A novel monoclonal antibody to characterize pathogenic polymers in liver disease associated with α1-antitrypsin deficiency

Authors

  • Elena Miranda,

    Corresponding author
    1. Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/Medical Research Council Building, Cambridge, UK
    2. Department of Cellular Biology and Development, Università di Roma La Sapienza, Rome, Italy
    • Dipartimento Biologia Cellulare e dello Sviluppo, Università di Roma La Sapienza, Piazzale Aldo Moro 5, Roma, 00185, Italy
    Search for more papers by this author
    • fax: +39 06 4991 2351

  • Juan Pérez,

    1. Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/Medical Research Council Building, Cambridge, UK
    2. Department of Cellular Biology, Genetics, and Physiology, Universidad de Málaga, Facultad de Ciencias, Málaga, Spain
    Search for more papers by this author
  • Ugo I. Ekeowa,

    1. Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/Medical Research Council Building, Cambridge, UK
    Search for more papers by this author
  • Nedim Hadzic,

    1. Institute of Liver Studies, King's College Hospital, London, UK
    Search for more papers by this author
  • Noor Kalsheker,

    1. Division of Clinical Chemistry, Queen's Medical Centre, University of Nottingham, Nottingham, UK
    Search for more papers by this author
  • Bibek Gooptu,

    1. Crystallography, Birkbeck College, Institute of Structural and Molecular Biology, London, UK
    Search for more papers by this author
  • Bernard Portmann,

    1. Institute of Liver Studies, King's College Hospital, London, UK
    Search for more papers by this author
  • Didier Belorgey,

    1. Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/Medical Research Council Building, Cambridge, UK
    Search for more papers by this author
  • Marian Hill,

    1. Division of Clinical Chemistry, Queen's Medical Centre, University of Nottingham, Nottingham, UK
    Search for more papers by this author
  • Susan Chambers,

    1. Department of Clinical Biochemistry, King's College Hospital, London, UK
    Search for more papers by this author
  • Jeff Teckman,

    1. St Louis University School of Medicine, Cardinal Glennon Children's Hospital, St Louis, MO
    Search for more papers by this author
  • Graeme J. Alexander,

    1. Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/Medical Research Council Building, Cambridge, UK
    Search for more papers by this author
  • Stefan J. Marciniak,

    1. Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/Medical Research Council Building, Cambridge, UK
    Search for more papers by this author
  • David A. Lomas

    1. Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/Medical Research Council Building, Cambridge, UK
    Search for more papers by this author

  • Potential conflict of interest: Dr. Lomas is a consultant for, advises, is on the speakers' bureau of, and received grants from GlaxoSmithKline. He is also a consultant for Novartis and Amicus.

Abstract

Alpha1-antitrypsin is the most abundant circulating protease inhibitor. The severe Z deficiency allele (Glu342Lys) causes the protein to undergo a conformational transition and form ordered polymers that are retained within hepatocytes. This causes neonatal hepatitis, cirrhosis, and hepatocellular carcinoma. We have developed a conformation-specific monoclonal antibody (2C1) that recognizes the pathological polymers formed by α1-antitrypsin. This antibody was used to characterize the Z variant and a novel shutter domain mutant (His334Asp; α1-antitrypsin King's) identified in a 6-week-old boy who presented with prolonged jaundice. His334Asp α1-antitrypsin rapidly forms polymers that accumulate within the endoplasmic reticulum and show delayed secretion when compared to the wild-type M α1-antitrypsin. The 2C1 antibody recognizes polymers formed by Z and His334Asp α1-antitrypsin despite the mutations directing their effects on different parts of the protein. This antibody also recognized polymers formed by the Siiyama (Ser53Phe) and Brescia (Gly225Arg) mutants, which also mediate their effects on the shutter region of α1-antitrypsin. Conclusion: Z and shutter domain mutants of α1-antitrypsin form polymers with a shared epitope and so are likely to have a similar structure. HEPATOLOGY 2010

The serpinopathies are conformational diseases characterized by the polymerization and intracellular retention of members of the serine protease inhibitor or serpin superfamily of proteins.1 The best known is α1-antitrypsin deficiency, with the most common severe deficiency allele being the Z mutation (Glu342Lys). This mutation results in the retention of ordered polymers of α1-antitrypsin as periodic acid Schiff positive inclusion bodies within the endoplasmic reticulum (ER) of hepatocytes.2 These inclusions predispose the individual homozygous for the Z variant of the α1-antitrypsin protease inhibitor (PI*Z) to neonatal hepatitis, cirrhosis, and rarely, hepatocellular carcinoma.3 Deficiency of circulating α1-antitrypsin results in early onset panlobular emphysema.4

