Portal de Periódicos da CAPES

Location of the Actual Signal in the Negatively Charged Leader Sequence Involved in the Import into the Mitochondrial Matrix Space

2003; Elsevier BV; Volume: 278; Issue: 16 Linguagem: Inglês

10.1074/jbc.m212743200

1083-351X

Abhijit Mukhopadhyay, Thomas S. Heard, Xiaohui Wen, Philip K. Hammen, Henry Weiner,

Ion channel regulation and function

Proteins destined for the mitochondrial matrix space have leader sequences that are typically present at the most N-terminal end of the nuclear-encoded precursor protein. The leaders are rich in positive charges and usually deficient of negative charges. This observation led to the acid-chain hypothesis to explain how the leader sequences interact with negatively charged receptor proteins. Here we show using both chimeric leaders and one from isopropyl malate synthase that possesses a negative charge that the leader need not be at the very N terminus of the precursor. Experiments were performed with modified non-functioning leader sequences fused to either the native or a non-functioning leader of aldehyde dehydrogenase so that an internal leader sequence could exist. The internal leader is sufficient for the import of the modified precursor protein. It appears that this leader still needs to form an amphipathic helix just like the normal N-terminal leaders do. This internal leader could function even if the most N-terminal portion contained negative charges in the first 7–11 residues. If the first 11 residues were deleted from isopropyl malate synthase, the resulting protein was imported more successfully than the native protein. It appears that precursors that carry negatively charged leaders use an internal signal sequence to compensate for the non-functional segment at the most N-terminal portion of the protein. Proteins destined for the mitochondrial matrix space have leader sequences that are typically present at the most N-terminal end of the nuclear-encoded precursor protein. The leaders are rich in positive charges and usually deficient of negative charges. This observation led to the acid-chain hypothesis to explain how the leader sequences interact with negatively charged receptor proteins. Here we show using both chimeric leaders and one from isopropyl malate synthase that possesses a negative charge that the leader need not be at the very N terminus of the precursor. Experiments were performed with modified non-functioning leader sequences fused to either the native or a non-functioning leader of aldehyde dehydrogenase so that an internal leader sequence could exist. The internal leader is sufficient for the import of the modified precursor protein. It appears that this leader still needs to form an amphipathic helix just like the normal N-terminal leaders do. This internal leader could function even if the most N-terminal portion contained negative charges in the first 7–11 residues. If the first 11 residues were deleted from isopropyl malate synthase, the resulting protein was imported more successfully than the native protein. It appears that precursors that carry negatively charged leaders use an internal signal sequence to compensate for the non-functional segment at the most N-terminal portion of the protein. Most proteins found in the mitochondrial matrix space are encoded by nuclear genes and are synthesized on cytosolic polysomes. The preproteins typically carry N-terminal targeting sequences (leader sequence) that direct the proteins to the translocase of the outer mitochondrial membrane (TOM), 1The abbreviations used are: TOMtranslocase of the mitochondrial outer membranepALDHprecursor aldehyde dehydrogenaseIPMSisopropylmalate synthaseMPPmitochondrial processing peptidaseTFEtrifluoroethanollong formpreprotein containing the leader plus the mature part of IPMSshort formIPMS containing only the mature portion of IPMSIPMS4(1–21)a synthetic peptide corresponding to the N-terminal 21 residues of Leu4NOESYNuclear Overhauser effect spectroscopyTOCSYtotal correlation spectroscopy1The abbreviations used are: TOMtranslocase of the mitochondrial outer membranepALDHprecursor aldehyde dehydrogenaseIPMSisopropylmalate synthaseMPPmitochondrial processing peptidaseTFEtrifluoroethanollong formpreprotein containing the leader plus the mature part of IPMSshort formIPMS containing only the mature portion of IPMSIPMS4(1–21)a synthetic peptide corresponding to the N-terminal 21 residues of Leu4NOESYNuclear Overhauser effect spectroscopyTOCSYtotal correlation spectroscopy which consists of receptors and a general import pore (1Moczko M. Dietmeier K. Sollner T. Segui B. Steger H.F. Neupert W. Pfanner N. FEBS Lett. 1992; 310: 265-268Crossref PubMed Scopus (55) Google Scholar, 2Moczko M. Gartner F. Pfanner N. FEBS Lett. 1993; 326: 251-254Crossref PubMed Scopus (46) Google Scholar, 3Neupert W. Annu. Rev. Biochem. 1997; 66: 863-917Crossref PubMed Scopus (974) Google Scholar, 4Pfanner N. Craig E.A. Honlinger A. Annu. Rev. Cell Dev. Biol. 1997; 13: 25-51Crossref PubMed Scopus (146) Google Scholar, 5Endo T. Kohda D. Biochim. Biophys. Acta. 2002; 1592: 3-14Crossref PubMed Scopus (107) Google Scholar). A common feature of leader sequences is the frequent occurrence of basic residues, particularly arginine, a residue that is found three times more often in the leaders than proteins (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). Because TOM proteins are negatively charged, an acid-chain hypothesis was developed to suggest that positive charges in the leader sequences are necessary for TOM to interact with negatively charged receptor proteins (7Schatz G. Nature. 1997; 388: 121-122Crossref PubMed Scopus (88) Google Scholar).The leader sequence for pALDH has been studied extensively in our laboratory. As shown in Fig. 1, the leader is composed of 19 amino acids that can be induced into a helix-linker-helix amphiphilic structure in the presence of trifluoroethanol or detergent micelles (8Karslake C. Piotto M.E. Pak Y.K. Weiner H. Gorenstein D.G. Biochemistry. 