Does the Rewarmed Heart Restore the Myocardial Proteome to That of the Pre-Cooled State?　– A Proteomic Analysis of Surgical Samples –

Abstract

Background: Hypothermia is utilized in cardiac and aortic surgery to protect organs from ischemic reperfusion injury. Although the cooled body is invariably rewarmed after the procedure, it is still unknown whether the rewarmed body regains its former biological state. This study determined the modulatory effects of hypothermia on the human myocardial proteome and whether subsequent rewarming restores the proteome to the state prior to cooling.

Methods and Results: A quantitative proteomic analysis was performed using isobaric tags for relative and absolute quantification labeling tandem mass spectrometry. Right atrial samples were taken 3 times (pre, during and post cooling) during deep hypothermic cardiopulmonary bypass (CPB) from 8 patients with aortic arch aneurysms and 3 corresponding time points during normothermic CPB from 8 patients with ascending aortic or valsalva aneurysms. In total, 697 proteins were identified, with 222 proteins having high protein confidence. Bioinformatic analyses revealed significant downregulation of 19 proteins associated with energy production at hypothermic cardioplegic arrest. On rewarmed beating, 10 proteins remained downregulated, including those regulating cardiac contraction and adaptor proteins, although levels of the aforementioned 19 downregulated proteins returned to their initial values. Additional echocardiographic evaluation demonstrated that hypothermia preserved the variables of diastolic function to a greater extent than normothermic surgery.

Conclusions: Rewarming restores the human myocardial proteome to the pre-cooled state, except for proteins regulating cardiac contraction and adaptor proteins. (Circ J 2015; 79: 2648–2658)

Hypothermia protects organs from ischemic reperfusion injury, mainly by reducing energy metabolism.^1,2 Myocardial oxygen consumption decreases to 50% of normothermic control by reducing temperature by 10℃ (ie, Q₁₀=2).^2–4 However, it remains unknown how and what biochemical changes support such a dynamic reduction of energy metabolism, although there have been some studies focusing on myocardial energy substrate selection (ie, either fatty acid or glucose utilization).^5,6 Several studies have focused on well-recognized pathways or targeted proteins, demonstrating that downregulation of apoptotic pathways could be a major mechanism of protection. However, research in each case was restricted to a small portion of the known whole system regulating apoptosis according to the well-known KEGG (Kyoto Encyclopedia of Genes and Genomes, http://kegg.jp/kegg) pathway such as the apoptosis map (map04210).^7–10 To date, there has not been a proteomic study leading to discovery of an unknown protein biomarker or global proteomic change. After hypothermia is induced by body cooling and maintained for a certain period of time, rewarming is an invariable procedure. This simply implies that the protective effect of hypothermia may be canceled by rewarming. Whether hypothermia-induced global proteomic changes could be fully restored by rewarming has never been investigated. Isobaric tagbased quantitative mass spectrometry has been developed and utilized as a reliable highthroughput proteomics technology.¹¹ Human plasma or myocardial proteomes have been successfully studied using the isobaric tag for relative and absolute quantification (iTRAQ) labeling technology.^12–15 This new technology enables us to unbiasedly identify and quantitatively analyze the abundance of hundreds of proteins using small amounts of clinical samples taken multiple times during surgery.^11,16 Therefore, we can make comparisons among groups, and before and after treatment, and thereby differentiate the effects of hypothermia (cooling) and rewarming using this technology.

Editorial p 2549

The study objective was to clarify the effects of both deep hypothermia (22℃) and rewarming on the myocardial proteome during cardioplegic arrest and reperfusion by comparison with normothermic counterparts. By quantitatively analyzing and comparing 3 sets of iTRAQ labeling samples (ie, before cooling, during cooling, and after rewarming), we sought to verify whether the changes in the proteome profile induced by hypothermia could be fully corrected by rewarming.

Methods

Patient Selection

This prospective cohort study received approval from the Ethics Committee of Shimane University Faculty of Medicine and was carried out at Shimane University Hospital. We included 16 patients undergoing thoracic aortic repair in this study: 8 patients underwent aortic arch repair using deep hypothermic cardiopulmonary bypass (CPB) combined with circulatory arrest and selective cerebral perfusion (DHCA+SCP); the remaining 8 patients underwent aortic root-ascending aortic repair using normothermic CPB (Table 1). All participants gave written informed consent. To evaluate myocardial injury brought about by warm or hypothermic cardioplegic arrest and CPB, we retrospectively analyzed myocardial injury markers and changes in echocardiographic variables in patients who had undergone thoracic aortic repair in the same period, including the cohort cases. These included 17 patients who had undergone ascending aortic or Valsalva sinus replacement (Bentall)±aortic valve replacement (AVR) with normothermic CPB, and 76 patients undergoing aortic arch or ascending aortic replacement±Bentall or AVR with deep hypothermic CPB (Table 1).

Table 1. Baseline Characteristics, Surgical Procedures and Intraoperative Variables

	Normothermic CPB; n=17 (proteomic analysis; n=8)	Hypothermic CPB; n=76 (proteomic analysis; n=8)	P value
Age (years)	60.7±15.2 (64.9±8.7)	72.3±12.5 (68.6±6.4)	0.001 (0.349)
Sex (male)	9 (3)	41 (3)	1.000 (1.0000)
Hypertension	4 (4)	36 (6)	0.104 (0.608)
Diabetes mellitus	1 (1)	13 (0)	0.453 (1.000)
Dyslipidemia	6 (4)	11 (2)	0.076 (0.608)
Ischemic heart diseases	2 (1)	16 (2)	0.510 (1.000)
Ascending aorta replacement	7 (5)	32 (1)	1.000 (0.119)
Total arch replacement	0	32 (1)	0.0004 (1.000)
Hemi-arch replacement	0	9 (3)	0.203 (.200)
Bentall operation	10 (3)	6 (2)	<0.0001 (1.000)
Aortic valve replacement	5 (3)	1 (0)	0.0006 (0.200)
Coronary artery bypass grafting	2 (1)	4 (0)	0.301 (1.000)
Minimal core temperature (℃)	34.4±0.9 (34.9±1.1)	22.9±2.5 (22.0±1.1)	<0.0001 (<0.0001)
Aortic cross-clamping time (min)	128.0±45.8 (111.6±32.1)	151.8±62.3 (120.0±64.1)	0.083 (0.7452)
CPB time (min)	179.5±60.7 (150.0±37.6)	244.0±79.8 (217.1±66.4)	0.002 (0.0261)

Data are number of patients or given as mean±SD in patients undergoing either normothermic CPB or hypothermic CPB. Number, mean±SD and P value in each line indicate those patients analyzed retrospectively, and number in parenthesis are the proteomic study cohort group. CPB, cardiopulmonary bypass.

