Appl Microbiol Biotechnol DOI 10.1007/s00253-014-5877-2

BIOTECHNOLOGICALLY RELEVANT ENZYMES AND PROTEINS

Characterization of an extensin-modifying metalloprotease: N-terminal processing and substrate cleavage pattern of Pectobacterium carotovorum Prt1 Tao Feng & Christian Nyffenegger & Peter Højrup & Silvia Vidal-Melgosa & Kok-Phen Yan & Jonatan Ulrik Fangel & Anne S. Meyer & Finn Kirpekar & William G. Willats & Jørn D. Mikkelsen

Received: 19 February 2014 / Revised: 27 May 2014 / Accepted: 29 May 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract Compared to other plant cell wall-degrading enzymes, proteases are less well understood. In this study, the extracellular metalloprotease Prt1 from Pectobacterium carotovorum (formerly Erwinia carotovora) was expressed in Escherichia coli and characterized with respect to N-terminal processing, thermal stability, substrate targets, and cleavage patterns. Prt1 is an autoprocessing protease with an Nterminal signal pre-peptide and a pro-peptide which has to be removed in order to activate the protease. The sequential cleavage of the N-terminus was confirmed by mass spectrometry (MS) fingerprinting and N-terminus analysis. The optimal reaction conditions for the activity of Prt1 on azocasein were at pH 6.0, 50 °C. At these reaction conditions, KM was 1.81 mg/mL and kcat was 1.82×107 U M−1. The enzyme was relatively stable at 50 °C with a half-life of 20 min. Ethylenediaminetetraacetic acid (EDTA) treatment abolished activity; Zn2+ addition caused regain of the activity, but Zn2+addition decreased the thermal stability of the Prt1 enzyme presumably as a result of increased proteolytic autolysis. Electronic supplementary material The online version of this article (doi:10.1007/s00253-014-5877-2) contains supplementary material, which is available to authorized users. T. Feng : C. Nyffenegger : A. S. Meyer (*) : J. D. Mikkelsen Center for BioProcess Engineering, Department of Chemical and Biochemical Engineering, Technical University of Denmark, Søltofts Plads, Building 229, 2800 Kgs. Lyngby, Denmark e-mail: [email protected] P. Højrup : K.100 % of the initial activity (Fig. 4b); the increase in activity to 200 % relative activity with Zn2+ addition is presumed to be due to the base case not being completely saturated with

Appl Microbiol Biotechnol Table 1 Digestion of pre-pro-Prt1 by trypsin: peptide identification m/z

Position

Sequence

Obs

4,949.7 1,107.7 1,116.6 3,057.5 1,449.7 2,256.9 826.5 1,946.9 3,142.4 4,644.2 4,630.1 3,029.5 1,868.9 3,293.6 3,374.5 1,578.5 1,733.9

2–43 35–43 44–53 54–79 68–79 68–87 80–87 88–104 105–132 133–174 137–178 179–205 206–223 224–253 254–283 284–298 299–314

KHHHHHHPMSDYDIPTTENLYFQGAMASRPICSVIPPYILHR VIPPYILHR IIANGTDEQR HCAQQTLMHVQSLMVSHHPRPEPHEK VSHHPRPEPHEK VSHHPRPEPHEKLPAGQANR LPAGQANR SIHDAEQQQQLPGKLVR AEGQPSNGDIAVDEAYSYLGVTYDFFWK IFQRNSLDAEGLPLAGTVHYGQDYQNAFWNGQQMVFGDGDGK NSLDAEGLPLAGTVHYGQDYQNAFWNGQQMVFGDGDGKIFNR FTIALDVVAHELTHGITENEAGLIYFR QSGALNESLSDVFGSMVK QYHLGQTTEQADWLIGAELLADGIHGMGLR SMSHPGTAYDDELLGIDPQPSHMNEYVNTR EDNGGVHLNSGIPNR AFYLAAIALGGHSWEK

