Histamine, putrescine, spermine, spermidine, agmatine, cadaverine, trimethylamine, O-phthalaldehyde, 4-(1-pyrene)butyric acid N-hydroxysuccinimide ester (PSE) were purchased from Sigma (Sigma Aldrich, St. Louis, MO). 1-Pyrenebutyry Chloride (PBC) was purchased from Tronto Research Chemicals Inc., (Ontario, Canada). The purity of all reagents was HPLC grade or highest quality.

Histamine, putrescine, spermine, spermidine, agmatine, cadaverine and trimethylamine were prepared in 1 M, and diluted to the required concentration by HPLC grade water (Wako pure chemical, Osaka, Japan) before use. OPA solution was prepared by dissolving 20 mg OPA into 10 mL of ethanol. PSE solution was prepared in 5 mM by acetonitrile of highest grade. PBC solution was prepared in 30 mM in DMSO and diluted to 6 mM by acetonitrile. All the solutions were stored in polypropylene micro tubes protected from light and stored at –20˚C, and used within 1 week.

Fresh sardine Sardinops melanostictus (Temminck and Schlegel, 1846) and mackerel Scomber japonicus (Houttuyn, 1782) were purchased from a supermarket in Tokyo. The fish in polystyrene foam box containing ice were immediately transported to the laboratory. Some fish were stored at 30˚C for 24 h and 72 h to be decomposed by microbial decarboxylation. After peeling, 5.0 g of fish meat was homogenated with 8.0 mL of water and centrifuged at 430 × g for 10 min at 4˚C. The supernatant volume was re-adjusted to 8.0 mL and mixed with 2.4 mL of 30% trichloroacetic acid, followed by centrifugation at 13,000 × g for 10 min to remove crude protein. The supernatant was filtered with 0.45 μm syringe filter DIS-MIC (Advantec, Japan), stored at –80˚C until analysis. As control, fresh fish extracts were prepared just after thawing. For OPA-derivatization, the extract was further purified by mixing 3 N NaOH containing 100 mg/mL of NaCl with N-butanol at ratio of 20:25:2 (vol/vol) followed by centrifugation at 10,000 × g for 10 min. The supernatant was added with 0.12 N HCl and N-heptane at 5:3:9 (vol/vol). After shaking, the lower phase was collected and stored at –80˚C until analysis.

OPA derivatization was performed as follow. 1200 µL of purified fish extract was mixed with 60 µL of OPA solution and 240 µL of 1 N NaCl. After incubation at 4˚C for 40 min, 120 µL of 3 N HCl was added. PSE derivatization was performed as described with slight modification [33]. Briefly, 150 µL of each amine standard solution or fish solution were mixed with 1200 µL of 5 mM PSE solution and 300 µL of 1.5 mM potassium carbonate in 2-mL screw cap polypropylene tubes (Axygen scientific, Unicon city, California). The tubes were tightly sealed and heated for 90 min in boiling water. The reacted solutions were diluted 1000 times with acetonitrile to decrease emission from monomer. PBC-labeling was performed as described, with slight modification [31]. Briefly, sample solutions, 6 mM PBC solution and 0.5 mM potassium carbonate were mixed at ratio (volume) of 10:20:1 (600 µL:1200 µL:60 µL), followed by incubation at 25˚C for 5 min. The fluorescence intensity from OPA-, PSE, PBCderivatized biogenic amines were measured by emission 350 - 600 nm and excitation of 340, 350 and 355 nm, respectively.

Fluorescence spectral was measured with RF-5300 spectrofluorometer (Shimadzu, Tokyo, Japan) in 10 × 10 mM quartz cuvette. Spectral band width of 1.5 nm was used for both the excitation and emission. Data were collected and analyzed with Ion Prove Program (version 1.00) (Shimadzu, Tokyo, Japan).

We investigated the fluorescence from OPA labeled histamine at the concentration between 10 pM - 1 M, and found peak fluorescence at 445 nm with dose-dependent increase in a range of 100 nm - 100 μm (Figure 2(a)).

Fluorescence intensity from distilled water, 100 nm and 1 M of histamine were 1.03, 1.56 and 4.31, respectively. The solution of histamine above 1 mM was yellow color, and therefore it is impossible to take proper intensity. We also measured OPA-derivatized fish extracts prepared from fresh sardine, mackerel and their spoiled fishes incubated at 30˚C for 3 days. Both spoiled sardine and mackerel obviously increased fluorescence compared with fresh one, suggesting the histamine production through decomposition (Figure 2(b)).

We next analyzed the excimer fluorescence from biogenic amines mainly produced in decayed fish. Each concentration of amine standard solutions (0, 0.1, 1, 10 mM) was reacted with weakly alkalinized PSE solution, and diluted 1000-times with acetonitrile to minimize intermolecular excimer. Neither histamine nor trimethylamine altered the fluorescence (Figures 3(a) and (b)). Agmatine minimally increased fluorescence around 450 nm (Figure 3(c)). Excimer fluorescence with broad peak at around 470 nm was clearly appeared in putrescine, cadaverine, spermidine and spermine at the concentration of 10 mM (Figures 3(d)-(g)). As for PBC-derivatized amines, there is no change of peak and intensity (data not shown). Although the detection of PSE-derivatized histamine from rat brain by using HPLC was reported [33], we could detect the excimer fluorescence from only lin ear type diamines, namely putrescine, cadaverine, spermidine and spermine (structures shown in Figure 1, right side), but not histamine in our analysis. Trimethylamine, itself tertiary amine, did not emit the fluorescence. In addition, increase of the exicimer fluorescence seen in PSE-labeled agmatine seems false-positive, since stokes shift was smaller than that of other polyamine. In case of amine having primary amine at the branched side chain,

normally excited dimer may not be composed because of an inadequate tertiary structure. In fact, it is known that two or more fluorophores in a molecule cause quenching of fluorescence, called self-quenching. Therefore, it is supposed that only linear diamines containing primary amines at both ends are detectable with PSE in our method.

Freshness of fish was evaluated by measurement of excimer fluorescence from PSE-derivatized fish extract after one-step purification with trichloroacetic acid. The

fluorescence wave profile in the extracts prepared from fresh fish was the same as water (Figure 4(a), line 4-6). Spoiled mackerel released the significant emission with broad peak at around 470 nm (Figure 4(a), line 2). The fluorescence intensity was also increased in the extract from sardine incubated for 72 h, and was higher than that from that for 24 h (Figure 4(a), line 1 and 3).

Figure 4(b) shows the fluorescence intensity value of each samples subtracted that of blank (HPLC grade water). The fluorescence intensity in fresh sardine and mackerel were 0.65 and 1.43. Similarly, histamine, agmatine and trimethylamine showed the intensity of 0.77, 0.14 and 0.30, respectively. Mackerel incubated for 72 h at 30˚C increased the intensity to 8.39. In sardine, the intensity in 24 h and 72 h at 30˚C incubation were 3.60 and 21.4, showing polyamine production during improper storage. Therefore, polyamine in fish extract can be estimated as 1 - 10 mM level of putrescine, cadaverine, spermidine and spermine, suggesting significant detection level and practical application for evaluation of freshness (Figure 4(b)).