Heme oxygenase 1 (HO-1) catalyzes the degradation of heme to biliverdin, carbon monoxide (CO), and free iron (1). Biliverdin is subsequently reduced to bilirubin by biliverdin reductase, and iron is sequestered by ferritin (1). The HO-1 is a cytoprotective enzyme with an important role in cellular defense by virtue of the antioxidative, anti-inflammatory, antiapoptotic, and antiproliferative effects mediated by HO reaction products (2). The HO-1 is upregulated by various critical illness-related stress stimuli, such as oxidative stress, cytokines, endotoxin, hypoxia, and heme (2). Indeed, increased expression of HO-1 protein has been found in the lungs of patients with acute respiratory distress syndrome (3), in monocytes of patients with systemic inflammatory response syndrome (4), and in the bone marrow macrophages of patients dying from severe sepsis or septic shock (5). In addition, HO-1 mRNA expression is increased in the blood cells of premature infants with respiratory distress syndrome and pediatric cancer patients with systemic inflammatory response syndrome (6, 7). Upregulated HO-1 mRNA and protein levels have also been found in the livers of acute liver failure patients (8).
Several studies have shown that in humans, the magnitude of HO-1 stress response is modulated by HO-1 gene polymorphisms. Short microsatellite GTn repeat length allele and −413A allele have been shown to enhance transcriptional activity of HO-1 gene compared with long GTn and −413T alleles (9-11). The HO-1 polymorphisms have been associated with various clinical conditions, such as outcome of organ transplantation and susceptibility to coronary artery disease, restenosis after peripheral angioplasty and coronary stenting, or emphysema (10, 12-17).
Because HO-1 is a ubiquitously expressed enzyme present in tissues and organs of critically ill patients and highly induced in response to various stresses, the magnitude of HO-1 increase could be associated with the disease severity of critically ill patients and thereby be related with their outcome. Despite the potential clinical applicability of measuring HO-1 levels among the critically ill, the studies investigating HO-1 plasma concentrations and HO-1 polymorphisms in intensive care unit (ICU) patients are few. We therefore evaluated the association of HO-1 polymorphisms and plasma concentrations with the outcome and disease severity in critically ill ICU patients. As HO-1 polymorphisms were recently shown toassociate with the development of adult respiratory distress syndrome (18), we also investigated whether certain HO-1 polymorphisms would be protective against multiple-organ dysfunction (MOD), and whether HO-1 plasma concentrations would differ between patients with and without MOD.
MATERIALS AND METHODS
Patients and blood sampling
The study was approved by the local ethics committee. A written informed consent was required before study inclusion and drawing blood samples. Between January 2004 and July 2005, 244 patients were enrolled from a nine-bed medical ICU and two 8-bed medical-surgical ICUs in the tertiary care Helsinki University Central Hospital. The patient information was collected and stored through the Finnish intensive care quality consortium's database (Intensium, Kuopio, Finland) including demographic data, diagnosis by International Statistical Classification of Diseases, 10th Revision (ICD-10), Simplified Acute Physiology Score II (SAPS II) (19), Acute Physiology and Chronic Health Evaluation II (APACHE II) score (20), Sequential Organ Failure Assessment (SOFA) score on days 1 to 4 in the ICU (21), and hospital mortality.
Blood samples were collected into 10-mL EDTA tubes on the first, second, and third to fourth days in the ICU. Plasma and cells were separated immediately by centrifugation at 1,600g (+4°C) for 10 min, and plasma samples were stored at −80°C.
HO-1 plasma concentration
The HO-1 plasma concentration was measured using the human HO-1 enzyme-linked immunosorbent assay kit (Stressgen, Victoria, Canada) according to manufacturer's instructions. Plasma samples of 58 healthy volunteers (28 women and 30 men) were used for determining the reference range of plasma HO-1. This population was a part of a larger cohort designed to provide reference values for routine laboratory parameters (the Nordic Reference Interval Project) (22).