+ + + + + + + + + + + + + + + + +

303.2 1,156.7 1,521.8 905.5 1,061.6 2,414.2 2,931.5 Sequence coverage

315–317 318–326 327–339 340–347 340–348 349–369 349–373

AGR IWYDTLCDK TLPQNADFEIFAR HTIQHAAK HTIQHAAKR FNHTVADIVQQSWETVGVEVR FNHTVADIVQQSWETVGVEVRQEFL

+ + + + + + 99.2 %

Characters in italic (starting from KH enduing with CS) represent the sequence of the pre-peptide of pre-pro-Prt1; bold characters indicate the pro-peptide sequence of pro-Prt1 (starting with VI ending with LM). Plus sign indicates a positive observation in the reflector mode mass spectrum of the trypsindigested protein Obs observed mass and confirmed amino acid sequence

Zn2+. Although Zn2+ did not seem to affect the yields of the reaction after 10 min (Fig. 2), it was observed that the initial rates of the reactions with Zn2+ added were higher (Fig. 4b). Taken together with the thermal stability data, this result implies that Zn2+ ions not only activated the proteolytic activity of Prt1 but also activated the autolytic activity, so that while Prt1 becomes more active, it also becomes less stable in the presence of Zn2+. Calcium was not able to reconstitute the activity of Prt1 pre-treated by EDTA (Fig. 4b). Prt1 cleavage patterns on β-casein, ribonuclease A, and BSA

Fig. 1 Separation by SDS-PAGE (4–20 %) of pure potato lectin subjected to hydrolysis by the Prt1 protease. Protein markers (lane 1); potato lectin control (buffer) (lane 2); potato lectin incubated with Prt1 for 1, 5, 10, and 60 min (lane 3–6); and potato lectin after incubation with chymosin (lane 7)

To investigate the cleavage patterns of Prt1, three substrates (β-casein, ribonuclease A, and BSA) were used for digestion by the protease, and all three substrates were effectively cleaved by this enzyme. After mapping the sequence of the peptides detected by MALDI TOF/TOF analysis, 16, 23, and 17 cleavage sites were identified for

Appl Microbiol Biotechnol Fig. 2 Surface response as a function of temperature and pH on the Prt1 activity. Ptr1 was incubated with azocasein, and the activity was measured by recording the release of soluble azocasein peptides by 440 nm after 10-min incubation: a at 0.5 mM ZnCl2 and b at 50 °C. The optimal conditions were determined to be as follows: pH 6.1, temperature 50 °C, and 0.5 mM ZnCl2. Further information on the surface response is given in Supplementary Material, Table S1

β-casein, ribonuclease A and BSA respectively, i.e., a total of 56 cleavage sites for all three substrates. Based on these 56 cleavage sites, the frequency of each amino acid occurrence at the cleavage site (P4, P3, P2, P1, P1′, P2′, P3′, and P4’, nomenclature of Schechter and Berger (1967)) was determined (Table 2). A consensus sequence of amino acids around the cleavage site (P4-P3-P2-P1-

P1′-P2′-P3′-P4′) as being (Ala)-(Pro/Glu/Gln)-(Lys/Glu/ Gln)-(Pro)-(Ala/Val/Phe)-(Ala/Leu)-(Ser)-(Ser/Glu) could be proposed (Fig. 5). The results thus suggest that Prt1 has a preference for cleaving peptide bonds constituted by a proline followed by a hydrophobic residue. For comparison, in a previous study, the metalloprotease Bacillus thermolysin was found to be much less specific

Fig. 3 Plot of residual activity of Prt1 after heat inactivation (LnU) versus heating time: a mature Prt1, b Prt1 with 1 mM of Zn2+, c Prt1 with 1 mM of Ca2+, and d Prt1 with 1 mM of Zn2+ and Ca2+

Appl Microbiol Biotechnol

a Relative activity (%)

Fig. 4 a The effects of various protease inhibitors and chelators on relative Prt1 activity. b The effects of Zn2+ and Ca2+ (1 mM) on the recovery of Prt1 activity after EDTA treatment: E* is Prt1 after incubation and subsequent removal of EDTA. E* + Zn is the Prt1 reconstituted with ZnCl2, and E* + Ca is the Prt1 reconstituted with CaCl2. Relative activities are recorded in

