Peripheral Blood Inflammation Profile of Patients with Pulmonary Arterial Hypertension Using the High-Throughput Olink Proteomics Platform

To the Editor:

Inflammation is a driver of pulmonary arterial hypertension (PAH), a chronic and progressive disease of increased pulmonary vascular resistance that results in right ventricular failure (1, 2). Infiltrating inflammatory cells are increased around and within pulmonary vascular lesions and are important sources of cytokines, chemokines, and growth factors, which dictate the proremodeling pulmonary vascular microenvironment (1, 3). Furthermore, perivascular inflammation contributes to intima and media remodeling in PAH (4, 5). Because increased cytokines and chemokines in the peripheral blood of patients with PAH correlate with survival and disease severity (1), peripheral biomarkers may reflect PAH pathobiology (6, 7). Sweatt and colleagues identified four distinguished proteomic immune clusters within patients with idiopathic pulmonary arterial hypertension (IPAH) using a multiplex immunoassay composed of 48 cytokines, chemokines, and growth factors (8). Their findings underscore the need for high-throughput approaches allowing simultaneous measurements of inflammation markers.

In this pilot study, we addressed whether a high-throughput protein detection methodology, such as Olink, would reveal novel biomarkers or distinct protein signatures that may reflect pulmonary hypertension (PH) pathobiology. Instead of a broad approach used by other high-throughput proteomic platforms, we used a targeted approach by analyzing protein biomarkers previously described to be involved in other inflammation-driven diseases. Simultaneous analysis of 92 biomarkers in the peripheral blood of patients with clinically characterized IPAH and patients with scleroderma-associated pulmonary arterial hypertension (SSc-PAH) identified 20 differentially expressed markers, including potential novel ones, such as AXIN-1 (axis inhibition protein 1), IL-15RA (interleukin 15 receptor subunit alpha), X4EBP1 (eukaryotic translation initiation factor 4E-binding protein), ST1A1 (sulfotransferase 1A1), and STAMBP (signal-transducing adaptor molecule binding protein).

Peripheral blood of patients with group-1 PAH was biobanked from patients from the PH clinics at either the University of Colorado Anschutz Medical Campus (n = 31) or Stanford University (n = 23) (Table 1). Blood from healthy control subjects (CTL) from both sites was also collected. Plasma was processed within 2 hours of collection, never thawed before testing, and kept in a monitored −80°C freezer for a period of 6 months to 6 years. The samples were tested using the Olink inflammation marker panel (https://www.olink.com/products/inflammation/#). Briefly, plasma was added to the plate, where labeled antibody probe pairs bind to their target proteins. A PCR reporter sequence was amplified, detected, and quantified using high-throughput real-time PCR. Results were calculated by relative quantification and presented using normalized protein expression values. To validate our Olink findings, we used standard AXIN-1 and CXCL1 ELISAs (LSBio and Abcam, respectively). Batch effects were detected by principal component analysis and corrected using the empirical Bayes method (ComBat algorithm) (9). ANOVA was used with a Benjamini-Hochberg procedure to decrease the false discovery rate. A heatmap was calculated on the basis of Euclidean pairwise distances for rows and Hamming distances for columns. For the ELISA, the different groups were compared using ANOVA (with Tukey multiple comparisons test).

Table 1.

Summary Table with Subject Ages, Pulmonary Arterial Hypertension Etiology, and Clinical Features

	Healthy Control Subjects (n = 14)	Subjects with IPAH (n = 27)	Subjects with SSc-PAH (n = 27)
Sex, n	10 F/4 M	16 F/11 M	25 F/2 M
Race/ethnicity, n	10 White/4 non-White*	18 White/9 non-White*	17 White/10 non-White*
Age, yr, median (range)	53.64 (48–64)	61.75 (32–70)	66 (21–83)
6MWD, ft, median (range)	N/A	431.9 (227.07–2,060)	420.6 (15–1,798)
BNP, pg/ml, median (range)	N/A	120.5 (12–13,622)	211 (7–12,565)
mPAP, mm Hg, median (range)	N/A	43 (26–75)	36.5 (20–70)
FC, median (range)	N/A	3 (1–4)	3 (1–4)

Definition of abbreviations: 6MWD = 6-minute walk distance; BNP = brain natriuretic peptide; F = female; FC = functional class; IPAH = idiopathic pulmonary arterial hypertension; M = male; mPAP = mean pulmonary arterial pressure; N/A = not applicable; SSc-PAH = scleroderma-associated pulmonary arterial hypertension.

