Proteomic identification of altered apolipoprotein patterns in pulmonary hypertension and vasculopathy of sickle cell disease

Abstract

Pulmonary arterial hypertension (PAH) is emerging as a major complication and independent risk factor for death among adults with sickle cell disease (SCD). Using surface-enhanced laser desorption/ionization time of flight mass spectrometry (SELDI-TOF MS), we searched for biomarkers of PAH in plasma specimens from 27 homozygous sickle cell anemia (HbSS) patients with PAH and 28 without PAH. In PAH patients, analysis consistently showed lower abundance of a 28.1-kDa peak (P < .001), identified by high-resolution mass spectrometry as the oxidant-scavenging protein apolipoprotein A-I (apoA-I), which correlated with clinical assays of apoA-I (r = .58, P < .001) and high-density lipoprotein (HDL) levels (r = .50, P = .001). Consistent with endothelial dysfunction that may mediate this effect in PAH, HbSS patients with lower apoA-I levels also displayed impaired vasodilatory responses to acetylcholine (mean +/- SEM, 189% +/- 34% [n = 13] vs 339% +/- 51% [n = 13], P < .001). As a group, patients with SCD demonstrated significantly lower apoA-I levels than African-American control subjects. The PAH cohort was further characterized by high levels of apolipoproteins A-II and B and serum amyloid A, and low levels of haptoglobin dimers and plasminogen. These results imply a relationship of apolipoproteins to the development of PAH vasculopathy in SCD, potentially involving an unexpected mechanistic parallel to atherosclerosis, another proliferative vasculopathy.

Figure 1

Representative SELDI-TOF MS spectra from plasma from patients with SCD with and without PAH. The vertical axes represent intensity of peaks in arbitrary units, and the horizontal axes represent mass to charge ratio (m/z). These 4 spectra were obtained from plasma eluted from ion anion exchange resin, bound to IMAC30-Cu⁺⁺ matrix, and ionized for SELDI-TOF MS. Two specimens are from SCD patients without PAH (A,B) and 2 are from SCD patients with PAH (C,D). A peak at m/z 28.1 kDa was observed at lower average intensity in SCD patients with PAH ().

Figure 2

Peptide mapping of apoA-I by MALDI mass spectrometry. (A) MS and MS/MS analysis of a trypsin-digested gel splice resulted in matching of most of the major peaks as either apoA-I (peaks indicated by nos. 1-21), immunoglobulin kappa light chain (IgK, *), or trypsin autolysis products (•). No other proteins were identified as significant matches. (B) A total of 21 different peaks matched to apoA-I with MS/MS sequencing of 3 of these matching 19, 12, and 11 fragment ion peaks. (C) Matching peptides indicated in bold represent 66% of the sequence and are distributed throughout the sequence.

Figure 3

Correlation of the m/z 28 101 peak with apoA-I and HDL-C assays. The intensities of the m/z 28 101 from all patients with or without PAH peak are plotted against each clinical laboratory measurement of apolipoprotein A-I (apoA-I) and high-density lipoprotein cholesterol (HDL-C) from the same specimens. The Spearman correlation coefficients and P values are indicated.

Figure 4

Prevalence of pulmonary arterial hypertension (PAH) by apolipoprotein quartiles. (A) Low apoA-I levels trend with high prevalence of PAH. The apoA-I levels from Figure 1 were grouped into the lowest, highest, and middle 2 quartiles, and the percentage of patients with TRV 2.5 m/s or higher was calculated. There is a strong trend toward higher prevalence of PAH in the lower apoA-I group in this analysis (P = .06, χ² test for trend). (B) A similar analysis for apoB shows the highest prevalence of PAH in the highest quartile of apoB levels (P = .05) and (C) even more strikingly, with high apoB/apoA-I ratio (P = .006).

Figure 5

Low apoA-I level is a marker for endothelial dysfunction. Forearm blood flow was measured in 26 patients with SCD with venous occlusion plethysmography following test doses of acetylcholine (ACh) and sodium nitroprusside (SNP) infused into the brachial artery. Patients with higher than median levels of apoA-I () demonstrated dose-dependent vasodilation to ACh close to previously published normal values. In sharp contrast, those with lower than median levels of apoA-I had markedly blunted responses measured as (A) absolute blood flow (P < .001, 2-way ANOVA) or as (B) percentage of increase from baseline (P = .001). Although the (C) absolute blood flow was lower at baseline and all doses of SNP in low apoA-I patients (P = .007), the (D) percentage increase from baseline did not differ by apoA-I status.