A Naturally Occurring Non-Polymerogenic Mutant of aa 1-Antitrypsin Characterized by Prolonged Retention in the Endoplasmic Reticulum

liver cells but the mechanism by which of glycoprotein elicits liver By expressing mutant a 1-ATZ in skin fibroblasts from a 1-AT deficient patients with liver disease (“susceptible hosts”) and comparing them to fibroblasts from a 1-AT deficient individuals without liver disease (“protected hosts”) we have found that there is more efficient degradation of the mutant a 1-ATZ molecule in the ER of protected hosts. These results have provided evidence for the concept that genetic/environmental mechanisms which impair the degradation pathways that constitute the quality control apparatus of the ER increase susceptibility to liver disease in a 1-AT-deficient individuals In this study we examined a family with two sibs having severe liver disease associated with homozygous a 1-AT deficiency. The mother had died of presumed idiopathic liver disease many years earlier ( a 1-AT phenotype unknown). We found that one of the sibs was a compound heterozygote, with one PI Z allele and one allele with both the substitution that characterizes the PI Z allele and the substitution that characterizes the PI saarbrucken (saar) allele. Although the studies of this family do not substantiate the idea that the PI saar allele is associated with liver injury, studies of the intracellular fate of this mutant have important implications for understanding the mechanism of liver disease in a 1-AT deficiency. Immunofluorescence. Cells were plated on glass slides, fixed, permeabilized and stained with anti- a 1-AT IgG as primary and rhodamine-conjugated anti-Ig as secondary and anti-GRP78/BiP Ig as primary with Cy2-conjugated anti-Ig as secondary, exactly as previously described (16). Cells stained with primary or secondary antibody alone were included as controls for background.


Introduction
A subgroup of patients (10%) with the classical form of α1-AT deficiency (homozygous for the PI Z allele) are predisposed to chronic hepatitis, cirrhosis and hepatocellular carcinoma.
Because the incidence of homozygous α1-AT deficiency in Caucasian populations is relatively high (1 in 1600 to 1 in 2000 live births), this deficiency constitutes the most common genetic cause of liver disease in children and the most frequent genetic diagnosis for which children undergo liver transplantation (reviewed in 1). Deficiency of α1-AT is also associated with premature development of destructive lung disease/emphysema (1).
The PI Z allele is characterized by a single nucleotide substitution which results in the substitution of lysine for glutamate at amino acid 342 (2-4). The mutant α1-ATZ protein is retained in the endoplasmic reticulum (ER) of liver cells rather than secreted into the blood and body fluids (5,6). Circulating blood levels of α1-AT in deficient patients reach 10-15% of the levels present in the general population. Studies by Lomas and Carrell have shown that the substitution of lysine for glutamate 342 restricts the mobility of the reactive-site loop of the α1-AT molecule and brings out a tendency for α1-AT molecules to polymerize in the ER (7)(8)(9).
Studies which show that the secretory defect is partially reversed by introducing another mutation which counteracts the polymerogenic properties of the α1-AT molecule (10)(11)(12) provide evidence that the polymerization is, at least in part, responsible for the retention of α1-ATZ in the ER.
Emphysema is caused by lack of α1-AT in the lung, permitting neutrophil elastase to destroy the connective tissue matrix (reviewed in 13,14). In contrast to this loss-of-function mechanism for lung injury, liver disease is thought to involve a gain-of-function mechanism wherein the retention of the mutant polymerized α1-ATZ protein in the ER invokes damage to by guest on March 24, 2020 http://www.jbc.org/ Downloaded from ER Retention of Mutant α1-AT Saarbrucken 4 the affected liver cells (reviewed in 15) but the mechanism by which retention of this aggregated glycoprotein elicits liver injury is unknown. By expressing mutant α1-ATZ in skin fibroblasts from α1-AT deficient patients with liver disease ("susceptible hosts") and comparing them to fibroblasts from α1-AT deficient individuals without liver disease ("protected hosts") we have found that there is more efficient degradation of the mutant α1-ATZ molecule in the ER of protected hosts. These results have provided evidence for the concept that genetic/environmental mechanisms which impair the degradation pathways that constitute the quality control apparatus of the ER increase susceptibility to liver disease in α1-AT-deficient individuals (16).
In this study we examined a family with two sibs having severe liver disease associated with homozygous α1-AT deficiency. The mother had died of presumed idiopathic liver disease many years earlier (α1-AT phenotype unknown). We found that one of the sibs was a compound heterozygote, with one PI Z allele and one allele with both the substitution that characterizes the PI Z allele and the substitution that characterizes the PI saarbrucken (saar) allele. Although the studies of this family do not substantiate the idea that the PI saar allele is associated with liver injury, studies of the intracellular fate of this mutant have important implications for understanding the mechanism of liver disease in α1-AT deficiency.

