Identification and characterisation of eight novel SERPINA1 Null mutations

Alpha-1 antitrypsin (AAT) is a serine protease inhibitor, encoded by the SERPINA1 gene on the long arm of chromosome 14 at 14q32.1. The gene is comprised of four coding exons (II, III, IV, and V), three untranslated exons (Ia, Ib, and Ic) in the 5′ region and six introns. Following translation, the 24 amino acid signal peptide is removed and the mature polypeptide is a 394 amino acid, 52 kDa glycoprotein with three asparagine-linked carbohydrate side chains [1]. AAT is an acute phase protein produced predominantly by hepatocytes, but AAT synthesis also occurs in mononuclear phagocytes, neutrophils, and airway and intestinal epithelial cells [2]. Consistent with a role as an important acute phase reactant, hepatocytes express approximately 200 times more AAT mRNA than other cells [3] and serum levels rapidly increase several-fold during the acute phase response [4]. The primary function of AAT is the regulation of serine proteases, and the chief site of action is the lungs where it protects the fragile alveolar tissues from proteolytic degradation during inflammatory responses. In addition to its undoubted anti-protease properties, there is accumulating evidence that AAT plays a key anti-inflammatory role [5].

Alpha-1 antitrypsin deficiency (AATD) (MIM # 613490) is an inherited condition caused by mutations within the polymorphic SERPINA1 gene and is characterised by decreased serum AAT concentrations. AATD is an under-diagnosed condition and the majority of cases remain undiagnosed. The World Health Organisation (WHO), the American Thoracic Society (ATS), and the European Respiratory Society (ERS) advocate a targeted screening approach for the detection of AATD in at risk populations, specifically chronic obstructive pulmonary disease (COPD), non-responsive asthma, cryptogenic liver disease and in first degree relatives of known AATD patients. Over 100 mutations leading to AAT deficiency have been identified to date and are associated with varying degrees of risk for lung and liver disease. AATD is associated with increased risk of cutaneous panniculitis [6] and case reports have linked AATD to vasculitis [7], and Wegener’s granulomatosis [8] with the Z allele over-represented in subsets of ANCA-associated vasculitis [9]. The most common mutations known to cause AATD are the dysfunctional Z (Glu342Lys) and S (Glu264Val) mutations. The Z mutation leads to a severe plasma deficiency and is the most common clinically significant allele. The majority of individuals diagnosed with severe AATD are homozygous for the Z mutation, and have circulating AAT levels reduced to 10-15% of normal. This is because the Z mutation prompts the AAT protein to polymerise and accumulate within the endoplasmic reticulum of hepatocytes, thus causing impaired secretion [10]. The rate of polymer formation for S is much slower than Z AAT, leading to reduced retention of protein within hepatocytes, milder plasma deficiency, and a negligible risk of disease in MS heterozygotes [11],[12]. However, there is a risk of lung disease in compound heterozygotes. For example, if the slowly polymerising S variant of AAT is inherited with a rapidly polymerising variant such as Z, the two variants when co-expressed can interact to form heteropolymers, leading to cirrhosis and plasma deficiency [13].

The ultra rare family of SERPINA1 mutations termed silent or Null are characterised by a complete absence of AAT in the plasma. Null (also called Q0) mutations are caused by a variety of different mechanisms including large gene deletions [14], intron mutations [15], nonsense mutations [16], and frameshift mutations [17]. In some cases, Null variants are synthesised in the hepatocytes, but they are rapidly cleared by intracellular degradation pathways [18]. As Null mutations do not induce AAT polymerisation, they confer no risk of liver disease but do confer a particularly high risk of lung disease [19]. The exact prevalence of Null mutations is unclear, and is hampered by a lack of general awareness of AATD and inherent flaws in diagnostic strategies.

The diagnostic algorithm for diagnosis of AATD was applied as previously reported [20]. The probands were referred to the Italian or Irish National Reference Centres for the Diagnosis of AATD, situated in Pavia and Brescia (Italy), and Dublin (Ireland), respectively. Where possible, relatives were analysed and family trees were created (online Additional file 1). Family members included in the study or their parents gave written informed consent. All procedures were in accordance with the declaration of Helsinki and approved by the local ethics committees. Clinical data were obtained from direct observation or medical charts.

AAT measurements were performed by a rate immune nephelometric method (Array 360 System; Beckman-Coulter) or by immune turbidimetry (Beckman Coulter AU5400). The phenotype was determined by isoelectric focusing analysis (IEF) on agarose gel with specific immunological detection [21]. DNA was isolated from whole peripheral blood or dried blood spot (DBS) samples using a commercial extraction kit (DNA IQ System, Promega or PAXgene Blood DNA kit, PreAnalytix or DNA Blood Mini kit, Qiagen). The new mutations were identified by sequencing all coding exons (II-V) of the AAT gene (SERPINA1, RefSeq: NG_008290), as previously described [20],[22], using the CEQ 8800 genetic analysis System (Beckman Coulter) or the Big Dye Terminator Cycle Sequencing Kit 3.1 (Applied Biosystem) with the 3130 Genetic Analyzer.

The specific mutations are summarized in Table 1. The eight new Null mutations have been conventionally named Q0_cork, Q0_perugia, Q0_brescia, Q0_torino, Q0_cosenza, Q0_pordenone, Q0_lampedusa, and Q0_dublin according to the birthplaces of the oldest subject carrying each mutation. Q0_brescia, Q0_torino and Q0_cosenza consist of point mutations in the sequence of coding DNA that result in a premature stop codon (nonsense mutation). Q0_cork, Q0_perugia, Q0_pordenone, Q0_lampedusa and Q0_dublin were caused by deletions, resulting in frameshift of the reading frame, and creating premature stop codons (Figure 1).

