- Email: [email protected]
The ex-vivo intestinal absorption rate of uranium is a two-phase function of supply
Regulatory Toxicology and Pharmacology 69 (2014) 256–262
Contents lists available at ScienceDirect
Regulatory Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/yrtph
The ex-vivo intestinal absorption rate of uranium is a two-phase function of supply Rainer Konietzka ⇑, Rita Heinze, Margarete Seiwert, Hermann H. Dieter Federal Environment Agency, Toxicology of Drinking Water and Swimming Pool Water and Health-related Environmental Monitoring, Wörlitzer Platz 1, 06844 Dessau-Roßlau, Germany
a r t i c l e
i n f o
Article history: Received 25 September 2013 Available online 2 May 2014 Keywords: Uranium Ex vivo Small intestine Absorption
a b s t r a c t The concentration-dependent absorption behaviour of uranium was investigated with surviving intestinal segments of rat jejunums, using an ex-vivo model. The results showed a monotonic slightly nonlinear increase in absorption as uranium concentrations increased. This trend was observed over the entire concentration range tested. In the lower concentration range a slower linear ascent was observed while a steeper linear ascent was found for the higher concentration range. Statistical fit was only slightly poorer for an exponential function in the range of lower values and a logarithmic function in the range of higher values. The proportion of uranium absorbed expressed as percent of uranium concentrations in the perfusion solutions followed a monotonically increasing trend from 20 to around 200 lg/l uranium in the perfusion solutions, which thereafter appears to reach a plateau, as further increase towards concentrations around 400 lg/l is not substantial. The uranium concentration administered had no effect on the vitality and consequently the functionality of the intestinal segments, measured in terms of active glucose transport. The results imply that uranium concentrations of more than 20 lg/l in drinking water, for example, could lead to elevated absorption rates and thus to higher internal exposures to consider when setting of Guideline values in this concentration range. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Uranium is a natural radioactive element which can be present in the ground water in concentrations of a few hundred lg/l, depending on the geological conditions. In the lower concentration range, health effects are expected from its chemical toxicity as a heavy metal rather than from its radioactivity (Tasat et al., 2012). The chemical toxicity of uranium depends on the chemical species as which it occurs, particularly on its water solubility (Harrison and Stather, 1981). Uptake can be through the lung, the skin or the gastrointestinal tract and results of a study with volunteers in Canada (Zamora et al., 2002) show the relevance of uptake through drinking-water. As highlighted in the World Health Organisation’s (WHO) Guidelines for Drinking-water Quality, ‘‘in circumstances in which uranium is present in a drinking-water source, the majority of intake can be through drinking-water’’ (WHO, 2011, p. 3). While concentrations are low in many areas, more than 20 lg/l have particularly been reported from a range of small drinkingwater supplies in a number of regions of the world, and the WHO is including uranium in its plan of work of the rolling revision ⇑ Corresponding author. Fax: +49 30 8903 1830. E-mail address: [email protected] (R. Konietzka). http://dx.doi.org/10.1016/j.yrtph.2014.04.012 0273-2300/Ó 2014 Elsevier Inc. All rights reserved.
of the WHO Guidelines for Drinking-water Quality (WHO, 2012). Germany introduced a limit of 10 lg/l in its Drinking Water Ordinance in 2011 (TrinkwV, 2011). Uranium mainly damages the kidneys, especially the glomeruli and proximal tubuli. It causes histological changes of the epithelial cells in the lower segment of the proximal tubuli, which functionally lead to increased excretion of glucose, amino acids and proteins (e.g., b-micro globulin). Glomerulus damage can lead to a decrease of the glomerular filtration capacity, recognizable through changes in clearance rates and through proteinuria (EFSA, 2009; WHO, 2012). Furthermore moderate effects on liver function were demonstrated experimentally (Gueguen et al., 2006), and Dublineau et al. (2006) report indication for altered intestinal immunological reactions after repeated exposure. Data on the various absorption sites along the gastrointenstinal tract as well as for cellular pathways (para- or trans-cellular) and the transport systems involved in uranium absorption are limited (Dublineau et al., 2005). Experimental results obtained with rats indicate the trans-cellular pathway to be the primary one for gastrointestinal absorption of dissolved uranium compounds in the small intestine (Dublineau et al., 2006). As in-vivo experiments are under ethical debate, cost-intensive and the information they provide is limited, it is essential to include ex-vivo and in-vitro
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
methods to clarify specific questions on e.g. absorption concentration relations. Ex-vivo investigations of the intestinal lining take the interaction between natural cell populations into account and can thus provide insights in addition to those gleaned from in-vitro investigations using intestinal cell cultures. Data available to date on dose-dependent uranium (U) absorption are inconsistent (Konietzka, submitted for publication). The collation of percentage absorption levels for various species published by Leggett and Harrison (1995) shows no increase with increasing amounts of soluble uranium compounds administered. However, among the reports used by these authors, a trend is only clear in the data from La Touche et al. (1987), and the increase of absorption rates they observed following a hundredfold increase in the dose administered to rats was only barely fourfold. This may be due to exhaustion of the capacity of active transport systems (as known for other divalent metals, Illing et al. (2012) and which are likely to be relevant for uranium as well), eg. the exhaustion of transport proteins or their expression. The question therefore arises as to whether uranium that occurs naturally in the environment, for example in ground, drinking and mineral water, follows the same absorption pattern at higher concentrations, or whether there is no need to regulate higher environmentally-relevant concentrations as they do not lead to higher intake. The results presented here targeted experimental clarification of this issue by examining the absorption of uranium from a range of uranium concentrations using an ex-vivo model consisting of isolated surviving jejunal segments of rats, as part of the small intestine where the uranium adsorption takes place (Dublineau et al., 2005). 2. Material and methods 2.1. Experimental animals The tests, which were notified to Chemnitz Regional Council in accordance with animal protection legislation, were conducted on male Sprague–Dawley rats (supplier: Charles River, Sulzfeld) with 220–250 g body weight. The animals were kept in a filter cabinet (supplier: Ehret, Emmendingen), five animals per type IV macrolon cage, in a dark/light cycle of 12 h darkness alternating with 12 h light, and at a temperature of 24 ± 2 °C. Unlimited drinking water
257
and feed (supplier: Altromin, Lage) was available. After a one-week familiarisation phase, feed was withdrawn from the animals 12 h before intestinal preparation, but they continued to have unlimited access to drinking water. 2.2. Preparation of the surviving jejunal segments The rats were killed by cervical dislocation while under ether anaesthesia. The abdominal cavity was opened along the median line, the upper small intestine segment was released and severed behind the Flexura duodenojejunalis. A glass capillary tube was inserted into the intestinal lumen and made liquid-tight. An intestinal segment approximately 10 cm long lying in the caudal direction was severed and flushed through via the capillary tube in place with an electrolyte solution (Tyrode’s solution, pH 7.2) at a temperature of 37 °C that had been enriched with oxygen by means of carbogen gasification (95% O2, 5% CO2). Another glass capillary tube was then attached at the other end of the segment. Finally the blood vessels of the intestine were severed, and the segment was placed in a vessel and flushed internally and externally with oxygen-enriched Tyrode’s solution at a temperature of 37 °C before being transferred to the perfusion equipment. 2.3. Perfusion equipment The intestinal segments were placed in perfusion equipment constructed according to the design devised by Fisher and Parson (1949), with the modifications introduced by Schümann et al. (1986). The equipment (Fig. 1) consisted of two double-walled sealable glass vessels placed one on top of the other, with water at 37 °C flowing through the space between them, as described by Richter and Strugala (1985). The upper glass vessel, which acted as the reservoir for the perfusion solution, ran into a small glass tube at its bottom end leading into the lower vessel via a Teflon tube fitted with a three-way valve. It was connected to one end of the prepared intestinal segment at this point, under liquid-tight conditions. The other end of the intestinal segment was connected with the small glass tube used to supply carbogen, also under liquid-tight conditions. As a result, the intestinal segment hung free in the space of the lower glass vessel, which was kept moist by means of blotting paper soaked in Tyrode’s solution placed on
Fig. 1. Ex-vivo perfusion equipment constructed according to the design devised by Fisher and Parson (1949), modified from Schümann et al. (1986), (left, Perfusat = perfusate, Resorbat = absorbate; Fig. from Richter and Strugala, 1985) and the laboratory set-up of five sets of perfusion equipment forming an ex-vivo perfusion battery (right).
