4.3.1. Adenosine Triphosphate-Binding Cassette Family A Member 3 (ABCA3) Protein
The biology of ABCA3 protein is very complex. ABCA3 (Adenosine Triphosphate-binding Cassette Family A Member 3) protein belongs to proteins’ ATP binding cassette transporter superfamily. ABCA3 is a phospholipidic glycoprotein of 1704 amino acids localized in the external limiting membrane of the lamellar bodies [
10,
11,
50,
52,
73,
79]. Six transmembrane structures mediate ABCA3 function by forming an ATP channel for lipids (disaturated-phosphatidylcholine, phosphatidylglycerol, phosphatidylethanolamine, and cholesterol) transportation from the cytosol into the LBs [
1,
15,
52,
64,
80,
81,
82]. ABCA3 is also involved in lung surfactant transcription and assembly, SFTPB and SFTPC translation, lung surfactant structural transformation and production in AEC II, and epithelial lung cell apoptosis [
4]. A possible role of ABCA3 in the metabolism of lung surfactant phospholipids was also described [
79]. Intracellular metabolism of cholesterol may also be influenced by ABCA3[
15,
81]. The decreased pool of mature SFTPB and SFTPB aggregates into the LBs, accumulation of large quantities of pro-SFTPB in LBs with leaks in the alveolar spaces, and abnormal processing of SFTPC were described in association with ABCA3 deficiency [
71].
Consequently, ABCA3 deficiency is characterized by abnormal structure LBs, abnormal lipid composition of the lung surfactant, and abnormal processing of surfactant proteins B and C [
68,
71,
75,
83,
84]. Reduced ability to decrease surface alveolar tension was demonstrated in patients with ABCA3 deficiency[
11]. Accumulation of dysfunctional, inefficient lung surfactant is associated with compromised gas exchanges through reduced diffusion barrier and increased discordance between ventilation and perfusion, decreased activity of the macrophages, and secondary lung lesions [
1,
17,
64,
72]. The biological mechanism of the lung lesions associated with ABCA3 deficiency is unknown [
15,
85,
86]. However, AEC II lesions are the final pathway to the associated lung disease, as AEC II represents the key factor for alveolar maintenance and repair[
24]. Inadequate lung surfactant production leads to recurrent atelectasis and hypoxemia followed by secondary chronic inflammation; coupled with abnormal intracellular surfactant metabolism, these changes lead to chronic AEC II lesions [
87]. Based on in vitro mechanistic studies, three classes of
ABCA3 gene variants can be identified: type 1 -
ABCA3 trafficking variants, characterized by abnormal protein folding, abnormal intracellular localization and trafficking; type 2 - complete deficiency of lipid transportation and variants affecting only the lipid transport due to deficient ATP hydrolysis, with normal trafficking and localization of the ABCA3 protein [
3,
60,
88,
89,
90,
91]; type 3 – a compound heterozygous of type 1 and 2, often associated with a more severe phenotype, early onset, neonatal RDS and neonatal death [
54].
On lung histology, ABCA3 deficiency is characterized by AEC II hyperplasia, variable degrees of interstitial thickening, prominent macrophages and proteinaceous material in the alveolar spaces, aspect frequently described as chronic or desquamative or non-specific interstitial pneumonitis or alveolar lung proteinosis; lung fibrosis is associated in fatal cases; however, these changes are frequently seen in other surfactant metabolism conditions, including SFTPB and SFTPC deficiency [
11].
As the clinical picture of ABCA3 deficiency is undistinguishable from SFTPB and SFTPC deficiency, even with lung histology [
13,
15], electron microscopy may help. The absence of normal LBs strongly suggests ABCA3 deficiency [
11,
25,
51]. Small, markedly abnormal LBs with dense phospholipidic membranes are characteristic of ABCA3 deficiency, as compared to SFTPB deficiency, which is characterized by disorganized LBs, with multiple vesicular inclusions dispersed in AEC II cytoplasm and alveolar lumen [
1,
24].
ABCA3 is expressed not only in the AEC II but also in the brain, kidney, and platelets [
22]. ABCA3 expression occurs in normal fetuses at 26-27 weeks of gestation [
4], even at 23-24 weeks, associated with lung inflammation [
71], and is developmentally regulated. It increases with the gestational age to reach a peak around term [
11,
13] under the influence of steroids and TTF 1 [
11,
71,
92,
93,
94].
4.3.2. Adenosine Triphosphate-Binding Cassette Family A Member 3 (ABCA3) Gene
ABCA3 synthesis is encoded by the
ABCA3 gene, intensely expressed in AEC II [
8,
15,
51].