The Z mutation of α1-antitrypsin lies between the head of strand 5A and the base of the mobile reactive center loop5 (Fig. 1). Other mutations that cause α1-antitrypsin deficiency cluster around the shutter region of the protein (Fig. 1). In the classical model of serpin polymerization, these mutations are believed to open β-sheet A, giving rise to a polymerogenic intermediate that has been termed M*.6, 7 The patent β-sheet A then accepts the reactive loop of a second α1-antitrypsin molecule to form a dimer, which can extend into chains of reactive center loop-β-sheet A polymers.2, 6, 8-12 The recent crystal structure of a dimer of another serpin, antithrombin, demonstrated a linkage between a β-hairpin of the reactive loop and strand 5A of one molecule and β-sheet A of another. This dimer was used as the basis of a novel model for the polymer in which helix I is unravelled and the proteins are linked by a β-hairpin containing the reactive center loop and strand 5A.13

Figure 1.

Pathogenic mutations and polymerization in α1-antitrypsin. In the classical model of serpin polymerization, pathogenic mutations destabilize the native fold and increase the population of a highly polymerogenic monomeric intermediate species. Intermolecular linkage occurs via insertion of the reactive site loop (red) of one molecule into the opened β-sheet A (blue) of a second molecule. Propagation of this single β-strand–mediated linkage (red dotted arrows) leads to the formation of polymer chains. The most severe polymerogenic mutations in α1-antitrypsin are Glu342Lys (or Z) (black arrow) and mutations in the shutter region that regulates opening of β-sheet A (enlarged box, upper panel). The highly conserved nature of the shutter region across the serpin superfamily is seen by comparison with the neuron-specific serpin neuroserpin (box, lower panel). Critical residues in this region are vulnerable to mutations that cause severe disease. The novel mutation in the index case affects a key site of shutter region stabilization (His334). This is homologous to one that affects neuroserpin to give severe, juvenile-onset dementia.

The ATZ11 monoclonal antibody (mAb) recognizes polymers and inclusions of α1-antitrypsin formed by the Z mutation but not those that result from mutations in the shutter region.14 It is unclear whether this is a consequence of the Z and shutter domain mutants forming different types of polymers. We have addressed this issue by generating a novel conformational mAb (mAb 2C1) that is specific for polymers of α1-antitrypsin. Our antibody recognized polymers formed by Z α1-antitrypsin in vivo. It also recognizes polymers formed by the Siiyama (Ser53Phe) and Brescia (Gly225Arg) mutants, and the novel His334Asp shutter domain mutant of α1-antitrypsin that is associated with prolonged neonatal jaundice in a 6-week-old boy. These data show that Z and shutter domain mutants form polymers with a shared epitope and so they are likely to have a similar structure.

Abbreviations

ELISA, enzyme-linked immunosorbent assay; ER, endoplasmic reticulum; HRP, horseradish peroxidase; IgG, immunoglobulin G; mAb, monoclonal antibody; PI*Z, Z variant of the α1-antitrypsin protease inhibitor; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis.

Materials and Methods

Reagents and Antibodies.

Reagents, buffers, culture media, and serum for cell cultures were from Sigma-Aldrich Co. (Dorset, UK) unless stated otherwise. Goat polyclonal anti-calreticulin antibody was from Santa Cruz Biotechnology (through Autogen Bioclear, Mile Elm Calne, UK). Mouse monoclonal anti-GM130 antibody was from BD Biosciences Pharmingen (Oxford, UK). The eukaryotic initiation factor 2 alpha (eIF2α) antibody15 was a kind gift from David Ron, New York University. Goat polyclonal anti-rabbit IgG (horseradish peroxidase [HRP]) and goat and rabbit polyclonal anti-mouse IgG (HRP) antibodies were from Sigma-Aldrich Co. Mouse monoclonal (704) anti-α1-antitrypsin, anti-rabbit IgG (conjugated with tetramethyl rhodamine isothiocyanate) and anti-mouse IgG (conjugated with fluorescein isothiocyanate) antibodies were from Abcam (Cambridge, UK).

Expression of α1-Antitrypsin in COS-7 Cells.

The plasmids expressing human Z (Glu342Lys), King's (His334Asp), and Siiyama (Ser53Phe) α1-antitrypsin were generated using the Quikchange mutagenesis kit (Stratagene, La Jolla, CA) from pcDNA containing wild-type α1-antitrypsin.16 The plasmid expressing Brescia (Gly225Arg) α1-antitrypsin was a kind gift from Anna Fra, University of Brescia. The transfection of COS-7 cells, sodium dodecyl sulfate (SDS), and nondenaturing PAGE followed by western blot analysis, confocal microscopy, metabolic labeling, and immunoprecipitation were performed as detailed previously.17

Production of the Anti-α1-Antitrypsin mAbs.