258
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
the walls of the vessel and at a temperature of 37 °C. The flow of the perfusion solution was controlled by the carbogen supply via three-way valves. The absorbate extracted from the intestines passed through a funnel placed in the lower opening of the second perfusion vessel and was collected in Eppendorf tubes. Five sets of perfusion equipment were set up to form a battery for parallel testing (Fig. 1).
ing a glucose calibration curve (based on a 45% standard glucose solution, supplier: Sigma). In addition to absorbate glucose level measurements, the levels of glucose in the perfusion solutions were also measured immediately before the start of perfusion (P0) and after 120 min of perfusion (PE).
2.4. Uranium concentrations used
Uranium levels in the perfusion solutions and absorbates were measured in appropriate dilutions using quadropole mass spectrometry, with inductively coupled plasma as the ion source (ICP-MS). The samples which had been acidified with 2% nitric acid (Merck, SuprapurÒ) were compared directly with standard 2% nitric acid (single standard, supplier: Accu Standard Inc. New Haven, CT). All the analytical values were produced from double measurements with three repeats each. The limit of quantification (6 times standard deviation from the standard blank from 10 repeats) was 0.002 lg/l.
A perfusion solution consisting of Tyrode’s electrolyte solution (pH 7.2 and osmolality = 295 mOsm/kg) was used as the control. For the tests involving uranium, each batch of Tyrode’s solution was made up with a defined volume of water containing uranium instead of deionised water, with all other ingredients remaining the same. This water, obtained from a spring in Nürtingen, in the southern German district of Neckar, contained approximately 400 lg/l uranium (as shown by in-house analyses; see below) and approximately 20% equivalent proportions of sodium, sulphate and hydrogen carbonate (supplier: Labour Jäger, Tübingen). The sodium content was reduced to approximately 3.15 g/l in total (remaining concentration in commercial Tyrode’s solution) in order to ensure that the osmolality of the solution did not rise above approximately 295 mOsm/kg once the water containing uranium had been added. The pH value of all perfusion solutions was set at 7.2. The perfusion solutions were made up fresh each day and the actual uranium levels were determined by analysis before the start of perfusion (P0) and after the end of perfusion (PE). 2.5. Performance of the ex-vivo perfusion On one study day five intestinal segments were perfused in parallel with varying uranium concentrations. The first stage of this process involved placing the corresponding solutions into the upper vessel of the equipment, heating it to 37 °C and enriching it with oxygen by passing carbogen through the solution. The prepared duodenum segments were then inserted, and the shut-off valves above the carbogen inlet were opened to start the perfusion liquid circulating. The first drop of absorbate which passed through the intestine still contained blood and was discarded. Immediately after this the absorbate was collected in 30-min intervals over a 120-min period. The volume of each of the four absorbates was determined separately and centrifugated at 13,000 rpm for 10 min to remove tissue components and any remaining erythrocytes. The residue was removed and a part was immediately taken to measure the glucose content. The remaining volumes were pooled to form a single sample and topped up to 5 ml with 2% HNO3 (SuprapurÒ, Merck 65%). Uranium concentrations were then measured for these samples without any further preparation. This procedure was repeated on 6 to 7 intestinal segments for each uranium concentration. Most of the samples prepared for uranium concentration measurement were affected by clouding, probably due to suspended matter, despite centrifugation and careful removal of the supernatant. 2.6. Test of the vitality of the ex-vivo intestinal segments that underwent perfusion The vitality of the duodenum segments is an essential condition for ex-vivo tests of absorption. The vitality of the intestinal segments was tested by measuring the volume of perfusion solution passing through the intestinal wall (volume of absorbate) and the glucose levels in the absorbate. High glucose levels in the absorbate show that the volume of absorbate has not been distorted by intestinal lesions. A test kit (supplier: Xenometrix AG, Allschwil, Switzerland) was used to measure glucose levels, apply-
2.7. Uranium analysis
2.8. Statistics Samples with <50 lg/l U and samples with >80 lg/l U in the perfusion solution were evaluated separately, as the analytical findings suggested that the absorption could be significantly different in these two concentration ranges. Within the lower concentration range, a distinction was made between groups A (around 5 lg/ l), B (around 21 lg/l) and C (around 43 lg/l), and for groups in the higher concentration range between D (around 99 lg/l), E (around 133 lg/l) F (around 199 lg/l) and G (around 422 lg/l). Arithmetic means (AM) and standard deviations (SD) were calculated for all samples within each group, along with the number of perfused intestinal segments (n, see in Tables 2 and 3, column Group). Absorbate uranium concentrations below the limit of quantification were set at half the limit of determination (0.001 lg/l) for computations. The dependency of the absorbate volume (ml absorbate/cm intestine) after 120 min perfusion on the uranium concentration in the perfusion solution was investigated as a vitality test. We expected these two parameters to be independent, i.e. that neither a linear relation (significant Pearson’s correlation) nor a monotonic relation (significant Spearman’s rank correlation) would be observed. In order to investigate the vitality, we also evaluated the change in glucose concentrations in the absorbate over the test period (after 30, 60, 90 and 120 min), using a two-factorial analysis of variance with repeated measurements, testing the linear and the quadratic trend components for significance. Samples with lower and higher uranium concentrations in the perfusion solution (<50 lg/l vs. >80 lg/l) were tested separately to elucidate the supposed differences in trend. We described the relation between uranium concentrations in the perfusion solution and those in the absorbate with linear regression analyses. The analyses were performed separately for the samples in the lower concentration range (<50 lg/l U in the perfusion solution) and the higher concentration range (>80 lg/l) so as to be able to identify differences in absorption. The strength of the relation was indicated by the determination coefficient R2, and the form by the regression line. The regression constant was set at 0 when performing the regression analysis for the lower concentration range. Notably, each regression equation only applies for the range of values which was analysed in this study. If the slopes of the regression lines are significantly different in the lower and higher concentration ranges, this means that a smaller amount of the uranium administered in the perfusion solution passed into the absorbate in one concentration range than in the other. The regression lines are significantly different if the confidence intervals of the slope coefficients do not overlap.
259
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
Further analyses were carried out to discover whether a common non-linear function described the relationship better than a linear function (higher R2). We tested a second- and third-degree polynomial, a logarithmic function, an exponential function and a power function. Microsoft Office Excel 2003 was used to compute the statistical parameters and to generate the scatter plots. The statistical tests and regression analyses were performed with SPSS 17 for Windows (SPSS Inc., Chicago, Illinois).
3. Results 3.1. Vitality test parameters Absorbate volumes do not linearly follow the uranium concentration changes in the perfusion solutions (for the data shown in Table 1 r = 0.093, p = 0.518), but they do show a significant, though weak, monotonic relation (Spearman’s rank correlation rS = 0.333, p = 0.018). However, this is only due to the two lowest concentration groups (around 5 and 21 lg/l U) and is not observed if these are excluded from the correlation analysis. Glucose concentrations in the perfusion solutions (15.1 mM at the start of perfusion, P0) were slightly lower at the end of the test (PE, after 120 min perfusion) because of the active removal of glucose with the absorbate (Table 2). The higher glucose concentrations observed in the absorbate already after the first 30-min perfusion interval can be attributed to the active transport mechanisms of the intestinal segments which had retained their vitality. The glucose absorption trends over the course of perfusion (30, 60, 90 or 120 min) varied depending on the uranium concentrations in the samples. In samples with higher uranium levels, a significant linear increase of glucose levels in the absorbate was observed over time (linear trend component: p = 0.029; quadratic trend component: not significant). In contrast, a curvilinear trend was observed in samples with lower uranium levels, with a slight rise and subsequent fall to the level observed after 30 min perfusion time (quadratic trend component: p = 0.006; linear trend component: not significant).
