Plasma levels of α₁-antitrypsin-derived C-terminal peptides in PiMM and PiZZ COPD patients

Extract

α₁-Antitrypsin (α₁-AT) is an acute-phase glycoprotein that antagonises the activity of various proteases and performs broad immunomodulatory functions [1, 2]. One of the well-recognised functions of α₁-AT is to protect the lungs against the development of COPD and emphysema. Consequently, people with severe inherited α₁-antitrypsin deficiency (α₁-ATD), and especially smokers, are at a higher risk of developing COPD with emphysema in the third or fourth decades of life [3, 4]. Most clinically recognised α₁-ATD patients carry the Z-allele homozygously (PiZZ, Glu342Lys mutation in SERPINA1 gene) and have mean serum levels of ∼32 mg·dL⁻¹, while individuals with a normal, PiMM genotype have α₁-AT levels of ∼130 mg·dL⁻¹ [5].

Tweetable abstract

Plasma levels of α₁-antitrypsin-derived C-terminal peptides might be valid as novel biomarkers to predict and/or characterise exacerbations in PiMM and PiZZ COPD patients, or to reflect the efficiency of augmentation therapy in PiZZ patients https://bit.ly/3rNJeLd

To the Editor:

α₁-Antitrypsin (α₁-AT) is an acute-phase glycoprotein that antagonises the activity of various proteases and performs broad immunomodulatory functions [1, 2]. One of the well-recognised functions of α₁-AT is to protect the lungs against the development of COPD and emphysema. Consequently, people with severe inherited α₁-antitrypsin deficiency (α₁-ATD), and especially smokers, are at a higher risk of developing COPD with emphysema in the third or fourth decades of life [3, 4]. Most clinically recognised α₁-ATD patients carry the Z-allele homozygously (PiZZ, Glu342Lys mutation in SERPINA1 gene) and have mean serum levels of ∼32 mg·dL⁻¹, while individuals with a normal, PiMM genotype have α₁-AT levels of ∼130 mg·dL⁻¹ [5]. The dominant theory for the pathogenesis of α₁-ATD-related emphysema is an imbalance between proteases and antiproteases towards protease activity [6, 7]. Therefore, in addition to the usual treatment options for COPD and emphysema, patients with α₁-ATD-related emphysema are treated with human plasma-purified pharmaceutical preparations of α₁-AT as an augmentation therapy.

There are reports suggesting that α₁-ATD might arise not only due to inherited mutations of the SERPINA1 gene but also due to post-translational modifications of α₁-AT causing an “acquired” α₁-ATD. In vivo, α₁-AT can undergo oxidation, degradation, complex formation with other substances, self-assembly or other modifications. Some of these may result in “acquired” deficiency of native α₁-AT and in the generation of new molecular forms [8, 9]. For instance, active metalloproteases, like MMP-13, can inactivate α₁-AT by cleavage [10] and generate fragments with novel biological activities [11]. As yet, post-translationally modified forms of α₁-AT and their putative relationship with acquired α₁-ATD have received little attention in COPD and other clinical research areas.

We previously demonstrated that the content of urinary peptides differs between COPD patients with PiMM and PiZZ genotypes [12]. More recent studies found that plasma levels of carboxyl (C)-terminal peptides of α₁-AT are significantly elevated in patients with acute respiratory distress syndrome, severe COVID-19 and bacterial pulmonary sepsis [13–15]. Since peptides of α₁-AT are generated under inflammatory conditions and COPD is characterised by the persistent systemic inflammation [16], we aimed to investigate whether peptides of α₁-AT are present in plasma of COPD patients with PiMM and PiZZ genotypes.

We enrolled 111 clinically stable COPD patients, 67 PiMM and 44 PiZZ, of whom 21 were on intravenous α₁-AT augmentation therapy (i.v. α₁-AT) (Prolastin, 60 mg·kg⁻¹ body weight). Phenotyping and genotyping were performed to confirm PiMM and PiZZ genotypes. Plasma samples of PiZZ patients on i.v. α₁-AT were taken 1 week after therapy prior to the next i.v. α₁-AT infusion. All EDTA-treated plasma samples were stored at −80°C until analysis. The PiZZ and PiMM patients were homogeneous regarding age (median (interquartile range) 57 (52–62) versus 59 (49–71) years, p=0.161), gender (female/male 16/28 versus 32/35, p=0.161) and smoking habits (smoker/non-smoker 7/37 versus 20/46, p=0.114). Relative to PiMM, PiZZ patients had mild emphysema and lower gas transfer (mean±sd diffusing capacity of the lung for carbon monoxide (D_LCO) 68.5±18.9% (n=28) versus 47.1±18.5% predicted (n=39), p<0.001). All participants provided informed consent. The ethics committee of the Institute of Tuberculosis and Lung Diseases (ITLD), Warsaw, Poland, approved the study (KB-23/2019 and KB-79/2020).

