Small RNA mediated repression of subtilisin production in Bacillus licheniformis

Abstract

The species Bacillus licheniformis includes important strains that are used in industrial production processes. Currently the physiological model used to adapt these processes is based on the closely related model organism B. subtilis. In this study we found that both organisms reveal significant differences in the regulation of subtilisin, their main natural protease and a product of industrial fermentation processes. We identified and characterized a novel antisense sRNA AprAs, which represents an RNA based repressor of apr, the gene encoding for the industrial relevant subtilisin protease. Reduction of the AprAs level leads to an enhanced proteolytic activity and an increase of Apr protein expression in the mutant strain. A vector based complementation of the AprAs deficient mutant confirmed this effect and demonstrated the necessity of cis transcription for full efficiency. A comparative analysis of the corresponding genome loci from B. licheniformis and B. subtilis revealed the absence of an aprAs promoter in B. subtilis and indicates that AprAs is a B. licheniformis species specific phenomenon. The discovery of AprAs is of great biotechnological interest since subtilisin Carlsberg is one of the main products of industrial fermentation by B. licheniformis.

Figure 1

AprAs transcription profile, promoter region and sequence analysis. (a) Transcription profile of AprAs in the early stage of an industrial subtilisin production. Transcriptome data were visualised in a logarithmic scale using TraV. The red graph represents the transcription activity of aprAs and the blue of apr. The black arrow on the complementary strand represents the coding region of apr and the yellow arrow of aprAs. The dark red arrows represent the identified promoters and green arrows predicted transcription terminators. (b) Promoter region of aprAs. The −10 box (TAGAAT) of the potential SigA promoter is highlighted in red and the exchanged sequence (AscI pattern) is given. Predicted transcription terminators are framed with green arrows. Protein coding sequences are shown in black bold letters and the stop codons are underlined. The aprAs coding sequence is given in orange letters and the experimentally determined transcription start site of AprAs is framed in red. (c) AprAs sequence. The 144 base large AprAs sequence was analyzed via the RNAfold WebServer, using “The Vienna RNA Websuite”. The calculation of secondary structures was possible and the optimal structure with a minimum free energy of −51.9 kcal/mol is shown in dot-bracket notation.

Figure 2

Slot blot analysis of AprAs transcription in B. licheniformis MW3 and B. licheniformis MW3 mutant strains. (a) The quality of the total RNA of B. licheniformis MW3 (lane 1), B. licheniformis MW3 AprAs⁻ (lane 2), B. licheniformis MW3 pV2 (lane 3), B. licheniformis MW3 AprAs⁻ pV2 (lane 4), B. licheniformis MW3 pV2::aprAs (lane 5) and B. licheniformis MW3 AprAs⁻ pV2::aprAs (lane 6) was determined using an Agilent Bioanalyzer. Lane 1 and 2 were analysed on a separate chip and the respective ladders (L) are shown. (b) The AprAs transcript was detected with an AprAs specific probe in B. licheniformis MW3 total RNA. B. licheniformis MW3 (lane 1) and B. licheniformis MW3 pV2 (lane 3) show an AprAs specific signal, whereas B. licheniformis MW3 AprAs⁻ and B. licheniformis MW3 AprAs⁻ pV2 (lane 4) possess no or only a very faint AprAs specific signal. B. licheniformis MW3 pV2::aprAs (lane 5) and B. licheniformis MW3 AprAs- pV2::aprAs (lane 6) show a strong AprAs signal dependent on the AprAs transcription from vector pV2::aprAs.

