Effect of carbohydrate position on lysosomal transport of procathepsin L

Abstract

To study the role of carbohydrate in lysosomal protein transport, we engineered two novel glycosylation signals (Asn-X-Ser/Thr) into the cDNA of human procathepsin L, a lysosomal acid protease. We constructed six mutant cDNAs encoding glycosylation signals at mutant sites Asn-138, Asn-175, or both sites together, in the presence or absence of the wild-type Asn-204 site. We stably transfected wild-type and mutant cDNAs into NIH3T3 mouse fibroblasts and then used species-specific antibodies to determine the glycosylation status, phosphorylation, localization, and transport kinetics of recombinant human procathepsin L containing one, two, or three glycosylation sites. Both novel glycosylation sites were capable of being glycosylated, although Asn-175 was utilized only 30-50% of the time. Like the wild-type glycosylation at Asn-204, carbohydrates at Asn-138 and Asn-175 were completely sensitive to endoglycosidase H, and they were phosphorylated. Mutant proteins containing two carbohydrates were capable of being delivered to lysosomes, but there was not a consistent relationship between the efficiency of lysosomal delivery and carbohydrate content of the protein. Pulse-chase labeling revealed a unique biosynthetic pattern for proteins carrying the Asn-175 glycosylation sequence. Whereas wild-type procathepsin L and mutants bearing carbohydrate at Asn-138 appeared in lysosomes by about 60 min, proteins with carbohydrate at Asn-175 were processed to a lysosome-like polypeptide within 15 min. Temperature shift, brefeldin A, and NH4Cl experiments suggested that the rapid processing did not occur in the endoplasmic reticulum and that Asn-175 mutants could interact with the mannose 6-phosphate receptor. Taken together, our results are consistent with the interpretation that Asn-175 carbohydrate confers rapid transport to lysosomes. We may have identified a recognition domain in procathepsin L that is important for its interactions with the cellular transport machinery.

Figure 1

Representations of (pro)cathepsin L. (A) Linear map of proCL showing the profragment (pro) and both segments of the mature enzyme (25K + 5K). On top of the map are the amino acid positions of the glycosylation sites used in this study: 204,N3 is the single wild-type site in human proCL; 138 (N1) and 175 (N2) are two novel sites introduced in this work. Directly below the map are the wild-type sequences at residues 138–140, 175–177, and 204–206 and the mutations at those residues that either create (138 and 175) or destroy (204) glycosylation signals. The bottom of the panel shows the names of the relevant mutants and their corresponding amino acids encoded at positions 138, 175, and 204, with mutations shown in boldface. (B) Drawing of a three-dimensional computer model of cathepsin L. The model was developed as described in MATERIALS AND METHODS. Numbers and arrows designate the positions of amino acids relevant to this study.

Figure 2

EndoH digestion of recombinant proCLs. Cell lines stably expressing wild type (proCL) or mutants encoding the wild-type glycosylation site plus a novel glycosylation site at Asn-138 (N1N3) or Asn-175 (N2N3). Cells were radiolabeled for 5 h with [³⁵S]methionine, cell lysates were incubated in the presence (+) or absence (−) of endoH to remove high-mannose carbohydrate, and immunoprecipitations were performed with antiserum specific for human proCL. Immunoprecipitated proteins were run on SDS-PAGE and processed for fluorography as described in MATERIALS AND METHODS. Proenzyme (pro), lysosomal single-chain (mature), and two-chain (25) forms of the proteins are indicated. The 5-kDa carboxy-terminal fragment from the two-chain enzyme was run off the gel. The position of the 29-kDa size standard is shown on the left and lane numbers are provided on the bottom.

Figure 3

Phosphorylation of novel carbohydrates. Cell lines stably expressing wild type (proCL) or mutants containing only the novel glycosylation sites at Asn-138 (N1Q3) or Asn-175 (N2Q3). Cells were radiolabeled for 4–5 h with either [³⁵S]methionine (S) or [³²P]orthophosphate (P). Cell lysates were subjected to immunoprecipitation and SDS-PAGE, as previously, and processed for fluorography or autoradiography, respectively. The proenzymes (pro) and the processed, single-chain proteins (mature) are indicated. The position of the 29-kDa size standard is shown on the left and lane numbers are provided on the bottom.