^*Non-White comprises Hispanic, African American, and Asian.

We found 11 inflammatory markers augmented in patients with SSc-PAH and patients with IPAH: VEGFA, CD8A, CXCL9, MMP1, IL-15RA, PD-L1, HGF, TNF, X4EBP1, CCL28, and CD40. Other markers were increased only in IPAH (AXIN-1, ST1A1, CXCL1, STAMBP, SIRT2, and CASP8) or SSc-PH (TNF-β, CX3CL1, and uPA) (Figure 1 and Figure E2 in the data supplement). To confirm our findings, we performed AXIN-1 and CXCL1 ELISAs on the University of Colorado specimens and found that both markers were also increased in the patients with IPAH compared with the CTL and SSc-PAH groups (P = 0.0394 and 0.0153 for AXIN-1 and P = 0.0084 and 0.0242 for CXCL1, respectively) (Figure E1). We were surprised to find that traditional markers such as IL-6 were not significantly elevated. In fact, IL-6 was more abundant in IPAH and SSc-PAH than in CTL (P = 0.05167) before false discovery rate adjustment, but this increase fell below statistical significance after accounting for multiple comparisons. This could be owing to a limited number of samples in this preliminary study.

Figure 1.

Heatmap of the 20 inflammation markers differentially expressed in control subjects (CTL), subjects with idiopathic pulmonary arterial hypertension (IPAH), and subjects with scleroderma-associated pulmonary arterial hypertension (SSc-PAH), clustered by proteins and subjects (z-scores).

The differentially expressed cytokines, chemokines, and growth factors participate in leukocyte recruitment, angiogenesis, tissue remodeling, and apoptosis inhibition. AXIN-1, IL-15RA, X4EBP1, ST1A1, and STAMBP have not previously been investigated as being involved in PAH pathogenesis. AXIN-1 is part of the β-catenin degradation complex in the cell cytoplasm, activated when Wnt does not bind to the cell receptor Frizzled. Loss of Wnt signaling results in vascular remodeling and impaired pulmonary vascular regeneration (10). IL-15RA is the receptor of the cytokine IL-15 and is present not only on cell membranes but also in a soluble form. When bound to IL-15, it affects proliferation, survival, and differentiation and targets a wide range of cell populations (11). ST1A1 is a sulfotransferase, a cytoplasmic enzyme involved in metabolism of circulating compounds, including drugs, carcinogens, and hormones, and has not been described previously in PH (12). X4EBP1 regulates protein translation and is involved in tumorigenesis (13). Finally, STAMBP is a metalloprotease and a member of the Jab1/MPN metalloenzyme family of deubiquitinases, reported to be involved in inflammasome regulation and cytokine secretion (14).

An unsupervised cluster analysis of the 20 different markers demonstrated that different PH etiologies tend to cluster. It is also notable that CXCL1, ST1A1, AXIN-1, SIRT2, STAMBP, and CASP8 tend to be more elevated only in patients with IPAH and may reflect its pathogenesis. Further studies are warranted to better understand the role of these particular biomarkers in this disease.

A limitation of the present study is the need to further validate the uncovered potential markers in larger cohorts of PAH lungs (5). Validation of expression of these cytokines and chemokines in the pulmonary vasculature may strengthen the concept that peripheral blood analysis may be a suitable approach to studying the inflammatory phenotype of living patients, and give insight into the pathophysiology of the disease.

In summary, we observed 20 inflammation-related markers distributed differently among IPAH and SSc-PAH etiologies, which may assist in characterizing disease subgroups. Further studies are required to clarify their roles in PAH pathogenesis.