Identification and confirmation of haptoglobin as a potential serum biomarker in hypertrophic cardiomyopathy using proteomic approaches

Figures & data

Figure 1. Experimental design. A: Serum samples handling was optimized by removing salts and other potentially interfering substances. B: DIGE analyses revealed proteins differentially expressed in a significant manner between serum specimens from HCM and control patients. C: Preparative new gels were visualized using Coomassie staining for picking up selected spots. D: Proteins of interest were identified using liquid chromatography coupled with MS/MS. E: Nephelometry served to validate the HCM diagnostic properties of identified proteins on independent series of serum samples.

Table I. Information of patients and controls providing serum samples included for 2D-DIGE analysis.

	Patients ID	Age	Controls ID	Age
Pool 1	163207	21	161737	21
	212773	22	983257	23
	212811	22	21025844	23
	163212	23	21025841	24
Pool 2	163216	25	161702	25
	163203	27	163123	27
	163078	29	161700	29
	163066	30	163095	30
Pool 3	163208	30	161419	30
	163173	33	1390519	33
	212792	33	163089	33
	212775	39	161848	40
Pool 4	163246	41	163090	41
	163171	42	161814	43
	212784	43	161816	43
	212803	43	161863	43
Pool 5	212770	44	163134	44
	212726	46	161822	47
	212779	48	161859	49
	163189	51	161802	49

Table II. Demographic and clinical characteristics of patients included in the 2D-DIGE analysis.

	Patients (20)
Age	34.3 ± 13.6
Sex male (%)	20 (100)
Hypertension (%)	3 (15)
Diabetes mellitus (%)	0 (0)
NYHA class II–IV (%)	4 (20)
Previous syncope (%)	2 (10)
Previous AF (%)	1 (5)
Previous angina (%)	4 (20)
Previous dyspnoea (%)	7 (35)
Maximum LV wall thickness (mm)	23.58 ± 8.18
Obstructive (gradient > 30 mmHg) (%)	4 (20)
LV ejection fraction (%)	69.9 ± 8.8
Non-sustained ventricular tachycardia (%)	6 (30)
LGE (%)	13 (65%)
Number of LGE segments/patient	2.50 (0.00–5.75)
Haptoglobin (g/L)	0.91 ± 0.57
Hemopexin (g/L)	0.76 (0.68–0.82)

NYHA = New York Heart Association; AF = atrial fibrillation; LV = left ventricle; LGE = late gadolinium enhancement in cardiac magnetic resonance.

Figure 2. Representative 2D-DIGE staining of proteomic profiling from serum samples. A: Protein extract from HCM patients; and B: proteome extract from sex- and age-matched controls. Spots marked with a number indicate identified proteins whose abundance changes between HCM and control specimens reached statistical significance.

Table III. Demographic and clinical characteristics of patients included in the nephelometric immunoassay.

	Patients (181)
Age	49.4 ± 15.1
Sex male (%)	126 (69.6)
Hypertension (%)	64 (35.4)
Diabetes mellitus (%)	16 (8.8)
NYHA class II–IV	37 (20.4)
Previous syncope (%)	19 (10.5)
Previous AF (%)	30 (16.6)
Previous angina (%)	40 (22.1)
Previous dyspnoea (%)	80 (44.2)
Maximum LV wall thickness (mm)	20.5 ± 5.5
Obstructive (gradient > 30 mmHg) (%)	60 (33.1)
LV ejection fraction (%)	69.1 ± 9.9
Non-sustained ventricular tachycardia (%)	61 (33.7)
LGE (%)	119 (65.7)
Number of LGE segments/patient	2.00 (0.00–6.00)
Haptoglobin (g/L)	0.91 ± 0.68
Hemopexin (g/L)	0.82 (0.70–0.93)

NYHA = New York Heart Association; AF = atrial fibrillation; LV = left ventricle; LGE = late gadolinium enhancement in cardiac magnetic resonance.

Figure 3. Three-dimensional view of differentially expressed proteins in serum samples from HCM patients compared with controls. Changes in protein expression from HCM and controls are shown by each line. The average ratio of expression for each selected protein grouping in HCM and control specimens as obtained by computational analysis with DeCyder Software is represented by lines marked with a cross. Statistical analysis allowed detection of significant abundance changes (95th confidence level) based on the variance of the mean change within the cohort: A: haptoglobin; B: hemopexin; C: Ig alpha-1 chain C region isoform; D: albumin.

Table IV. Significant differentially expressed proteins identified by reverse-phase liquid chromatography coupled with MS/MS after 2D-DIGE analysis of HCM patients versus control serum samples.

Spot number	Accession number (Swiss Prot)	Name	Size (Da)	pI	t test	Unique peptides	Sequence coverage	Score	Average fold range ratio	Function (NCBI and Swiss Prot)
1296	P00738	Haptoglobin	45205	6.13	0.015	8	22.91%	1469.8	–1.55	Haemoglobin metabolic process and cellular iron ion homeostasis
665	P02768	Albumin	47360	5.97	0.0056	23	41.22%	4132.4	+ 3.32	Different molecules binding and transport
612					0.0004	18	39.74%	5114.5	+ 2.44
656					0.043	15	33.28%	1654.9	+ 2.14
643					0.0086	11	24.91%	653.2	+ 1.95
677					0.016	9	16.58%	52.38	+ 2.49
680					0.018	3	5.25%	16.65	+ 1.90
693	P02790	Hemopexin	49295	6.43	0.018	1	2.38%	5.82	+ 2.44	Haemoglobin metabolic process and cellular iron ion homeostasis
743	P01876	Ig alpha-1 chain C region	37654	6.08	0.035	3	11.65%	23.42	+ 1.54	Immune response, antigen binding
1062		Unknown
1067		Unknown

Table V. Correlation of haptoglobin and hemopexin serum levels with HCM.

	Patients Average ± SD or Median (IQR)	Controls Average ± SD or Median (IQR)	P
Haptoglobin	0.91 ± 0.68	1.237 ± 0.441	< 0.001
Hemopexin	0.824 (0.700–0.925)	0.837 (0.751–0.954)	0.113

SD = standard deviation; IQR = interquartile range.