3.2. Uranium concentrations of the perfusion solutions and of the absorbates Table 3 shows the uranium concentrations of the perfusion solutions as determined by analysis before perfusion (P0) and after 120 min perfusion (PE), and the uranium concentrations of the absorbates. The percentages of uranium observed in the absorbate showed a monotonic trend from around 6% for 5 (group A) and 21 lg/l U (group B) in the perfusion liquid, rising to 29% for around 200 lg/l U (group F, Table 3). However, from 200 to 420 lg/l
(group G) the uranium absorption rate did not show a statistically significant further increase. The uranium absorption results are summarised in Fig. 2. They show a smaller increase parameter for samples in the lower concentration range (<50 lg/l) (slope of regression lines 0.123, 95% CI: 0.079–0.167) as compared to the higher concentration range (> 80–442 lg/l, increase in regression lines here 0.286, 95% CI: 0.208–0.365). This indicates that at lower concentrations (<50 lg/l U) a smaller proportion of the uranium administered in the perfusion liquid passes into the absorbate than at concentrations >80 lg/l U. As there is no overlap between the confidence intervals of the parameters of the two concentration ranges, the difference between the slope coefficients is statistically significant. The regression analyses showed a significant linear relation between the uranium concentrations of perfusions solutions and of absorbates. However, the correlation found for the lower level range is not quite as strong as that found for the higher level range (determination coefficients R2 = 0.4593 vs. R2 = 0.649). As Fig. 2 (dotted lines) shows, the fit to the data is only slightly inferior for the non-linear functions, i.e. an exponential function in the lower value range and a logarithmic function in the higher value range (determination coefficients R2 = 0.4389 and R2 = 0.6182). The proportion of uranium absorbed expressed as percent of uranium concentrations in the perfusion solutions followed a monotonically increasing trend from 20 to around 200 lg/l uranium in the perfusion solutions, which thereafter appears to reach a plateau at concentrations of more than 400 lg/l (see right column Table 3 and Fig. 3). The 7 group mean percentages shown in Fig. 3 could be fitted similarly well by other functions than a logarithmic one. To further elucidate this trend, a study with groups equally distributed over a wider range of uranium concentrations in the perfusion solutions would be valuable.
4. Discussion and conclusions The purpose of the investigations presented here, performed on an ex-vivo model of surviving rat intestinal segments, was to clarify whether uranium absorption rates are dose-dependent, which would imply a nonlinear increase of risk at higher exposure levels. Although the data show some variation in each dose group, the overall result is a non-linear increase of absorption rates and the proportion of administered uranium absorbed with increasing concentrations to which the animals were exposed: For the lower uranium concentrations adsorption rates were around 5 lg/l were around 6%, and they increased at higher uranium concentrations administered, gradually in the lower uranium concentration range (around 40 lg/l) and more sharply above 80 lg/l, but reaching a plateau of about 30% at 200 lg/l U and levelling off at further increase of the uranium concentration towards 400 lg/l. The proportion of uranium absorbed also shows this plateau, leveling off
Table 1 Absorbate volume after 120 min of perfusion. Group
Concentration of perfusion solution lg uranium/l (AM ± SD)
ml absorbate/cm intestine over 120 min of perfusion (AM ± SD)
<50 lg uranium/l perfusion solution (P0)
A (n = 6) B (n = 6) C (n = 6)
5.15 ± 2.24 21.0 ± 0.44 43.1 ± 2.68
0.170 ± 0.029 0.181 ± 0.044 0.223 ± 0.077
>80 lg uranium/l perfusion solution (P0)
D (n = 9) E (n = 9) F (n = 7) G (n = 7)
99.1 ± 12.2 133 ± 9.36 199 ± 18.5 422 ± 24.2
0.347 ± 0.177 0.252 ± 0.092 0.239 ± 0.067 0.236 ± 0.047
Notes: n = number of intestinal segments perfused; AM = arithmetic mean; SD = standard deviation.
260
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
Table 2 Glucose concentration of perfusion solutions after the end of the 120-min perfusion period (PE) and of absorbates at 30-min intervals (glucose concentration of the perfusion solution before the start of perfusion: P0 = 15.1 mM). Group
Perfusion solution mM glucose (AM SD) PEafter 120 min perfusion
Absorbate mM glucose (AM ± SD) 30-min interval
60-min interval
90-min interval
120-min interval
<50 lg uranium/l perfusion solution (P0)
0 (n = 3) A (n = 6) B (n = 6) C (n = 6)
13.8 ± 0.6 11.8 ± 1.0 13.2 ± 2.4 12.4 ± 1.7
17.9 ± 3.3 34.1 ± 9.0 23.3 ± 6.0 28.2 ± 6.3
20.2 ± 3.3 38.4 ± 9.1 25.1 ± 3.1 29.1 ± 7.2
18.7 ± 2.2 34.4 ± 10.0 26.0 ± 4.0 26.3 ± 6.3
17.2 ± 4.4 34.0 ± 10.2 23.2 ± 6.1 23.4 ± 4.8
>80 lg uranium/l perfusion solution (P0)
D (n = 9) E (n = 9) F (n = 7) G (n = 7)
12.5 ± 1.5 13.9 ± 4.7 14.2 ± 1.2 13.6 ± 1.7
18.5 ± 5.0 17.7 ± 5.7 15.8 ± 2.2 18.1 ± 3.0
19.1 ± 4.1 18.2 ± 3.8 17.4 ± 2.9 18.6 ± 2.4
18.4 ± 4.9 19.5 ± 3.8 19.1 ± 5.0 18.8 ± 3.8
18.5 ± 4.8 21.1 ± 5.0 20.8 ± 5.8 19.2 ± 3.8
Notes: n = number of intestinal segments perfused; AM = arithmetic mean; SD = standard deviation.
Table 3 Uranium (U) concentrations of perfusion solutions before the start of perfusion (P0) and after 120 min perfusion (PE) and of absorbates. Group
<50 lg uranium/l perfusion solution (P0)
O (n = 3) A (n = 6) B (n = 6) C (n = 6)
>80 lg uranium/l perfusion solution (P0)
D (n = 9) E (n = 9) F (n = 7) G (n = 7)
lg/l U [AM ± SD (min–max)]
% uranium of P0 in absorbate
Perfusion solution (P0)
Perfusion solution (PE)
Absorbate (120 min)
0±0 5.15 ± 2.24 (3.47–8.47) 21.0 ± 0.44 (20.4–21.6) 43.1 ± 2.68 (39.9–47.7)
0 ± 0.0 2.35 ± 1.05 (1.47–4.13) 15.6 ± 2.68 (12.2–18.7) 35.5 ± 4.65 (29.5–41.3)
0 ± 0.0 0.41 ± 0.51 (0–1.25) 1.07 ± 0.75 (0.15–1.88) 6.13 ± 3.93 (2.22–11.3)
– 7% ± 17% (1%–47%) 5% ± 4% (1–9%) 14% ± 9% (5–26%)
99.1 ± 12.2 (88.6–118) 133 ± 9.36 (123–152) 199 ± 18.5 (179–224) 422 ± 24.2 (375–442)
99.7 ± 17.3 (73.5–134) 135 ± 12.1 (120–159) 199 ± 25.2 (172–241) 451 ± 22.1 (412–481)
19.8 ± 18.6 (2.88–57.6) 25.0 ± 16.6 (5.2–52.9) 56.8 ± 16.9 (24.1–74.7) 112 ± 47.9 (39.9–185)
20% ± 19% (3–55%) 19% ± 13% (4–42%) 29% ± 10% (11–42%) 26% ± 11% (9–43%)
Notes: n = number of intestinal segments perfused; AM = arithmetic mean; SD = standard deviation; min = minimum; max = maximum.
Fig. 2. Uranium concentrations in absorbates at low and higher perfusion solution concentrations.
