A Tale of Two Cell Factories
Neef, Jolanda
DOI:
10.33612/diss.99279788
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Neef, J. (2019). A Tale of Two Cell Factories: Heterologous protein secretion in Bacillus subtilis and Lactococcus lactis. University of Groningen. https://doi.org/10.33612/diss.99279788
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
A Tale of Two Cell Factories
Heterologous protein secretion in
Bacillus subtilis and Lactococcus lactis
Jolanda Neef
of Groningen (The Netherlands).
The studies presented in this thesis were financially supported by the Top Institute Pharma (Project T4‐213 and T4‐502) and Genencor‐DuPont (Palo Alto, CA, USA).
Publication of this thesis was partly supported by the Graduate School of Medical Sciences of the University of Groningen.
A Tale of Two Cell Factories – Heterologous protein secretion in Bacillus subtilis and Lactococcus
lactis Dissertation of the University of Groningen ISBN: 978‐94‐034‐2101‐8 (printed book) ISBN: 978‐94‐034‐2100‐1 (digital) Cover: Abstract interpretation of the Two Cell Factories, Bacillus subtilis and Lactococcus lactis Cover design: Amber Nieuwenweg and Jolanda Neef Printed by: Ipskamp Printing, Enschede Copyright © Jolanda Neef, 2019
A Tale of Two Cell Factories
Heterologous protein secretion in Bacillus subtilis and Lactococcus lactis
PhD thesis
to obtain the degree of PhD at the
University of Groningen
on the authority of the
Rector Magnificus Prof. C. Wijmenga
and in accordance with
the decision of the College of Deans
This thesis will be defended in public on
Wednesday 6 November 2019 at 12:45 hours
by
Jolanda Neef
born on 14 August 1978
Co‐supervisor
Dr. G. Buist
Assessment committee
Prof. J. Kok
Prof. D.J. Scheffers
Prof. D. van Sinderen
Dr. Sjouke Piersma
Drs. Rocío Aguilar Suárez
Voor Joep en NienkeChapter 1
General introduction and scope of this thesis
9
Manuscript to be submitted
Chapter 2
Efficient production of secreted staphylococcal antigens in a
43
non‐lysing and proteolytically reduced Lactococcus lactis strain
Published in Appl Microbiol Biotechnol. 2014, 98: 10131‐41
Chapter 3
Versatile vector suite for the extracytoplasmic production and
67
purification of heterologous His‐tagged proteins in Lactococcus
lactis
Published in Appl Microbiol Biotechnol. 2015, 99: 9037‐48
Chapter 4
A Lactococcus lactis expression vector set with multiple affinity
89
tags to facilitate isolation and direct labeling of heterologous secreted
proteins
Published in Appl Microbiol Biotechnol. 2017, 101: 8139–49
Chapter 5
Human antibody responses against non‐covalently cell wall‐bound
107
Staphylococcus aureus proteins
Published in Sci Rep. 2018, 8: 3234
Chapter 6
Intramembrane protease RasP boosts protein production in Bacillus
131
Published in Microb Cell Fact. 2017, 16: 57
Chapter 7
Relative contributions of individual Sec pathway components to high‐
149
level enzyme secretion by Bacillus subtilis
Manuscript to be submitted
Chapter 8
General summary and discussion
169
Chapter 9
Nederlandse samenvatting (voor de leek)
179
Chapter 10
Dankwoord – Acknowledgements
189
Curriculum Vitae (NL)
197
Curriculum Vitae (ENG)
198
List of publications
199
Chapter 1
General introduction and scope of this thesis
A Tale of Two Cell Factories
Heterologous protein secretion in Bacillus subtilis and Lactococcus lactis
Jolanda Neef, Jan Maarten van Dijl*, Girbe Buist*
*Authors contributed equally to this work
Manuscript to be submitted
Supplementary material available at
https://www.dropbox.com/sh/ncxgv6d6vl505f6/AABEOJjZepDTijRqFVkyFsFda?dl=0
Abstract
Secreted recombinant proteins are of great significance for industry, healthcare and a sustainable bio‐ based economy. Consequently, there is an ever‐increasing need for efficient production platforms to deliver such proteins in high amounts and high quality. Gram‐positive bacteria, particularly bacilli such as Bacillus subtilis, are favored for the production of secreted industrial enzymes. Nevertheless, protein production in the B. subtilis cell factory can be very challenging due to bottlenecks in the general (Sec) secretion pathway as well as this bacterium’s intrinsic capability to secrete a cocktail of highly potent proteases. This has placed another Gram‐positive bacterium, Lactococcus lactis, in the focus of attention as an alternative, non‐proteolytic, cell factory for secreted proteins. Here we review our current understanding of the general secretion machinery employed by B. subtilis and L. lactis to guide proteins from their site of synthesis, the cytoplasm, into the fermentation broth. In addition, we address species‐specific signatures of signal peptide that direct proteins into the respective secretory pathways. An advantage of this cell factory comparison is that it identifies new opportunities for protein secretion pathway engineering to remove or bypass current production bottlenecks. Noteworthy new developments in cell factory engineering are the mini‐Bacillus concept, highlighting potential advantages of massive genome minimization, and the application of thus far untapped ‘non‐ classical’ protein secretion routes. Altogether, it is foreseen that engineered lactococci will find future applications in the production of high‐quality proteins at the relatively small pilot scale, while engineered bacilli will remain a favored choice for protein production in bulk.Introduction
To thrive and survive in different ecological niches, bacteria secrete a wide range of different proteins. This allows them to take optimal advantage of their habitat, as exemplified by the secretion of proteases that facilitate the acquisition of peptides and amino acids, be it in the soil through the degradation of dead organic matter or in the human body upon invasive disease (Richardson et al. 2015). The secretion of proteins from their site of synthesis, the cytoplasm, to the extracellular milieu is not a spontaneous process, but it requires membrane channels and an intricate machinery that converts metabolic energy into a force that drives proteins through the membrane. Importantly, many integral membrane proteins are inserted into the bacterial cytoplasmic membrane via the same export pathway that is followed by secreted proteins. Consequently, protein translocation across the cytoplasmic membrane is invariably an essential process (Yuan et al. 2010). In most cases, the translated proteins that are transported across the membrane need to stay in a soluble unfolded state, which is supported by so‐called chaperones. In addition, proteins destined for export from the cytoplasm are labeled with an N‐terminal signal peptide, which serves different roles in membrane targeting and translocation (Dalbey et al. 2012). Generally, nascent signal peptides emerging from the ribosome are recognized by dedicated chaperones that enhance membrane targeting and association with the machinery for membrane translocation (Pallen et al. 2003; Zanen et al. 2006; Tjalsma et al. 2000; Bechtluft et al. 2010; Tsirigotaki et al. 2017). During or shortly after membrane translocation, the signal peptide is separated from the exported protein by signal peptidases and the protein will fold into its stable and active conformation at the trans side of the membrane (Dalbey et al. 2012). The latter folding step could in principle occur spontaneously but, in most cases, it involves particular folding catalysts (De Geyter et al. 2016; Sarvas et al. 2004). The folding step is considered important for release of the exported protein into the extracellular environment, if only to avoid unwanted interactions with the cell envelope, to prevent aggregation, or to limit degradation as unfolded proteins are highly susceptible to proteases. If the exported proteins are synthesized with particular retention signals, they will be retained in the bacterial cell envelope although increasing evidence indicates that this retention is often only temporal (Tjalsma et al. 2000; Tjalsma et al. 2004; Dreisbach et al. 2011).
