Novel SERPINA1 Alleles Identified Through a Large Alpha-1 Antitrypsin Deficiency Screening Program and Review of Known Variants

Abstract

The SERPINA1 gene encodes the serine protease inhibitor alpha-1 antitrypsin (AAT) and is located on chromosome 14q31-32.3 in a cluster of homologous genes likely formed by exon duplication. AAT has a variety of anti-inflammatory properties. Its clinical relevance is best illustrated by the genetic disease alpha-1 antitrypsin deficiency (AATD) which is associated with an increased risk for chronic obstructive pulmonary disease (COPD) and cirrhosis. While 2 single nucleotide polymorphisms (SNPs) , S and Z, are responsible for more than 95% of all individuals with AATD, there are a number of rare variants associated with deficiency and dysfunction, as well as those associated with normal levels and function. Our laboratory has identified a number of novel AAT alleles that we report in this manuscript. We screened more than 500,000 individuals for AATD alleles through our testing program over the past 20 years. The characterization of these alleles was accomplished by DNA sequencing, measurement of AAT plasma levels and isoelectric focusing at pH 4-5. We report 22 novel AAT alleles discovered through our screening programs, such as Z_{little rock} and QO_chillicothe, and review the current literature of known AAT genetic variants.

Introduction

The alpha-1 antitrypsin (AAT) gene (SERPINA1) is located on chromosome 14q31-32.3. This serine protease inhibitor is predominantly produced by hepatocytes,¹ but also expressed by macrophages,² neutrophils,³ monocytes,⁴ and epithelial cells.⁵ AAT is synthesized as a single polypeptide chain that undergoes co/post-translational modification which includes a 24 amino acid N-terminal clip and the addition of 3 N-linked glycosylated oligosaccharides to produce the di-, tri-, and tetra-antennary structure.⁶

AAT is an acute phase reactant protein known for its anti-inflammatory properties. This is demonstrated by increases in AAT levels within hours after inflammation or infection begins. Allelic variations in this genetic disease can lead to deficiency/dysfunction of the AAT protein. The protein may misfold and accumulate in the endoplasmic reticulum of hepatocytes leading to increased susceptibility for development of cirrhosis.⁷ Since levels of circulating AAT are decreased, less of this molecule can reach the lungs and inhibit neutrophil elastase. The balance of protease to anti-protease is shifted towards lung destruction in deficiency. This is classically seen with the development of chronic obstructive pulmonary disease (COPD) at earlier ages even with minimal or no tobacco history in affected individuals.⁸

AAT’s clinical relevance is best demonstrated by the genetic disease alpha-1 antitrypsin deficiency (AATD) which predisposes individuals to developing COPD and cirrhosis. In the homozygous state, it is associated with development of emphysema at an early age⁸ and with an increased incidence of hepatitis, usually progressing to cirrhosis.⁷ There are 2 major alleles, S and Z, which encompass greater than 95% of all known AAT mutations. The S mutation results from the substitution of valine for glutamic acid at amino acid position 264 in exon III (g.9628 A>T). The Z mutation results from the substitution of lysine for glutamic acid at amino acid position 342 in exon V (g.11940 G>A). However, hundreds of mutations exist, including variants associated with normal circulating plasma levels and those associated with deficiency and/or dysfunction.

Our laboratory has identified several novel mutations through our nationwide testing program where we screened more than 500,000 individuals. Individuals with abnormal or ambiguous screening results were invited to join the Alpha-1 Foundation DNA and Tissue Bank and enrolled at their own discretion. The Alpha-1 Foundation DNA and Tissue Bank was established in 2002 and contains approximately 2400 DNA and plasma samples from AATD patients and their families. These individuals were initially screened based on medical indications from liver, lung, or family history that pointed towards a possible diagnosis of AATD. Characterization of novel alleles was accomplished using DNA sequencing, measuring AAT levels in plasma, and isoelectric focusing (IEF) at pH 4–5. Pathogenic variants were determined using PolyPhen-2, a program which estimates the probability that an amino acid change significantly affects protein structure. In this report we characterized 22 alleles discovered at the University of Florida AAT Genetics Laboratory and provide a comprehensive review of known AAT allelic variants.

