Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Surfactant protein (SP)-C is one of the four specific proteins in pulmonary surfactant, which is secreted by the alveolar type II cell. The main function of surfactant is to reduce the surface tension at the air-water interface of the lung alveoli by forming a surface active film, thereby preventing alveoli from collapsing and facilitating the work of breathing (for review, see Refs. 1 and 2). SP-C is important in promoting the spreading of the surfactant lipids to the air-water interface (1, 2). This protein is synthesized as a propeptide (proSP-C) with a molecular mass of 21 kD, which probably adopts a type II orientation in the membrane (3). Mature SP-C is formed by several proteolytic steps at both the N- and C-terminal ends of proSP-C and is finally stored in the lamellar body of alveolar type II cells (1, 4, 5). It is a small and very hydrophobic protein of 4.2 kD and its primary amino acid sequence is highly conserved among species (6). SP-C isolated from bronchoalveolar lavage (BAL) is mainly dipalmitoylated (6). Palmitoyl chains are attached to cysteine residues 5 and 6 (numbering is based on the sequence of the mature protein) of proSP-C before its processing to mature SP-C distal to the trans-Golgi network (5, 7). Gustafsson and colleagues (8) described a tripalmitoylated isoform of SP-C, constituting approximately 4% of the main form, and found that the third palmitoyl chain was linked to the  -amino group of lysine 11. Other characteristics of SP-C are two prolines that flank the palmitoylated cysteine residues and two positively charged amino acids, lysine and arginine, at positions 11 and 12 (6). The two basic residues, which probably interact with the head groups of negatively charged phospholipids (9), are followed by a transmembrane domain consisting of 23 hydrophobic amino acids (6). The function of SP-C palmitoylation is not known, but a role of the acyl chains in membrane association of SP-C is unlikely because this occurs even in their absence (10, 11). The palmitoyl chains are thought to serve as an anchor between two lipid layers, with the palmitoyl chains in one bilayer and the transmembrane domain of the peptide in the other, in this way keeping lipids that have been squeezed out during compression of the surface film near the surfactant monolayer for rapid insertion upon the next inhalation (12, 13).

Palmitoylation is a post-translational modification in which palmitate, a 16-carbon fatty acid, is attached to the sulfhydryl group of a cysteine residue through a thioester linkage. It occurs on a wide variety of viral and cellular proteins, many of which play key roles in regulating cellular structure and function (for review, see Refs. 14 and 15). Palmitoylation has been reported to occur at various subcellular sites, such as endoplasmic reticulum (ER), Golgi apparatus, and plasma membrane (16), and is very likely an enzymatic process, involving a palmitoyltransferase (20, 21). Several palmitoyltransferase activities have been partially or completely purified (for review, see Ref. 15), but their involvement in the palmitoylation of proteins under physiologic conditions remains to be established. Palmitoylation can be a reversible modification: cycles of acylation and deacylation have been described to occur, e.g., for the  subunit of G proteins and nitric oxide synthase (15).

Palmitoylated proteins can be grouped into four categories (15): category 1 palmitoylated proteins are membrane proteins that are palmitoylated on one or several cysteine residues located adjacent to or just within the membrane domain; category 2 proteins include members of the Ras family that are palmitoylated in the C-terminal region after isoprenylation of their C-terminal CAAX-box; category 3 proteins are proteins palmitoylated at cysteine residues near the N- or C-terminus; and category 4 proteins are dually fatty acylated with myristate and palmitate. For the first category of palmitoylated proteins, to which SP-C also belongs, no consensus signal for palmitoylation has been identified, in contrast to what has been found for the second and fourth groups (15). In a given palmitoylated membrane protein (category 1) only distinct cysteines are palmitoylated. Although the position of the palmitoylated cysteine residues varies considerably in different acyl proteins (14, 22), they are mostly located at the cytoplasmic side near transmembrane domains, or in the transmembrane domain itself. Schweizer and associates (23), who studied the structural requirements for the palmitoylation of the transmembrane protein p63, came to the conclusion that palmitoylation occurs without a primary sequence motif: only the six-amino-acid spacing between the cysteine to be palmitoylated and the transmembrane segment allowed efficient palmitoylation, whereas the identity of neither the amino acids surrounding this cysteine nor the transmembrane domain was critical for palmitoylation. However, Ponimaskin and Schmidt (24), who investigated the palmitoylation of influenza virus hemagglutinin, argued that being in the vicinity of a transmembrane domain is not enough for a cysteine to become palmitoylated, and thus additional structural features must be required.

