OUP user menu

Purification and characterisation of a serine peptidase from the marine psychrophile strain PA-43

Jane A. Irwin , Gudni A. Alfredsson , Anthony J. Lanzetti , Haflidi M. Gudmundsson , Paul C. Engel
DOI: http://dx.doi.org/10.1111/j.1574-6968.2001.tb10770.x 285-290 First published online: 1 July 2001


An extracellular serine peptidase, purified from the culture supernatant of the sub-Arctic psychrophilic bacterium strain PA-43, is monomeric, with a relative molecular mass of 76 000, and an unusually low pI of 3.8. The peptidase is active towards N-succinyl AAPF p-nitroanilide and N-succinyl AAPL p-nitroanilide, indicating a chymotrypsin-like substrate specificity. It is inhibited by the serine peptidase inactivator phenylmethylsulfonyl fluoride, but not by EDTA or EGTA, suggesting that added metal ions are not necessary for activity. The enzyme is most active at pH 8.3 and at 55–60°C, although it is unstable at 60°C. It is nevertheless remarkably stable for an enzyme from a psychrophilic microorganism, remaining active after 1 week at 20°C and after five freeze–thaw cycles. Comparison of the N-terminal 40 amino acid residues with other archived sequences revealed highest similarity to the alkaline serine protease (aprx) from Bacillus subtilis.

  • Serine peptidase
  • Psychrophilic
  • Halophilic
  • Cold-adapted enzyme
  • Vibrio
  • Shewanella

1 Introduction

Psychrophilic organisms live at the lowest temperatures which allow life to exist, and, given the extent of permanently cold environments on Earth, such as polar regions, mountains, and deep oceans, they comprise a sizeable part of the biosphere. In recent years, the molecular adaptations that allow these organisms to survive in such harsh conditions have come under investigation [1,2].

Cold-adapted enzymes produced by psychrophilic microorganisms or ectothermic animals inhabiting low-temperature environments, such as Antarctic fish, generally have a low temperature of optimum activity, unusually high values for kcat and for catalytic efficiency (kcat/Km) relative to their mesophilic and thermophilic orthologs over the temperature range 0–30°C, and comparatively low thermostability [3]. These characteristics may reflect exceptional molecular flexibility, probably resulting from a decrease in non-covalent interactions. Beyond such general ideas, however, the factors determining the high activity of cold-adapted enzymes at low temperatures and their thermal instability at moderate temperatures are not well understood.

Studies on cold-adapted peptidases have centred on those from fish [4,5] and marine bacteria. Only a few cold-adapted extracellular bacterial peptidases have been investigated, including examples from Vibrio [6] and Bacillus [7] species. Recently, a cold-active serine peptidase from a Gram-negative psychrotrophic bacterium, Shewanella Ac10, has been cloned, sequenced, and characterised [8]. We report here the purification and detailed characterisation of an alkaline serine peptidase excreted by a sub-Arctic psychrophilic microorganism, strain PA-43. Some properties have been briefly reported elsewhere [9]. Despite its source, the peptidase does not show typically psychrophilic thermal instability. Such examples may be especially useful in testing emerging ideas about psychrophilic adaptation at the protein level.

2 Materials and methods

2.1 Chemicals

p-Nitroanilide peptidase substrates, peptidase inhibitors, azocasein, Mr markers for gel-filtration chromatography, Q Sepharose, Sephacryl S-300, sea salts, hydrolysed casein, and sodium caseinate were purchased from Sigma-Aldrich Chemical Co. A Superose 6 HR 10/30 fast protein liquid chromatography (FPLC) pre-packed column and PBE 94 resin were obtained from Amersham Pharmacia Biotech Ltd.

2.2 Isolation, growth and maintenance

The bacterium, initially described as a Vibrio sp., but now placed by 16S rRNA gene sequencing among the genus Shewanella [10], was isolated from a sea urchin harvested off the south-west coast of Iceland, and grows optimally at 16.5°C. Strain PA-43 was initially grown at 15°C in a clear liquid medium (CHYSS) containing 2 g l−1 yeast extract (Oxoid L21), 5 g l−1 casein (enzymatic hydrolysate, Sigma), and 20 g l−1 sea salts (Sigma) (pH 7.0). Details of growth experiments are given elsewhere [10]. Highest proteolytic activity however, was obtained with 5 g l−1 sodium caseinate, 4 g l−1 yeast extract and 1.5% NaCl (w/v) (C2Y+1.5% NaCl). This medium was used for subsequent peptidase production. Substitution of sea salts with NaCl made little difference to the proteolytic activity.

