Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Bacillus SEVA siblings: A Golden Gate-based toolbox to create personalized integrative vectors for Bacillus subtilis

A Publisher Correction to this article was published on 06 March 2018

A Publisher Correction to this article was published on 17 January 2018

This article has been updated


Bacillus subtilis combines natural competence for genetic transformation with highly efficient homologous recombination. These features allow using vectors that integrate into the genome via double homologous recombination. So far, their utilization is restricted by the fixed combination of resistance markers and integration loci, as well as species- or strain-specific regions of homology. To overcome these limitations, we developed a toolbox for the creation of personalized Bacillus vectors in a standardized manner with a focus on fast and easy adaptation of the sequences specifying the integration loci. We based our vector toolkit on the Standard European Vector Architecture (SEVA) to allow the usage of their vector parts. The Bacillus SEVA siblings are assembled via efficient one-pot Golden Gate reactions from four entry parts with the choice of four different enzymes. The toolbox contains seven Bacillus resistance markers, two Escherichia coli origins of replication, and a free choice of integration loci. Vectors can be customized with a cargo, before or after vector assembly, and could be used in different B. subtilis strains and potentially beyond. Our adaptation of the SEVA-standard provides a powerful and standardized toolkit for the convenient creation of personalized Bacillus vectors.


The diversity of plasmid vectors in molecular cloning

In molecular cloning, vectors are vehicles to transfer foreign nucleic acids into a living cell. In the context of this article, we restrict the term “vector” to “plasmid vectors”, small circular DNA molecules that originate from bacteria. They are easy to handle for inserts up to 10–15 kb, replicate independently of the bacterial chromosome and can be isolated in large amounts through standard plasmid preparation procedures. Over the last decades, vectors were increasingly modified to meet custom needs. For instance, they may contain promoters for ready-to-use gene expression, or reporter genes to measure transcription or translation rates, respectively. More elaborated vector types can contain biosafety features, internal measuring standards or may be used for clean chromosomal gene deletions or dual expression systems1,2,3,4. Synthetic biology strives to implement and extend the principles of engineering (standardization, decoupling, abstraction) into biology5. Especially standardization in vector and plasmid construction offers the clear advantage of comparability, compatibility, flexibility and reusability of single parts and whole vectors, as exemplified by the Standard European Vector Architecture (SEVA).

The Standard European Vector Architecture provides standardized vectors for Gram-negative bacteria

In 2013, the group of Victor de Lorenzo developed a standardized vector toolbox for the use in Gram-negative bacteria, with a special interest in Pseudomonas putida. With SEVA, they set the stage for a community-driven development platform and for evolving a standardized vector collection. This platform facilitates finding, creating, and naming of suitable vectors as well as their downstream handling6,7. Currently, the SEVA database lists 135 SEVA vectors and 49 SEVA siblings8 (, March 2017). Subsequently, linker sequences were developed to make the vectors compatible to different cloning methods9.

Each SEVA vector contains at least three functional elements (Fig. 1a), (i) the origin of replication (ori) for vector replication in the host cell, (ii) a selectable marker, e.g. an antibiotic resistance cassette to select for vector uptake and maintenance, and (iii) a multiple cloning site (MCS) for insertion of the DNA of interest. In the SEVA standard, the latter part is called cargo, irrespective of the size and nature of the insert. The cargo is isolated by flanking double transcriptional terminators (T1, T0), to avoid unwanted transcriptional read-through into other elements of the backbone. All parts are flanked by defined rare endonuclease restriction sites and assemble in a fixed order and orientation (Fig. 1a). They must not contain these and further restriction endonuclease recognition sites. Additionally, the transcriptional terminators and the origin of transfer (oriT, for plasmid conjugation) are predefined, whereas the selectable marker, ori and cargo can be chosen freely, as long as they adhere to the standard. An easy number-based nomenclature assures the fast determination of vector features from the vector’s name6,7.

