Lipid Peroxidation and Etheno DNA Adducts in White Blood Cells of Liver Fluke-Infected Patients: Protection by Plasma α-Tocopherol and Praziquantel

Infection by Opisthorchis viverrini, a strong risk factor for cholangiocarcinoma, probably causes cancer through growth stimuli produced by chronic inflammation. Studies from our group (1, 2) and others (3-5) have provided evidence that persistent oxidative/nitrative stress and excess lipid peroxidation are induced by chronic inflammatory processes that lead to massive DNA damage in target organs and represent an important step in carcinogenesis. The oxidation of lipids by reactive oxygen and nitrogen species results in byproducts such as trans-4-hydroxy-2-nonenal, malondialdehyde, and crotonaldehyde. These intermediates can react with DNA bases to form exocyclic DNA adducts (6-8). Reaction of DNA bases with the major lipid peroxidation product trans-4-hydroxy-2-nonenal yields inter alia the etheno-DNA adducts 1,N⁶-etheno-2′-deoxyadenosine (ϵdA) and 3,N⁴-etheno-2′-deoxycytidine (ϵdC). These etheno-modified DNA bases are highly miscoding lesions, and are thought to initiate the carcinogenic process through specific point mutations. ϵdA and ϵdC, which are also produced in DNA by the human carcinogen vinyl chloride, induce specific base pair substitution mutations in mammalian cells (9, 10). The high frequency of mutations and the deregulation of cell homeostasis are events thought to play an important role in human chronic disease pathogenesis (11).

Our previous work provided evidence that etheno DNA adducts may serve as potential risk markers for cancer-prone diseases caused by viral infections, chronic inflammatory processes such as chronic pancreatitis, inflammatory bowel diseases, and thalassemia-related metal storage (12-14).

Previously we reported that excretion of ϵdA and ϵdC in urine was highly increased in O. viverrini–infected Thai patients (15). We proposed that chronic O. viverrini infection leads through oxidative/nitrative stress to a massive urinary excretion of etheno-bridged deoxyribonucleosides, reflecting high lipid peroxidation–derived DNA damage in affected internal organs. These promutagenic etheno adducts, together with other lesions in the DNA of bile duct epithelial cells, may increase the risk of O. viverrini–infected patients developing cholangiocarcinoma later in life.

There is supportive evidence for a mechanistic link between oxidative stress and lipid peroxidation–induced DNA damage and protection by cancer chemopreventive dietary components (16). A phase II chemoprevention trial using the oxidized DNA base 5-hydroxymethyl-2′-deoxyuridine as a damage marker, showed an inverse relationship with plasma α-tocopherol levels in female volunteers after a short-term dietary vitamin E supplementation (17). A pilot study in healthy women revealed an inverse correlation of ϵdA levels in WBC DNA with vitamin E consumption, calculated from a food frequency questionnaire (18). Previous studies suggest that α-tocopherols (vitamin E compounds) may be inversely associated with cancer risk (19, 20). Therefore, the lipid soluble α-tocopherol seems to be effective for trapping lipid-derived free radicals in membranes and by thus inhibiting lipid peroxidation–mediated adverse reactions.

The major aim of this study was to clarify whether chronic O. viverrini infection can induce lipid peroxidation–derived DNA damage in infested patients as a consequence of persistent oxidative/nitrative stress generated in vivo. As damage marker we measured by our ultrasensitive immunoaffinity/³²P-postlabeling method two lipid peroxidation–derived etheno adducts (ϵdA, ϵdC) in WBC samples. We hypothesized that when WBC circulates through the infected hepatobiliary tract DNA damage may occur at the site of O. viverrini infection (Fig. 1). Another aim was to verify whether a single dose of praziquantel, a known effective antiparasitic drug when administered to O. viverrini–infected patients, could lower ϵdA and ϵdC levels in WBC, concomitantly with those of 8-isoprostane, malondialdehyde, and nitrate/nitrite measured in plasma as inflammatory indicators. The potential of plasma α-tocopherol to lower oxidative stress and lipid peroxidation–induced DNA damage in O. viverrini–infected patients was also investigated.