2.3 Enzyme assay

Proteolytic activity was assayed routinely at 25°C with 0.36 mM N-succinyl Ala-Ala-Pro-Phe (AAPF) p-nitroanilide as substrate in 0.1 M Tris–HCl buffer, pH 8.5, using a Uvikon 922 or 941 spectrophotometer (Kontron) at 410 nm and an extinction coefficient for p-nitroanilide of 8480 M−1 cm−1. The enzyme was also assayed with the macromolecular substrate azocasein in a reaction mixture (500 μl) containing 0.1 M Tris–HCl, 200 mM NaCl, pH 8.5, 100 μl 3% (w/v) azocasein in buffer, and 20 μl enzyme. The reaction was terminated after 20 min by adding 500 μl 10% trichloroacetic acid. Precipitated protein was removed by centrifuging (10 min at 13 000×g) and the activity in the supernatant was calculated using an extinction coefficient of 900 M−1 cm−1 at 366 nm for the chromophore. Rates were constant during 1 min assays. Kinetic parameters were estimated according to Wilkinson [11] and rate data (for 0.18–7.2 mM substrate) were fitted using the programme Enzpack 3.0 (Biosoft, Cambridge, UK).

2.4 Enzyme purification

After harvesting cells by centrifugation, the cell-free supernatant (500 ml) was applied to a Q Sepharose column (12 ml), equilibrated with 50 mM Tris–HCl buffer, pH 7.0. All steps were performed at 4°C. The column was washed with buffer and eluted with a linear gradient of 0–1 M KCl in buffer (12 column volumes). Pooled active fractions were concentrated to approximately 5 ml (Centriplus 10 concentrators) and applied to a Sephacryl S-300 gel filtration column (1.6×77 cm) equilibrated with the same buffer. Active fractions were again pooled and concentrated. Some preparations were further purified by FPLC using a Superose 6 HR 10/30 gel filtration column equilibrated with the same buffer. Chromatofocusing was carried out on a PBE 94 column (1 ml) equilibrated with 20 column volumes of 25 mM Tris–HCl buffer, pH 7.4, and eluted with Polybuffer 74 (10 ml) to give a pH gradient from 7.0 to 4.0.

2.5 Analytical procedures

Protein elution was monitored by A280 measurement and precise concentrations were assayed with the Bio-Rad protein assay dye reagent (Bio-Rad Laboratories), and bovine serum albumin (BSA) as standard.

The enzyme's Mr value was determined on an FPLC system (Amersham Pharmacia Biotech) by analytical gel filtration on a Superose 6 HR 10/30 pre-packed column, equilibrated with 50 mM potassium phosphate buffer containing 0.15 M KCl. Marker proteins were apoferritin, 443 kDa; glutamate dehydrogenase, 300 kDa; β-amylase, 200 kDa; BSA, 66.2 kDa; cytochrome c, 12.4 kDa. Samples (100 μl, approx. 1 mg ml−1) were loaded separately and the flow rate was 0.5 ml min−1.

A Mini-Protean apparatus (Bio-Rad) was used with 10% SDS–PAGE gels [12] to monitor purification and to determine the subunit Mr of the enzyme. The calibration proteins were phosphorylase b, 97.4 kDa; BSA, 66.2 kDa; aldolase, 39.2 kDa; triose phosphate isomerase, 26.6 kDa; trypsin inhibitor, 21.5 kDa and lysozyme, 14.4 kDa.

The Multiphor II electrophoresis unit (Pharmacia) for isoelectric focusing had a thermostatted circulator for the cooling plate. Precast polyacrylamide gels (Pharmacia, Ampholine R PAG plate, pH range 3.5–9.5) were calibrated with 11 pre-stained proteins covering the pH range 3.5–9.3. Samples (15 μl), applied to the gel surface at suitable positions along the gradient, were run for 1.5 h and stained with Coomassie blue G-250 (0.12% w/v) in 25% ethanol/8% acetic acid after 45 min fixation in an aqueous solution of 11.6% trichloroacetic acid/3.4% sulfosalicylic acid and brief equilibration in 25% ethanol/8% acetic acid.