140 120 100 80 60 40 20 0

Relative activity (%)

b

200 150 100 50 0

(Keil 1992). The relatively high frequency of preference for proline at or near the cleavage site corroborated that Prt1 may have high affinity for attacking proline-rich and hydroxy-proline-rich plant glycoproteins. Prt1 activity on extensin extracted from plants The immuno-glycan microarrays study showed that Prt1 could catalyze the degradation of extensin extracted from various plant species (Fig. 6). This type of assay is based on the principle that enzyme activities can be detected by a loss or reduction of antibody binding to the substrate after enzyme treatment. Reduction in antibody binding may result in modification or degradation of epitopes (Fangel et al. 2012). In this case, a set of well-established anti-extensin antibodies were used, and the substrates were array-immobilized crude extracts from a variety of plant species, all of which apart from Ginkgo biloba contained extensin, as evidenced by antibody binding to buffer-treated control arrays (Fig. 6a). In all the

other species, the signals obtained from all the antiextensin antibodies were reduced (and in some cases abolished, Fig. 6b) on arrays treated with Prt1 compared to buffer-treated arrays. The fact that xyloglucanase did not lead to a general decrease in antibody binding indicates that the observed reduction of signals was a Prt1-specific effect (Fig. 6a). Although the antibodies used have all been characterized as extensin-specific, their epitopes have not been fully resolved. However, all are thought to recognize carbohydrate moieties (Smallwood et al. 1995). The observed decreases in antibody binding therefore indicate an indirect effect, namely, that degradation of core proteins by Prt1 caused release of epitope-bearing glycan side chains. These data are significant because they indicate that Prt1 is capable of accessing protein substrates within the context of native extensin glycosylation. This may have important implications for the practical utilization of Prt1 during biomass processing.

Appl Microbiol Biotechnol Table 2 Frequency of residues (%) around the substrate cleavage site when cleaved by Prt1

Residue G A V L I F Y W P S C T M H K R D E N Q

P4 5 11 7 7 4 2 9 2 7 5 0 5 0 4 5 7 5 4 2 9

P3 2 5 4 9 4 2 5 0 14 5 4 4 0 4 2 4 4 11 4 16

Cleavage Site P2 P1 P1' P2' P3' P4' 2 0 0 0 7 0 7 7 13 13 7 5 9 2 20 7 5 5 4 4 9 11 5 7 0 0 7 5 4 5 2 7 14 4 5 4 7 9 7 5 4 5 0 2 2 2 0 0 4 13 4 0 9 9 5 4 5 9 11 11 2 5 0 4 5 0 7 5 0 2 7 2 5 2 4 5 0 0 4 7 2 7 4 5 11 9 4 9 7 5 2 4 0 5 9 5 0 2 4 2 2 5 11 9 5 4 5 11 9 7 0 2 0 4 11 4 2 5 4 9

Observations are based on a total of 56 cleavage events when using βcasein, ribonuclease A, and BSA as substrates. Frequencies between 5 and 10 % are presented in light green and those above 10 % in dark green. Nomenclature follows that of Schechter and Berger (1967): PN and PN′ being the Nth residue before and the Nth residue after the cleavage site, respectively

Secondary structure elements by CD CD spectra were measured for Prt1 and are shown in Fig. 7a. A comparison of the predicted secondary structure contents from the spectra to the structural content of thermolysin showed that 15 % of the residues in Prt1 were predicted to be in a helix conformation, 33 % in a sheet, 22 % in a turn, and 30 % in an unordered conformation (Table 3). The secondary structure contents deviated significantly from the structure of thermolysin, which exhibits higher helix content, but lower sheet content (Table 3). Thermal denaturation analysis of Prt1 by assessment of the residual native state mean residue molar ellipticity as a function of temperature revealed that the Tm, i.e., the

temperature at the transition midpoint (where 50 % of the native state signal remains) of the enzyme, was 52.6 °C (Fig. 7b).