The study was done in healthy volunteers (n = 20, 9 male and 11 female) and O. viverrini–infected patients (n = 50, 26 male and 24 femal). Healthy noninfected subjects (range and mean age ± SE, 35-59 and 47.6 ± 1.7 y) were recruited as persons who lived in a nonendemic area of O. viverrini infection and showed negative results (both at present and in the past) for O. viverrini egg counts in feces [range and mean egg count per gram of feces (epg) ± SE, 15-1,091 and 283.81 ± 52.34 epg). O. viverrini–infected patients (range and mean age ± SE, 28-78 and 53.0 ± 2.0 y) were persons who lived in an endemic area in Northeast Thailand (Phuvieng District near Ubonrat Dam, Khon Kaen Province) and had positive O. viverrini egg counts in feces. All individuals positive for O. viverrini infection were treated with a single dose of praziquantel (40 mg/body weight). There were no positive O. viverrini egg counts in feces of all subjects after 2 mo of follow-up. Other chronic inflammatory processes caused by infection and chronic diseases may influence the levels of etheno DNA adducts. Therefore, persons with acute (non–O. viverrini) infections showing nitrate/nitrite and positive leukocyte counts in urine were excluded; patients with chronic inflammatory conditions caused by hepatitis B virus and tuberculosis infections and those with diabetic mellitus were also excluded. An informed consent form was signed by each study subject. The study protocol was approved by the Ethics Group of the Human Research Committee (HE 480316), Khon Kaen University, Thailand.

Peripheral 10 mL blood samples were obtained by sterilized venipuncture, collected in tubes containing acid citrate/dextrose and centrifuged at 2,000 rpm at 4°C for 15 min; plasma and buffy coat were then collected and stored at −80°C until used. The isolation of genomic DNA from buffy coats was achieved using the blood and cell culture Midi kit (Qiagen) according to the manufacturer's protocol, except with the modification of adjusting the pH to 7.4 and the NaCl concentration in the elution buffer to 1.4 mol/L.

ϵdA and ϵdC were analyzed in DNA by an immunoaffinity/³²P-postlabeling method (21). In brief, ∼25 μg of DNA were hydrolyzed to nucleotide 3′-monophosphates using micrococcal nuclease and spleen phosphodiesterase. Normal nucleotides were quantitated by high performance liquid chromatography, and the adducts were enriched on immunoaffinity columns prepared from the monoclonal antibodies EM-A-1 (ϵdA) and EM-C-1 (ϵdC). The antibodies used in this study were provided by Dr. Manfred Rajewsky (Institute of Cell Biology, University of Essen, Essen, Germany). The adducts and the internal standard deoxyuridine 3′-monophosphate was labeled with [γ-³²P]ATP (10 μCi; Amersham Buchler) and T4 polynucleotide kinase (Fermentus). The adducts were resolved on polyethyleneimine-TLC plates using two-directional chromatography [D1 = 1 mol/L acetic acid (pH 3.5); D2 = saturated ammonium sulphate (pH 3.5)]. After autoradiogram, the adduct spots and the internal standard were marked and cut, and the radioactivity was measured in a liquid scintillation counter. The absolute adduct levels were quantitated using standards, and the relative adduct level per parent nucleotides was determined with the amount of deoxycytidine (dC) and deoxyadenosine (dA) obtained from high performance liquid chromatography analysis as described (21).

Plasma α-tocopherol was measured by high performance liquid chromatography (22). Plasma (120 μL) was extracted with 120 μL of 0.1% butylated hydroxytoluene in ethanol containing 10 μmol/L α-tocopherol acetate (Sigma-Aldrich) as an internal standard. To extract α-tocopherol, 120 μL of 10 mmol/L SDS and 3.0 mL of n-heptane were added and 2.8 mL of n-heptane supernatant containing α-tocopherol were transfered into a new tube. The supernatant was concentrated by evaporation under N₂, at 45°C. The dried residue was reconstituted with 120 μL of the mobile phase (methanol:acetonitrile:methylene chloride = 200:200:50 v/v/v) and injected into a high performance liquid chromatography device (Waters 2487) for at a flow rate of 1 mL/min.