Figure 3

Determination of exoprotease activity by qualitative and quantitative measurement in B. licheniformis MW3 and mutant strains. For qualitative determination of exoprotease activity (a,c), cultures were adjusted to OD₆₀₀ of 0.1. Of each culture 3 µl were dropped on a M9 skim milk agar plate and incubated for 5 days at 37 °C. Exoprotease activity led to the digestion of milk protein resulting in a surrounding halo. The size of the respective halos is indicated with a red bar. For quantitative determination of exoprotease activity (b,d), all cultures were inoculated to an OD₆₀₀ of 0,1 in liquid M9 skim milk medium and incubated for 24 h under intensive shaking. The exoprotease activity was determined using the culture supernatant. (a) B. licheniformis MW3 AprAs⁻ reveals an increased exoprotease activity visible through a stronger halo formation compared to MW3. (b) The four fold increased exoprotease activity of B. licheniformis MW3 AprAs⁻ confirms the exoprotease phenotype observed in Fig. 3a. (c) B. licheniformis MW3 pV2::aprAs shows the smallest halo formation. The halo of the complemented B. licheniformis MW3 AprAs⁻ pV2::aprAs is comparable to MW3 pV2. B. licheniformis MW3 AprAs⁻ pV2 shows the strongest proteolytic activity indicated by the largest halo formation. The ectopic transcription of aprAs results in a reduction of exoprotease activity in both MW3 and mutant AprAs⁻. (d) The protease activity of the complemented strain B. licheniformis MW3 AprAs⁻ pV2::aprAs is strongly decreased compared to B. licheniformis MW3 AprAs⁻ and comparable to B. licheniformis MW3 pV2. B. licheniformis MW3 pV2::aprAs shows an approximately 50% reduced exoprotease activity.

Figure 4

Protease activity and Apr protein expression in B. licheniformis MW3, MW3 AprAs⁻ and MW3 AprAs⁻ pV2::aprAs. The strains were grown in 400 ml M9 skim milk medium for 48 h. (a) The cell culture supernatants of samples from time points 12 h, 24 h, 36 h and 48 h were used for quantitative exoprotease activity determination. The absolute value of protease activity is shown as the average of three independent experiments, each with triplicate measurements. The exoprotease activity increased from 12 h to 48 h and is clearly stronger in B. licheniformis MW3 AprAs⁻ in all experiments compared to B. licheniformis MW3 and MW3 AprAs⁻ pV2::aprAs. The error bars display the standard deviation and the colour legend is shown in the figure. (b) Extracellular proteins were isolated from supernatants of 400 ml M9 skim milk cultures after 48 h of growth. The protein fractions were separated by 2D-gelelectrophoresis. A gel image of B. licheniformis MW3 (green) was overlaid with the respective image of MW3 AprAs⁻. The volume of the three identified Apr protein spots is strongly increased in B. licheniformis MW3 AprAs⁻ (red) compared to the original strain B. licheniformis MW3 (green). In addition, the gel image of B. licheniformis MW3 (green) was overlaid with the respective image of MW3 AprAs⁻ pV2::aprAs (red). After vector complementation, the effect of the mutation is reversed. B. licheniformis MW3 AprAs⁻ pV2::aprAs shows only a slight increase in Apr protein expression compared to B. licheniformis MW3. (c) To quantify the Apr expression the spot volumes (in percentage of the whole protein spot volume) were calculated. The average of the normalized spot volumes of three replicates is shown and the standard deviation is given. The Apr protein expression in the extracellular proteome of B. licheniformis MW3 AprAs⁻ is approximately three times increased compared to the original strain B. licheniformis MW3. The Apr expression of B. licheniformis MW3 AprAs⁻ pV2::aprAs is reduced compared to the precursor strain B. licheniformis MW3 AprAs⁻, but still higher than in B. licheniformis MW3.

Figure 5

Alignment of AprAs promoter sequences from B. licheniformis and B. subtilis strains. Protein coding sequences are marked in grey. Transcription terminator sequences are marked in green and the promoter regions (−10 box) in red. The alignment was performed using MUSCLE. (a) Sequence alignment of B. licheniformis DSM13 (NC_006270) and B. subtilis 168 (NC_000964). B. subtilis 168 lacks the −10 box upstream of the aprAs gene of B. licheniformis DSM13. (b) Sequence alignment of 17 B. licheniformis strains. The AprAs −10 box is present in all strains. (c) Sequence alignment of 35 B. subtilis strains. The AprAs −10 box is absent in all strains.