help button home button
AJRCMB
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Published ahead of print on September 11, 2003, doi:10.1165/rcmb.2003-0220OC
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2003-0220OCv1
30/4/530    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Umland, S. P.
Right arrow Articles by Del Mastro, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Umland, S. P.
Right arrow Articles by Del Mastro, R.
American Journal of Respiratory Cell and Molecular Biology. Vol. 30, pp. 530-539, 2004
© 2004 American Thoracic Society
DOI: 10.1165/rcmb.2003-0220OC

Mouse ADAM33

Two Splice Variants Differ in Protein Maturation and Localization

Shelby P. Umland, Yuntao Wan, Himanshu Shah, Charles G. Garlisi, Kristine E. Devito, Karen Braunschweiger, Ferdous Gheyas and Richard Del Mastro

Departments of Allergy and of Statistics, Schering-Plough Research Institute, Kenilworth, New Jersey; and Genome Therapeutics Corporation, Waltham, Massachusetts

Address correspondence to: Shelby P. Umland, Ph.D., Schering-Plough Research Institute, 2015 Galloping Hill Road, K15-1-1700, Kenilworth, NJ 07033. E-mail: shelby.umland{at}spcorp.com


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
We compared the tissue mRNA prevalence and protein maturation of two splice variants of mouse ADAM33, a metalloprotease implicated in airway hyperresponsiveness. These variant cDNAs, designated 914 ({alpha}) and 906 (ß), encode membrane-bound forms that differ primarily in 26 residues (exon 17) between the cysteine-rich and epidermal growth factor–like domains. Proteins of ~ 120 and 103 kD, detectable by anti-ADAM33 antibodies, were expressed in 914-transfected HEK293 cells. The time-dependent appearance of the ~ 100-kD form and its inhibition by a peptidyl chloromethylketone, or the calcium ionophore, A23187, indicated that this was mature ADAM33, which was processed by a furin-like convertase. One form, ~ 110 kD, was detected in 906-transfected cell lysates. Trypsin and biotinylation treatment of transfected cells demonstrated that all of the mature ~ 100-kD, a minority of the ~ 120-kD pro-form, and none of the 906-expressed 110-kD form localized to the cell surface. The mature form was resistant to endoglycosidase Hf. The ~ 110-kD form was endoglycosidase Hf-sensitive, indicating retention proximal to the trans-Golgi, consistent with a lack of maturation. Quantitation of transcripts demonstrated that those containing exon 17 predominate, whereas those lacking exon 17 are negligible in the mouse lung, although detectable at low levels in mouse testis, heart, and brain. Thus, potential dominant-negative effects exerted by the nonprocessed 906-encoded ß splice variant are unlikely to occur in mouse lung.

Abbreviations: antibody, Ab • a disintegrin and metalloprotease, ADAM • dec-Arg-Val-Lys-Arg-chloromethyl ketone, decRVKR • epidermal growth factor, EGF • endoglycosidase Hf–resistant, Endo Hr • endoglycosidase Hf–sensitive, Endo Hs • endoplasmic reticulum, ER • human embryonic kidney cells, HEK • histidine, His • phosphate-buffered saline, PBS • polymerase chain reaction, PCR • peptide:N-glycosidase F-resistant, PNGase Fr • peptide:N-glycosidase–sensitive, PNGase Fs • treatment, Rx • streptavidin, SA • Tris-buffered saline, TBS • TBS with 0.1% Tween 20, TBS-T


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Recently human ADAM33, a new member of the ADAM (a disintegrin and metalloprotease) family, was identified as an asthma susceptibility gene through a genetic linkage and association study of families with asthma (1). Before this, a quantitative trait locus (bhr1) for airway hyperresponsiveness in mice was mapped to a chromosomal region in which mouse ADAM33 is located (2). Thus, the study of mouse ADAM33 can provide insights into the dysregulation of its human homolog in asthma.

ADAMs are a family of type I transmembrane zymogen proteins, which typically contain an N-terminal secretion signal and pro-domain, as well as metalloprotease, disintegrin-like, cysteine-rich, epidermal growth factor (EGF)-like, transmembrane, and cytoplasmic domains (3). A proposed major function of those ADAMs containing the active site consensus sequence (HEXXHXXGXXH) is the shedding of the ectodomains of membrane proteins (4). However, for very few ADAMs have specific substrates been identified and biologically validated (5). The predicted amino acid sequence of human and mouse ADAM33 (1, 6) includes the active site sequence, 345HEIGHSLGLSHD356, in the metalloprotease domain, suggesting catalytic activity. In addition, ADAM33 belongs to a subfamily (1, 6) in which others (ADAM12 [7], ADAM19 [9]) were shown to be catalytically active.

Biological functions have also been identified for other domains of ADAM proteins. These metalloproteases are synthesized intially as inactive zymogens that require post-translational processing to mature active forms. The pro-domain maintains latency by a cysteine-switch residue in which the thiol group is coordinated with the active site Zn2+ in the catalytic domain; catalytic activity subsequent to removal of the pro-domain has been demonstrated for ADAM9 (8), -12 (7), -17, and -19 (9). Processing to remove the pro-domain involves endoproteolytic cleavage at a RX(K/R)R consensus sequence located at the junction of the pro- and catalytic domains by a furin or furin-like proprotein convertase(s), which are serine peptidases (10, 11), and occurs in the secretory pathway. The pro-domain also functions to allow proper folding and transit through the secretory pathway (8, 12, 13).

Some discrete functions have been attributed to extracellular domains downstream of the catalytic domain of ADAMs. The disintegrin domain of several ADAMs can associate with receptors of the integrin family (14). Among those ADAMs that are largely expressed in somatic cells and are demonstrated or predicted to be catalytically active (14), the disintegrin domains of ADAM12 and ADAM15, which lack RGD motifs, bind to {alpha}9ß1 integrins and support cell–cell interactions (15). In addition, the disintegrin domain of ADAM9 mediates binding to the integrin {alpha}6ß1 on fibroblasts with subsequent changes in morphology and motility (16). Less is known regarding the cysteine-rich domain, and virtually nothing about the EGF-like domain, of ADAMs. Adhesion of the cysteine-rich domain to cell surface molecules, including fibronectin, syndecans, and integrins, has been described for ADAM12 and ADAM13 (14), and thus this domain, in concert with the disintegrin domain, may facilitate cell–cell and cell–extracellular matrix interactions (3). Of importance to catalytic activity, the cysteine-rich domain of ADAM13 has been shown to directly regulate ADAM13 protease activity in vivo (17). A role for the cysteine-rich domain in protease activation is also suggested by the finding that this domain of ADAM17 facilitates the dissociation of the prodomain from the catalytic domain in soluble recombinant forms of this metalloprotease (13).

Because ADAMs are modular proteins with several domains conferring and regulating function, splice variants lacking the consensus domain structure may differ functionally from full-length forms. Splice variants of several ADAMs have been described that have altered domain structure or lack specific domains. Splice variants encoding soluble forms lacking the transmembrane and cytoplasmic domains have been described for ADAM9 (18), ADAM12 (19) and ADAM28 (20, 21). In addition, an alternatively spliced form of mouse ADAM19 (meltrin ß) that lacks the prodomain, metalloprotease, and disintegrin domains (22) mediates a function distinct from the full-length form. Here, we compare the protein maturation and intracellular localization of full-length mouse ADAM33 to a splice variant, isolated from the lung, that lacks an exon at the junction of the predicted cysteine-rich and EGF-like domains. In addition, the prevalence of these mRNAs in mouse tissue was quantitated.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
A/J, AKR/J, BALB/cJ, DBA/2J, C57BL/6J, and C3H/HeJ 4- to 6-wk-old male mice were obtained from Jackson Laboratories (Bar Harbor, ME).