Figure 1

Configuration of a SEVA vector and its Bacillus SEVA sibling. (a) The basic SEVA vector is composed of six parts, separated by defined endonuclease restriction sites. The transcriptional terminators T0 and T1 as well as the origin of transfer oriT are fixed, whereas the cargo, the ori and the antibiotic marker can be chosen freely from a pool of SEVA-compatible parts. (b) For genomic integration, three parts were added to the SEVA layout to create the Bacillus SEVA sibling pBS: Flanking homology regions up and down, as well as an antibiotic marker for Bacillus. Vector verification and propagation occurs in E. coli and only the part in between the homology regions (dashed line) will integrate into the genome. Vectors are drawn not to scale. Functional transcriptional units are indicated with an arrow (promoter) and black bar (terminator).

Distinct features of vectors for Bacillus subtilis

Bacillus subtilis is the best studied low-G + C Gram-positive bacterium (phylum Firmicutes) and one of the leading workhorses of the biotechnological industry10,11,12. Its ease of genetic manipulation is based on its natural competence, which includes the active uptake of (any) DNA and recombination of homologous regions into the chromosome13. These features enable to routinely use integrative instead of replicative vectors. This ensures genetic stability even without maintaining a constant selective pressure. Moreover, copy number effects are avoided, which can otherwise influence the promoter activity on replicative vectors, particularly at the single cell level14,15.

For cloning convenience, vectors designed for integration into the B. subtilis genome usually replicate in E. coli but do not contain an ori for B. subtilis. Instead, two homology regions of at least 400 bp, which define the insertion locus, flank the cargo and the resistance marker for B. subtilis. The plasmids are linearized in the E. coli part of the vector before B. subtilis transformation to avoid integration via single crossing over events. Consequently, only the DNA segment that is located between the homology regions will integrate into the chromosome, resulting in genetic stability of single copy number inserts even in the absence of selective pressure.

The currently available vectors combine known and effective selection cassettes with well-characterized integration loci16,17,18,19,20. This status quo limits the combination of resistance markers and integration sites, and does not allow targeting entirely new chromosomal regions. Moreover, existing vectors cannot be used in another Bacillus strain, which differs in the nucleotide sequence at the specified integration site.

Due to the genetic accessibility of B. subtilis, PCR-based methods, such as long-flanking homology (LFH)-PCR21, can be used to target new or strain-specific loci. While this approach is very convenient for generating knock-out mutants, it is restricted with respect to the cargo: fusing several PCR products, and obtaining large fragments, e.g. the luxABCDE bioluminescence reporter operon (5.7 kb) that allows online measurement17,19, with sufficient yield for efficient transformation (>1 µg22) can be tedious. In these cases, transformants have to be sequenced each time to ensure the integrity of the cargo (e.g. lack of mutations). In order to combine the advantages of vectors (stability and reusability) with those of PCR-based methods (flexibility) for the use in B. subtilis and related species, we designed, constructed and tested a new vector concept: the Bacillus SEVA siblings.

Concept of Bacillus SEVA siblings as customized vectors

Here, we describe a toolkit that was developed for the fast and easy generation of genomic integration vectors for B. subtilis, thereby overcoming the traditional limitations described above. Each vector contains a resistance marker, cargo and flanking homology regions of choice. To ensure compatibility and reusability of already existing parts for replication in E. coli, we based our system on SEVA and added flanking homology regions as well as Bacillus resistance markers at defined positions (Fig. 1b). Our vectors will be named Bacillus SEVA siblings, which is in line with the current SEVA regulations6.

Our toolbox offers seven functional antibiotic resistance cassettes for selection in B. subtilis and high and medium copy number E. coli vectors modified to be assembled with your homology regions of interest in a one-pot Golden Gate assembly23,24. We tested the assembly efficiencies for five different enzymes and the B. subtilis transformability of every part provided. As proof of concept for efficient assembly and functionality, we analyzed the expression levels of the red fluorescing protein mKate2 under control of the xylose-inducible promoter PxylA at different chromosomal insertion loci.