Genomic DNA was isolated from whole blood using Puregene DNA Isolation Kit (Gentra Systems, Minneapolis, Minn). Two single nucleotide polymorphisms (SNPs), −413A/T (rs2071746) and +99G/C (rs2071747), were determined by allelic discrimination using the 5′ nuclease TaqMan assay on an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, Calif). The following primers and probes were used for −413A/T: forward 5′-TGACATTTTAGGGAGCTGGAGACA-3′, reverse 5′-AGGCGTCCCAGAAGGTTCCA-3′, probe for A allele 5′-FAM-CCCACCAGGCTATTGCTCTGAGCA-Tamra-3′, and probe for T allele 5′-VIC-CCCACCAGGCTTTTGCTCTGAGC-Tamra-3′ (10). TaqMan SNP genotyping assay was used for +99G/C (Applied Biosystems).
Genotyping of the GTn repeat length polymorphism in the HO-1 promoter was performed by polymerase chain reaction using a 5′FAM-labeled forward primer 5′-FAM-AGAGCCTGCAGCTTCTCAGA-3′ and reverse primer 5′-ACAAAGTCTGGCCATAGGAC-3′. The length of polymerase chain reaction products was determined by fragment analysis using a capillary sequencer (ABI3730xl DNA Analyzer; Applied Biosystems). GeneScan-500 LIZ (Applied Biosystems) was used as a size marker. The number of GTn repeats was determined using the GeneMaker 1.4 software (SoftGenetics).
Data are presented as median and interquartile range ([IQR] 25th-75th percentiles) or as absolute values and percentages. Nonparametric Mann-Whitney U and Kruskal-Wallis tests were used for comparisons of differences in continuous variables. Chi-square and Fisher exact test were used for categorical variables. Nonparametric Spearman correlation was used in determining bivariate correlations. In linear regression analysis, we used log-transformed plasma HO-1 concentration as dependent variable to screen which variables had independent effect on first-day HO-1 concentrations. To determine the discriminative power for survival by the HO-1 plasma concentrations, receiver operating characteristic (ROC) curves were constructed, and the area under the curve was calculated with 95% confidence interval (CI). In all tests, P < 0.05 was considered significant. Statistical procedures were performed using the SPSS 12.0 statistical software (SPSS, Chicago, Ill).
Hardy-Weinberg equilibrium was calculated with Pearson correlation and Fisher exact tests. Pairwise linkage disequilibrium (LD) measures were determined using the Haploview 4.0 software. Haplotypes were constructed using the solid spine of LD algorithm and the Haploview software (23). In addition to the marker-by-marker analysis, we evaluated the associations also by haplotype-based analysis (24, 25). The Genetic Power Calculator was used for power calculations (26). Assuming a relative risk of 2.5 for the heterozygotes and 3 for minor allele homozygotes regarding hospital mortality, our sample had 83% power to detect association with the +99G/C polymorphism.
Originally, 244 patients were included in the study, but 13 patients were excluded because adequate blood sample was not obtained within 96 h, thus the final patient population was 231 patients. The SNP information was available for all 231 study patients and the GTn repeat length result from 230 patients. The analysis of GTn repeat length failed from one patient. The SNP polymorphisms were in Hardy-Weinberg equilibrium. For the HO-1 plasma concentration analysis, the first-day sample was available in 160 study patients, the second-day sample in 162 patients, and the third- to fourth-day sample in 187 patients. The hospital mortality rate was 19.5% (45/231). Baseline characteristics of the study patients are presented in Table 1.
Frequencies of HO-1 polymorphisms
The +99GC genotype was found in 24 (10.4%) of 231 study patients and the +99GG genotype in 207 patients (89.6%). The frequencies of the −413A/T genotypes were AA 35.5% (82/231), AT 48.5% (112/231), and TT 16.0% (37/231). The GTn repeat length range was 18 to 40. The distribution is presented in Figure 1. The most frequent repeat lengths were 30 (47.6%) and 23 (18.6%). The GTn repeat length polymorphism was divided into short (S) (≤26 GT repeats), medium (M) (27-33 GT repeats), and long (L) alleles (>33 GT repeats) by its trimodal distribution. The S allele frequency was 36%, the M allele frequency was 59%, and the L allele frequency was 5%. The allele frequencies did not differ between men and women.