The present study was carried out to obtain more insight into the structural requirements for the palmitoylation of proSP-C and into the subcellular localization of the palmitoylation. We found that of the cysteine residues in proSP-C, only the ones at positions 5 and 6 of mature SP-C are palmitoylated; and that substitution of the two basic residues, arginine and lysine, which are located at the N-terminal end of the transmembrane domain, by glutamine abolishes palmitoylation of proSP-C by preventing the transport of proSP-C to the Golgi compartment.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Ham's F12 medium and enzymes used in molecular cloning were obtained from Life Technologies (Gaithersburg, MD), [³H]palmitate and [¹²⁵I]-labeled goat antimouse immunoglobulin (Ig)G from NEN Life Science Products (Boston, MA), and oligonucleotides from Amersham Pharmacia Biotech (Roosendaal, The Netherlands). Anti-Xpress antibody (Invitrogen, Carlsbad, CA) was used for the detection of the his tag. Antibody 68514, recognizing amino acids 1 to 20 of proSP-C (25), was a kind gift of Dr. T. E. Weaver (University of Cincinnati, Cincinnati, OH).

Recombinant DNA Procedures

All basic DNA procedures were performed as described (26). Complementary DNA (cDNA) encoding proSP-C was amplified by polymerase chain reaction (PCR) using a plasmid containing full-length cDNA encoding human proSP-C (27) as a template and 5'-cgggatccatggatgtgggcagcaaagaggtc-3' and 5'-cggaattcccg gaggcgtcctagatgtag-3' as left- and right-hand primers, respectively. The PCR product was cloned into pcDNA3.1His (Invitrogen) or pcDNA3.1 using the BamHI and EcoRI sites present in the primers. These constructs, designated pHisproSP-C and pproSP-C, were verified by DNA sequencing. Mutations in pHisproSP-C were made with the QuikChange Site-Directed Mutagenesis kit (Stratagene, Cedar Creek, TX) as described by the manufacturer using the following oligonucleotide primers: 5'- gatttggcattccctcctccccagtgcacctg-3' and 5'-caggtgcactggggaggagg gaatgccaaatc-3' (CC^5,6SS); 5'-ggcattccctcctgcccagtgcacctg-3' and 5'-caggtgcactgggcaggagggaatgcc-3' (C⁵S); 5'-ggcattccctgctccccagtg cacctg-3' and 5'-caggtgcactggggagcagggaatgcc-3' (C⁶S); 5'-gcccagtgc acctgcaacaacttcttatcgtggtggtgg-3' and 5'-ccaccaccacgataagaagttgt gcaggtgcactgggc-3' (KR^11,12QQ); 5'-gcccagtgcacctgaaacaacttcttat cgtggtggtgg-3' and 5'-ccaccaccacgataagaagttgtttcaggtgcactgggc-3' (R¹²Q); 5'-gcccagtgcacctgcaacgccttcttatcgtgg-3' and 5'-ccacgataagaa ggcgttgcaggtgcactgggc-3' (K¹¹Q); and 5'-gcccagtgcacctgagacaa cttcttatcgtgg-3' and 5'-ccacgataagaagttgtctcaggtgcactggg-3' (K¹¹RR¹²Q). All mutations were confirmed by DNA sequencing.

Cell Culture and Transfection

Chinese hamster ovary (CHO)-K1 cells (CRL-9618; American Type Culture Collection, Manassas, VA) were cultured in Ham's F12 medium supplemented with 7.5% fetal calf serum, 100 U/ml penicillin, and 100 µg/ml streptomycin in a humidified 5% CO₂ atmosphere at 37°C. CHO cells, grown in 35-mm dishes, were transiently transfected with Lipofectamine Plus (Life Technologies) as recommended by the manufacturer.

Metabolic Labeling with [³H]Palmitate

At 24 h after the introduction of plasmid DNA, transiently transfected CHO cells were rinsed with phosphate-buffered saline (PBS) and labeled in 800 µl serum-free Ham's F12 medium containing 500 µCi [³H]palmitate (30 to 60 Ci/mmol) in the presence or absence of 10 µg/ml brefeldin A (BFA) for 2 h at 37°C. Labeled cells were washed twice with ice-cold PBS and lysed with sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer (62.5 mM Tris, 2% SDS, 10% glycerol, and 0.003% bromophenol blue, pH 6.8) for direct analysis on SDS-PAGE or with lysis buffer (0.1% SDS, 1% Triton-X-100, 1% deoxycholate, 0.15 M NaCl, 20 mM Tris, 10 mM ethylenediaminetetraacetic acid, 10 mM iodacetamide, and 1 mM phenylmethylsulfonyl fluoride, pH 7.4) for immunoprecipitation.