1990; 29: 9872-9878Crossref PubMed Scopus (85) Google Scholar). Although each helical segment contains an equal number of positive charges, the N-terminal helical segment has been shown to be necessary to efficiently target the leader to the matrix space (9Wang Y. Weiner H. J. Biol. Chem. 1993; 268: 4759-4765Abstract Full Text PDF PubMed Google Scholar). If both of the arginines of the N-helical segment are substituted with glutamine, the ability to be imported is essentially eliminated, showing that the net positive charge in the N-helical segment is more important than the total positive charge throughout the leader (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). Deletion of the linker results in a non-processable leader that is longer in helical content (10Thornton K. Wang Y. Weiner H. Gorenstein D.G. J. Biol. Chem. 1993; 268: 19906-19914Abstract Full Text PDF PubMed Google Scholar) and is more capable of importing ALDH than is the native sequence in a cell-free import assay (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). The enhanced stability of the linker-deleted leader has been used successfully to study structural aspects as a compensating factor for the loss of positive charges resulting from arginine to glutamine mutational substitutions. The linker-deleted structure can import a "passenger" protein even if both of the N-terminal arginine residues (Arg-3 and Arg-10) are mutated to glutamines (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). On the basis of the positive charge versus structural compensation model developed from our previous studies, glutamic and aspartic acid residues were systematically introduced into the pALDH leader to ascertain how negative charges may be tolerated. It was observed that when serine was replaced with glutamic acid, the S7E mutant pALDH was imported to an extent similar to pALDH; however, the R3Q,S7E double mutant was poorly imported (11Heard T.S. Weiner H. J. Biol. Chem. 1998; 273: 29389-29393Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). This result implies that only a net positive charge was required for the proper import of a precursor into the matrix. Although negatively charged amino acids typically are not found in leader sequences, chaperonin 10 (12Hartman D.J. Hoogenraad N.J. Condron R. Hoj P.B. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 3394-3398Crossref PubMed Scopus (104) Google Scholar) and rhodanese (13Miller D.M. Delgado R. Chirgwin J.M. Hardies S.C. Horowitz P.M. J. Biol. Chem. 1991; 266: 4686-4691Abstract Full Text PDF PubMed Google Scholar) each contain one negative charge. Different forms of yeast IPMS (coded byLeu4 and Leu9) contain one and two negative charges, respectively, in their leader sequences (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 15Casalone E. Barberio C. Cavalieri D. Polsinelli M. Yeast. 2000; 16: 539-545Crossref PubMed Scopus (21) Google Scholar, 16Voss H. Benes V. Andrade M.A. Valencia A. Rechmann S. Teodoru C. Schwager C. Paces V. Yeast. 1997; 13: 655-672Crossref PubMed Scopus (17) Google Scholar). Mitochondrial IPMS provided us with a good model to study the mechanism of import of a natural protein with negative charges within the mitochondrial leader sequence.In this study, we used several chimeric and mutant forms of pALDH and IPMS containing neutral or negative charges in their leader sequences. We will show that some precursors could be imported to the mitochondrial matrix with a leader sequence that is not present at the most N-terminal portion.DISCUSSIONDuring the past decade much has been learned about the leader sequences of precursor proteins that are destined for the mitochondrial matrix space (3Neupert W. Annu. Rev. Biochem. 1997; 66: 863-917Crossref PubMed Scopus (974) Google Scholar, 27von Heijne G. EMBO J. 1986; 5: 1335-1342Crossref PubMed Scopus (705) Google Scholar). No reports exist, however, to explain how a leader that does not possess a net positive charge at its N terminus is imported. During our studies with the positive charge in the ALDH leader, we often found that if the leader was made more helical it would compensate for the removal of positive charges (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). No actual structure for a precursor protein that possesses a cleavable leader sequence has been determined. NMR has been used to show that a number of peptides corresponding to leader sequences can indeed form α helices when bound to micelles or when in a hydrophobic environment (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar,8Karslake C. Piotto M.E. Pak Y.K. Weiner H. Gorenstein D.G. Biochemistry. 1990; 29: 9872-9878Crossref PubMed Scopus (85) Google Scholar, 10Thornton K. Wang Y. Weiner H. Gorenstein D.G. J. Biol. Chem. 1993; 268: 19906-19914Abstract Full Text PDF PubMed Google Scholar, 28Hammen P.K. Gorenstein D.G. Weiner H. Biochemistry. 1996; 35: 3772-3781Crossref PubMed Scopus (38) Google Scholar). Recently the structure of the aldehyde dehydrogenase leader bound to TOM20 was shown to be helical, supporting the notion that helicity is indispensable for its function (29Muto T. Obita T. Abe Y. Shodai T. Endo T. Kohda D. J. Mol. Biol. 2001; 306: 137-143Crossref PubMed Scopus (82) Google Scholar).A few precursor proteins do not possess positive charges in what could be considered to be the classical N-terminal leader sequence. For example, the non-processed precursors rhodanese (30Waltner M. Weiner H. J. Biol. Chem. 1995; 270: 26311-26317Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar) and chaperonin10 (31Ryan M.T. Hoogenraad N.J. Hoj P.B. FEBS Lett. 1994; 337: 152-156Crossref PubMed Scopus (36) Google Scholar, 32Jarvis J.A. Ryan M.T. Hoogenraad N.J. Craik D.J. Hoj P.B. J. Biol. Chem. 1995; 270: 1323-1331Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar) have no positive charges in the first eight residues and actually have a negative charge in the leader. Leu4 leader has one lysine residue but two negative charges in its first 11 residues. These leader peptides were all found to be helical, consistent with our notion that structure can somehow overcome positive charge deficiency. Even if structure could overcome this situation, it is not apparent how the TOM complex, which contains patches of negative charges (7Schatz G. Nature. 