Isolation of Mouse ADAM33 cDNA
PCR primers were designed to generate RT-PCR products from the entire mouse ADAM33 gene based on consensus sequence assembled from sequence in the public domain (Unigene Mm. 44960) (primer 1: GCAAGAGTGAGAGGGAAAGGCACT; primer 2: ACAGCTCTCAGGGCAGGCGGGAT). cDNA was generated from mouse lung and brain polyA+ RNA (Clontech, Palo Alto, CA) from 8- to 12-wk-old Balb/c mice using a Superscript First Strand synthesis system (Lifetech). DNA was amplified using a GC Advantage kit (Clontech) with a 60°C annealing temperature. Products were separated on a 1% low melting agarose gel and bands of the expected size were excised from the gel. The purified products were cloned into the pCR II-TOPO vector and transformed into TOP10 chemically competent cells (Invitrogen Life Technologies, Carlsbad, CA) according to the manufacturer's recommendations. Clones were sequence verified. GenBank accession numbers for the 914 and 906 sequences, respectively, are AY382193 and AY382194.

cDNA Constructs
For expression in eukaryotic cells, several constructs were prepared as follows. The 914 and 906 cDNAs (GenBank accession no.) were subcloned into pcDNA3.1(-) (Invitrogen) containing the 6-His and V5 epitope tags and designated 914 or 906. Using the 914 pcDNA3.1(-) plasmid, the active site glutamic acid (347Glu)was mutated to an alanine (347Ala) using oligonucleotide primers and PfuTurbo DNA polymerase as described by the manufacturer (QuickChange Site-Directed Mutagenesis Kit; Stratagene, La Jolla, CA). The single base pair change from GAG to GCG generated a PvuI site that was used for diagnostic purposes. This plasmid was designated 914E347A. Constructs were verified by restriction endonuclease digestion and DNA sequencing. Plasmids were purified from Escherichia coli extracts using DNA affinity columns (Qiagen, Valencia, CA) as described by the manufacturer.

Cell Culture
Human embryonic kidney cells (HEK293) were obtained from Canji, Inc. (San Diego, CA). HEK293 cells were grown in complete media (Dulbecco's modified Eagle's medium; Gibco Invitrogen, Inc., Grand Island, NY), 10 mM Hepes, 10% fetal bovine serum, 50 U/ml penicillin, 50 µg/ml streptomycin, and 2 mM L-glutamine.

Transfections
One day before plasmid transfection, HEK293 cells were cultured in 6- or 12-well plates at 3.0 or 1.5 x 105 cells/well, respectively, in complete media at 37°C in 5% CO2. On the day of transfection, cells were washed with Dulbecco's modified Eagle's medium (serum- and antibiotic-free) and transfected with 0.8 µg plasmid DNA using Lipofectamine and Plus reagent (Invitrogen) as described by the manufacturer. After 5 h at 37°C, cells were rinsed and cultured in complete media containing 1x nonessential amino acids (Gibco) instead of serum. For transfected cell treatment with the furin inhibitor (23), decanoyl-Arg-Val-Lys-Arg-chloromethyl ketone (decRVKR; Bachem, Philadelphia, PA), two-thirds of the culture medium was replaced every 8 h with medium containing decRVKR to a final concentration of 30 µM and 0.1% methanol. Transfected cells were alternatively treated with the indicated concentrations of calcium ionophore, A23187 (Sigma Aldrich, St. Louis, MO), which was added only at the initiation of culture after transfection. Control cultures received the same treatment with medium containing a final concentration of 0.1% methanol (vol/vol) only. Transiently transfected cells were maintained in culture for 24–48 h, as indicated before harvest.

Antibody Production
Peptides containing amino-terminal cysteines were synthesized (Zymed Laboratories, Inc., South San Francisco, CA) corresponding to regions within the catalytic domain (amino acids 304 to 320; NH2-[C]GRTFQGTTVGLAPVEGIY-COOH) and cytoplasmic domain (amino acids 784 to 805; NH2-CQQRSHPPSLDLLSD-PGNSELT-COOH) of mouse ADAM33 (6). The peptides were chosen from regions with favorable predicted antigenicity, hydrophilicity, surface probability and flexibility (MacVector software; Eastman Kodak Co., Rochester, NY) and low similarity to other ADAMs. Peptides were covalently conjugated to keyhole limpet hemocyanin, mixed with adjuvant and each peptide injected into two NZW rabbits (Zymed). Antiserum collected from catalytic domain peptide-injected rabbits was purified with affinity columns prepared from unconjugated peptide. For preparation of a mouse monoclonal antibody to the cytoplasmic domain, KLH-conjugated peptide (NH2-CQQRSHPPSLDLLSDPGNSEL-COOH) was mixed with adjuvant and injected into BALB/c mice (Zymed). Serum was tested 8 wk later for reactivity to murine ADAM33. Splenocytes from the most reactive mouse were fused with myeloma cells and limiting dilution was used to purify antigen-reactive hybridoma cell lines. Ascites was prepared from mice injected with hybridoma cells and used without purification. The antibody designations are as follows: ASP2, rabbit polyclonal affinity purified antisera reactive to the catalytic domain peptide 304 -320; Cyt2, mouse monoclonal antibody (ascites) reactive to the cytoplasmic domain peptide 784–805.

Western Blotting
After transfection, the cell supernatants were collected after centrifugation and cell extracts were prepared from HEK293 cells as described (24). Lysis buffer contained 1% NP-40, 10 mM 1,10-phenanthroline, and 1x Complete Protease Inhibitor Cocktail (Roche, Indianapolis, IN) in Tris-buffered saline (TBS; 20 mM Tris-HCl, 500 mM NaCl, pH 7.5). Cell culture medium were concentrated 5 - 10-fold using Centriprep YM10 filter devices (Millipore, Bedford, MA), and lysates were used directly. Where indicated, samples were enriched for glycoproteins by ConA-sepharose (Amersham Pharmacia Biotech, Piscataway, NJ) as described (24). Samples were heat-denatured in gel loading buffer (50 mM Tris-HCl, pH 6.8, 5% glycerol, 2% SDS, 0.002% bromophenol blue dye, and 100 mM dithiolothreitol [DTT]) unless noted otherwise, and equivalent protein separated by electrophoresis through 7.5%, 10%, or 4–20% SDS-polyacrylamide gels (Bio-Rad, Hercules, CA). Molecular mass was estimated by comparison to Precision Plus (Bio-Rad), Kaleidoscope (Bio-Rad), or BenchMark (Invitrogen) prestained standards included on every gel. Proteins were transferred to PVDF membranes by electroblotting. For immunodetection, membranes were blocked with 5% ECL blocking reagent (Amersham) in TBS containing 0.1% Tween 20 (TBS-T) for 1 h at room temperature. After rinsing three times with TBS-T, membranes were incubated with antibodies (1–2 µg/ml affinity-purified anti-peptide antibodies (ASP2), 1:30,000 of ascites Cyt2, and 1:5,000 horseradish peroxidase–conjugated anti-histidine (-His) or anti-V5 (1:5,000) antibody (Invitrogen) for 1 h. Unbound antibodies were removed by washing with TBS-T, and if required, membranes were incubated with goat anti-rabbit IgG conjugated to horseradish peroxidase (1:25,000; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) for 30 min. Unbound antibodies were removed by washing and immobilized antigen–antibody complexes were detected with ECL chemiluminescence detecting reagents (Amersham) as described by the manufacturer. For quantitation of the ratio of the pro-form to mature forms of ADAM33, digital scanning of developed films was done with analysis using Quantity One software (Bio-Rad). Multiple films were acquired to assure bands did not saturate the film.

Antibodies described above were tested for reactivity with recombinantly expressed forms of ADAM33 and compared with reactivity of pre-immune serum (data not shown) and reactivity to lysates from cells transfected with vector control plasmid. Cyt2 but not preimmune sera detected nonprocessed and processed full-length ADAM33 but not recombinant versions of ADAM33 that did not contain the cytoplasmic domain. ASP2 but not preimmune sera detected all forms of ADAM33 that contained the catalytic domain. The ADAM33 forms detected were not observed in vector control samples.

