Introduction

Many countries in Europe have reported a downward trend in their populations’ dietary intakes of selenium [1–3]. This is due, in part, to changes in the dietary habits of certain sub-groups within the general population, e.g. reduced calorific intakes and increased consumption of highly processed foods. However, the most significant impact is due to the substitution of North American wheat imports by produce from within Europe [3].

The health implications of this trend are deemed significant because of the reported involvement of selenium in many structural, metabolic and enzymatic processes within the body [4]. The element has been reported to have an involvement in restraining the progression and virulence of certain viral infections [5–7] and, possibly most significantly, it is widely reported for its anti-cancer effects [8, 9]. It has also been connected with correct immune function [10], anti-ageing [11] and male reproductive performance [12]. Any reduction in the dietary intakes of selenium may therefore be seen to be of nutritional and/or clinical concern.

In this study rye/wheat sourdough bread was identified as a potential vehicle for selenium supplementation because it is a dietary staple in many Eastern European countries where, incidentally, intakes of the element can sometimes be low [2]. Advances in the technology associated with the production of the sourdough bread have facilitated enrichment of the product with elevated levels of selenium, mainly in an organic form [13]. Because bread is a staple food, it is often used for dietary supplementation, e.g., with folic acid [14].

Therefore, the aim of the human study described here was to investigate whether consumption of the selenium-enriched sourdough bread had a positive effect on the body’s selenium status and was, therefore, an efficacious dietary source of the element.

Materials and Methods

Subjects

Twenty four female volunteers, aged between 24 and 25 years, were recruited from students at the Technical University of Łόdz, Poland. Informed, written consent was obtained prior to taking a screening blood sample to eliminate those volunteers whose biochemical and haematological indices fell outside of the normal range. The volunteers agreed to exclude selenium-rich supplements and foods during the course of the study but, in all other respects, they were allowed to maintain their normal lifestyle and dietary habits. The Ethical Committee of the Military Medical Academy (WAM) in Łόdz, Poland approved the study.

Test Meals

Two mixed wheat/rye sourdough breads were produced. The rye flour was made from dried and ground rye seedlings which had been grown in either a selenium-rich environment or one containing only normal levels of the element (referred to in this paper as enriched and control material, respectively). Details of the bread making process, including the incorporation and biotransformation of the selenium, are described elsewhere [13].

Bread rolls were baked in sufficient amounts so as to provide enough test material for the duration of the 28-day feeding trial. The breads were dried, at 40°C, homogenised and stored until required for feeding to the volunteers. This removed the possibility of between-roll variability affecting the precision of the selenium utilisation data. The levels of selenium in the control and enriched breads were 0.04 and 3.58 μg g−1 respectively.

Study Design

On each test day, following an overnight fast, the volunteers were fed 19 g of the bread assigned to their particular group, equating to an additional 0.84 and 68.02 μg of selenium for the control and enriched group, respectively. Feeding of the bread was stopped after t = 4 weeks. On the morning of the first test meal (and at the same time 1, 2, 4 and 6 weeks later) samples of blood were drawn by venepuncture into tubes. The tubes (Medlab, Raszyn, Poland) contained sodium citrate (3.2% w/v), and the blood to citrate ratio was 9:1.

Analytical Methods

All reagents were analytical-grade or better (Sigma-Aldrich Company Ltd, Poole, Dorset, UK) and the water was of Millipore-grade (>18.2 MΩ cm−1).

Preparation of Blood Fractions

Fresh blood samples were centrifuged to separate the plasma and plasma platelets from the whole blood. To obtain the plasma fraction, blood was centrifuged at 1,500×g for 5 min. To obtain the platelet fraction, samples of blood had to undergo a two-stage separation. Firstly, centrifugation at 100×g for 10 min, followed by collection of the supernatant which was then centrifuged again, at 900×g for 15 min. The platelet-containing pellet was washed 3 times using Hanks buffer (each washing was followed by centrifugation at 900×g for 10 min). The washed pellet was suspended in saccharose solution (2 ml, 0.32 M). All samples were stored at −70°C until required for analysis.