Nitrate/nitrite was determined by simple Griess reaction with some modifications (23, 24). Nitrate concentration in plasma and standard nitrate were measured after reducing nitrate into nitrite using the catalyst, by vanadium three chloride (VCl₃; Sigma-Aldrich). Briefly, a saturated VCl₃ solution was prepared from 400 mg of VCl₃ in 50 mL of 1 mol/L HCl and stored at 4°C for <2 wk. Plasma (100 μL) was deproteinized by adding 300 μL of methanol:diethyl ether (3:1 v/v), followed by incubation at −80°C for 30 min and centrifuging at 10,000× g for 10 min at 4°C. Of the supernatant, 100 μL were pipetted into a new micro centrifuge tube for nitrate/nitrite determination. The 100 μL of each supernatant, 50 μL of N-(1-naphthyl) ethylenediamine dihydrochloride (0.1% w/v in water), and sulfanilamide (2% w/v in 5 % HCl) were thoroughly mixed and allowed to react for 20 min at 37°C. The absorbance was read against a reagent blank at 540 nm in a spectrophotometer (Spectra max GEMINI XPS, Molecular Devices). The sodium nitrate was used as a standard at various concentrations ranging from 6.25 to 100 μmol/L. Pooled plasma from a blood bank was used for testing within-run and between-run assay precision. A coefficient of variability of <10% was acceptable.

Thiobarbituric acid-reactive substances (TBARS) were measured (as reaction equivalent to malondialdehyde) with a slightly modified method by Nowak et al. (25). Briefly, 200 μL of plasma and 8.1% SDS, 1.5 mL of 0.5 mol/L HCl, 1.5 mL of 20 mmol/L thiobarbituric acid, 50 μL of 7.2% butylated hydroxytoluene in 95% ethanol, and 550 μL of deionized water were added to a screw-capped test tube. The sample was stirred and heated by boiling in a heater (90°C) for 1 h. After immediately cooling for 10 min in an ice box, the chromogen was extracted by adding 1 mL of deionized water and 5 mL of n-butanol in pyridine (15:1, v/v). The sample was then mixed thoroughly and centrifuged at 3,000× g for 15 min. The reaction product of TBARS with malondialdehyde (TBARS-malondialdehyde) was determined against reagent blank in a spectrofluorometer (Sunrise Tecan) with 520 nm excitation and 550 nm emission. The 1,1,3,3-tetramethoxypropane (Sigma-Aldrich) was used as standard. Pooled plasma from a blood bank was used for testing within-run and between-run assay precision. A coefficient of variability of <10% was acceptable.

Plasma 8-isoprostane was measured by the ELISA-based method, using a commercial kit (Cayman Chemical) according to the supplier's protocol.

Statistical analysis of the data was done with the StatView 5.0 program (SAS Institute). The comparison between O. viverrini–infected and healthy control groups was done by the Mann Whitney U-test; O. viverrini–infected groups before and after treatment with praziquantel were compared by Wilcoxon's rank test; the correlations between parameters were determined by using Spearman's rank test.

To clarify the role of oxidative stress and lipid peroxidation–derived DNA damage in the pathogenesis of opisthorchiasis-associated cholangiocarcinoma, the accumulation of two lipid peroxidation–derived etheno DNA adducts (ϵdA, ϵdC) in WBC was measured in 20 healthy volunteers and 50 O. viverrini–infected patients. WBC circulates through the infected hepatobiliary tract and damage of DNA may occur at the site of O. viverrini infection (Fig. 1). ϵdA and ϵdC were quantified in WBC DNA by our ultrasensitive and specific immunoaffinity/³²P-postlabeling method.

Figure 2A and B shows box-plots and mean levels of ϵdA and ϵdC in WBC DNA of O. viverrini–infected patients as compared with healthy subjects. The means of ϵdA and ϵdC levels (expressed as ϵdA/10⁸dA and ϵdC/10⁸dC, respectively) were significantly three to five times higher in O. viverrini–infected patients than in healthy controls (ϵdA, 5.66 ± 0.89 versus 1.81 ± 0.43; P = 0.0002; ϵdC, 27.85 ± 7.61 versus 5.40 ± 1.02; P = 0.0001).

Several inflammatory indicators were also measured in O. viverrini–infected patients and compared with those found in healthy volunteers. Figure 2C to E shows box-plots and mean levels of plasma 8-isoprostane, malondialdehyde, and nitrate/nitrite in patients versus healthy subjects. Similarly to ϵdA and ϵdC levels in WBC, all above inflammatory indicators in plasma were concomitantly increased by O. viverrini infection: 8-isoprostane levels (pg/mL), 59.60 ± 4.35 versus 23.70 ± 4.25 (P < 0.0001); malondialdehyde (μmol/L), 0.66 ± 0.22 versus 0.58 ± 0.15 (not significant, P = 0.45); and nitrate/nitrite (μmol/L), 26.56 ± 2.24 versus 13.90 ± 1.33 (P = 0.0001). Plasma C-reactive protein levels did not differ between O. viverrini–infected (1.33 ± 0.24 μg/dL) and healthy subjects (1.42 ± 0.55 μg/dL).