Six mice were immunized with Z α1-antitrypsin polymers prepared from α1-antitrypsin purified from the plasma of PI*Z homozygotes.18 The production of hybridoma cell lines was carried out as described.19 The hybridoma clones were screened by antigen-mediated enzyme-linked immunosorbent assay (ELISA) using purified Z α1-antitrypsin monomer and polymer as the antigen. Selected clones were subcloned by limited dilution and expanded as cell lines.

Sandwich ELISA.

All steps were carried out at room temperature and using 50 μL/well unless otherwise stated. Plates (Costar 3590; Corning Inc., Lowell, MA) were coated overnight with antigen purified rabbit polyclonal anti-α1-antitrypsin antibody at 2 μg/mL in 0.2 M phosphate-buffered saline (PBS), pH 7.4. The wells were then washed with 0.9% wt/vol NaCl, 0.05% vol/vol Tween 20 and blocked for 2 hours with 300 μL/well of blocking buffer (PBS, 0.25% wt/vol bovine serum albumin, 0.05% vol/vol Tween20, 0.025% wt/vol sodium azide). Standards (plasma purified M or Z α1-antitrypsin) and unknown samples were diluted in blocking buffer and incubated for 2 hours. After washing, the wells were incubated with either the 9C5 or 2C1 mAbs (1 μg/mL) diluted in blocking buffer for 2 hours. Bound mAbs were detected with rabbit anti-mouse HRP antibody (1:20,000 in blocking buffer without sodium azide) for 1 hour. The reaction was developed for 10 minutes with TMB substrate solution (Sigma-Aldrich Co., Dorset, UK), stopped with 1 M H2SO4, and then HRP activity was measured in a plate reader (Molecular Devices, Thermo-max microplate reader) at 450 nm.

Preparation of α1-Antitrypsin Conformers.

Complexes were prepared by incubating Z α1-antitrypsin at a 1:1 ratio with bovine trypsin for 30 minutes at room temperature. Reactive center loop cleaved Z α1-antitrypsin was prepared by incubation with porcine pancreatic elastase at a 10:1 (elastase:α1-antitrypsin) ratio for 30 minutes at room temperature. M and Z α1-antitrypsin polymers were prepared by heating the monomeric protein (0.2 mg/mL) at 60°C in PBS (pH 7.4) for 1 hour. All α1-antitrypsin polymers were confirmed by nondenaturing PAGE.

Purification of Z α1-Antitrypsin from Hepatic Inclusions.

Z α1-antitrypsin was purified from hepatic inclusions as described20 and confirmed by nondenaturing PAGE.

Results

Development of the 2C1 mAb that Is Specific for α1-Antitrypsin Polymers.

Mice were immunized with polymers of Z α1-antitrypsin prepared with protein purified from human plasma and polymerized by heating at 60°C for 1 hour. Hybridoma clones were screened for the recognition of Z α1-antitrypsin polymers in antigen-mediated ELISA, and 10 mAbs were selected for further characterization by sandwich ELISA. Most of these recognized all conformers of α1-antitrypsin but two were of particular interest: mAb 9C5 recognized all conformers with high affinity (Fig. 2A, left graph), whereas mAb 2C1 only showed high affinity for polymers formed by heating M or Z α1-antitrypsin at 60°C (Fig. 2A, middle graph) or Z α1-antitrypsin at 41°C (data not shown). It did not recognize α1-antitrypsin as a native monomer, complexed with trypsin or cleaved at the reactive center loop. None of the mAbs detected polymers of another serpin, neuroserpin (data not shown). The 2C1 mAb was more sensitive and more specific than the existing ATZ11 mAb,8 which recognized Z α1-antitrypsin complexed with trypsin or cleaved at the reactive center loop with a greater affinity than polymers. The ATZ11 mAb did not recognize polymers of M α1-antitrypsin (Fig. 2A, right graph). Furthermore, the 2C1 mAb was able to detect polymers in serum samples from PI*Z α1-antitrypsin homozygote individuals, whereas it gave no signal in serum samples from control subjects (results not shown).

Figure 2.

A new mAb specific for polymers of α1-antitrypsin. (A) Two new mAbs to α1-antitrypsin (α1AT) polymers were tested in sandwich ELISA with increasing concentrations of different antigens (see graph legends). The previously described mAb ATZ11 was included for comparison. The mAb 9C5 showed similar affinities for all the antigens, but only mAb 2C1 showed high affinity for polymers. The mAb ATZ11 detected α1-antitrypsin complexed with trypsin and cleaved at the reactive center loop more readily than polymers of Z α1-antitrypsin. It did not detect polymers of M α1-antitrypsin. (B) Equal quantities of monomeric (mon) and polymeric (pol) M and Z α1-antitrypsin and α1-antitrypsin extracted from liver inclusions from a PI*ZZ individual were separated by nondenaturing PAGE and subjected to western blot analysis. The membrane was first blotted with mAb 2C1, which recognized polymers from all three sources but not the M or Z monomers (left panel), and then with mAb 2D1, which recognized all the species (right panel).

