- Email: [email protected]
Serum selenium levels in cirrhotic patients are not influenced by the disease severity index
Available online at www.sciencedirect.com
Nutrition Research 30 (2010) 574 – 578 www.nrjournal.com
Serum selenium levels in cirrhotic patients are not influenced by the disease severity index Manuel Martínez-Peinado a , Flor Nogueras-López b , Antonio Arcos-Cebriánc , Ahmad Agil d , Miguel Navarro-Alarcón a,⁎ a
Department of Nutrition and Food Science, School of Pharmacy, University of Granada, E-18071 Granada, Spain b Department of Statistics, School of Pharmacy, University of Granada, E-18071 Granada, Spain c Department of Hepatology, Virgen de las Nieves University Hospital, E-18014 Granada, Spain d Deparment of Pharmacology, School of Medicine, University of Granada, 18071-Granada, Spain Received 9 June 2010; revised 13 August 2010; accepted 17 August 2010
Abstract Selenium (Se) is an antioxidant element that protects against cellular damage by reactive oxygen species. Therefore, total serum Se concentration may reflect protection during the development of cirrhosis, an oxidative stress-related disease. We hypothesized that serum Se levels are diminished in cirrhotic patients due to their enhanced oxidative stress, and serum Se levels are reduced the most in patients with the highest severity of cirrhosis. A case-control study was performed to determine whether cirrhosis is associated with changes in serum Se levels. Blood samples from 30 healthy controls and 93 cirrhotic patients were analyzed for total serum Se by hydride generation atomic absorption spectrometry. The Child-Pugh index score was used to evaluate the severity of liver disease. The mean serum Se concentration was significantly lower in patients vs controls (0.721 ± 0.239 vs 0.926 ± 0.241 μmol/L; P = .001). Mean serum Se levels were not significantly lower in patients with higher severity of cirrhosis (0.691 ± 0.229 vs 0.755 ± 0.255 μmol/L; P = .144). A positive and significant correlation was found between age and serum Se levels in patients (r = 0.277, P = .007). Patients showed significant sex differences in serum Se level (higher in male) and severity index (higher in female). The significantly decreased serum Se level in patients indicates that the Se component of the antioxidant system is severely impaired in cirrhosis. However, serum Se levels were not influenced by the severity of the disease. © 2010 Elsevier Inc. All rights reserved. Keywords: Abbreviations:
Serum Se; Cirrhosis; Severity of liver disease; Case-control study, Female and male patients Se, selenium; GSH-Px-3, glutathione peroxidase; CPI, Child-Pugh index.
1. Introduction Animals are known to synthesize numerous different intermediary metabolites during the conversion of inorganic selenium (Se) to organic forms and vice versa. The native Se forms in human plasma are reported to be selenotrisulfides, ⁎ Corresponding author. Tel.: +34 58 243865; fax: +34 58 243869. E-mail address: [email protected] (M. Navarro-Alarcón). 0271-5317/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.nutres.2010.08.004
Se in other oxidation states (eg, protein-bound selenide, selenite, and selenate), and organically bound Se [1]. Selenium is present in the form of selenoamino acids in the primary structure of proteins distributed among 3 compartments: glutathione peroxidase (isoform GSH-Px-3), containing ≈12% of plasma Se; selenoprotein P, containing 44% to 60%; and a nonspecific fraction (mostly selenomethionine), containing the remaining 44% [2,3]. Some of the Se is also present in reactive forms, such Se-diglutathione
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578
[4]. Due to its chemical similarity to sulfur, selenomethionine can replace methionine in native proteins [1]. Among Se-containing proteins, selenoprotein P was reported to be the best index of human nutritional status [5]. Most selenoproteins serve as enzymes that catalyze redox reactions. Selenoprotein P contains numerous potential redox centers in its cysteine and selenocysteine residues [5,6]. The reactive oxygen species that result from alteration of the cell redox state has been implicated [7] in the liver damage induced by different factors (eg, alcohol, virus, lipid alterations, or carbohydrate metabolism disorders). Most studies in humans observed significantly lower serum and plasma Se levels in patients with different degrees of hepatocellular injury than in healthy controls [8-13]. This difference was especially marked when the liver injury was due to the intake of toxic substances such as alcohol [5,14]. One group [15] reported that Se supplementation in cirrhotic individuals with a background of nutritional element deficiency did not restore hepatic glutathione peroxidase or Se to normal levels. As noted above, levels of nutritional Se biomarker are generally significantly reduced in patients with liver disease [13], but it remains unclear whether this low Se status is a preexisting factor. Selenium has also been found to have an antagonistic effect on fluoride-induced oxidative stress in human hepatocytes [16] and a hepatoprotective effect in experimental thioacetamide-induced liver cirrhosis [17]. Other researchers observed reduced Se levels in carbon tetrachloride–induced liver cirrhosis [18]. It has been proposed that selenium protects cell components from damage by reactive oxygen species through selenium-containing proteins, such as GSH-Px-3 [6,13]. A previous study reported a significant decrease in mean total serum Se levels in patients with liver disease [12] and indicated a significant decrease in these levels with greater intensity of liver damage, but only 12 cirrhotic patients were included in the study [12]. There remains a need to accurately establish the role of Se protection in the development of cirrhosis, examining the influence of the impairment of cirrhosis on serum Se levels in these patients. With the above background, we hypothesized that the antioxidant defense exerted by Se is impaired in cirrhotic patients and that the disease severity index has a negative influence on serum Se levels. Hence, the first objective of this case-control study was to assess the diminishment of serum Se level in cirrhosis. A second objective was to test the influence of the cirrhosis severity index on the cirrhosis impairs Se status of patients. With these objectives, blood samples from 30 healthy controls and 93 cirrhotic patients were analyzed for total serum Se by hydride generation atomic absorption spectrometry. The Child-Pugh index (CPI) score was used to evaluate the severity of liver disease The aim of this study was to increase understanding of the involvement of this antioxidant trace element in oxidative stress-related diseases
575