1997; 388: 121-122Crossref PubMed Scopus (88) Google Scholar), can recognize the atypical leader.We took advantage of the two-leader strategy that we used previously to investigate the processing of leader peptides by MPP (18Mukhopadhyay A. Hammen P. Waltner-Law M. Weiner H. Protein Sci. 2002; 11: 1026-1035Crossref PubMed Scopus (8) Google Scholar). Here a modified non-functional leader was fused to the N termini of a native or modified precursor protein. In this way, it was possible to ascertain what part of the leader was actually needed by the import apparatus.Previously, it was shown that the R3Q,R10Q double mutant pALDH was not imported (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). Here we report that if this modified leader was fused to the R3Q,R10Q pALDH, import was restored. This shows that two non-functional leaders could actually be used to import a protein. The reason this double leader was functional in import could be that the internal leader-like structures shown in Fig. 1 actually were the portion of the precursor that was involved in the translocation process. No such structural motif existed in the original R3Q,R10Q double mutant pALDH we previously investigated. From the various chimeric precursors used, it is concluded that an internal helix that is either a long continuous amphipathic helix or part of the helix-linker-helix motif could act as the actual leader. No attempt was made to determine how distal from the N terminus such a structure would function.Using the strategy described above, it was possible to show that the most N-terminal, non-functional segment could even possess a negative charge such as found in the R3Q,S7E double mutant of pALDH. This finding can then be used to explain how a natural precursor protein could possess a net negative charge at the start of the N-terminal sequence. It can be argued that the import apparatus simply ignores the unfavorable region and interacts with the positively charged stable helical domain that follows.To test for the above mentioned model we used Leu4 and Leu9, natural yeast proteins that possess negatively charged amino acids in the most N-terminal segment of their leader sequence. NMR spectroscopy was used to show that, as expected, the peptide corresponding to the leader of Leu4 formed a stable α helix. Before the yeast genome was obtained, only one form of the enzyme was known. The second gene product, Leu9 (which has 80% identity with Leu4), was identified later, but the protein was not characterized (15Casalone E. Barberio C. Cavalieri D. Polsinelli M. Yeast. 2000; 16: 539-545Crossref PubMed Scopus (21) Google Scholar). First, in vitro import was used to show that Leu9, like Leu4, is a mitochondrial protein. Unexpectedly, the Leu4 precursor was not imported under conditions that allowed the Leu9 to be imported. A chimeric protein was made by fusing the purported leader segment of Leu4 or Leu9 to mature ALDH. When the first 21 residues of Leu4 (those found to be helical) were used, no import was found. But if the first 31 residues were used, import could occur. In a similar manner, the first 30 residues of Leu9 could import of mature ALDH, whereas the first 20 could not.We were concerned with the fact that we could not obtain import with Leu4 as others had reported (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 24Hampsey D.M. Lewin A.S. Kohlhaw G.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 1270-1274Crossref PubMed Scopus (16) Google Scholar). The previously published experimental procedure used total RNA for translation (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 24Hampsey D.M. Lewin A.S. Kohlhaw G.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 1270-1274Crossref PubMed Scopus (16) Google Scholar) and antibodies (33Gasser S.M. Daum G. Schatz G. J. Biol. Chem. 1982; 257: 13034-13041Abstract Full Text PDF PubMed Google Scholar) to precipitate imported proteins. In contrast, the current experiments all used pure cDNAs. Given the fact that theLeu9 gene was not identified at the time the initial experiments were performed and that total RNA was used, it is possible that the Leu9 gene product was detected rather than the product for Leu4. We conclude that under the in vitro experimental procedures used in this study, Leu4 was not imported to the mitochondrial matrix space.The N-terminal portion of Leu4 and Leu9 provided a 21- or 20-residue helix thought to provide the essential structure for a region that does not possess a net positive charge. Our data revealed that the N-terminal region containing the negative charges formed the most stable part of the helix. However, the region that seemed most responsible for import of Leu4 or Leu9 was not the most helical portion of the leader, because the targeting signal seemed to appear after residue 11. It is generally accepted that before import to the mitochondria preproteins remain unfolded. A proteolysis assay was performed with the TNT-synthesized Leu4. Proteinase K proteolysis results indicated that the Leu4 was more resistant to digestion than was the Leu4(Δ1–11) protein (data not shown). A possible explanation for this phenomenon is that the most N-terminal 11 residues of Leu4 contribute to the overall structure of the protein. Without these residues the native folding of these proteins might be changed, as reflected by the instability of the protein and the proteinase K sensitivity. Therefore, the proteolysis results showed that Leu4 was folded and hence not imported to the mitochondria in an in vitro experiment. Some proteins that fold rapidly have been found to be imported to the mitochondria by a co-translocation pathway. The pALDH leader fused to green fluorescent protein (a protein that folds rapidly) was shown to be imported to the mitochondrial matrix space by a co-translational pathway (34Ni L. Heard T.S. Weiner H. J. Biol. Chem. 1999; 274: 12685-12691Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). Similarly, in vivo Leu4 could be translocated to the mitochondria through a co-translational