Introduction
Chronic obstructive pulmonary disease (COPD) is a major cause of morbidity and mortality. Long term inhalation of noxious gases including cigarette smoke leads to airflow obstruction and emphysema in the lungs. On the other hand persistent systemic inflammation and oxidative stress are common features of this disease. Currently, soluble Receptor for Advanced Glycation End products (sRAGE) is considered as a potential biomarker for emphysema, based on data showing emphysema as independent predictor of decreased levels of sRAGE in the circulation of COPD patients [
1,
2].
RAGE is a transmembrane, multi ligand, pattern-recognition receptor that belongs to the immunoglobulin super family of cell surface receptors [
3,
4]. The human RAGE gene (
AGER) consisting of 11 exons, produces membrane RAGE (mRAGE) which comprises an extracellular, a transmembrane, and cytosolic domain. Alternative splicing of the
AGER gene leads to the formation of endogenous soluble RAGE (esRAGE) [
5]. On the other hand, ectodomain shedding of mRAGE by metalloproteinase such as MMP9 and A disintegrin and metalloprotease (ADAM) 10 generates soluble RAGE (sRAGE) [
6,
7]. Soluble forms of RAGE found in the circulation can act as decoy receptors for Advanced Glycation Endproducts (AGEs), preventing binding of AGEs to cell bound full length RAGE and downstream activation of NF-kB [
8].
Among many different extracellular ligands of RAGE, we previously examined levels of three different protein-bound AGEs in the plasma of COPD patients. In summary, N
ϵ-(carboxymethyl) lysine (CML) was decreased, N
ϵ-(carboxyethyl) lysine increased and pentosidine was not different in plasma of COPD patients compared to controls [
9]. Literature survey shows that among several RAGE ligands investigated serum amyloid A (SAA), S100 protein A12 (S100A12) and High- mobility group protein B1 (HMGB1) were found to be increased in COPD patients compared to controls [
10,
11].
Although plasma sRAGE is considered as a biomarker for emphysema, no mechanisms have been proven to date that underlie this association. The most straightforward mechanism is that the lower plasma level of sRAGE is a reflection of the disappearance of pneumocytes which typically display very high RAGE expression [
12]. A complicating factor is that it is assumed that the major soluble form of RAGE in the circulation is sRAGE. Recently however, a study identified an important contribution of esRAGE to total sRAGE as well as positive correlations of esRAGE to total sRAGE in BAL fluid and serum of asthmatics and COPD patients [
13] Unfortunately, this manuscript did not report on differences in esRAGE levels between COPD patients and controls, or relations of esRAGE to disease markers.
In order to gain further evidence for plasma sRAGE as a potential biomarker in COPD, we measured plasma levels of esRAGE in patients with COPD and a control group. Furthermore, we established possible relations of esRAGE with different circulating AGEs, and disease specific characteristics such as lung function, DLCO and use of external oxygen. Circulating sRAGE levels were measured as well to identify the contribution of esRAGE to total sRAGE.
Discussion
The present study demonstrates decreased levels of esRAGE and total sRAGE in plasma of COPD patients compared to controls. Importantly, in COPD, total sRAGE and not esRAGE showed positive associations with FEV1, FEV1/VC, and DLCO, and only sRAGE levels were negatively and independently associated with LTOT. Further, only total sRAGE levels were positively associated with CML, a major ligand of RAGE.
Most of the studies that report plasma sRAGE levels measured using a commercially available ELISA that detects total sRAGE, thus including esRAGE. Only one recent publication measured esRAGE separately in COPD. This study identified an important contribution of esRAGE to total sRAGE and positive correlations of esRAGE to sRAGE in BAL fluid and serum of asthmatics and COPD patients [
13]. However, it did not report on differences in esRAGE levels between patients and controls, or relations to disease markers. Data from the present study confirm a moderate positive correlation of plasma esRAGE with total sRAGE levels and for the first time demonstrate decreased plasma esRAGE levels in COPD.
Two different mechanisms producing total sRAGE are by alternative splicing of
AGER gene which results in secretion esRAGE, and by ectodomain shedding of membrane RAGE by MMP9/ADAM10 which leads to the secretion of sRAGE [
21]. Both esRAGE and total sRAGE were observed to be lower in the circulation of COPD patients. Smoking per se seemed to affect total sRAGE levels as there was a significant difference between never and ex-smoking controls. This was however not the case for esRAGE (Figures
2A and
1A respectively), indicating that smoking affects shedding but not alternative splicing. Still it is important to further evaluate the effect of smoking as a cause for decreased sRAGE/esRAGE. Furthermore, the pattern of plasma sRAGE and esRAGE levels in the circulation of individual COPD patients and controls were found to be different. Based on these findings it is difficult to conclude whether alternative splicing or shedding or both are operational, since there is no literature on altered
AGER gene expression and ADAM10 in the pulmonary/extra pulmonary compartment in COPD. However there is evidence in the literature of increased MMP9 protein levels in BAL of COPD patients [
22,
23], whereas an absence of difference in MMP9 levels and activity in the circulation of COPD patients was reported [
24]. Further research is needed to determine whether altered gene expression, alternative splicing, and/or sheddases activity are responsible for decreased sRAGE levels in circulation.