Gram‐positive bacteria are particularly well known for their intrinsic capacity to secrete proteins directly into the extracellular milieu (Anne et al. 2017). This relates to the relatively simple structure of their cell envelope where, in the most elementary form, the membrane is surrounded by a relatively thick cell wall composed of peptidoglycan and several other polymers. Importantly, the cell wall has a porous structure, allowing the passage of proteins upon membrane translocation. A Gram‐positive bacterium that is well‐known for its high capacity to secrete proteins is Bacillus subtilis. In nature, B.
subtilis flourishes in the challenging niches provided by the soil and plant rhizosphere, where it secretes a cocktail of different degradative enzymes to obtain nutrients (e.g. sugars, amino acids, phosphate and metal ions), and to defend itself from chemical and biological insults (Earl et al. 2008). Due to its well‐developed and highly efficient secretion machinery, B. subtilis has also become a popular biotechnological ‘cell factory’ for the bulk production of commercially relevant secreted proteins, in particular technical enzymes such as proteases, amylases and xylanases (Van Dijl and Hecker 2013; Harwood and Cranenburgh 2007). Industrial enzymes are used in many different markets including personal care, food and beverages, detergents, textiles, animal feed, chemicals and biofuels. They permeate every aspect of our daily life and the markets are consequently large. In particular, the global market for food enzymes is forecast to grow from $1.8 billion in 2017 to about $2.2 billion in 2022, whereas the global industrial enzymes market is forecast to growth from $5.5 billion in 2018 to $7.0 billion by 2023 (Business Communications Company Inc., 2018). Today there is a need for new, improved and more versatile enzymes in order to develop more novel, sustainable and economically competitive production processes. Accordingly, the various secretion systems of B. subtilis and the full complement of secreted proteins, collectively termed the ‘secretome’, have been intensely investigated. Importantly, the studies on B. subtilis also provided new insights into the secretion pathways that are present in other Gram‐positive bacteria, especially bacilli, lactic acid bacteria, streptococci and staphylococci (Tjalsma et al. 2004; Sibbald et al. 2006).
In the biotechnological context, it should be noted that Gram‐positive bacteria lack the outer membrane that is present in Gram‐negative bacteria, such as Escherichia coli. This absence of an outer membrane not only simplifies the secretion process, but it also has the great advantage that Gram‐ positive bacteria such as B. subtilis and Lactococcus lactis lack lipopolysaccharides, also known as endotoxins (Hohmann et al. 2016). To avoid the barrier imposed by the outer membrane, E. coli‐based expression systems often aim at protein production in the cytoplasm. This allows the accumulation of massive amounts of product in the cytoplasm but, at the same time, it significantly complicates the downstream processing of the produced proteins. This is further complicated by the fact that proteins overproduced in the cytoplasm often aggregate, forming so‐called inclusion bodies from which the product can only be recovered upon cell disruption and treatment with strong denaturing agents, such as urea. Conversely, the downstream processing of proteins secreted into the growth medium of Gram‐positive bacteria is generally easy, non‐denaturing and cost‐effective (Hohmann et al. 2016). Amongst the Gram‐positive bacteria, B. subtilis and related bacilli, such as Bacillus amyloliquefaciens and Bacillus licheniformis, offer the additional advantage that they can be readily fermented at large scale using cheap carbon sources, which adds to the cost‐effectiveness. This is important as industrial enzymes, in contrast to biopharmaceuticals, are marketed at relatively low prices (Hohmann et al. 2016). A particular advantage of B. subtilis is that, due to the absence of toxins, products from this cell
factory have obtained the Generally Regarded as Safe (GRAS) status from the United States Food and Drug Administration (FDA). In addition, B. subtilis is on the list of organisms with Qualified Presumption of Safety (QPS) status assembled by the European Food Safety Authority (EFSA) (Hohmann et al. 2016). This makes B. subtilis very attractive as a cell factory, not only for food enzymes, but perhaps even more so for biotherapeutics that are used to treat an increasingly wide range of serious diseases. The worldwide sales for such protein drugs reached $138 billion in 2010 and are estimated to reach $320 billion by 2020 (PharmiWeb.com). Despite all the advantages, production of secreted heterologous proteins in B. subtilis is frequently challenging. In particular, the fact that B. subtilis naturally secretes multiple proteases can interfere with the production of proteins that are sensitive to degradation (Harwood and Cranenburgh 2007; Pohl and Harwood 2010). Even though strains lacking up to ten different proteases have been developed, this problem has still not been completely overcome (Supplementary Table 1). Moreover, various other secretion bottlenecks are also known to exist at the levels of membrane targeting, translocation and post‐translocational protein folding (Hohmann et al. 2016). Thus, while B. subtilis is overall a highly attractive cell factory that is intensively exploited in the industry (Pohl et al. 2013), there is room for additional, preferably Gram‐positive, bacterial platforms for secretory protein production in which the afore‐mentioned disadvantages can be circumvented. One such platform could be L. lactis, a food‐grade Gram‐positive bacterium thus far used mostly in the dairy industry. Recent studies have shown that L. lactis can be applied in the production of various protease‐sensitive proteins, including potential vaccines (Neef et al. 2014; Neef et al. 2015; Samazan et al. 2015; Romero Pastrana et al. 2017). Intriguingly, despite its important role in the degradation of casein during cheese ripening, L. lactis produces no more than three major proteases that can have an impact on protein production. The major secreted protease needed for cheese production is PrtP, but this plasmid‐ encoded enzyme is dispensable for growth of L. lactis on media other than milk (Kok et al. 1985; Mierau et al. 1997). L. lactis possesses two additional house‐keeping proteases, namely the extracytoplasmic protease HtrA and the cytoplasmic protease ClpP. Deletion of the htrA gene from L. lactis results in highly reduced proteolysis of various heterologously produced and secreted proteins (Supplementary Table 2), showing a major function in extracytoplasmic protein turnover (Cortes‐Perez et al. 2006). In contrast, deletion of the clpP gene either by itself or in combination with an htrA deletion does not further enhance extracellular protein production. This was exemplified with the secreted nuclease Nuc of Staphylococcus aureus and the human papillomavirus E7 protein fused to Nuc (Cortes‐Perez et al. 2006). Nonetheless, a clpP htrA double mutant was shown to be more resistant to temperature and ethanol stress than the respective single mutants. This indicates that the double mutant may be more robust and, therefore, better suited for protein production (Cortes‐Perez et al. 2006). The notion that