Anesthesia and CPB

The anesthesia and CPB protocols used were similar for both groups of patients: anesthesia was induced and maintained with fentanyl and propofol; catecholamines, vasodilators and antiarrhythmic drugs were administered similarly in both groups of patients. The CPB circuit was the same for both groups, consisting of a hollow-fiber polymer-coated membrane oxygenator (Quadrox-I, HM070000, Maquet, Germany), centrifugal pump (Rotaflow, BO-RF32, Maquet), arterial filter (40 μm Pall arterial line filter, AL6, Pall, Port Washington, NY, USA) and heparin-coated CPB circuit (MERA or JMS, Tokyo, Japan). The same cardioplegic solution (miniplegia, non-diluted cardioplegia including glutamate and aspartate as “substrates”) was administered at the same temperature as the circulating blood in both groups, namely warm solution in the normothermic group, low temperature solution in the hypothermic CPB group.^17–19 In DHCA+SCP, the body was cooled to 22℃ with topical head cooling. After cardioplegic cardiac arrest and initiation of HCA, SCP was performed following cannulation into the 3 arch vessels and maintained by monitoring pressures and brain tissue oximetry. Rewarming was initiated after completion of open distal anastomosis and initiation of distal perfusion.

CK-MB Measurement and Echocardiographic Examination

Postoperative CK-MB levels were measured to evaluate myocardial injury. Transthoracic echocardiography was performed before and 2 weeks after surgery by the same cardiologist specialized in echocardiography. To determine diastolic function, spectral Doppler tracing of diastolic transmitral inflow velocity was obtained using pulsed Doppler imaging, and peak velocities were measured in early (E) and late diastole (A). The tissue Doppler imaging technique was used to record mitral annular velocity. Early diastolic mitral annular velocity (E’) was measured by placing the sample volume at both the septal and lateral sides of the mitral annulus; the E/E’ ratio was calculated to estimate left ventricular (LV) filling pressure and stiffness.^20,21

Tissue Sampling

The 3 right atrial wall tissue samples were excised as follows: T1 (before cooling, 5 min after pump initiation), T2 (during normothermic cardioplegic arrest, 107.5±52.1 min after pump initiation in normothermic CPB or during deep hypothermic cardioplegic arrest or ventricular fibrillation (VF), 67.9±38.6 min after pump initiation in DHCA+SCP), T3 (just before termination of CPB, 150.0±37.6 min after pump initiation in normothermic CPB or 217.1±66.4 min after pump initiation in DHCA+SCP) (Table 1).

iTRAQ Labeling and Strong Cation Exchange (SCX) Chromatography

Samples were prepared according to the manual published by AB Sciex (Foster, CA, USA) and as described previously.^13,14 In brief, equal amounts of the T1, T2, and T3 samples from each patient were denatured and reduced, the cysteines alkylated, and finally digested with trypsin (AB Sciex). Each digest was labeled with a different iTRAQ tag using the iTRAQ reagent 4-plex kit (AB Sciex). iTRAQ label 114 was used for the T1 sample, and iTRAQ labels, 116 or 117 were selected for the T2 and T3 samples; the 3 samples from each patient were then combined. The combined samples were fractionated into 6 fractions by SCX chromatography according to the manufacturer’s instructions (AB Sciex) and each fraction was desalted, also according to the manufacturer’s instructions (Waters, Milford, MA, USA).

NanoLC and MALDI-TOF/TOF MS/MS Analysis

One fraction from the SCX chromatography was fractionated to 171 spots using a DiNa nanoLC system (KYA Tech, Tokyo, Japan) and collected onto an Opti-TOF LC/MALDI 384 target plate (AB Sciex) according to the manufacturer’s instructions and as described previously.^13,14 Spotted peptide samples were analyzed by a 5800 MALDI-TOF/TOF MS/MS Analyzer with TOF/TOF Series software (version 4.0, AB Sciex). MS/MS data were analyzed using ProteinPilot^TM software (version 3.0) and the Paragon^TM protein database (AB Sciex). Quantitative changes of proteins at T2 or T3 were calculated using iTRAQ ratios T2:T1 or T3:T1, respectively.

Western Blot Analysis

Western blot analyses were performed as described previously.^13,14 In brief, myocardial samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and immunoblotted using rabbit polyclonal IDH3A antibody (Proteintech, Chicago, IL, USA), mouse monoclonal gelsolin (GSN) antibody (Sigma, St. Louis, MO, USA) and anti-rabbit IR dye 680-conjugated IgG (LI-COR, Lincoln, NE, USA). Protein bands were visualized using an Odyssey (LI-COR) infrared imaging system and their intensities measured for densitometric analyses of IDHA3 and GSN. Data from duplicate or triplicate experiments were normalized as the ratio T2:T1 or T3:T1 and analyzed statistically.

iTRAQ Data Analysis and Bioinformatic Analysis

Proteins identified as showing expression changes were examined for conformity to the following conditions: (1) false discovery rate (FDR) <5% (FDR was estimated by “decoy database searching” using ProteinPilot Software); (2) protein confidence >99% (“unused ProtScore” >2). Unused ProtScore is defined as -log (1-% confidence/100). Proteins satisfying these criteria were regarded as having statistical significance.^13,14 The KEGG was used to assess whether these significant proteins could be associated with well-defined canonical or signaling pathways. PANTHER software (version 9.0, http://www.pantherdb.org) was used to test for statistical overrepresentation of gene ontology (GO, http://www.geneontology.org) terms as described in detail elsewhere.^22,23 If the number of identified genes in a GO term was significantly larger than that in the whole genome classified by the same GO term (ie, the number of observed genes in a GO term was significantly larger than the number of expected genes in the same GO term by the binomial test), the GO term was described here as ‘overrepresented’ with statistical significance after Bonferroni correction for multiple testing. The annotations of identified proteins were obtained from the Uniprot database (http://www.uniprot.org/).