pathway.The acid-chain hypothesis was based on the fact that the leader possessed positive charges at its N terminus, which could be important for binding with the TOM complex. However, it has been shown that the leader sequence of pALDH binds TOM20 through hydrophobic interactions (29Muto T. Obita T. Abe Y. Shodai T. Endo T. Kohda D. J. Mol. Biol. 2001; 306: 137-143Crossref PubMed Scopus (82) Google Scholar). TOM20 is the first receptor to be recognized by leader sequence and does not appear to depend on the positive charges of the leader sequence. It is possible that the other TOM protein may interact with positive charges present in leader sequence. Here we show that if the N-terminal segment is not positive, the translocator appears to recognize not the N termini but the region of the leader sequences that are positively charged. It has been shown that inner and intermembrane space proteins use an internal signal after the most N-terminal matrix space targeting signal (35Arnold I. Folsch H. Neupert W. Stuart R.A. J. Biol. Chem. 1998; 273: 1469-1476Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar, 36Bomer U. Meijer M. Guiard B. Dietmeier K. Pfanner N. Rassow J. J. Biol. Chem. 1997; 272: 30439-30446Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). How the first 11 residues in the case of Leu4 enter the translocation pore cannot be explained at this time. If the model of import were co-translational and not post-translational as we suggest, then it is possible that an unknown cytosolic factor is necessary to insert the proteins into the TOM apparatus. Although the data presented in this study cannot explain how the precursor with negative charges interacts with the TOM proteins, they do show that the information for import need not lie at the very N terminus of the precursor protein. Most proteins found in the mitochondrial matrix space are encoded by nuclear genes and are synthesized on cytosolic polysomes. The preproteins typically carry N-terminal targeting sequences (leader sequence) that direct the proteins to the translocase of the outer mitochondrial membrane (TOM), 1The abbreviations used are: TOMtranslocase of the mitochondrial outer membranepALDHprecursor aldehyde dehydrogenaseIPMSisopropylmalate synthaseMPPmitochondrial processing peptidaseTFEtrifluoroethanollong formpreprotein containing the leader plus the mature part of IPMSshort formIPMS containing only the mature portion of IPMSIPMS4(1–21)a synthetic peptide corresponding to the N-terminal 21 residues of Leu4NOESYNuclear Overhauser effect spectroscopyTOCSYtotal correlation spectroscopy1The abbreviations used are: TOMtranslocase of the mitochondrial outer membranepALDHprecursor aldehyde dehydrogenaseIPMSisopropylmalate synthaseMPPmitochondrial processing peptidaseTFEtrifluoroethanollong formpreprotein containing the leader plus the mature part of IPMSshort formIPMS containing only the mature portion of IPMSIPMS4(1–21)a synthetic peptide corresponding to the N-terminal 21 residues of Leu4NOESYNuclear Overhauser effect spectroscopyTOCSYtotal correlation spectroscopy which consists of receptors and a general import pore (1Moczko M. Dietmeier K. Sollner T. Segui B. Steger H.F. Neupert W. Pfanner N. FEBS Lett. 1992; 310: 265-268Crossref PubMed Scopus (55) Google Scholar, 2Moczko M. Gartner F. Pfanner N. FEBS Lett. 1993; 326: 251-254Crossref PubMed Scopus (46) Google Scholar, 3Neupert W. Annu. Rev. Biochem. 1997; 66: 863-917Crossref PubMed Scopus (974) Google Scholar, 4Pfanner N. Craig E.A. Honlinger A. Annu. Rev. Cell Dev. Biol. 1997; 13: 25-51Crossref PubMed Scopus (146) Google Scholar, 5Endo T. Kohda D. Biochim. Biophys. Acta. 2002; 1592: 3-14Crossref PubMed Scopus (107) Google Scholar). A common feature of leader sequences is the frequent occurrence of basic residues, particularly arginine, a residue that is found three times more often in the leaders than proteins (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). Because TOM proteins are negatively charged, an acid-chain hypothesis was developed to suggest that positive charges in the leader sequences are necessary for TOM to interact with negatively charged receptor proteins (7Schatz G. Nature. 1997; 388: 121-122Crossref PubMed Scopus (88) Google Scholar). translocase of the mitochondrial outer membrane precursor aldehyde dehydrogenase isopropylmalate synthase mitochondrial processing peptidase trifluoroethanol preprotein containing the leader plus the mature part of IPMS IPMS containing only the mature portion of IPMS a synthetic peptide corresponding to the N-terminal 21 residues of Leu4 Nuclear Overhauser effect spectroscopy total correlation spectroscopy translocase of the mitochondrial outer membrane precursor aldehyde dehydrogenase isopropylmalate synthase mitochondrial processing peptidase trifluoroethanol preprotein containing the leader plus the mature part of IPMS IPMS containing only the mature portion of IPMS a synthetic peptide corresponding to the N-terminal 21 residues of Leu4 Nuclear Overhauser effect spectroscopy total correlation spectroscopy The leader sequence for pALDH has been studied extensively in our laboratory. As shown in Fig. 1, the leader is composed of 19 amino acids that can be induced into a helix-linker-helix amphiphilic structure in the presence of trifluoroethanol or detergent micelles (8Karslake C. Piotto M.E. Pak Y.K. Weiner H. Gorenstein D.G. Biochemistry. 