several proteinaceous antigens from S. aureus can be produced and secreted in L. lactis, but not in wild‐type B. subtilis or mutant strains lacking particular protease genes. This was only recently shown to be possible in an engineered ‘mini‐Bacillus’ strain lacking 36% of the genome (Aguilar Suarez et al. 2018). On the other hand, although L. lactis displays less proteolytic activity than B. subtilis, it has to be noted that the overall production yields achieved with L. lactis remain significantly lower than those achieved with B. subtilis (Neef et al. 2014). At the laboratory scale, the difference in yields is about 100‐fold with B. subtilis secreting proteins in the gram per liter range. Under optimized fermentation conditions B. subtilis can easily produce 25 grams of protein per liter culture (Van Dijl and Hecker 2013; Hohmann et al. 2016). This difference in productivity can be attributed to multiple factors. In the first place, B. subtilis can be grown to much higher cell densities than L. lactis. Further, L. lactis has the tendency to acidify its growth medium due to lactate production, which may set a limit to growth. Importantly, there also appear to be substantial differences in the secretion capacity of both organisms, which may relate to differences in the machinery for protein export and the evolutionary pressures exerted on the respective organisms. While the secretion machinery of L. lactis has become highly geared towards growth in milk, B. subtilis has evolved in an environment where nutrients come in highly variable composition, quality and structure, depending on available dead organic matter. From a protein production point of view, the use of both B. subtilis and L. lactis is attractive due to their complementarity and the possibility to gain important insights how either production platform can be further improved. In particular, it seems that L. lactis can be very useful for the production of protease‐sensitive high‐value proteins, such as vaccine components and biopharmaceuticals, while B. subtilis is currently better suited as an industrial work horse for the production of technical enzymes in bulk amounts. The objective of this overview is, thus, to compare both production platforms with focus on the respective strengths and weaknesses.
Signal peptides
Protein production in all living cells relies on the transcription of genes, and the subsequent ribosomal translation of mRNA into polypeptides. To specifically transport proteins from the ribosome in the cytosol to a translocation channel in the cytoplasmic membrane, the respective proteins are synthesized with N‐terminal signal peptides (SPs). The different known SPs are variable in length and show little amino acid sequence similarity (Dalbey et al. 2012). Nevertheless, they show structural conservation as they invariably consist of (i) a positively charged N‐terminal region with lysine or arginine residues and, incidentally, histidine residues, (ii) a central H‐region, consisting of mostly hydrophobic residues, with the potential to adopt an α‐helical conformation, and (iii) a hydrophilic C‐ region with a signal peptidase recognition site including the Ala‐x‐Ala consensus motif. The C‐domainhas a β‐stranded confirmation to allow recognition by signal peptidase and subsequent cleavage C‐ terminally of the Ala‐x‐Ala (Dalbey et al. 2012; Tjalsma et al. 2000).
Of note, a range of proteins synthesized with signal peptides for export from the cytoplasm is transiently or permanently retained by the cell after membrane translocation. These include lipid‐ modified proteins (in short lipoproteins), and proteins that are either covalently or non‐covalently bound to cell wall components like peptidoglycan or (lipo)teichoic acids (Tjalsma et al. 2000; Tjalsma et al. 2004; Dreisbach et al. 2011; Dreisbach et al. 2010). The membrane retention of lipoproteins is facilitated by a diacyl‐glyceryl modification. This modification is catalyzed by the prolipoprotein diacylglyceryl transferase (Lgt), which attaches the diacyl‐glyceryl moiety to an invariant Cys residue in the so‐called lipo‐box lipoprotein precursors (Leskela et al. 1999). Subsequently, the lipid‐modified precursor protein is processed by the lipoprotein‐specific signal peptidase Lsp (Tjalsma et al. 1999; Pragai et al. 1997). To attach translocated proteins covalently to the cell wall peptidoglycan, a C‐ terminal LPxTG recognition motif is recognized by a transpeptidase that is referred to as ‘sortase’, which hydrolyses the bond between the Thr and Gly residues and attaches the Thr residue to the peptidoglycan (Dramsi and Bierne 2017; Schneewind and Missiakas 2014; Siegel et al. 2017; Marraffini et al. 2006). In B. subtilis and L. lactis this process is catalyzed by the sortase SrtA (YhcS in B. subtilis) (Fasehee et al. 2011; Dieye et al. 2010). Proteins that are non‐covalently attached to the cell wall are characterized by one or more conserved motifs that enable specific binding to the cell wall components by hydrogen bonding, van der Waals interactions, or electrostatic interactions. Well‐characterized domains for non‐covalent cell wall binding are the LysM domains (Buist et al. 2008), SH3b domains (Mitkowski et al. 2019), the sporulation‐related SPOR‐domains (Yahashiri et al. 2017), and CHAP domains (Bateman and Rawlings 2003).
Table 1: Numbers of potentially exported proteins of B. subtilis 168 and L. lactis IL1403 and their predicted extracytoplasmic localization.
Localization Species Number of proteins % of total number of proteinsa
All B. subtilis 270 6.5 L. lactis 107 5 Extracellular B. subtilis 129 3 (Secreted) L. lactis 55 2.5 Cell wall B. subtilis 38 1 L. lactis 19 1 Membrane – cell wall interface (Lipoproteins) B. subtilis L. lactis 103 33 2.5 1.5 aThe % of the total number of extracytoplasmic proteins at particular subcellular locations was calculated based on the total number of proteins of B. subtilis 168 or L. lactis IL1403 present in their respective proteomes (UP000001570 for B. subtilis 168 and UP000002196 for L. lactis IL1403).
