Introduction

Nutrition will affect functional outcomes in many studies performed with rodent models. This may already be the case with strong pharmaceutical or toxicological interventions, or when dominant gene knock-outs or knock-ins are considered. When we take obesity, type 2 diabetes, or cardiovascular diseases into consideration, which are multifactorial diseases with a strong dietary background, nutrition becomes even more important. Indeed, dietary intervention studies in wild-type mice and rats were found to give relatively small effects on a variety of parameters such as lipoprotein profiles, biomarkers of inflammation and gene expression responses (Keijer et al. 2010; Patsouris et al. 2006). In addition, “standard” chow diets usually contain poorly specified ingredients that may strongly vary in composition between batches and between providers. For example, a 5,000-fold difference in the level of phytoestrogens has been encountered (Leiter 2009; Thigpen et al. 1999). Moreover, when a chow diet is compared with a semi-purified diet, a remarkable difference in digestibility and thus effective energy intake has been described (Ortmann et al. 2003).

To be able to compare functional effects of different nutrients, it is important to use standardized, semi-purified diets; both as experimental as well as its reference, control diet. This necessity is now widely recognized among researchers employing rodents to unravel molecular mechanisms underlying functional effects of nutritional components (Baur et al. 2006; Daniel and Dieck 2004). Indeed, semi-purified control and experimental diets based on the recommendations of the American Institute of Nutrition (AIN93) are increasingly used. While of great importance, these guidelines leave substantial room for variation in a number of dietary constituents that are of significant relevance to molecular and health outcomes of dietary intervention studies (Keijer et al. 2010).

A need to further standardize diets is strongly felt in order to improve comparison of study outcomes and to increase efficiency of resources and animals (Barnard et al. 2009). It is, however, difficult to reach consensus since it implies that individual laboratories have to perform additional initial comparative analyses in ongoing research lines. Despite this, partners in two European research consortia, MitofoodFootnote 1 and BIOCLAIMS,Footnote 2 consisting of, respectively, 38 and 11 partners spread throughout 24 countries in Europe have proposed a common, defined semi-purified AIN93-based reference diet, the BIOCLAIMS Standard Diet (BIOsd), as well as an experimental high-fat diet. Here, we present the choices that have been made, including the underlying rationale, as well as a number of dietary analyses that have been performed on the final formulation of the standard diet.

Materials and methods

Gross energy determination

Approximately 10 g of BIOsd (Research Diets Services, Wijk bij Duurstede, NL) and standard chow (Teklad global 16% protein rodent diet, 2016) were grinded on a 1.0-mm sieve to prepare a homogenate sample for gross energy determination. Samples were left at room temperature for 24 h to acclimatize for humidity. Samples were measured by IKA C7000 calorimeter (Staufen, Germany) in triplicate or until the deviation was less than 2%. Polyethylene bags were used as standard.

Kahl pellet hardness tester

Hardness of pellets was assessed by using a Kahl device (Amandus Kahl Nachf., Reinbek/Hamburg, Germany). A pellet is inserted between two bars and due to increasing static pressure applied by a spring, the force (H, kg) necessary for breakage of the pellet is determined. The average of 12 measurements minus the hardest Hmax and the least hardest Hmin of pellets randomly chosen from the batch is referred to as “Kahl-hardness” of the pellet, which is given by the following formulae:

$$ {\text{Hardness}}( {\text{kg}} ) = \frac{\sum {n_{1 - 12} } \,{\text{Hardness }}({{\text{H}})_{1 - 12}} - {\text{Hmax}} - {\text{Hmin}} }{10}. $$

Here, measured pellets differed however in shape. Regular chow is oval shaped (16 × 10 mm) while the semi-purified diet is round (9 mm diameter). Chow was tested for breakage on the “flat” side.