Lowered levels of both esRAGE and sRAGE in the circulation of COPD patients could also be associated with the presence of functional single nucleotide polymorphisms (SNPs) in the
AGER gene. Sequence variation studies have shown that there are 50 SNPs in 11 exons and 10 introns in the
AGER gene. Among them rs2070600 [
15] was associated with lower levels of plasma sRAGE in patients with diabetes and also in COPD [
25,
26]. Information regarding associations of circulating esRAGE levels with polymorphisms is limited. In type 2 diabetes it has been shown that rs2070600 and rs1800625 are associated with esRAGE levels in a Chinese cohort [
27]. Given these genetic associations to circulating sRAGE levels in diseases affecting different primary organs and the contrasting increased pulmonary expression of RAGE in COPD patients, it remains questionable if lower systemic sRAGE levels indeed arise from altered pulmonary levels, and if so how [
26,
28]. There is one recent manuscript showing no decrease in plasma sRAGE level in patients with COPD with or without chronic heart failure [
29], findings that are in contrast with many others, including our own showing decreased plasma total sRAGE levels in COPD [
1,
10,
30].
Findings of the study presented here are in line with literature showing that plasma total sRAGE levels are associated with lung function in COPD [
1,
10,
30,
31]. In addition we evaluated whether comparable trends could be observed for esRAGE as both receptors have similar structures and functions, originating from the same gene through different mechanisms. In contrast to total sRAGE, we did however not find significant associations of plasma esRAGE with lung function parameters. Although the lack of CT data is a limitation of the current study, the finding that D
LCO in most previous studies also correlated with plasma sRAGE levels, was replicated here as well. In addition we provide evidence that plasma total sRAGE levels are associated with the use of LTOT, even after adjustment for possible confounders and lung function. Again, such association with D
LCO or oxygen use was not present for esRAGE. The mechanism of decreased systemic total sRAGE levels in COPD patients, and its association to alveolar damage is still not clear. From the results of the present study we can concluded that total sRAGE and not esRAGE resulting from splice variation of the
AGER gene contributes to this association. Mechanisms involved in shedding of the receptor are likely involved, including via activation of sheddases like ADAM10 and MMP9, or via ligand engagement.
We investigated correlations of total sRAGE and esRAGE with measured plasma AGEs. The only significant correlation observed was between sRAGE and CML. RAGE ligands other than AGEs have been measured in the systemic circulation of patients with COPD. HMGB1, S100A12and SAA were found to be increased in COPD [
10,
11,
32], Among these ligands, only SAA showed a negative association with sRAGE. We previously reported decreased plasma CML levels in COPD and we showed here that this is positively correlated with sRAGE. Accelerated formation and accumulation of AGEs is due to inflammation and oxidative stress [
33]. However, we previously found that alterations in plasma levels of CML were independent of CRP, as a marker of inflammation [
9]. With our previous [
30] and present data we furthermore did not find significant correlations of either sRAGE or esRAGE with CRP. Only one study was able to demonstrate a negative association of sRAGE with CRP in COPD [
10], but a clear relationship between sRAGE and inflammation and oxidative stress has not been found so far. Further investigations are clearly needed that assess the association of sRAGE/esRAGE with comprehensive panels of RAGE ligands as well as inflammatory markers in large cohorts of patients.
In contrast to the recent attention given to sRAGE in lung research, little is known about the role of esRAGE in pulmonary physiology and pathology. It has been shown that not only vascular endothelial cells express esRAGE, but also neuronal, thyroid follicular, hepatocytes and pancreatic beta cells could be a possible source of circulating esRAGE [
34]. Plasma levels are in addition determined by excretion via the kidneys. Plasma esRAGE is known to be strongly affected by renal insufficiency [
35]; in agreement we observe a significant negative correlation of esRAGE with GFR. The current study shows decreased esRAGE levels in COPD, but no associations with disease characteristics. The observed decrease in esRAGE levels in COPD is important for future investigations into molecular mechanism behind the decreased levels of total sRAGE demonstrated often already and proposed as a biomarker for the disease, but still warrants further confirmation. Secondly, to establish whether systemic sRAGE alterations are causatively related to the development of structural emphysema or a systemically reflective consequence further longitudinal research is warranted in non-obstructed cigarette smokers.
In summary, both esRAGE and sRAGE are decreased in the circulation of COPD patients compared to controls. Only sRAGE showed a positive association with lung function and DLCO, a negative association with CML and was affected by oxygen treatment in COPD. We conclude that, although these molecules are products of the same gene they may have different roles in the pathogenesis of COPD and that primarily total sRAGE is a candidate biomarker for COPD phenotypes.
Acknowledgement
This work was performed at CIRO +, centre of expertise for Chronic Organ Failure, in Horn, The Netherlands. We thank the COPD patients and the healthy subjects who volunteered to participatein the study. Moreover we are grateful to Trineke Hofstra, BSc; Ans Suntjens, BSc; MarcoAkkermans, BSc; Linda Op ‘t Veld, MSc; Koen Stakenborg, BSc; Jos Peeters, BSc; MartijnCuijpers, MSc; Annie van de Kruijs, RN; Irma Timmermans, RN; Miriam Groenen, MSc and Riny van Kessel, RN for planning and performing all the tests.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ER and NR designed the study and ER provided the samples. PG and JS performed the analysis; CGS provided the equipment and reagents for AGE analysis. FF was a study physician. PG contributed for statistical analysis, PG, NR and ER drafted the manuscript, ER, NR, FF, EW, CGS and PG contributed to acquisition and interpretation of data. All authors read and approved the final version of manuscript.