The affinity of the 2C1 antibody for Z α1-antitrypsin polymers was entirely abolished by the Gly117Phe mutation when the double variant (Gly117Phe/Z) was transiently expressed in COS-7 cells (data not shown). This was despite the expression of the Gly117Phe/Z mutant resulting in comparable levels of polymeric material as Z α1-antitrypsin in cell lysates and supernatants of transfected cells.21

When used in western blot analysis of nondenaturing PAGE, the 2C1 antibody detected polymers of M and Z α1-antitrypsin formed by heating in vitro and, most importantly, polymers of Z α1-antitrypsin isolated from the liver of an individual with α1-antitrypsin deficiency (Fig. 2B, left panel). It did not detect the monomeric form of either variant. Another mAb generated at the same time, 2D1, detected all the species present on the same membrane with similar intensities (Fig. 2B, right panel). The 2C1 mAb did not recognize the stable monomeric latent conformer of α1-antitrypsin22 when assessed by western blot analysis (results not shown).

The mAb 2C1 was also assessed in immunocytochemistry. COS-7 cells were transfected with cDNA for M or Z α1-antitrypsin and costained with a rabbit polyclonal antibody that detects all α1-antitrypsin (Fig. 3A, red) and the 2C1 mAb (Fig. 3A, green, overlap in yellow). Although all cells showed strong staining with the polyclonal antibody, only cells expressing polymerogenic Z α1-antitrypsin showed a strong signal with mAb 2C1, indicating that it also recognized polymers in this technique. Moreover, the 2C1 mAb was able to detect pathological polymers in paraffin-embedded liver sections from PI*Z homozygotes (Fig. 3B, middle panel).

Figure 3.

The mAb 2C1 detects polymers of α1-antitrypsin in immunocytochemistry. (A) COS-7 cells were transiently transfected with M, Z, or H334D α1-antitrypsin (α1AT), fixed, and immunostained with mAb 2C1 (green) and with a polyclonal antibody that recognizes all forms of α1-antitrypsin (red). Only the merged panels are shown, with overlapping signals in yellow. The DNA is stained blue (DAPI [4′,6-diamidino-2-phenylindole]). The polyclonal antibody detected α1-antitrypsin in cells expressing the M, Z His334Asp protein, whereas mAb 2C1 only reacted with Z and H334D α1-antitrypsin contained in perinuclear inclusions (green and yellow staining middle and right panels). Scale bar: 10 μm. (B) Immunostaining of paraffin-embedded liver sections with mAb 2C1. The mAb detected polymers of α1-antitrypsin within hepatocytes of a PI*ZZ individual (middle panel, black arrows) and the index case for His334Asp α1-antitrypsin (right panel, black arrowheads) but gave no signal in a control liver (left panel). Scale bar: 10 μm.

Our new mAb 2C1 is thus a powerful tool for the study of polymerization of α1-antitrypsin both in vivo and in experimental models of α1-antitrypsin deficiency.

Polymers Formed by a Shutter Region Mutant of α1-Antitrypsin that Causes Liver Disease Are Similar to Those Formed by Z α1-Antitrypsin.

Inclusions of α1-antitrypsin can also form in the liver as a result of mutations in the shutter region, which is distinct from the Z variant (Fig. 1). The 2C1 antibody was used to investigate a novel severe shutter region mutant that we have termed α1-antitrypsin King's. This mutation came to clinical attention in a 6-week-old Caucasian boy presented with prolonged neonatal jaundice. He was born after an uneventful dichorionic and diamniotic twin pregnancy and elective caesarean section at 34 weeks gestation. Physical examination revealed mild jaundice but was otherwise normal. His paternal grandfather was recently diagnosed with unresectable hepatocellular carcinoma. Laboratory tests revealed total bilirubin of 117 μmol/L (reference range: <20 μmol/L), conjugated bilirubin of 30 μmol/L, alanine aminotransferase of 25 IU/L (reference range: <50 IU/L), gamma-glutamyl transferase of 212 IU/L (reference range: <55 IU/L), alkaline phosphatase of 797 IU/L (reference range: <350 IU/L), albumin of 42 g/L (reference range: 35-50 g/L), and international normalized prothrombin time of 0.96 (reference range: 0.8-1.2). His abdominal ultrasound scan was unremarkable and serological tests for toxoplasmosis, herpes group viruses, rubella, and syphilis were negative. The serum α1-antitrypsin concentration was 15.7 mg/dL (reference range 100-200 mg/dL). All other investigations were unremarkable. Percutaneous liver biopsy was performed at 118 days of age and showed mild nonspecific portal inflammation, minimal portal fibrosis and discrete focal periportal steatosis. Periodic acid-Schiff (PAS) diastase–resistant staining revealed the presence of granules in keeping with the retention of α1-antitrypsin in a periportal distribution (Fig. 4). The jaundice resolved by 4 months of age and at 12-month follow-up, the boy continued to develop normally, with normal biochemical indices, apart from a mild elevation of alanine aminotransferase (70 IU/L). There were no clinical signs of chronic liver disease.