Material and Methods

Approach to Detection of Abnormal SERPINA1Alleles

The majority of samples were screened for abnormal alleles using dried blood spot cards containing whole blood collected in three 12mM circles on 903 paper. Punches from the 903 Whatman filter paper containing whole blood were used to determine AAT levels by nephelometry. DNA was extracted from the blood spots and genotyped by TaqMan polymerase chain reaction (PCR) using primers directed to the Z and S alleles. When a sample had a low AAT level inconsistent with the genotype of MZ, SZ, and ZZ, a letter was sent to either the patient or the patient’s physician (depending on the screening program) inviting the individual to join the Alpha-1 Foundation DNA and Tissue Bank and submit a clinical questionnaire and a whole blood sample.

Samples

Genomic DNA was isolated from whole blood samples obtained from participants in the Alpha-1 Foundation DNA and Tissue Bank at the University of Florida, protocol UF-IRB 201500842, after providing written consent. Candidate samples were prescreened and selected for DNA melting, DNA sequencing, or both based on 3 criteria, which, when considered together, suggested the existence of a rare or novel allele. In most cases TaqMan allelic discrimination had to indicate the existence of a non-S or non-Z allele, the AAT protein level had to be lower than 10μM by nephelometry, and IEF had to present an unusual protein migration signature. For melt experiments, DNA concentration was adjusted to 10μg/mL. An M₁M₁ (rs 6647) sample from the Alpha-1 Foundation DNA and Tissue Bank was used for a control.

Primers

Two different sets of primers were designed for the pre-melt PCR amplifications and pre-sequencing PCR amplifications due to the inability of the melt experiment to accurately detect mutations with high fidelity in amplicons greater than 400 base pairs. Exons II and V were subdivided into 3 and 2 sections, respectively, for the pre-melt amplification (Figure 1). Primers were designed to begin approximately 20 base pairs upstream of intron-exon junctions so splice site mutations, as well as intra-exon mutations, could be identified. We did not routinely screen the promotor regions of the AAT gene for 2 reasons: (1) in our previous sequencing studies of AATD participants we have identified promotor single nucleotide polymorphisms (SNPs) but were unable to attribute any reductions in plasma AAT to the SNPs and, (2) all participants we screened using this approach had SNPs that explained the decreases in plasma AAT, e.g., nonsense mutations, frameshifts, stop codon, and splicing mutations.

Pre-melt PCR

The pre-melt amplification solution included 1 μL of genomic DNA and 9 μL of a PCR master mix that included a Klentaq enzyme and LCGreen Plus (Idaho Technology, Inc., Salt Lake City, Utah). Reaction mixtures were pipetted into opaque black and white 96-well plates with a 20μL mineral oil overlay, sealed with optical adhesive tape and amplified. The PCR had an initial denaturation step of 95°C for 2 minutes, 35 cycles of denaturation at 95°C for 15 seconds, reannealing at 65°C for 20 seconds, and elongation at 72°C for 15 seconds, followed by a 95°C hold for 30 seconds and a 26°C hold for 30 seconds.

Melt Acquisition

High resolution melt scanning was performed on an Idaho Tech Lightscanner. After completion of PCR amplification, the instrument’s heating block was warmed to a holding temperature of 70°C, at which point the 96-well tray was inserted. Samples were melted within a temperature range of 74°C to 94°C with fluorescence levels measured over the interval.

Melt Analysis

Results were analyzed using the light scanner software package, which presented each melt event as a curve plotted as fluorescence versus temperature. All melt curves corresponding to a single amplicon section were grouped together and normalized by declaring 100% and 0% fluorescence levels at regions before and after the denaturation event. The temperature shift was set to 5% fluorescence. Using the M₁M₁ control sample as the baseline, the software generated -dF/dT derivative plots that gave steep parabolic curves for samples containing heteroduplexes.