Immunoprecipitation

Cell lysate was incubated with antibody 68514 (final dilution 1:100) for 2 h at room temperature. After this incubation, 2.5 mg preswollen protein A-sepharose beads were added and the lysate was again incubated for 2 h at room temperature. Immunoprecipitation complexes were pelleted and washed six times with lysis buffer. Washed immunoprecipitation complexes were boiled for 2 min in SDS-PAGE sample buffer before analysis by SDS-PAGE (28).

Quantification of [³H]Palmitate Incorporation

Samples were boiled for 2 min and separated on two identical 12% SDS polyacrylamide gels (28). One gel was fixed with 10% acetic acid/50% methanol, impregnated with 1 M sodium salicylate to enhance the ³H signal, dried, and exposed to Kodak X-Omat AR films (Kodak, Rochester, NY) to visualize radiolabeled proteins. Quantification of fluorograms was carried out by densitometry using a Personal densitometer SI and ImageQuant version 5.1 software (Molecular Dynamics, Sunnyvale, CA). To quantify the amount of his-proSP-C expressed, the second gel was subjected to immunoblot analysis using the anti-Xpress antibody directed against the his tag. [¹²⁵I]-labeled goat antimouse IgG was used as a second antibody for detection. Quantification was performed using the Fujix BAS 1000 bioimaging analyzer system (Fuji Photo Film, Düsseldorf, Germany), BAS-MP imaging plates (Fuji Photo Film), and Aida 2.11 software (Raytest Isotopenmessgeräte, Straubenhardt, Germany). The amount of [³H]palmitate incorporated was expressed relative to the amount of SP-C expressed.

Immunocytochemistry

CHO cells were grown on glass coverslips. At 24 h after transfection, cells were fixed with 2% formaldehyde in PBS and permeabilized with 0.1% Triton X-100 in PBS. After preincubation with PBS containing 0.1% bovine serum albumin (BSA), the cells were incubated with the primary antibodies for 2 h at room temperature, washed three times for five min each time with PBS containing 0.1% BSA, incubated with fluorophore-conjugated secondary antibodies overnight at 4°C, and washed five times. PBS/glycerol 1:9 was used as an antifading agent. Fluorescently labeled proteins were visualized by placing the specimen under a spectral confocal microscope (Leica TCS SP; Leica GmbH, Heidelberg, Germany). For visualization of his-proSP-C, mouse anti-Xpress antibody (Invitrogen) was used in combination with Alexa Fluor 568 rabbit antimouse IgG (1:1,000, Molecular Probes, Leiden, The Netherlands). For visualization of the ER, a goat polyclonal antibody against BiP (1:1,000, GRP 78 [C-20]; Santa Cruz Biotechnology, Santa Cruz, CA) was used in combination with Alexa Fluor 488 rabbit antigoat IgG (1:1,000; Molecular Probes). Colocalization of Alexa Fluor 488 and Alexa Fluor 568 was detected simultaneously. For Alexa Fluor 488-labeled proteins, the 488-nm argon laser line was used for excitation and fluorescence was detected with a photomultiplier tube (emission selected in the wavelength range of 500 to 550 nm). For Alexa Fluor 568- labeled proteins, the 568-nm krypton laser line was used for excitation and fluorescence was detected with a photomultiplier tube (emission selected in the wavelength range of 585 to 650 nm). Serial 0.243-µm optical sections running through the entire cell (double scans per section) were made.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cysteines 5 and 6 of Mature SP-C Are the Only Cysteines Palmitoylated in his-ProSP-C Expressed in CHO Cells