Measurement of Platelet Glutathione Peroxidase-1 (GPx1) Activity

Immediately prior to measurement of platelet GPx1 activity, careful defrosting of the samples was performed (2 h, 4°C) so as to minimise the possibility of changes due to sample-handling differences. The platelets were disrupted by sonication prior to centrifugation (20,000×g for 10 min, 4°C). Measurement of the platelet GPx1 was performed using the method described by Belsten and Wright [15]. Plasma protein concentrations were determined using a colorimetric method [16] and the data used in the normalisation of the GPx1 activity (expressed as U/mg protein).

Measurement of Selenium

Details of measurement of total selenium by hydride generation-inductively coupled plasma-mass spectrometry (HG-ICP-MS) are described elsewhere [13].

Selenium Speciation Analysis

The methods used to characterise the selenium species in the bread samples are described in detail elsewhere [13].

Quality Assurance and Quality Control (QA/QC) Procedures Undertaken during the Measurement of Total Selenium in the Plasma Samples

All stages of the analytical procedures used in the determination of total selenium in the plasma and bread samples were accredited by the United Kingdom Accreditation Scheme (UKAS) to ISO-17025. Each acid digestion batch included the following QA and QC materials: Seronorm-human serum and NIST 2670-freeze-dried urine (both from LGC Promochem, Teddington, UK), at least 4 procedural blanks and 1 spiked procedural blank.

QA/QC Procedures Undertaken during the Speciation Analysis of the Bread Samples

There are no certified reference materials available for individual selenium species in foods. However, candidate reference materials (wheat flour and yeast—EU funded project: G6RD-CT-2001-00473/SEAS) were used as part of the QA procedures accompanying the selenium speciation analysis of the experimental breads [13].

QA/QC Procedures Undertaken during the Measurement of Platelet GPx1 Activity

A quantity of human blood was taken prior to the start of the human study, for use as an in-house reference material (IHRM) in the measurement of platelet GPx1 activity. The platelet fraction was collected, prepared as above, sub-sampled into vials and stored at −70°C. Aliquots of the IHRM were analysed alongside the volunteers’ platelet samples, and the data used to normalise the test data, thereby minimising between-batch analytical variability.

Statistical Analysis

A repeated measures analysis of variance was performed with the applied treatment as a factor. This analysis takes into account the within-subject variation, i.e. the correlation between successive time points on the same subject. In addition, formal comparisons between the controlled and enriched groups were made at each time point of the study, using the estimate of variation calculated in the repeated measures analyses. Analyses of selenium and GPx1 were performed separately in the form of concentration and percentage change from baseline (week 0), using GenStat8.

Results

Human Study Data

Table 1 presents the plasma selenium concentration and the plasma platelet GPx1 activity data for the 24 volunteers, during the 6 weeks (5 time points) of the intervention study.

Table 1 Plasma selenium levels and platelet GPx1 activities; mean ± SD and (range), measured in the volunteers fed either enriched or control sourdough breads, and statistical data relating to the change observed in the two selenium status markers
Full size table

Table 2 reports the statistical data resulting from the analysis of the mean percentage changes observed in the two selenium status markers. The important values to be noted in the table are highlighted in bold (F-prob for time by treatment interaction), and reflect the effect of feeding the enriched bread over the whole period of the study.

Table 2 Statistical data from the analysis of the percentage change calculated in the enriched group volunteers’ plasma selenium concentrations and plasma platelet GPx1 activities
Full size table

Speciation of Selenium in the Breads Fed to the Human Study Volunteers

Figure 1 shows the HPLC-ICP-MS chromatograms obtained from the enriched and control breads. In Fig. 1a the selenomethionine peak is at t = 212 s, but is not present in Fig. 1b, i.e., the chromatogram obtained from the non-enriched control bread. The identity of this peak was confirmed by electrospray-mass spectrometric (ES-MS) analysis, and was shown to account for 42% of the total Se-containing peak areas. It was not possible to establish the identity of the other peaks in the chromatogram due to their low concentrations in the breads. However, what can be said is that their retention times do not correspond to those of the inorganic selenium forms, selenate or selenite [13].