N-terminal sequence of the purified enzyme was obtained by Western blotting 10% SDS–PAGE gels onto polyvinylidene difluoride membranes (Sequi-Blot™, Bio-Rad, 0.2 μm). Automated Edman-type analysis was performed using a Perkin-Elmer Model 494 sequencer and Procise Model 610 software.

3 Results

3.1 Enzyme purification and characterisation of the peptidase

Anion-exchange chromatography on Q Sepharose, followed by gel filtration on Sephacryl S-300, gave a peptidase that was almost homogeneous (Fig. 1). Table 1 records a typical purification. The enzyme eluted from Q Sepharose at KCl concentrations above 500 mM and the yield at this stage varied from 50 to 75%. One small contaminant (24 kDa on SDS–PAGE) persisted even after chromatofocusing, which also led to significant inactivation. FPLC gel filtration similarly failed to eliminate the contaminant. As the 24-kDa band remained when the sample was treated with phenylmethylsulfonyl fluoride (PMSF) prior to SDS–PAGE, it was probably not an autolytic fragment.

Figure 1

SDS–PAGE (10%) of the fractions obtained during the purification of the peptidase from the culture medium (C2Y). The lanes are as follows: 1, marker proteins with relative molecular masses indicated on the left; 2, culture medium; 3, Q Sepharose eluate; 4, Sephacryl S-300 peak; 5, 1 M NaCl eluate from PBE 94 chromatofocusing gel.

View this table:
Table 1

Summary of purification procedure for strain PA-43 proteinase

Purification stepVolume (ml)Total protein (mg)Total activity (U)Specific activity (U mg−1)Yield (%)Purification (fold)
Culture supernatant49516.865.23.881001
Q Sepharose chromatography40.41.7848.927.575.07.1
Sephacryl S-300 chromatography10.20.5132.864.350.316.5
Chromatofocusing on PBE 940.640.1917.0289.626.125.0
  • The activity at each step was measured using 0.36 mM AAPF at 25°C in 0.1 M Tris–HCl buffer, pH 8.5.

The enzyme was not eluted from the PBE 94 column by pH gradients of 9.0–6.0 or 7.0–4.0, and 1 M KCl was required for elution, suggesting a pI below 4.0. Isoelectric focusing subsequently confirmed that the pI of the enzyme was 3.8.

The relative molecular mass of the purified peptidase was estimated at 75 000 by FPLC gel filtration. Since SDS–PAGE gave an Mr of 76 000, the enzyme is a monomer. The activity was stable at −20°C in 0.1 M Tris–HCl, pH 8.5, for at least 3 months, with or without glycerol. The concentrated solution (0.15–0.25 mg ml−1) could be thawed and re-frozen at least five times without significant inactivation. At low concentrations (<10 μg ml−1), the enzyme was less stable at −20°C and lost approximately 50% of its activity after one freeze–thaw cycle.

3.2 Substrate specificity

The enzyme displayed high activity towards N-succinyl AAPF p-nitroanilide. The Km for AAPF was 3.2 mM and the kcat/Km value at 25°C was 145 s−1 mM−1. N-succinyl AAPL p-nitroanilide was also a relatively good substrate, displaying 41% of the activity with AAPF at the normal assay concentration of 0.36 mM (Km, 5.8 mM; kcat/Km, 59 s−1 mM−1 at 25°C). N-succinyl AAVA p-nitroanilide showed about 11% of the AAPF activity (Km, 2.7 mM; kcat/Km, 11 s−1 mM−1, 25°C). N-succinyl GGF p-nitroanilide, N-succinyl AAA p-nitroanilide and N-succinyl AAPD p-nitroanilide were poor substrates, and there was no activity with N-succinyl AAV p-nitroanilide, N-succinyl GFG p-nitroanilide, N-succinyl l-Phe p-nitroanilide, or N-benzoyl dl-Arg p-nitroanilide.