There was strong LD between all the studied polymorphisms. The D′ measure of LD between the two SNPs was 1.0, whereas the r2 value, which takes allele frequencies into account, was 0.08. The most common allele of the repeat length polymorphism, GT(30), was in complete LD (D′ = 1) with the −413A and +99G alleles. The next frequent alleles, GT(23) and GT(24), were in complete LD with the −413T and +99G alleles. Interestingly, the group of L alleles was in perfect LD (both D′ and r2 = 1) with the +99C allele. The different haplotypes formed by the three studied polymorphisms and their relative frequencies are presented in Table 2.
HO-1 plasma concentrations
The reference range for plasma HO-1 was 0.66 to 2.39 ng/mL (95% CI), and the concentrations did not differ between men and women. The median first-day plasma HO-1 concentrations of the study patients were significantly higher than the reference values (6.9 ng/mL; IQR, 5.1-9.9 ng/mL vs. 1.5 ng/mL; IQR, 1.13-1.77 ng/mL; P < 0.001) (Fig. 2). A significant difference persisted at later time points (P < 0.001) (Fig. 2). One patient had autoimmune hemolytic anemia and extremely high HO-1 concentration (407 ng/mL) and was therefore excluded from further analysis.
The first-day HO-1 concentrations in the ICD-10 diagnosis groups are shown in Table 3. The first-day HO-1 levels were increased in all critically ill patients regardless of the diagnosis. Patients with pancreatitis had the highest first-day HO-1 plasma levels (median, 9.0 ng/mL) and patients with sepsis or respiratory tract infection had a median first-day HO-1 plasma concentration of 8.4 ng/mL. Patients with heart arrest or acute myocardial infarction had lower first-day median level of plasma HO-1 of 6.3 and 6.4 ng/mL, respectively.
Association of HO-1 polymorphisms with its plasma concentrations and outcome
The HO-1 polymorphisms had a significant effect on HO-1 plasma concentrations. Patients with the +99C and the long GT allele (included in the −413T/GT[L]/+99C haplotype) had significantly lower HO-1 concentrations at all time points compared with the other haplotypes (Fig. 3). From the 24 patients carrying the −413T/GT(L)/+99C haplotype, 14 had the first-day plasma HO-1 value. The −413T allele was also associated with lower third- to fourth-day HO-1 plasma concentrations compared with AA homozygotes (6.1 ng/mL vs. 6.9 ng/mL, P = 0.021).
Patients with the −413T/GT(L)/+99C haplotype had less frequent MOD (SOFA score, >6) (46%) than patients with other haplotypes (70%) (P = 0.017) on the first day in ICU. The HO-1 polymorphisms did not associate with the hospital mortality or the severity of disease measured with APACHE II and SAPS II scores (data not shown). We also investigated the associations between haplotypes, plasma HO-1, and outcome in major subgroups according to ICD-10 diagnosis but did not find any in patients with cardiac problems (cardiac arrest, acute myocardial infarction, and heart insufficiency; n = 71) (Table 3) or patients with infection (sepsis, respiratory tract infection, meningitis, and other bacterial infections; n = 67) (Table 3).
Association of HO-1 plasma concentrations with bilirubin, C-reactive protein, and illness severity
The first- and second-day HO-1 concentrations correlated significantly with the first- and second-day bilirubin values (r = 0.27, P = 0.001; r = 0.26, P = 0.001, respectively). The first-day HO-1 concentration was also significantly associated with the first-day C-reactive protein (CRP) concentration (r = 0.18, P = 0.02), the disease severity scores APACHE II and SAPS II (r = 0.27, P = 0.001; r = 0.17, P = 0.038), as well as the first-day and the maximum SOFA scores (r = 0.32, P < 0.001; r = 0.34, P < 0.001; Fig. 4). In the linear regression analysis, renal failure (P = 0.02) and ventilation failure (P = 0.02) associated independently with HO-1 values when organ failures were analyzed separately. Patients with MOD on the first day of intensive care had significantly higher HO-1 concentrations at all time points than patients with no MOD (first-day HO-1, 7.8 ng/mL vs. 6.1 ng/mL, P = 0.015; second day, 6.9 vs. 5.6 ng/mL, P = 0.003; third to fourth days 6.5 ng/mL vs. 5.6 ng/mL, P = 0.004; respectively).