The genotype, AAT levels, and clinical details of each proband and their relatives bearing Null mutations are listed in Table 2.

Null alleles result from different molecular mechanisms, including large gene deletions, intron mutations, nonsense mutations, frameshift mutations due to small insertions or deletions, and missense mutations associated with amino acid substitutions in potentially critical structural elements [23]. The common trait of Null mutations is the total absence of serum AAT. These mutations are extremely rare and can be difficult to diagnose, mainly because isoelectric focusing (IEF), a commonly used diagnostic method, although not preferred technique for screening of AATD [24], is not able to detect Null variants, as they do not produce protein. Therefore, the M/Null and MM phenotypes are identical when analysed by isoelectric focusing with only the normal M protein evident. Secondly, M/Null genotypes can be misclassified as M homozygotes in many common genotyping assays [25]. Sequence analysis of SERPINA1 gene is the optimal technique to detect Null mutations and only the application of an efficient and cost-effective diagnostic algorithm can ensure the diagnosis of a subject heterozygous or homozygous for Null mutations [20].

The existence of AAT Null alleles was first noted in the early 1970s by several investigators. The first published report of a Null SERPINA1 mutation described the case of a 24 year old man who had advanced pulmonary emphysema and no detectable serum AAT [26]. The first report of a probable Null SERPINA1 mutation in Ireland was a case report in 1974 describing a pedigree in which the proband was Z/Null, a son S/Null and the mother M/Null [27]. The precise Null mutation was not identified and the diagnosis was based on the discordant AAT concentrations in his son and mother when compared to phenotype identified by starch gel electrophoresis. The first report of a Null mutation of Italian origin was Q0_trastevere, which was detected in an Italian individual with asthma and emphysema [16].

To date, a total of 26 different Null alleles have been detected and characterized (Table 3). Many are caused by premature stop codons, mainly due to nonsense mutations or insertion/deletion of one-two nucleotides that cause frameshift of the reading frame and lead to a premature stop codon. A second group of Null mutations lie in introns; some of these have been identified in mRNA splicing sites: Null_west is characterized by a single G > T base substitution at position 1 of intron II, which generally is highly conserved; Null_{bonny blue} has been described as a deletion of the previously reported G. Other mutations are caused by large deletions; examples are Null_{isola di procida}, a deletion of a 17Kb fragment that includes exons II-V [14], and Null_riedenburg, caused by the complete deletion of the gene [28]. It is well known that an almost full length molecule is essential for the secretion of AAT, therefore a truncated protein prevents the secretion itself [18].

Interestingly, Null mutations can also be induced by a simple amino acid substitution, like in Null_ludwigshafen (Ile⁹² > Asn⁹²). This substitution of a polar for a non-polar amino acid leads to folding impairment, with destruction of tertiary structure and therefore intracellular degradation [42]. In most Null mutations belonging to this group, it is not clear whether the altered glycoprotein is unstable and therefore recognized as defective by intracellular methabolic pathways and degraded, or if it is secreted but, due to a very short half-life with rapid turnover, it cannot be detected by routine diagnostic assays. In addition, some Null mutations may yet turn out to be “secreted” Null. For example, Null_{new hope} and Null_newport, were defined as Null on the basis of IEF and protein quantification in a period when molecular diagnosis was not widely available. A precedent for incorrect Null alleles does exist, and includes the well known M_heerlen, which was originally classified as PiQ0 on the basis of IEF and protein quantification [44], and P_lowell, previously called Q0_cardiff [45].

We describe here eight novel Null mutations in the coding regions of the SERPINA1 gene. Three (Q0_brescia, Q0_torino and Q0_cosenza) are nonsense mutations, the others (Q0_cork, Q0_perugia, Q0_pordenone, Q0_lampedusa and Q0_dublin) are frameshift mutations caused by deletion of one or two nucleotides.

It is worth noting most of the new mutations reported in this study occur close to other mutations, supporting the concept of mutational hot spots in the SERPINA1 gene [40]. In fact, Q0_brescia occurs in a portion of 27 nucleotides (nine amino acids) in exon III of the gene, where it is possible to find a conspicuous number of other mutations: P_lowell /P_duarte/Y_barcelona at codon 256, Q0_cairo and M_pisa [46] at codon 259, T/S at codon 264, and the normal variant L_frankfurt at codon 255. Q0_pordenone lies in another region of 27 nucleotides together with other Null (Q0_hongkong, Q0_{new hope}) and normal (P_lyon, P_saltlake,) mutations. Q0_lampedusa occurs in the region of 21 nucleotides where, in addition to Z, other deficient (King, W_bethesda) and normal (E_tokyo, P_st.albans) mutations lie. Lastly, Q0_dublin is only one nucleotide from M_heerlen and M_wurzburg mutations and two nucleotides from E_taurisano [46] deficient alleles.

While Null mutations are extremely rare, the recurrence of Q0_pordenone and Q0_brescia in certain localized areas, without evidence of consanguinity, may indicate a relatively high prevalence of each Null allele in these geographic regions.

Although a discussion of the clinical characteristics of the Null-bearing subjects presented herein is not the main purpose of this study, we can draw some interesting conclusions. Subjects with Null mutations should be considered a subgroup at particularly high risk of emphysema within the spectrum of AATD [19]. In support of this, we report three probands homozygous for Null alleles, with early onset lung disease, despite absent or modest smoking history. Interestingly, the clinical importance of Null heterozygosity has never been investigated. Here we report evidence of the recurrence of lung symptoms (dyspnoea, cough) and lung diseases (emphysema, asthma, chronic bronchitis) in M/Null subjects, over 45 years of age, irrespective of their smoking habit (Table 2).