Plasma levels of α₁-AT and high-sensitivity C-reactive protein (hs-CRP) were determined by nephelometry (IMMAGE 800 Protein Chemistry Analyzer; Beckman Coulter Inc., Brea, CA, USA) at the Department of Genetics and Clinical Immunology, ITLD. The lower detection limit was 10 mg·dL⁻¹ for α₁-AT and 0.02 mg·dL⁻¹ for hs-CRP. Plasma levels of C-terminal peptides of α₁-AT differing in the number of amino acids (C22, C36, C37, C39, C40, C42, C43, C44 and C45) were determined by an improved version of the previously published liquid chromatography–tandem mass spectrometry method, validated according to US Food and Drug Administration criteria [17]. The concentrations of peptides were determined in relation to the respective internal standards (C22^IS for C22; C37^IS for C36 and C37; and C42^IS for C40, C42, C43, C44 and C45) (0.8 µM each; sb-PEPTIDE, Saint Egrève, France) and 1/x² weighted quadratic regression using separate calibration curves for each peptide. Data acquisition and processing was performed with Analyst Software (version 1.6.2 and 1.7.1). One common single-nucleotide polymorphism (SNP) within M α₁-AT alleles (M3 allele; Asp376, rs1303) affects the mass of C-terminal peptides. Therefore, we applied a parallel quantification of C-terminal peptides from M-alleles with and without this polymorphism. The concentrations of wild type (wt) and SNP variants were first determined in the most abundant peptide, C42. Then, single values of wt and SNP measurements were summarised to give a final concentration for each peptide in carriers of M-alleles. Values of peptides in PiMM patient plasma, which were below lower limit of quantification (LLOQ) (0.025 µM for C36 and C42, and 0.01 µM for other peptides), were imputed (six values of C36, 11 of C37, three of C40 and two of C42) using the imputeLCMD (version 2.1) package of R Statistical Software (version 4.1.0, R Core Team 2021).

As expected, plasma α₁-AT levels were lower in PiZZ than in PiMM patients whereas PiZZ patients on i.v. α₁-AT had higher α₁-AT levels those off i.v. α₁-AT (figure 1a). Plasma hs-CRP levels varied in between 0.2 (0.1–0.9) mg·dL⁻¹ for PiMM, 0.5 (0.1–0.9) mg·dL⁻¹ for PiZZ off i.v. α₁-AT and 0.3 (0.2–0.5) mg·dL⁻¹ for PiZZ on i.v. α₁-AT. In entire cohort, we found no correlation between α₁-AT and hs-CRP levels (Spearman's rank correlation). In PiZZ patients off i.v. α₁-AT, plasma levels of all analysed peptides were below the LLOQ. However, C36, C37 C40, and C42 peptides were measurable in PiMM and in PiZZ patients on i.v. α₁-AT. As shown in figure 1b, in PiMM and PiZZ on i.v. α₁-AT, levels of C36 and C42 peptides were higher than those of C37 or C40. Moreover, the level of C36 and C42 peptides in PiMM (C36 0.068 (0.041–0.096) µM and C42 0.082 (0.062–0.105) µM were found to be about twice those in PiZZ on i.v. α₁-AT (C36 0.035 (0.029–0.051) µM and C42 0.042 (0.034–0.050) µM) with p=0.0008 for C36 and p<0.0001 for C42 (Mann–Whitney test). Positive correlations were found between α₁-AT and C36 or C42 levels in 102 patients; unfortunately, nine samples were not available for the peptide analysis (figure 1c and d).

FIGURE 1.