Men had higher second- and third- to fourth-day HO-1 concentrations than women (P < 0.001, P = 0.001, respectively), although their disease and organ failure scores did not differ from women's scores. Age did not correlate with HO-1 concentrations. In the linear regression analysis to identify the variables independently associating with the first-day HO-1 concentration, carriage of T/GT(L)/C haplotype, first-day CRP values, sex, SAPS II, and SOFA scores were included. First-day SOFA score (P = 0.001) and T/GT(L)/C haplotype (P = 0.03) were found to associate independently with first-day HO-1 concentrations.
HO-1 plasma concentrations and mortality
The hospital mortality for those 160 patients with the day-1 plasma HO-1 was 21% (33/160). The patients with the day-1 plasma HO-1 did not differ from the patients without the plasma sample regarding hospital mortality, SAPS II, sex, SOFA, or the carriage of the T/GT(L)/C haplotype (data not shown), but they were slightly younger (median, 57 years vs. 61 years, P = 0.04). The first-day HO-1 concentrations did not differ significantly between hospital survivors and nonsurvivors. For hospital mortality, the first-day HO-1 value produced an area under the curve of 0.59 (95% CI, 0.47-0.71).
In this study, we have analyzed two SNPs and the GTn repeat length polymorphism of the HO-1 gene as well as HO-1 plasma levels in critically ill patients. We found that the HO-1 −413T/GT(L)/+99C haplotype had an independent effect on HO-1 plasma concentrations and associated with the appearance of MOD. Patients with the −413T/GT(L)/+99C haplotype had lower HO-1 plasma concentrations and lower appearance of MOD compared with other haplotypes on the first day of intensive care. In addition, HO-1 plasma levels were associated with disease severity and organ dysfunction.
HO-1 polymorphisms have been associated with different diseases and shown to affect the transcriptional activity of HO-1 gene (9, 10, 13). The HO-1 GTn allele distribution in our study is in agreement with the earlier studies performed in white patient population (27), especially with the one performed in Finnish population (28). We divided our study population into three categories according to its trimodal distribution. In previous studies, the GTn repeat length has been divided into two or three categories, with the lowest L allele length varying from 26 to 33 repeats. This inconsistency in the classification of GTn alleles may have affected the interpretation of the results and of the functional importance of this polymorphism. The +99G/C polymorphism causes a change of aspartic acid to histidine at amino acid position 7 of the HO-1 protein. Gulesserian et al. have also found linkage disequilibrium between the GTn and the +99G/C polymorphism (29), but neither the frequency nor the functionality of the +99G/C polymorphism has been investigated in other clinical studies. Because of the perfect LD of the +99C with the long GT allele, it remains to be elucidated which one of these two polymorphisms is functionally more important. Few studies suggest a greater functional importance of −413A/T polymorphism than the GTn repeat length polymorphism (10, 13). In our study, the T allele had only a minor effect on the third- to fourth-day HO-1 plasma concentration compared with the AA homozygote, indicating a greater significance of GTn repeat length polymorphism. Few studies have also reported linkage disequilibrium between GTn microsatellite polymorphism and −413A/T SNP (13, 18, 30); but to our knowledge, this is the first study to evaluate GTn microsatellite polymorphism together with both −413A/T and +99G/C SNPs.