Figure 4.

Inclusions of α1-antitrypsin in the index case of a new variant associated with liver disease. Histological analysis of a liver biopsy from the index case. Left panels: Diastase-periodic acid-Schiff (D-PAS) staining showing intracellular diastase-resistant inclusions (examples are highlighted with black arrows) in hepatic cells. Right panels: Staining with an anti-α1-antitrypsin (α1AT) polyclonal antibody showing the retention of α1-antitrypsin within hepatocytes (general brown staining and open arrows highlighting inclusions). The bottom panels show the same specimens as the top panels, but at higher magnification. Scale bar: 10 μm.

Alpha1-antitrypsin phenotyping of the child by isoelectric focusing was consistent with the PI*Z phenotype, but genotyping by the Elucigene ARMS technique (Tepnel Diagnostics) confirmed only a single Z allele; this discrepancy suggested the presence of an unusual variant. Gene sequencing revealed a novel His334Asp mutation in exon 5 (1072C→G) in both the index case and his mother (Table 1). Phenotyping of the father demonstrated a Pi*SZ phenotype with the paternal grandmother and grandfather being PI*MS and PI*MZ, respectively.

Table 1. Genotyping of the Index Case and His Parents
 Mutation AssessedIndex Case (Z/King's)Father (S/Z)Mother (M/King's)
  1. The novel His334Asp mutation was inherited with a Pi*M1Val haplotype. The origin of the alleles in the index case is shown in bold.

S alleleGlu264ValGlu264 + Glu264Val264 + Glu264Glu264 + Glu264
Z alleleGlu342LysGlu342 + Lys342Glu342 + Lys342Glu342 + Glu342
Novel changeHis334AspAsp334 + His334His334 + His334His334 + Asp334

The His334Asp variant is striking in that it is homologous to a mutant of the neurone specific serpin neuroserpin that causes polymerization, ER retention and the dementia familial encephalopathy with neuroserpin inclusion bodies or FENIB19, 23 (Fig. 1). His334Asp α1-antitrypsin was therefore assessed by transient transfection in COS-7 cells in comparison to wild-type M and mutant Z α1-antitrypsin. Cell lysates and culture medium supernatants were assessed by western blot analysis of SDS and nondenaturing PAGE (Fig. 5A). After SDS-PAGE, α1-antitrypsin was present as a 52 kDa band (in keeping with ER glycosylation) in the cell lysates (C), and a fully glycosylated 55 kDa band in the supernatant (M) (Fig. 5A, upper panel). Cells expressing M α1-antitrypsin showed little signal in the cell lysates and intense signal in the supernatant, demonstrating the efficient secretion of the wild-type protein. Cells expressing either Z or His334Asp α1-antitrypsin gave a strong signal in the lysates as the protein was retained within the cells. Nondenaturing PAGE showed α1-antitrypsin polymers in both cell lysates and culture media of cells expressing either mutant, but remarkably, the His334Asp variant formed more polymers than Z α1-antitrypsin (Fig. 5A, lower panel). This is most likely due to less efficient degradation of His334Asp α1-antitrypsin as a result of faster polymerization. It is unlikely to be due to higher levels of expression as COS-7 cells transfected with M, Z, or His334Asp α1-antitrypsin and pulsed with [35S]methionine and cysteine all show similar levels of expression at time 0 after the pulse (Fig. 5B, total α1-antitrypsin).

Figure 5.