Figure 2F shows box-plots and mean levels of plasma α-tocopherol in O. viverrini–infected patients as compared with healthy subjects. α-Tocopherol levels (μg/mL) were significantly lower in O. viverrini–infected patients than in healthy controls (7.62 ± 0.35 versus 10.12 ± 0.50; P = 0.0004).

Table 1 compares ϵdA and ϵdC levels in WBC of 8-isoprostane, malondialdehyde, and nitrate/nitrite levels in plasma of O. viverrini–infected patients, before and 2 months after treatment by a single dose (40 mg/kg body weight) of the antiparasitic drug praziquantel. ϵdA and ϵdC levels were significantly two times lowered by O. viverrini elimination (ϵdA, 5.66.70 ± 0.89 adducts/10⁸ dA versus 3.29 ± 0.68; P < 0.02; ϵdC, 27.85 ± 7.61 adducts/10⁸ dC versus 15.60 ± 3.90; P < 0.03). Remarkably, after antiparasitic drug treatment adduct levels were not statistically different from those of healthy controls (ϵdA, 1.81 ± 0.43 adducts/10⁸ dA; P = 0.058; ϵdC, 5.40 ± 1.02 adducts/10⁸ dC; P = 0.056). Praziquantel treatment significantly decreased plasma 8-isoprostane (59.60 ± 4.35 versus 50.60 ± 4.43 pg/mL; P = 0.01) but its concentration still remained higher than in healthy controls (23.70 ± 4.25; P < 0.0001). Praziquantel treatment of O. viverrini–infected patients significantly decreased plasma levels of malondialdehyde (0.66 ± 0.04 versus 0.45 ± 0.02 μmol/L; P < 0.0001) and of plasma nitrate/nitrite (26.56 ± 2.24 versus 17.09 ± 1.20 μmol/L; P = 0.001). Plasma malondialdehyde (P = 0.0002) remained at a lower level than in controls. Treatment of patients by praziquantel led only to a marginal increase in plasma α-tocopherol (7.61 ± 0.35 versus 7.27 ± 0.31 μg/mL; P = 0.06) being still significantly lower than in healthy controls (10.12 ± 0.50; P < 0.0001).

Table 2 lists the correlations among parameters measured in plasma and WBC DNA before and after treatment of O. viverrini–infected patients with praziquantel. To increase statistical power data were combined with those from untreated healthy controls. Plasma ALP activity has already been reported in our previous study (15) but data were included for correlation analyses. There was a significant positive correlation between ϵdA and ϵdC levels in WBC (r = 0.631; P < 0.0001). ϵdA and ϵdC levels in WBC were positively and significantly correlated with 8-isoprostane, malondialdehyde, and nitrate/nitrite levels (except ϵdC) and with ALP activity in plasma. Positive and significant correlations were found for (a) 8-isoprostane versus malondialdehyde, versus nitrate/nitrite levels, and versus ALP activity; (b) malondialdehyde levels versus plasma nitrate/nitrite levels and versus ALP activity; and (c) nitrate/nitrite levels versus ALP activity. Plasma α-tocopherol levels were inversely and significantly correlated with 8-isoprostane and nitrate/nitrite levels and with ALP activity. Inverse correlations of plasma α-tocopherol levels with WBC ϵdA and with WBC ϵdC levels were observed but did not reach statistical significance. For all investigated parameters, neither gender- nor age-related differences were apparent; egg counts in feces and plasma C-reactive protein were also not correlated with any parameters analyzed (data not shown). The data on alcohol consumption and smoking behavior were not available in this study.

Comparing data with our previous study (15), there was a significant, positive correlation between plasma malondialdehyde and urinary malondialdehyde levels (r = 0.842; P < 0.001). The plasma nitrate/nitrite levels significantly correlated with urinary levels of nitrate/nitrite (r = 0.702; P < 0.001). The WBC ϵdC levels correlated significantly but negatively with urinary ϵdC levels (r = −0.248; P = 0.04). There was a positive correlation between the levels of WBC ϵdA and urinary ϵdA levels, but this did not reach statistical significance (r = 0.118; P = 0.17).