Experimental Procedures
Patients. The index patient, patient #1, developed GI bleeding at 48 years of age. She was found to have esophageal varices and a diagnosis of α1-AT deficiency with a PI Z genotype and blood levels of α1-AT that were 8% normal. She initially underwent variceal sclerotherapy but later required a distal splenorenal shunt at age 50 for recurrent variceal bleeding. Symptoms of chronic liver failure necessitated a liver transplant at age 56 years. There was no history of significant alcohol exposure. The native liver was characterized by cirrhosis and PAS positive, diastase-resistant globules in liver parenchymal cells.
Her brother, patient #2, was discovered to have splenomegaly at age 38 years. For unknown reasons he underwent a splenectomy shortly thereafter. At age 41 years he developed intestinal bleeding and was found to have esophageal and gastric varices. He was found to be homozygous for PI Z allele with blood α1-AT levels 11% normal. He underwent a gastric devascularization procedures at that time. A liver biopsy showed chronic hepatitis with PASpositive, diastase-resistant globules in liver cells. He had recurrent gastrointestinal bleeding, developed hepatic encephalopathy and died at age 46 after declining liver transplantation. He was not known to have much in the way of alcohol exposure. The mother of patients #1 and #2 died from severe liver failure of unknown etiology. There was no history of alcohol use for the mother. Another brother (patient #3) and two other sisters (patients #4 and #5) were available for further studies.
Sequence analysis. For the index patient, patient #1, total cellular RNA was isolated from her native liver by guanidine isothiocyanate extraction and ethanol precipitation (17). Poly (antisense) --5'ATCACATGCAGGCAGGGACCA-3'. The 1.4 kb PCR fragment was gel purified and subjected to direct sequence analysis. This PCR fragment was also subcloned into the pGME-T vector so that the sequence of each of the two α1-AT alleles could be determined.
A total of eight subclones were subjected to sequence analysis, confirming the presence of compound heterozygousity with four subclones having identical sequence for each of the two alleles.
Genomic DNA was isolated from skin fibroblasts of patient #2 and from blood of patients #3, 4, and 5 with the DNeasy Tissue Kit (QIAGEN). This DNA was subjected to PCR for exon V using primer A1-G5 (sense) --5'-CTGGGATCAGCCTTACAACGTGT-3' and A1-GR were separately transfected by the calcium phosphate DNA-precipitation method with each of these vectors and candidate clones selected in G418. For pulse labeling, each of these cell lines was incubated at 37°C for 3 hr in Trans 35 S label (250 µcl/ml DMEM lacking methionine). For pulse-chase radiolabeling, each of the cell lines was incubated at 37°C for 1 hr in Trans 35 S label (250 µcl/ml DMEM lacking methionine) for the pulse period, rinsed extensively and then incubated at 37°C in complete DMEM for the chase period. At specified time intervals extracellular (EC) medium was harvested and cells (IC) lysed in PBS/1% Triton X-100/0.5% deoxycholic acid/2mM phenylmethylsulfonic acid. EC and IC samples were clarified, immunoprecipitated and immunoprecipitates analyzed by SDS-PAGE/fluorography exactly as described previously (16,19). Aliquots of each sample were subjected to trichloroacetic acid precipitation to ensure that there were equivalent levels of incorporation of radiolabel in each cell line at time 0 IC. Separate gels were also subjected to PhosphorImager (Molecular Dynamics) analysis.
For Western blot analysis cell were lysed in PBS/1% Triton X-100/0.5% deoxycholic acid/2mM phenylmethylsulfonic acid and cell lysates subjected to SDS-PAGE directly or to immunoprecipitation followed by SDS-PAGE. Blots were incubated with biotinylated goatantihuman α1-AT, then with peroxidase labeled strepavidin and developed with the Supersignal substrate (Pierce Chemicals, Rockford, IL).

Preparation of soluble and insoluble fractions from cell lysates. Cell lysates were
passed through a 25-gauge needle ten times on ice (20). Insoluble material was recovered by centrifugation at 16,000 xg for 15 minutes. Pellets were solubilized in 50 µl 50 mM Tris-HCl, pH 6.8, 5% SDS, 10% glycerol with 1 minute of sonication followed by 10 minutes of boiling. Immunofluorescence. Cells were plated on glass slides, fixed, permeabilized and stained with anti-α1-AT IgG as primary and rhodamine-conjugated anti-Ig as secondary and anti-GRP78/BiP Ig as primary with Cy2-conjugated anti-Ig as secondary, exactly as previously described (16). Cells stained with primary or secondary antibody alone were included as controls for background. In order to more definitively compare the fate of α1-ATZ and α1-AT saar, we carried out pulsechase radiolabeling for even longer time intervals (Figure 6). For α1-ATZ the results show slow but progressive disappearance from IC between 2 and 8 hrs of the chase period with a small amount of mature α1-AT appearing EC between 5 and 8 hrs. For α1-AT saar there is very relatively little disappearance from IC throughout the chase period but some disappearance does occur between 5 and 8 hrs indicating that intracellular retention of this mutant is as great as, or even greater than, that of α1-ATZ. Again, no detectable α1-AT saar is observed in the EC fluid even after 8 hrs of the chase period. Results for α1-ATZ + saar were similar to these for α1-AT saar (data not shown).