such as cirrhosis, as well as in healthy nutrition. Few data have been published to date on the effect of the severity of liver disease on Se metabolism. 2. Methods and materials 2.1. Subjects The initial study population comprised patients under follow-up by the Hepatology Department of our reference hospital. The diagnosis of cirrhosis was based on clinical and ultrasound criteria. Healthy controls were enrolled from among blood donors in the same geographic area. All subjects were fully informed of the aims of the study, and an informed consent was obtained from each one before blood was drawn. Exclusion criteria for study participation were the presence or history of non–insulin-dependent diabetes mellitus, wide bowel resection, gastrectomy, inflammatory bowel disease, cancer, chronic renal failure, active alcoholism, corticoid therapy, or hypercortisolism and the receipt of drugs that interfere with muscle function. In addition, subjects under dietary supplements or drugs with a known antioxidant capacity were excluded. Serum Se levels found in these patients were compared with those measured in 30 healthy controls from the same area. The patients were divided into 2 groups according to their CPI score for severity of the liver disease: CPI A (patients with a well-balanced disease; between 5 and 6 points) and B (patients with a significant functional compromise; between 7 and 9 points) plus C (patients with an unbalanced disease; between 10 and 15 points). Therefore, the CPI A and B plus C corresponded to low and medium plus high severity, respectively [19,20]. The final CPI score was obtained as a sum of the assigned points by the physician to the ascites degree, bilirubin and albumin plasma concentrations, prothrombin time, and the encephalopathy degree. This CPI score was a model initially proposed with the aim to stratify the surgical risk in patients with portal blood imbalance [19] and later modified by others [20] when the status nutritional parameter was changed by the prothrombin time [21]. The patients were also divided into 3 groups according to age (b30, 31-50, and N50 years) to have a balanced number of patients in each group. The institutional approval for using human subjects in the study was obtained from the Virgen de las Nieves University Hospital. The study, including the release of human serum samples, was approved by the ethics committee of the hospital and was conducted in accordance with the Helsinki Declaration. 2.2. Blood samples Blood was drawn from the antecubital vein, with the subject seated after overnight fasting. Samples were left to coagulate spontaneously and then centrifuged at 3000g to obtain serum separated by gelose. Aliquots (1 mL) of serum
576
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578
were frozen at −25°C for transportation and storage in the laboratory of the Department of Nutrition and Food Science (University of Granada) until analyzed [19]. 2.3. Total Se measurement Sample mineralization (200 μL of serum) was performed using a HNO3/HClO4 mixture heated to 110°C in a thermostated block following a previously optimized procedure [22]. We then added 1.25-mL HClcc and heated to 80°C for 15 minutes to reduce the Se (VI) to Se (IV). When cool, this was diluted to 10 mL with a 0.5% (vol/vol) HCl solution. This 10-mL aliquot was then transferred to a reaction vessel, which was placed in the MHS-10 system. Selenium determination was carried out using the hydride generation atomic absorption spectrometry technique (Perkin-Elmer model 1100B atomic absorption spectrometer equipped with a Perkin-Elmer MHS-10 hydride generator; Perkin-Elmer, Norwalk, CT, USA). Hydride generation was carried out using a solution of 3% (wt/vol) NaBH4 in 1% (wt/vol) NaOH. All the samples were analyzed in triplicate. The absorbance measured in each sample (peak height mode) was correlated with the Se concentration by the addition calibration method. The accuracy (100.35%) and precision (5.31%) of the method were also tested with the reference material 0148 (Contox trace metal serum control Panel C from the Kaulson Laboratories Inc, West Caldwell, NY). The Se concentrations obtained for the reference material by the technique used (1.928 ± 0.110 μmol/L) were not significantly different (P N .05) from certified levels (1.900 ± 0.076 μmol/L). 2.4. Statistical analyses Statistical analyses were performed with SPSS software (version 15.0 for Windows program; SPSS, Chicago, Ill). The Kolmogorov-Smirnov test was used to examine the normal distribution of variables and the Levene's test to study the homogeneity of variances. Data were compared by applying the Student t test for parametric variables and the Kruskal-Wallis test for nonparametric variables. A nonparametric distribution was found for serum Se, age, and cirrhosis severity variables (P values by KolmogorovSmirnov test were .014, .003, and .004, respectively). Correlation analyses using the Spearman coefficient were also applied for assessing linear relationships between the evaluated parameters. For the relationships between selenium and age and between patients' age and CPI score, a simple linear model was chosen because of its explanatory power and because it does not differ significantly from any other models. Serum Se and cirrhosis severity results are expressed as arithmetic mean ± SD. A 3-way variance analysis to discern the relative or combined roles of the 3 factors (sex, age, and severity of cirrhosis) on serum Se levels of the patients was performed. In this case, the values were presented as medians ± SE. Results were considered statistically significant at P b .05.
3. Results and discussion The final patient sample comprised 93 patients (39 females and 54 males) with a diagnosis of cirrhosis; their mean (SD) age was 41.6 (16.6) years. The control group comprised 30 healthy individuals (13 females and 17 males) with a mean (SD) age of 40.7 (18.8) years. Serum Se concentrations were significantly higher (P b .05) in the oldest patient group (N50 years) compared with the younger age groups (b30 and 31-50 years; Table 1). In addition, serum Se levels showed a linear correlation with age (r = 0.277, P = .007) in cirrhotic patients, similar to previous reports in healthy individuals [23]. The same general tendency was also observed by other authors in several countries, but conclusions about age dependency are indeed questionable due to the fact that the group studied is sometimes poorly defined (ie, the age-range too small or too large) [24]. Serum Se concentrations were previously found to be significantly higher in the oldest age group in a study of healthy primary school children from northeast Thailand [25]. In the present study, higher patient age was inversely correlated with the severity (index) of the disease (P b .001, r = −0.456), with younger cirrhotics having more severe disease. This is not a common finding in alcoholic cirrhotics, reflecting the distinct etiology of the disease in our patients. Patients with active alcoholism were not included in the study. One group reported a nonsignificant reduction in serum Se levels with higher age in patients with liver disease, commenting that published findings on this issue have been controversial [12]. Mean serum Se concentrations of the patient and control groups are shown in Table 1. Serum Se levels were significantly lower (P b .001) in cirrhotic patients than in healthy controls, as previously reported in patients with liver disease [5,12,13,15,26-28]. One group who found significantly reduced levels of antioxidants, including Se, in alcoholic cirrhotics (vs sex-matched controls) described a relationship between cirrhosis and oxidative stress. Other researchers [26] who also found a significantly lower Se levels in cirrhotics (vs noncirrhotics and controls) reported that an unbalanced nutritional status was the strongest predictor of low Se values in alcoholics.