3.3 Effect of inhibitors on activity

The PA-43 peptidase was not inhibited by 10 mM EDTA or EGTA, suggesting that calcium, or other metal ions, are not essential for activity. The irreversible serine peptidase inactivator PMSF (1 mM) completely inhibited the enzyme, but the enzyme was not particularly sensitive to the irreversible trypsin inhibitor tosyl-l-lysine chloromethyl ketone or the chymotrypsin inhibitor tosyl-l-phenylalanine chloromethyl ketone, displaying 26% and 14% inhibition, respectively, after incubation for 10 min with 1 mM inhibitor. The enzyme was strongly inhibited by chymostatin (90% inhibition with 10 μM) and showed 73% inhibition with 50 μM antipain. Leupeptin and pepstatin were less effective (29% and 25% inhibition, respectively, with 10 μM inhibitor). Inhibition by leupeptin was concentration-dependent (53% with 40 μM and 68% with 100 μM leupeptin). Iodoacetamide (5 mM) failed to inhibit the enzyme, suggesting that it is not a cysteine peptidase. Aprotinin, even at 300 μM, did not inhibit. Neither soybean trypsin inhibitor (100 μM) nor the antibiotic peptide bacitracin (1 mM), which inhibits a wide range of peptidases [13], inhibited the enzyme, which also failed to bind to bacitracin–Sepharose.

3.4 Effect of pH and temperature on activity and stability

Optimal activity towards AAPF occurred at approximately pH 8.3. The enzyme was inactive at pH values below 5.0, and retained 24% of maximal activity at pH 10.5. The enzyme, incubated for 16 h in different buffers over the pH range 4.0.–2.0, was stable at pH values from 6.0 to 11.0, but not outside this range. This may explain why so much activity was lost at low pH values (pH<6.0) during chromatofocusing.

The enzyme's activity with AAPF was measured at 16 temperatures between 4 and 72°C. In all cases the reaction trace was linear over the period of observation, and maximum activity was measured at 60°C (270 U mg−1). Activities at 4, 20, 40 and 72°C were, respectively, 10%, 23%, 67% and 42% of the maximum. With azocasein as a substrate (incubations were carried out at 10°C intervals from 0–80°C), an apparent optimum temperature for activity was observed at 60°C. The enzyme was stable at room temperature (20°C) for at least a week and for at least 3 h at 37°C, although it lost most of its activity after 18 h at this temperature. First-order decay plots (Fig. 2) gave half-life values for the peptidase of 125, 42, 4 min, and 35 s at 50, 55, 60, and 70°C, respectively.

Figure 2

Semilogarithmic plot of peptidase activity as a function of time at different temperatures. The purified enzyme (0.5 mg ml−1) was incubated at different temperatures (37, 50, 55, 60°C) in 50 mM Tris–HCl, pH 7.0, and aliquots were withdrawn, cooled, and assayed at 25°C in 0.1 M Tris–HCl buffer containing 0.36 mM AAPF. The residual activity is defined as the activity at a given time divided by the activity at time 0 min.

No autolytic cleavage to recognisable fragments was seen on SDS–PAGE after the enzyme was incubated for 30 min at a range of temperatures. However, after the 60°C incubation the band was much fainter than for other temperatures and no activity remained, suggesting extensive autolysis to small fragments at this temperature.

3.5 N-terminal sequence

An N-terminal sequence of 40 amino acid residues was obtained by Edman degradation. Similarity searches using BLAST algorithms, accessed through the National Centre of Biotechnology Information (http://www.ncbi.nlm.nih.gov), yielded matches with Bacillus subtilis alkaline serine peptidase (aprx), the Streptomyces coelicolor serine peptidase, the minor extracellular peptidase EPR precursor from B. subtilis, and the intracellular serine peptidase from Paenibacillus polymyxa. Sequences that displayed closest similarity to the PA-43 sequence were aligned (Fig. 3). The highest identity to the PA-43 peptidase sequence was found with the B. subtilis alkaline peptidase (57% identity, but over only 21 residues). The PA-43 peptidase N-terminal sequence did not show any homology to the N-terminal sequence or internal sequences from any other bacterial peptidases, including peptidases from Vibrio or Shewanella species.

Figure 3

Alignment between the N-terminal sequence of the peptidase and other peptidase sequences. Alignments were carried out with the programme Clustal W (version 1.8). The following EMBL/GenBank accession numbers were used to obtain these sequences; B. subtilis alkaline protease aprx, Z99113; S. coelicolor serine protease, U33176; minor extracellular proteinase VPR precursor from B. subtilis, P29141; intracellular serine protease from P. polymyxa, P29139. An asterisk denotes residue identity and a colon indicates similarity.