Characterization of His334Asp (H334D) α1-antitrypsin. (A) Cell lysates (C) and culture medium (M) from COS-7 cells transfected with M, Z, and H334D α1-antitrypsin (α1AT) were collected after 24 hours and analyzed by 10% SDS (upper panel) or nondenaturing PAGE (lower panel) and western blot with a mAb to α1-antitrypsin (black arrowhead: intracellular α1-antitrypsin; black arrow: extracellular α1-antitrypsin; vertical line: polymers; open arrowhead: monomer). In the same membrane, luciferase was used as a transfection control and eIF2α as a loading control. (B) COS-7 cells transfected as in (A) were pulse-labeled with [35S]methionine and chased as shown (0, 1, 2, or 4 hours). Alpha-1-antitrypsin was immunoprecipitated with an antibody to total α1-antitrypsin (Total α1AT panels) or α1-antitrypsin polymers (mAb 2C1; α1AT polymers panels) by splitting each sample in two equal aliquots. M, Z, and H334D: α1-antitrypsin variants. Arrowhead: intracellular 52 kDa species; arrow: intracellular and extracellular 55 kDa species. M α1-antitrypsin was cleared from the cell lysates and appeared in the culture medium faster than Z or His334Asp α1-antitrypsin, as seen in the top three panels and in the quantification graphs on the right (top graphs, average of three repeats). Polymers were detected in cells expressing Z or His334Asp α1-antitrypsin only, as shown in the three bottom panels and the quantification graphs on the right (bottom graphs). (C) COS-7 cells transfected as in (A) were fixed after 48 hours and immunostained with an anti-α1-antitrypsin polyclonal antibody (red) and either an ER protein (calreticulin) or a Golgi protein (GM 130) (green). Only merged panels are shown with overlapping signals in yellow. DNA is blue (DAPI). Scale bar: 10 μm.

The trafficking of His334Asp α1-antitrypsin was investigated in more detail by metabolic pulse labeling of transiently transfected COS-7 cells with [35S]methionine and cysteine, followed by immunoprecipitation and analysis by SDS-PAGE and autoradiography (Fig. 5B). Alpha1-antitrypsin was readily detected in cell lysates at every time point as a 52 kDa band (Fig. 5B, total α1AT panels), and in the culture medium as a fully processed form of 55 kDa. Phosphoimaging quantification showed that the intracellular clearance of α1-antitrypsin was slower for both Z and His334Asp α1-antitrypsin (Fig. 5B, top left graph), whereas secretion into the culture medium was most efficient for M α1-antitrypsin and slightly faster for His334Asp than Z α1-antitrypsin (Fig. 5B, top right graph). This is in keeping with previous reports for M and Z α1-antitrypsin,24, 25 and with the liver disease caused by the new His334Asp mutation of α1-antitrypsin. We also examined early polymer formation by immunoprecipitation with mAb 2C1 from the same samples analyzed above. Polymers of His334Asp α1-antitrypsin accumulated faster than Z polymers inside the cells (Fig. 5B, α1AT polymers panels and left histogram), in keeping with the results from the steady state western blot analysis (Fig. 5A).

The intracellular localization of the mutant proteins was assessed by confocal microscopy (Fig. 5C). Both mutants produced a punctate pattern that colocalized with a marker for the ER (calreticulin), whereas only M α1-antitrypsin colocalized with the Golgi marker (GM130), in agreement with its efficient forward trafficking. The presence of M α1-antitrypsin within the ER is expected due to the continuous synthesis of the protein at this time after transfection, but the formation of low levels of M α1-antitrypsin polymers in our transient transfection cell model can not be ruled out. This is discussed in more detail in the results section for Fig. 6. The lack of staining for α1-antitrypsin in the Golgi of the mutant expressing cells indicates that the transport from the ER to the Golgi is the limiting step for the secretion of both Z and His334Asp α1-antitrypsin. When cells expressing His334Asp α1-antitrypsin were costained for total α1-antitrypsin with a rabbit polyclonal antibody (Fig. 3A, right panel, red) and the mAb 2C1 (Fig. 3A, right panel, green) a strong overlap was seen (yellow). This demonstrates that mAb 2C1 can detect the intracellular polymers formed by His334Asp α1-antitrypsin. The mAb 2C1 also detected polymers in paraffin-embedded liver sections of the index case (Fig. 3B, right panel), where the staining represents detection of both Z and His334Asp α1-antitrypsin polymers. The finding that the 2C1 antibody recognizes His334Asp α1-antitrypsin polymers in transfected cells by immunoprecipitation (Fig. 5B), immunocytochemistry (Fig. 5C), western blot, and ELISA (Fig. 6B,C) demonstrates that they share the same epitope present in Z α1-antitrypsin, even though the mutations mediate their effects via different parts of the protein (Fig. 1).

Figure 6.

Polymers formed by different mutants of α1-antitrypsin are recognized by mAb 2C1. (A) COS-7 cells were transiently transfected with M, Z, H334D (HD), Siiyama (Sii), or Brescia (Bs) α1-antitrypsin (α1AT). Cell lysates were collected after 48 hours of expression and analyzed by 10% SDS (α1AT panel) or nondenaturing PAGE (α1AT non denat. panel) and western blot with a commercial mAb to α1-antitrypsin (vertical line: polymers; square bracket: monomers; black arrow: intracellular α1-antitrypsin). In the same SDS membrane, luciferase was used as a transfection control (luciferase panel) and glyceraldehyde 3-phosphate dehydrogenase was used as a loading control (GAPDH panel). (B) A parallel western blot loaded equally to that in (A) was blotted with 2C1 mAb, which recognized only the polymers in each lane (vertical line). (C) The same cell lysates were analyzed by sandwich ELISA with the 2C1 mAb. The graph shows relative values, calculated by normalizing α1-antitrypsin polymer content (in nanograms per milliliter) to total protein in each lysate and then to polymers contained in the lysate from Z α1-antitrypsin expressing cells (n = 3, graph shows average and standard deviation).