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
261
Fig. 3. The proportion of uranium absorbed expressed as percent of uranium concentrations in the perfusion solutions (Data points are the group means shown in Table 3). For comparison, the regression functions for low and higher uranium concentrations in the perfusion solutions and the absorbates from Fig. 2 are included in this figure.
around 200 lg/l. The results of parallel tests of glucose absorption ruled out a concentration-dependent impairment of intestinal segment vitality. The advantage of these findings obtained with ex-vivo investigations is that they are independent of the physiological condition of the test animal. For example, withdrawal of food 12 h before preparation of the intestines may in principle increase the absorption measured in this study, as fasting increases uranium absorption in vivo (Sullivan et al., 1986; La Touche et al., 1987 and Bhattacharyya et al., 1989). This study excluded the possible impact of factors such as the diet and its iron content, as described by Sullivan and Ruemmler (1988), and the resulting concentration/ absorption ratios are therefore affected only by the relationships within the intestine and by the uranium. As the absorption of divalent cations is an active process, e.g. involving transport through the apical cell membrane (Powell et al., 1999), the marked increase in absorption at higher concentrations points either to an additional absorption mechanism or mode, or to elevated expression of the bound proteins, such as the divalent metal transporter 1 (DMT1). DMT1 is known to be triggered by the iron content of the diet (Park et al., 2002; Gunshin et al., 1997; Canonne-Hergaux et al., 1999), and it may also be triggered by fasting (Sullivan et al., 1986; Konietzka, submitted for publication). As these factors were excluded in the case of the investigation method used here, two possible explanations remain: at the higher uranium concentrations an additional absorption mode or absorption mechanism, such as passive diffusion (Dublineau et al., 2005), may be more prominent, or transport proteins that may be involved (Okazaki et al., 2012; Nadadur et al., 2008; Au et al., 2008; Yokel et al., 2006) are triggered by the uranium itself as part of a concentration-dependent process. The fact that non-linear functions describe the data only slightly poorer and also seem biologically plausible tends to support the second possibility. However, both hypotheses require further testing. For regulatory purposes, the World Health Organisation and other authors assumed relatively low in-vivo absorption rates ranging from 0.3% to 1.5% (WHO, 2012; Konietzka, 2014; Konietzka et al., 2005). However, for the purposes of risk quantification our results indicate that absorption rates could be signifi-
cantly higher for uranium concentrations above 20 lg/l than at lower concentrations. The resulting higher internal exposures appear relevant for the setting of Guideline values in this concentration range because higher absorptions rates will cause higher levels in the target organ (kidneys). A lower Guideline value would be appropriate to account for this mechanism. Conflicts of interest None. Acknowledgments We thank Mr. Matthias Skerswetat for his experimental work and Mr. Ulrich Lippold for the uranium analyses. We are grateful to Dr. Ingrid Chorus and Dr. Tamara Grummt for helpful comments. References Au, C., Benedetto, A., Aschner, M., 2008. Manganese transport in eukaryotes: the role of DMT1. Neuro Toxicology 29, 569–576. Bhattacharyya, M.H., Larsen, R.P., Cohen, N., Ralston, L.G., Moretti, E.S., Oldham, R.D., Ayrest, L., 1989. Gastrointestinal absorption of Plutonium and Uranium in feed and fasted adult Baboons and mice: application to humans. Radiat. Prot. Dosimetry 26, 159–165. Canonne-Hergaux, F., Gruenheid, S., Ponka, P., Gros, P., 1999. Cellular and subcellular localization of the Nramp2 iron transporter in the intestinal brush border and regulation by dietary iron. Blood 93 (12), 4406–4417. Dublineau, I., Grison, S., Baudelin, C., Dudoignon, N., Souidi, M., Marquette, C., Paquet, F., Aigueperese, J., Gourmelon, P., 2005. Absorption of uranium through the entire gastrointestinal tract of the rat. Int. J. Radiat. Biol. 81, 473–482. Dublineau, I., Grison, S., Linard, C., Baudelin, C., Dudoignon, N., Souidi, M., Marquette, C., Paquet, F., Aigueperse, J., Gourmelon, P., 2006. Short-term effects of depleted Uranium on immune status in rat intestine. J. Toxicol Environ. Health, Part A: Current Issues 69, 1613–1628. EFSA, European Food Safety Authority, 2009. Uranium in foodstuffs, in particular mineral water. Scientific opinion of the Panel on Contaminants in the Food Chain.. Fisher, R.B., Parsons, D.S., 1949. A preparation of surviving rat small intestine for the study of absorption. J. Physiol. 110, 36–46. Gueguen, Y., Souidi, M., Baudelin, C., Dudoignon, N., Grison, N., Dublineau, I., Marquette, C., Voisin, P., Gourmelon, P., Aigueperse, J., 2006. Short-term hepatic effects of depleted uranium on xenobiotic and bile acid metabolizing cytochrome P450 enzymes in the rat. Arch. Toxicol. 80, 187–195. Gunshin, H., Mackenzie, B., Berger, U.V., Gunshin, Y., Romero, M.F., Boron, W.F., Nussberger, S., Gollan, J.L., Hediger, M.A., 1997. Cloning and characterization of a mammal-ian proton-coupled metal-ion transporter. Nature 388, 482–488. Harrison, J.D., Stather, J.W., 1981. The gastrointestinal absorption of Protactinium, Uranium and Neptunium in the hamster. Radiat. Res. 88 (1), 47–55. Illing, A.C., Shawki, A., Cunninghamm, C.L., Mackenzie, B., 2012. Substrate profile and metal-ion selectivity of human divalent metal-ion transporter-1. J. Biol. Chem. 287, 30485–30496.
262
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
Konietzka, R., 2014. Gastrointestinal absorption of uranium compounds – a review, Regul. Toxicol. Pharmacol., submitted for publication. Konietzka, R., Dieter, H.H., Voss, J.-U., 2005. Vorschlag für einen gesundheitlichen Leitwert für Uran in Trinkwasser. Umweltmed Forsch Prax 10, 133–143. La Touche, Y.D., Willis, D.L., Dawydiak, O.I., 1987. Absorption and biokinetics of U in rats following an oral administration of Uranyl Nitrate solution. Health Phys. 53, 147–162. Leggett, R.W., Harrison, J.D., 1995. Fractional absorption of ingested uranium in humans. Health Phys. 68, 484–498. Nadadur, S.S., Srirama, K., Mudipa, A., 2008. Iron transport & homeostasis mechanisms: their role in health & disease. Indian J. Med. Res. 128, 533–544. Okazaki, Y., Ma, Y., Yeh, M., Yin, H., Li, Z., Yeh, K., Glass, J., 2012. DMT1 (IRE) expression in intestinal and erythroid cells is regulated by peripheral benzodiazepine receptor-associated protein 7. Am. J. Physiol. Gastrointest. Liver Physiol. 302, G1180–G1190. Park, J.D., Cherrington, N.J., Klaassen, C.D., 2002. Intestinal absorption of cadmium is associated with divalent metal transporter 1 in rats. Toxicol. Sci. 68, 288–294. Powell, J.J., Jugdaohsingh, R., Thompson, R.P.H., 1999. The regulation of mineral absorption in the gastrointestinal tract. Proc. Nutr. Soc. 58, 147–153. Richter, E., Strugala, G.J., 1985. An all-glass perfusator for investigation of the intestinal transport and metabolism of foreign compounds in vitro. J. Pharmacol. Methods 14, 297–304. Schümann, K., Osterloh, K., Forth, W., 1986. Independence of in vitro iron absorption from mucosal transferrin content in rat jejunal and ileal segments. Blut 53, 391– 400.