Statistical Analysis

Continuous variables are expressed as mean±standard deviation; they were tested for statistically significant differences between patient groups using Student’s t-test or Welch’s t-test. Categorical variables were compared using Fisher’s exact test. Echocardiographic variables were tested for statistically significant differences between preoperative and postoperative values using paired t-tests. For analysis of iTRAQ ratios (ie, T2:T1, T3:T1 ratios) for each significantly identified protein, P values were calculated by 1-sample t-test of averaged protein ratio against 1 to assess the validity of the protein expression change.^14,23 To assess the western blot analysis data, statistical comparisons were performed by 2way analysis of variance (general linear model) with repeated measures followed by a posthoc Bonferroni test to identify individual differences. P<0.05 was defined as statistically significant.

Results

Proteomic Profiling by PANTHER and KEGG Software

Mass spectrometry identified 697 proteins satisfying the criteria FDR <5% (Table S1), of which 222 fulfilled the criteria of protein confidence >99% (“unused ProtScore” >2) in at least 5 of 8 samples in either group. These proteins were classified according to the GO biological process using PANTHER software. The “metabolic process” was the most prominent GO category (Figure 1). Of the 222 proteins, 203 were matched to the genome annotation (k number) in the KEGG and searched using the “Search&Color Pathway” tool in the “KEGG Mapper.” Among 178 pathways detected, metabolic pathway (ko 01100) included the largest number of proteins (66 proteins) (Figure 2A); glycolysis/gluconeogenesis (ko 00010) included 14 proteins; citrate cycle (tricarboxylic acid (TCA) cycle, ko 00020) had 13 proteins (Figure 2B); oxidative phosphorylation (ko 00190) had 27 proteins (Figure 2C).

Figure 1.

The 222 proteins identified with high protein confidence classified according to GO (gene ontology) biological process by PANTHER software.

Figure 2.

Among the 222 identified proteins, 203 were also matched to the annotation number (k number) and subsequently searched using the “Search & Color Pathway” tool in the “KEGG Mapper”. (A) The 66 identified proteins were matched to a metabolic pathway (ko 01100) and highlighted by black in the pathway map, indicating that identified proteins were intensively located on glycolysis (marked by “G”), TCA cycle (marked by “T”) and oxidative phosphorylation (marked by “O”) in the map. (B) TCA cycle (“T,” ko 00020) map is illustrated and 13 matched proteins/associated enzyme commission numbers (EC) are highlighted in pink. In these, 6 proteins were downregulated with statistical significance (indicated by red arrows) during deep hypothermia (T2) in DHCA patients. These were MDH2 (EC: 1.1.1.37), IDH3 (EC: 1.1.1.41), SDHA (EC: 1.3.5.1), DLD (EC: 1.8.1.4), DLAT (EC: 2.3.1.12) and ACO (EC: 4.2.1.3). (C) Oxidative phosphorylation (“O”, ko 00190) map is illustrated and 27 matched proteins/associated EC numbers are shown in pink. In these, 6 proteins were downregulated with statistical significances (indicated by red arrows) during deep hypothermia (T2) in DHCA patients. These proteins were SDHA, CYC1, UQCRC1, ATP5C1, COX5B, and NDUFV2. DHCA, deep hypothermic cardiopulmonary bypass combined with circulatory arrest.

Proteomic Changes During DHCA and Normothermic CPB

To demonstrate both the fold changes and P values of proteins having statistical significance, fold changes (T2/T1, T3/T1) were combined with their P values and demonstrated by log-transformed iTRAQ ratio and P values, with the 2 vertical lines showing a fold change=1.2 and =0.833, respectively, and the horizontal line showing P<0.05 (Figure 3). At a glance, a scatterplot of proteins had a V-shaped distribution in the normothermic CPB patients, indicating there was an even distribution among up- and downregulated proteins (Figures 3A,B; Table S2). Immunoglobulin and acute phase protein expressions were significantly decreased midway (T2) of CPB (Figure 3A); heat shock protein 1 (HSPE1) and acute phase proteins, including α-1-acid glycoprotein 1 (ORM1), α-1-antitripsin (SERPINA1), and transthyretin (TTR), were significantly decreased on termination (T3) of CPB (Figure 3B; Table S2). Conversely, protein scatterplots were located in the left half field only, indicating that most proteins were uniformly downregulated in DHCA patients, with 50 downregulated proteins having statistical significance (Figure 3C; Table S3). These downregulated proteins are also indicated by the red arrow on the enzyme commission (EC) number or gene name in Figure 2B,C. After rewarming, the number of significantly downregulated proteins decreased and the number of upregulated proteins increased (shown in the left upper corner and right lower corner in Figure 3D). Cardiac samples were taken at deep hypothermic VF before aortic clamping in 3 patients in order to examine the effects of simple deep hypothermia; the remaining 5 patients had cardiac samples taken at deep hypothermic cardioplegic arrest. However, there were no significant differences in proteome between the 2 groups. In DHCA patients, levels of downregulated proteins mostly returned to the pre-cooling values after rewarming and declamping (T3). However, 10 proteins remained downregulated at T3 (Figure 3D; Table S3). These included myosin light chain 3 (MYL3), tropomodulin-1 (TMOD1), GSN and adaptor proteins regulating many signaling pathways.

Figure 3.