The 2C1 mAb Recognizes Polymers Formed by Different Shutter Domain Mutants of α1-Antitrypsin.

The 2C1 antibody was used to assess polymers formed by two other shutter domain mutants of α1-antitrypsin: Siiyama (Ser53Phe)26 and Brescia (Gly225Arg).27 These were transiently expressed in COS-7 cells, in parallel to M, Z, and H334D α1-antitrypsin. The cell lysates were assessed by SDS and nondenaturing PAGE followed by western blot analysis and also by sandwich ELISA with the 2C1 mAb. All variants were efficiently expressed by COS-7 cells and showed clear intracellular signals in western blot analysis of the SDS-PAGE (Fig. 6A). The total amount of α1-antitrypsin polymers in each lysate was determined by western blot of nondenaturing PAGE with a commercial mAb that recognizes all α1-antitrypsin (Fig. 6A, line). When the same lysates were analyzed with the 2C1 antibody (Fig. 6B), the bands corresponding to the monomeric forms were not recognized (compare to Fig. 6A, bracket), but the polymers formed by each variant were detected with similar intensities to the commercial mAb used in Fig. 6A. When quantified by sandwich ELISA with the 2C1 mAb, each α1-antitrypsin variant showed a signal relative to Z α1-antitrypsin that reflected the intensity seen in the western blot analysis (Fig. 6C). In this series of experiments, the high expression levels obtained in our transient transfection system caused a small amount of polymerization of M α1-antitrypsin that could be seen in the western blot in Fig. 6A (α1-AT nondenaturing panel, M lane), and detected with the 2C1 mAb both by western blot (Fig. 6B, M lane) and by ELISA (Fig. 6C, M lane). We have previously observed polymers in COS-7 cells transfected with M α1-antitrypsin (E.M. and D.A.L., unpublished results) whereas others have reported the polymerization of the wild-type serpin megsin when expressed at high levels.28

These results demonstrate that the 2C1 mAb recognizes polymers formed by different disease associated variants of α1-antitrypsin, including Z and a range of shutter domain mutants.

The 2C1 mAb Recognizes All Polymers Formed by Cells Expressing His334Asp α1-Antitrypsin.

It is possible that a mixture of different polymer types form in vivo and that mAb 2C1 detects only a portion of them in ELISA, western blot and immunocytochemistry. This possibility was assessed by using the 2C1 mAb to immunodeplete polymers formed by His334Asp α1-antitrypsin (Fig. 7). COS-7 cells were transiently transfected with either M or His334Asp α1-antitrypsin and cell lysates were collected after 24 hours expression. The polyclonal antibody removed all the α1-antitrypsin from the supernatants of both M and His334Asp α1-antitrypsin expressing cells after the second round of immunoprecipitation (Fig. 7, top panel, M and H334D α1AT, S2). In contrast, the mAb 2C1 immunoprecipitated minimal amounts of α1-antitrypsin from cells expressing the M variant (Fig. 7, bottom panel, M α1AT, P1 and P2) but large amounts of His334Asp α1-antitrypsin (Fig. 7, bottom panel, H334D α1-antitrypsin, P1, P2, and P3), virtually depleting the sample after three rounds of immunoprecipitation (Fig. 7, bottom panel, H334D α1AT, S3). Similar results were obtained when the experiment was repeated using cells expressing Z α1-antitrypsin. These data show that only one type of polymer, recognized by the 2C1 mAb, is detectable in the lysates of cells expressing His334Asp and Z α1-antitrypsin.

Figure 7.

All polymers formed by His334Asp α1-antitrypsin are recognized by mAb 2C1. COS-7 cells were transiently transfected with either M or H334D α1-antitrypsin (α1AT), and cell lysates were collected after 24 hours expression. A polyclonal antibody against α1-antitrypsin (top panel) or mAb 2C1 (bottom panel) were used in parallel samples to immunoprecipitate (IP) α1-antitrypsin. The starting lysate (S0), pellet (P), and supernatant after immunoprecipitation (S) were analyzed by 10% SDS-PAGE and western blot. The polyclonal antibody immunoprecipitated α1-antitrypsin from both M and H334D α1-antitrypsin–expressing cells (P1 and P2, in M and H334D α1AT panels), depleting all signal after the second round of immunoprecipitation (S2, in M and H334D α1AT panels). The mAb 2C1 did not immunoprecipitate α1-antitrypsin from cells expressing the M variant (P1 and P2, in M α1AT panel), but it did precipitate H334D α1-antitrypsin, depleting the sample after three rounds of immunoprecipitation (S3, in H334D α1AT panel).