To obtain an overview of the different properties of the secretion machinery of B. subtilis and L. lactis, we revisited previously published genome data by a bioinformatics analysis, predicting all signal peptides as well as the subcellular location of the respective proteins using the decision tree represented in Figure 1. To this end, the proteomes of the type strains B. subtilis 168 and L. lactis IL1403, UP000001570 and UP000002196 respectively, were used because these are currently annotated best for the two species. Further, a pipeline of freely available prediction tools for signal peptides and subcellular protein localization was applied (Figure 1). Based on this combinatorial analysis, the localization of the various exported proteins was assessed, as detailed in Supplementary Tables 3 and 4, and summarized in Table 1. Relative to the total numbers of proteins per organism, B. subtilis 168 potentially produces more extracellular proteins than L. lactis IL1403. On the other hand, the relative numbers of predicted cell wall‐bound proteins in L. lactis and B. subtilis are comparable (Table 1). Figure 1: Decision tree for prediction of signal peptides and subcellular protein localization. To predict signal peptides (SPs) and subcellular protein localization in B. subtilis and L. lactis, a pipeline of freely available prediction tools for SPs and subcellular protein localization was applied. These included LipoP1.0a for detection of SPs of lipoproteins (http://www.cbs.dtu.dk/services/LipoP/) (Juncker et al. 2003), Signal P4.1 for prediction of secretory SPs (http://www.cbs.dtu.dk/services/SignalP/) (Petersen et al. 2011), CW PRED for the determination of LPxTG motifs of covalently cell wall bound proteins (http://bioinformatics.biol.uoa.gr/CW‐PRED/input.jsp). Analysis of cell wall binding domains in proteins exported from the cytoplasm was performed using CDD‐batch (https://www.ncbi. nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi) (Marchler‐Bauer et al. 2015) and TMHMM2.0c for transmembrane helixes (http://www.cbs.dtu.dk/services/TMHMM/) (Krogh et al. 2001). Although LipoP was originally trained for the prediction of lipoprotein signal peptides from Gram‐negative bacteria, it can also be applied for the prediction of such signal peptides from Gram‐positive bacteria (Rahman et al. 2007; Tjalsma and van Dijl 2005). Comparison of the lengths of identified SPs from both species showed that the average lengths of SPs from all secreted B. subtilis and L. lactis proteins differ considerably. As shown in Figure 2a, B. subtilis extracytoplasmic proteins have SPs with an average length of 24.7 amino acids (aa), and for extracytoplasmic proteins of L. lactis the average SP length is 27.7 aa. For proteins that are directed to
the cell wall, the average length of SPs in L. lactis is longer (33.8 aa) than for the equivalent proteins of B. subtilis (27.5 aa). L. lactis contains one extremely long SP of 57 aa, which belongs to the cell wall bound protein AcmA (Buist et al. 1995). Why this SP needs to be this long is presently not clear, but it is suggestive of an additional function of the signal peptide or usage of an alternative secretion mechanism (Berks 1996). Also, the SPs from lipo‐ and secreted proteins of L. lactis are longer compared to B. subtilis, in L. lactis IL1403 they have lengths of 21.4 aa and 29.4 aa respectively, and in B. subtilis
Figure 2: General properties of signal peptides in B. subtilis and L. lactis. The prediction of signal peptides and subcellular protein localization in B. subtilis 168 and L. lactis IL1403 was performed through a combinatorial analysis with the LipoP, Signal P4.1, TMHMM, CW‐PRED and CDD‐batch algorithms, using the respective translated open reading frames (light grey, B. subtilis 168; dark grey, L. lactis IL1403). Error bars represent the Standard Error of the Mean. A distinction was made between signal peptides of all collectively analyzed exported proteins (All), or signal peptides of secreted (Secreted), cell wall‐bound (Cell wall) and lipid‐modified proteins (Lipo). (a) Signal peptide lengths, (b) pI distribution of SPs, and (c) number of positively charged amino acids in the signal peptide. a 0 5 10 15 20 25 30 35 40
All Secreted Cell wall Lipo
Len gth of SP (# am ino ac ids ) c 0 1 2 3 4 5 6
All Secreted Cell wall Lipo
# pos it iv e ly ch ar ge d re si d u e s 0 10 20 30 40 50 60 70 80 90 100
B. subtilis L. lactis B. subtilis L. lactis B. subtilis L. lactis
Secreted Cell wall Lipo
12.50‐13.49 11.50‐12.49 10.50‐11.49 9.50‐10.49 8.50‐9.49 7.50‐8.49 6.50‐7.49 5.50‐6.49 4.50‐5.49 3.50‐4.49
b
pI168 this is 19.8 aa and 27.9 aa, respectively. Of note, these differences are less pronounced than obtained for the cell wall bound proteins.
The isoelectric point (pI) of most SPs of L. lactis and B. subtilis lies in the range of 9.5‐10.5 and for SPs from lipoproteins and cell wall‐bound proteins in L. lactis this is also the dominant pI range. Notably, there is a higher variation in the pI of SPs from B. subtilis than observed for SPs from L. lactis (Figure
2b). Related to this, we determined the number of positively charged aa in the complete SP (i.e.