Scatterplot of proteins identified with high protein confidence demonstrating log-transformed iTRAQ ratio and P values. Fold changes >1.2 or 0.833 were defined as significant up- or downregulation, shown by the 2 vertical lines. P<0.05 was defined as statistically significant and shown by the horizontal line. Proteins located in the upper left or right fields were statistically down- or upregulated. (A) Scatterplot of identified proteins at midway (T2) in normothermic cardiopulmonary bypass (CPB) patients; 10 proteins in the left upper corner were downregulated with statistical significance. (B) Scatterplot of identified proteins at termination of CPB (T3) in normothermic CPB patients. There were 6 downregulated proteins and 1 upregulated protein with statistical significance. (C) Scatterplot of identified proteins during cooling (T2) in DHCA patients. There were 50 downregulated proteins with statistical significance, showing an enlarged view on the right. Numbers indicate proteins shown in Table S3. (D) Scatterplot of identified proteins after rewarming (T3) in DHCA patients. There were 10 downregulated proteins with statistical significances. DHCA, deep hypothermic CPB combined with circulatory arrest.

Bioinformatic Analysis by PANTHER Software

Proteins decreased or increased (<0.833fold or >1.2fold) with statistical significance (P<0.05) found in at least 5 of 8 samples, which were located in the upper left or right field in Figures 3A–D, were identified to obtain both an adequate sample size and statistical power,^14,24 and were subsequently analyzed by PANTHER software. The statistical overrepresentation test identified no protein group (GO) overrepresented statistically in normothermic CPB patients; however, there were 8 statistically overrepresented GO categories (biological processes) mainly related to energy production in DHCA patients, in which 19 significantly downregulated proteins were directly related to energy production. These downregulated proteins mostly returned to the pre-cooling levels after rewarming and declamping (T3); thus there were only 2 GO groups relating to cardiac structure morphogenesis statistically overrepresented at T3 in DHCA patients (Table 2).

Table 2. GO Categories (Biological Process) Showing Statistical Overrepresentation for Myocardial Proteins Taken Under Conditions of Both Hypothermic Cardioplegic Arrest and Rewarmed Beating

GO term	Description	Homo sapiens (reference)			Myocardial tissue sample
		No. of genes (n=21,804)*	No. of expected genes**	No. of observed genes (no. of mapped genes=57)†	P value	Genes
At hypothermic arrest
0006091	Generation of precursor metabolite and energy	280	0.72	12	1.16E-09	CYB5R3 UQCRC1 SDHA AC02 DLD COX5B IDH3A NDUFV2 MDH2 NNT CYC1 ETFA
0022904	Respiratory electron transport chain	232	0.6	8	2.59E-05	CYB5R3 UQCRC1 SDHA DLD COX5B NDUFV2 CYC1 ETFA
0006631	Fatty acid metabolic process	188	0.48	6	1.61E-03	ANXA1 CYB5R3 ANXA5 ETFA ECHS1 DECR1
0006119	Oxidative Phosphorylation	57	0.15	4	2.69E-03	SDHA COX5B NDUFV2 NNT
0006099	Tricarboxylic acid cycles	23	0.06	3	5.46-E03	AC02 IDH3A MDH2
0008152	Metabolic process	8,613	22.12	37	9.20E-03	MYOM3 CYB5R3 UQCRC1 APOA1 RPS27A HK1 PHB PRDX2 ACO2 SDHA CKB HSPA8 DECR1 HSPD1 MYOM1 ARL6IP5 IDH3A DLD ATP5C1 SERPINA1 S100A1 ANXA1 MDH2 HSPA1A EFTA DLAT HSP90B1 CKM ALDH2 COX5B MYH9 SORBS2 SOD1 NNT NDUFV2 TUFM ANXA5
0006635	Fatty acid β-oxidation	35	0.09	3	1.88E-02	EFTA ECHS1 DECR1
0005975	Carbohydrate metabolic process	650	1.67	8	4.32E-02	MDH2 SORBS2 HK1 DLAT ECHS1 SDHA ACO2 IDH3A
At rewarmed beating
0032989	Cellular component morphogenesis	646	0.3	4	2.45E-02	MYH9 SORBS2 TMOD1 GSN
0009653	Anatomical structure morphogenesis	691	0.32	4	3.18E-02	MYH9 SORBS2 TMOD1 GSN

Of 222 proteins fulfilling the criteria, proteins that changed (<0.83-fold or >1.2-fold) significantly were analyzed by PANTHER software (version 9.0). The test for statistical overrepresentation (see Methods for more detail) of each GO (see Methods for more detail) term (biological process) identified no proteins group (GO) in normothermic patients, but 8 GO categories in hypothermic patients. *Total number of genes in the whole genome (Homo sapiens, n=21,804) classified by the GO term; **number of genes that would be detected in the input list for a particular GO category on the basis of the reference list (Homo sapiens); ^†number of genes from the input list (n=57) classified by the GO term. GO, gene ontology.

Western Blot Analysis

Among the protein levels that decreased after cooling in DHCA patients, IDH3A, a protein that recovered to its pre-cooling level (protein no. 46 in Table S3), and GSN, a protein that did not recover its level (no. 26 protein in Table S3), were both validated by western blotting. The IDH3A level decreased after cooling with statistical significance (P=0.013, between T1 and T2) and increased after rewarming to some extent in DHCA patients, although there were no significant changes in normothermic CPB (Figure 4A). On the other hand, GSN levels decreased during both cooling (T2) and rewarming (T3) similarly to the iTRAQ ratio, but without reaching statistical significance, because of large variations in the DHCA patients (Figure 4B); however, these levels decreased at T2 and increased to the initial level at T3 in normothermic CPB patients.

Figure 4.

Representative photographs and relative intensities (mean±standard deviation) of western blot analyses of isocitrate dehydrogenase [NAD] subunit α, mitochondrial (IDHA3) (A) and gelsolin (GSN) (B). Sampling times were as follows: T1: 5 min after cardiopulmonary bypass (CPB), T2: midway during CPB (hypothermic cardioplegic arrest or VF), T3: before termination of CPB. *P=0.031 between T1 and T2 in the DHCA group. DHCA, deep hypothermic CPB combined with circulatory arrest.

