Extraction and recovery of inositol phosphates from tissues.

The recent interest in inositol phosphates as second messengers (e.g. Berridge & Irvine, 1984) has led to many experiments being performed using myo-['H]inositol to label cells, followed by stimulation and subsequent quantification of radiolabelled inositol phosphates (e.g. Berridge et al., 1982, 1983). We have recently become aware of two instances in which considerable loss of radiolabelled inositol phosphates can occur during extraction and analysis, and have found a simple remedy which we suggest could be used routinely to circumvent any future problems. When extracting inositol phosphates from concanavalin A-treated murine thymocytes labelled with myo['H]inositol (Moore et al., 1984) we found that inositol tetrakisphosphate (InsP4) levels were very much reduced if instead of a chloroform/methanol extraction we used trichloroacetic acid. This was caused by absorbance of [3H]InsP4 onto plastic surfaces and cellulose acetate filters used in our procedure; [3H]Ins(1,3,4,5)P4 was lost in this way, but [3H]Ins(1,4,5)P3 was not. Thus a specific loss of InsP4 occurred which distorted the relative amounts detectable in the tissue. In view of the probable second messenger role of InsP4 (Irvine & Moor, 1986) this loss is undesirable. Another extraction method which under some circumstances could lead to loss of radiolabelled inositol phosphates was that of Sharpes & McCarl (1982) using Freon and tri-n-octylamine. When [3H]Ins(1,4,5)P3 (Amersham) was extracted from water using this method, < 10% remained in the aqueous upper phase. Maleic acid in the original InsP3 solution (i.e. having the InsP3 either in a Tris/maleate buffer rather than a Tris/HCl buffer, or in the presence of 75 mm-maleate, pH 7) prevented this loss, which we ascribe to the trapping of InsP3 as the octylamine salt, and its displacement by the maleate anion (Wreggett & Irvine, 1987). We believe both these losses during extraction are caused by the well-known phenomenon of trace amounts of compounds being bound to non-specific sites, this problem typically being alleviated by increasing the mass of the compound concerned. Thus, in this particular case, the problems may only be confined to very small amounts of tissue, and only to particular pieces of apparatus (type of filter etc.); for example, we know (C. P. Downes & P. T. Hawkins, personal communication) that no loss of any inositol phosphates occurs during extraction with trichloroacetic acid of carbacholstimulated parotid slices (Hawkins et al., 1986) and this may be perhaps because of the greater mass of tissue, or slightly different techniques. Rather than investigate this in detail, we preferred to find a simple solution: if this