Genetic improvement of Bacillus licheniformis strains for efficient deproteinization of shrimp shells and production of high-molecular-mass chitin and chitosan

Abstract

By targeted deletion of the polyglutamate operon (pga) in Bacillus licheniformis F11, a derivative form, F11.1 (Δpga), was obtained that, along with lacking polyglutamate (PGA) formation, displayed enhanced proteolytic activities. The phenotypic properties were maintained in a strain in which the chiBA operon was additionally deleted: F11.4 (ΔchiBA Δpga). These genetically modified strains, carrying the Δpga deletion either alone (F11.1) or together with the ΔchiBA (F11.4) deletion, were used in fermentations (20-liter scale) aiming at the deproteinization of shrimp shells in order to obtain long-chain chitin. After chemical deacetylation, the resulting chitosan samples were analyzed by nuclear magnetic resonance spectroscopy, size exclusion chromatography, and viscometry and compared to a chitosan preparation that was produced in parallel by chemical methods by a commercial chitosan supplier (GSRmbH). Though faint lipid impurities were present in the fermented polysaccharides, the viscosity of the material produced with the double-deletion mutant F11.4 (Δpga ΔchiBA) was higher than that of the chemically produced and commercially available samples (Cognis GmbH). Thus, enhanced proteolytic activities and a lack of chitinase activity render the double mutant F11.4 a powerful tool for the production of long-chain chitosan.

FIG. 1.

Schematic representation of PGA synthetase operon A and chitinase operon B from wild-type (WT) B. licheniformis F11; deletion mutants B. licheniformis F11.1 (Δpga), B. licheniformis F11.3 (ΔchiBA), and B. licheniformis F11.4 (Δpga ΔchiBA); and complementation mutant B. licheniformis F11.2 with an adenosine insertion inside the chiA sequence and thereby reactivated chitinase A. Open reading frames are illustrated as arrows, the direction of which correlates with the transcriptional orientation. Positions of PCR primers are marked by solid arrows. The positions of the PCR primers to amplify the genomic region, pga3/pgaS20 for the pga operon and Anker2chiBA/Anker2bchiBA for the chiBA operon, as well as of the PCR primers to amplify the flanking regions, are given. (A) Region of the pga operon. rbsR codes for the transcriptional repressor of the ribose operon; ywsC codes for PGA synthetase; ywtA, ywtB, and ywtC are also part of the membrane-bound PGA synthetase complex; and ywtD codes for PGA depolymerase. (B) Region of the chiBA operon. Bli00337 codes for a hypothetical protein, chiB codes for chitinase B, chiA codes for chitinase A, mpr codes for glutamate-specific endopeptidase, and ycdF codes for a putative glucose-1-dehydrogenase.

FIG. 2.

Microscopic visualization by negative capsular staining of the slime capsules of B. licheniformis F11 (wild type) and B. licheniformis F11.1 (Δpga mutant) grown on PGA production medium (for details, see Materials and Methods).

FIG. 3.

Analysis of extracellular enzyme activities of wild-type B. licheniformis F11 and mutants B. licheniformis F11.1 (Δpga), B. licheniformis F11.2 (active chiA), B. licheniformis F11.3 (ΔchiBA), and B. licheniformis F11.4 (Δpga ΔchiBA) and determination of specific protease activities. (A) B. licheniformis F11 and F11.1 (Δpga) cultures containing equal numbers of cells (same optical density [OD]) were spotted onto agar plates containing lichenin (cellulase), skim milk (protease), and starch (amylase); clearing halos around the colonies illustrate enzyme activities. (B) Extracellular protease activity after 12, 24, and 36 h of batch fermentation in M9 minimal medium. Mean values were calculated from duplicates of at least five independent experiments. Standard deviations are given. (C) Analysis of extracellular chitinase activity of the parental strain B. licheniformis F11 (wild type) in comparison to those of the respective mutants. Equal numbers of cells were spotted on chitin plates and incubated for 6 days at 37°C.

FIG. 4.

Proteolytic activity in fermentation broth measured with azocasein as protease substrate. (A) Fermentations were done in a 1.2-liter working volume. Values given are means of duplicates. (B) Fermentations were done in a 20-liter working volume. The values shown are means of duplicates.

FIG. 5.

Results of structural analysis. (A) ¹H NMR spectra of chitosans F11.1 CSN, F11.4 CSN, and SN 21 (400 MHz; solvent, D₂O plus DCl; H-Ac, CH₃-acetyl; D, deacetylated unit; A, acetylated unit). (B) H,H-COSY NMR spectra of chitosan F11.1 CSN (400 MHz; solvent, D₂O plus DCl). (C) Broad-band-decoupled ¹³C NMR spectrum of chitosan F11.1 CSN as an example of the chitosans examined (100 MHz; solvent, D₂O plus DCl).

FIG. 6.

Viscometric measurements of chitosan samples F11.1 CSN, F11.4 CSN, SN 21, and CGN (Cognis chitosan) in a 0.5 M acetic acid-0.2 M sodium acetate buffer at 25°C.

FIG. 7.

Absolute molar mass distributions of chitosan samples F11.1 CSN, F11.4 CSN, and SN 21 (solvent, 0.1 vol% CF₃COOH plus 0.1 M NaCl; dn/dc, 1.85 ml/g).