Table 1 Serum Se levels in cirrhosis patients by age group and in healthy controls Individuals
n
Total Se (μmol/L) Mean ± SD
Cirrhosis patients (y) b30 31-50 N50 Healthy controls
93 23 38 32 30
a
0.721 ± 0.239 0.662 b ± 0.236 0.680 b ± 0.198 0.811 b ± 0.268 0.926 a ± 0.241
Range 0.157-1.311 0.248-1.258 0.157-1.311 0.157-1.311 0.488-1.369
a P b .001. Mean total serum Se level measured in patients was significantly lower. b P b .012. Mean total serum Se level measured in patients older than 50 years was significantly higher.
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578 Table 2 Serum Se levels in cirrhosis patients by sex Sex
n
Total Se (μmol/L) a
Female Male All
39 54 93
Mean ± SD
Range
0.665 ± 0.228 0.761 ± 0.243 0.721 ± 0.239
0.157-1.311 0.157-1.311 0.157-1.311
a
P = .047. Mean total serum Se level measured in male was significantly higher.
The Se plays an important antioxidant role in various biochemical reactions as a cofactor of GSH-Px [25-29]. This enzyme acts against the onset and progression of events producing liver injury by protecting cells against free radical formation from hydroperoxide ions. A recent study found that Se supplementation to cirrhotic patients did not restore their serum Se and glutathione peroxidase to normal levels [15]; this finding that could be related with the fact that plasma Se levels take months to return to normal in proteinenergy undernourishment, after nutritional rehabilitation. However, there have been few human trials on the prophylactic value of Se against the processes implicated in cirrhosis. A study in rats [30] showed that Se may play an important role in preventing the induction of hepatic cellular injury by carbon tetrachloride. Other researchers [31] found that the Se supplementation increased liver expression of manganese superoxide dismutase in rats, improving their antioxidant defense, thus supporting the anti-inflammatory and hepatoprotective effects claimed for Se. Table 2 shows the mean serum Se concentrations in cirrhotic patients by sex. Concentrations were significantly higher in the males than in the females (P b .05), similar to previous reports even in healthy individuals [23]. The severity index was significantly higher in the female vs male cirrhotic patients (CPI: 7.81 ± 1.71 vs 6.28 ± 1.31, respectively; P b .001). As shown in Table 3, serum Se levels of the patients were not significantly correlated with the CPI-determined severity of their disease (P N .05). Therefore, as the CPI can take values from 5 to 15, this regression analysis is semicontinuous. This finding is not in agreement with previous reports of an association between higher severity of liver disease and greater impairment of body Se status [13,32]. In contrast, another study [33] concluded that decreased serum Se levels result from liver function Table 3 Correlation of serum Se levels in cirrhosis patients with severity of liver disease (CPI) CPI
n
Child-Pugh A Child-Pugh B + C a
42 51
Total Se (μmol/L) Mean a ± SD
Range
0.755 ± 0.255 0.691 ± 0.229
0.157-1.311 0.248-1.311
P = .144. No significant difference was found.
577
impairment of any etiology. Taking the present results into account, dietary or prescribed Se supplementation for cirrhotic patients should not be ruled out. We performed a 3-way variance analysis to discern the relative and combined roles of the 3 factors (sex [male and female], age [b30, 31-50, and N50 years], and severity of cirrhosis [CPI A vs B + C]) on the serum Se levels of the patients, finding no interaction effect of these factors on these levels (P N .05). The severity of cirrhosis was not related to serum Se levels (P = .763), but sex (difference between medians, 7.66; SE, 3.86) and overall age were independently related to these levels (P b .05). Thus, serum Se concentrations significantly differed between the oldest patient group (N50 years) and the younger patients (b30 years: difference between medians, 11.9; SE, 4.94; 31-50 years: difference between medians, 11.6; SE, 4.8). In our opinion, the main limitation of the study was the number of studied cirrhosis patients; in fact, for the CPI C group, only 1 patient was considered. Although serum Se levels were not significantly influenced by the severity index of the liver disease (consequently, we reject the main hypothesis for the research), future studies should be performed with a higher, more balanced number of cirrhosis patients for the CPI A, B, and C groups to elucidate definitively the hypothesis. In conclusion, serum Se levels were significantly lower in cirrhotic patients than in healthy controls but were not significantly influenced by the CPI-determined severity of liver disease. Serum Se concentrations were significantly higher in male vs female cirrhotics, whereas the severity index was significantly higher in the female vs male cirrhotics. Serum Se levels were significantly lower in younger vs older patients. Our results demonstrate that the Se component of the antioxidant system is severely impaired in cirrhosis.
Acknowledgment This study was partially supported by a grant: AGR-109 (Junta de Andalucía, Spain).
References [1] Rhead WJ, Evans GA, Schrauzer GN. Selenium in plasma: level sin blood proteins and behavior upon dialysis, acidification, and reduction. Bioinorg Chem 1974;3:217-23. [2] Burk RF, Early DS, Hill KE, Palmer IS, Boeglin ME. Plasma selenium in patients with cirrhosis. Hepatology 1998;27:794-8. [3] Speckmann B, Walter PL, Alili L, Reihner R, Sies H, Klotz LO, et al. Selenoprotein P expressions is controlled through interaction of the coactivator PGC-1α with fox01a and hepatocyte nuclear factor 4α transcription factors. Hepatology 2008;48:1998-2006. [4] Mas A, Sarkar B. Role of glutathione in selenite binding by human plasma. Biol Trace Elem Res 1988;20:95-104. [5] Burk RF, Hill KE. Selenoprotein P: an extracellular protein with unique physical characteristics and a role in selenium homeostasis. Ann Rev Nutr 2005;25:215-35.