Mutations generated to improve thermal stability To assess the options for generating a calcium binding site, in order to improve the thermal stability of Prt1, we introduced four site mutations based on the sequence alignment between the N-terminus of thermolysin and Prt1. Both proteases share a conserved motif with the sequence LPGX1LX2 where X2 is a hydrophobic residue. This is the most N-terminal conserved motif in the consensus sequence of Prt1 and thermolysin and is followed in the latter by a calcium binding site. In Prt1, the amino acid sequence following the conserved motif seems shortened and lacks the calcium complexing aspartic acid residues. We therefore chose sites located in the immediate C-terminal vicinity of the conserved motif to mutate or insert amino acids (InsD, InsF, MutA, or loop). All four enzyme mutants were successfully expressed and purified, but their activity was lower than that of the wild type (Fig. S6a in the Supplementary Material). The CD spectra of the mutants InsD and InsF deviated strongly from the wild-type signal (Fig. 8a), exhibiting random coil spectra with little helix but significant sheet contents. Accordingly, the activity of these mutated enzymes was significantly decreased (Fig. S6a in the Supplementary Material). In contrast, the CD measurements showed that the MutA and loop mutant enzymes had secondary structure contents closer to those of thermolysin than the native wild-type Prt1, i.e., they contained more helix than the Prt1 wild type but less sheet structures, while the percentage of turns stayed the same (Table S4 in the Supplementary Material and Fig. 8a). Accordingly, these two mutants MutA and loop had slightly higher Tm values than the wt (Fig. 8b). For MutA, a Tm of 58.2 °C was determined, while it had only 75 % of the wild-type activity under optimal conditions (for the wild type) (Fig. 7a). The Tm for the loop-mutant was determined to be 54.3 °C, but its activity was only 50 % of the wild-type activity when assessed at 50 °C (Fig. S6a in the Supplementary Material and Fig. 8b). Compared to the wild type, the two mutant enzymes MutA and loop also lost activity significantly

Fig. 5 Cleavage patterns of Prt1 indicating the proposed elevated preference for cleavage with P in position P1 and A, V, or F in position P1′ in the substrate (re. Table 2). PN and PN′ being the Nth residue, respectively, found before and after the cleavage site

Appl Microbiol Biotechnol

Fig. 6 a CDTA extracts from six different plant species were printed in a microarray format as described in Moller et al. (2007). Each extract was represented by two replicates and four dilutions resulting in a total of eight spots per sample. Identical arrays were treated with Prt1 at 0.5 mg/mL, Prt1 at 0.05 mg/mL, commercial xyloglucanase at 0.5 U/mL, and buffer, respectively. Afterward, arrays were probed with anti-extensin monoclonal antibodies (JIM11, JIM12, JIM20, and LM3), and detected binding for each antibody is presented as a sub-heat map where color intensities are proportional to mean spot signal values. The highest mean signal Fig. 7 a Far-UV circular dichroism spectra of Prt1 at 25 °C, recorded in 20 mM phosphate buffer, pH 7.0. b Temperature transitions of Prt1. The fraction of the native state mean residue molar ellipticity is shown as a function of temperature

value was set to 100, and all other values were adjusted accordingly. Different array treatments are indicated at the top, while extracted plant species are listed on the right of the heat map. b JIM12 and LM3 antibody signals detected for Buxus sempervirens CDTA extracts are represented in two sub-heat maps. The normalized mean signal values for the four dilutions are shown individually. Below each heat map, a part of the corresponding arrays is shown in order to visualize the correlation between heat map values and spot intensities observed on the arrays

Appl Microbiol Biotechnol Table 3 Prt1 secondary structure determination from far-UV-CD measurements

Prt1 Thermolysin

Helix (%)

Sheet (%)

Turn (%)

Unordered (%)

Total (%)

15 41

33 13

22 47

30

100 100

The percentage of α-helix, β-sheets, turns, and unordered loops were determined from far-UV-CD measurements using CDSSTR (Sreerama and Woody 2000) through the Dichroweb analysis web server (Whitmore and Wallace 2007). Secondary structure contents of thermolysin were determined from the pdb file 1OS0

faster than the wild-type Prt1 when incubated at 40 °C or above for 30 min (Fig. S6b in the Supplementary Material). Mutant enzymes with higher Tm but lowered activity and thermal stability may immediately seem contradictory; however, taken together, the data essentially suggest that elevated temperatures did not affect the secondary structure but distorted the catalytic site. Our small set of mutations thus suggests that the N-terminus of Prt1 defines its thermal stability and helps maintain the enzymatic activity at elevated temperatures.