We investigated the structural requirements for the palmitoylation of proSP-C rather than mature SP-C because SP-C becomes palmitoylated already in its propeptide form (7). ProSP-C constructs with a his tag on the N-terminus were used in this study. Immunocytochemistry showed that proSP-C, with or without the his tag, had a similar subcellular localization in the cell (not shown), indicating that proSP-C with the tag behaves the same as wild-type (WT). In total cell lysates of his-proSP-C-transfected cells, a protein of approximately 28 kD in which [³H]palmitate was incorporated was readily observed (Figure 1A). This protein was not present in mock-transfected cells (Figure 1A), and its size corresponded to the size of his-proSP-C as detected on immunoblots (Figure 1B). The 28-kD polypeptide could be immunoprecipitated by a proSP-C-specific antibody, further proving its identity as proSP-C (Figure 1A). As shown in Figure 1A, [³H]palmitate-labeled his- proSP-C was well separated from other labeled proteins. Therefore, the palmitoylation levels of his-proSP-C were determined in subsequent experiments by fluorography after electrophoresis of whole-cell lysates.

View larger version (80K): [in this window] [in a new window]   Figure 1.   His-proSP-C is expressed in CHO cells as a 28-kD palmitoylated protein. pHisproSP-C (WT) or mock transfected CHO cells were labeled for 2 h with [3H]palmitate, after which cells were lysed with lysis buffer. His-proSP-C was immunoprecipitated (IP) from the cell lysates with proSP-C-specific antibody 68514. The samples were subjected to SDS-PAGE and analyzed by (A) fluorography and (B) Western blotting with the anti-Xpress antibody which recognizes the his tag. The numbers at the left margin of the gel indicate known molecular mass in kilodaltons.

ProSP-C contains several cysteine residues in addition to the two that are palmitoylated in vivo. To see whether his- proSP-C in CHO cells was palmitoylated on the same residues as SP-C isolated from BAL, i.e., at cysteine residues 5 and 6 (29, 30), substitution mutants of proSP-C were made in which both cysteines (CC^5,6SS) or only one of the cysteine residues (C⁵S and C⁶S) were substituted by a serine residue (Figure 2A). As shown in Figures 2B and 2C, no incorporation of [³H]palmitate was detected in the double cysteine mutant CC^5,6SS, whereas the incorporation in both single cysteine mutants, C⁵S and C⁶S, was reduced to about 50% of that of WT. Incorporation of label was corrected for proSP-C expression, although the expression levels did not differ much among the various constructs. These results show that in proSP-C only cysteine residues 5 and 6 (numbering is based on the mature protein) are palmitoylated when expressed in CHO cells, and thus no palmitoylation could be observed in any other cysteine residue or lysine 11 of proSP-C. This makes cells transfected with his-proSP-C a relevant model to study palmitoylation of SP-C.

View larger version (30K): [in this window] [in a new window]   Figure 2.   Cysteines 5 and 6 of mature SP-C are the only cysteines palmitoylated in his-proSP-C expressed in CHO cells. (A) A schematic presentation of his-proSP-C and the cysteine mutants. His-proSP-C is represented by a bar, which is subdivided into the his-tag (grey), the part of proSP-C which forms mature SP-C (black), and the N- and C-terminal propeptides (white). Above the bar the lengths in amino acids of the different parts are indicated. The amino acid sequence of mature SP-C is shown in single letter code and is numbered. For the mutants, the transmembrane domain of mature SP-C is represented by a line. Substituted amino acids are underlined and bold. The names of the mutants are depicted at the right, next to the amino acid sequence of the mature SP-C part of these his-proSP-C mutants. (B) CHO cells transiently transfected with WT, CC5,6SS, C5S, and C6S were labeled with [3H]palmitate. Cell lysates were subjected to SDS-PAGE and fluorography for the detection of 3H radioactivity. The band representing 28-kD his-proSP-C on the fluorogram is shown. (C) The amount of [3H]palmitate incorporated is expressed relative to the amount of proSP-C (see MATERIALS AND METHODS). The value obtained with WT was taken as 100%. Data are means ± standard deviation (SD) of three separate experiments.

Substitution of the Juxtamembrane Lysine and Arginine by Glutamine Reduces the Level of Palmitoylation of ProSP-C

The two juxtamembrane basic amino acids, lysine and arginine, are conserved in the primary amino acid sequence of SP-C of different species. The two basic residues might be a part of a signal directing the palmitoylation of SP-C, inasmuch as lysine and arginine residues are often found near palmitoylated cysteines (14). To investigate the possible requirement of the basic amino acids for palmitoylation of proSP-C, mutants were constructed in which either one (R¹²Q or K¹¹Q) or both (KR^11,12QQ) of the two basic residues were substituted by the uncharged amino acid glutamine (Figure 3A). Transiently transfected CHO cells expressing these constructs were analyzed by [³H]palmitate labeling, and all the mutants showed dramatic decreases in the levels of palmitoylation (Figures 3C and 3D), although their expressions were comparable to that of WT (Figure 3B). Compared with WT proSP-C, incorporation of [³H]palmitate into KR^11,12QQ, R¹²Q, and K¹¹Q was reduced to 11, 25, and 70%, respectively.