arginine, lysine and histidine), where it should be noted that the charge of histidine depends on the pH. This analysis showed that SPs from extracellular proteins of L. lactis contain more positively charged aa than the SPs of such proteins in B. subtilis (Figure 2c). To analyze this in more detail, we assessed the positive charge relative to the SP length, as presented in Figure 3. This showed that SPs of secreted and cell wall‐bound proteins contain relatively more positively charged amino acids in L. lactis than in B. subtilis. However, for SPs from lipoproteins this is the other way round with relatively more positively charged aa in SPs of B. subtilis lipoproteins. This appears mainly due to the bias introduced by the B. subtilis lipoprotein YdhK SP of 38 aa, which contains no less than 9 positively charged aa. The YdhK protein is involved in survival of oxidative stress conditions (Reder et al. 2012). Figure 3: Numbers of positively charged residues per SP length. The numbers of positively charged aa in the SPs of (a) secreted proteins, (b) cell wall‐bound proteins, (c) lipoproteins of B. subtilis (‐) and L. lactis (x) are plotted against the length of the SP. The trendline through the data points is represented by straight lines for B. subtilis and dotted lines for L. lactis. c a b 0 2 4 6 8 10 12 14 16 0 10 20 30 40 50 60 # pos it iv e ly ch ar ge d r e si due s SP length (#AA) 0 2 4 6 8 10 12 14 0 10 20 30 40 50 60 # pos it iv e ly ch ar ge d r e si due s SP length (# AA) 0 2 4 6 8 10 0 10 20 30 40 # pos it iv e ly ch ar ge d r e si due s SP length (#AA)
To assess the general capabilities of the identified signal peptides, we compared the mature cargo proteins of L. lactis and B. subtilis. This showed that the mature proteins exported with the help of the identified signal peptides are, overall, comparable in size (Supplementary Figure 1). Nevertheless, the cell wall‐bound proteins of L. lactis are on average larger (59.5 kDa) than the respective proteins from B. subtilis (48.8 kDa), while their pI is lower (5.5 vs 7.8, respectively; Supplementary Figure 1, a and b). The majority of cell wall‐bound proteins of L. lactis has a pI between, 4.5‐5.5, while in B. subtilis the majority of proteins belonging to this group has a pI between 8.5‐9.5 (Supplementary Fig. 1c). The average pI of the secreted proteins appears to be the same in both bacteria (both 6.6), which is also reflected in the comparable pI distribution. In contrast, the lipoproteins of L. lactis have a significantly higher pI (7.2) than lipoproteins of B. subtilis (6). As described by (Tjalsma et al. 2004), secretory SPs of B. subtilis contain a cleavage site for type I signal peptidases (SipS‐SipW) with a consensus Ala‐x‐Ala motif at the ‐3 to ‐1 positions relative to the cleavage site. Lipoproteins contain a consensus Leu‐x‐x‐Cys motif for recognition by type II signal peptidase (LspA), where cleavage takes place N‐terminally of the strictly conserved Cys residue (Tjalsma et al. 1999). To determine if this is also true for L. lactis IL1403, the cleavage sites of the different SPs were assessed using the web‐based sequence logo generator Seq2Logo (http://www.cbs.dtu.dk/biotools/Seq2Logo/) (Thomsen and Nielsen 2012). As is depicted in Figure 4, most secreted and cell wall‐located proteins of B. subtilis contain the Ala‐x‐Ala consensus motif, whereas proteins of L. lactis IL1403 with the equivalent localization contain Val or Ala residues at the ‐3 position. B. subtilis prefers an Ala residue at the +1 position after the cleavage site, and at +2 it prefers a negatively charged residue with Glu most abundantly represented in secretory proteins and Asp in cell wall proteins. In contrast, L. lactis has a preference for Ala at +1 in secretory proteins and Asp in cell wall‐bound proteins, whereas at +2 a Thr is preferred. For lipoproteins the differences in cleavage site motifs are less pronounced with a slight preference for Gly at the ‐1 position in B. subtilis 168. While many different signal peptides are applied for protein production in B. subtilis (Supplementary
Table 1), the preferred signal peptide for protein production in L. lactis is derived from the major
secreted lactococcal protein Usp45. The latter signal peptide has a length of only 27 aa (van Asseldonk et al. 1990) (Supplementary Table 2). Interestingly, while the native Usp45 protein is abundantly secreted by L. lactis, its SP contains a cleavage site that is more similar to the ones encountered in cell wall‐bound proteins (VYA_DT), which is consistent with the presence of a CHAP domain for cell wall binding. Interestingly the Usp45 signal peptide was optimized for heterologous protein expression and secretion by random mutations of the H‐region (Ng and Sarkar 2013), but our present bioinformatics analysis suggests that there may be additional benefit in modifying the C‐region as well.
Figure 4: Signal peptidase cleavage sites. To assess differences in the preferred SP cleavage sites for signal peptidase I (a and b) and signal peptidase II (c) in B. subtilis 168 and L. lactis IL1403 a Seq2logo analysis was performed (http://www.cbs.dtu.dk/biotools/Seq2Logo/). Specifically, this involved the signal peptidase cleavage sites (dotted line) in (a) secreted, (b) cell wall‐bound and (c) lipoproteins based on aa residues covering the predicted ‐4 to +4 positions. The analysis was performed using the Hobohm1 Clustering method with a threshold of 1, resulting in a P‐weigthed Lullback‐Leibler logo of the respective regions.
The Sec pathway
In bacteria, the most frequently used secretion pathway is the Sec‐dependent protein translocation pathway. This pathway translocates proteins across the cytoplasmic membrane in an unfolded state. It has been estimated that ~96% of the proteins exported from the E. coli cytoplasm follow the Sec pathway (Tsirigotaki et al. 2017). The latter proteins are predominantly destined for the outer surface of the inner membrane, the periplasm and the outer membrane. For B. subtilis it has been estimated that ~97% of the proteins with a signal peptide are exported via the Sec pathway (Tjalsma et al. 2000; Tjalsma et al. 2004). As discussed in the above section on signal peptides, the translocated proteins of B. subtilis may end up at the outer surface of the inner membrane, in the cell wall or in the extracellular milieu. The Bacillus cell wall is a porous network, consisting of peptidoglycan, wall‐ and lipo‐teichoic acids, teichuronic acids and polysaccharides. Although B. subtilis seems to contain the equivalent of a periplasmic space between the membrane and cell wall (Matias and Beveridge 2008; Matias and Beveridge 2005), the majority of translocated proteins will end up in the extracellular milieu. Overall, the cell wall of L. lactis is composed of the same macromolecular components as that of B. subtilis (Chapot‐Chartier and Kulakauskas 2014), but it was recently discovered that L. lactis IL1403 also contains a linear backbone of repeated α‐l‐Rhamnose disaccharide subunits (Vinogradov et al. 2018). D‐Alanylation of the lipo‐teichoic acids in the cell wall of L. lactis results in the entrapment of secreted staphylococcal nuclease at the cell surface meaning that the composition of the cell wall is of importance for optimal protein secretion (Nouaille et al. 2004).Components of the Sec‐dependent secretion machinery can be divided into seven groups: (i) cytosolic chaperones, (ii) the translocase consisting of the translocation motor (SecA) and components of the translocation channel (SecYEG, SecDF‐YajC, SpoIIIJ), (iii) signal peptidases, (v) signal peptide peptidases, (vi) folding factors that function at the trans‐side of the membrane, and (vii) proteases that have a function in protein quality control. The main components of the secretion machinery of B. subtilis have been investigated extensively, and database similarity searches show that many but not all of these components are conserved in L. lactis. The similarities and differences in the secretion machinery of both organisms are schematically represented in Figure 5. The Sec and Tat pathways for protein secretion in B. subtilis are best understood (Tjalsma et al. 2000; Tjalsma et al. 2004; Van Dijl and Hecker 2013; Hohmann et al. 2016; Song et al. 2015). In contrast to the Sec pathway, the Tat pathway secretes fully folded proteins. While B. subtilis contains two distinct Tat translocases with different substrate specificities, the Tat pathway is absent from L. lactis (Goosens et al. 2014; Goosens and van Dijl 2017). For this reason and also because the B. subtilis Tat pathway is not yet readily accessible for applications in protein production, the following sections of this review are focused on the Sec pathway only.