Sanger Sequencing

Exons determined to contain mutations were amplified using in-house sequencing primers. The PCR included an initial denaturation step of 94°C for 1 minute, 40 cycles of denaturation at 94°C for 10 seconds, reannealing at 56°C for 30 seconds, and elongation at 72°C for 1 minute. Sequencing was performed by the UF ICBR sequencing core using an Applied Biosystems Model 3130 Genetic Analyzer or by GeneWiz (South Plainfield, New Jersey). Returned sequences were aligned against National Center for Biotechnology Information-Gene consensus sequences using the MacVector ClustalW/Multiple Sequence Alignment (Apex, North Carolina).

PolyPhen-2

PolyPhen-2 is a program designed to predict the impact of amino acid substitutions on the structure and function of human proteins.⁹ Position of the variant within the protein, along with the specific amino acid substitution was inputted to determine a score from 0–1 to indicate if the overall protein would be deficient, dysfunctional, normal, or null. A score of 0.8 or greater is considered probably damaging.

Naming

Novel alleles were named according to the birthplace of the individual with the novel variant. Designation of M, Z, S, QO, etc., were implemented based on the pattern of the protein on the IEF gel. Novel alleles were named according to the Human Genome Variation Society and described at the DNA level with the format “position substituted” “reference nucleotide” > “new nucleotide.” The reference transcript does not include the 24 residues of the signal peptide. Much of the variation in AAT variants is based on amino acid substitutions that alter the electrophoretic migration in an IEF gel at pH 4–5. The most common alleles are the M alleles, M1–3, where the differences are based on a combination of SNPs encoding differentially charged amino acids that can be identified by IEF (see example in Figure 2 ). We used M1(val213) as the base allele to compare all variants.

Approach to Literature Review

Construction of tables of known variants of AAT was accomplished by using the search term “alpha-1 antitrypsin” in the PubMed database. The search was accomplished in early 2020 and there were approximately 14,000 articles with the search word in the title and/or abstract. All 14,000 article abstracts and titles available on PubMed were screened for terms indicating a report of novel AAT alleles. Following the identification of these articles the authors reviewed the articles for accuracy and availability of sufficient data to support the variant.

Results

From the Alpha-1 Foundation DNA and Tissue Bank, we identified 22 novel alleles (Figure 1 and Table 1). The majority of these novel alleles were discovered in asymptomatic individuals who underwent genetic testing due to a family history of AATD. Three individuals with novel mutations who presented with pulmonary symptoms at a younger age are discussed in greater detail below.

I_orange

The proband was a 33-year-old female with a history of asthma. She was a non-smoker with a family history of COPD. During initial screening, she was determined to be carrying a Z allele and subsequent DNA sequencing revealed a heterozygous mutation at codon 39 (Arg CGC > Cys TGC) on an M2 (Arg101 CGT > His CAT) background which causes it to run cathodal to the known I mutation by IEF.

QO_{san francisco}

This patient was a 41-year-old female with a history of hepatitis, asthma, and emphysema. She was on supplemental oxygen and receiving AAT augmentation therapy at the time of sample collection, though she reported a very low plasma level of AAT prior to initiation of augmentation therapy. DNA sequencing appeared to be homozygous for a null allele in exon II (Pro28 CCC, insertion of T, shift to stop TGA 32) though there is a possibility she has a complete deletion of her second AAT allele. The patient had 2 children who were tested at the time, a 16-year-old with a plasma AAT level of 12.8mM and a 7-year-old with a level of 15.4mM, both are heterozygous for this novel null allele.