578
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578
[6] Koyama H, Abdulah R, Ohkubo T, Imai Y, Satoh H, Nagai K. Depressed serum selenoprotein P: possible new predicator of increased risk for cerebrovascular events. Nutr Res 2009;29:94-9. [7] Morisco F, Vitaglione P, Amoruso D, Russo B, Fogliano V, Caparoso N. Foods and liver health. Mol Aspects Med 2008;29: 144-50. [8] Lecomte E, Herberth B, Pirrollet P, Chancerelle Y, Arnaud J, Musse N, et al. Effect of alcohol consumption on blood antioxidant nutrients and oxidative stress indicators. Am J Clin Nutr 1994;60: 255-61. [9] Loguercio C, De Girolamo V, Federico A, Feng SL, Cataldi V, Del Vecchio Blanco C, et al. Trace elements and chronic liver diseases. J Trace Elem Med Biol 1997;11:158-61. [10] Guarini P, Stanzial AM, Olivieri O, Casaril M, Galvani S, Pantalena M, et al. Erythrocyte membrane lipids and serum selenium in post-viral and alcoholic cirrhosis. Clin Chim Acta 1998;270:139-50. [11] Navarro-Alarcón M, López-Martínez MC. Essentiality of selenium in the human body: relationship with different diseases. Sci Total Environ 2000;249:347-71. [12] Navarro-Alarcon M, Lopez-Ga de la Serrana H, Perez-Valero V, Lopez Martinez MC. Selenium concentrations in serum of individuals with liver diseases (cirrhosis or hepatitis): relationship with some nutritional and biochemical markers. Sci Total Environ 2002;291:135-41. [13] Navarro-Alarcon M, Cabrera-Vique. Selenium in food and the human body: a review. Sci Total Environ 2008;400:115-41. [14] Thuluvath PJ, Triger DR. Selenium in chronic liver disease. J Hepathol 1992;14:176-82. [15] Zhang J, Wang H, Yu H. Thioacetamide-induced cirrhosis in seleniumadequate mice displays rapid and persistent abnormity of hepatic selenoenzymes which are mute to selenium supplementation. University of Science and Technology of China, Southern Campus; 2007. [16] Wang A, Xia T, Yuan J, Chen X, Wang KY. Antagonistic effect of selenium on oxidative stress, DNA damage, and apoptosis induced by fluoride in human hepatocytes. Fluoride 2004;37:107-16. [17] Al-Bader A, Abul H, Hussain T, Al-Moosawi M, Mathew TC, Dashti H. Selenium and liver cirrhosis. Mol Cel Biochem 1998;185:1-6. [18] González-Reimers E, López-Lirola A, Martín Olivera R, SantolariaFernández F, Galindo-Martín L, Abreu-González P, et al. Effects of protein deficiency on liver trace elements and antioxidant activity in carbon tetrachloride-induced liver cirrhosis. Biol Trace Elem Res 2003;93:127-39. [19] Child GG, Turcotte JG. Surgery and portal hypertension. In: Child GG, editor. The liver and portal hypertension. Philadelphia: Saunders; 1964. p. 50-2.
[20] Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-9. [21] Meijide Miguez HM. Child Pugh. Clasificación pronóstica de la hepatopatía. http://www.meiga.info/Escalas/CHILD-PUGH.pdf. [22] Navarro M, Lopez H, Ruiz ML, Gonzalez S, Perez V, Lopez MC. Determination of selenium in serum by hydride generation atomic absorption spectrometry for calculation of daily dietary intake. Sci Total Environ 1995;175:245-52. [23] Lopes PA, Santos MC, Vicente L, Rodrigues MO, Pavao ML, Neve J, et al. Trace element status (Se, Cu Zn) in healthy Portuguese subjects from Lisbon population. Biol Trace Elem Res 2004;101:1-17. [24] Robberecht H, Deelstra H. Factors influencing blood selenium concentration values—a literature review. J Trace Elem Electrolytes Health Dis 1994;8:129-43. [25] Krittaphol W, Bailey KB, Pongcharoen T, Winichagoon P, Thomson C, Gibson RS. Primary school children from northeast Thailand are not at risk of selenium deficiency. Asia Pac J Clin Nutr 2006;15:474-81. [26] Gonzalez-Reimers E, Galindo-Martin L, Santolaria-Fernandez F, Sanchez-Perez MJ, Alvisa-Negrin J, Garcia-Valdecasas- Campelo EG, et al. Prognostic value of serum selenium levels in alcoholics. Biol Trace Elem Res 2008;125:22-9. [27] Rai M, Pal MR, Sabihta P, Pai P. Hepatoprotective influence of selenium in experimental liver cirrhosis. Kuwait Med J 2001;33:333-6. [28] Pemberton PW, Smith A, Warnes TW. Non-invasive monitoring of oxidant stress in alcoholic liver disease. Scand J Gastroenterol 2005; 40:1102-8. [29] Zulet MA, Puchau B, Hermsdorff HHM, Navarro C, Martínez JA. Dietary selenium is negatively associated with serum sialic acid and metabolic syndrome features in healthy young adults. Nutr Res 2009; 29:41-8. [30] Ozardali I, Bitiren M, Karakilcik AZ, Zerin M, Aksoy N, Musa D. Effects of selenium on histopathological and enzymatic changes in experimental liver injury of rats. ExpToxicol Pathol 2004;56:59-64. [31] Shilo S, Pardo M, Aharon-Simon M. Selenium supplementation increases liver MnSOD expression: molecular mechanism for hepatoprotection. J Inorg Biochem 2008;102:110-8. [32] Valimaki MJ, Harju KJ, Ylikahri RH. Decreased serum selenium in alcoholics, a consequence of liver dysfunction. Br J Nutr 1986;55: 227-33. [33] Lin CC, Huang JF, Tsai LY, Huang YL. Selenium, iron, copper, and zinc levels and copper-to zinc ratios in serum of patients at different stages of viral hepatic diseases. Biol Trace Elem Res 2006;109: 15-23.