Discussion In this study, Prt1 was demonstrated to undergo a pre-pro-auto activation process, which appeared to be similar to that of other extracellular proteases (Wu and Chen 2011). Autoprocessing of protease N-termini, as observed for the pre-pro-Prt1 N-terminus, is very common in bacterial extracellular metalloproteases (Hase and Finkelstein 1993). The pre-peptide is automatically cleaved off during secretion from the host, while the pro-peptide is removed after secretion, leading to the active protease (Eijsink et al. 2011). Although the E. coli host did not secrete the pre-pro-Prt1, the prepeptide was removed during the folding of the enzyme as proved by the SDS-PAGE study (Fig. S1: Extract in the Supplementary Material). Although the N-terminal His6 tag Fig. 8 Far-UV circular dichroism spectra of wild-type Prt1 in comparison with four designed Prt1 variants at 25 °C (a). The data was recorded in 20 mM phosphate buffer, pH 7.0. Temperature transitions of wildtype Prt1 and the variants MutA and loop (b). The fraction of the native state mean residue molar ellipticity is shown as a function of temperature. The color code is the same as in (a), and the Tm values are given in the figure

in the pre-peptide was lost early in the process, the seven adjacent histidines located in the N-terminus of the pro-Prt1 apparently provided sufficient binding to the Ni-NTA column, and the pro-Prt1 eluted at a relatively low concentration of imidazole (68 mM) (data not shown). The predicted N-terminal signal pre-peptide in the UniProt database (www.uniprot.org) was about 20 amino acids long, which means that the cleavage site would be A46. However, the N-terminal study revealed that the pre-peptide was cleaved prior to V35, indicating that the UniProt prediction may need adjustment. In the present work, the positive effect of the Nterminal processing was evident from the increase in activity taking place during incubation at 37 °C (Fig. S3 in the Supplementary Material). However, extended incubation of the enzyme also caused degradation that resulted in the loss of activity. It has previously been reported that the Nterminal pre-pro peptide is functioning as a chaperone for correct folding of the protease, before it is excised automatically after expression (Inouye 1991). In order to verify that theory, the gene of the active protease without the pre-pro peptides was designed and ligated back to the pETM11 vector. When expressed, the majority of the truncated protease was insoluble (data not shown). Based on sequence homology assessment, the Prt1 from P. carotovorum EC14 belongs to the thermolysin-like proteases (Kyostio et al. 1991). Prt1 has six active-site residues, three zinc-binding residues, and five out of seven substratebinding residues as compared with that of thermolysin

Appl Microbiol Biotechnol

(Kyostio et al. 1991). Prt1 has 68.3 and 20 % sequence similarity with Bacillus thermolysin Fbk2 (UniProt G3GBT6 (www.uniprot.org)) and Bacillus thermolysin (Uniprot P00800 (www.uniprot.org)), respectively. The data expand the previous studies on the Prt1 enzyme published by Kyostio et al. (1991) mainly with respect to providing new knowledge about enzyme maturation, thermal stability, substrate specificity, and cleavage pattern. For thermolysin, it has been shown that during maturation, a stretch of amino acids comprising the pro-peptide/N-terminal boundary needs to insert productively into the catalytic center, a process that obviously requires conformational flexibility (Sauter et al. 1998). During maturation, the N-termini of the mature protease need to relocate from a position in the catalytic center to their final position in the mature enzyme (Sauter et al. 1998). Intensive studies have shown that the N-terminal β-sheets as well as the binding site of calcium ions are the most deleterious points of mutations. Mutations taking place in the N-terminal calcium binding site will cause dramatic changes of the thermal stability (Eijsink et al. 2011; Van den Burg et al. 1998). Thermolysin requires one zinc ion and four calcium ions to stabilize its structure. It is also known that thermolysin and thermolysin-like proteases (TLPs) all require calcium ions for proper folding and structure stability (Veltman et al. 1998; Alexander et al. 2001). However, Ca2+ was apparently not required for activation and did not induce activation (Fig. 4b) or stabilize (or destabilize) the thermal robustness of the Prt1 protease (Fig. 3c). The sequence alignment of N-terminus of Prt1 with thermolysin and TLPs showed that Prt1 lacks the N-terminal calcium binding site (D289, D291, and Q293 from thermolysin), which is a significant difference between Prt1 and thermolysin and some TLPs. This difference also explains the lower thermal stability of Prt1, since the N-terminus of Prt1 is not stabilized by calcium and is thus vulnerable to destabilization. Hence, when the Zn2+ ions activate the proteolytic as well as the autolytic activity of Prt1, the unstable N-terminus of Prt1 may very likely undergo autolytic digestion after translocation to the active pocket. The sequence analysis of the Prt1 indicates that it belongs to the same family as thermolysin (classified as family M4 in the MEROPS peptidase database). According to the MEROPS database, this family of proteases usually has a cleavage preference with valine, leucine, isoleucine, and phenylalanine at P1′. However, our data on the Prt1 metalloprotease were not in accordance with this general rule. With β-casein as a substrate, proline showed the highest frequency in position P1, which may indicate a slightly increased preference for cleaving after proline residues. However, it must be noted that β-casein in itself is particularly rich in proline residues. Proline is also present in position P1 when using ribonuclease A and BSA as substrates, although to a much lesser extent than with β-casein. The cleavage of potato