Myocardial Injury and Echocardiographic Findings

Additional retrospective analysis revealed that there were 2 hospital deaths, from septic shock and rupture of abdominal aortic aneurysm after thoracic aortic repair, in the deep hypothermic CPB group (n=76), but there were no hospital deaths in the normothermic group (n=17). There were 3 major complications, comprising 2 strokes and 1 graft infection, in the deep hypothermic group, and there was 1 case of mediastinitis in the normothermic group. There was no incidence of severe low output syndrome necessitating IABP or any assisted circulation in either group. The postoperative peak CK-MB value was lower in the deep hypothermic group than in the normothermic group (34.2±22.2 ng/ml, n=70 vs. 44.7±33.3 ng/ml, n=16, P=0.246), although aortic cross-clamping time was longer in the deep hypothermic group (Table 1, P=0.083). Echocardiographic evaluation did not demonstrate significant changes in ejection fraction pre- and post- surgery in either group, although left ventricular end-diastolic volume index (LVEDVI) significantly decreased after surgery in both groups (Table 3). E/A similarly increased post surgery in both groups with statistical significance. E/E’ slightly decreased postoperatively in the deep hypothermic group (P=0.239, Figure 5A), but this ratio significantly increased postoperatively in the normothermic group (P=0.017, Figure 5B).

Figure 5.

Representative Doppler tracings pre and post surgery. E/E’ slightly decreased from 12.1 to 8.4 in a patient who underwent hypothermic surgery (A), but increased from 11.9 to 18.2 in a patient who underwent normothermic surgery (B).

Table 3. Echocardiographic Variables Before and After Surgery

	Normothermic (n=17†)		Hypothermic (n=29†)
	Before surgery	After surgery	Before surgery	After surgery
LVMI (g/cm)	149.3±43.9	133.7±47.1	119.7±36.2	106.5±33.2**
LVEDVI (ml/m2)	79.1±36.0	58.0±15.6**	55.7±23.2	42.7±11.9**
LVMI/LVEDVI (g/ml)	2.2±0.7	2.6±0.7	2.4±0.6	2.6±0.8
LVEF (%)	60.5±8.0	56.8±13.4	63.1±9.0	60.4±8.1
E (cm/s)	53.6±14.1	67.3±14.6**	60.6±18.4	66.7±21.4
A (cm/s)	70.1±21.3	70.7±21.7	82.0±21.5	76.0±22.8*
E/A	0.80±0.34	1.00±0.29*	0.76±0.22	0.97±0.49*
DcT (ms)	235.2±75.7	191.1±39.3	238.4±43.1	201.8±45.7***
E’ (cm/s)	5.2±2.6	4.9±2.0	4.7±1.5	5.9±2.2**
E/E’	12.4±6.2	15.6±7.1*	13.5±6.2	12.0±5.4

Data are given as mean±SD. ^†Patients evaluated by echocardiography pre- and postoperatively. Paired t-tests were carried out between “before surgery” and “after surgery” in each group, *P<0.05; **P<0.01; ***P<0.001. DcT, deceleration time; LVEDVI, left ventricular end-diastolic volume index; LVEDVI, left ventricular end-diastolic volume index; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index.

Discussion

Downregulation of Metabolic Proteins by Deep Hypothermia

Bioinformatic analyses demonstrated that deep hypothermic cardioplegic arrest/VF produced downregulation of metabolic proteins (Table 2). However, no such downregulation of proteins was observed in normothermic CPB counterparts that were similar to the DHCA group regarding cardioplegic solution, ischemic time, but not the temperature and T3 sampling time. These are unexpectedly new findings that have not been previously reported, although several studies have demonstrated reduced oxygen consumption or relatively constant ATP levels in cardiac tissue during hypothermic cardioplegic arrest compared with normothermic cardioplegic arrest.^3,4,25 Chitwood et al have demonstrated in the potassium-arrested heart model that asystole at 37℃ decreased MV̇O₂ from 5.18±0.55 to 1.85±0.20 ml O₂/min/100 g of left ventricle (35% of normal value); asystole at 22℃ further decreased MV̇O₂ to 0.54±0.05 ml O₂/min/100 g of left ventricle (10% of normal value).⁴ In the present proteomic analysis, levels of proteins belonging to the TCA cycle and ETC did not decrease in the normothermic arrested heart, implying that enzymatic protein levels might be too high relative to the decreased oxygen consumption (35% of normal); however, enzymatic activity would decrease accordingly. In the deep hypothermic (22℃) arrested heart, the levels of proteins were downregulated but remained much higher than the levels of myocardial oxygen consumption (10% of normal). Aragones et al recently demonstrated that restricting movement of glycolytic intermediates into the TCA cycle could induce hypoxia tolerance by inactivating the pyruvate dehydrogenase complex (PDC).²⁶ In our study, we revealed that expression of the dihydrolipoyllysine-residue acetyltransferase component of the PDC (DLAT) was decreased at hypothermic arrest and returned to initial values at rewarmed beating (carbohydrate metabolic process in Table 2; no. 19 protein in Table S3), together with decreased expression of many other proteins in the TCA cycle and ETC. It can be speculated that hypothermia could induce ischemia tolerance by globally reducing the expression of proteins related to oxygen consumption, which could constitute a distinct approach from that targeting a specific protein for inhibition.