Vector layout for double cross-over homologous recombination in B. subtilis

We designed and constructed our vector building toolkit with the main goal to easily exchange the target loci for double homologous integration into the genome. This is important for two major reasons: Current vector collections do not allow the free combination of integration loci and selectable markers. Moreover, they omit the use of Bacillus strains if they differ from the reference lab strains in their nucleotide sequence, so that the integration sites of standard vectors are not compatible. For our collection of customizable vectors, we focus on allelic replacement via double recombination, where the DNA-sequence to be integrated into the chromosome (here termed integration part) is flanked by two regions of homology (Fig. 1b). For B. subtilis, approximately 400 bp of identical sequence are required for efficient integration. Shorter sequences (70 bp) can sometimes be sufficient, but are not recommended due to the significantly reduced efficiency25. For ease of construction, the integration part is combined with an ori and selectable marker for E. coli (here termed replication part). Before B. subtilis transformation, the vector is linearized in the replication part to avoid single cross-over events and ensure double cross-over integration.

Although our toolkit was optimized for the exchange of integration sites, it can be customized in manifold ways as will be outlined in the discussion. A quick user manual with helpful information for vector construction can be found in the supplemental material (Text S1).

The customizable vector collection is based on SEVA to allow re-use of E. coli parts

Currently, most vectors and plasmids used for genetic manipulation of B. subtilis are propagated in E. coli. As demonstrated for NarK and β-carotene production, the E. coli vector propagation strongly depends on the ori and resistance markers in ways that cannot be foreseen yet9. Consequently, we designed our vectors to be flexible not only with respect to their cargo, but also their ori and E. coli resistance marker. For this purpose, we based our vector toolkit on SEVA. This standard was designed for vectors to allow exchanging of the cargo, resistance marker and ori, using defined rare type II restriction enzymes7. The accompanying vector collection was very well-received by scientists and is still growing. It can be used to adapt the vectors of our toolbox, which are therefore named Bacillus SEVA siblings. SEVA restriction sites were removed from critical parts of the entry vectors, so that the final vectors adhere to the SEVA standard, if no forbidden restriction sites are present in the customized parts.

Bacillus SEVA siblings are assembled de novo from multiple fragments via Golden Gate cloning

As depicted in Fig. 1b, the integration and replication parts of an integration vector are separated by the regions of homology, called up and down. Consequently, the exchange of both up and down fragments with standard cloning techniques would involve two cloning and verification steps without a selectable marker. To avoid laborious stepwise cloning, advanced cloning techniques for the easy, fast, efficient and directed one-pot assembly of multiple fragments are available, such as Gibson assembly26,27 or Golden Gate cloning23,24. The latter is based on a ligase and type IIS endonucleases (e.g. BsaI), which cut outside (next to) their recognition site. Consequently, the restriction site can be designed according to need and separated from the recognition site during the cloning procedure. The reaction mix can include linear and circular DNA, containing restriction sites for the same enzyme but different overhangs. This allows the easy and efficient assembly of up to ten parts in the correct order and loss of the recognition sites in the final vector23. For our vector toolbox, we chose Golden Gate assembly for the following reasons: (i) Gibson assembly usually must be established in a lab to run smoothly, whereas restriction based cloning is more robust. (ii) The up and down regions need to be PCR-generated and Gibson assembly asks for longer overhangs thus increasing the primer costs.

Instead of re-using and modifying preexisting vectors, each new vector will be freshly assembled by combining four fragments via Golden Gate assembly (Fig. 2): the up and down fragments, the replication part, and the cargo which includes an antibiotic marker for selection in Bacillus and a MCS. Therefore, all vectors and sequences offered through the Bacillus Genetic Stock Center (BGSC) and the SEVA collection (see Table 1 and supplementary data S2) are entry vectors to allow customized assembly, but no final Bacillus vectors.

Figure 2

Assembly of a Bacillus SEVA sibling pBS. (a) Collection of entry parts needed for the assembly of a pBS vector: one cargo vector, one destination vector, one up and one down flanking homology fragment. The latter two are depicted as PCR fragments, but can also be located on a vector. Each of the desired fragments is flanked by IIS-restriction sites where the recognition site (R) is located outside the desired fragment. The compatibility of the resulting overhangs is indicated with letters and a color gradient, e.g. E1 and E2 overhangs can anneal. (b) Intermediate stage of the Golden Gate assembly, showing the desired fragments and some of the by-products (grey). (c) Creation of the final vector, including some possible by-products (grey). Only the destination vector and the final vector carry the ampicillin resistance marker and will be selected for after transformation of the reaction mix into E. coli. The destination vector will be counter-selected by a red/white screen based on an mRFP-marker.