lectin and plant extensin further verified the ability of Prt1 to catalyze the hydrolysis of highly glycosylated proteins, signifying the prospects of using the Prt1 protease for degradation of plant cell wall structural proteins.

References Alexander PA, Ruan B, Bryan PN (2001) Cation-dependent stability of subtilisin. Biogeosciences 40:10634–10639 Bauer S, Vasu P, Persson S, Mort AJ, Somerville CR (2006) Development and application of a suite of polysaccharide-degrading enzymes for analyzing plant cell walls. Proc Natl Acad Sci U S A 103:11417– 11422 Beg QK, Kapoor M, Mahajan L, Hoondal GS (2001) Microbial xylanases and their industrial applications: a review. Appl Microbiol Biotechnol 56:326–338 Bell KS, Sebaihia M, Pritchard L, Holden MTG, Hyman LJ, Holeva MC, Thomson NR, Bentley SD, Churcher LJC, Mungall K, Atkin R, Bason N, Brooks K, Chillingworth T, Clark K, Doggett J, Fraser A, Hance Z, Hauser H, Jagels K, Moule S, Norbertczak H, Ormond D, Price C, Quail MA, Sanders M, Walker D, Whitehead S, Salmond GPC, Birch PRJ, Parkhill J, Toth IK (2004) Genome sequence of the enterobacterial phytopathogen Erwinia carotovora subsp. atroseptica and characterization of virulence factors. Proc Natl Acad Sci U S A 101(30):11105–11110 Bhat MK (2000) Cellulases and related enzymes in biotechnology. Biotechnol Adv 18:355–383 Brinckmann J (2005) Collagens at a glance. Top Curr Chem 247:1–6 Buckeridge MS, Souza TA, Murakami MT, Ward RJ, Prade RA, Polizeli ML (2012) Functional characterization and oligomerization of a recombinant xyloglucan-specific endo-β-1,4-glucanase (GH12) from Aspergillus niveus. Biochim Biophys Acta 1824:461–467 Craig R, Beavis RC (2004) TANDEM: matching proteins with mass spectra. Bioinformatics 20:1466–1467 de Vries RP, Visser J (2001) Aspergillus enzymes involved in degradation of plant cell wall polysaccharides. Microbiol Mol Biol Rev 65:497– 522 Dow JM, Davies HA, Daniels MJ (1998) A metalloprotease from Xanthomonas campestris that specifically degrades proline/ hydroxyproline-rich glycoproteins of the plant extracellular matrix. Mol Plant Microbe Interact 11:1085–1093 Eijsink VG, Matthews BW, Vriend G (2011) The role of calcium ions in the stability and instability of a thermolysin-like protease. Protein Sci 20:1346–1355 Fangel JU, Pedersen HL, Vidal-Melgosa S, Ahl LI, Salmean AA, Egelund J, Rydahl MG, Clausen MH, Willats WG (2012) Carbohydrate microarrays in plant science. Methods Mol Biol 918:351 Fedatto LM, Silva-Stenico ME, Etchegaray A, Pacheco FTH, Rodrigues JLM, Tsai SM (2006) Detection and characterization of protease secreted by the plant pathogen Xylella fastidiosa. Microbiol Res 161:263–272 Hanes CS (1932) Studies on plant amylases: the effect of starch concentration upon the velocity of hydrolysis by the amylase of germinated barley. Biochem J 26:1406–1421 Hase CC, Finkelstein RA (1993) Bacterial extracellular zinc-containing metalloproteases. Microbiol Rev 57:823–837 Inouye M (1991) Intramolecular chaperone: the role of the pro-peptide in protein folding. Enzyme 45:314–321 Keil B (1992) Specificity of proteolysis. Springer-Verlag, BerlinHeidelberg-New York, p 335