Our recent animal proteomic study demonstrated that two-thirds of myocardial proteins in the left ventricle were downregulated after surface cooling (23℃, unpublished data). Among these proteins, ATP5C1 was downregulated, and also decreased in the right atrial tissue from DHCA patients. Analyses of protein abundance in the left ventricle are clinically valuable; however, it is quite difficult to take LV tissue samples safely during surgery. Therefore we analyzed right atrial tissue as a surrogate myocardial sample, as has been previously done in other studies. However, recent proteomic and transcriptomic analyses showed that >90% of proteins belong to both ventricles and atria, while only 6.7% of proteins were atria-specific and 9.1% were ventricle-specific.^27,28 In the ventricles, protein expression preferentially relates to muscular contraction and energy production. In contrast, fibrosis and apoptotic pathways are concentrated in atrial myocardium, indicating the higher susceptibility of atrial myocardium to apoptosis.^27,28 Although the present study analyzed changes in protein expression during a short CPB time, key regulatory enzymes may have t_1/2 values as low as 0.5 to 2 h.²⁹

Rewarming and the Myocardial Proteome

After reperfusion and rewarming, the myocardial proteome recovered to pre-cooling levels, especially in energy production (Figure 3D; Table S3). However, there were 10 proteins that had reduced expressions. Among these, MYL3 and TMOD1 have important cardiac regulatory functions. Myosin light chain can decrease myosin stiffness, in vitro motility and modulate cardiac function.³⁰ TMOD are dynamic caps that inhibit actin monomer association and dissociation from actin filament pointed ends and thereby precisely regulate actin filament length. Inhibition of TMOD-mediated capping of the pointed end of actin filaments results in inappropriate thin filament elongation.³¹ Downregulation of both MYL3 and TMOD could lead to dysregulation in thin filament length and myosin stiffness, although further research is required to evaluate such a hypothesis. There have been studies demonstrating acute deterioration in systolic function just after rewarming, because of decreased calcium sensitivity.³² Although GSN expression decreased after rewarming and reperfusion in the present DHCA patients, a recent study demonstrated that GSN may promote apoptosis through the GSN/HIF-1α/DNAase 1 pathway and downregulation of survival factors.^33,34 Basic studies have demonstrated that rewarming could exacerbate preexisting ischemia-reperfusion injury by several mechanisms, such as activation of apoptotic processes, complement activation, or chelatable iron oxygen injury.^14,35–37 These findings have been clinically translated into the proposal that rewarming should be slow (0.1–0.3℃/h) in therapeutic hypothermia.³⁸

Adaptor Proteins and IR Injury

Expression of 2 major adaptor proteins remain decreased after rewarming: sorbin and SH3 domain-containing protein 2 (SORBS2) did not change during deep hypothermia, but their expressions decreased after rewarming (protein no. 2 in Table S3); 14-3-3 protein zeta/delta (YWHAZ) decreased during deep hypothermia and its expression did not recover after rewarming (protein no. 40 in Table S3). To our knowledge, this is the first demonstration that adaptor proteins are downregulated by hypothermia and rewarming. SORBS 2 can form a complex with tyrosine-protein kinase ABL1, which activates apoptotic signaling resulting from DNA injury, is abundantly expressed in the heart, and is located in the Z-disk of the cardiac sarcomere.³⁹ When severe IR injury occurs, SORBS2 proteins are released from the heart into the bloodstream after reperfusion.⁴⁰ However, we could not detect SORBS2 protein in plasma taken after reperfusion in our previous proteomic analysis.¹⁴ The 7 isoforms of 14-3-3 proteins have multiple roles in the apoptotic signaling network and work for signaling integration.⁴¹ 14-3-3 proteins can contribute to cell survival by sequestering pro-apoptotic proteins, such as BAD, BAD and c-ABl.⁴¹ Apoptotic activation is facilitated by JNK-mediated phosphorylation of 14-3-3, triggering release of these apoptotic effectors.⁴¹ It is unclear how hypothermia-induced downregulation of 14-3-3 protein and SORBS2 could be implicated in regulating the abovementioned apoptotic process. A further study including protein phosphorylation is needed.

Temperature of Cardioplegia/CPB and Postoperative Cardiac Function

In terms of postoperative peak CK-MB measurement and echocardiographic systolic function, there was no significant difference in cardioprotective effect between warm and hypothermic cardioplegia/CPB. However, in the normothermic patients, peak CK-MB levels slightly increased over those in the hypothermic patients. Furthermore, E/E’ (most relevant predictor of LV filling pressure) significantly increased postoperatively, suggesting diastolic dysfunction caused by inadequate protection during normothermic surgery and subsequent fibrosis. Pathological examination of LV samples demonstrated such a chronic fibrosis with normal cardiac function 50 days after transplantation in a prolonged (4–8 h) myocardial preservation model.⁴² Although there were several differences, including age and surgical procedure, between the present groups except for temperature (Table 1), differences in surgical procedures could not have significantly impacted E/E’ because there was no change in E/E’ value pre and post AVR.²¹ Although this ratio is a clinically useful tool, its accuracy has some limitations.⁴³ Whether hypothermic (cold) cardioplegia is superior to its warm counterpart is still undetermined.^19,44 Because a uniform study design has not been followed in these studies, many different conditions other than temperature have been used, such as antegrade/retrograde or intermittent/continuous perfusion, crystalloid or blood cardioplegia.

Study Limitations

Because the amounts of the clinical samples of cardiac tissue were very limited, it was not possible to fully validate large proteomic data sets. In fact, western blotting was limited to examination of 2 important proteins. iTRAQ labeling and repeat samplings could be effective in serial measurements of proteomic changes during the course of surgery. However, many factors relating to surgical intervention, such as deep hypothermic circulatory arrest of the lower body, may have affected the cardiac proteome in this study. Clinical data were retrospectively analyzed; furthermore, there were several differences between groups. Therefore, the effect of hypothermia/rewarming on the cardiac proteome and function would best be validated by animal studies.

Conclusions

Deep hypothermia produced downregulation of cardiac proteins related to energy production and rewarming restored these proteins to pre-cooling levels, except for adaptor proteins and regulators of cardiac contraction. To our knowledge, these are novel findings that may contribute to a global understanding of the effects of hypothermia and rewarming.

Acknowledgments

We thank Dr Hiroyuki Yoshitomi for his professional assistance with echocardiography and for providing echocardiographic images.

Grants

This study was supported by a Grantin Aid for Scientific Research (B) from the Japan Society for the Promotion of Science (nos. 21390389 and 23659845).