Table 1 Vectors of the Bacillus SEVA siblings toolbox.

Architecture of Bacillus SEVA siblings

Figure 1 compares a final Bacillus SEVA sibling (pBS), with a standard SEVA vector for E. coli. SEVA suggests a designated location for the insertion of special features that are not part of the cargo. They are positioned at the terminator sequences and next to, but not obstructing the AscI and SwaI recognition sites which allow the exchange of SEVA vector parts. In line with this regulation, the homology regions and resistance marker are located at both terminator sequences. The respective Bacillus resistance cassette is placed between the cargo and the T0-terminator and directed counter-clockwise in order to not interfere with the transcription of the cargo. SEVA-vectors are named according to their features in a number-based code, see SEVA 2.0 for a comprehensive description6. We suggest the naming of final vectors to be based on the SEVA-standard, in which the E. coli features are specified in three digits: the first digit describes the E. coli resistance marker, e.g. 1 for ampicillin resistance or 2 for kanamycin. The second digit indicates the ori, e.g. 4 for pRO1600/ColE1 or 9 for pBR322/ROP. The third digit encodes the cargo, e.g. 1 for the default MCS or 3 for the lacZα-pUC18 MCS. pSEVA243 consequently is a vector mediating kanamycin resistance [2] with a high copy number [4 = pRO1600/ColE1] that carries a MCS for blue/white screening [3 = lacZα-pUC18]. Bacillus-specific features should be added behind, so pBS143K-amyE is a vector with an E. coli ampicillin resistance marker [1], pRO1600/ColE1 ori [4] and lacZα-pUC18 MCS [3] that carries a kanamycin Bacillus resistance marker [K] and integrates into the amyE-locus [amyE].

We conceptualized the assembly so that SEVA-parts can be exchanged before or after the assembly of the final vectors, e.g. to accommodate the presence of either SEVA- or Golden Gate-forbidden restriction sites.

Golden Gate shuffling to assemble Bacillus SEVA siblings with type IIS restriction enzymes

In our current set-up, four different fragments are needed to assemble the final vector: the up and down homology fragments, the cargo with the Bacillus resistance cassette and the part for replication and selection in E. coli, which we call destination vector. These entry parts can be combined using Golden Gate assembly as detailed in Fig. 2. The up and down homology fragments can be used either as PCR products (as depicted) or as cargo of the specialized vectors containing the up (pSEVA243X) and down (pSEVA243Y) fragments, respectively. Each entry part is flanked by recognition sites for a type IIS restriction endonuclease, which creates overhangs that allow for the directional assembly of all entry parts. Compatible overhangs are named with the same capital letter in Fig. 2a.

As necessary for Golden Gate assembly, the recognition sites are located “outside” of the part desired for the assembly, so that correctly assembled parts cannot be re-cleaved – in contrast to the re-ligation products of entry vectors. By this means, assembly of mostly correct final vectors is ensured. The destination vector carries a different antibiotic marker than all other entry vectors to select for vectors carrying the correct backbone. An mRFP1-cassette present on the original destination vector is used for red/white screening. This part is removed during assembly of the final vector so that colonies carrying the original vector appear red and colonies carrying the correct final vector appear white.

Classic Golden Gate assembly uses BsaI and BbsI (=BpiI) as type IIS restriction enzymes23,28, but BsmBI and BtgZI were recently found to also be suitable29,30. Our Bacillus SEVA siblings toolbox accommodates all four of them for assembly to circumvent compatibility issues with the desired genomic region, e.g. the presence of one or more recognition sites in the up and down fragments. In addition to BsaI, BbsI, BsmBI and BtgZI, we also included a fifth restriction enzyme (AarI)– not reported previously for its use in Golden Gate assembly.

Golden Gate assembly restriction sites are arranged in special MCS-IIS

All four entry parts are flanked by MCSs that contain recognition sites for all five type IIS restriction enzymes (MCS-IIS). Inside each MCS-IIS, recognition and restriction sites are designed so that the same overhang sequence is created, independent of the enzyme used. The overhangs are non-palindromic and differ in at least two nucleotides to ensure the correct assembly of the desired vector. For ease of understanding, fusion sites were named with capital letters B, C, E, and F as indicated in Fig. 2. Figure 3 shows the annotated MCS-IIS C2, in which all five enzymes create the overhang GCGA. For the detailed sequence of all 8 MCS-IIS, see Fig. S1.