We found that the carriers of the −413T/GT(L)/+99C haplotype had significantly lower HO-1 plasma concentrations, and this haplotype had an independent effect on the first-day plasma HO-1 levels. If plasma HO-1 is considered as a surrogate marker of HO-1 expression, this result is in line with the previous finding that transcriptional activity and ability to enhance the expression of the HO-1 gene decreases with increasing numbers of GTn repeats (9, 11, 30). Although many studies have found the short GTn to associate with a better outcome in many diseases related to intensive care, we found that the carriers of the long GT allele, included in the −413T/GT(L)/+99C haplotype, had lower appearance of MOD. Likewise, the long GT allele has recently been shown to protect against acute respiratory distress syndrome (18). Melley et al. (31) have suggested an optimal range of HO-1 induction in critically ill patients based on carboxyhemoglobin measurements. However, whether there is a possible therapeutic window for HO-1 expression and whether the −413T/GT(L)/+99C haplotype or lower plasma HO-1 concentration may be protective against MOD remain to be elucidated.
The higher HO-1 plasma levels were associated with MOD and more severe disease. Therefore, the lower appearance of MOD and lower HO-1 plasma levels in the carriers of −413T/GT(L)/+99C haplotype may partly explain the fact that we found no association between hospital mortality and the HO-1 polymorphisms or HO-1 plasma concentrations. High heterogeneity of the study population could be another explanation for this result. The investigated polymorphisms may also be in LD with another genetic marker with greater functional importance. In addition, the exact mechanism by which the polymorphisms affect the HO-1 expression is also unresolved.
The HO-1 plasma concentrations were remarkably higher in critically ill ICU patients compared with healthy controls, although the association between HO-1 levels and hospital mortality remained nonsignificant. Concentrations of circulating HO-1 have been investigated in few earlier studies (18, 32-34). The plasma HO-1 levels in our study were comparable to the serum or plasma HO-1 levels in the earlier studies using the same enzyme-linked immunosorbent assay method. The HO-1 plasma levels associated with bilirubin and CRP levels, although the association was not very strong. As a degradation product of heme, bilirubin could be considered as an indirect measure of HO-1 activity, although impaired elimination increases bilirubin levels as well. The weak association of HO-1 levels with bilirubin suggests that bilirubin may rather reflect the overall HO activity than the activity of HO-1, as bilirubin is produced by both HO-1 and the constitutive isoform HO-2. The association of HO-1 with CRP was evaluated because CRP is a commonly used marker of infection, inflammation, and tissue injury, and HO-1 levels were at highest in patients with infection or inflammation. The weak association of HO-1 with CRP suggests different roles for HO-1 and CRP in the critically ill. The CRP may represent more a marker of infection and inflammation, whereas HO-1 may instead represent a marker of oxidative stress and illness severity, as the association of HO-1 with APACHE II and SOFA scores was stronger.
The source of increased plasma HO-1 is unclear. HO-1 is an intracellular protein expressed in many cell types and tissues (2). In our study population, the patients with pancreatitis or bacterial infection had higher HO-1 concentrations compared with patients with cardiac diseases. The HO-1 production is also activated in acute myocardial infarction but the response may be rather local than systemic. The HO-1 response may also differ between tissues and stimuli. Pancreatitis is characterized by necrosis of pancreatic tissue and activation of systemic inflammation response (35). Hence, increased levels of plasma HO-1 may derive from injured organs or cells in these patients. This is supported by an earlier finding that pregnant women with pre-eclampsia had increased HO-1 mRNA and protein expression in decidua basalis as well as elevated serum HO-1 levels compared with healthy pregnant controls (36). Linearly, we found that the degree of organ dysfunction associated independently with first-day plasma HO-1, especially the renal and ventilation failure. The production of HO-1 is also increased to provide protection against oxidative and inflammatory stimuli. Monocytes are the main producers of HO-1 among peripheral blood mononuclear cells. Because their HO-1 production is upregulated in vivo in various acute inflammatory illnesses in children, they could serve as a source of plasma HO-1 (37).