Protein Deglycosylation
Proteins in cell lysates from transiently transfected cells were denatured by the addition of 0.5% SDS, 1% ß-mercaptoethanol for 10 min at 100°C before digestion with peptide:N-glycosidase F (PNGase F; New England BioLabs, Inc., Beverly, MA). Denatured samples were deglycosylated for 4 h at 37°C in reactions supplemented with 50 mM sodium citrate (pH 5.5) for endoglycosidase Hf (Endo H; New England Biolabs) or with 50 mM sodium phosphate (pH 7.5), 1% NP-40 for PNGase F. Control samples were handled similarly but without enzyme. Following deglycosylation, samples were heat-denatured with SDS, reduced with DTT and used for Western blotting.

Cell Surface Trypsin Treatment or Biotinylation
Forty-eight hours after transfection, washed cells were treated with trypsin as described (24). Cells were washed with cold phosphate-buffered saline (PBS) and treated with 500 µg/ml bovine pancreas trypsin (Sigma Aldrich) in PBS for 30 min at 4°C. Following a wash with 500 µg/ml soybean trypsin inhibitor (Sigma Aldrich) in PBS, the cells were collected and lysed as described above. Alternatively, washed cells were treated with sulfosuccinimidyl 6-(biotinamido)hexanoate (NHS-LC-biotin; Pierce Chemical, Rockford, IL), a non–membrane-permeant biotinylation reagent, at 0.5 mg/ml in PBS on ice for 45 min. Treated cells were washed first with 0.1 M glycine in TBS, followed by three washes with PBS and cell lysis (21). Biotinylated proteins were selected with strepavidin (SA)-Sepharose (Pharmacia).

Quantitative mRNA TaqMan Analysis
TaqMan analysis—mouse lung tissue. Total RNA was isolated from lung from several mouse strains using Qiagen RNeasy Midi kit (Qiagen) and tested for quality (ratio of 28S to 18S rRNA) and quantity using an Agilent 2,100 Bioanalyzer and RNA Labchips (Waldbroun, Germany). In addition, QUICK-CLONE cDNA from BALB/c mouse spleen, heart, smooth muscle, kidney, testis, and brain was purchased from Clontech. Real-time quantitative polymerase chain reaction (PCR) was done by the fluorogenic 5'-nuclease PCR method (TaqMan) with an ABI Prism 7,700 Sequence Detection System (Applied Biosystems). The TaqMan probes were labeled with the dye 6-carboxyfluorescein at the 5'-end of the sequence and with the quencher 6-carboxytetramethylrhodamine at the 3'-end; all were synthesized by Applied Biosystems or Biosource International (Camarillo, CA). The TaqMan reagents (Table 1) for mouse ADAM33 were designed using the 914 and 906 cDNA sequences. cDNA was synthesized using Superscript II reverse transcriptase (Invitrogen) with 5 µM oligo d(T)15 and 1.25 µM oligo p(dN)6 (Roche). A quantity of 10 ng of cDNA per sample was analyzed in a 50-µl reaction mixture. The final concentrations of the primers and probe in the PCR reactions were 200 nM and 100 nM, respectively. The PCRs were performed in duplicate in 96-well plates. The cycling conditions were: 50°C, 2 min; 95°C, 10 min; followed by 45 cycles at 95°C for 15 s and 60°C for 1 min. To compare the prevalence of ADAM33 splice variant mRNAs in a given tissue, absolute quantitation was done using a standard curve generated with five 10-fold serial dilutions of the pcDNA3.1(-) (Invitrogen) plasmid containing either the 914 or 906 cDNA starting at 0.1 or 1 ng plasmid per reaction. The amount of amplicon product was derived from the standard curve and is expressed as pg. To compare the amount of ADAM33 mRNA between mouse strains, 18 s RNA was quantitated using reagents from Applied Biosystems and used for normalization of cDNA input. The quantification and normalization of the amplicons per reaction was done according to the methods described in the User Bulletin No. 2 of ABI Prism 7,700 Sequence Detection System and expressed as 2(-{Delta}Ct). Normalized expression values were log-transformed. Statistical analysis was done by one-way ANOVA followed by pairwise comparison.


View this table:
[in this window]
[in a new window]
  		TABLE 1 Nonconservative amino acid differences in mouse and human ADAM33 cDNAs

 

   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
ADAM33 cDNAs were cloned by RT-PCR from BALB/c mouse brain and lung mRNA. Two clones were obtained, designated 914 and 906, which differed most prominently in the inclusion of exon 17 at the junction of the predicted cysteine-rich and EGF-like domains, with the brain-derived cDNA, 914, containing this exon and the lung-derived 906 cDNA lacking this exon. Nonconservative amino acid differences between the 914- and 906-encoded proteins occurred at amino acids 85 in the pro-domain, and 321 and 336 in the catalytic domain, and are summarized in Table 1. Our brain-derived 914 clone is nearly identical to a previously identified cDNA obtained from C57BL/6 mouse brain (Genbank AB059632) and designated the {alpha} form (6). Differences between these two clones resulting in nonconservative amino acid changes (Table 1) were at amino acids 68 and 85 in the pro-domain and at amino acid 409 at the 3' end of the catalytic domain. At each of these sites, the 914 sequence was identical to the amino acid sequence of human ADAM33 (Genbank AF466287 [1]). Similarly, our lung-derived 906 cDNA was nearly identical to a previously identified cDNA obtained from C57BL/6 brain (Genbank AB059633) lacking exon 17 and designated the ß splice variant (6). Several nonconservative amino acid differences occurred between these sequences at amino acids 68 and 85 in the pro-domain, and amino acids 321, 336, and 409 in the catalytic domain (Table 1). Importantly, the exon contained in 914 (and AB059632) and missing in 906 (and AB059633) contains 26 amino acids including four cysteines as well as a predicted N-linked glycosylation site. Hence, differences between these two variants should be observable at the protein level.

The 914 and 906 cDNAs were subcloned into the mammalian expression vector, pcDNA3.1(-). Transfection was done in HEK293 cells, which do not express ADAM33 mRNA. Using an antibody (ASP2) to a peptide within the catalytic domain of ADAM33, two proteins with the apparent mass of ~ 120 and ~ 100 kD were detected by SDS-PAGE in the 914-transfected cell lysates (Figure 1A, lane 2). These proteins were also detected by antibodies to the ADAM33 cytoplasmic domain, Cyt2, (data not shown) and to the carboxyl terminal 6-His tag (Figure 1A, lane 5). A similar pattern of proteins was expressed in HEK293 cells using the 914 construct in which the catalytic active site glutamic acid, 347Glu, was mutated to alanine, 347Ala (914E347A, Figure 1A, lanes 3 and 6). This mutation in human ADAM33 resulted in an inactive ADAM33 protease (25). It was observed that the two protein forms derived from the mutant 914E347A form migrated slightly faster, at ~ 114 and 98 kD, than those forms derived from the 914 construct. A similar observation has been made for E/A mutant form of the ADAM19 ectodomain compared with wild type (9). No ASP2-, anti-His-, Cyt2-reactive proteins were detected in concentrated cell culture medium from 914-, 914E347A-, or 906-transfected HEK293 cells (data not shown).



