modified prolipoprotein in the soluble fraction was dramaticallyhigher in the case of 1-4 (Fig. 3B, lane 5). The amounts of pro-lipoprotein in the soluble fraction were 90%, 113%, 96%, and338% of that ofpKEN125 for I-1, 1-2, 1-3, and 1-4, respectively.The prolipoprotein of 1-4 remained in the supernatant evenafter prolonged centrifugation (for 4 hr instead of 30 min at 105x g), while almost all of the lipoprotein was eliminated (data

1 2 3 4

PLPE-LPPm

FIG. 3. Autoradiogram of NaDodSO4/polyacrylamide gel analysisof immunoprecipitates from a pulse experiment. Cells were inducedwith iPr-S-Gal and pulse labeled with [35S]methionine for 10 sec.Immunoprecipitates with antilipoprotein serum from the membranefractions (A) and the soluble fractions (B) were subjected to NaDodSO4/polyacrylamide gel electrophoresis. Lanes: 1, pKEN125; 2, I-1; 3, 1-2,4, I-3;,5, 1-4. PLP, unmodified prolipoprotein; LPP, mature lipoprotein.

h

FIG. 4. Autoradiogram of NaDodSO4/polyacrylamide gel analysisof immunoprecipitates from a pulse-chase experiment. Lanes: 1, pulseexperiment with membrane fraction; 2, pulse experiment with solublefraction; 3, 2-min chase experiment with membrane fraction; 4, 2-minchase experiment with soluble fraction. Only a part of the gel is shown.PLP, unmodified prolipoprotein; LPP, mature lipoprotein.

Proc. Natl. Acad. Sci. USA 79 (1982)

Opop

qb4b aft 40

4WD 4*4400 %.ml

Proc. Natl. Acad. Sci. USA 79 (1982) 3441

not shown). Furthermore, only a smtll amount of the solubleprolipoprotein (15%) was found in the periplasmic fraction,which was prepared by osmotic shock without using EDTA (12)(data not shown). These results indicate that the prolipoproteinof 1-4 accumulated as a soluble protein in the cytoplasm.

Pulse-Chase Experiment. To determine whether a largeamount of the unmodified prolipoprotein found in the solublefraction of 1-4 could be secreted across the membrane and pro-cessed to mature lipoprotein in the outer membrane, a pulse-chase experiment was carried out. As shown in Fig. 4, theamount of membrane lipoprotein of I-4 increased substantiallyduring the 2-min chase (352%; Fig. 4, lanes 1 and 3). This in-crease in membrane lipoprotein of 1-4 can approximately ac-count for the decrease in soluble prolipoprotein during thechase (Fig. 4, lanes 2 and 4). The total amount of immunopre-cipitable radioactivity in the cells did not change appreciablyduring the chase.

DISCUSSIONWe have shown that the positively charged amino-terminal re-gion ofthe prolipoprotein plays an important role during an ini-tial step of protein secretion across the membrane. Assumingthat a formyl group is attached at the amino-terminal methio-nine (Fig. 1), there is a +2 charge on the amino-terminal regionof the prolipoprotein due to lysine residues at positions 2 and5. We kept the lysine residue at position 5 intact in all mutantsso that the length of the hydrophobic section or the distancebetween the charged amino-terminal region and the cleavagesite was unchanged.

As we changed the net charge on the amino-terminal regionfrom +2 to + 1, 0, and -1, we noted two distinct effects in themutations. First, the total lipoprotein production due to theinducible lpp gene (prolipoprotein plus lipoprotein in the wholecells) or the rate of lipoprotein production appears to decrease;the cases of + 1 charge (I-1 and 1-2) show a lipoprotein produc-tion that is 70-80% that of pKEN125 and those of 0 and -1charge (1-3 and 1-4) show productions 30-45% of pKEN125.These mutation effects on lipoprotein production cannot be at-tributed to the formation of a more stable secondary structurein this region of the lpp mRNA, which would inhibit the rateof protein synthesis as in the case of the lamB gene (13). Thus,lower lipoprotein production in the mutants may be due to anincreased susceptibility of mutant prolipoproteins to proteasedigestion. However, this is unlikely due to the observed sta-bility of the 1-4 prolipoprotein.

The second effect of the mutations is to inhibit the rate oftranslocation of the mutant prolipoprotein across the mem-

brane, especially in the case of 1-4. The change ofthe net chargeon the amino-terminal region from 0 (1-3) to -1 (1-4) had littleeffect on the overall rate of the lipoprotein production (prolip-oprotein plus lipoprotein). However, this change severely re-duced the rate of translocation of the prolipoprotein across themembrane, resulting in accumulation of the glycerol-unmodi-fied prolipoprotein in the cytoplasm. The soluble form of theunmodified prolipoprotein appears to be stable and to be grad-ually and posttranslationally translocated across the membraneand processed to the mature lipoprotein; the final yield of ma-ture lipoprotein of I-4 was nearly identical to that of 1-3 in along-term labeling experiment.

These results are consistent with the loop model (1, 3): Neg-ative charge on the amino-terminal region of the signal peptidehampers the initial interaction of the signal peptide with thenegatively charged inner surface ofthe cytoplasmic membrane,resulting in accumulation of secretory precursor molecules inthe cytoplasmic fraction. However, alternative explanations ofour results are possible. Further mutagenesis experimentsshould give us more detailed information concerning the func-tion of this region.

This work was supported by National Institute of General MedicalSciences Grants GM 19043 (to M.I.) and GM 30395 (to K.I.) andAmerican Cancer Society Grant BC-67D (to M.I.).

1. Inouye, M. & Halegoua, S. (1980) CRC Crit. Rev. Biochem. 7,339-371.

2. Inouye, S., Wang, S., Sekizawa, J., Halegoua, S. & Inouye, M.(1977) Proc. Natl. Acad. Sci. USA 74, 1004-1008.

3. Halegoua, S. & Inouye, M. (1979) in Bacterial Outer Mem-branes: Biogenesis and Functions, ed. Inouve, M. (Wiley, NewYork), pp. 67-113.

4. Blobel, G. & Dobberstein, B. (1975)J. Cell Biol. 67, 835-851.5. Nakamura, K. & Inouye, M. (1982) J. Mol. Applied Genet., in

press.6. Nakamura, K. & Inouye, M. (1979) Cell 18, 1109-1117.7. Wallace, R. B., Johnson, P. F., Tanaka, S., Schold, M., Itakura,

K. & Abelson, J. (1980) Science 209, 1396-1400.8. Miyoshi, K., Arentzen, R., Huang, T. & Itakura, K. (1980) Nu-

cle* Acids Res. 8, 5507-5517.9. Inouye, M., Shaw, J. & Shen, C. (1972) J. Biol. Chem. 247,

8154-8159.10. Inouye, S., Takeishi, K., Lee, N., DeMartini, M., Hirashima, A.

& Inouye, M. (1976)J. Bacteriol. 127, 555-563.11. Maxam, A. M. & Gilbert, W. (1977) Proc. Natl. Acad. Sci. USA

74, 560-564.12. Neu, H. C. & Heppel, L. A. (1965) J. Biol Chem. 240,

3685-3692.13. Schwartz, 'M., Roa, M. & Debarbouille, M. (1981) Proc. Natl.

Acad. Sci. USA 78, 2937-2941.

Biochemistry: Inouye et al.