1990; 29: 9872-9878Crossref PubMed Scopus (85) Google Scholar). Although each helical segment contains an equal number of positive charges, the N-terminal helical segment has been shown to be necessary to efficiently target the leader to the matrix space (9Wang Y. Weiner H. J. Biol. Chem. 1993; 268: 4759-4765Abstract Full Text PDF PubMed Google Scholar). If both of the arginines of the N-helical segment are substituted with glutamine, the ability to be imported is essentially eliminated, showing that the net positive charge in the N-helical segment is more important than the total positive charge throughout the leader (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). Deletion of the linker results in a non-processable leader that is longer in helical content (10Thornton K. Wang Y. Weiner H. Gorenstein D.G. J. Biol. Chem. 1993; 268: 19906-19914Abstract Full Text PDF PubMed Google Scholar) and is more capable of importing ALDH than is the native sequence in a cell-free import assay (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). The enhanced stability of the linker-deleted leader has been used successfully to study structural aspects as a compensating factor for the loss of positive charges resulting from arginine to glutamine mutational substitutions. The linker-deleted structure can import a "passenger" protein even if both of the N-terminal arginine residues (Arg-3 and Arg-10) are mutated to glutamines (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). On the basis of the positive charge versus structural compensation model developed from our previous studies, glutamic and aspartic acid residues were systematically introduced into the pALDH leader to ascertain how negative charges may be tolerated. It was observed that when serine was replaced with glutamic acid, the S7E mutant pALDH was imported to an extent similar to pALDH; however, the R3Q,S7E double mutant was poorly imported (11Heard T.S. Weiner H. J. Biol. Chem. 1998; 273: 29389-29393Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). This result implies that only a net positive charge was required for the proper import of a precursor into the matrix. Although negatively charged amino acids typically are not found in leader sequences, chaperonin 10 (12Hartman D.J. Hoogenraad N.J. Condron R. Hoj P.B. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 3394-3398Crossref PubMed Scopus (104) Google Scholar) and rhodanese (13Miller D.M. Delgado R. Chirgwin J.M. Hardies S.C. Horowitz P.M. J. Biol. Chem. 1991; 266: 4686-4691Abstract Full Text PDF PubMed Google Scholar) each contain one negative charge. Different forms of yeast IPMS (coded byLeu4 and Leu9) contain one and two negative charges, respectively, in their leader sequences (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 15Casalone E. Barberio C. Cavalieri D. Polsinelli M. Yeast. 2000; 16: 539-545Crossref PubMed Scopus (21) Google Scholar, 16Voss H. Benes V. Andrade M.A. Valencia A. Rechmann S. Teodoru C. Schwager C. Paces V. Yeast. 1997; 13: 655-672Crossref PubMed Scopus (17) Google Scholar). Mitochondrial IPMS provided us with a good model to study the mechanism of import of a natural protein with negative charges within the mitochondrial leader sequence. In this study, we used several chimeric and mutant forms of pALDH and IPMS containing neutral or negative charges in their leader sequences. We will show that some precursors could be imported to the mitochondrial matrix with a leader sequence that is not present at the most N-terminal portion. DISCUSSIONDuring the past decade much has been learned about the leader sequences of precursor proteins that are destined for the mitochondrial matrix space (3Neupert W. Annu. Rev. Biochem. 1997; 66: 863-917Crossref PubMed Scopus (974) Google Scholar, 27von Heijne G. EMBO J. 1986; 5: 1335-1342Crossref PubMed Scopus (705) Google Scholar). No reports exist, however, to explain how a leader that does not possess a net positive charge at its N terminus is imported. During our studies with the positive charge in the ALDH leader, we often found that if the leader was made more helical it would compensate for the removal of positive charges (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). No actual structure for a precursor protein that possesses a cleavable leader sequence has been determined. NMR has been used to show that a number of peptides corresponding to leader sequences can indeed form α helices when bound to micelles or when in a hydrophobic environment (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar,8Karslake C. Piotto M.E. Pak Y.K. Weiner H. Gorenstein D.G. Biochemistry. 1990; 29: 9872-9878Crossref PubMed Scopus (85) Google Scholar, 10Thornton K. Wang Y. Weiner H. Gorenstein D.G. J. Biol. Chem. 1993; 268: 19906-19914Abstract Full Text PDF PubMed Google Scholar, 28Hammen P.K. Gorenstein D.G. Weiner H. Biochemistry. 1996; 35: 3772-3781Crossref PubMed Scopus (38) Google Scholar). Recently the structure of the aldehyde dehydrogenase leader bound to TOM20 was shown to be helical, supporting the notion that helicity is indispensable for its function (29Muto T. Obita T. Abe Y. Shodai T. Endo T. Kohda D. J. Mol. Biol. 2001; 306: 137-143Crossref PubMed Scopus (82) Google Scholar).A few precursor proteins do not possess positive charges in what could be considered to be the classical N-terminal leader sequence. For example, the non-processed precursors rhodanese (30Waltner M. Weiner H. J. Biol. Chem. 1995; 270: 26311-26317Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar) and chaperonin10 (31Ryan M.T. Hoogenraad N.J. Hoj P.B. FEBS Lett. 1994; 337: 152-156Crossref PubMed Scopus (36) Google Scholar, 32Jarvis J.A. Ryan M.T. Hoogenraad N.J. Craik D.J. Hoj P.B. J. Biol. Chem. 1995; 270: 1323-1331Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar) have no positive charges in the first eight residues and actually have a negative charge in the leader. Leu4 leader has one lysine residue but two negative charges in its first 11 residues. These leader peptides were all found to be helical, consistent with our notion that structure can somehow overcome positive charge deficiency. Even if structure could overcome this situation, it is not apparent how the TOM complex, which contains patches of negative charges (7Schatz G. Nature. 