Plasma levels of α₁-antitrypsin (α₁-AT) and C-terminal α₁-AT peptides in PiMM and PiZZ COPD patients. a) Plasma levels of α₁-AT in PiMM patients (n=67), and PiZZ paitents off (n=23) and on intravenous α₁-AT (n=21). Data were calculated by using one-way ANOVA. Values passed Shapiro–Wilk normality test and are presented as mean±sd. b) Peptide levels in PiMM and PiZZ COPD patients on i.v. α₁-AT. Peptide concentrations below the lower limit of quantification (LLOQ) were imputed or set to zero. Values failed Shapiro–Wilk normality test and are presented as median (interquartile range). Peptide levels in PiMM and PiZZ on i.v. α₁-AT were calculated using Mann–Whitney test. In PiZZ patients off i.v. α₁-AT, plasma levels of all analysed peptides were below the LLOQ. Correlations between α₁-AT and c) C36 and d) C42 levels. Correlations were calculated using Pearson's test; n indicates the number of available data pairs. The correlation factor r is given. For statistical analysis and data presentation, Prism (version 9.1.2, GraphPad Software) was used. A p-value <0.05 indicates significance.

There was no relationship between C36 and C42 levels and patient age, gender or spirometry tests (Mann–Whitney tests were used for age and spirometry tests, and Fisher's exact test was employed for categorical variables; data not shown). Among 60 patients for whom paired data were available, a weak positive correlation was found between C42 and D_LCO % predicted (Spearman's test: r=0.38, p=0.02).

Taken together, we provide evidence that C-terminal peptides of α₁-AT, notably C36 and C42, are present in plasma of stable PiMM but not in PiZZ COPD patients off i.v. α₁-AT. Since these peptides occur in plasma of PiZZ patients on i.v. α₁-AT and strongly correlate with α₁-AT levels, it is reasonable to assume that the peptides originate from α₁-AT protein cleavage rather than from previously suggested alternative transcripts of the SERPINA1 gene [18]. Confirming this, PiZZ patients off i.v. α₁-AT had very low plasma levels of α₁-AT relative to PiMM patients (28±9 mg·dL⁻¹ (n=23) versus 152±26 mg·dL⁻¹ (n=67), respectively; p<0.001) (figure 1a) and therefore, peptide levels in these patients are undetectable. However, we cannot exclude that proteolytic cleavage of misfolded Z α₁-AT generates the measured peptides and/or hydrophobic C-terminal peptides are hidden within the polymeric structures of circulating Z α₁-AT.

We also found that commercial α₁-AT preparations contain small amounts of C36 and C42 peptides. Based on the analyses of three different lots of Prolastin preparations (2.5 mg·mL⁻¹), peptide concentrations ranged from 0.152 to 0.445 µM for C36 and values just above the LLOQ (0.026 µM) for C42. Hence, in PiZZ patients on i.v. α₁-AT, small amounts of α₁-AT peptides can be delivered with therapy. In our case, plasma samples of PiZZ on i.v. α₁-AT were obtained 1 week after administration of i.v. α₁-AT. Therefore, we assume that peptides in these patients arise from endogenous cleavage of administered α₁-AT rather than from the α₁-AT preparation per se. Unfortunately, nothing is known about peptide pharmacodynamic and pharmacokinetic properties, and we hope our pilot study will encourage further investigations in this field.

The interest is high in circulating peptides as diagnostic and prognostic markers for a variety of diseases. Peptides of α₁-AT have been identified in human urine, bronchoalveolar lavage fluid, gingival crevicular fluid, spleen and bile, and nipple aspiration fluids [17]. For example, the C36 peptide of α₁-AT (typical cleavage product of serine proteases) has been reported as a regulator of bile acid synthesis in a rat model [19] and as a pro-inflammatory activator of human monocytes [11]; the C42 peptide of α₁-AT (generated by metalloprotease cleavage) was suggested as a putative biomarker of sepsis [14] and acute respiratory distress syndrome severity [15]. Other peptides were proposed as biomarkers for glomerular kidney diseases, pulmonary fibrosis, gingivitis and carotid artery stenosis [17]. Whether peptides of α₁-AT per se or in relation to α₁-AT protein can be clinically useful to characterise chronic systemic inflammation [20], exacerbation severity and/or effects of therapeutics in COPD patients, remains to be answered.

Our data provide further evidence that in vivo cleavage of α₁-AT results in a generation of specific profiles and levels of peptides. This post-translational modification may not only lead to acquired deficiency of α₁-AT but also to the generation of byproducts with novel biological activities. We believe that a ratio between α₁-AT and the peptides generated after α₁-AT cleavage may help us better understand protease/antiprotease imbalance mechanisms in PiZZ and PiMM COPD.