Figure 4

Steady-state distribution of recombinant proteins. Individual clones of cells stably expressing wild-type (proCL) or mutant proteins, as indicated. Cells were radiolabeled for 4 h with [³²P]orthophosphate (left-hand panel in A and B) or [³⁵S]methionine (right-hand panel in A and B, both panels in C). Cell lysates (C) and culture media (M) were collected and subjected to immunoprecipitation, as previously described. The proenzymes (pro) and the processed lysosomal proteins (mature) are indicated by brackets. The positions of size markers are indicated on the left of each panel and lane numbers are provided on the bottom.

Figure 5

Pulse-chase analysis. (A–E) Cell lines stably expressing recombinant proteins were pulse-labeled for 15 min with [³⁵S]methionine and then chased with unlabeled methionine for the indicated times. Cell lysates were collected at each time point and subjected to immunoprecipitation as previously. (A) Wild-type proCL; (B) N1N3; (C) N1Q3; (D) N2N3; (E) N2Q3. (F) Lysates from the 2h time point of proCL and N2Q3 cells were treatd with endoF (+F), endoH (+H), or left untreated (−) before immunoprecipitation. Brackets on the right of each panel indicate the positions of proenzymes. The number of carbohydrates on each proenzyme are included in panels D and E. Asterisks indicate the processed, lysosome-like bands for the respective recombinants, along with the number of carbohydrates present on each of those proteins. Positions of the 29-kDa standard are shown.

Figure 6

BFA treatment. Cell lines stably expressing the indicated recombinants were radiolabeled for 3 h with [³⁵S]methionine in the absence (−) or presence (+) of BFA (10 μg/ml) to block transport in the endoplasmic reticulum. Cell lysates were collected and subjected to immunoprecipitation, as previously. Proenzyme (pro) and processed (mature) proteins are indicated by brackets on both sides of the image. The position of the 29-kDa size standard is shown and lane numbers are provided on the bottom.

Figure 7

Ammonium chloride treatment. Cell lines stably expressing proCL, N2N3, or N2Q3 recombinants. Cells were radiolabeled for 5 h with [³⁵S]methionine in the absence (−) or presence (+) of NH₄Cl (10 mM). Cell lysates (C) and culture media (M) were collected and subjected to immunoprecipitation with anti-human proCL antiserum (last three panels). In the left panel (Mouse), anti-mouse proCL antiserum was used to immunoprecipitate endogenous protein from lysates of cells expressing the N2N3 recombinant, using 10-fold less sample than in the N2N3 panel. Positions of proenzymes and processed proteins (along with the number of carbohydrates on each protein) are indicated, as previously. The 29-kDa size standard is shown on the left and lane numbers are provided below each panel.

Figure 8

Tunicamycin treatment. (A) Transfected cell lines were incubated in the presence (+) or absence (−) of tunicamycin (tunic.) for 4 h and then radiolabeled for 3 h with [³⁵S]methionine (+/- tunic.). Cell lysates were immunoprecipitated with anti-human proCL antiserum or anti-mouse proCL antiserum, as indicated, followed by SDS-PAGE and fluorography. (B) Three human cell lines, as indicated, were treated with tunicamycin (+) or left untreated (−) and radiolabeled with [³⁵S]methionine. Cell lysates (C) or media (M) were immunoprecipitated with anti-human proCL antiserum, followed by SDS-PAGE. (C) Human SK-Hep cells were transfected with wild-type mouse proCL (mproCL) or with a mutant lacking both of its normal glycosylation signals (mQ204/251). Left, Populations of cells (pop.) or two individual clones from each population (c1 and c2) were radiolabeled and cell lysates were immunoprecipitated with anti-mouse proCL antiserum. Right, NIH3T3 cells expressing authentic mouse proCL or populations of SK-Hep cells expressing wild-type or mutant recombinant mouse proCLs, as indicated, were radiolabeled with [³⁵S]methionine. Cell lysates (C) and media (M) were immunoprecipitated with anti-mouse proCL antiserum. Brackets mark the positions of proenzymes (pro). The positions of the small amount of unglycosylated, lysosomal proteins are indicated (*,0). The 29-kDa size standard is shown on the left and lane numbers are provided below the panels.