Figure 5: Comparison of the main components of the Sec‐dependent protein translocation machinery of B. subtilis and L. lactis. Overview of components of the Sec‐dependent protein export machinery of (a) B. subtilis and (b) L. lactis. (i) Upon ribosomal (R) translation, exported membrane and secretory precursor proteins will be specifically targeted to the Sec pathway by means of their signal peptide (SP). In addition, these proteins will be recognized by cytoplasmic chaperones and targeting factors, including SRP/FtsY, CsaA and SecA, which may keep the exported proteins in a translocation‐competent state. (ii) The Sec translocase consists of the ATP‐dependent SecA motor protein and channel components SecY, SecE and SecG. The Sec translocase of B. subtilis also includes the proton‐motive force‐dependent translocation motor SecDF‐YajC. Particular membrane proteins require the
homologous SpoIIIJ/YqjG or YidC insertases for membrane biogenesis. (iii) Retention of particular proteins in the membrane or cell wall may be achieved by Lgt‐mediated diacyl‐glyceryl modification of the lipo‐box in lipoproteins, or by sortase (YhcS/SrtA)‐mediated transpeptidation of the LPxTG moiety to the cell wall peptidoglycan. (iv) Shortly after or during the translocation process, the precursor protein is cleaved by one of the type I signal peptidases (SPase; SipS‐V, SipL or SipW), or the type II signal peptidase LspA. (v) TepA, SppA, and RasP have been implicated in the degradation of cleft signal peptides. (vi) The folding of various exported proteins is dependent on the action of the folding catalysts PrsA, PmpA, PrtM, or BdbB‐D. (vii) HtrA, HtrB, HtrC, WprA and PrsW are implicated in quality control of the membrane and secretory proteins. Of note, HtrA and HtrB of B. subtilis have a dual localization in the membrane and growth medium. Extracellular proteases AprE, Bpr, Epr, Mpr, NprB, NprE and Vpr of B. subtilis can have roles in the degradation of secreted proteins. HtrA and PrtP are the major proteases of L. lactis. Intracellular proteases ClpC‐ClpP, ClpE‐ClpP, ClpX‐ClpP, AprX, IspA and TepA may be involved in the degradation of mis‐targeted exported proteins. Secretion machinery components that have been engineered for improved protein secretion are marked in bold. Of note, staphylococcal DsbA in B. subtilis and B. subtilis SecDF in L. lactis were heterologously expressed in B. subtilis and L. lactis, respectively (dotted line). To target proteins specifically into the Sec‐pathway, the N‐terminal signal peptide of a (heterologous) protein has to be recognized by the secretion machinery. In the case of secreted proteins, the precise nature of soluble cytoplasmic factors that facilitate membrane targeting is still enigmatic as a dedicated factor, such as SecB of E. coli, is apparently absent (Tjalsma et al. 2000). It was therefore proposed that signal peptide recognition may involve the highly conserved signal recognition particle (SRP), consisting of the small cytoplasmic RNA (scRNA), the Ffh protein and the histon‐like protein HBsu (Nakamura et al. 1999; Bunai et al. 1996). Assisted by a homologue of the cytoplasmic SRP receptor, in bacteria known as FtsY, the SRP could in principle keep the precursor protein in a translocation‐ competent state (Zanen et al. 2006). However, the roles of SRP and FtsY in protein secretion is yet to be demonstrated unambiguously. This has been difficult, because both SRP and FtsY are, in general, of major importance for membrane protein biogenesis and, consequently, essential for growth and cell viability (Kuhn et al. 2017). Thus, the observed effects of SRP or FtsY depletion on protein secretion may be indirectly caused by impaired membrane protein biogenesis leading to membrane perturbations (Zanen et al. 2006).
Another factor that has been implicated in dedicated membrane targeting of secretory protein precursors to the translocase in the membrane is the chaperone CsaA of B. subtilis. Similar to SRP and FtsY, CsaA is essential for growth and cell viability, which made it difficult to distinguish direct and indirect effects following its depletion. Nonetheless, it has been shown that CsaA interacts with the SecA proteins of B. subtilis and E. coli (Muller, Ozegowski et al. 2000), and is induced in response to severe secretion stress (Hyyrylainen et al. 2005). In addition, it was shown that CsaA has binding affinity for the mature part of the secretory protein SdpC (previously known as YvaY) in B. subtilis, but not for the signal peptide of this protein (Linde et al. 2003; Muller, Bron et al. 2000). Of note, CsaA is not conserved in L. lactis. Other chaperones with a function in protein folding, that have been implicated in protein secretion by B. subtilis are the heat‐shock proteins DnaK (Hsp70) and GroEL (Hsp60)
the proteins to fold properly in an isolated environment (Moliere and Turgay 2009). Most likely, the effects of dnaK or groEL mutations on protein secretion relate to the general and universally conserved chaperone functions of these proteins. Lastly, the SecA protein of B. subtilis has been implicated in targeting secretory precursor proteins to the membrane‐embedded translocase complex. This idea is based on the observation that B. subtilis contains a pool of soluble SecA, the fact that SecA has affinity for signal peptides and secretory precursors, and the well‐established function of SecA as the translocation motor in the Sec machinery (Bauer et al. 2014; Gouridis et al. 2009; Xie and Dalbey 2018). The Sec machinery in the membrane consists of the afore‐mentioned translocation ATPase SecA (Chatzi et al. 2014), the SecYEG protein translocation channel (Zimmer et al. 2008), and the SecDF‐YajC complex (Bolhuis et al. 1998). Best‐described is the role of SecA and SecYEG in preprotein translocation across the membrane, where SecA drives the translocation of the precursor through the SecYEG channel by cycles of ATP binding and hydrolysis (Tsirigotaki et al. 2017; Prabudiansyah and Driessen 2017; Kedrov et al. 2013). The vast majority of the studies on SecA‐SecYEG function has been performed in E. coli, but based on the high similarity of the respective components, the translocation mechanism in B. subtilis and L. lactis is most likely well‐conserved. Importantly, SecA, SecY and SecE are essential for protein translocation, cell growth and cell viability, whereas the non‐essential SecG is merely needed for efficient protein transport. The importance of the non‐essential secretion machinery component SecDF‐YajC in protein translocation has been known for a long time, but its function was only recently identified through structural analyses. The available data imply that the SecDF‐YajC complex functions as a proton‐driven secretion motor that allows pre‐protein translocation to be energized by the proton‐motive force (Furukawa et al. 2017). Of note, SecDF is absent from L. lactis, whereas a homologue of YajC (WP_003130588) is present in this bacterium. The L. lactis YajC homologue (110 residues) is somewhat larger than its homologue in B. subtilis (i.e. YrbF, 89 residues). Two additional membrane proteins that associate with the Sec channel are SpoIIIJ and YqjG of B. subtilis. These proteins are primarily involved in membrane protein biogenesis (Hennon et al. 2015). Deletion of spoIIIJ combined with a depletion of yqjG led to reduced levels of protein secretion, but this effect may be indirectly due to mis‐assembly of membrane proteins (Tjalsma et al. 2003; Saller et al. 2009). Intriguingly, SpoIIIJ of B. subtilis seems also involved in the assembly of a pathway for intercellular signaling between the mother cell and the fore spore during early stages of sporulation (Camp and Losick 2008). A homologue of SpoIIIJ and YqjG exists in L. lactis (YidC) but its possible involvement in protein secretion has not been investigated.