QO_knoxville

The proband was a 57-year-old male with an extensive family history of lung disease, including emphysema, COPD, chronic bronchitis, and asthma. In addition to a Z allele, he was determined to have a novel null allele (a frame shift deletion in exon V resulting in a stop codon at 373). Family testing facilitated by the Alpha-1 Foundation DNA and Tissue Bank showed his son, a 28-year-old with asthma was also heterozygous for this null allele. More family members were screened and a 34-year-old niece of the proband with no active lung disease was found to have the SQO_knoxville genotype.

Other Novel Alleles

Because the genetic screening program focused predominately on individuals identified as AAT deficient, the majority of alleles identified represent disease-associated mutations and fall into 2 major categories of AATD: null (n=13) and deficient (n=7). Only 2 normal alleles were identified, P_{connelllsville} and N_{hartford city} (Table 1, Figures 1 and 2). Two of the deficient alleles, I_orange and S_{north adams}, show altered IEF migration patterns of the I and S alleles, respectively, and only differ from them with respect to their base alleles which in these novel cases appear on the M₂ and M₄ backgrounds rather than on the more common M₁ background. Electrophoretic differences such as these have previously been reported in the P family of alleles.

Novel deficiency alleles were called “disease associated” based on a PolyPhen-2 score above 0.8 and a clinical history of respiratory and/or liver disease. The molecular mechanisms of abnormal secretion of AAT typically were associated with amino acid substitutions that cause a significant charge alteration, such as a neutral amino acid to a charged amino acid or vice versa. Novel null alleles were most commonly caused by single base deletions and subsequent sequence frameshifts, leading to a premature stop codon. One null allele, QO_cincinnati, was the result of a base change in a splice junction in intron III (Figure 1). Two other null mutations were the result of a base substitution that created a stop codon (Figure 1 and Table 1).

We have grouped known variants into similar categories as the reported novel variants of normal, deficient/dysfunctional, and null variants (Table 2)^10-31 (Table 3)^32-59 (Table 4).^60-84 Alleles listed without a name were discovered via sequencing alone and were not given a name based on their pattern of IEF. We included a table that lists variants that have been identified by IEF and did not have DNA sequencing (Table 5).^85-106

Discussion

Our laboratory has been screening individuals for AAT variants for several years using a series of improving and more efficient technologies to simplify the accurate identification of new alleles. In the process, we have identified a number of alleles that provide insight into the molecular basis of AATD-based key relationships between structure and function. While rare deficiency alleles do not play a significant role in the vast majority of AATD individuals, they may play an important role in guiding novel therapies involving chaperones, gene editing, and gene silencing to modulate the consequences of misfolded AAT.

While more than 95% of all disease-affected AATD individuals have severe deficiency due to the presence of the Z allele, novel mutations in the AAT gene provide insight into the key structural elements of the AAT protein. Modeling of Z_{little rock} (Figure 3) and QO_Chillicothe (Figure 4) demonstrate how amino acid changes may create different interactions that alter the structural stability of the protein. These observations have allowed structural biologists to identify key mechanisms of misfolding including the serpin shutter disruption, importance of the C-terminus in structure, and our understanding of the mechanisms of polymerization of variants.^107,108

Centers throughout the world devoted to screening for known and novel AAT alleles need to continue their work. While the ease of sequencing DNA has made major leaps, specialized protein analysis and retaining key clinical information of alleles remain very important resources for the AATD community and requires specialists. The Alpha-1 Foundation has been one of the most generous funders of these specialized detection centers. There is much to do before there is a cure for AATD, a condition that affects nearly half a million individuals world-wide. A major step towards developing a cure for AATD includes screening for deficient individuals with informative structural changes and using this information to better understand the structural basis of AATD. As has been said by more than one geneticist, nature through its rich variation has done all the interesting experiments, we just need to determine what we can learn from them.

Abbreviations

Abbreviations: AAT=alpha-1 antitrypsin; AATD= alpha-1 antitrypsin deficiency; COPD=chronic obstructive pulmonary disease; IEF=isoelectric focusing; PCR=polymerase chain reaction; SNPs=single nucleotide polymorphisms

Funding Statement

This work was supported by the Alpha-1 Foundation.