Intracellular localization and solubility of α α1-AT saar and α α1-ATZ + saar. In order
to determine whether α1-AT saar and α1-ATZ + saar accumulate in the ER, the stable transfected CHO cell lines were subjected to immunofluorescence with antibody to α1-AT and antibody to GRP78/BiP as a marker for the ER ( Figure 7A). The results show that antibody to α1-AT only faintly stains the perinuclear region in CHO cells transfected with wild type α1-AT.
However, in CHO cells transfected with α1-ATZ, α1-AT saar and α1-ATZ + saar, antibody to α1-AT brightly stains the cytoplasm in a reticular pattern accentuated in the perinuclear region.
Staining with antibody to GRP78/BiP stains an identical region in CHO cells transfected with wild type α1-AT, α1-ATZ, α1-AT saar and α1-ATZ + saar. In Figure 7B, staining for α1-AT co-localized with that for BiP by double-label immunofluorescence in CHO cells transfected with α1-ATZ + saar. Staining for α1-AT also co-localized with that for BiP in CHO cells transfected with α1-ATZ and α1-AT saar (data not shown). These results provide evidence that α1-AT saar and α1-ATZ + saar are retained in the ER.

Discussion
Extensive clinical experience with α1-AT deficiency has shown that there is marked variability in the development and severity of liver disease among homozygotes for the PI Z allele, patients with the classical form of this disease. Prospective nationwide screening studies carried out in Sweden have confirmed that only ∼10% of this population develops clinically significant liver disease (21,22). Using a genetic complementation approach we have found that there is less efficient ER degradation of α1-ATZ in cells from some α1-AT-deficient individuals that are susceptible to liver disease (16), and therein, that these individuals have a lesser inherent capacity to respond to, and/or protect themselves from, the hepatotoxic circumstances posed by ER retention of mutant α1-ATZ. In this study we examined a potentially informative family in which a pair of sibs had severe liver disease associated with the classical form of α1-AT deficiency. There was also a history of severe idiopathic liver disease in the mother which could not be attributed to known hepatotoxic exposures including alcohol. Unfortunately, we have been unable to establish a skin fibroblast cell line from the index patient to determine whether she has delayed ER degradation of α1-ATZ, but we do know that skin fibroblasts from her brother have delayed ER degradation of α1-ATZ (16). We decided to analyze the sequence of the α1-AT gene in these sibs. To our surprise the index patient from the family has an unusual compound heterozygous mutation of the α1-AT with one PI Z allele and a second allele with a combination of the Z and saar mutations, the PI Z + saar allele. This result raised the possibility that the presence of a PI saar allele, or a PI Z + saar allele, represented a genetically linked mechanism for susceptibility to liver disease. Although we cannot completely exclude this possibility it is unlikely because one of these alleles was not found in the brother, patient #2, in by guest on March 24, 2020 http://www.jbc.org/ Downloaded from whose fibroblasts there is delayed ER degradation of α1-ATZ. Thus, one of these alleles is not even linked to the liver disease phenotype in this family.
The presence of the saar mutation on one of the α1-AT alleles of patient #1 also prompted us to examine the fate of α1-AT saar in a heterologous mammalian cell expression system. The results show that there is prolonged ER retention of α1-AT saar, perhaps even more prolonged than that of α1-ATZ, even though it does not form insoluble polymers. Previous studies have shown that there is ER retention of the PI Clayton allele, which has the same truncation on the background of another polymorphic variant of α1-AT (23), α1-AT Hong Kong in which the carboxyl terminus is truncated by sixty-one amino acids (24), and a recombinant α1-AT molecule in which the carboxyl terminus is truncated by four amino acids (25). It was particularly interesting to observe that when the α1-ATZ molecule is truncated, as it would be in the PI Z + saar allele found in our index patient, it no longer forms insoluble aggregates but is retained in the ER for a prolonged, perhaps every more prolonged, time interval.
These results provide further evidence that there are mechanisms by which mutant α1-AT molecules are retained in the ER other than the polymerization mechanism which has been previously correlated with ER retention of α1-ATZ (7-9). By comparing the effects of the polymerogenic mutant α1-ATZ to those of nonpolymerogenic mutants, such as α1-AT saar and α1-ATZ + saar, which are retained in the ER to a similar extent, we may be able in future studies to ascertain whether it is the intrinsic polymerogenic properties that are hepatotoxic in some patients with α1-AT deficiency.     hrs. The right panel shows the EC samples for the same gel exposed to fluorography for 3 days.