View larger version (57K):
[in this window]
[in a new window]
  		Figure 1. Expression of mouse ADAM33 protein in 914-transfected HEK293 cells. (A) HEK293 cells were transfected and 48 h cell lysates analyzed by SDS-PAGE (10% wt/vol polyacrylamide) and Western blotting with antibodies that recognize the catalytic domain, ASP2 (lanes 1–3), or anti-his antibody (lanes 4–6). X indicates presence of the indicated vector. Vector control, lanes 1 and 4; 914, lanes 2 and 5; 914E347A, lanes 3 and 6. (B) HEK293 cells were transfected; cell lyates were collected 24 and 48 h after transfection and analyzed by SDS-PAGE (7.5% polyacrylamide wt/vol) and Western blotting with anti-V5 Ab (lanes 1–4) and ASP2 (lanes 5–8). Vector control, lanes 3, 4, 7, and 8; 914, lanes 1, 2, 5, and 6. The film exposures for the anti-V5 Ab and ASP2 were 1 and 30 min, respectively. Asterisk indicates a nonspecific band. Molecular mass was estimated by comparison to size standards included on each gel where n = the number of independent SDS-PAGE gels analyzed and are as follows: 914 proform = 120 ± 3.8 kD, n = 8; 914 mature form = 103 ± 4, n = 8; 914E/A proform = 114 ± 0.6 kD, n = 3; 914E/A mature form = 98 ± 2.3 kD, n = 3. The ratio of the mature ~ 100-kD form to ~ 120-kD pro-form in 914-transfected HEK293 cell lysates at 48 h after transfection was 31 ± 4%, n = 3.

 
A comparison of cell lysates at 24 and 48 h after transfection showed significantly more of the smaller, heterogenous ~ 100 kD band, detected by both anti-V5 Ab (Figure 1B, lane 2 versus lane 1) and ASP2 (Figure 1B, lane 6 versus lane 5), at 48 than at 24 h. The ~ 100-kD band consisted of two bands of similar but distinguishable size with the slower or faster migrating bands predominating at 24 or 48 h, respectively, suggesting differential glycosylation. No proteins of these sizes were detected by these antibodies in similarly treated control vector transfected cell lysates (Figure 1B, lanes 3, 4, 7, and 8). This pattern of time-related increases in the expression of the ~ 100-kD proteins in cell lysates in concert with similar kinetics for human ADAM33, where the larger form was recognized by an antibody to the human pro-domain (25), indicates that the larger ~ 120-kD form is likely to be the full-length unprocessed mouse ADAM33, whereas the smaller ~ 100-kD bands are the mature forms generated by proteolytic cleavage of the pro-domain.

To provide further evidence that these forms differed due to the presence of the pro-domain, cells transfected with 914 and 914E347A were treated with decRVKR, a peptidyl chloromethylketone inhibitor of the protease activity of furin-like proprotein convertases (23). Inhibition of the appearance of the ~ 100-kD forms was seen at both 24 h (Figure 2, lanes 1–6) and 48 h (Figure 2, lanes 7–12). Due to the minimal presence of the ~ 100-kD forms at 24 h, the decRVKR-mediated inhibition of these forms, although detectable (Figure 2, lane 3 versus lane 2), was less evident than at 48 h (lane 9 versus lane 8), but at each time point, the faster migrating of the two ~ 100-kD bands was most inhibited. At both 24 (lane 6 versus lane 5) and 48 h (lane 12 versus lane 11), inhibition of the formation of the ~ 100-kD forms in the 914E347A transfected samples was very significant. The appearance of the ~ 100-kD forms in 914E347A transfected cells and their inhibition by decRVKR indicate that autocatalysis is not involved in pro-domain removal of ADAM33. A dose-dependent inhibition by the calcium ionophore, A23187, (Figure 2B, lanes 3–6) of the ~ 100-kD form was also observed. This is consistent with furin-mediated maturation of mouse ADAM33 because the autoactivation of furin to an active peptidase is calcium-dependent and inhibitable by A23187 (26).



View larger version (46K):
[in this window]
[in a new window]
  		Figure 2. Maturation of mouse ADAM33 is mediated by a calcium-dependent furin-like convertase. (A) decRVKR-CMK inhibits the maturation of mouse ADAM33. HEK293 cells were transfected and cultured in the absence (lanes 1, 2, 5, 7, 8, and 11) and presence (lanes 3, 4, 6, 9, 10, and 12) of the furin inhibitor, decRVKR-CMK (30 µM in 0.1% methanol) for 24 (lanes 1–6) or 48 h (lanes 7–12). Control vector cultures contained 0.1% methanol only. X indicates presence of the indicated vector. Vector control, lanes 1, 4, 7, and 10; 914, lanes 2, 3, 8, and 9; 914E347A, lanes 5, 6, 11, and 12. Asterisk indicates a nonspecific band. (B) Maturation of mouse ADAM33 is calcium-dependent. 914-transfected HEK293 cells were cultured in the absence (lane 1) or presence of the calcium ionphore, A23187 (lanes 3–6), or the solvent control, 0.1% methanol (lanes 2) for 48 h. Cell lysates analyzed by SDS-PAGE (7.5%) and Western blotting with Cyt2. The results in A and B are each representative of two independent experiments.

 
Cell lysates from HEK293 cells transfected with the splice form, 906, were compared with that of 914-transfected cells. With the very sensitive anti-V5 Ab, only one band of ~ 110 kD was detected in cell lysates from 906-transfected cells (Figure 3A, lane 2), which had slightly increased mobility compared with the ~ 120-kD band (Figure 3A, lane 1) detected in 914-transfected cell lysates. This is consistent with its correspondence to a pro-form with a predicted smaller size due to lacking 26 amino acids in the EGF-like domain. The 26 amino acids encoded by exon 17 predicts a mass of 2.9 kD; an N-linked glycosylation site (NAT) occurs within this sequence. Anti-His (data not shown) and -Cyt2 Abs (see Figure 4, lane 1) also detected only one form. Further comparison of the forms detected in 914- and 906-transfected cell lysates in nonreduced and reduced states was done because anti-ADAM antibodies (27) are variously effective and detect different forms under these conditions. In lysates from 914- and 914E347A-transfected cells, the ~ 120- and ~100-kD proteins detected by anti-V5 (Figure 3B, lanes 1 and 3), migrated significantly faster in the absence of DTT (Figure 3B, lanes 2 and 4). A similar pattern was found with Cyt2 (data not shown), but not ASP2, which did not detect nonreduced ADAM33. The anti-V5–reactive protein in nonreduced 906 cell lysate also migrated faster than did the reduced protein (Figure 3B, lane 6 versus lane 5). This mobility difference is typical of cysteine-rich proteins and has been described for other ADAMs (24). Interestingly, the proportion of mature form to pro-form detected was lower in the reduced 914 and 914E347A lysates (Figure 3B, lanes 1 and 3) than in the corresponding nonreduced samples (Figure 3B, lanes 2 and 4), indicating that the anti-V5 Ab efficiently recognized nonreduced mature mouse ADAM33. Additionally, two forms of > 199 kD were detected in the absence of DTT in both the 914- and 914E347A-transfected cell lysates with anti-V5 Ab (Figure 3B, lanes 2 and 4) and Cyt2 (data not shown). These forms may represent multimers of mouse ADAM33 or association with other proteins.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 3. Comparison of 914- and 906-expressed reduced and nonreduced ADAM33 protein. (A) HEK293 cells were transfected and 48 h cell lysates analyzed by Western blotting with anti-V5 Ab. X indicates presence of the indicated vector. 914, lane 1; 906, lane 2. (B) Before SDS-PAGE, cell lysates were boiled in gel loading buffer with (+, lanes 1, 3, and 5) or without DTT (-, lanes 2, 4, and 6) and the samples analyzed by Western blotting with anti-V5 Ab. X indicates presence of the indicated vector. 914, lanes 1 and 2; 914E347A, lanes 3 and 4; 906, lanes 5 and 6. Molecular mass of the 906-encoded protein form was estimated by comparison to size standards included on each gel where n = the number of independent SDS-PAGE gels analyzed: 906 proform = 111 ± 2.3 kD, n = 3.

 


View larger version (26K):
[in this window]
[in a new window]
  		Figure 4. Deglycosylation of 906- and 914-expressed ADAM33 protein. Cell lysates obtained 48 h after transfection of HEK293 cells with 906 (lanes 1–3) or 914 (lanes 4–6) were treated with PNGase F (lanes 2 and 5) or Endo H (lanes 3 and 6) or mock treated (Control, lanes 1 and 4) as described in MATERIALS AND METHODS. SDS-PAGE (7.5% wt/vol polyacrylamide) was followed by Western blotting with Cyt2. Film exposure was 1 min. The Western blot shown is representative of at least three independent deglycoslyation treatments of lysates from two independent tranfection cell lysates.