Nutrition Research 30 (2010) 574 – 578 www.nrjournal.com
Serum selenium levels in cirrhotic patients are not influenced by the disease severity index Manuel Martínez-Peinado a , Flor Nogueras-López b , Antonio Arcos-Cebriánc , Ahmad Agil d , Miguel Navarro-Alarcón a,⁎ a
Department of Nutrition and Food Science, School of Pharmacy, University of Granada, E-18071 Granada, Spain b Department of Statistics, School of Pharmacy, University of Granada, E-18071 Granada, Spain c Department of Hepatology, Virgen de las Nieves University Hospital, E-18014 Granada, Spain d Deparment of Pharmacology, School of Medicine, University of Granada, 18071-Granada, Spain Received 9 June 2010; revised 13 August 2010; accepted 17 August 2010
Abstract Selenium (Se) is an antioxidant element that protects against cellular damage by reactive oxygen species. Therefore, total serum Se concentration may reflect protection during the development of cirrhosis, an oxidative stress-related disease. We hypothesized that serum Se levels are diminished in cirrhotic patients due to their enhanced oxidative stress, and serum Se levels are reduced the most in patients with the highest severity of cirrhosis. A case-control study was performed to determine whether cirrhosis is associated with changes in serum Se levels. Blood samples from 30 healthy controls and 93 cirrhotic patients were analyzed for total serum Se by hydride generation atomic absorption spectrometry. The Child-Pugh index score was used to evaluate the severity of liver disease. The mean serum Se concentration was significantly lower in patients vs controls (0.721 ± 0.239 vs 0.926 ± 0.241 μmol/L; P = .001). Mean serum Se levels were not significantly lower in patients with higher severity of cirrhosis (0.691 ± 0.229 vs 0.755 ± 0.255 μmol/L; P = .144). A positive and significant correlation was found between age and serum Se levels in patients (r = 0.277, P = .007). Patients showed significant sex differences in serum Se level (higher in male) and severity index (higher in female). The significantly decreased serum Se level in patients indicates that the Se component of the antioxidant system is severely impaired in cirrhosis. However, serum Se levels were not influenced by the severity of the disease. © 2010 Elsevier Inc. All rights reserved. Keywords: Abbreviations:
Serum Se; Cirrhosis; Severity of liver disease; Case-control study, Female and male patients Se, selenium; GSH-Px-3, glutathione peroxidase; CPI, Child-Pugh index.
1. Introduction Animals are known to synthesize numerous different intermediary metabolites during the conversion of inorganic selenium (Se) to organic forms and vice versa. The native Se forms in human plasma are reported to be selenotrisulfides, ⁎ Corresponding author. Tel.: +34 58 243865; fax: +34 58 243869. E-mail address: [email protected] (M. Navarro-Alarcón). 0271-5317/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.nutres.2010.08.004
Se in other oxidation states (eg, protein-bound selenide, selenite, and selenate), and organically bound Se [1]. Selenium is present in the form of selenoamino acids in the primary structure of proteins distributed among 3 compartments: glutathione peroxidase (isoform GSH-Px-3), containing ≈12% of plasma Se; selenoprotein P, containing 44% to 60%; and a nonspecific fraction (mostly selenomethionine), containing the remaining 44% [2,3]. Some of the Se is also present in reactive forms, such Se-diglutathione
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578
[4]. Due to its chemical similarity to sulfur, selenomethionine can replace methionine in native proteins [1]. Among Se-containing proteins, selenoprotein P was reported to be the best index of human nutritional status [5]. Most selenoproteins serve as enzymes that catalyze redox reactions. Selenoprotein P contains numerous potential redox centers in its cysteine and selenocysteine residues [5,6]. The reactive oxygen species that result from alteration of the cell redox state has been implicated [7] in the liver damage induced by different factors (eg, alcohol, virus, lipid alterations, or carbohydrate metabolism disorders). Most studies in humans observed significantly lower serum and plasma Se levels in patients with different degrees of hepatocellular injury than in healthy controls [8-13]. This difference was especially marked when the liver injury was due to the intake of toxic substances such as alcohol [5,14]. One group [15] reported that Se supplementation in cirrhotic individuals with a background of nutritional element deficiency did not restore hepatic glutathione peroxidase or Se to normal levels. As noted above, levels of nutritional Se biomarker are generally significantly reduced in patients with liver disease [13], but it remains unclear whether this low Se status is a preexisting factor. Selenium has also been found to have an antagonistic effect on fluoride-induced oxidative stress in human hepatocytes [16] and a hepatoprotective effect in experimental thioacetamide-induced liver cirrhosis [17]. Other researchers observed reduced Se levels in carbon tetrachloride–induced liver cirrhosis [18]. It has been proposed that selenium protects cell components from damage by reactive oxygen species through selenium-containing proteins, such as GSH-Px-3 [6,13]. A previous study reported a significant decrease in mean total serum Se levels in patients with liver disease [12] and indicated a significant decrease in these levels with greater intensity of liver damage, but only 12 cirrhotic patients were included in the study [12]. There remains a need to accurately establish the role of Se protection in the development of cirrhosis, examining the influence of the impairment of cirrhosis on serum Se levels in these patients. With the above background, we hypothesized that the antioxidant defense exerted by Se is impaired in cirrhotic patients and that the disease severity index has a negative influence on serum Se levels. Hence, the first objective of this case-control study was to assess the diminishment of serum Se level in cirrhosis. A second objective was to test the influence of the cirrhosis severity index on the cirrhosis impairs Se status of patients. With these objectives, blood samples from 30 healthy controls and 93 cirrhotic patients were analyzed for total serum Se by hydride generation atomic absorption spectrometry. The Child-Pugh index (CPI) score was used to evaluate the severity of liver disease The aim of this study was to increase understanding of the involvement of this antioxidant trace element in oxidative stress-related diseases
575