Figure 3

Architecture of the MCS-IIS C2. This DNA-sequence is located on the cargo vector between the E. coli ori and the Bacillus antibiotic marker. The recognition sites for five type IIS restriction enzymes (AarI, BtgZI, BbsI, BsaI, BsmBI), each designed to create a 5′ GCGA-overhang are encoded on the DNA stretch. Architecture of all MCS-IIS can be found in Fig. S1.

To compare the assembly efficiency as a function of the enzyme of choice, we used one set of entry vectors to construct final vectors with identical features. However, nucleotide sequences differ at the assembly scar sites, due to the MCS-IIS. We used pSEVA23X-amyE (up), pSEVA243Y-amyE (down), pBSc243M and pBSd141R as entry vectors. In this case, the cargo carries a lacZα-fragment that allows for blue-white screening and colonies carrying the correct vector appear blue in the presence of 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal). Red colonies carry the original destination vector and white colonies an incorrect vector. We checked colony color, test digest and sequencing results and found good assembly efficiencies for four enzymes: AarI, BbsI, BsaI and BsmBI (Table 2). For BtgzI however, we failed in finding conditions allowing the correct assembly of the final vector. This was surprising, since its use in Golden Gate assembly has been described previously in combination with BsmBI30. Even if the desired fragments were digested and gel purified separately, the ligation yielded no correct vector. But in principle, BtgZI can be used for assembly once suitable conditions are found. In contrast, BsaI, BpiI and BsmBI were particularly well-suited with efficiencies of >95% in routinely vector assembly (data not shown). The optimized conditions we used for each enzyme are given briefly in the Methods section and are described in more detail in the Supplemental Text S1.

Table 2 Assembly efficiencies* depending on restriction enzyme.

Taken together, Bacillus SEVA siblings vectors can be efficiently assembled using one of four type IIS restriction endonucleases. In addition to the established enzymes (BsaI, BbsI and BsmBI), AarI was also found to be suitable for Golden Gate assembly. Because of its 7 bp recognition site, it should be found less frequently in genomic sequences compared to the “classical” enzymes with a 6 bp recognition site.

Our collection of entry parts for the assembly of Bacillus SEVA siblings

For the assembly of Bacillus SEVA siblings vectors, four different categories of entry parts are needed: cargo and resistance, up, down, and the replication part. They will be described in more detail below. The entry vectors offered with our toolbox are depicted in Fig. 4 and listed with detailed descriptions in Table 1.

Figure 4

Vector suite for the generation of Bacillus SEVA siblings. Schematic representation of the vector architectures, details are listed in Table 1. (a) Vectors for flanking homology regions. Up fragments (PCR product) can be stored in pSEVA243X and down in pSEVA243Y, each linearized with EcoRV. The respective MCS-IIS are encoded on the vectors. If required restriction site are already encoded on the primer overhangs, fragments can be stored in pSEVA243 or used directly for Golden Gate assembly. (b) Cargo vectors carry one of the following Bacillus antibiotic markers: Ble, Cat, Kan, MLS, Spc, Tet, Zeo and either the default MCS (pBSc241res) or the lacZα*-pUC18 MCS for blue/white-screening. Vectors carrying the Bacillus kanamycin resistance marker utilize the medium copy number pBR322/ROP ori, all others the high copy number pRO1600/ColE1. Backbones are also available without Bacillus marker to allow insertion of a new or customized marker. (c) Destination vectors carry an mRFP1-cassette as cargo for red/white screening and an ampicillin resistance marker for selection in E. coli. They are available with high (pRO1600/ColE1) or medium (pBR322/ROP) copy number origins of replication.

