Transforming growth factor-beta2 (TGF-beta2) is one of the most important immunosuppressive cytokines in the anterior chamber of the eye. It is secreted as a complex with latency-associated peptide as an inactive precursor. Only the activated form of TGF-beta2 can bind to its receptor and induce signaling. To date, the concentration of active TGF-beta2 in aqueous humor was exclusively determined using samples that had been preserved at -80 degrees C. Quantitative measurements of the activated form directly after sampling have not yet been taken. The aim of this study was to investigate the effect of cryopreservation on the concentration of active TGF-beta2 in the aqueous humor. Samples of aqueous humor were drawn from patients with either cataract or a corneal disorder for determination of TGF-beta2 using a Sandwich-ELISA. In group I (n=30, patients with corneal disorders or cataract), one part of each sample was tested for active TGF-beta2 directly after sampling, whereas the remaining material was stored at -80 degrees C for later analysis. Group II consisted of patients undergoing a simple cataract extraction (n=38), and active TGF-beta2 levels were determined within 3 h after sampling. In Group III (n=34, patients with corneal disorders or cataract), active TGF-beta2 was determined within 3 h after puncture, as were total TGF-beta2 levels after acidic activation for each sample. The average level of active TGF-beta2 in the aqueous humor of group I analyzed directly after sampling was 35+/-31 (median 37) pg/ml. In contrast, frozen samples from the same patients showed an average concentration of 155+/-103 (median 152) pg/ml. The average level of active TGF-beta2 in aqueous humor of 38 cataract (group II) eyes was 40+/-24 (median 41) pg/ml determined within 3 h after puncture. The average level of total TGF-beta2 in group III was 1,654+/-631 (median 1,542) pg/ml compared to 33+/-39 (median 28) pg/ml of active TGF-beta2 determined directly after sampling, yielding a ratio of 2% of active to total TGF-beta2. Levels of active TGF-beta2 in aqueous humor determined directly after sampling were 4.4 fold lower than those measured in frozen samples. Thus, samples meant for determining active TGF-beta2 levels should not be kept frozen.