such as cirrhosis, as well as in healthy nutrition. Few data have been published to date on the effect of the severity of liver disease on Se metabolism. 2. Methods and materials 2.1. Subjects The initial study population comprised patients under follow-up by the Hepatology Department of our reference hospital. The diagnosis of cirrhosis was based on clinical and ultrasound criteria. Healthy controls were enrolled from among blood donors in the same geographic area. All subjects were fully informed of the aims of the study, and an informed consent was obtained from each one before blood was drawn. Exclusion criteria for study participation were the presence or history of non–insulin-dependent diabetes mellitus, wide bowel resection, gastrectomy, inflammatory bowel disease, cancer, chronic renal failure, active alcoholism, corticoid therapy, or hypercortisolism and the receipt of drugs that interfere with muscle function. In addition, subjects under dietary supplements or drugs with a known antioxidant capacity were excluded. Serum Se levels found in these patients were compared with those measured in 30 healthy controls from the same area. The patients were divided into 2 groups according to their CPI score for severity of the liver disease: CPI A (patients with a well-balanced disease; between 5 and 6 points) and B (patients with a significant functional compromise; between 7 and 9 points) plus C (patients with an unbalanced disease; between 10 and 15 points). Therefore, the CPI A and B plus C corresponded to low and medium plus high severity, respectively [19,20]. The final CPI score was obtained as a sum of the assigned points by the physician to the ascites degree, bilirubin and albumin plasma concentrations, prothrombin time, and the encephalopathy degree. This CPI score was a model initially proposed with the aim to stratify the surgical risk in patients with portal blood imbalance [19] and later modified by others [20] when the status nutritional parameter was changed by the prothrombin time [21]. The patients were also divided into 3 groups according to age (b30, 31-50, and N50 years) to have a balanced number of patients in each group. The institutional approval for using human subjects in the study was obtained from the Virgen de las Nieves University Hospital. The study, including the release of human serum samples, was approved by the ethics committee of the hospital and was conducted in accordance with the Helsinki Declaration. 2.2. Blood samples Blood was drawn from the antecubital vein, with the subject seated after overnight fasting. Samples were left to coagulate spontaneously and then centrifuged at 3000g to obtain serum separated by gelose. Aliquots (1 mL) of serum
576
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578
were frozen at −25°C for transportation and storage in the laboratory of the Department of Nutrition and Food Science (University of Granada) until analyzed [19]. 2.3. Total Se measurement Sample mineralization (200 μL of serum) was performed using a HNO3/HClO4 mixture heated to 110°C in a thermostated block following a previously optimized procedure [22]. We then added 1.25-mL HClcc and heated to 80°C for 15 minutes to reduce the Se (VI) to Se (IV). When cool, this was diluted to 10 mL with a 0.5% (vol/vol) HCl solution. This 10-mL aliquot was then transferred to a reaction vessel, which was placed in the MHS-10 system. Selenium determination was carried out using the hydride generation atomic absorption spectrometry technique (Perkin-Elmer model 1100B atomic absorption spectrometer equipped with a Perkin-Elmer MHS-10 hydride generator; Perkin-Elmer, Norwalk, CT, USA). Hydride generation was carried out using a solution of 3% (wt/vol) NaBH4 in 1% (wt/vol) NaOH. All the samples were analyzed in triplicate. The absorbance measured in each sample (peak height mode) was correlated with the Se concentration by the addition calibration method. The accuracy (100.35%) and precision (5.31%) of the method were also tested with the reference material 0148 (Contox trace metal serum control Panel C from the Kaulson Laboratories Inc, West Caldwell, NY). The Se concentrations obtained for the reference material by the technique used (1.928 ± 0.110 μmol/L) were not significantly different (P N .05) from certified levels (1.900 ± 0.076 μmol/L). 2.4. Statistical analyses Statistical analyses were performed with SPSS software (version 15.0 for Windows program; SPSS, Chicago, Ill). The Kolmogorov-Smirnov test was used to examine the normal distribution of variables and the Levene's test to study the homogeneity of variances. Data were compared by applying the Student t test for parametric variables and the Kruskal-Wallis test for nonparametric variables. A nonparametric distribution was found for serum Se, age, and cirrhosis severity variables (P values by KolmogorovSmirnov test were .014, .003, and .004, respectively). Correlation analyses using the Spearman coefficient were also applied for assessing linear relationships between the evaluated parameters. For the relationships between selenium and age and between patients' age and CPI score, a simple linear model was chosen because of its explanatory power and because it does not differ significantly from any other models. Serum Se and cirrhosis severity results are expressed as arithmetic mean ± SD. A 3-way variance analysis to discern the relative or combined roles of the 3 factors (sex, age, and severity of cirrhosis) on serum Se levels of the patients was performed. In this case, the values were presented as medians ± SE. Results were considered statistically significant at P b .05.
3. Results and discussion The final patient sample comprised 93 patients (39 females and 54 males) with a diagnosis of cirrhosis; their mean (SD) age was 41.6 (16.6) years. The control group comprised 30 healthy individuals (13 females and 17 males) with a mean (SD) age of 40.7 (18.8) years. Serum Se concentrations were significantly higher (P b .05) in the oldest patient group (N50 years) compared with the younger age groups (b30 and 31-50 years; Table 1). In addition, serum Se levels showed a linear correlation with age (r = 0.277, P = .007) in cirrhotic patients, similar to previous reports in healthy individuals [23]. The same general tendency was also observed by other authors in several countries, but conclusions about age dependency are indeed questionable due to the fact that the group studied is sometimes poorly defined (ie, the age-range too small or too large) [24]. Serum Se concentrations were previously found to be significantly higher in the oldest age group in a study of healthy primary school children from northeast Thailand [25]. In the present study, higher patient age was inversely correlated with the severity (index) of the disease (P b .001, r = −0.456), with younger cirrhotics having more severe disease. This is not a common finding in alcoholic cirrhotics, reflecting the distinct etiology of the disease in our patients. Patients with active alcoholism were not included in the study. One group reported a nonsignificant reduction in serum Se levels with higher age in patients with liver disease, commenting that published findings on this issue have been controversial [12]. Mean serum Se concentrations of the patient and control groups are shown in Table 1. Serum Se levels were significantly lower (P b .001) in cirrhotic patients than in healthy controls, as previously reported in patients with liver disease [5,12,13,15,26-28]. One group who found significantly reduced levels of antioxidants, including Se, in alcoholic cirrhotics (vs sex-matched controls) described a relationship between cirrhosis and oxidative stress. Other researchers [26] who also found a significantly lower Se levels in cirrhotics (vs noncirrhotics and controls) reported that an unbalanced nutritional status was the strongest predictor of low Se values in alcoholics.