pBSc: Cargo and resistance

Since for most experiments the same cargo will usually be combined with the same resistance marker, both are located on the cargo vector pBSc. As a result, only four instead of five parts have to be assembled, thereby increasing cloning efficiency. For our collection of cargo vectors, we used well-established antibiotic resistance markers to enable selection on phleomycin D, chloramphenicol, kanamycin, macrolide and streptogramin B antibiotics (MLS), spectinomycin, tetracycline, and zeocin (Table 3). Forbidden restriction sites were removed and transcriptional terminators were added where necessary. Transcription occurs in the opposite direction to the cargo. Per default, the cargo vector uses the high copy number ori pRO1600/ColE1. For the kanamycin resistance marker, this vector was unstable, so the medium copy number ori pBR322/ROP was used. All cargo vectors are offered with the default SEVA MCS (pUC18-related polylinker without lacZα,7) or the lacZα*-pUC18 MCS (Table 1). The cargo of choice can be inserted into the cargo vector or into the final vector, depending on enzyme compatibility and cloning strategy.

Table 3 Description of antibiotic markers used in this study.

Up and down flanking homology regions

The choice of up and down flanking homology regions depends on and needs to be adjusted to the Bacillus strain and experiment. The regions need specific overhangs for the subsequent assembly. This can be achieved by either of the following two strategies: (i) PCR-amplified fragments of choice (without overhangs) are cloned blunt end into the EcoRV-linearized pSEVA243X (up) or pSEVA243Y (down), respectively. As a result, the insert receives the matching MCS-IIS encoded on these vectors which can be used for Golden Gate assembly. (ii) Incorporation of one restriction site to the primers will directly allow the use of the PCR product for Golden Gate assembly.

pBSd: replication part in E. coli

The destination vector carries the replication part of the final vector. Its resistance (ampicillin marker) differs from all other entry vectors to allow selection for the correct backbone after vector assembly. A high copy number (pBSd141R) and medium copy number version (pBSd191R) are provided. Both carry an mRFP cassette in their default MCS, thereby allowing red/white screening: After Golden Gate assembly, colonies with the original destination vector will appear red and therefore can be discarded. If different E. coli features are required for the final vector, they can be exchanged in the destination vector or the final vector, depending on needs and enzyme compatibilities.

Assembly, transformation and integration of pBS vectors

The reaction for the vector assembly contains the entry parts, one of the type IIS restriction enzymes, ligase and buffer. The specific protocols depend on the enzyme used and can be found in the Materials and Method section. Competent E. coli cells are subsequently transformed with the reaction mix and plated on selective media containing ampicillin (and X-Gal in case of blue/white screening). Verification of the final vector can be achieved in a two-step procedure using test digests followed by sequencing. For the latter, primers TM3782 and TM3783 (if pBSd141R was used as destination vector) or TM3783 and TM5128 (pBSd191R) are recommended. The sequencing results should cover the up and down region as well as the adjacent assembly scars to ensure correct assembly of all parts. We tested all entry parts of the toolbox (Table 1) for their assembly efficiencies (Tables S4S6) and found that in all cases it was more than sufficient to test six colonies of the correct color to obtain a correct final vector. If the high copy number destination vector is used, or selection for the correct cargo is possible, e.g. via blue/white screening, assembly efficiencies are even higher.

Prior to transformation of B. subtilis W168, all vectors have to be linearized, e.g. using ApaI, to ensure chromosomal integration by double homologous recombination. All transformations were successful – except for vectors carrying the tetracycline resistance marker where only for one of two loci could be targeted (Tables S4, S5). The transformation efficiencies depended on the selection marker (Tables 2, 3 and S4S6). Correct integration was verified by colony PCR or physiological tests; e.g. a starch test for integration into the amyE locus.