1997; 388: 121-122Crossref PubMed Scopus (88) Google Scholar), can recognize the atypical leader.We took advantage of the two-leader strategy that we used previously to investigate the processing of leader peptides by MPP (18Mukhopadhyay A. Hammen P. Waltner-Law M. Weiner H. Protein Sci. 2002; 11: 1026-1035Crossref PubMed Scopus (8) Google Scholar). Here a modified non-functional leader was fused to the N termini of a native or modified precursor protein. In this way, it was possible to ascertain what part of the leader was actually needed by the import apparatus.Previously, it was shown that the R3Q,R10Q double mutant pALDH was not imported (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). Here we report that if this modified leader was fused to the R3Q,R10Q pALDH, import was restored. This shows that two non-functional leaders could actually be used to import a protein. The reason this double leader was functional in import could be that the internal leader-like structures shown in Fig. 1 actually were the portion of the precursor that was involved in the translocation process. No such structural motif existed in the original R3Q,R10Q double mutant pALDH we previously investigated. From the various chimeric precursors used, it is concluded that an internal helix that is either a long continuous amphipathic helix or part of the helix-linker-helix motif could act as the actual leader. No attempt was made to determine how distal from the N terminus such a structure would function.Using the strategy described above, it was possible to show that the most N-terminal, non-functional segment could even possess a negative charge such as found in the R3Q,S7E double mutant of pALDH. This finding can then be used to explain how a natural precursor protein could possess a net negative charge at the start of the N-terminal sequence. It can be argued that the import apparatus simply ignores the unfavorable region and interacts with the positively charged stable helical domain that follows.To test for the above mentioned model we used Leu4 and Leu9, natural yeast proteins that possess negatively charged amino acids in the most N-terminal segment of their leader sequence. NMR spectroscopy was used to show that, as expected, the peptide corresponding to the leader of Leu4 formed a stable α helix. Before the yeast genome was obtained, only one form of the enzyme was known. The second gene product, Leu9 (which has 80% identity with Leu4), was identified later, but the protein was not characterized (15Casalone E. Barberio C. Cavalieri D. Polsinelli M. Yeast. 2000; 16: 539-545Crossref PubMed Scopus (21) Google Scholar). First, in vitro import was used to show that Leu9, like Leu4, is a mitochondrial protein. Unexpectedly, the Leu4 precursor was not imported under conditions that allowed the Leu9 to be imported. A chimeric protein was made by fusing the purported leader segment of Leu4 or Leu9 to mature ALDH. When the first 21 residues of Leu4 (those found to be helical) were used, no import was found. But if the first 31 residues were used, import could occur. In a similar manner, the first 30 residues of Leu9 could import of mature ALDH, whereas the first 20 could not.We were concerned with the fact that we could not obtain import with Leu4 as others had reported (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 24Hampsey D.M. Lewin A.S. Kohlhaw G.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 1270-1274Crossref PubMed Scopus (16) Google Scholar). The previously published experimental procedure used total RNA for translation (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 24Hampsey D.M. Lewin A.S. Kohlhaw G.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 1270-1274Crossref PubMed Scopus (16) Google Scholar) and antibodies (33Gasser S.M. Daum G. Schatz G. J. Biol. Chem. 1982; 257: 13034-13041Abstract Full Text PDF PubMed Google Scholar) to precipitate imported proteins. In contrast, the current experiments all used pure cDNAs. Given the fact that theLeu9 gene was not identified at the time the initial experiments were performed and that total RNA was used, it is possible that the Leu9 gene product was detected rather than the product for Leu4. We conclude that under the in vitro experimental procedures used in this study, Leu4 was not imported to the mitochondrial matrix space.The N-terminal portion of Leu4 and Leu9 provided a 21- or 20-residue helix thought to provide the essential structure for a region that does not possess a net positive charge. Our data revealed that the N-terminal region containing the negative charges formed the most stable part of the helix. However, the region that seemed most responsible for import of Leu4 or Leu9 was not the most helical portion of the leader, because the targeting signal seemed to appear after residue 11. It is generally accepted that before import to the mitochondria preproteins remain unfolded. A proteolysis assay was performed with the TNT-synthesized Leu4. Proteinase K proteolysis results indicated that the Leu4 was more resistant to digestion than was the Leu4(Δ1–11) protein (data not shown). A possible explanation for this phenomenon is that the most N-terminal 11 residues of Leu4 contribute to the overall structure of the protein. Without these residues the native folding of these proteins might be changed, as reflected by the instability of the protein and the proteinase K sensitivity. Therefore, the proteolysis results showed that Leu4 was folded and hence not imported to the mitochondria in an in vitro experiment. Some proteins that fold rapidly have been found to be imported to the mitochondria by a co-translocation pathway. The pALDH leader fused to green fluorescent protein (a protein that folds rapidly) was shown to be imported to the mitochondrial matrix space by a co-translational pathway (34Ni L. Heard T.S. Weiner H. J. Biol. Chem. 1999; 274: 12685-12691Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). Similarly, in vivo Leu4 could be translocated to the mitochondria through a co-translational pathway.The acid-chain hypothesis was based on the fact that the leader possessed positive charges at its N terminus, which could be important for binding with the TOM complex. However, it has been shown that the leader sequence of pALDH binds TOM20 through hydrophobic interactions (29Muto T. Obita T. Abe Y. Shodai T. Endo T. Kohda D. J. Mol. Biol. 2001; 306: 137-143Crossref PubMed Scopus (82) Google Scholar). TOM20 is the first receptor to be recognized by leader sequence and does not appear to depend on the positive charges of the leader sequence. It is possible that the other TOM protein