An excellent example of a crazy protocol for unreliable results

"Preparation of experimental diets

​Cholesterol (C75209; Sigma) was dissolved in ​diethyl ether (Sigma) to create a 10% (or a 5%) solution, of which 400 μl was added to 0.5 g of standard zebrafish larval food (ZM; ZM Systems, ingredients: protein 52%, oil 12%, ash 8%, moisture 7%, fibre 3%) or adult food (Hikari, Hikari, ingredients: protein 49%, oil 7.8%, ash 11%, moisture>10%, fibre 0.9%, phosphorus 1.7%), to create ​cholesterol-enriched diets for acute and extended feeding, respectively. Four hundred microlitres of ​diethyl ether was added to 0.5 g of ZM or Hikari to serve as a control diet. The diets were left overnight for the ether to evaporate and ground up the following day into fine particles using a pestle and mortar (adapted from Stoletov et al.³⁶). For the preparation of sterile food (sZM, sHCD) for GF experiments, ZM was autoclaved before the supplementation with ​cholesterol as described above, which was performed under sterile conditions.

Feeding of zebrafish larvae and adults

Before experimental procedures, larvae were initially reared at 28.5 °C at a maximum density of 50 larvae per Petri dish in system water containing 3 × 10⁻⁵% ​methylene blue (Sigma) as an antifungal agent and 30 mg l⁻¹ ​1-phenyl 2-thiourea (Sigma) to prevent melanization (except for ​Tra^−/−;​Nac^−/− mutant). Great care was taken to remove unfertilized eggs and chorions post hatching. Zebrafish larvae (6 dpf) were placed in system water containing clotted cream (63.5 g of fat, 170 mg of ​cholesterol, 2.2 g of carbohydrate and 1.6 g of protein per 100 g) used at a 1:10 dilution, ​cholesterol, or a standard diet (ZM) for 6 h at 28.5 °C¹⁸. The number of neutrophils and macrophages was quantified at indicated time points after the 6-h-feeding period by counting cells (in a blinded manner) that were present in the intestine of live Tg(​mpx:GFP), Tg(​lyz:dsRed) andTg(​fms:mCherry) transgenic zebrafish, respectively, whereas myeloid cells were quantified in the intestine of WT zebrafish after fixation and immunolabelling with an ​L-plastin antibody. Significant differences in myeloid cell accumulation were found using doses of ≥4% ​cholesterol with groups of about 30 larvae and power calculations determined significant differences for 2% ​cholesterolwhen using n=190 (Supplementary Fig. 2a) These concentrations of ​cholesterol correspond to doses given to genetically non-predisposed mice²⁷. Adult zebrafish were starved for 48 h before feeding for 6 h with ​cholesterol or standard diet (Hikari). For extended feeding, zebrafish larvae were fed twice a day for 10 days with ​cholesterol-enriched or standard ZM diet (seeSupplementary Fig. 13a for protocol). For experiments studying lipid deposition in the caudal vein, ZM and HCD were supplemented with 10 μl ml⁻¹ of a fluorescent lipid probe, cholesteryl BODIPY 576/589 C11 (Invitrogen) according to Stoletov et al.³⁶ For all experiments involving feeding, zebrafish larvae were randomly assigned to the various treatment groups. After randomized allocation of animals, treatments were coded until the experiment was finished and data were analysed. A sample size of about 15 zebrafish larvae per treatment group in which the readout was quantification of intestinal inflammatory cells was calculated based on the magnitude ±s.d. (time zero before feeding: 8.3±4.1; 18 h post ZM feeding: 11.09±5.05; 18 h post HCD feeding: 19.57±9.11) identified in a pilot study and was used for all further experiments involving feeding."

http://www.nature.com/ncomms/2014/141223/ncomms6864/pdf/ncomms6864.pdf