Both epidemiologic and experimental evidence implicate liver fluke (O. viverrini) infection, which is endemic in the region of this study, as a major carcinogenic risk factor for cholangiocarcinoma (relative risk ∼5; refs. 26, 27). The main histopathologic features of opisthorchiasis both in man and in animal models are inflammatory processes causing persistent oxidative/nitrative stress and excess lipid peroxidation (28-30). We have reported a high excretion of two lipid peroxidation–derived etheno DNA adducts (ϵdA, ϵdC) in the urine of O. viverrini–infected Thai subjects, likely reflecting high DNA damage in affected internal tissues, i.e., the hepatobiliary tract (15). As adduct levels in urine decreased significantly after a single dose of the antiparasitic and cholangiocarcinoma-protective drug praziquantel we proposed that these adducts could be explored as noninvasive risk markers for developing opisthorchiasis-related cholangiocarcinoma.

In the present study, we measured the steady-state levels of ϵdA and ϵdC in DNA of WBC samples by our ultrasensitive detection method. Using this procedure, the existence of variable background levels of adducts in liver and other tissues from unexposed rodents and humans could be unambiguously and quantitatively revealed (31). This background likely reflects the physiologic level of lipid peroxidation–caused DNA damage from normal endogenous processes. Although our method is time consuming and labor intensive, its high sensitivity and specificity allows to reliably determine any disease-related increase in adduct levels in human WBC and tissues, requiring only a few micrograms of DNA sample for biomonitoring.

WBC from healthy volunteers and O. viverrini–infested patients before and after treatment by the antiparasitic drug praziquantel were analyzed. We hypothesized that when WBC circulate through the infected hepatobiliary tract, DNA damage may occur at the site of O. viverrini infection (Fig. 1). Our results indeed showed a significant increase of etheno adduct levels in WBC of O. viverrini–infected patients when compared with healthy, noninfected subjects. Together with the observed significant adduct reduction after treatment with the antiparasitic drug praziquantel, our results provide strong support that O. viverrini–related infection and inflammation are key events for causing lipid peroxidation–derived DNA damage in vivo, which may play an important role in opisthorchiasis-associated cholangiocarcinoma. This assumption is further supported by previous results showing that O. viverrini infection in hamsters also induced other types of DNA damage, i.e., 8-nitroguanine and 8-oxo-deoxyguanosine in inflammatory and bile duct epithelial cells (32). In patients recruited from the same area as in this study, urinary levels of 8-oxodG, another oxidative damage marker were found to be significantly higher in cholangiocarcinoma patients than in O. viverrini–infected and healthy subjects, and higher in O. viverrini–infected subjects than in healthy subjects. Urinary 8-oxodG levels were significantly correlated with 8-oxodG levels in leukocyte DNA (33). Other results in cholangiocarcinoma patients suggested that 8-nitroguanine and 8-oxo-deoxyguanosine formation and reciprocal activation of the hypoxia-inducible factor-1α likely contribute to multiple genetic changes, tumor progression, and invasiveness of cholangiocarcinoma (34).

The antiparasitic and cholangiocarcinoma-protective drug praziquantel prevents O. viverrini–induced cholangiocarcinoma not only by elimination of parasites but also by its anti-inflammatory effect as an inhibitor of inducible nitric oxide synthase-dependent DNA damage (35). From our study in O. viverrini–infected patients we also must assume that the reduction by praziquantel of lipid peroxidation–derived etheno-DNA adducts in WBC and of the inflammatory indicators in plasma, 8-isoprostane, malondialdehyde, and nitrate/nitrite is caused by its suppression of oxidative/nitrative stress, acting together with its known antiparasitic effectiveness (Fig. 1). Three decades ago we were the first to show that praziquantel is devoid of genotoxic activity, unlike other antiparasitic drugs such as hycanthone in use at that time (36).