 
Mouse ADAM33 encoded by the 914 cDNA contains six potential sites for asparagine-linked (N-linked) oligosaccharide modification: one within the pro-domain, three within the catalytic domain, one in the disintegrin domain, and one within the EGF-like domain, which is encoded by exon 17. The splice form encoded by 906 cDNA contains these predicted glycosylation sites with the exception of that in the EGF-like domain, encoded by exon 17, which is lacking in this splice form. To observe this experimentally, cell lysates from 914- and 906-transfected cells were treated with peptide:N-glycosidase F (PNGase F) and endoglycosidase Hf (Endo H). PNGase F removes most or all N-linked carbohydrate residues between the innermost N-acetylglucosamine and asparagine (28). Endo H cleaves high mannose-type oligosaccharides but not complex glycans (29); acquisition of resistance to Endo H is indicative of a glycoprotein reaching the medial-Golgi compartment, where conversion of high mannose to complex glycans occurs (28). Treatment of lysates from 906-transfected cells with either Endo H (Figure 4, lane 3 versus lane 1) or PNGase F (Figure 4, lane 2 versus lane 1) demonstrated one band of ~ 100 kD; in contrast to the untreated lysate (Figure 4, lane 1), a ~ 110-kD form was not detected by Cyt2 (Figure 4, lanes 2 and 3) or ASP2 (data not shown) in either the Endo H– or PNGase F–treated samples. This indicated that the 906-expressed protein was both Endo Hs and PNGase Fs. Endo H treatment of lysates from 914-transfected cells (Figure 4, lane 6 versus lane 4) showed one clear band of ~ 100 kD, suggesting that the 914 proform was Endo Hs and the mature form was Endo Hr. The ~ 100-kD band in the Endo H–treated sample comigrated with that of the untreated sample and exhibited increased intensity. With PNGase F treatment, forms of ~ 100 kD and 80 kD were detected by Cyt2 (Figure 4, lane 5) and ASP2 (data not shown); this is consistent with these forms corresponding to the pro-form and mature forms, respectively. This is supported by the lack of the 80-kD form in the PNGase F–treated 906 cell lysates, which is consistent with the absence of exon 17 and a potential N-linked glycosylation site present in that exon.

To determine whether ADAM33 was present on the cell surface or was located intracellularly, 914- and 906-transfected cells were treated with trypsin on ice and the cell lysates immunoblotted as above. Untreated and trypsin-treated lysates from 906-transfected cells (Figure 5A, lane 3 versus lane 4) both showed a single band of ~ 110 kD reactive with anti-V5 Ab. Mobility did not differ between the 906 trypsin-treated and untreated cells, although the intensity of the trypsin-treated was slightly less. In contrast, trypsin treatment of 914-transfected cells resulted in the disappearance of the ~ 100-kD band (Figure 5A, lane 5 versus lane 6). The ~ 120-kD band decreased in intensity in trypsin-treated cells, but showed no change in mobility (lane 6) compared with untreated cells (lane 5). With ASP2 as the detecting Ab, identical results were obtained in trypsin-treated versus untreated 906 cell lysates (Figure 5, lane 8 versus lane 7), whereas neither the pro-form nor the mature form in 914 cell lysates (Figure 5, lane 10 versus lane 9) was detected. The apparent discrepancy in detecting the pro-form in 914 cell lysates (lane 6 versus lane 10) is likely due to the much greater sensitivity provided by the anti-V5 (lane 6) compared with the ASP2 Ab (lane 10). Each of these results was confirmed in additional independent experiments. Disappearance of anti-V5 reactivity with the carboxyl terminal V5 epitope suggests destabilization of the protein following trypsin digestion. Biotinylation of only cell surface–expressed proteins in 906- and 914-transfected HEK293 cells was also done to determine the localization of the ADAM33 forms (Figure 5B). No protein forms were detected following SA-sepharose selection of biotinylated cell-surface proteins in 906-transfected cell lysates (Figure 5B, lane 4 versus lane 3). In contrast, in lysates from similarly treated 914-transfected cells, a prominent band of ~ 100 kD and a minor band of ~ 120 kD (Figure 5B, lane 6 versus lane 5) were detected. These results, together with those from cell trypsinization, strongly suggest that 906- and 914-expressed ADAM33 protein differ in their cellular localization, with the former being largely intracellular and inaccessible to cell trypsinization or biotinylation, whereas all of the mature form and a minority of the immature form expressed by 914 was located on the cell surface.



View larger version (43K):
[in this window]
[in a new window]
  		Figure 5. Cell surface trypsinization or biotinylation of 906- or 914-transfected HEK293 cells. (A) Forty-eight hours after transfection with 906 or 914 constructs, cells were treated on ice with bovine pancreas trypsin or PBS (mock) followed by washing with soybean trypsin inhibitor. Mock treated (lanes 1, 3, 5, 7, and 9) or trypsin-treated cell lysates (lanes 2, 4, 6, 8, 10, and 11) were analyzed by SDS-PAGE (7.5% wt/vol polyacrylamide) and Western blotting with anti-V5 (lanes 1–6) or ASP2 (lanes 7–12) Ab. Control vector (indicated by X), lanes 1, 2, and 11; 906, lanes 3, 4, 7, and 8; 914, lanes 5, 6, and 9–11. (B) Cell surface biotinylation of 906- or 914-transfected HEK293 cells. Forty-eight hours after transfection with 906 or 914 constructs, cells were treated in PBS on ice with NHS-LC-biotin for 45 min. Control vector (indicated by X), lanes 1 and 2; 906, lanes 3 and 4; and 914, lanes 5 and 6. Transfected cells were treated with NHS-LC-biotin (lanes 2, 4, and 6) or not (lanes 1, 3, and 5); Biotinylated proteins in the transfected cell lysates were selected by SA-sepharose. SDS-PAGE (7.5% wt/vol polyacrylamide) was followed by Western blotting with Cyt2. Film exposure was 30 min. The results in A and B are each representative of two independent trypsinization or biotinylation experiments in which the cell lysates were analyzed with at least two antibodies.

 
The studies above predict that the 906-encoded ADAM33 protein, unlike the 914-encoded protein, would be catalytically inactive and retained in the endoplasmic reticulum (ER) or early-Golgi compartments. Thus, the prevalence of these two forms could determine the overall activity of mouse ADAM33. To assess this quantitatively, we measured levels of ADAM33 mRNA transcripts containing, or not, exon 17. This was done by the real-time quantitative fluorogenic 5'-nuclease polymerase chain reaction (PCR) method (TaqMan) using two diagnostic sets of primers and probes (Figure 6A, Table 2). The exon 17–specific set (exon 17+) consisted of primers (1948F/2020R) that amplified a fragment detectable by a probe (1973T) within exon 17 of the transcript (Figure 6A, lower panel; Table 2), and therefore would not detect transcripts lacking exon 17. The second set of primers (1903F/1978R), designated exon 17-, was designed using the 906 sequence to amplify a fragment detectable by a probe (exonPR-jctT) spanning the exon 16–to–exon 18 junction (Figure 6A, upper panel; Table 2). This set detects the exon 16/18 junction as in 906 or ß cDNAs (6) and does not detect cDNAs containing exon 17. The exon 17- and exon 17+ reagents had similar amplification efficiency (slope = -3.7 ± 0.09 and -3.5 ± 0.09, n = 3, respectively) as did a third primer (2059F/2123R) and probe (2081T) set detecting a sequence in exon 18 in the cysteine-rich domain of mouse ADAM33 (slope = -3.6 ± 0.16, n = 4; Table 2). The relative prevalence of transcripts containing or lacking exon 17 was determined using total lung RNA obtained from each of three mice of six different mouse strains (A/J, C3H/HeJ, AKR/J, DBA/2J, BALB/cJ, and C57BL/6J). Figure 6B shows the comparison of lung mRNA levels in the six different strains of mice using the exon 17+ and exon 18 reagents. (Here, the 914 plasmid was used for the standard curve.) A small difference in the levels detected with the exon 17+ and exon 18+ reagents was noted for each strain, although the expression pattern across the mouse strains was similar for the two reagent sets. A similar comparison of mRNA levels of the six different strains of mice was done using the exon 17- reagents and those to the cysteine-rich region (Figure 6C). (Here, the 906 plasmid was used for the standard curve.) Transcripts lacking exon 17 were rare in the lung of all strains of mice (Figure 6C), whereas the same pattern was obtained using the exon 18+ reagents as with the 914 plasmid as the standard curve (Figure 6B). Statistical analysis of the exon 18+ transcripts indicated that the levels in A/J mice (geometric mean, 95% confidence interval = 5.6 x 104, 4.48 - 7.0 x 104) did not differ significantly (P = 0.15) from those in C3H/HeJmice (7.0 x 104, 5.6 - 8.8 x 104), which are two strains that differ in native (30) or allergic (31) airway responsiveness.