View larger version (33K): [in this window] [in a new window]   Figure 3.   Substitution of lysine 11 and arginine 12 by glutamine residues reduces the level of palmitoylation of proSP-C. (A) A schematic presentation of his-proSP-C and the lysine and arginine mutants. Only the amino acid sequence of the part encoding mature SP-C is shown. The transmembrane domain is represented by a line. The substituted amino acids are underlined and bold. The names of the mutants are depicted at the right, next to the amino acid sequences. CHO cells transiently transfected with WT his-proSP-C and the mutants thereof (KR11,12QQ, R12Q and K11Q) were labeled with [3H]palmitate and analyzed by (B) Western blotting with the anti-Xpress antibody, which recognizes the his tag, and (C) fluorography as described in Figure 2. (D) The amount of [3H]palmitate incorporated is expressed relative to the amount of proSP-C (see MATERIALS AND METHODS). The value obtained with WT was taken as 100%. Data are means ± SD of three separate experiments.

Mutants Lacking the Juxtamembrane Positive Charges Are Retained in the ER

To investigate whether the observed decrease in palmitoylation of the proSP-C mutants lacking the juxtamembrane positive charge was caused by a direct effect of the mutations on the palmitoylation reaction itself or by an indirect effect on the folding, and thereby on the subcellular localization of proSP-C, we studied the localization of the proSP-C mutants in CHO cells with immunohistochemistry. WT proSP-C was localized in punctate vesicular structures throughout the cell, whereas the double-charge mutant KR^11,12QQ showed a more perinuclear labeling (Figure 4). This perinuclear labeling colocalized predominantly with the lumenal ER marker BiP (31) (Figure 4). WT proSP-C showed only little colocalization with BiP (Figure 4). Mutants R¹²Q and K¹¹Q, missing only one basic residue, showed a labeling pattern that was a mixture of the labeling of WT and the KR^11,12QQ mutant (not shown). The nonpalmitoylated proSP-C mutant CC^5,6SS showed a labeling pattern comparable to WT (not shown), indicating that the difference in localization between WT and the mutants in which the juxtamembrane basic residues are replaced is not caused by the difference in the degree of palmitoylation.

View larger version (24K): [in this window] [in a new window]   Figure 4.   Mutants lacking the juxtamembrane positive charges are retained in the ER. At 24 h after transfection of CHO cells with WT his-proSP-C (A-C) or KR11,12QQ (D-F) the cells were fixed and double-immunostained for his tag and BiP. His tag- Alexa 568-specific fluorescence is depicted in A and D; BiP- Alexa 488-specific fluorescence is depicted in B and E; and an overlay of the fluorescence by the two fluorophores is depicted in C and F. Colocalization appears as yellow.

Effect of BFA on Palmitoylation of the Mutants Lacking the Juxtamembrane Positive Charges

Immunohistochemistry showed that the charge mutants lacking the juxtamembrane positive charge are probably located in the ER compartment. Because palmitoylation of proSP-C may take place in a compartment distal to the ER, the loss in palmitoylation might be caused by the different localization of the mutants. Therefore BFA, which induces redistribution of the Golgi and ER to one compartment (32), was added to the cells during labeling with [³H]palmitate. As shown in Figure 5, the palmitoylation of his- proSP-C was not influenced by addition of BFA. However, BFA abolished the reduction of the palmitoylation of charge mutants KR^11,12QQ and K¹¹Q (compare Figure 5 with Figure 3). Palmitoylation of R¹²Q was restored to only about 65% of that of the WT in the presence of BFA (Figure 5). This indicates that the positively charged amino acids have a predominantly indirect effect on the palmitoylation level of proSP-C by affecting the localization of the protein, and not a direct effect by affecting the recognition of the cysteines residues as palmitoyl acceptors.