In WBC of O. viverrini–infected patients, we found a positive correlation between the levels of ϵdA and ϵdC, but in all samples ϵdC levels were 5- to 15-fold higher than those of ϵdA. Alkyl-N-purine-DNA glycosylase (ANPG) and AlkB protein remove ϵdA, the mismatch-specific and AlkB protein remove ϵdC. Previous studies showed that ϵdA adducts in cellular DNA in vitro are repaired more quickly by base excision repair than are ϵdC adducts (37-39). A third repair pathway by epoxidation of the etheno bond by the repair AlkB proteins also occurs at a faster rate for ϵdA than ϵdC (40, 41). Another explanation for the higher ϵdC levels is provided by the hijacking of the human ANPG by ϵdC lesions when present in DNA (42). As a consequence ANPG that normally repairs ϵdA, binds to ϵdC, suggesting that ϵdC cannot be rapidly repaired by thymidine-DNA glycosylase and tends to accumulate.

Several mechanisms by which O. viverrini infection enhances the pathogenesis of cholangiocarcinoma have been proposed by Sripa and Pirojkul (28). Moreover, immunopathologic processes and mechanical injury are thought to mediate hepatobiliary damage in opisthorchiasis (43). Our results in infected patients showing strong positive associations of WBC ϵdA and ϵdC levels with plasma ALP activity are in support of the above proposed mechanism. We conclude that inflammation-induced lipid peroxidation and ensuing DNA damage together with mechanical injury in target hepatobiliary cells act as a driving force to malignancy.

Using etheno-DNA adducts as markers it should be possible to identify effective chemopreventive agents that protect against lipid peroxidation–associated damage. We focused on α-tocopherol as dietary vitamin E intake was already found to lower etheno-DNA adduct levels in WBC of healthy female volunteers (18). Also in a human hepatocellular carcinoma cell line α-tocopherol conferred protection against lipid peroxidation and oxidative DNA damage induced by ionizing radiation (44). Adequate intake of vitamin E and protein prevented an increase of oxidative damage to DNA, lipids, and protein after total body irradiation of mice (45).

In our study, mean plasma α-tocopherol levels were significantly lower in O. viverrini–infected patients than in healthy controls. α-Tocopherol levels were inversely and significantly correlated with plasma 8-isoprostane and nitrate/nitrite levels and with ALP activity. Inverse correlations of plasma α-tocopherol levels with ϵdA and ϵdC levels in WBC were observed, but did not reach statistical significance. Our results provide some support for α-tocopherol as a protective factor, and dietary supplementation may help to reduce lipid peroxidation–induced damage in vivo and possibly also the risk for opisthorchiasis-associated cholangiocarcinoma. Among the parameters analyzed, the levels of plasma 8-isoprostane and α-tocopherol were well correlated and showed a statistical difference from healthy controls, even after praziquantel treatment. Supplementations with α-tocopherol have been shown to reduce the levels of isoprostane both in plasma and urine (46, 47). Therefore, plasma 8-isoprostane levels can be explored as a biomarker for verifying the efficacy of the supplementation.

In conclusion, chronic O. viverrini infection of subjects via inflammation-induced oxidative/nitrative stress leads to accumulation of lipid peroxidation–derived DNA damage in vivo, which may be lowered by dietary vitamin E supplementation. Promutagenic etheno adducts, together with other such DNA lesions in bile duct epithelial cells, most likely increase the risk of O. viverrini–infected patients to later develop cholangiocarcinoma. A relationship between these adduct markers and disease causation is further supported by the protective effect of the antiparasitic drug praziquantel both against DNA damage (this study, refs. 33, 35) and against cholangiocarcinoma (48). Levels of ϵdA and ϵdC in WBC and urine (15) should be explored (a) as putative risk markers for developing opisthorchiasis-related cholangiocarcinoma, (b) as biomarkers to identify disease protective agents, and (c) to assess the efficacy of preventive and therapeutic interventions in at risk populations.

No potential conflicts of interest were disclosed.

The authors thank all patients and volunteers who participated in this study.

Grant Support: Thailand Research Fund through the Royal Golden Jubilee Ph.D. Programme (to S. Dechakhamphu and P. Yongvanit), the Invitation Research Grant (No. i50214) from Faculty of Medicine, Khon Kaen University. Research was in part supported by the Division of Toxicology and Cancer Risk Factors, German Cancer Research Center, Heidelberg and through a grant from Dr. F.P. Armbruster (to S. Dechakhamphu) by Immundiagnostik, Bensheim, Germany (SHeP-GMS-0001).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.