View larger version (25K):
[in this window]
[in a new window]
  		Figure 6. Quantitation of ADAM33 transcripts, containing or not exon 17, in mouse lung. (A) Map of TaqMan primers and probes (Table 2) in relationship to exon organization. Top panel shows the exon 17- primer set with exon organization of 906 sequence. Bottom panel shows the exon 17+ primer set, exon 17- primers and exon 18+ primer set (Table 2) in relationship to exon organization of 914 sequence. Arrows of similar pattern detect exon 17-, exon 17+, or exon 18+ transcripts. The number and letters within the exon boxes indicates the exon according to (6) and (1), respectively. (B) Quantitation of exon 17+ (striped bars) and exon 18+ (open bars) transcripts in lung from six strains of mice using the 914 plasmid for the standard curve. (C) Quantitation of exon 17- (solid bars) and exon 18+ (open bars) transcripts in lung from six strains of mice using the 906 plasmid for the standard curve. Mouse strains (n = 3) are indicated on the x-axis. Values are expressed as pg amplicon product derived from the standard curves. (D) Quantitation of exon 17- (solid bars) and exon 18+ (open bars) transcripts in BALB/c mouse spleen, heart, smooth muscle, kidney, testis, and brain using the 906 plasmid for the standard curve.

 

View this table:
[in this window]
[in a new window]
  		TABLE 2 Diagnostic primers used to quantitate ADAM33 mRNAs containing or not containing exon 17

 
Analysis of other mouse tissues with exon 18+ and exon 17- primer sets (Figure 6D) indicated low levels of transcripts lacking exon 17 in testis, heart, and brain, and much higher levels of transcripts containing exon 17. Of the tissues tested, transcripts containing exon 17 were highest in smooth muscle and testis, somewhat lower in brain, heart, and spleen, and lowest in kidney. Interestingly, smooth muscle, which expressed the highest level of exon 18+ transcripts, expressed no exon 17- transcripts. Because no sequence variation was noted within the region detected by the exon 17- reagents in the 906, 914, or {alpha} and ß splice forms to suggest that the exon 17- reagents did not work, this strongly suggests that the exon 17+ containing transcripts predominate in mouse tissue relative to transcripts lacking this exon.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We present here a comparative study of two forms of mouse ADAM33, a new member of the ADAM family of membrane-anchored glycoprotein metalloproteases. These splice variants encode membrane-bound forms that differ in their nucleotide sequences primarily by their inclusion or not of an exon between the cysteine-rich and EGF-like regions. Human homologs of each of these mouse forms have been described (6). We provide evidence that these splice variants differ at the protein level (i) in the extent of processing to a mature form, (ii) in their sensitivity to enzymatic deglycosylation and trafficking in the secretory pathway, (iii) in their pattern of cell surface or intracellular location, and (iv) in their prevalence in several mouse tissues, at the mRNA level. This comparative study is of importance because human ADAM33 was recently identified as an asthma gene by a genetic linkage and association study of asthmatic sib-pairs (1). Moreover, in earlier studies of the genetics of airway hyperresponsiveness in mice, a quantitative trait locus (bhr1) for this phenotype was mapped to a chromosomal region within which mouse ADAM33 is located (2).

The two ADAM33 cDNA splice variants studied here were isolated from mouse brain and lung tissue and designated 914 and 906, respectively. These clones differed most prominently in the inclusion of exon 17 in the EGF-like domain, with the brain-derived cDNA, 914, containing this exon and the lung-derived 906 cDNA lacking this exon. The 906 cDNA, is nearly identical to a previously identified cDNA also lacking exon 17, and designated the ß splice variant (6). It is unlikely that these latter forms represent tissue- or mouse strain–restricted splice variants, because the 906 cDNA was cloned from BALB/c lung, whereas the ß splice variant (6) was cloned from C57BL/6 brain (Genbank AB059633). Our brain-derived 914 clone, containing exon 17, is also nearly identical to a previously identified cDNA isolated from C57BL/6 mouse brain (Genbank AB059633) and designated the {alpha} form (6). Several differences between these two sequences resulting in nonconservative amino acid changes exist in the predicted pro-domain (amino acid 68: Lys in 914, Pro in AB059632; amino acid 85: Asn in 914, Lys in AB059632) and at the 3' end of the predicted catalytic domain (amino acid 409: Val in 914, Thr in AB059633). It is noteworthy that at each of these sites, the 914 (and 906) sequence was identical (or similar at 409) to the amino acid sequence of human ADAM33 (Genbank AF466287).

ADAM33 protein, expressed following transfection of HEK293 cells with the 914 construct, consisted of two major forms of ~ 120 and 100 kD, each detectable with antibodies specific to peptides in the catalytic or cytoplasmic domains, or to the carboxyl-terminal 6-His or V5 epitope tags. These molecular masses are consistent with those expressed by the human homolog of mouse 914 (25), of which the larger but not the smaller was detected by an antibody to the pro-domain of human ADAM33, indicating that it was the nonprocessed pro-form. Further evidence that the ~ 120- and 100-kD forms correspond to the latent pro-form and processed mature forms of mouse ADAM33 was provided by the time-dependent disappearance and appearance of the ~ 120-kD and ~ 100-kD forms, respectively. Pulse-chase studies to further define the temporal relationship between the ~ 120-kD and ~ 100-kD forms were precluded because the anti-peptide antibodies ASP2 and Cyt2 were not effective for immunoprecipitation. However, additional evidence was provided by the dec-RVKR–mediated inhibition of the ~ 100-kD form. decRVKR, a peptidyl chloromethylketone containing a consensus furin cleavage sequence, covalently modifies the substrate-binding site and inhibits the protease activity of furin-like proprotein convertases that are active in the secretory pathway (23). Autoactivation of furin-like proprotein convertases is calcium-dependent (26) and mouse ADAM33 maturation was inhibited by a calcium ionophore. Importantly, the ~ 100-kD form was also expressed by the 914E347A mutant, was more pronounced at 48 than 24 h, and was inhibitable by decRVKR. This mutation in the active site renders human ADAM33 catalytically inactive (25) as does a similar mutation in ADAM9 (8), ADAM12 (7), ADAM19 (9). A role for autocatalysis in this processing, demonstrated for ADAM8 (32) and ADAM28 (21) which lack the recognition sequence for furin-like proteases, is not evident. A minor point of interest is the observation that the ratio of total cell-associated mature to proform protein for both mouse and human ADAM33 (25) was always higher for the E/A mutant than for the non-mutated construct (914E/A, Figure 1A, lane 3 versus lane 2; Figure 2, lane 5 versus lane 2, lane 11 versus lane 8; Figure 3B, lane 3 versus lane 2). Most simply, this suggests that some autodegradation of the mature form occurs.