Animal experiment

Male C57BL/6JRccHsd mice were delivered from Harlan Laboratories GmbH to the German Institute of Human Nutrition (DIfE) at 4 weeks of age. Mice were housed individually with ad libitum access to food and water at a temperature of 22°C on a 12 h light–dark cycle. At 6 weeks of age, mice were switched from standard chow diet (Ssniff) to the BIOsd (Research Diets Services, Wijk bij Duurstede, NL). Body weight and composition was determined weekly for 12 weeks using quantitative magnetic resonance (QMR) (Bruker’s Minispec MQ10, Housten Texas, USA) as described (Klaus et al. 2005). Animal maintenance and experiments were approved by the animal welfare committee of the Ministry of Agriculture and Environment (State of Brandenburg, Germany).

All experiments were performed in accordance with the directive of the European Communities Council (68/609/EEC) and the Principles of Laboratory Animal Care (NIH publication no. 85-23, revised 1985).

Results and discussion

An inventory of changes in nutrient composition compared with the original AIN93 diet was made, and the grounds for making these changes are given below. The general formula of the BIOsd in comparison with AIN diets is shown in Table 1, and the mineral and vitamin mixes of the BIOsd are shown in Tables 2 and 3, respectively.

Table 1 BIOCLAIMSsd
Full size table
Table 2 Contribution of minerals
Full size table
Table 3 Contribution of vitamins
Full size table

Fiber

Cellulose is used at 50 g/kg according to AIN-93G and AIN-93M (Table 1).

Macro nutrients

Carbohydrates

In the BIOsd, corn starch has been replaced by wheat starch. This has been done due to a high prevalence of natural stable isotope 13C in corn. This change makes the diet suitable for metabolic tracer studies using natural enriched corn (Friedrich et al. 2011; Isken et al. 2010; Schmidt and Metges 1985). Furthermore, a combination of maltodextrin, dextrose and sucrose has been used to obtain a pellet with a pleasant hardness for mice (Koopman et al. 1989) (BIOsd vs. chow, Table 4).

Table 4 Pellet hardness (kg)
Full size table

Protein

For the BIOsd a choice has been made to use 220 g protein/kg diet, supplemented with 3 g l-Cysteine, corresponding to a total of 23en% of protein. This protein content avoids potential hyperphagia due to lower protein content (Du et al. 2000). l-Cysteine is added to meet the requirements of sulfur amino acid levels (Reeves et al. 1993). The increase in casein also covers the increase in recommendation of sulfur amino acids according to the National Research Council nutrient requirements of Laboratory Animals (National Research Counsil 1995).

Fat

The choice for fats was based on a number of criteria. First, the diet should provide a balance of the essential fatty acids, linoleic acid (omega-6 fatty acid, n-6;18:2) and linolenic acid (omega-3 fatty acid, n-3;18:3) acid. Bourre et al. suggest a minimal intake of 12 g linoleic acid and 2 g α-linolenic acid/kg diet for rats, which has been recommended after revision of the original AIN-93 (Bourre et al. 1989; Reeves 1997). Secondly, a healthy polyunsaturated to saturated fat (P/S) ratio of at least 2 should be attained (Lee et al. 1989). Thirdly, we selected oils containing neither polyphenols nor carotenes to be able to sensitively intervene with these food bioactive compounds. To meet all these criteria, a combination of sunflower oil (70%), coconut oil (18%) and flaxseed oil (12%) has been selected. This results in a (n-6)/(n-3) ratio of 7.8 and a P/S ratio of 3.1 (Table 5). The use of soy oil was considered not appropriate due to its variable polyphenol content. Finally, the gross energy content of the BIOsd diet versus chow is shown in Table 6.

Table 5 Fatty acid profile (% of g/kg diet)
Full size table
Table 6 Gross energy content kJ/g
Full size table

In addition, it was decided to optionally add 30 mg/kg cholesterol to the diet (investigator depended), since human diets usually contain cholesterol. This amount of cholesterol corresponds to the amount of cholesterol present in approximately 30 g lard/kg diet and represents average Dutch human intake (Van Schothorst et al. 2005). To our knowledge, there are no known side effects if cholesterol is excluded from the diets.