may interact with positive charges present in leader sequence. Here we show that if the N-terminal segment is not positive, the translocator appears to recognize not the N termini but the region of the leader sequences that are positively charged. It has been shown that inner and intermembrane space proteins use an internal signal after the most N-terminal matrix space targeting signal (35Arnold I. Folsch H. Neupert W. Stuart R.A. J. Biol. Chem. 1998; 273: 1469-1476Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar, 36Bomer U. Meijer M. Guiard B. Dietmeier K. Pfanner N. Rassow J. J. Biol. Chem. 1997; 272: 30439-30446Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). How the first 11 residues in the case of Leu4 enter the translocation pore cannot be explained at this time. If the model of import were co-translational and not post-translational as we suggest, then it is possible that an unknown cytosolic factor is necessary to insert the proteins into the TOM apparatus. Although the data presented in this study cannot explain how the precursor with negative charges interacts with the TOM proteins, they do show that the information for import need not lie at the very N terminus of the precursor protein. During the past decade much has been learned about the leader sequences of precursor proteins that are destined for the mitochondrial matrix space (3Neupert W. Annu. Rev. Biochem. 1997; 66: 863-917Crossref PubMed Scopus (974) Google Scholar, 27von Heijne G. EMBO J. 1986; 5: 1335-1342Crossref PubMed Scopus (705) Google Scholar). No reports exist, however, to explain how a leader that does not possess a net positive charge at its N terminus is imported. During our studies with the positive charge in the ALDH leader, we often found that if the leader was made more helical it would compensate for the removal of positive charges (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). No actual structure for a precursor protein that possesses a cleavable leader sequence has been determined. NMR has been used to show that a number of peptides corresponding to leader sequences can indeed form α helices when bound to micelles or when in a hydrophobic environment (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar,8Karslake C. Piotto M.E. Pak Y.K. Weiner H. Gorenstein D.G. Biochemistry. 1990; 29: 9872-9878Crossref PubMed Scopus (85) Google Scholar, 10Thornton K. Wang Y. Weiner H. Gorenstein D.G. J. Biol. Chem. 1993; 268: 19906-19914Abstract Full Text PDF PubMed Google Scholar, 28Hammen P.K. Gorenstein D.G. Weiner H. Biochemistry. 1996; 35: 3772-3781Crossref PubMed Scopus (38) Google Scholar). Recently the structure of the aldehyde dehydrogenase leader bound to TOM20 was shown to be helical, supporting the notion that helicity is indispensable for its function (29Muto T. Obita T. Abe Y. Shodai T. Endo T. Kohda D. J. Mol. Biol. 2001; 306: 137-143Crossref PubMed Scopus (82) Google Scholar). A few precursor proteins do not possess positive charges in what could be considered to be the classical N-terminal leader sequence. For example, the non-processed precursors rhodanese (30Waltner M. Weiner H. J. Biol. Chem. 1995; 270: 26311-26317Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar) and chaperonin10 (31Ryan M.T. Hoogenraad N.J. Hoj P.B. FEBS Lett. 1994; 337: 152-156Crossref PubMed Scopus (36) Google Scholar, 32Jarvis J.A. Ryan M.T. Hoogenraad N.J. Craik D.J. Hoj P.B. J. Biol. Chem. 1995; 270: 1323-1331Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar) have no positive charges in the first eight residues and actually have a negative charge in the leader. Leu4 leader has one lysine residue but two negative charges in its first 11 residues. These leader peptides were all found to be helical, consistent with our notion that structure can somehow overcome positive charge deficiency. Even if structure could overcome this situation, it is not apparent how the TOM complex, which contains patches of negative charges (7Schatz G. Nature. 1997; 388: 121-122Crossref PubMed Scopus (88) Google Scholar), can recognize the atypical leader. We took advantage of the two-leader strategy that we used previously to investigate the processing of leader peptides by MPP (18Mukhopadhyay A. Hammen P. Waltner-Law M. Weiner H. Protein Sci. 2002; 11: 1026-1035Crossref PubMed Scopus (8) Google Scholar). Here a modified non-functional leader was fused to the N termini of a native or modified precursor protein. In this way, it was possible to ascertain what part of the leader was actually needed by the import apparatus. Previously, it was shown that the R3Q,R10Q double mutant pALDH was not imported (6Hammen P.K. Waltner M. Hahnemann B. Heard T.S. Weiner H. J. Biol. Chem. 1996; 271: 21041-21048Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). Here we report that if this modified leader was fused to the R3Q,R10Q pALDH, import was restored. This shows that two non-functional leaders could actually be used to import a protein. The reason this double leader was functional in import could be that the internal leader-like structures shown in Fig. 1 actually were the portion of the precursor that was involved in the translocation process. No such structural motif existed in the original R3Q,R10Q double mutant pALDH we previously investigated. From the various chimeric precursors used, it is concluded that an internal helix that is either a long continuous amphipathic helix or part of the helix-linker-helix motif could act as the actual leader. No attempt was made to determine how distal from the N terminus such a structure would function. Using the strategy described above, it was possible to show that the most N-terminal, non-functional segment could even possess a negative charge such as found in the R3Q,S7E double mutant of pALDH. This finding can then be used to explain how a natural precursor protein could possess a net negative charge at the start of the N-terminal sequence. It can be argued that the import apparatus simply ignores the unfavorable region and interacts with the positively charged stable helical domain that follows. To test for the above mentioned model we used Leu4 and Leu9, natural yeast proteins that possess negatively charged amino acids in the most N-terminal segment of their leader sequence. NMR spectroscopy was used to show that, as expected, the peptide corresponding to the leader of Leu4 formed a stable α helix. Before the yeast genome was obtained, only one form of the enzyme was known. The second gene product, Leu9 (which has 80% identity with Leu4), was identified later, but the protein was not characterized (15Casalone E. Barberio C. Cavalieri D. Polsinelli M. Yeast. 