Sullivan, M.F., Ruemmler, P.S., 1988. Absorption of 233U, 237Np, 238Pu, 241Am and 244Cm from gastro intestinal tract of rats fed an iron-deficient diet. Health Phys. 54, 311–316. Sullivan, M.F., Ruemmler, P.S., Ryan, J.L., Buschbom, R.L., 1986. Influence of oxidizing or reducing agents on gastrointestinal absorption of U, Am, Cm and Pm by rats. Health Phys. 50, 223–232. Tasat, D.R., Orona, N.S., Bozal, C., Ubios, A.M., Cabrini, R.L., 2012. Intracellular metabolism of uranium and the effects of bisphosphonates on its toxicity. In: Paula Bubulya (Ed.), Cell Metabolism – Cell Homeostasis and Stress Response. pp. 116–148. Rijeka, Croatia: InTech. ISBN 978-953-307-978-3. TrinkwV, TrinkwasseränderungsV (2011): Erste Verordnung zur Änderung der Trinkwasserverordnung vom 03. Mai 2011. Bundesgesetzblatt, Jahrgang 2011, Teil I Nr. 21, ausgegeben zu Bonn am 11. Mai 2011. WHO, World Health Organization, 2011. Guidelines for drinking-water quality, fourth ed.
Contents lists available at ScienceDirect
Regulatory Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/yrtph
The ex-vivo intestinal absorption rate of uranium is a two-phase function of supply Rainer Konietzka ⇑, Rita Heinze, Margarete Seiwert, Hermann H. Dieter Federal Environment Agency, Toxicology of Drinking Water and Swimming Pool Water and Health-related Environmental Monitoring, Wörlitzer Platz 1, 06844 Dessau-Roßlau, Germany
a r t i c l e
i n f o
Article history: Received 25 September 2013 Available online 2 May 2014 Keywords: Uranium Ex vivo Small intestine Absorption
a b s t r a c t The concentration-dependent absorption behaviour of uranium was investigated with surviving intestinal segments of rat jejunums, using an ex-vivo model. The results showed a monotonic slightly nonlinear increase in absorption as uranium concentrations increased. This trend was observed over the entire concentration range tested. In the lower concentration range a slower linear ascent was observed while a steeper linear ascent was found for the higher concentration range. Statistical fit was only slightly poorer for an exponential function in the range of lower values and a logarithmic function in the range of higher values. The proportion of uranium absorbed expressed as percent of uranium concentrations in the perfusion solutions followed a monotonically increasing trend from 20 to around 200 lg/l uranium in the perfusion solutions, which thereafter appears to reach a plateau, as further increase towards concentrations around 400 lg/l is not substantial. The uranium concentration administered had no effect on the vitality and consequently the functionality of the intestinal segments, measured in terms of active glucose transport. The results imply that uranium concentrations of more than 20 lg/l in drinking water, for example, could lead to elevated absorption rates and thus to higher internal exposures to consider when setting of Guideline values in this concentration range. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Uranium is a natural radioactive element which can be present in the ground water in concentrations of a few hundred lg/l, depending on the geological conditions. In the lower concentration range, health effects are expected from its chemical toxicity as a heavy metal rather than from its radioactivity (Tasat et al., 2012). The chemical toxicity of uranium depends on the chemical species as which it occurs, particularly on its water solubility (Harrison and Stather, 1981). Uptake can be through the lung, the skin or the gastrointestinal tract and results of a study with volunteers in Canada (Zamora et al., 2002) show the relevance of uptake through drinking-water. As highlighted in the World Health Organisation’s (WHO) Guidelines for Drinking-water Quality, ‘‘in circumstances in which uranium is present in a drinking-water source, the majority of intake can be through drinking-water’’ (WHO, 2011, p. 3). While concentrations are low in many areas, more than 20 lg/l have particularly been reported from a range of small drinkingwater supplies in a number of regions of the world, and the WHO is including uranium in its plan of work of the rolling revision ⇑ Corresponding author. Fax: +49 30 8903 1830. E-mail address: [email protected] (R. Konietzka). http://dx.doi.org/10.1016/j.yrtph.2014.04.012 0273-2300/Ó 2014 Elsevier Inc. All rights reserved.
of the WHO Guidelines for Drinking-water Quality (WHO, 2012). Germany introduced a limit of 10 lg/l in its Drinking Water Ordinance in 2011 (TrinkwV, 2011). Uranium mainly damages the kidneys, especially the glomeruli and proximal tubuli. It causes histological changes of the epithelial cells in the lower segment of the proximal tubuli, which functionally lead to increased excretion of glucose, amino acids and proteins (e.g., b-micro globulin). Glomerulus damage can lead to a decrease of the glomerular filtration capacity, recognizable through changes in clearance rates and through proteinuria (EFSA, 2009; WHO, 2012). Furthermore moderate effects on liver function were demonstrated experimentally (Gueguen et al., 2006), and Dublineau et al. (2006) report indication for altered intestinal immunological reactions after repeated exposure. Data on the various absorption sites along the gastrointenstinal tract as well as for cellular pathways (para- or trans-cellular) and the transport systems involved in uranium absorption are limited (Dublineau et al., 2005). Experimental results obtained with rats indicate the trans-cellular pathway to be the primary one for gastrointestinal absorption of dissolved uranium compounds in the small intestine (Dublineau et al., 2006). As in-vivo experiments are under ethical debate, cost-intensive and the information they provide is limited, it is essential to include ex-vivo and in-vitro
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
methods to clarify specific questions on e.g. absorption concentration relations. Ex-vivo investigations of the intestinal lining take the interaction between natural cell populations into account and can thus provide insights in addition to those gleaned from in-vitro investigations using intestinal cell cultures. Data available to date on dose-dependent uranium (U) absorption are inconsistent (Konietzka, submitted for publication). The collation of percentage absorption levels for various species published by Leggett and Harrison (1995) shows no increase with increasing amounts of soluble uranium compounds administered. However, among the reports used by these authors, a trend is only clear in the data from La Touche et al. (1987), and the increase of absorption rates they observed following a hundredfold increase in the dose administered to rats was only barely fourfold. This may be due to exhaustion of the capacity of active transport systems (as known for other divalent metals, Illing et al. (2012) and which are likely to be relevant for uranium as well), eg. the exhaustion of transport proteins or their expression. The question therefore arises as to whether uranium that occurs naturally in the environment, for example in ground, drinking and mineral water, follows the same absorption pattern at higher concentrations, or whether there is no need to regulate higher environmentally-relevant concentrations as they do not lead to higher intake. The results presented here targeted experimental clarification of this issue by examining the absorption of uranium from a range of uranium concentrations using an ex-vivo model consisting of isolated surviving jejunal segments of rats, as part of the small intestine where the uranium adsorption takes place (Dublineau et al., 2005). 2. Material and methods 2.1. Experimental animals The tests, which were notified to Chemnitz Regional Council in accordance with animal protection legislation, were conducted on male Sprague–Dawley rats (supplier: Charles River, Sulzfeld) with 220–250 g body weight. The animals were kept in a filter cabinet (supplier: Ehret, Emmendingen), five animals per type IV macrolon cage, in a dark/light cycle of 12 h darkness alternating with 12 h light, and at a temperature of 24 ± 2 °C. Unlimited drinking water
257
and feed (supplier: Altromin, Lage) was available. After a one-week familiarisation phase, feed was withdrawn from the animals 12 h before intestinal preparation, but they continued to have unlimited access to drinking water. 2.2. Preparation of the surviving jejunal segments The rats were killed by cervical dislocation while under ether anaesthesia. The abdominal cavity was opened along the median line, the upper small intestine segment was released and severed behind the Flexura duodenojejunalis. A glass capillary tube was inserted into the intestinal lumen and made liquid-tight. An intestinal segment approximately 10 cm long lying in the caudal direction was severed and flushed through via the capillary tube in place with an electrolyte solution (Tyrode’s solution, pH 7.2) at a temperature of 37 °C that had been enriched with oxygen by means of carbogen gasification (95% O2, 5% CO2). Another glass capillary tube was then attached at the other end of the segment. Finally the blood vessels of the intestine were severed, and the segment was placed in a vessel and flushed internally and externally with oxygen-enriched Tyrode’s solution at a temperature of 37 °C before being transferred to the perfusion equipment. 2.3. Perfusion equipment The intestinal segments were placed in perfusion equipment constructed according to the design devised by Fisher and Parson (1949), with the modifications introduced by Schümann et al. (1986). The equipment (Fig. 1) consisted of two double-walled sealable glass vessels placed one on top of the other, with water at 37 °C flowing through the space between them, as described by Richter and Strugala (1985). The upper glass vessel, which acted as the reservoir for the perfusion solution, ran into a small glass tube at its bottom end leading into the lower vessel via a Teflon tube fitted with a three-way valve. It was connected to one end of the prepared intestinal segment at this point, under liquid-tight conditions. The other end of the intestinal segment was connected with the small glass tube used to supply carbogen, also under liquid-tight conditions. As a result, the intestinal segment hung free in the space of the lower glass vessel, which was kept moist by means of blotting paper soaked in Tyrode’s solution placed on
Fig. 1. Ex-vivo perfusion equipment constructed according to the design devised by Fisher and Parson (1949), modified from Schümann et al. (1986), (left, Perfusat = perfusate, Resorbat = absorbate; Fig. from Richter and Strugala, 1985) and the laboratory set-up of five sets of perfusion equipment forming an ex-vivo perfusion battery (right).