3.6 Effects of salts, metal ions, organic solvents and chaotropic reagents on activity and stability

The effect of salts on the enzyme was investigated because PA-43 is moderately halophilic, growing well in up to 6% NaCl. The enzyme was activated by NaCl, KCl, and LiCl and maximally (3.6-fold) with 4.5 M NaCl. 3.65 M KCl gave 2.95-fold activation, but LiCl was far less effective, giving a maximum activation of 47% at 2.5 M, and inhibition above 3.5 M. Both (NH4)2SO4 and NH4Cl activated the enzyme (8.2-fold activation in 2 M (NH4)2SO4 and 2.1-fold in 2 M NH4Cl). However, AAPF insolubility at (NH4)2SO4 concentrations above 2.5 M hindered activity assays. Above 3 M, NH4Cl inhibited the peptidase.

Subtilases, in general, have 4 Ca2+ binding sites and require Ca2+ for activity [9]. However, the activity of this peptidase was unaffected by EDTA or EGTA and was not increased by adding 2 mM CaCl2 to the assay mixture. No significant activation or inhibition was observed in the presence of 1 mM Mg2+, Mn2+, Fe2+, or Zn2+. To determine if the stability of the enzyme was affected by Ca2+, the enzyme was incubated at 40, 50, and 60°C in the presence of 2 mM CaCl2 for 30 min and the activity compared to controls incubated at the same temperature, as well as a control incubated at 25°C. The enzyme was neither stabilised by Ca2+ nor destabilised by 10 mM EDTA or EGTA in the incubation buffer. Thus, Ca2+ is either not required or extremely tightly bound. The enzyme was stabilised by NaCl (optimally 1 M) when incubated at 50°C.

The peptidase retained 85% and 91% activity in 10% ethanol and 10% methanol, respectively. It was remarkably stable in dimethyl sulfoxide (DMSO), retaining 87% and 65% of its activity in 50% DMSO at 20°C after 10 and 15 days, respectively. After 26 h in 6 M urea, 80±10% of the activity remained, and 52% remained even after 6 days, but the enzyme lost all activity after 1 day in 7 M urea.

4 Discussion

We have purified an alkaline peptidase from the culture supernatant of a marine psychrophilic bacterium, strain PA-43. This enzyme appears to be a serine peptidase, on the basis of its sensitivity to PMSF, and its preference for substrates with large hydrophobic amino acids in the P1 position, e.g. AAPF and AAPL, is characteristic of subtilisins [14]. Studies with inhibitors also indicated a serine peptidase with no requirement for added metal ions. The data on thermostability and variation of activity with temperature, as well as resistance to autolysis, indicate that this is a more stable enzyme than might be expected in a cold-adapted organism. The temperature optima for AAPF hydrolysis (55–60°C) and 60°C for azocasein hydrolysis compare with a topt of 60°C for the mesophilic peptidase subtilisin Carlsberg [7]. The stable activity in some organic solvents and insensitivity to freeze–thawing suggest that the peptidase might be used for industrial processes which require not only relatively low temperatures, but considerable durability.

The peptidase was also found to be activated by neutral salts, especially chlorides, in the order Na+>K+>Li+. The thermostable neutral metallo-endopeptidase thermolysin is also activated by salts, and the order of efficiency of ions in thermolysin's activation is the same as that observed for this enzyme [15]. It has been reported that proteins from halophilic bacteria have surfaces rich in acidic residues, which may interact with hydrated salt cations and enhance stability [16]. The low pI of this enzyme (3.8) suggests that it too may have an acidic surface able to interact with cations. This property, together with the activation by salts, indicates a halophilic character.

If further sequence information confirms this enzyme as a subtilisin-like peptidase, it will be possible to make use of an abundance of data available for mesophilic and thermophilic subtilases with respect to structure, catalytic properties, and stability. Over 200 distinct sequences of these enzymes from a wide range of organisms have been determined [17]. This, together with three-dimensional structures, makes them ideal candidates for structure–function comparisons. One may now predict that those features that strictly correlate with the characteristic features of psychrophilic enzymes should be missing from this peptidase since, though derived from a psychrophile, it lacks those features.


This work was supported by the European Union and was within the project ‘Extremophiles as Cell Factories’ (Project PL960488) in the Biotechnology Programme of the 4th Framework (Contract No. BIO4-CT96-0488). Partial support was provided by the Technology Fund of Iceland. We are grateful to Dr Suren Aghajanian for assistance with analytical gel filtration.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
View Abstract