Increased expression of HO-1 is considered to have a protective role in various stress situations and diseases (2). The HO-1 protects renal tubuli from oxidative injuries in various renal diseases (38). In the rat liver, induction of HO-1 protects against ischemia/reperfusion injury and improves microcirculation (39). In addition, HO-1 and CO are cytoprotective in various lung injury models (40). However, the role of increased HO-1 in the critically ill patients remains to be resolved. Because plasma HO-1 concentrations were associated with the degree of organ failure, and the carriers of −413T/GT(L)/+99C haplotype had lower plasma HO-1 levels together with lower frequency of MOD, we cannot exclude the possibility that increased HO-1 may also have detrimental effects. The release of pro-oxidative free iron by HO-1 may cause adverse effects as well as increased bilirubin (41, 42). Nevertheless, the increase of HO-1 in critically ill patients is likely to represent a defensive response against acute illness and organ injury to repair the underlying injury. The expression of HO-1 is commonly considered as a marker of oxidative stress. The higher plasma concentrations of HO-1 in more severely ill patients may thus simply reflect the magnitude of stress of these patients and suggest a possible use of HO-1 as a general marker of disease severity. This is supported by the increased HO-1 levels in all critically ill patients regardless of the diagnosis. Accordingly, arterial blood CO, bilirubin levels, and monocyte HO-1 protein expression were recently shown to be higher in severe septic shock compared with nonseptic patients (43). The role of HO-1 regarding MOD and the role of HO-1 polymorphisms in the regulation of HO-1 response need specification in further studies, however.
In conclusion, the HO-1 +99C and long GT allele (>33 repeats) were in perfect linkage disequilibrium in this Finnish critically ill patient population. The −413T/GT(L)/+99C haplotype was independently associated with the HO-1 plasma levels in these patients. As the significantly increased HO-1 concentrations in critically ill patients were associated with the degree of organ dysfunction, plasma HO-1 may originate from cell injury.
The authors thank Maritta Putkiranta and Jarkko Lakkisto for expert technical assistance.
1. Maines M: Heme oxygenase: function, multiplicity, regulatory mechanisms and clinical implications. FASEB J
2. Otterbein LE, Choi AM: Heme oxygenase: colors of defence against cellular stress. Am J Physiol Lung Cell Mol Physiol
3. Mumby S, Upton RL, Chen Y, Stanford SJ, Quinlan GJ, Nicholson AG, Gutteridge JM, Lamb NJ, Evans TW: Lung heme oxygenase-1 is elevated in acute respiratory distress syndrome. Crit Care Med
4. Mohri T, Ogura H, Koh T, Fujita K, Sumi Y, Yoshiya K, Matsushima A, Hosotsubo H, Kuwagata Y, Tanaka H, et al.: Enhanced expression of intracellular heme oxygenase-1 in deactivated monocytes from patients with severe systemic inflammation response syndrome. J Trauma
5. Schaer DJ, Schaer CA, Schoedon G, Imhof A, Kurrer MO: Hemophagocytic macrophages constitute a major compartment of heme oxygenase expression in sepsis. Eur J Haematol
6. Farkas I, Maroti Z, Katona M, Endreffy E, Monostori P, Máder K, Túri S: Increased heme oxygenase-1 expression in premature infants with respiratory distress syndrome. Eur J Pediatr
7. Schmidt JE, Morgan JI, Rodriguez-Galindo C, Webb DL, Liang H, Tamburro RF: Heme oxygenase-1 messenger RNA expression is induced in peripheral blood mononuclear cells of pediatric cancer patients with systemic inflammatory response syndrome. Pediatr Crit Care Med