2000; 16: 539-545Crossref PubMed Scopus (21) Google Scholar). First, in vitro import was used to show that Leu9, like Leu4, is a mitochondrial protein. Unexpectedly, the Leu4 precursor was not imported under conditions that allowed the Leu9 to be imported. A chimeric protein was made by fusing the purported leader segment of Leu4 or Leu9 to mature ALDH. When the first 21 residues of Leu4 (those found to be helical) were used, no import was found. But if the first 31 residues were used, import could occur. In a similar manner, the first 30 residues of Leu9 could import of mature ALDH, whereas the first 20 could not. We were concerned with the fact that we could not obtain import with Leu4 as others had reported (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 24Hampsey D.M. Lewin A.S. Kohlhaw G.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 1270-1274Crossref PubMed Scopus (16) Google Scholar). The previously published experimental procedure used total RNA for translation (14Beltzer J.P. Morris S.R. Kohlhaw G.B. J. Biol. Chem. 1988; 263: 368-374Abstract Full Text PDF PubMed Google Scholar, 24Hampsey D.M. Lewin A.S. Kohlhaw G.B. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 1270-1274Crossref PubMed Scopus (16) Google Scholar) and antibodies (33Gasser S.M. Daum G. Schatz G. J. Biol. Chem. 1982; 257: 13034-13041Abstract Full Text PDF PubMed Google Scholar) to precipitate imported proteins. In contrast, the current experiments all used pure cDNAs. Given the fact that theLeu9 gene was not identified at the time the initial experiments were performed and that total RNA was used, it is possible that the Leu9 gene product was detected rather than the product for Leu4. We conclude that under the in vitro experimental procedures used in this study, Leu4 was not imported to the mitochondrial matrix space. The N-terminal portion of Leu4 and Leu9 provided a 21- or 20-residue helix thought to provide the essential structure for a region that does not possess a net positive charge. Our data revealed that the N-terminal region containing the negative charges formed the most stable part of the helix. However, the region that seemed most responsible for import of Leu4 or Leu9 was not the most helical portion of the leader, because the targeting signal seemed to appear after residue 11. It is generally accepted that before import to the mitochondria preproteins remain unfolded. A proteolysis assay was performed with the TNT-synthesized Leu4. Proteinase K proteolysis results indicated that the Leu4 was more resistant to digestion than was the Leu4(Δ1–11) protein (data not shown). A possible explanation for this phenomenon is that the most N-terminal 11 residues of Leu4 contribute to the overall structure of the protein. Without these residues the native folding of these proteins might be changed, as reflected by the instability of the protein and the proteinase K sensitivity. Therefore, the proteolysis results showed that Leu4 was folded and hence not imported to the mitochondria in an in vitro experiment. Some proteins that fold rapidly have been found to be imported to the mitochondria by a co-translocation pathway. The pALDH leader fused to green fluorescent protein (a protein that folds rapidly) was shown to be imported to the mitochondrial matrix space by a co-translational pathway (34Ni L. Heard T.S. Weiner H. J. Biol. Chem. 1999; 274: 12685-12691Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). Similarly, in vivo Leu4 could be translocated to the mitochondria through a co-translational pathway. The acid-chain hypothesis was based on the fact that the leader possessed positive charges at its N terminus, which could be important for binding with the TOM complex. However, it has been shown that the leader sequence of pALDH binds TOM20 through hydrophobic interactions (29Muto T. Obita T. Abe Y. Shodai T. Endo T. Kohda D. J. Mol. Biol. 2001; 306: 137-143Crossref PubMed Scopus (82) Google Scholar). TOM20 is the first receptor to be recognized by leader sequence and does not appear to depend on the positive charges of the leader sequence. It is possible that the other TOM protein may interact with positive charges present in leader sequence. Here we show that if the N-terminal segment is not positive, the translocator appears to recognize not the N termini but the region of the leader sequences that are positively charged. It has been shown that inner and intermembrane space proteins use an internal signal after the most N-terminal matrix space targeting signal (35Arnold I. Folsch H. Neupert W. Stuart R.A. J. Biol. Chem. 1998; 273: 1469-1476Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar, 36Bomer U. Meijer M. Guiard B. Dietmeier K. Pfanner N. Rassow J. J. Biol. Chem. 1997; 272: 30439-30446Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). How the first 11 residues in the case of Leu4 enter the translocation pore cannot be explained at this time. If the model of import were co-translational and not post-translational as we suggest, then it is possible that an unknown cytosolic factor is necessary to insert the proteins into the TOM apparatus. Although the data presented in this study cannot explain how the precursor with negative charges interacts with the TOM proteins, they do show that the information for import need not lie at the very N terminus of the precursor protein. We thank Professor Guntar Kohlhaw, Purdue University, for his help and advice when working with IPMS. Supplementary Material Download .pdf (.02 MB) Help with pdf files Download .pdf (.02 MB) Help with pdf files