Supplementary Files

Supplementary File 1

Table S1. Raw proteomic data of right atrial tissue samples taken at either the midpoint or just before termination of normothermic or hypothermic CPB after FDR analysis

Table S2. Significantly increased or decreased myocardial proteins at the midpoint (normothermic arrest) or termination (beating) of normothermic CPB

Table S3. Significantly increased or decreased myocardial proteins at the midpoint of CPB (hypothermic arrest) or termination of CPB (rewarmed beating) in patients undergoing deep hypothermic CPB

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-15-0541

References

• 1.    Polderman  KH. Mechanisms of action, physiological effects, and complications of hypothermia. Crit Care Med 2009; 37: S186–S202.
• 2.    Tissier  R,  Chenoune  M,  Ghaleh  B,  Cohen  MV,  Downey  JM,  Berdeaux  A. The small chill: Mild hypothermia for cardioprotection? Cardiovasc Res 2010; 88: 406–414.
• 3.    Buckberg  GD,  Brazier  JR,  Nelson  RL,  Goldstein  SM,  McConnell  DH,  Cooper  N. Studies of the effects of hypothermia on regional myocardial blood flow and metabolism during cardiopulmonary bypass. I: The adequately perfused beating, fibrillating, and arrested heart. J Thorac Cardiovasc Surg 1977; 73: 87–94.
• 4.    Chitwood  WR Jr,  Sink  JD,  Hill  RC,  Wechsler  A,  Sabiston  DC Jr. The effects of hypothermia on myocardial oxygen consumption and transmural coronary blood flow in the potassium-arrested heart. Ann Surg 1979; 190: 106–116.
• 5.    Russ  C,  Lee  JC. Effect of hypothermia on myocardial metabolism. Am J Physiol 1965; 208: 1253–1258.
• 6.    Gilbert  NF,  Meyer  PE,  Tauriainen  MP,  Chao  RY,  Patel  JB,  Malloy  CR, et al. Effects of hypothermia on myocardial substrate selection. Ann Thorac Surg 2002; 74: 1208–1212.
• 7.    Ning  XH,  Chen  SH,  Xu  CS,  Li  L,  Yao  LY,  Qian  K, et al. Hypothermic protection of the ischemic heart via alterations in apoptotic pathways as assessed by gene array analysis. J Appl Physiol 2002; 92: 2207–2220.
• 8.    Babu  PP,  Suzuki  G,  Ono  Y,  Yoshida  Y. Attenuation of ischemic and/or reperfusion injury during myocardial infarction using mild hypothermia in rats: An immunohistochemical study of Bcl-2, Bax, Bak and TUNEL. Pathol Int 2004; 54: 896–903.
• 9.    Castedo  E,  Castejon  R,  Monguio  E,  Ramis  S,  Montero  CG,  Serrano-Fiz  S, et al. Influence of hypothermia on right atrial cardiomyocyte apoptosis in patients undergoing aortic valve replacement. J Cardiothoracic Surg 2007; 2: 7.
• 10.    Shuja  F,  Tabbara  M,  Li  Y,  Lin  B,  Butt  MU,  Velmahos  GC, et al. Profound hypothermia decreases cardiac apoptosis through Akt survival pathway. J Am Coll Surg 2009; 209: 89–99.
• 11.    Wu  L,  Candille  SI,  Choi  Y,  Xie  D,  Jiang  L,  Li-Pook-Than  J, et al. Variation and genetic control of protein abundance in humans. Nature 2013; 499: 79–82.
• 12.    Pottiez  G,  Wiederin  J,  Fox  HS,  Ciborowski  P. Comparison of 4plex to 8plex iTRAQ quantitative measurements of proteins in human plasma samples. J Proteome Res 2012; 11: 3774–3781.
• 13.    Satoh  K,  Maniwa  T,  Oda  T,  Matsumoto  K. Proteomic profiling for the identification of serum diagnostic biomarkers for abdominal and thoracic aortic aneurysms. Proteome Sci 2013; 11: 27.
• 14.    Oda  T,  Yamaguchi  A,  Yokoyama  M,  Shimizu  K,  Toyota  K,  Nikai  T, et al. Plasma proteomic changes during hypothermic and normothermic cardiopulmonary bypass in aortic surgeries. Int J Mol Med 2014; 34: 947–956.
• 15.    Tu  T,  Zhou  S,  Liu  Z,  Li  X,  Liu  Q. Quantitative proteomics of changes in energy metabolism-related proteins in atrial tissue from valvular disease patients with permanent atrial fibrillation. Circ J 2014; 78: 993–1001.
• 16.    Oda  T,  Matsumoto  K. Proteomic analysis in cardiovascular research. Surg Today 2015 April 19, doi:10.1007/s00595-015-1169-4.
• 17.    Menasche  P. Blood cardioplegia: Do we still need to dilute? Ann Thorac Surg 1996; 62: 957–960.
• 18.    Petrucci  O,  Vieira  RW,  do Carmo  MR,  de Oliveira  PPM,  Antunes  N,  Braile  DM. Use of (all-blood) miniplegia versus crystalloid cardioplegia in an experimental model of acute myocardial ischemia. J Card Surg 2008; 23: 361–365.
• 19.    Yamamoto  H,  Yamamoto  F. Myocardial protection in cardiac surgery: A historical review from the beginning to the current topics. Gen Thorac Cardiovasc Surg 2013; 61: 485–496.
• 20.    Nagueh  SF,  Appleton  CP,  Gillebert  TC,  Marino  PN,  Oh  JK,  Smiseth  OA, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr 2009; 22: 107–133.
• 21.    Seo  JS,  Jang  MK,  Lee  EY,  Yun  SC,  Kim  DH,  Song  JM, et al. Evaluation of left ventricular diastolic function after valve replacement in aortic stenosis using exercise Doppler echocardiography. Circ J 2012; 76: 2792–2798.