8. Fujii H, Takahashi T, Matsumi M, Kaku R, Shimizu H, Yokoyama M, Ohmori E, Yagi T, Sadamori H, Tanaka N, et al: Increased heme oxygenase-1 and decreased delta-aminolevulinate synthase expression in the liver of patients with acute liver failure. Int J Mol Med
9. Hirai H, Kubo H, Yamaya M, Nakayama K, Numasaki M, Kobayashi S, Suzuki S, Shibahara S, Sasaki H: Microsatellite polymorphism in heme oxygenase-1 gene promoter is associated with susceptibility to oxidant-induced apoptosis in lymphoblastoid cell lines. Blood
10. Ono K, Goto Y, Takagi S, Baba S, Tago N, Nonogi H, Iwai N: A promoter variant of the heme oxygenase-1 gene may reduce the incidence of ischemic heart disease in Japanese. Atherosclerosis
11. Brydun A, Watari Y, Yamamoto Y, Okuhara K, Teragawa H, Kono F, Chayama K, Oshima T, Ozono R: Reduced expression of heme oxygenase-1 in patients with coronary atherosclerosis. Hypertens Res
12. Exner M, Böhmig GA, Schillinger M, Regele H, Watschinger B, Hörl WH, Raith M, Mannhalter C, Wagner OF: Donor heme oxygenase-1 genotype is associated with renal allograft function. Transplantation
13. Buis CI, van der Steege G, Visser DS, Nolte IM, Hepkema BG, Nijsten M, Slooff MJ, Porte RJ: Heme oxygenase-1 genotype of the donor is associated with graft survival after liver transplantation. Am J Transplant
14. Kaneda H, Ohno M, Taguchi J, Togo M, Hashimoto H, Ogasawara K, Aizawa T, Ishizaka N, Nagai R: Heme oxygenase-1 gene promoter polymorphism is associated with coronary artery disease in Japanese patients with coronary risk factors. Arterioscler Thromb Vasc Biol
15. Tiroch K, Koch N, Beckerath N, Kastrati A, Schömig A: Heme oxygenase-1 gene promoter polymorphism and restenosis following coronary stenting. Eur Heart J
16. Schillinger M, Exner M, Minar E, Mlekusch W, Müllner M, Mannhalter C, Bach FH, Wagner O: Heme oxygenase-1 genotype and restenosis after balloon angioplasty: a novel vascular protective factor. J Am Coll Cardiol
17. Yamada N, Yamaya M, Okinaga S, Nakayama K, Sekizawa K, Shibahara S, Sasaki H: Microsatellite polymorphism in the heme oxygenase-1 gene promoter is associated with susceptibility to emphysema. Am J Hum Genet
18. Sheu CC, Zhai R, Wang Z, Gong MN, Tejera P, Chen F, Su L, Thompson BT, Christiani DC: Heme oxygenase-1 microsatellite polymorphism and haplotypes are associated with the development of acute respiratory distress syndrome. Intensive Care Med
19. Le Gall JR, Lemeshow S, Saumer F: A new simplified acute physiology score (SAPS II) based on a European/North American multicenter study. JAMA
20. Knaus WA, Drapper EA, Wagner DP, Zimmerman JE: APACHE II: a severity of disease classification system. Crit Care Med
21. Vincent J-L, De Mendonça A, Cantraine F, Moreno R, Takala J, Suter PM, Sprung CL, Colardyn F, Blecher S: Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on "sepsis-related problems" of the European Society of Intensive Care Medicine. Crit Care Med
22. Rustad P, Simonsson P, Felding P, Pedersen M: Nordic reference interval project bio-bank and database (NOBIDA): a source for future estimation and retrospective evaluation of reference intervals. Scand J Clin Lab Invest
23. Barrett JC, Fry B, Maller J, Daly MJ: Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics
24. Akey J, Jin L, Xiong M: Haplotypes vs single marker linkage disequilibrium tests: what do we gain? Eur J Hum Genet
25. Zhang K, Calabrese P, Nordborg M, Sun F: Haplotype block structure and its applications to association studies: power and study designs. Am J Hum Genet
