In order to understand the mechanism of chitin degradation by C. tainanensis, chitin degradation studies were performed in a 5 L fermenter. C. tainanensis was incubated in B-H broth containing 4% α-chitin powder. Dissolved oxygen (DO), pH, total aerobe count and reducing sugar were measured continuously for a period of 72 hours. The pH value of the broth decreased gradually from 7.4 to 5.3 in 14 hours (Figure 1A). DO decreased for 4 hours, after that, it increased up to 100% (Figure 1B). This implied that C. tainanensis proliferated rapidly for 14 hours, and some acidic metabolites were released and accumulated in the broth. After incubation, C. tainanensis was dead owing to the acidic broth, confirmed by counting of total aerobe number (Figure 1C).

It was interesting to notice that the concentration of reducing sugar in the broth was low within the first 14 hours; however, when C. tainanensis was incubated for longer than 14 hours, the level of reducing sugar increased over the incubation time of the tested period of 72 hours (Figure 1D). This suggested that some factors that were able to degrade chitin completely were produced simultaneously with C. tainanensis proliferation. The hydrolytic product was used as a carbon source for bacterial proliferation, and acidic metabolites were produced simultaneously. C. tainanensis was dead at pH 5.3; however, the assumed factors still existed in the broth. NAG, the final hydrolytic product of chitin degradation, was not consumed, but accumulated in the broth instead. The assumed factors were nominated “CDFs”.

The bacterial debris was obtained by centrifugation, and it was suspended in 0.1 M NaOAc, pH 5.0. After 4% β-chitin was added, the brix increased dramatically in the reaction containing bacterial debris, while that in the reaction containing supernatant increased moderately. In another experiments, β-chitin was hardly degraded by the bacterial debris obtained from the incubation broth using glucose as a carbon source (Figure 2). β-chitin was more susceptible than α-chitin, and higher reducing sugar was produced [21]. NAG concentration was directly measured using a hand refractometer due to its high concentration in the broth. These results reflected that CDFs were located on the bacterial debris, and were induced by chitin powder.

In order to understand how CDFs hydrolyzed chitin, endochitinase and NAGase activities were measured. Both enzyme activities were detected in the bacterial debris. The specific activity of NAGase was much higher than that of endochitinase (data not shown). These results were in agreement with previous reports [14,17,18].

Since assumed CDFs were located on the debris of C. tainanensis, it was reasonable to identify the structural information of CDFs. Bacterial debris (O), dramatically shaken fraction (FI), surfactant extracted fraction (FII) and residue (P) were obtained as described in the Experimental section. Chitin degrading effects and NAGase activities of the four fractions were compared, and those of bacterial debris were set as 100% (Figure 3). Chitin was not efficiently degraded in fraction FI which possessed high NAGase activity. Soluble surface components were almost extracted by shaking treatment due to low chitin degrading effect in fraction FII. In the residual fraction (P), chitin was efficiently degraded although lower NAGase activity existed in this fraction. These results suggested that CDFs strongly interacted with the bacterial surface, and were similar to the cellulosome concept in cellulose degradation [22]. In addition, NAGase activity did not play major roles in chitin degradation of C. tainanensis.

Chitin degrading effects and enzymatic activities of the bacterial debris were also performed in buffers with different pHs. It was surprising that the optimum pH value of the chitin degrading effect was about 5.3, while those of endochitinase and NAGase isolated from the bacterial debris were about 7.0 (Figure 4). These results suggested that the mechanism of chitin degradation of C. tainanensis was novel and different from that of the enzymatic method published before. Some unknown factors not yet explored played important roles in chitin degradation.

Powdered α-chitin and β-chitin used in this study were supplied from Charming and Beauty company (Taipei, Taiwan). Media for microorganism incubation were purchased from Difco Laboratories (Dentroit, MI). Colloidal chitin was prepared from α-chitin using the method of Jeniaux [23,24], and other chemicals used in this study were purchased from Sigma Chemical Co. (St. Louis, MO, USA).

The agar containing B-H broth, Difco Bushnell-Haas broth, and 0.2% colloidal chitin was used as a selective medium. The colonies with surrounding clear zones were obtained and grown in 400 mL Luria-Bertani medium at 30 °C overnight. A fresh overnight culture was transferred to a 5 L fermenter containing 4 L B-H broth supplemented with 4% α-chitin powder as a sole carbon source. The agitation rate and air feeding were set at 200 rpm and 4 L/min. All fermentations were performed at 30 °C. pH and DO were detected by an appropriate electrode, and the reducing sugar was analyzed with a spectrophotometer using potassium ferricyanide as substrate [25]. Numbers of bacteria in the fermenter were counted after spraying the diluted broth on nutrient agar following 48 hours incubation.

C. tainanensis was incubated in B-H broth containing 4% β-chitin powder for 48 hours. Bacterial debris was collected by centrifugation, and was suspended in an equal volume of 0.1 M NaOAc, pH 5.0. Bacterial debris (O) was shaken dramatically to obtain the surface component in supernatant (FI) after centrifugation, and the pellet was further treated with 0.1% N-lauroylsarcosine to isolate the other component on the surface of bacterial debris (FII). After centrifugation, the residues were suspended in equal volume of 0.1 M NaOAc, pH 5.0 for test.

Chitin was used as substrate to analyze the chitin degrading effect. Briefly, in the sterilized test tube, chitin was added to the bacterial debris or its components described above at a final concentration of 4%, and 10 μg/mL of tetracycline was added subsequently to prevent microorganism contamination. NAG released in the reaction was detected using a hand refractometer after centrifugation, and the result was expressed as brix (%). The yield was calculated from the equation: Yield (%) = NAG (g)/Chitin (g) added [14].

Endochitinase and NAGase activities were analyzed using 0.18 mM PNP-(GlcNAc)₃ and 0.18 mM PNP-GlcNAc as substrate, respectively [7,20,24,26,27]. After 30 min incubation, NaOH was added to a final concentration of 0.1 M to terminate the reaction. The absorbance at 405 nm was measured using a spectrophotometer. When enzyme activity of insoluble component was analyzed, the insoluble component was removed by centrifugation before colorimetric measurement.