Fig. 1
figure 1

HPLC-ICP-MS chromatograms showing the selenium profiles of the enriched and control breads; a is the chromatogram obtained from the selenium-enriched bread and b is the chromatogram obtained from the control bread

Full size image

Analytical QA/QC Data

The recovery of selenium from the CRMs was as follows; Seronorm = 97 ± 8.5% (n = 37) and NIST2670 = 98 ± 7.4% (n = 42). Mean values for the recovery of spike were 101 ± 8.2% (n = 18). A limit of detection (LOD) for the procedure was 5 ng g−1.

QA Data Obtained during the Speciation Analysis of the Bread Samples

The percentage extraction efficiency for the candidate reference material was 64.8 ± 2.5% and the LOD for the procedure was calculated to be 0.002 μg g−1 (reported as selenomethionine).

Discussion

Plasma Selenium Concentration

From Tables 1 and 2, it can be seen that the calculated mean percentage change value shows that the difference between the control and enriched groups was highly significant over the duration of the study (p < 0.001). The percentage change in the mean concentration of the two groups was similar at week 1, but was significantly greater at weeks 2, 4 and 6. It should be noted that even 2 weeks after cessation of the feeding stage, the samples at week 6 still showed elevated plasma selenium levels. This suggests that the absorbed selenium had been successfully incorporated into the volunteers’ body reservoirs and was then being slowly released back into the blood.

Platelet GPx1 Activity

Tables 1 and 2b shows that there was no statistically significant difference in the mean percentage change observed in the platelet GPx1 activity between the control and enriched groups during the study (p = 0.756). This may be due to the plasma selenium levels, at the start of the feeding regime, being sufficiently high [mean ± SD: 56.3 ± 6 (μg L−1) and range: 48–65 (μg L−1)] as to have already induced maximal expression of the enzyme. However, using the calculation used by Zachara and Wasowicz [17], the average plasma selenium level of 58.2 μg L−1 (measured in the t = 0 samples of the 24 volunteers) equates to a dietary selenium intake of ∼39 μg day−1. This value is significantly below the 60 μg day−1 Reference Nutrient Intake for females proposed by the UK government Department of Health in 1991 [18] and recommended in ‘Guidelines for Poland’ [19]. It is therefore uncertain whether the selenium status of the volunteers would have been sufficiently high to result in maximal GPx1 expression prior to the study onset.

An alternative explanation is that, as the chemical form of the selenium in the enriched bread was shown to be mainly present in organic forms of the element, i.e. selenomethionine and selenium-containing proteins [13, 20], the volunteers’ GPx1 status would not be significantly affected, particularly at the conservative levels of supplementation used in this study. This hypothesis is supported by Neve [21], who concluded that, although organic forms of selenium increased blood-selenium concentrations to a greater extent than inorganic forms, the latter had a greater effect on GPx production. Brown et al. [22], who fed volunteers pure selenomethionine, reported that GPx activity showed little or no response to supplementation, despite the plasma selenium concentration increasing. Similarly, from a study where long-term supplementation with selenate and selenomethionine had been carried out, Thomson et al. [23] concluded that selenomethionine was quickly incorporated into a general tissue-protein pool, while inorganic selenate was more directly available for GPx synthesis. Neve et al. [24] reported that GPx1 activity was only affected after 45 days of dietary intervention with selenomethionine (at 100 μg/day), a conclusion which may explain the response observed in the current study.

The work presented here was part of a multi-disciplinary investigation into the physicochemical properties and bioavailability of selenium from enriched sourdough bread, developed as a potential vehicle for dietary supplementation. If the issue surrounding the lack of effect on GPx1 status is to be addressed, any future study would need to be conducted over a much longer intervention period (possibly up to 3 months) and at a higher level of enrichment (probably 100 μg day−1).