Table 1 Serum Se levels in cirrhosis patients by age group and in healthy controls Individuals
n
Total Se (μmol/L) Mean ± SD
Cirrhosis patients (y) b30 31-50 N50 Healthy controls
93 23 38 32 30
a
0.721 ± 0.239 0.662 b ± 0.236 0.680 b ± 0.198 0.811 b ± 0.268 0.926 a ± 0.241
Range 0.157-1.311 0.248-1.258 0.157-1.311 0.157-1.311 0.488-1.369
a P b .001. Mean total serum Se level measured in patients was significantly lower. b P b .012. Mean total serum Se level measured in patients older than 50 years was significantly higher.
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578 Table 2 Serum Se levels in cirrhosis patients by sex Sex
n
Total Se (μmol/L) a
Female Male All
39 54 93
Mean ± SD
Range
0.665 ± 0.228 0.761 ± 0.243 0.721 ± 0.239
0.157-1.311 0.157-1.311 0.157-1.311
a
P = .047. Mean total serum Se level measured in male was significantly higher.
The Se plays an important antioxidant role in various biochemical reactions as a cofactor of GSH-Px [25-29]. This enzyme acts against the onset and progression of events producing liver injury by protecting cells against free radical formation from hydroperoxide ions. A recent study found that Se supplementation to cirrhotic patients did not restore their serum Se and glutathione peroxidase to normal levels [15]; this finding that could be related with the fact that plasma Se levels take months to return to normal in proteinenergy undernourishment, after nutritional rehabilitation. However, there have been few human trials on the prophylactic value of Se against the processes implicated in cirrhosis. A study in rats [30] showed that Se may play an important role in preventing the induction of hepatic cellular injury by carbon tetrachloride. Other researchers [31] found that the Se supplementation increased liver expression of manganese superoxide dismutase in rats, improving their antioxidant defense, thus supporting the anti-inflammatory and hepatoprotective effects claimed for Se. Table 2 shows the mean serum Se concentrations in cirrhotic patients by sex. Concentrations were significantly higher in the males than in the females (P b .05), similar to previous reports even in healthy individuals [23]. The severity index was significantly higher in the female vs male cirrhotic patients (CPI: 7.81 ± 1.71 vs 6.28 ± 1.31, respectively; P b .001). As shown in Table 3, serum Se levels of the patients were not significantly correlated with the CPI-determined severity of their disease (P N .05). Therefore, as the CPI can take values from 5 to 15, this regression analysis is semicontinuous. This finding is not in agreement with previous reports of an association between higher severity of liver disease and greater impairment of body Se status [13,32]. In contrast, another study [33] concluded that decreased serum Se levels result from liver function Table 3 Correlation of serum Se levels in cirrhosis patients with severity of liver disease (CPI) CPI
n
Child-Pugh A Child-Pugh B + C a
42 51
Total Se (μmol/L) Mean a ± SD
Range
0.755 ± 0.255 0.691 ± 0.229
0.157-1.311 0.248-1.311
P = .144. No significant difference was found.
577
impairment of any etiology. Taking the present results into account, dietary or prescribed Se supplementation for cirrhotic patients should not be ruled out. We performed a 3-way variance analysis to discern the relative and combined roles of the 3 factors (sex [male and female], age [b30, 31-50, and N50 years], and severity of cirrhosis [CPI A vs B + C]) on the serum Se levels of the patients, finding no interaction effect of these factors on these levels (P N .05). The severity of cirrhosis was not related to serum Se levels (P = .763), but sex (difference between medians, 7.66; SE, 3.86) and overall age were independently related to these levels (P b .05). Thus, serum Se concentrations significantly differed between the oldest patient group (N50 years) and the younger patients (b30 years: difference between medians, 11.9; SE, 4.94; 31-50 years: difference between medians, 11.6; SE, 4.8). In our opinion, the main limitation of the study was the number of studied cirrhosis patients; in fact, for the CPI C group, only 1 patient was considered. Although serum Se levels were not significantly influenced by the severity index of the liver disease (consequently, we reject the main hypothesis for the research), future studies should be performed with a higher, more balanced number of cirrhosis patients for the CPI A, B, and C groups to elucidate definitively the hypothesis. In conclusion, serum Se levels were significantly lower in cirrhotic patients than in healthy controls but were not significantly influenced by the CPI-determined severity of liver disease. Serum Se concentrations were significantly higher in male vs female cirrhotics, whereas the severity index was significantly higher in the female vs male cirrhotics. Serum Se levels were significantly lower in younger vs older patients. Our results demonstrate that the Se component of the antioxidant system is severely impaired in cirrhosis.
Acknowledgment This study was partially supported by a grant: AGR-109 (Junta de Andalucía, Spain).
References [1] Rhead WJ, Evans GA, Schrauzer GN. Selenium in plasma: level sin blood proteins and behavior upon dialysis, acidification, and reduction. Bioinorg Chem 1974;3:217-23. [2] Burk RF, Early DS, Hill KE, Palmer IS, Boeglin ME. Plasma selenium in patients with cirrhosis. Hepatology 1998;27:794-8. [3] Speckmann B, Walter PL, Alili L, Reihner R, Sies H, Klotz LO, et al. Selenoprotein P expressions is controlled through interaction of the coactivator PGC-1α with fox01a and hepatocyte nuclear factor 4α transcription factors. Hepatology 2008;48:1998-2006. [4] Mas A, Sarkar B. Role of glutathione in selenite binding by human plasma. Biol Trace Elem Res 1988;20:95-104. [5] Burk RF, Hill KE. Selenoprotein P: an extracellular protein with unique physical characteristics and a role in selenium homeostasis. Ann Rev Nutr 2005;25:215-35.