Discussion

It is well recognized that mutations in α1-antitrypsin cause the protein to form intracellular polymers that are associated with liver disease. The structure of these polymers is believed to result from the sequential linkage between the reactive center loop of one molecule and β-sheet A of another.2 However, this has recently been challenged by a model in which polymers are linked by a β-hairpin of both the reactive center loop and strand 5A of one molecule inserting into β-sheet A of another.13 The data in support of the classical model for α1-antitrypsin polymerization are based on polymers induced by heating purified α1-antitrypsin, whereas the new model is based on polymers formed at low pH or in the presence of chemical denaturants. It is not known if different disease related mutants of α1-antitrypsin form polymers by the same mechanism and with the same overall structure.

We have developed the novel 2C1 mAb to evaluate the conformation of polymers of α1-antitrypsin formed in vitro and in vivo. This antibody detected polymers prepared by heating purified M or Z α1-antitrypsin in vitro, polymers obtained from the liver of a Z α1-antitrypsin homozygote and polymers from transfected cells expressing the Z variant. It also detected polymers in fixed cells and tissue. The 2C1 mAb was specific for an epitope on polymers as it did not recognize the monomeric protein, the complex of α1-antitrypsin with trypsin, reactive center loop cleaved α1-antitrypsin or α1-antitrypsin in the monomeric, inactive latent conformer. We believe this to be the first mAb with such a high specificity for the pathological polymers of α1-antitrypsin.

The 2C1 antibody was then used to evaluate polymers formed by the novel His334Asp mutant of α1-antitrypsin identified in a 6-week-old boy who presented with prolonged jaundice. This mutant has striking homology to His338Arg neuroserpin, a highly polymerogenic mutant that causes intracellular polymerization, formation of inclusion bodies within the ER and the dementia FENIB.23 Our results show that His334Asp α1-antitrypsin forms polymers within the ER more rapidly than Z and indeed any other mutation of α1-antitrypsin described to date. Although separated by only eight residues, the effects of the Z (Glu342Lys) and His334Asp mutations are on different structural features of the protein. The Z mutation is in the hinge region and so perturbs the relationship between the reactive loop and β-sheet A (Fig. 1). The His334Asp mutation is in the shutter region where it will disrupt a critical hydrogen bond network that controls opening of β-sheet A.29 Both mutations result in polymers that are recognized by the 2C1 antibody suggesting that they share the same structure. Given the homology to the highly polymerogenic His338Arg variant of neuroserpin, it is likely that neuroserpin mutants form polymers with a similar structure to those formed by α1-antitrypsin.23 Our new mAb 2C1 similarly recognized polymers formed by the Siiyama (Ser53Phe)26 and Brescia (Gly225Arg)27 mutants that are also located within the shutter region of α1-antitrypsin.

The epitope that is recognized by the 2C1 antibody is unknown. However, its high affinity for polymers of Z α1-antitrypsin is completely abolished by the introduction of the Gly117Phe mutation. This mutation causes side chain repacking and a half turn downward displacement of the F-helix.21 These data suggest that the 2C1 antibody may recognize a neo-epitope formed as a result of F-helix remodeling during polymerization.

It is possible that a mix of different α1-antitrypsin polymers coexist in disease and that only one of them is detected by the 2C1 antibody. However, this is unlikely because the 2C1 antibody was able to immunoprecipitate all pathological polymers of α1-antitrypsin from cell lysates of transfected cells. Polymers of mutant α1-antitrypsin were also present within the extracellular media (Fig. 5). Similar data were obtained when we assessed polymers formed by mutants of neuroserpin.17 It is unclear if extracellular polymers are secreted as such or form in the culture medium from secreted monomer.

Taken together, our data show that polymers formed in vivo by the Z and shutter domain mutants of α1-antitrypsin share an epitope that is also present in polymers induced by heating purified M or Z α1-antitrypsin. This suggests that they have a similar overall structure. Understanding the structure of these polymers is essential to aid the development of small molecules to block the aberrant conformational transitions of mutant α1-antitrypsin and so prevent the associated liver and lung disease.

Acknowledgements

We are very grateful to Dr. Sabina Janciauskiene for providing the ATZ11 monoclonal antibody, to Dr. Hagosa Abraha for help with the genotyping of the index case, and to Dr. Anna Fra for the kind gift of the Brescia α1-antitrypsin DNA plasmid. We dedicate this article to Jesús Miranda Baños.

Ancillary