View larger version (42K): [in this window] [in a new window]   Figure 5.   Effect of BFA on palmitoylation of the mutants lacking the juxtamembrane positive charges. (A) CHO cells transiently transfected with WT his-proSP-C, KR11,12QQ, R12Q, and K11Q were labeled with [3H]palmitate in the presence of 10 µg/ml BFA and analyzed by fluorography. (B) The amount of [3H]palmitate incorporated is expressed relative to the amount of proSP-C (see MATERIALS AND METHODS). The value obtained with WT was taken as 100%. Data are means ± SD of three separate experiments.

Interestingly, BFA could not raise palmitoylation of R¹²Q to normal levels, as we found for KR^11,12QQ and K¹¹Q. To investigate this observation, we made another mutant of proSP-C in which lysine 11 was changed to an arginine and arginine 12 was mutated to a glutamine (K¹¹RR¹²Q), thus differing from R¹²Q only in the identity of the positively charged residue present at position 11. Compared with WT proSP-C, incorporation of [³H]palmitate into K¹¹RR¹²Q was reduced to 61 ± 2%. In the presence of BFA it was still reduced to 62 ± 17%.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study shows that substitution of the juxtamembrane basic residues lysine and arginine by glutamine residues has large effects on the palmitoylation level of proSP-C. By the addition of BFA this effect could nearly be abolished for the lysine and double mutant. The palmitoylation of the arginine mutant increased also, but not to WT levels. Further, these mutants were found to have a different subcellular localization in CHO cells than WT proSP-C. We can draw several conclusions from these results.

First, our results indicate that the juxtamembrane arginine and lysine residues are not structurally required for palmitoylation of proSP-C. In the presence of BFA the level of palmitoylation of this double-basic-charge mutant KR^11,12QQ is comparable to that of WT (Figure 5). Therefore, the mutant has the potential to become fully palmitoylated. Schweizer and coworkers (23) changed the amino acids surrounding the palmitoylated cysteine of p63 (including two arginine residues) to alanines and also found no effect on palmitoylation. Therefore, the often-observed presence of arginine and lysine residues near palmitoylated cysteines (14) may be a consequence of the fact that palmitoylated cysteines are often located near the transmembrane domain of a protein, which is, in general, preceded by positive charges (33) rather than a requirement for palmitoylation.

Positively charged amino acids in the region flanking the transmembrane domain of a protein have been shown to have significant effects on the membrane orientation of the protein. A type II transmembrane orientation is more likely when the net charge is higher at the N-terminal side (33). This predicts a type II orientation for proSP-C, which is in agreement with the conclusion drawn from proteinase protection studies by Keller and colleagues (3) but in contradiction with the results of Vorbroker and associates (34), who suggested a type III orientation after proteinase protection studies. A type II orientation of proSP-C in the membrane is also supported by the fact that the palmitoylated cysteine residues of proSP-C are located in the N-terminus, inasmuch as palmitoylation of proteins is thought to occur on the cytoplasmic side of the membrane (14, 35). Further, depending on its length and hydrophobicity, positively charged residues following a hydrophobic segment have been described to increase the efficiency of a stop-transfer signal of this hydrophobic segment (36, 37). However, after removal of the juxtamembrane positive charges, palmitoylation still occurred in the presence of BFA, indicating that the removal of these positive charges does not influence the orientation and translocation of proSP-C. Calculation of the net charge of proSP-C (considering the 15 residues flanking the transmembrane domain on both sides) shows that the N-terminus is more positive even in the absence of the two basic residues, explaining the lack of effect of the juxtamembrane positively charged residues on membrane orientation. To our surprise, the proSP-C mutant in which only the juxtamembrane arginine was mutated (R¹²Q) did not become completely palmitoylated after BFA addition, in contrast to the mutant in which only the lysine was changed (K¹¹Q) and the double mutant (KR^11,12QQ). To see whether this observation is related to the localization of the only positive charge present or to the type of basic amino acid (lysine versus arginine), we made K¹¹RR¹²Q, differing from R¹²Q in the positively charged residue present at position 11. K¹¹RR¹²Q also showed impaired palmitoylation in the presence of BFA, like R¹²Q. This suggests that removal of a positive charge on position 11 has a different effect compared with removal of a positive charge at position 12, regardless of the basic residue present. It is tempting to speculate that the palmitoylation process discriminates between the position of the positive charge (amino acid 11 versus 12) present in proSP-C when there is only one charged amino acid present. However, we cannot exclude the possibility that the removal of the positive charge at position 12 had influence on orientation or translocation of part of the expressed mutant protein and that consequently the cysteines were not available at the cytosolic side of the membrane for palmitoylation.

