Intraperitoneal injection involves five steps: fasting, weighing, anesthetizing, injection, and recovery. For each step there are best practices that can ensure success. Success includes a healthy fish patient as well as a good experimental outcome.

Fasting: A 24-hour fast should empty the intestinal bulb. This practice is taken from the fish veterinary literature (e.g., Brown 1993). Additional fasting considerations are discussed below.

Longer-term fasting: We have found that a 72-hour fast is required to decrease blood glucose to a baseline level prior to injection (Eames et al., 2010). We have also found that for glucose studies there are several procedures that are required to ensure that the fish are fasted properly. Start with a clean tank (no debris on the bottom). Tanks should be offline, clearly labeled as 'fasting', and in a location where enthusiastic fish care personnel will not feed them. Evaluate the external environment of the tank and take steps to prevent the fish from being stressed from disturbances, as stress is known to raise blood glucose (Chavin and Young, 1970; Groff et al., 1999). For example, we had a fasting experiment in which a radio was operated daily on the bench that was holding the fish tanks. We found that blood glucose was unusually high and concluded that the fish were stressed by the vibrations. Another stressor is overcrowding. Fish should be kept at a density that conforms with good fish husbandry practices. For recommendations, see Brand et al. (2002) and Westerfield (1995). We have had good results fasting our fish at a density of 10-12 fish in a 9 liter tank (with 3 layers of marbles taking up some of that volume). Separating the sexes may cause stress, so we recommend maintaining a mixed-sex population during the fast. This means that eggs can be laid, and the eggs need to be sequestered so that they will not be eaten. A simple way to sequester eggs is to cover the tank bottom with 2-3 layers of marbles. Water quality needs to be maintained by removing eggs and waste and by replacing about 10-15% of the tank water, daily. For removing eggs and waste, siphoning works well.

Weighing: When weighing fish that are not anesthetized, care should be taken to minimize water transfer from the net into the beaker, to ensure accurate weighing. If the net (with fish) is blotted on paper towels, the majority of the excess water can be removed, and the weight can be accurately measured. It may be easier to anesthetize the fish prior to weighing, but we have not tested the possible effects of anesthetizing a fish twice in one day. We have tested our technique by weighing the fish first with the netting/blotting method and then re-weighing the fish after it has been anesthetized, and gently blotted dry. We found no significant difference in weight between the methods (P = 0.7927, t-test). Additionally, we tested whether this netting/blotting method affected blood glucose, in comparison with simply transferring the fish to the beaker as soon as it is netted (no blotting). We found no significant difference in blood glucose level between the two transfer methods (P = 0.2241, t-test).

Anesthetizing: Chemical anesthesia may be suitable for many studies. Here we have demonstrated cold water anesthesia as an alternative, because many anesthetics (including tricaine/MS-222 (Brown, 1993)), raise blood glucose. In previous studies, we have determined that cold water does not raise blood glucose in zebrafish (Eames et al., 2010). For cold water anesthesia, the temperature should be decreased slowly. The rate of decrease seems to depend on the size of the fish, with smaller fish going under faster than larger fish. Following injection, you may observe that the fish is recovering too slowly from the anesthetic (see below). This can result when either the starting temperature is too low, or when the temperature is decreased too rapidly. The starting temperature is too low if the fish bends laterally upon entering the water. If the starting temperature is correct, the fish will maintain its balance initially. It will rotate its pectoral fins to a horizontal position, gasp, and have rapid operculum movements. Typically, it will swim. As temperature decreases, movements will decrease and the fish will lose equilibrium. A surgical plane of anesthesia is reached when the fish can be handled without reacting. To maintain the fish under surgical anesthesia, your fingers must be cold, so keep them in the water prior to handling the fish. The sponge must also be kept cold at the same temperature as the water used for anesthetizing the fish. It is important to saturate the sponge with water that is sufficiently cold to maintain anesthesia once the fish is placed onto it.

Injection: Prior to undertaking injections, you may want to dissect at least one fish to get a sense of body wall thickness. This can help you to judge how far the needle needs to insert to enter the abdominal cavity. Additionally, as you insert the needle, you can feel the body wall "give" when the needle enters the abdominal cavity. During the injection, take steps to keep the patient happy. Make sure the sponge is saturated with the correct temperature cold water to prevent the fish from reviving during injection. A well-saturated and soft sponge is important for minimizing damage to the scales and mucus covering of the skin. A well-saturated sponge is also important for keeping the gills aerated. We highly recommend the foam sponge listed below under Materials. Finally, once the fish is anesthetized, work quickly to minimize the time that the fish is under.

Recovery: The fish should recover from the anesthesia virtually upon entering the warm tank water. If the fish does not begin swimming immediately, gently swirl the water towards its gills to speed recovery. If recovery is slow, then the fish went under too quickly and you should adjust the anesthesia procedure appropriately. The possible causes of slow recovery are discussed under Anesthetizing.

No conflicts of interest declared.

This study was supported by Juvenile Diabetes Research Foundation grant 5-2007-97 (to V.E.P.), by National Institute of Diabetes and Digestive and Kidney Diseases grants R01DK064973 (to V.E.P.), R01DK48494 (to L.H.P.), T32DK07074 (supporting S.C.E.), K01DK083552 (to M.D.K), and by P60DK20595 to The University of Chicago Diabetes Research and Training Center. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIDDK or the NIH.