ABCA3 gene variants are the most common cause of congenital lung surfactant defects [
2,
3,
11,
22,
25,
54,
72,
75,
85]. The first cases of ABCA3 deficiency, secondary to a bi-allelic loss of function
ABCA3 variant, were reported by Shulenin et al. [
51] in 2004, in full-term infants with unexplained severe RDS. Most
ABCA3 variants are challenging to interpret as few of them have been studied in vitro to identify their intracellular expression and function [
15,
54,
62,
75]. According to Wambach et al. [
95], the incidence of
ABCA3 variants is estimated between 1:4.400 and 1:20.000 in European and African descendants, most of them compound heterozygous [
72]. The incidence is probably overestimated as not all missense variants are pathogenic, and the annual number of cases identified is lower than expected according to the prediction calculations [
8,
95]. It is also possible that mild cases may not be recognized [
54]. In humans, the
ABCA3 gene has 80 kb, is located on chromosome 16 (16p.13.3) [
85], and comprises 33 exons [
11,
96]. Only 0.15% of the 274
ABCA3 gene variants reported in the genome Aggregation Database (gnomAD) are classified as pathogenic, 0.21% as likely pathogenic, and 92.62% were VUS [
70]. Twenty-five pathogenic or likely pathogenic of the
ABCA3 variants were reported in the ClinVar database, 11 of these with loss of function of the protein; 47 disease-causing
ABCA3 variants are reported in the in silico tool SIFT, 46 of these being also included in Polyphen-2, while 49 pathogenic mutations can be found in MutationTaster2 [
52]. Most reported variants are located on the exons or at the limit between introns and exons.
The expression of the
ABCA3 gene is directly proportional to gestational age [
17,
71]. Most
ABCA3 variants have an autosomal recessive inheritance [
8,
85,
97], but uniparental disomy was also reported [
98]. The most frequent pathogenic
ABCA3 variant is p.Glu292Val (E292V), representing 10% of the reported pathogenic alleles [
99]. This variant occurs in gnomAD with 0.23% allelic frequency [
52]. Pathogenic
ABCA3 variants are reported anywhere on the gene [
17]. NGS – WES, WGS - identifies new variants, most of them with unknown significance (VUS). In this situation, a correlation genotype-phenotype is impossible [
22,
75]. Therefore, the effect cannot be predicted accurately for all the reported variants, complicating the clinical decision, patient management, and familial counseling [
17,
52,
75]. Interpretation and counseling are also difficult in patients exhibiting a unique
ABCA3 variant on one allele [
11,
13,
22,
62,
64,
75]. Recently, it was suggested that
ABCA3 variants may be responsible for the increased severity of RDS in preterm infants compared to as expected according to their gestational age [
25,
50,
54,
83,
100]. In 2004, Shulenin et al. [
51] reported 12 different causative variants of the
ABCA3 gene in 16/21 neonates with severe, unexplained RDS. A study comprising 68 preterm infants with a gestational age <32 weeks of gestation with unusually severe RDS identified 24/68 heterozygous for previously described rare or new
ABCA3,
SFTPB, and
SFTPC variants, all VUS; 21
ABCA3 variants were found in 18 of the patients; 11 deaths were noted between 2 and 6 months of age and one infant presented histological aspects suggestive for ABCA3 deficiency [
69].
According to Peca et al. [
71], the ABCA3 deficiency phenotype depends on the residual function of the ABCA3 protein, mutation type and severity, the activity of the intracellular stress pathways, general individual aspects, other associated mutations, and modifiable environmental factors. Variable genotype-phenotype correlation has been associated with
ABCA3 variants [
8,
22,
70,
75]; diverse symptoms, severity, and outcomes are associated with
ABCA3 variants [
63]. However, interactions with other variants (as, for example, variants of
SFTPB or
SFTPC) [
22,
62,
71,
95,
101,
102,
103] or with external, environmental factors (for example, respiratory infections or smoking) [
22,
50,
64,
79,
104] may induce changes of the phenotype. Similar mechanisms were suggested for mono-allelic variants of the
ABCA3 gene [
71,
95]. According to Yang et al. [
105], loss of over 50% of the ABCA3 protein function is associated with increased morbidity and mortality. A critical level for ABCA3 protein function of 20-30% was estimated by Wambach et al. [
19]. Usually, bi-allelic mutations of the
ABCA3 gene are associated with loss-of-function of ABCA3 protein and severe RDS with neonatal onset. Missense mutations, insertions, and small deletions are typically associated with the residual function of the protein [
52,
64]. Both bi-allelic and mono-allelic variants may present with RDS [
51,
63,
95,
101,
103]. Null/null mutations – nonsense and frameshift – result in a truncated, non-functional ABCA3 protein [
54,
71] associated with a more severe phenotype, the neonatal onset of RDS, death before one year of age, need for lung transplantation, or death even with lung transplantation [
9,
52,
62]. Late preterm and term infants with homozygous and compound heterozygous
ABCA3 variants were associated with earlier presentation of severe RDS, higher radiological scores, and increased mortality rates (all p<0.05) as compared to infants with single mutations or no genetic abnormalities identified [
14]. Beers et al. [
106] suggested a synergistic additive effect for compound heterozygous in the
cis region of the gene. Lack of gene expression, decreased expression, abnormal intracellular protein trafficking inside LBs, abnormal phospholipid folding, and functional defects, including ATP hydrolysis, were described as consequences depending on the
ABCA3 locus [
1]. Fatal cases of ABCA3 deficiency were described in association with abnormal trafficking, while defects in phosphatidylcholine were correlated with less severe lung disease [
88,
89].