258
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
the walls of the vessel and at a temperature of 37 °C. The flow of the perfusion solution was controlled by the carbogen supply via three-way valves. The absorbate extracted from the intestines passed through a funnel placed in the lower opening of the second perfusion vessel and was collected in Eppendorf tubes. Five sets of perfusion equipment were set up to form a battery for parallel testing (Fig. 1).
ing a glucose calibration curve (based on a 45% standard glucose solution, supplier: Sigma). In addition to absorbate glucose level measurements, the levels of glucose in the perfusion solutions were also measured immediately before the start of perfusion (P0) and after 120 min of perfusion (PE).
2.4. Uranium concentrations used
Uranium levels in the perfusion solutions and absorbates were measured in appropriate dilutions using quadropole mass spectrometry, with inductively coupled plasma as the ion source (ICP-MS). The samples which had been acidified with 2% nitric acid (Merck, SuprapurÒ) were compared directly with standard 2% nitric acid (single standard, supplier: Accu Standard Inc. New Haven, CT). All the analytical values were produced from double measurements with three repeats each. The limit of quantification (6 times standard deviation from the standard blank from 10 repeats) was 0.002 lg/l.
A perfusion solution consisting of Tyrode’s electrolyte solution (pH 7.2 and osmolality = 295 mOsm/kg) was used as the control. For the tests involving uranium, each batch of Tyrode’s solution was made up with a defined volume of water containing uranium instead of deionised water, with all other ingredients remaining the same. This water, obtained from a spring in Nürtingen, in the southern German district of Neckar, contained approximately 400 lg/l uranium (as shown by in-house analyses; see below) and approximately 20% equivalent proportions of sodium, sulphate and hydrogen carbonate (supplier: Labour Jäger, Tübingen). The sodium content was reduced to approximately 3.15 g/l in total (remaining concentration in commercial Tyrode’s solution) in order to ensure that the osmolality of the solution did not rise above approximately 295 mOsm/kg once the water containing uranium had been added. The pH value of all perfusion solutions was set at 7.2. The perfusion solutions were made up fresh each day and the actual uranium levels were determined by analysis before the start of perfusion (P0) and after the end of perfusion (PE). 2.5. Performance of the ex-vivo perfusion On one study day five intestinal segments were perfused in parallel with varying uranium concentrations. The first stage of this process involved placing the corresponding solutions into the upper vessel of the equipment, heating it to 37 °C and enriching it with oxygen by passing carbogen through the solution. The prepared duodenum segments were then inserted, and the shut-off valves above the carbogen inlet were opened to start the perfusion liquid circulating. The first drop of absorbate which passed through the intestine still contained blood and was discarded. Immediately after this the absorbate was collected in 30-min intervals over a 120-min period. The volume of each of the four absorbates was determined separately and centrifugated at 13,000 rpm for 10 min to remove tissue components and any remaining erythrocytes. The residue was removed and a part was immediately taken to measure the glucose content. The remaining volumes were pooled to form a single sample and topped up to 5 ml with 2% HNO3 (SuprapurÒ, Merck 65%). Uranium concentrations were then measured for these samples without any further preparation. This procedure was repeated on 6 to 7 intestinal segments for each uranium concentration. Most of the samples prepared for uranium concentration measurement were affected by clouding, probably due to suspended matter, despite centrifugation and careful removal of the supernatant. 2.6. Test of the vitality of the ex-vivo intestinal segments that underwent perfusion The vitality of the duodenum segments is an essential condition for ex-vivo tests of absorption. The vitality of the intestinal segments was tested by measuring the volume of perfusion solution passing through the intestinal wall (volume of absorbate) and the glucose levels in the absorbate. High glucose levels in the absorbate show that the volume of absorbate has not been distorted by intestinal lesions. A test kit (supplier: Xenometrix AG, Allschwil, Switzerland) was used to measure glucose levels, apply-
2.7. Uranium analysis
2.8. Statistics Samples with <50 lg/l U and samples with >80 lg/l U in the perfusion solution were evaluated separately, as the analytical findings suggested that the absorption could be significantly different in these two concentration ranges. Within the lower concentration range, a distinction was made between groups A (around 5 lg/ l), B (around 21 lg/l) and C (around 43 lg/l), and for groups in the higher concentration range between D (around 99 lg/l), E (around 133 lg/l) F (around 199 lg/l) and G (around 422 lg/l). Arithmetic means (AM) and standard deviations (SD) were calculated for all samples within each group, along with the number of perfused intestinal segments (n, see in Tables 2 and 3, column Group). Absorbate uranium concentrations below the limit of quantification were set at half the limit of determination (0.001 lg/l) for computations. The dependency of the absorbate volume (ml absorbate/cm intestine) after 120 min perfusion on the uranium concentration in the perfusion solution was investigated as a vitality test. We expected these two parameters to be independent, i.e. that neither a linear relation (significant Pearson’s correlation) nor a monotonic relation (significant Spearman’s rank correlation) would be observed. In order to investigate the vitality, we also evaluated the change in glucose concentrations in the absorbate over the test period (after 30, 60, 90 and 120 min), using a two-factorial analysis of variance with repeated measurements, testing the linear and the quadratic trend components for significance. Samples with lower and higher uranium concentrations in the perfusion solution (<50 lg/l vs. >80 lg/l) were tested separately to elucidate the supposed differences in trend. We described the relation between uranium concentrations in the perfusion solution and those in the absorbate with linear regression analyses. The analyses were performed separately for the samples in the lower concentration range (<50 lg/l U in the perfusion solution) and the higher concentration range (>80 lg/l) so as to be able to identify differences in absorption. The strength of the relation was indicated by the determination coefficient R2, and the form by the regression line. The regression constant was set at 0 when performing the regression analysis for the lower concentration range. Notably, each regression equation only applies for the range of values which was analysed in this study. If the slopes of the regression lines are significantly different in the lower and higher concentration ranges, this means that a smaller amount of the uranium administered in the perfusion solution passed into the absorbate in one concentration range than in the other. The regression lines are significantly different if the confidence intervals of the slope coefficients do not overlap.
259
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
Further analyses were carried out to discover whether a common non-linear function described the relationship better than a linear function (higher R2). We tested a second- and third-degree polynomial, a logarithmic function, an exponential function and a power function. Microsoft Office Excel 2003 was used to compute the statistical parameters and to generate the scatter plots. The statistical tests and regression analyses were performed with SPSS 17 for Windows (SPSS Inc., Chicago, Illinois).
3. Results 3.1. Vitality test parameters Absorbate volumes do not linearly follow the uranium concentration changes in the perfusion solutions (for the data shown in Table 1 r = 0.093, p = 0.518), but they do show a significant, though weak, monotonic relation (Spearman’s rank correlation rS = 0.333, p = 0.018). However, this is only due to the two lowest concentration groups (around 5 and 21 lg/l U) and is not observed if these are excluded from the correlation analysis. Glucose concentrations in the perfusion solutions (15.1 mM at the start of perfusion, P0) were slightly lower at the end of the test (PE, after 120 min perfusion) because of the active removal of glucose with the absorbate (Table 2). The higher glucose concentrations observed in the absorbate already after the first 30-min perfusion interval can be attributed to the active transport mechanisms of the intestinal segments which had retained their vitality. The glucose absorption trends over the course of perfusion (30, 60, 90 or 120 min) varied depending on the uranium concentrations in the samples. In samples with higher uranium levels, a significant linear increase of glucose levels in the absorbate was observed over time (linear trend component: p = 0.029; quadratic trend component: not significant). In contrast, a curvilinear trend was observed in samples with lower uranium levels, with a slight rise and subsequent fall to the level observed after 30 min perfusion time (quadratic trend component: p = 0.006; linear trend component: not significant).