As mentioned above, the signal peptide of the exported protein is removed by a signal peptidase, which takes place during or shortly after translocation (Tjalsma et al. 1997; van Roosmalen et al. 2004; Tjalsma et al. 1998). This processing step is necessary to liberate the translocated protein from the membrane, because the signal peptide would otherwise serve as a membrane anchor (Dalbey et al.
2012). Intriguingly, B. subtilis contains five chromosomally‐encoded signal peptidases, known as SipSTUVW (Tjalsma et al. 1998). Probably this functional redundancy reflects the lifestyle of B. subtilis, where high‐level secretion of proteins is needed to retain a competitive advantage over other soil‐ dwelling microorganisms. In contrast, L. lactis has only one signal peptidase‐encoding gene, which is apparently sufficient to sustain growth on milk and the media used to culture this bacterium in the laboratory. The cleft signal peptides are, subsequently, degraded by signal peptide peptidases. Three proteins have been implicated in signal peptide degradation, namely the cytoplasmic protein TepA, and the membrane proteases RasP and SppA (Zanen et al. 2006; Bolhuis, Tjalsma et al. 1999; Bolhuis, Matzen et al. 1999; Heinrich et al. 2008; Neef et al. 2017). Mutation of the respective genes impact to different extents on protein secretion, which would be consistent with the expected phenotype of reduced signal peptide turnover. This relates to previous findings showing that synthetic signal peptides can inhibit the function of SecA and signal peptidase (Cunningham and Wickner 1989; Wickner et al. 1987). In addition, accumulating signal peptides will probably perturb the integrity of the membrane. By far the strongest secretion phenotype was observed for a rasP mutation, which resulted in impaired secretion of the α‐amylases AmyE and AmyL, as well as the serine protease BPN’ (Heinrich et al. 2008; Neef et al. 2017). Importantly, overexpression of rasP facilitated the enhanced production of two secretory proteins that are otherwise hard to produce in B. subtilis, namely a serine protease from Bacillus clausii and the α‐amylase AmyAc from Paenibacillus curdlanolyticus. Likewise, it was shown in B. licheniformis that overexpression of sppA can enhance the production of an amylase and nattokinase (Cai et al. 2017). Thus, it seems that signal peptide peptidase activity can set a limit to protein secretion. In this respect, it is noteworthy that RasP cleaves its substrates within the plane of the membrane, while SppA cleaves substrates at the extracytoplasmic side of the membrane (Dalbey et al. 2012). While the effects of rasP deletion or overexpression are clear, the precise mechanism by which RasP impacts on protein secretion remains to be elucidated. This relates to the fact that RasP is known to be required for the activation of the SigW stress response by cleavage of the anti‐sigma factor RsiW, while another RasP substrate (FtsL) is involved in cell division (Heinrich et al. 2008; Bramkamp et al. 2006). Lastly, depending on the growth medium that was used, the secretion‐ enhancing effect of RasP overproduction was correlated with improved growth of B. subtilis (Neef et al. 2017). This suggests that, irrespective of RasP’s mode of action, the overproduction of this membrane protease enhances the fitness of cells that overproduce secretory proteins. Whether this is also the case in other cell factories like L. lactis remains to be investigated.
Upon passage of the Sec channel in an unfolded state, the translocated protein needs to fold into its stable and active conformation. This is known to require the lipoprotein PrsA, which has both peptidyl‐ prolyl cis/trans isomerase and chaperone activities (Kontinen and Sarvas 1993; Vitikainen et al. 2005;
of the dnaK operon with prsA facilitated the overproduction of the α‐amylases AmyL and AmyS. In fact, this resulted in a 9‐ and 12‐fold increased activity of AmyL and AmyS, respectively, while the productivity for AmyL and AmyS increased by 13‐ and 17‐fold (Chen et al. 2015). In L. lactis two PrsA orthologues are present, which are known as PmpA (protein maturation protein A) (Drouault et al. 2002) and PrtM (Venema et al. 2003). Although PmpA does not contain the PPIase domain, it still has a foldase activity, as was shown by improved production of a lipase from Staphylococcus hyicus upon pmpA overexpression (Drouault et al. 2002). In particular, degradation of the lipase was substantially reduced when pmpA was overexpressed. The PrtM protein is needed for maturation and activation of the major protease PrtP, both being encoded by the same plasmid. In the absence of PrtM, PrtP is exported from the L. lactis cells in an inactive state with the pro‐peptide still present. This also shows that PmpA and PrtM have distinct substrate specificities, which explains why PmpA cannot compensate for the absence of PrtM in PrtP activation (Venema et al. 2003).
In B. subtilis three different membrane‐bound thiol‐disulfide oxidoreductases (TDORs) have been implicated in the formation of disulfide bonds in translocated proteins. These are the ‘Bacillus disulfide bond’ (Bdb) proteins BdbB, BdbC and BdbD (Kouwen and van Dijl 2009b; Kouwen and van Dijl 2009a). A fourth TDOR, known as BdbA, is expressed but no function in disulfide bond formation has been demonstrated as yet. The bdbA and bdbB genes are located on the SPβ prophage, while bdbC and bdbD are located on the core genome of B. subtilis (Kouwen et al. 2007). The major oxidative TDOR complex in B. subtilis consists of BdbC and BdbD, the main substrates being the ComEC and ComGC pseudopilins that are needed for the binding and uptake of DNA during genetic competence (Meima et al. 2002; Draskovic and Dubnau 2005). Accordingly, a bdbCD mutant shows strongly reduced competence. Furthermore, it was shown by mass spectrometry that the abundance of 18 membrane‐associated proteins was altered in a bdbCD double mutant compared to the parental strain 168. One of the missing proteins was ProA, which is involved in osmoprotection. Consistent with this observation, it was found that the BdbCD‐deficient strain was more sensitive to osmotic shock (Goosens et al. 2013). Furthermore, it was shown that BdbB and BdbC cooperate in the oxidative folding of the secreted SPβ‐ encoded bacteriocin sublancin 168 (Dorenbos et al. 2002). Lastly, an important finding for heterologous protein production was that both BdbC and BdbD are needed for secretion of the alkaline phosphatase PhoA of E. coli in an active and protease‐resistant state (Kouwen et al. 2007). The secretion of this PhoA protein could be further enhanced by heterologous co‐expression of staphylococcal DsbA proteins and a reduction of the cytoplasmic levels of the thioredoxin A (Kouwen et al. 2008). It thus seems that secretion of disulfide‐bonded proteins can be enhanced by increasing the overall oxidative power of the B. subtilis cell factory. Although L. lactis IL1403 does not contain homologous TDORs, it is able to produce and secrete disulfide bond‐containing heterologous proteins
(Singh et al. 2018). This raises the question whether disulfide bond formation in proteins secreted by L. lactis is a spontaneous process, or whether it is catalyzed by as yet unidentified TDORs.