4.3.3. ABCA3 c.838C>T (p.Arg280Cys, R280C) Variant
In vitro functional studies performed by Weichert et al. [
86] have suggested that this variant can lead to partial retention of the ABCA3 protein in the endoplasmic reticulum and is involved in epithelial lung cells apoptosis at least one pathway, altering ABCA3 protein function (type 1, trafficking/folding defect based on in vitro studies). ABCA3 protein retention in the endoplasmic reticulum may increase reticulum endoplasmic stress and its susceptibility to stress; an adverse effect of R280C on LBs biogenesis and induced presence of apoptotic markers (glutathione on caspase 4 pathway) were demonstrated in the experimental epithelial lung cells in Weichert et al. experiments [
86], also suggestive for functional impairment on ABCA3 protein and lung disease pathogenesis. These experiments confirmed previous studies by Matsumura et al. [
88,
89] that defined the c.838C>T (p.Arg280Cys, R280C) variant as disruptive of ABCA3 folding and trafficking. Based on increased retention of ABCA3 protein in the endoplasmic reticulum, ABCA3 variants F1203del, N124Q, N140Q, and R280C may be classified as potentially pathogenic, according to other experts [
105].
Furthermore, multiple computational predictive in silico tools and conservational analysis also indicate the negative impact of the c.838C>T (p.Arg280Cys, R280C) variant on the ABCA3 protein function [
55,
65,
66]. Nevertheless, currently, the c.838C>T (p.Arg280Cys, R280C) variant is listed as VUS, considering the existent evidence insufficient to define the mutation’s pathogenicity. Pros and cons arguments on c.838C>T (p.Arg280Cys, R280C) pathogenicity are presented in
Table 2.
The initial lung imaging – usually resembling that seen in preterm infants with RDS [
3,
22,
25,
59,
71] – evolves throughout the disease and may vary over time, as it happened also in our patient [
8]. Later in the course of the disease, a nodular and consolidation pattern may be observed [
96]. Different lung imaging aspects are expected as even identical variants of the
ABCA3 gene may present with different phenotypes [
112]. The same observation applies to lung histology. All these aspects are described in the literature in association with ABCA3, SFTPB, and SFTPC deficiency [
13,
15]. Electron microscopy of the lung tissue was reported only in the patient presented by Jackson et al. [
64].
Familial history had no relevance for ABCA3 deficiency in the several reviewed cases, as in our family. This is not unexpected as ABCA3 deficiencies are rare diseases with autosomal recessive inheritance. Most patients were compound heterozygous for the
ABCA3 c.838C>T (p.Arg280Cys, R280C) variant, which, together with the association of various other
ABCA3 variants, may explain the different phenotypes of the subjects (
Table 1). Additionally, there is an urgent need for functional studies to quantify the impact of each variant on protein function, expressed as a percentage.
Our case, a compound heterozygous for ABCA3 c.838C>T (p.Arg280Cys, R280C), and ABCA3 c.697C>T (p.Gln233Ter, Q233X, Q233*) variants, presented with fatal RDS with neonatal onset. Most probably, a cumulative damaging effect of the c.697C>T (p.Gln233Ter, Q233X, Q233*) variant (in silico predicted as probably pathogenic) significantly contributed to the severe phenotype of our patient.
We support the recommendation that in cases of unexplained, early onset, severe neonatal RDS with persistent radiological and clinical symptoms persistent over one week, evolving to hypoxemic respiratory failure despite maximal conventional therapy and with transient response to surfactant administration in term and near-term infants, genetic surfactant metabolism dysfunctions should be suspected [
6,
8,
9,
13,
15,
79,
86]. NGS offers a crucial role in molecular medicine, a step forward to individualized medicine, as this genetic testing precisely detects single nucleotide variants [
70]. NGS, WES and WGS, can help the index case and their family identify the genetic defect and provide genetic counseling. Moreover, extended genetic sequencing may replace the information offered by lung biopsy, an invasive investigation previously recommended in assessing inherited surfactant metabolism disorders [
4,
6,
9,
13,
71], as a rapid and precise diagnosis is of utmost importance for genetic counseling [
10,
96]. These techniques may identify VUS that need predictive tools for clarifying the impact on protein function and pathogeny and further genetic counseling [
15,
70,
110].
Our study has several additional limitations. First, the presented patient and his family were investigated by an NGS panel that included a limited number of genes, thus not providing a comprehensive analysis of their genome. Additionally, we could not perform functional studies to quantify the effect of the ABCA3 variants, which collectively impacted protein function and led to RDS. Second, we did not assess gene expression or measure the concentration of ABCA3 protein. However, the clinical, paraclinical, imaging, and histological findings, predictive tools, and the results of previously reported studies strongly suggest the pathogenicity of the associated variants in our patient and the need to reclassify these ABCA3 variants.