3.2. Uranium concentrations of the perfusion solutions and of the absorbates Table 3 shows the uranium concentrations of the perfusion solutions as determined by analysis before perfusion (P0) and after 120 min perfusion (PE), and the uranium concentrations of the absorbates. The percentages of uranium observed in the absorbate showed a monotonic trend from around 6% for 5 (group A) and 21 lg/l U (group B) in the perfusion liquid, rising to 29% for around 200 lg/l U (group F, Table 3). However, from 200 to 420 lg/l
(group G) the uranium absorption rate did not show a statistically significant further increase. The uranium absorption results are summarised in Fig. 2. They show a smaller increase parameter for samples in the lower concentration range (<50 lg/l) (slope of regression lines 0.123, 95% CI: 0.079–0.167) as compared to the higher concentration range (> 80–442 lg/l, increase in regression lines here 0.286, 95% CI: 0.208–0.365). This indicates that at lower concentrations (<50 lg/l U) a smaller proportion of the uranium administered in the perfusion liquid passes into the absorbate than at concentrations >80 lg/l U. As there is no overlap between the confidence intervals of the parameters of the two concentration ranges, the difference between the slope coefficients is statistically significant. The regression analyses showed a significant linear relation between the uranium concentrations of perfusions solutions and of absorbates. However, the correlation found for the lower level range is not quite as strong as that found for the higher level range (determination coefficients R2 = 0.4593 vs. R2 = 0.649). As Fig. 2 (dotted lines) shows, the fit to the data is only slightly inferior for the non-linear functions, i.e. an exponential function in the lower value range and a logarithmic function in the higher value range (determination coefficients R2 = 0.4389 and R2 = 0.6182). The proportion of uranium absorbed expressed as percent of uranium concentrations in the perfusion solutions followed a monotonically increasing trend from 20 to around 200 lg/l uranium in the perfusion solutions, which thereafter appears to reach a plateau at concentrations of more than 400 lg/l (see right column Table 3 and Fig. 3). The 7 group mean percentages shown in Fig. 3 could be fitted similarly well by other functions than a logarithmic one. To further elucidate this trend, a study with groups equally distributed over a wider range of uranium concentrations in the perfusion solutions would be valuable.
4. Discussion and conclusions The purpose of the investigations presented here, performed on an ex-vivo model of surviving rat intestinal segments, was to clarify whether uranium absorption rates are dose-dependent, which would imply a nonlinear increase of risk at higher exposure levels. Although the data show some variation in each dose group, the overall result is a non-linear increase of absorption rates and the proportion of administered uranium absorbed with increasing concentrations to which the animals were exposed: For the lower uranium concentrations adsorption rates were around 5 lg/l were around 6%, and they increased at higher uranium concentrations administered, gradually in the lower uranium concentration range (around 40 lg/l) and more sharply above 80 lg/l, but reaching a plateau of about 30% at 200 lg/l U and levelling off at further increase of the uranium concentration towards 400 lg/l. The proportion of uranium absorbed also shows this plateau, leveling off
Table 1 Absorbate volume after 120 min of perfusion. Group
Concentration of perfusion solution lg uranium/l (AM ± SD)
ml absorbate/cm intestine over 120 min of perfusion (AM ± SD)
<50 lg uranium/l perfusion solution (P0)
A (n = 6) B (n = 6) C (n = 6)
5.15 ± 2.24 21.0 ± 0.44 43.1 ± 2.68
0.170 ± 0.029 0.181 ± 0.044 0.223 ± 0.077
>80 lg uranium/l perfusion solution (P0)
D (n = 9) E (n = 9) F (n = 7) G (n = 7)
99.1 ± 12.2 133 ± 9.36 199 ± 18.5 422 ± 24.2
0.347 ± 0.177 0.252 ± 0.092 0.239 ± 0.067 0.236 ± 0.047
Notes: n = number of intestinal segments perfused; AM = arithmetic mean; SD = standard deviation.
260
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
Table 2 Glucose concentration of perfusion solutions after the end of the 120-min perfusion period (PE) and of absorbates at 30-min intervals (glucose concentration of the perfusion solution before the start of perfusion: P0 = 15.1 mM). Group
Perfusion solution mM glucose (AM SD) PEafter 120 min perfusion
Absorbate mM glucose (AM ± SD) 30-min interval
60-min interval
90-min interval
120-min interval
<50 lg uranium/l perfusion solution (P0)
0 (n = 3) A (n = 6) B (n = 6) C (n = 6)
13.8 ± 0.6 11.8 ± 1.0 13.2 ± 2.4 12.4 ± 1.7
17.9 ± 3.3 34.1 ± 9.0 23.3 ± 6.0 28.2 ± 6.3
20.2 ± 3.3 38.4 ± 9.1 25.1 ± 3.1 29.1 ± 7.2
18.7 ± 2.2 34.4 ± 10.0 26.0 ± 4.0 26.3 ± 6.3
17.2 ± 4.4 34.0 ± 10.2 23.2 ± 6.1 23.4 ± 4.8
>80 lg uranium/l perfusion solution (P0)
D (n = 9) E (n = 9) F (n = 7) G (n = 7)
12.5 ± 1.5 13.9 ± 4.7 14.2 ± 1.2 13.6 ± 1.7
18.5 ± 5.0 17.7 ± 5.7 15.8 ± 2.2 18.1 ± 3.0
19.1 ± 4.1 18.2 ± 3.8 17.4 ± 2.9 18.6 ± 2.4
18.4 ± 4.9 19.5 ± 3.8 19.1 ± 5.0 18.8 ± 3.8
18.5 ± 4.8 21.1 ± 5.0 20.8 ± 5.8 19.2 ± 3.8
Notes: n = number of intestinal segments perfused; AM = arithmetic mean; SD = standard deviation.
Table 3 Uranium (U) concentrations of perfusion solutions before the start of perfusion (P0) and after 120 min perfusion (PE) and of absorbates. Group
<50 lg uranium/l perfusion solution (P0)
O (n = 3) A (n = 6) B (n = 6) C (n = 6)
>80 lg uranium/l perfusion solution (P0)
D (n = 9) E (n = 9) F (n = 7) G (n = 7)
lg/l U [AM ± SD (min–max)]
% uranium of P0 in absorbate
Perfusion solution (P0)
Perfusion solution (PE)
Absorbate (120 min)
0±0 5.15 ± 2.24 (3.47–8.47) 21.0 ± 0.44 (20.4–21.6) 43.1 ± 2.68 (39.9–47.7)
0 ± 0.0 2.35 ± 1.05 (1.47–4.13) 15.6 ± 2.68 (12.2–18.7) 35.5 ± 4.65 (29.5–41.3)
0 ± 0.0 0.41 ± 0.51 (0–1.25) 1.07 ± 0.75 (0.15–1.88) 6.13 ± 3.93 (2.22–11.3)
– 7% ± 17% (1%–47%) 5% ± 4% (1–9%) 14% ± 9% (5–26%)
99.1 ± 12.2 (88.6–118) 133 ± 9.36 (123–152) 199 ± 18.5 (179–224) 422 ± 24.2 (375–442)
99.7 ± 17.3 (73.5–134) 135 ± 12.1 (120–159) 199 ± 25.2 (172–241) 451 ± 22.1 (412–481)
19.8 ± 18.6 (2.88–57.6) 25.0 ± 16.6 (5.2–52.9) 56.8 ± 16.9 (24.1–74.7) 112 ± 47.9 (39.9–185)
20% ± 19% (3–55%) 19% ± 13% (4–42%) 29% ± 10% (11–42%) 26% ± 11% (9–43%)
Notes: n = number of intestinal segments perfused; AM = arithmetic mean; SD = standard deviation; min = minimum; max = maximum.