578
M. Martínez-Peinado et al. / Nutrition Research 30 (2010) 574–578
[6] Koyama H, Abdulah R, Ohkubo T, Imai Y, Satoh H, Nagai K. Depressed serum selenoprotein P: possible new predicator of increased risk for cerebrovascular events. Nutr Res 2009;29:94-9. [7] Morisco F, Vitaglione P, Amoruso D, Russo B, Fogliano V, Caparoso N. Foods and liver health. Mol Aspects Med 2008;29: 144-50. [8] Lecomte E, Herberth B, Pirrollet P, Chancerelle Y, Arnaud J, Musse N, et al. Effect of alcohol consumption on blood antioxidant nutrients and oxidative stress indicators. Am J Clin Nutr 1994;60: 255-61. [9] Loguercio C, De Girolamo V, Federico A, Feng SL, Cataldi V, Del Vecchio Blanco C, et al. Trace elements and chronic liver diseases. J Trace Elem Med Biol 1997;11:158-61. [10] Guarini P, Stanzial AM, Olivieri O, Casaril M, Galvani S, Pantalena M, et al. Erythrocyte membrane lipids and serum selenium in post-viral and alcoholic cirrhosis. Clin Chim Acta 1998;270:139-50. [11] Navarro-Alarcón M, López-Martínez MC. Essentiality of selenium in the human body: relationship with different diseases. Sci Total Environ 2000;249:347-71. [12] Navarro-Alarcon M, Lopez-Ga de la Serrana H, Perez-Valero V, Lopez Martinez MC. Selenium concentrations in serum of individuals with liver diseases (cirrhosis or hepatitis): relationship with some nutritional and biochemical markers. Sci Total Environ 2002;291:135-41. [13] Navarro-Alarcon M, Cabrera-Vique. Selenium in food and the human body: a review. Sci Total Environ 2008;400:115-41. [14] Thuluvath PJ, Triger DR. Selenium in chronic liver disease. J Hepathol 1992;14:176-82. [15] Zhang J, Wang H, Yu H. Thioacetamide-induced cirrhosis in seleniumadequate mice displays rapid and persistent abnormity of hepatic selenoenzymes which are mute to selenium supplementation. University of Science and Technology of China, Southern Campus; 2007. [16] Wang A, Xia T, Yuan J, Chen X, Wang KY. Antagonistic effect of selenium on oxidative stress, DNA damage, and apoptosis induced by fluoride in human hepatocytes. Fluoride 2004;37:107-16. [17] Al-Bader A, Abul H, Hussain T, Al-Moosawi M, Mathew TC, Dashti H. Selenium and liver cirrhosis. Mol Cel Biochem 1998;185:1-6. [18] González-Reimers E, López-Lirola A, Martín Olivera R, SantolariaFernández F, Galindo-Martín L, Abreu-González P, et al. Effects of protein deficiency on liver trace elements and antioxidant activity in carbon tetrachloride-induced liver cirrhosis. Biol Trace Elem Res 2003;93:127-39. [19] Child GG, Turcotte JG. Surgery and portal hypertension. In: Child GG, editor. The liver and portal hypertension. Philadelphia: Saunders; 1964. p. 50-2.
[20] Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-9. [21] Meijide Miguez HM. Child Pugh. Clasificación pronóstica de la hepatopatía. http://www.meiga.info/Escalas/CHILD-PUGH.pdf. [22] Navarro M, Lopez H, Ruiz ML, Gonzalez S, Perez V, Lopez MC. Determination of selenium in serum by hydride generation atomic absorption spectrometry for calculation of daily dietary intake. Sci Total Environ 1995;175:245-52. [23] Lopes PA, Santos MC, Vicente L, Rodrigues MO, Pavao ML, Neve J, et al. Trace element status (Se, Cu Zn) in healthy Portuguese subjects from Lisbon population. Biol Trace Elem Res 2004;101:1-17. [24] Robberecht H, Deelstra H. Factors influencing blood selenium concentration values—a literature review. J Trace Elem Electrolytes Health Dis 1994;8:129-43. [25] Krittaphol W, Bailey KB, Pongcharoen T, Winichagoon P, Thomson C, Gibson RS. Primary school children from northeast Thailand are not at risk of selenium deficiency. Asia Pac J Clin Nutr 2006;15:474-81. [26] Gonzalez-Reimers E, Galindo-Martin L, Santolaria-Fernandez F, Sanchez-Perez MJ, Alvisa-Negrin J, Garcia-Valdecasas- Campelo EG, et al. Prognostic value of serum selenium levels in alcoholics. Biol Trace Elem Res 2008;125:22-9. [27] Rai M, Pal MR, Sabihta P, Pai P. Hepatoprotective influence of selenium in experimental liver cirrhosis. Kuwait Med J 2001;33:333-6. [28] Pemberton PW, Smith A, Warnes TW. Non-invasive monitoring of oxidant stress in alcoholic liver disease. Scand J Gastroenterol 2005; 40:1102-8. [29] Zulet MA, Puchau B, Hermsdorff HHM, Navarro C, Martínez JA. Dietary selenium is negatively associated with serum sialic acid and metabolic syndrome features in healthy young adults. Nutr Res 2009; 29:41-8. [30] Ozardali I, Bitiren M, Karakilcik AZ, Zerin M, Aksoy N, Musa D. Effects of selenium on histopathological and enzymatic changes in experimental liver injury of rats. ExpToxicol Pathol 2004;56:59-64. [31] Shilo S, Pardo M, Aharon-Simon M. Selenium supplementation increases liver MnSOD expression: molecular mechanism for hepatoprotection. J Inorg Biochem 2008;102:110-8. [32] Valimaki MJ, Harju KJ, Ylikahri RH. Decreased serum selenium in alcoholics, a consequence of liver dysfunction. Br J Nutr 1986;55: 227-33. [33] Lin CC, Huang JF, Tsai LY, Huang YL. Selenium, iron, copper, and zinc levels and copper-to zinc ratios in serum of patients at different stages of viral hepatic diseases. Biol Trace Elem Res 2006;109: 15-23.