Micronutrients

Vitamin and mineral mix

Vitamin and mineral contents follow AIN-93 recommendations and are specified in Tables 2 and 3, respectively. The carbohydrate component that is used as a carrier for the vitamin provides 40 kcal/kg of energy, which has been taken into account for calculating the carbohydrate energy contribution and total energy content (Table 1).

Polyphenols and carotenoids

The oils that are used contain no carotenoids. Sunflower oil contains 1 mg polyphenols per 100 g oil (De Leonardis et al. 2003), according to the French National Institute for Agricultural Research (INRA) Phenol Explorer Database (INRA 2009). This results in 0.3 mg of polyphenols/kg diet (3 × 10−5%), which is considered to be a negligible amount. Sunflower oil contains 41.08 mg/100 g of α-tocopherol (vitamin E) (US Department of Agriculture and Service 2010).

Performance in mice

Using BIOsd during maintenance, similarly to the use of other purified diets, the mice initially lost some weight, as a consequence of the diet switch. After this, the mice displayed a fully normal development and behavior. This was also the case for rats (data not shown).

Figure 1 shows the development of body composition of C57BL/6J mice, which were fed the BIOsd diet from 6 weeks of age for 12 weeks. At 6 weeks (week 6) of age, mice are not yet fully grown. During feeding of the BIOsd, mice gained body weight due to an increase in lean body mass as well as body fat as expected (body mass (g) at week 5: 19.95 ± 0.24, at week 18: 24.45 ± 0.46; lean mass (g) at week 5: 14.65 ± 0.20, at week 18: 17.29 ± 0.34; body fat (g) at week 5: 5.30 ± 0.11, at week 18: 7.16 ± 0.17, data are mean ± SEM, n = 12).

Fig. 1
figure 1

Body weight and body composition development of C57BL/6J mice fed BIOsd ad libitum for 12 weeks from 6 weeks of age. Data are mean ± SEM, n = 12

Full size image

The emphasis on a purified diet to reduce experimental variation becomes more and more important due to the use of new technologies such as genomics, proteomics and metabolomics, where small differences can be of major significance for interpretation of the data. But enhanced standardization of general used diets is still necessary for the discovery of effects of phytoestrogens as a powerful variable. Phytoestrogen content of equal open-formula diets can differ up to threefold depending on mill date and use of different commodities (Thigpen et al. 1999). Genistein, a phytoestrogen present in soy for example, displays estrogen receptor activation properties, which are an unwanted effect in dietary studies (Ren et al. 2001). Not only the protein fraction, but also the fat fraction can contain phytoestrogens. Traditional diets often contain soybean oil for its adequate (n-6)/(n-3) ratio. Furthermore, for a proper health and physiological state of the rodents, general rules for n-6 and n-3 fatty acids and P/S ratio exclude usage of, for instance, palm oil as sole dietary fatty acid ingredient due to a (n-6)/(n-3) ratio of 25:1, which is considered to increase the probability of a number of diseases (Simopoulos 2002).

As many partners are interested in the use of a high-fat diet, we also propose a high-fat diet based on the BIOsd, in which a part of the carbohydrate content of wheat starch will be replaced by fat to mimic average daily western human intake (40 en% fat, 4,700 kcal/kg).

The initial studies that were performed, and those that are still ongoing, show that the diet can be used for maintenance. We have not examined long-term or trans-generational effects, and therefore at present, we do not advise to use the diet as a standard diet in breeding unless further tested. If used during breeding, we recommend to mesh the diet into porridge for easy access for the dams as the BIOsd is harder than standard chow (Table 4); however, it has been reported that this is not necessary for mice post-weaning (Koopman et al. 1989). Indeed, as the data shows, it is adequate for maintenance.