Fig. 2. Uranium concentrations in absorbates at low and higher perfusion solution concentrations.
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
261
Fig. 3. The proportion of uranium absorbed expressed as percent of uranium concentrations in the perfusion solutions (Data points are the group means shown in Table 3). For comparison, the regression functions for low and higher uranium concentrations in the perfusion solutions and the absorbates from Fig. 2 are included in this figure.
around 200 lg/l. The results of parallel tests of glucose absorption ruled out a concentration-dependent impairment of intestinal segment vitality. The advantage of these findings obtained with ex-vivo investigations is that they are independent of the physiological condition of the test animal. For example, withdrawal of food 12 h before preparation of the intestines may in principle increase the absorption measured in this study, as fasting increases uranium absorption in vivo (Sullivan et al., 1986; La Touche et al., 1987 and Bhattacharyya et al., 1989). This study excluded the possible impact of factors such as the diet and its iron content, as described by Sullivan and Ruemmler (1988), and the resulting concentration/ absorption ratios are therefore affected only by the relationships within the intestine and by the uranium. As the absorption of divalent cations is an active process, e.g. involving transport through the apical cell membrane (Powell et al., 1999), the marked increase in absorption at higher concentrations points either to an additional absorption mechanism or mode, or to elevated expression of the bound proteins, such as the divalent metal transporter 1 (DMT1). DMT1 is known to be triggered by the iron content of the diet (Park et al., 2002; Gunshin et al., 1997; Canonne-Hergaux et al., 1999), and it may also be triggered by fasting (Sullivan et al., 1986; Konietzka, submitted for publication). As these factors were excluded in the case of the investigation method used here, two possible explanations remain: at the higher uranium concentrations an additional absorption mode or absorption mechanism, such as passive diffusion (Dublineau et al., 2005), may be more prominent, or transport proteins that may be involved (Okazaki et al., 2012; Nadadur et al., 2008; Au et al., 2008; Yokel et al., 2006) are triggered by the uranium itself as part of a concentration-dependent process. The fact that non-linear functions describe the data only slightly poorer and also seem biologically plausible tends to support the second possibility. However, both hypotheses require further testing. For regulatory purposes, the World Health Organisation and other authors assumed relatively low in-vivo absorption rates ranging from 0.3% to 1.5% (WHO, 2012; Konietzka, 2014; Konietzka et al., 2005). However, for the purposes of risk quantification our results indicate that absorption rates could be signifi-
cantly higher for uranium concentrations above 20 lg/l than at lower concentrations. The resulting higher internal exposures appear relevant for the setting of Guideline values in this concentration range because higher absorptions rates will cause higher levels in the target organ (kidneys). A lower Guideline value would be appropriate to account for this mechanism. Conflicts of interest None. Acknowledgments We thank Mr. Matthias Skerswetat for his experimental work and Mr. Ulrich Lippold for the uranium analyses. We are grateful to Dr. Ingrid Chorus and Dr. Tamara Grummt for helpful comments. References Au, C., Benedetto, A., Aschner, M., 2008. Manganese transport in eukaryotes: the role of DMT1. Neuro Toxicology 29, 569–576. Bhattacharyya, M.H., Larsen, R.P., Cohen, N., Ralston, L.G., Moretti, E.S., Oldham, R.D., Ayrest, L., 1989. Gastrointestinal absorption of Plutonium and Uranium in feed and fasted adult Baboons and mice: application to humans. Radiat. Prot. Dosimetry 26, 159–165. Canonne-Hergaux, F., Gruenheid, S., Ponka, P., Gros, P., 1999. Cellular and subcellular localization of the Nramp2 iron transporter in the intestinal brush border and regulation by dietary iron. Blood 93 (12), 4406–4417. Dublineau, I., Grison, S., Baudelin, C., Dudoignon, N., Souidi, M., Marquette, C., Paquet, F., Aigueperese, J., Gourmelon, P., 2005. Absorption of uranium through the entire gastrointestinal tract of the rat. Int. J. Radiat. Biol. 81, 473–482. Dublineau, I., Grison, S., Linard, C., Baudelin, C., Dudoignon, N., Souidi, M., Marquette, C., Paquet, F., Aigueperse, J., Gourmelon, P., 2006. Short-term effects of depleted Uranium on immune status in rat intestine. J. Toxicol Environ. Health, Part A: Current Issues 69, 1613–1628. EFSA, European Food Safety Authority, 2009. Uranium in foodstuffs, in particular mineral water. Scientific opinion of the Panel on Contaminants in the Food Chain.
262
R. Konietzka et al. / Regulatory Toxicology and Pharmacology 69 (2014) 256–262
Konietzka, R., 2014. Gastrointestinal absorption of uranium compounds – a review, Regul. Toxicol. Pharmacol., submitted for publication. Konietzka, R., Dieter, H.H., Voss, J.-U., 2005. Vorschlag für einen gesundheitlichen Leitwert für Uran in Trinkwasser. Umweltmed Forsch Prax 10, 133–143. La Touche, Y.D., Willis, D.L., Dawydiak, O.I., 1987. Absorption and biokinetics of U in rats following an oral administration of Uranyl Nitrate solution. Health Phys. 53, 147–162. Leggett, R.W., Harrison, J.D., 1995. Fractional absorption of ingested uranium in humans. Health Phys. 68, 484–498. Nadadur, S.S., Srirama, K., Mudipa, A., 2008. Iron transport & homeostasis mechanisms: their role in health & disease. Indian J. Med. Res. 128, 533–544. Okazaki, Y., Ma, Y., Yeh, M., Yin, H., Li, Z., Yeh, K., Glass, J., 2012. DMT1 (IRE) expression in intestinal and erythroid cells is regulated by peripheral benzodiazepine receptor-associated protein 7. Am. J. Physiol. Gastrointest. Liver Physiol. 302, G1180–G1190. Park, J.D., Cherrington, N.J., Klaassen, C.D., 2002. Intestinal absorption of cadmium is associated with divalent metal transporter 1 in rats. Toxicol. Sci. 68, 288–294. Powell, J.J., Jugdaohsingh, R., Thompson, R.P.H., 1999. The regulation of mineral absorption in the gastrointestinal tract. Proc. Nutr. Soc. 58, 147–153. Richter, E., Strugala, G.J., 1985. An all-glass perfusator for investigation of the intestinal transport and metabolism of foreign compounds in vitro. J. Pharmacol. Methods 14, 297–304. Schümann, K., Osterloh, K., Forth, W., 1986. Independence of in vitro iron absorption from mucosal transferrin content in rat jejunal and ileal segments. Blut 53, 391– 400.
Sullivan, M.F., Ruemmler, P.S., 1988. Absorption of 233U, 237Np, 238Pu, 241Am and 244Cm from gastro intestinal tract of rats fed an iron-deficient diet. Health Phys. 54, 311–316. Sullivan, M.F., Ruemmler, P.S., Ryan, J.L., Buschbom, R.L., 1986. Influence of oxidizing or reducing agents on gastrointestinal absorption of U, Am, Cm and Pm by rats. Health Phys. 50, 223–232. Tasat, D.R., Orona, N.S., Bozal, C., Ubios, A.M., Cabrini, R.L., 2012. Intracellular metabolism of uranium and the effects of bisphosphonates on its toxicity. In: Paula Bubulya (Ed.), Cell Metabolism – Cell Homeostasis and Stress Response. pp. 116–148. Rijeka, Croatia: InTech. ISBN 978-953-307-978-3. TrinkwV, TrinkwasseränderungsV (2011): Erste Verordnung zur Änderung der Trinkwasserverordnung vom 03. Mai 2011. Bundesgesetzblatt, Jahrgang 2011, Teil I Nr. 21, ausgegeben zu Bonn am 11. Mai 2011. WHO, World Health Organization, 2011. Guidelines for drinking-water quality, fourth ed.