Second, our results suggest that the juxtamembrane positive charged amino acids are involved in subcellular localization of proSP-C. The mutant lacking the two basic residues clearly showed a more perinuclear, probably ER, localization in the cell than WT (Figure 4). The single-basic-residue mutants, R¹²Q and K¹¹Q, showed partly WT subcellular localization, corresponding with the higher levels of palmitoylation of the single-basic-residue mutants when compared with the double mutant (Figure 3). Russo and associates (38) found in their study with green fluorescent protein (EGFP)-proSP-C fusion constructs that mature SP-C (amino acids 24 to 59 of proSP-C) contains a signal sequence that is capable of inducing ER insertion of the EGFP fusion protein but is incapable of directing the fusion proteins to distal compartments. In the same study, a part of the N-terminal propeptide (amino acids 10 to 23) was found to be required for targeting proSP-C to late vesicular compartments. In earlier studies (39, 40), the distal COOH terminus was thought to contain this targeting signal because deletion of the last 10 amino acids resulted in ER/Golgi localization of the proteins. Recently, the retention of this proSP-C lacking the last 10 amino acids was ascribed to misfolding and subsequent formation of aggregates within aggresomes of unprocessed mutant protein (38, 41). Whether the perinuclear localization of the KR^11,12QQ mutant, and to a lesser extent that of K¹¹Q and R¹²Q, is caused by improper folding, by the introduction of an ER retention signal, or by affecting the targeting signal is not clear. We found no indication that the juxtamembrane positive charge mutants were broken down more rapidly compared with WT or that they were ubiquitinated as described for several C-terminal mutants by Kabore and coworkers (41), inasmuch as the mutants had the same intensity and appearance on Western blots as did WT proSP-C, and no bands with a higher molecular mass indicative of ubiquitination (42) were visible (Figure 3B). Further, the perinuclear structure in which the juxtamembrane positive charge mutants are located appeared to be different from the aggresomes described in Reference 41: around the whole nucleus instead of in the form of a single small area near the nucleus. The possibility of affecting the targeting signal would imply that mature sequences in SP-C are part of the targeting signal of proSP-C, possibly in combination with amino acids 10 to 23. The subcellular localization of the nonpalmitoylated proSP-C mutant (CC^5,6SS) is similar to the localization of the WT, so the palmitoyl chains do not seem to be important in subcellular targeting. Therefore, anchoring of different lipid layers of surfactant by the final processing product, mature SP-C, still remains the only function assigned to the palmitoyl chains (12, 13).

Third, proSP-C becomes palmitoylated in a compartment distal to the ER. We showed that the nonpalmitoylated mutants lacking the positive charges accumulated in the ER compartment and that they became palmitoylated after addition of BFA, a compound that causes a redistribution of the Golgi proteins into the ER compartment (32, 43). Further, palmitoylation of proSP-C was reported to take place before processing to mature SP-C starts distal to the trans-Golgi network (5). Combining these observations, we think that palmitoylation probably takes place in the ER-Golgi intermediate compartment (ERGIC) or cis-Golgi network. Also, p63, a protein which is localized in ER, was found to be palmitoylated only after the addition of BFA (23). The Golgi apparatus (44) and a pre-early Golgi compartment (16, 45) have both been implicated as sites where palmitoylation occurs. It seems likely that palmitoylation of proSP-C involves a palmitoyl acyltransferase, because palmitoylation was shown to occur presumably in the ERGIC or early Golgi compartment but not in the ER. The ER and Golgi compartment do not differ to such an extent in pH, lipid composition, etc., that it is likely to influence a nonenzymatic palmitoylation reaction. This strengthens the idea, which is still a matter of debate, that palmitoylation of proteins is generally an enzymatic process (20, 21, 46, 47).

In conclusion, we showed in this article that the reduction in palmitoylation of the proSP-C mutant in which the basic juxtamembrane residues, lysine and arginine, were substituted by glutamine residues is caused by an effect on the targeting of this mutant and not by a direct effect on the structural requirements for palmitoylation. Further research is required to determine which structural features of proSP-C are responsible for the fact that only cysteines residues 5 and 6 (numbering as in the sequence of mature SP-C) are palmitoylated.