• 22.    Mi  H,  Muruganujan  A,  Casagrande  JT,  Thomas  PD. Largescale gene function analysis with the PANTHER classification system. Nat Protoc 2013; 8: 1551–1566.
• 23.    Ashburner  M,  Ball  CA,  Blake  JA,  Blake  JA,  Botstein  D,  Butler  H, et al. Gene ontology: Tool for the unification of biology. Nat Genet 2000; 25: 25–29.
• 24.    Lancaster  TS,  Jefferson  SJ,  Hunter  JC,  Lopez  V,  Van Eyk  JE,  Lakatta  EG, et al. Quantitative proteomic analysis reveals novel mitochondrial targets of estrogen deficiency in the aged female rat heart. Physiol Genomics 2012; 44: 957–969.
• 25.    Hayashida  N,  Ikonomidis  JS,  Weisel  RD,  Shirai  T,  Ivanov  J,  Carson  SM, et al. The optimal cardioplegic temperature. Ann Thorac Surg 1994; 58: 961–971.
• 26.    Aragones  J,  Schneider  M,  Geyte  KV,  Fraisl  P,  Dresselaers  T,  Mazzone  M, et al. Deficiency or inhibition of oxygen sensor Phd1 induces hypoxia tolerance by programming basal metabolism. Nat Genet 2008; 40: 170–180.
• 27.    Barth  AS,  Merk  S,  Arnoldi  E,  Zwermann  L,  Kloos  P,  Gebauer  M, et al. Functional profiling of human atrial and ventricular gene expression. Pflugers Arch Eur J Physiol 2005; 450: 201–208.
• 28.    Lu  ZQ,  Sinha  A,  Sharma  P,  Kislinger  T,  Gramolini  AO. Proteomic analysis of human fetal atria and ventricle. J Proteome Res 2014; 13: 5869–5878.
• 29.    Rodwell  VW. Catabolism of proteins and amino acid nitrogen. In: Weitz M, Brown RY, editors. Harper’s illustrated biochemistry, 30^th edn. New York: McGraw-Hill Education, 2015; 287–296.
• 30.    Lossie  J,  Kohncke  C,  Mahmoodzadeh  S,  Steffen  W,  Canepari  M,  Maffei  M, et al. Molecular mechanism regulating myosin and cardiac function by ELC. Biochem Biophys Res Commun 2014; 450: 464–469.
• 31.    Gokhin  DS,  Fowler  VM. Tropomodulin capping of actin filaments in striated muscle development and physiology. J Biomed Biotechnol 2011; 2011: 103069.
• 32.    Han  YS,  Tveita  T,  Prakash  YS,  Sieck  GC. Mechanisms underlying hypothermia-induced cardiac contractile dysfunction. Am J Physiol Heart Circ Physiol 2010; 298: H890–H897.
• 33.    Li  GH,  Shi  Y,  Chen  Y,  Sun  M,  Sader  S,  Maekawa  Y, et al. Gelsolin regulates cardiac remodeling after myocardial infraction through DNase I-mediated apoptosis. Circ Res 2009; 104: 896–904.
• 34.    Nishio  R,  Matsumori  A. Gelsolin and cardiac myocyte apoptosis: A new target in the treatment of postinfarction remodeling. Circ Res 2009; 104: 829–831.
• 35.    Wang  B,  Armstrong  JS,  Lee  JH,  Bhalala  U,  Kulikowicz  E,  Zhang  H, et al. Rewarming from therapeutic hypothermia induces cortical neuron apoptosis in a swine model of neonatal hypoxic-ischemic encephalopathy. J Cereb Blood Flow Metab 2015; 35: 781–793.
• 36.    Bisschops  LL,  Hoedemaekers  CW,  Mollnes  TE,  van der Hoeven  JG. Rewarming after hypothermia after cardiac arrest shifts the inflammatory balance. Crit Care Med 2012; 40: 1136–1142.
• 37.    Rauen  U,  Petrat  F,  Li  T,  de Groot  H. Hypothermia injury/cold-induced apoptosis: Evidence of an increase in chelatable iron causing oxidative injury in spite of low O₂^–/H₂O₂ formation. FASEB J 2000; 14: 1953–1964.
• 38.    Rittenberger  JC,  Polderman  KH,  Smith  WS,  Weingart  SD. Emergency neurological life support: Resuscitation following cardiac arrest. Neurocrit Care 2012; 17: S21–S28.
• 39.    Wang  B,  Golemis  EA,  Kruh  GD. ArgBP2, a multiple Src homology 3 domain-containing, Arg/Abl-interacting protein, is phosphorylated in v-Abl-transformed cells and localized in stress fiber and cardiocyte z-disks. J Biol Chem 1997; 272: 17542–17550.
• 40.    Kakimoto  Y,  Ito  S,  Abiru  H,  Kotani  H,  Ozeki  M,  Tamaki  K, et al. Sorbin and SH3 domain-containing protein 2 is released from infarcted heart in the very early phase: Proteomic analysis of cardiac tissues from patients. J Am Heart Assoc 2013; 2: e000565, doi:10.1161/JAHA.113.
• 41.    Gardino  AK,  Yaffe  MB. 14-3-3 proteins as signaling integration points for cell cycle control and apoptosis. Sem Cell Dev Biol 2011; 22: 688–695.
• 42.    Lurie  KG,  Billinham  ME,  Masek  MA,  Ginsburg  R,  Bristow  MR,  Harrison  D, et al. Ultrastructural and functional studies on prolonged myocardial preservation in an experimental heart transplant model. J Thorac Cardiovasc Surg 1982; 84: 122–129.
• 43.    Chiang  SJ,  Daimon  M,  Ishii  K,  Kawata  T,  Miyazaki  S,  Hirose  K, et al. Assessment of elevation of and rapid change in left ventricular filling pressure using a novel global strain imaging diastolic index. Circ J 2014; 78: 419–427.
• 44.    Abah  U,  Roberts  PG,  Ishaq  M,  De Silva  R. Is cold or warm blood cardioplegia superior for myocardial protection? Interact CardioVasc Thorac Surg 2012; 14: 848–855.