Appl Microbiol Biotechnol Kieliszewski MJ, Lamport DT (1994) Extensin: repetitive motifs, functional sites, post-translational codes, and phylogeny. Plant J 5:157– 172 Kyostio SR, Cramer CL, Lacy GH (1991) Erwinia carotovora subsp. carotovora extracellular protease: characterization and nucleotide sequence of the gene. J Bact 173:6537–6546 Mei YZ, Chen YR, Zhai RY (2013) Cloning, purification and biochemical properties of a thermostable pectinase from Bacillus halodurans M29. J Mol Catal B: Enzymatic 94:77–81 Moller I, Sørensen I, Bernal AJ, Blaukopf C, Lee K, Øbro J, Pettolino F, Roberts A, Mikkelsen JD, Knox JP, Bacic A, Willats WG (2007) High-throughput mapping of cell-wall polymers within and between plants using novel microarrays. Plant J 50:1118–1128 Olsen DT, Peng L, Ellgaard L, Højrup P, Houen G (2013) Purification and characterization of a soluble calnexin from human placenta. Protein Express Purif 92:105–111 Riffel A, Brandelli AB, de Cláudia MS, Gustavo HMFE, Marcos NT, Flavio CA (2007) Purification and characterization of a keratinolytic metalloprotease from Chryseobacterium sp. kr6. J Biotechnol 128: 693–703 Sauter NK, Mau T, Rader SD, Agard DA (1998) Structure of α-lytic protease complexed with its pro region. Nature Struct Biol 5: 945–950 Schechter I, Berger A (1967) On the size of the active site in proteases. I. Papain. Biochem Biophys Res Commun 27(2): 159–162 Showalter AM (1993) Structure and function of plant cell wall proteins. Plant Cell 5:9–23 Silva IR, Larsen DM, Meyer AS, Mikkelsen JD (2011) Identification, expression, and characterization of a novel bacterial RGI lyase

enzyme for the production of bio-functional fibers. Enzyme Microb Technol 49:160–166 Silva IR, Larsen DM, Jers C, Derkx P, Meyer AS, Mikkelsen JD (2013) Enhancing RGI lyase thermostability by targeted single point mutations. Appl Microbiol Biotechnol 97(22):9727–9735 Smallwood M, Martin H, Knox JP (1995) An epitope of rice threonineand hydroxyproline-rich glycoprotein is common to cell wall and hydrophobic plasma membrane glycoproteins. Planta 196:510–522 Sreerama N, Woody RW (2000) Estimation of protein secondary structure from CD spectra: comparison of CONTIN, SELCON and CDSSTR methods with an expanded reference set. Anal Biochem 282:252–260 Toth IK, Bell KS, Holeva MC, Birch PRJ (2003) Soft rot erwiniae: from genes to genomes. Mol Plant Pathol 4:17–30 Van den Burg B, Vriend G, Veltman OR, Venema G, Eijsink VGH (1998) Engineering an enzyme to resist boiling. Proc Natl Acad Sci U S A 95:2056–2060 Veltman OR, Vriend G, Berendsen HJ, Van den Burg B, Venema G, Eijsink VGH (1998) A single calcium binding site is crucial for the calcium-dependent thermal stability of thermolysin-like proteases. Biochemistry 37:5312–5319 Whitmore L, Wallace BA (2007) Protein secondary structure analyses from circular dichroism spectroscopy: methods and reference databases. Biopolymers 89:392–400 Wu JW, Chen XL (2011) Extracellular metalloproteases from bacteria. Appl Microbiol Biotechnol 92:253–262 Zeuner B, Luo JQ, Nyffenegger C, Aumala V, Mikkelsen JD, Meyer AS (2014) Optimizing the biocatalytic productivity of an engineered sialidase from Trypanosomarangeli for 3′-sialyllactose production. Enzyme Microb Technol 55:85–93