Quality control
Despite the presence of extracytoplasmic folding catalysts and chaperones, the folding of translocated proteins may not be 100% efficient. If this is the case, the resulting malfolded proteins may either be degraded by quality control proteases, such as HtrA, HtrB and WprA, or they may accumulate in the cell wall environment. The latter was clearly shown for the α‐amylase AmyQ in PrsA‐depleted cells, but there is also evidence from pulse‐chase labeling studies that substantial amounts of α‐amylase are degraded upon overexpression in wild‐type B. subtilis (Hyyrylainen et al. 2001; Stephenson et al. 2002). This accumulation of misfolded and/or mislocalized proteins may be detrimental for the bacterial cell and has, therefore, been termed secretion stress. Like many other cell envelope stresses, also secretion stress is sensed by a dedicated two‐component regulatory system which was named CssRS (Hyyrylainen et al. 2001). The CssRS‐specific response is not only triggered by secretory protein overproduction but also by heat‐stress, suggesting that the system senses the accumulation of unfolded proteins either directly or indirectly (Darmon et al. 2002). To counteract the stress, CssRS activates the synthesis of HtrA and HtrB, which have both serine protease and chaperone functions (Hyyrylainen et al. 2001; Lulko et al. 2007; Antelmann et al. 2003). The sensor kinase CssS, mainly located at cell septa and the cell poles (Noone et al. 2012), senses the secretion stress and subsequently activates the response regulator CssR. The response to this activation results in autoregulation of CssRS, causing increased expression levels of this kinase and response regulator together with HtrA and HtrB. This may either lead to refolding or degradation of the protein that triggered the CssRS response. N‐terminally cleaved forms of HtrA and HtrB are also detectable in the growth medium (Antelmann et al. 2003), which seems to involve the action of the intramembrane protease RasP (Zweers et al. 2009; Dalbey et al. 2012). Strikingly, when the htrA gene is deleted, the expression of htrB is strongly increased in a CssRS‐dependent manner, and vice versa (Krishnappa et al. 2014; Noone et al. 2001). This suggests that a basal level of HtrA and HtrB is needed to avoid secretion stress caused by native non‐overexpressed exported proteins, a view that is supported by the fact that it is difficult to delete both htrA and htrB from the genome of wild‐type B. subtilis. Of note, the htrC gene, encoding a paralogue of HtrA and HtrB is not part of the CssRS regulon. Yet, it seems that HtrC can compensate to some extent for the absence of HtrA and HtrB, because the htrC gene was found to be upregulated in a 10‐fold protease‐negative mutant that lacks also the htrA and htrB genes (Pohl et al. 2013). The exact mechanism of this htrC upregulation is currently unclear but it seems to be triggered upon severely reduced extracytoplasmic proteolysis.Interestingly, in L. lactis the major chromosomally‐encoded extracytoplasmic protease is HtrA, while paralogous HtrB or HtrC proteases are absent. HtrA is responsible for the degradation of AcmA, the major autolysin of L. lactis, as well as a range of other cell surface proteins (Steen et al. 2005; Guillot et al. 2016). The degradation by HtrA was shown to be reduced upon D‐Alanine depletion of Lipoteichoic acids of the cell wall (Steen et al. 2005). A recent transcript profiling analysis of S. aureus gene expression under different stress conditions showed induction of htrA upon stress imposed by cell wall‐active antibiotics (Utaida et al. 2003). It is however not known whether or how the expression of htrA upon exposure to this type of stress is regulated in L. lactis, and whether this impacts on extracellular protein production. In any case, a BlastP search with the CssR or CssS sequences of B. subtilis did not result in the identification of specific homologs of these proteins in L. lactis. A highly active protease in the cell wall environment of B. subtilis is the serine protease WprA (wall protease A). Inactivation of WprA can increase the yield of industrially relevant secreted proteins, like α‐amylases (Stephenson and Harwood 1998; Stephenson et al. 2002). Recently it was discovered that WprA also has a major role in the degradation of cell envelope proteins of B. subtilis, in particular PrsA, which suggests that WprA may be active in close proximity to the Sec channel (Krishnappa et al. 2014). This view is supported by earlier findings, showing that WprA is responsible for the degradation of an unstable mutant form of the signal peptidase SipS (Bolhuis, Tjalsma, Stephenson et al. 1999). In addition, WprA is needed for functional protein translocation via the Tat translocase, providing further evidence for its activity in close proximity to membrane‐embedded preprotein translocation systems (Monteferrante et al. 2013). Of note, the effect of a wprA deletion on the degradation of PrsA was only observed in mutants lacking genes for additional secreted proteases, suggesting that also some of the other secreted proteases of B. subtilis are active in the membrane‐cell wall environment. Interestingly, a single wprA mutation also had a negative effect on the cellular and secreted levels of HtrA and HtrB, which was restored when WprA was expressed from an inducible plasmid. The latter observation suggests that the expression of HtrA, HtrB and WprA is inter‐related, but the exact mechanism is currently not known. A possible explanation is that the protease WprA also has a chaperone function, which is needed to stabilize HtrA and HtrB when other extracytoplasmic proteases are expressed (Krishnappa et al. 2014). Like HtrB and HtrC, the WprA protease is not conserved in L. lactis, which is probably one of the reasons why heterologously expressed proteins are fairly stable in this organism. In addition to WprA, B. subtilis produces at least seven other secreted proteases, named AprE, Bpr, Epr, Mpr, NprB, NpreE and Vpr. Together, these proteases impact strongly on the productivity of heterologous proteins, which was shown by Pohl et al. by sequentially constructing a set of multiple protease mutants (Pohl et al. 2013). Of note, WprA and the minor extracellular protease Epr of B. subtilis contribute to the prevention of autolysis in the stationary growth phase by controlling the cellular level of the autolysins LytE and LytF (Yamamoto et al. 2003). This is probably one of the reasons