26. Purcell S, Cherny SS, Sham PC: Genetic Power Calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics
27. Guénégou A, Leynaert B, Bénessiano J, Pin I, Demoly P, Neukirch F, Boczkowski J, Aubier M: Association of lung function decline with the heme oxygenase-1 gene promoter microsatellite polymorphism in a general population sample. Results from the European Community Respiratory Health Survey (ECRHS), France. J Med Genet
28. Turpeinen H, Kyllönen L, Parkkinen J, Laine J, Salmela KT, Partanen J: Heme oxygenase 1 gene polymorphisms and outcome of renal transplantation. Int J Immunogenet
29. Gulesserian T, Wenzel C, Endler G, Sunder-Plassmann R, Marsik C, Mannhalter C, Iordanova N, Gyöngyösi M, Wojta J, Mustafa S, et al: Clinical restenosis after coronary stent implantation is associated with the heme oxygenase-1 gene promoter polymorphism and the heme oxygenase-1 +99G/C variant. Clin Chem
30. Song F, Li X, Zhang M, Yao P, Yang N, Sun X, Hu FB, Liu L: Association between heme oxygenase-1 gene promoter polymorphisms and type 2 diabetes in a Chinese population. Am J Epidemiol
31. Melley DD, Finney SJ, Elia A, Lagan AL, Quinlan GJ, Evans TW: Arterial carboxyhemoglobin level and outcome in critically ill patients. Crit Care Med
32. Schipper HM, Chertkow H, Mehindate K, Frankel D, Melmed C, Bergman H: Evaluation of heme oxygenase-1 as a systemic biological marker of sporadic AD. Neurology
33. Kirino Y, Takeno M, Iwasaki M, Ueda A, Ohno S, Shirai A, Kanamori H, Tanaka K, Ishigatsubo Y: Increased serum HO-1 in hemophagocytic syndrome and adult-onset Still's disease: use in the differential diagnosis of hyperferritinemia. Arthritis Res Ther
34. Sato T, Takeno M, Honma K, Yamauchi H, Saito Y, Sasaki T, Morikubo H, Nagashima Y, Takagi S, Yamanaka K, et al: Heme oxygenase-1, a potential biomarker of chronic silicosis, attenuates silica-induced lung injury. Am J Respir Crit Care Med
35. Pandol SJ, Saluja AK, Imrie CW, Banks PA: Acute pancreatitis: bench to the bedside. Gastroenterology
36. Eide IP, Isaksen CV, Salvesen KA, Langaas M, Schonberg SA, Austgulen R: Decidual expression and maternal serum levels of heme oxygenase 1 are increased in pre-eclampsia. Acta Obstet Gynecol Scand
37. Yachie A, Toma T, Mizuno K, Okamoto H, Shimura S, Ohta K, Kasahara Y, Koizumi S: Heme oxygenase-1 production by peripheral blood monocytes during acute inflammatory illnesses of children. Exp Biol Med
38. Morimoto K, Ohta K, Yachie A, Yang Y, Shimizu M, Goto C, Toma T, Kasahara Y, Yokoyama H, Miyata T, et al: Cytoprotective role of heme oxygenase (HO)-1 in human kidney with various renal diseases. Kidney Int
39. Schmidt R, Trischler E, Hoetzel A, Loop T, Humar M, Halverscheid L, Geiger KK, Pannen BH: Heme oxygenase-1 induction by the clinically used anesthetic isoflurane protects rat livers from ischemia/reperfusion injury. Ann Surg
40. Jin Y, Choi AMK: Cytoprotection of heme oxygenase-1/carbon monoxide in lung injury. Proc Am Thorac Soc
41. Emerit J, Beaumont C, Trivin F: Iron metabolism, free radicals, and oxidative injury. Biomed Pharmacother
42. Immenschuh S, Shan Y, Kroll H, Santoso S, Wössmann W, Bein G, Bonkovsky HL: Marked hyperbilirubinemia associated with the heme oxygenase-1 gene promoter microsatellite polymorphism in a boy with autoimmune hemolytic anemia. Pediatrics
43. Takaki S, Takeyama N, Kajita Y, Yabuki T, Noguchi H, Miki Y, Inoue Y, Nakagawa T, Noguchi H: Beneficial effects of the heme oxygenase-1/carbon monoxide system in patients with severe sepsis/septic shock. Intensive Care Med