Thus, mouse ADAM33, like ADAM9 (8), -10 (33), -12 (7), -15 (24), and -17 (27) is processed to a mature form by removal of the N-terminal pro-domain, by a furin-like proprotein convertase. This cleavage is likely to occur in the mouse ADAM33 protein at the preferred site, RX(K/R)R, containing consecutive basic residues 200RVRREAR, within a sequence with overlapping furin cleavage sites. Supporting this is the finding that the amino-terminal sequence of soluble catalytic domain protein of human ADAM33 was 209EARRT, where cleavage occurred at the first of three overlapping cleavage sites (25).

Evidence was provided by both cell-surface trypsinization and biotinylation that the ~ 100-kD mature form of mouse ADAM33 was largely located on the cell surface. The mature form was accessible to cell trypsinization with loss of Ab reactivity, including Ab reactivity to the C-terminal intracellular V5 epitope, suggesting destabilization of the protein following trypsinization. Biotinylation of cell surface proteins resulted in the detection of a significant amount of the ~ 100-kD mature form and a minimal amount of the ~ 120-kD pro-form, a pattern also displayed by ADAM17 (27). These findings, together with detection of the pro-form in cell lysates following cell trypsinization, indicates that a majority of the mouse ADAM33 pro-form was located intracellularly when expressed ectopically in HEK293 cells, whereas a majority of the mature form was expressed on the cell surface. The latter differs from the mature form of human ADAM33, which was largely located intracellularly (25), and points to possible differences in function between the homologs. However, for both mouse and human ADAM33, the subcellular location of each remains to be defined in nontransfected cells.

In contrast, ADAM33 protein expressed from the splice variant 906 consisted of only one form of ~ 110 kD, whose intensity and mobility was unaffected by cell trypsinization or biotinylation. Thus, this form was exclusively located intracellularly. This is further supported by the sensitivity of this ~ 110-kD form to the deglycosidase activity of Endo H. Because acquisition of resistance to Endo H is indicative of a glycoprotein reaching the medial-Golgi compartment, the 906-expressed ~ 110-kD form is localized in the ER and proximal Golgi. In contrast, the 914-expressed ~ 100-kD mature form was Endo Hr, indicating that it had traversed the medial-Golgi compartment and ultimately a significant fraction of this form located to the plasma membrane. The lack of maturation of 906-encoded protein and its retention in early Golgi compartments is likely caused by the absence of the twenty-six amino acids encoded by exon 17. This exon contains four cysteines which may be involved in disulfide bond formation and correct folding of the protein, upon which trafficking through the secretory pathway depends (8). It is also worth noting that the 906- and 914-encoded proteins differ in the number of cysteine residues within the catalytic domain, which may also affect correct folding and ultimately catalytic activity.

The lack of maturation of the 906-encoded protein and its probable localization to the ER suggests that it may have dominant-negative activity. For example, dominant negative effects on basal processing of neuregulin-1 were mediated by a mutant form of ADAM19, with the catalytic site-inactivating mutations H346A and H350A, which did not undergo processing (34). Thus, the biological activity of mouse ADAM33 could be affected by the prevalence and tissue distribution of the 906-encoded form relative to 914-encoded ADAM33. To address the question of prevalence of these forms, quantitation of ADAM33 mRNAs in mouse lung was done. This analysis was conducted in mouse lung from six genetically different strains of mice because airway responsiveness in mice has a genetic component and a quantitative trait locus (bhr1) for noninflammatory airway hyperresponsiveness was mapped to a region on chromosome 2 in which mouse ADAM33 is located (2). Although we did not directly quantitate full-length native transcripts encoded by mRNAs corresponding to the 906 and 914 cDNAs, nevertheless, transcripts lacking exon 17 were rare in lung from all six mouse strains, compared with those containing exon 17 or exon 18. In contrast to the lung, the splice variant lacking exon 17 was detectable in testis, heart, and brain, although each had much higher levels of transcripts containing exon 17. Further analysis of these transcripts and encoded ADAM33 protein isoforms in specific cell types within these organs will be necessary to understand the importance of this splice variant. Because we have been unable to detect endogenous ADAM33 protein in tissues expressing the highest levels of mRNA (heart and lung), even after pooling multiple organ lysates followed by enrichment by ConA sepharose, further studies will necessitate the development of other antibodies.

In conclusion, several points can be made. With regard to a role for ADAM33 in mouse lung function, it is unlikely that the ß splice variant is a contributor based upon its negligible expression, despite its intriguing potential to have dominant-negative activity. However, it is likely other splice variants exist and will require mRNA and protein analyses similar to those described here. Second, a correlation of ADAM33 mRNA levels in the lung and airway hyperresponsivness (30, 31) was not observed. Demonstration of a correlation of ADAM33 and strain-dependent differences in airway hyperresponsiveness may require further study at the level of protein function.


   Acknowledgments
 
The authors thank Dr. George Kong for statistical analyses and Drs. Motasim Billah and Robert Egan for their support and many useful suggestions.

Received in original form June 12, 2003

Received in final form August 25, 2003


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Van Eerdewegh, P., R. D. Little, J. Dupuis, R. G. Del Mastro, K. Falls, J. Simon, D. Torrey, S. Pandit, J. McKenny, K. Braunschweiger, A. Walsh, Z. Liu, B. Hayward, C. Folz, S. P. Manning, A. Bawa, L. Saracino, M. Thackston, Y. Benchekroun, N. Capparell, M. Wang, R. Adair, Y. Feng, J. Dubois, M. G. Fitzferald, H. Huang, R. Gibson, K. M. Allen, A. Pedan, M. R. Danzig, S. P. Umland, R. W. Egan, F. M. Cuss, S. Rorke, J. B. Clough, J. W. Holloway, S. T. Holgate, and T. P. Keith. 2002. Association of the ADAM33 gene with asthma and bronchial hyperresponsiveness. Nature 418:426–430.[CrossRef][Medline]
  2. De Sanctis, G. T. 1995. Quantitative locus analysis of airway hyperresponsiveness in A/J and C57BL/6J mice. Nat. Genet. 11:150–154.[CrossRef][Medline]
  3. Schlöndorff, J., and C. P. Blobel. 1999. Metalloprotease-disintegrins: modular proteins capable of promoting cell-cell interactions and triggering signals by protein-ectodomain shedding. J. Cell Sci. 112:3603–3617.[Abstract]
  4. Primakoff, P., and D. G. Myles. 2000. The ADAM gene family. Surface proteins with adhesion and protease activity. Trends Genet. 16:83–87.[CrossRef][Medline]
  5. Blobel, C. P. 2000. Remakable roles of proteolysis on and beyond the cell surface. Curr. Opin. Cell Biol. 12:606–612.[CrossRef][Medline]
  6. Yoshinaka, T., K. Nishii, K. Yamada, H. Sawada, E. Nishiwaki, K. Smith, K. Yoshino, H. Ishiguro, and S. Higashiyama. 2002. Identification and characterization of novel mouse and human ADAM33s with potential metalloprotease activity. Gene 282:227–236.[CrossRef][Medline]
  7. Loechel, F., B. J. Gilpen, E. Engvall, R. Albrechtsen, and U. M. Wewer. 1998. Human ADAM12 (meltrin {alpha}) is an active metalloprotease. J. Biol. Chem. 273:16993–16997.[Abstract/Free Full Text]
  8. Roghani, M., J. D. Becherer, M. L. Moss, R. E. Atherton, H. Erdjument-Bromage, J. Arribas, R. K. Blackburn, G. Weskamp, P. Tempst, and C. P. Blobel. 1999. Metalloprotease-disintegrin MDC9: intracellular maturation and catalytic activity. J. Biol. Chem. 274:3531–3540.[Abstract/Free Full Text]
  9. Chesneau, V., J. D. Becherer, Y. Zheng, H. Erdjument-Bromage, P. Tempst, and C. P. Blobel. 2003. Catalytic properties of ADAM19. J. Biol. Chem. 278:25933–25939.[Abstract/Free Full Text]
  10. Nakayama, K. 1997. Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing a wide variey of precursor proteins. Biochem. J. 327:625–635.
  11. Steiner, D. F. 1998. The proprotein convertases. Curr. Opin. Chem. Biol. 2:31–39.[CrossRef][Medline]
  12. Loechel, F., M. T. Overgaard, C. Oxvig, R. Albrechtsen, and U. M. Wewer. 1999. Regulation of human ADAM12 protease by the prodomain: evidence for a functional cysteine switch. J. Biol. Chem. 274:13427–13433.[Abstract/Free Full Text]
  13. Milla, M. E., M. A. Leesnitzer, M. L. Moss, W. C. Clay, H. L. Carter, A. B. Miller, J.-L. Su, M. H. Lambert, D. H. Willard, D. M. Sheeley, T. A. Kost, W. Burkhart, M. Moyer, R. K. Blackburn, G. L. Pahel, J. L. Mitchell, C. R. Hoffman, and J. D. Becherer. 1999. Specific sequence elements are required for the expression of functional tumor necrosis factor-{alpha}-converting enzyme (TACE). J. Biol. Chem. 274:30563–30570.[Abstract/Free Full Text]
  14. Seals, D. F., and S. A. Courtneidge. 2003. The ADAMs family of metalloproteases: multidomain proteins with multiple functions. Genes Dev. 17:7–30.[Free Full Text]
  15. Eto, K., W. Puzon-McLaughlin, D. Sheppard, A. Sehara-Fujisawa, X.-P. Zhang, and Y. Takada. 2000. RGD-independent binding of integrin {alpha}9ß1 to the ADAM-12 and -15 disintegrin domains mediates cell-cell interactions. J. Biol. Chem. 275:34922–34930.[Abstract/Free Full Text]
  16. Nath, D., P. M. Slocombe, A. Webster, P. E. Stephens, A. J. P. Docherty, and G. Murphy. 2000. Meltrin {gamma} (ADAM-9) mediates cellular adhesion through {alpha}6ß1 integrin, leading to a marked induction of fibroblast motility. J. Cell Sci. 113:2319–2328.[Abstract]
  17. Smith, K. M., A. Gaultier, H. Cousin, D. Alfandari, J. M. White, and D. W. DeSimone. 2002. The cysteine-rich domain regulates ADAM protease function in vivo. J. Cell Biol. 159:893–902.[Abstract/Free Full Text]
  18. Hotada, N., H. Koike, N. Sasagawa, and S. Ishiura. 2002. A secreted from of human ADAM19 has an {alpha}-secretase activity for APP. Biochem. Biophys. Res. Commun. 293:800–805.[CrossRef][Medline]
  19. Gilpen, B. J., F. Loechel, M.-G. Mattei, E. Engvall, R. Albrechtsen, and U. M. Wewer. 1998. A novel, secreted form of human ADAM 12 (Meltrin {alpha}) provokes myogenesis in vivo. J. Biol. Chem. 273:157–166.[Abstract/Free Full Text]
  20. Roberts, C. M., P. H. Tani, L. C. Bridges, Z. Laszik, and R. D. Bowditch. 1999. MDC-L, a novel metalloprotease disintegrin cysteine-rich protein family member expressed in human lymphocytes. J. Biol. Chem. 274:29251–29259.[Abstract/Free Full Text]
  21. Howard, L., R. A. Maciewicz, and C. P. Blobel. 2000. Cloning and characterization of ADAM28: evidence for autocatalytic pro-domain removal and for cell surface localization of mature ADAM28. Biochem. J. 348:21–27.
  22. Kurisaki, T., S. Wakatsuki, and A. Sehara-Fujisawa. 2002. Meltrin ß mini, a new ADAM19 isoform lacking metalloprotease and disintegrin domains, induces morphological changes in neuronal cells. FEBS Lett. 532:419–422.[CrossRef][Medline]
  23. Stieneke-Grober, A., M. Vey, H. Angliker, E. Shaw, G. Thomas, C. Roberts, H.-D. Klenk, and W. Garten. 1992. Influenza virus hemagglutinin with multibasic cleavage site is activated by furin, a subtilisin-like endoprotease. EMBO J. 11:2407–2424.[Medline]
  24. Lum, L., M. S. Reid, and C. P. Blobel. 1998. Intracellular maturation of the mouse metalloprotease disintegrin MDC15. J. Biol. Chem. 273:26236–26247.[Abstract/Free Full Text]
  25. Garlisi, C. G., J. Zou, K. E. Devito, F. Tian, F. X. Zhu, J. Liu, H. Shah, Y. Wan, M. M. Billah, R. W. Egan, and S. P. Umland. 2003. Human ADAM33: protein maturation and localization. Biochem. Biophys. Res. Commun. 301:35–43.[CrossRef][Medline]
  26. Fujiwara, T., T. K. Oda, S. Yokota, A. Takatsuki, and I. Y. 1988. Brefeldin A causes disassembly of the Golgi complex and accumulation of secretory proteins in the endoplasmic reticulum. J. Biol. Chem. 263:18545–18552.[Abstract/Free Full Text]
  27. Schlöndorff, J., J. D. Becherer, and C. P. Blobel. 2000. Intracellular maturation and localization of the tumour necrosis factor {alpha} convertase (TACE). Biochem. J. 347:131–138.
  28. Maley, F., R. B. Trimble, A. L. Tarantino, and T. H. Plummer. 1989. Characterization of glycoproteins and their associated oligosaccharides through use of endoglycosidases. Anal. Biochem. 180:195–204.[CrossRef][Medline]
  29. Robbins, P. W., R. B. Trimble, D. F. Wirth, C. Hering, F. Maley, G. F. Maley, R. Das, B. W. Gibson, N. Royal, and K. Biemann. 1984. Primary structure of the Streptomyces enzyme endo-bet-N-acetylglucosaminidase H. J. Biol. Chem. 259:7577–7583.[Abstract/Free Full Text]
  30. De Sanctis, G. T., and J. M. Drazen. 1996. Genetics of native airway responsiveness in mice. Am. J. Respir. Crit. Care Med. 156:582–588.
  31. Ewart, S. L., D. Kuperman, E. Schadt, A. Grupe, D. M. Shubitowski, G. Peltz, and M. Wills-Karp. 2000. Quantitative trait loci controlling allergen-induced hyperresponsiveness in inbred mice. Am. J. Respir. Cell Mol. Biol. 23:537–545.[Abstract/Free Full Text]
  32. Schlomann, U., D. Wildeboer, A. Webster, O. Antropova, D. Zeuschner, C. G. Knight, A. J. P. Docherty, M. Lambert, L. Skelton, H. Jockusch, and J. W. Bartsch. 2002. The metalloprotease-disintegrin ADAM8: processing by autocatalysis is required for proteolytic activity and cell adhesion. J. Biol. Chem. 277:48210–48219.[Abstract/Free Full Text]
  33. Anders, A., S. Gilbert, W. Garten, R. Postina, and F. Fahrenholz. 2001. Regulation of the alpha-secretase ADAM10 by its prodomain and proprotein convertases. FASEB J. 15:1837–1839.[Free Full Text]
  34. Shirakabe, K., S. Wakatsuki, T. Kurisaki, and A. Fujisawa-Sehara. 2001. Roles of meltrin ß/ADAM19 in the processing of neuregulin. J. Biol. Chem. 276:9352–9358.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2003-0220OCv1
30/4/530    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Umland, S. P.
Right arrow Articles by Del Mastro, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Umland, S. P.
Right arrow Articles by Del Mastro, R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Proc. Am. Thorac. Soc. Am. J. Respir. Crit. Care Med.
Copyright © 2004 American Thoracic Society.