As a proof of concept, we analyzed the expression of the reporter gene mkate2 under control of the xylose-inducible promoter PxylA at five loci spread across the chromosome (Table 4): amyE, ykoS, ypqP (prophage SPβ), nkd, and thrC at positions 28, 111, 183–195, 203, and 283° on the circular chromosome, respectively. amyE and thrC are early-discovered and frequently used integration loci close to the ori, encoding a starch-hydrolyzing alpha-amylase and the threonine synthase, respectively. The 130 kb-large prophage SPβ, which itself is inserted at the ypqP locus, is known to be not essential and was targeted to demonstrate the possibility of deleting large genomic regions. The two remaining genes, ndk and ykoS, encode a nucleoside diphosphate kinase and a gene of unknown function, respectively. Those non-essential genes were chosen based on their chromosomal location. Since the same reporter was used in all vector constructs, the reporter-cargo was added at the entry vector level to pBSc241M, resulting in pBSc241M-PxylA-mkate2. pBSd141R (ori pRO/ColE1, high copy number) was chosen as the destination vector and primers were designed for PCR-product assembly of up and down fragments via BsaI. Assembly and verification of four final vectors was achieved within five working days. The vector carrying homology regions for integration into ndk was instable (small colonies), necessitating a change to pBSd191R (ori pBR322/ROP, medium copy number) as a destination vector. After transformation of B. subtilis W168, the resulting integrants were verified by starch or threonine auxotrophy tests (amyE and thrC, respectively), or colony PCR for the remaining loci (Table 4). The xylose-inducible promoter PxylA was fully induced with xylose in exponentially growing cultures. The fluorescence intensity of B. subtilis cells was quantified in triplicates in a microtiter plate reader as a measure for mKate2 production (Fig. 5). The results highlight a dependence of the expression levels on the chromosomal location: As demonstrated previously, genes close to the ori (0°/360°) tend to be expressed at higher levels than those close to the termination region (180°). One exception was the ykoS-locus, located at the replication termination site, which had a higher reporter activity than the constructs inserted at the ori-proximal ypqP or ndk sites. Nevertheless, these results are in good agreement with a recent study31, thereby demonstrating that our vector toolbox can for example be used to study the effect of chromosome location on expression levels in a simple and straightforward manner.

Table 4 Assembly efficiencies* and reporter activity depending on integration sites.
Figure 5

Maximal promoter activity of PxylA depends on chromosomal location. The reporter construct PxylA-mkate2 was integrated into the B. subtilis genome at five different chromosomal loci, as indicated. mKate expression was maximally induced with the addition of 0.2% xylose and fluorescence intensity was measured as an indicator for mKate abundance. The fluorescence intensity is given as a function of the chromosome position. The error bars show the standard deviation of three independent biological replicates. The data was fitted to two second order polynomial functions: dashed line: no constrains, R2 = 0.85; dotted line: minimum was set to X = 180, R2 = 0.72.


B. subtilis is a versatile heterologous host with powerful genetics allowing precise genomic manipulation. But so far, the available integrative vectors could not easily be adapted to different strains or, more importantly, other bacilli, since no modular toolbox was available to allow free combination of resistance markers and integration sites. To fill this gap, we present a rigorously evaluated toolbox for the construction of integrative vectors with customizable flanking homology regions for Bacillus sp. The final pBS vector can be assembled within one week from four entry parts via efficient Golden Gate cloning. Our vector suite offers the choice of four type IIS restriction endonucleases (AarI, BbsI, BsaI and BsmBI) for Golden Gate assembly, seven Bacillus antibiotic resistance markers, and two different MCSs, available on either a high or medium copy number E. coli backbone. The toolbox is adjusted to and widely compatible with the E. coli SEVA standard to allow reusability of its cargos or vector parts. All components provided in this toolbox were successfully tested for assembly efficiency, functionality, and usability in B. subtilis. The supplementary assembly guide (Text S1) provides necessary information to facilitate a fast and efficient cloning process.

In the course of developing our vectors, a few combinations were discovered which were difficult to handle on high copy number vectors (ori pRO1600/ColE1): (i) The Bacillus kanamycin resistance marker was prone to mutations and is therefore offered with the medium copy number ori pBR322/ROP. (ii) The lacZα*-pUC18 MCS differs from the original lacZα-pUC18 MCS in a single-nucleotide polymorphism causing a premature stop codon in lacZα that does not impair blue/white screening. There was a strong selection pressure against the original lacZα-pUC18 MCS (very small colonies), which caused transposon integrations and single nucleotide polymorphisms in lacZα. Consequently, we used the mutated but fully functional lacZα*-pUC18 MCS for our constructs. (iii) The final ndk-vectors only contained the ndk down fragment when using the high copy number pBSd141R as destination vector. However, assembly with the medium copy number vector pBSd191R was successful in the first attempt.

Based on our experience, the use of a medium copy number vector is highly recommended as the first trouble shooting strategy in case of cloning issues, especially if slowly growing colonies occur. Also, low copy number oris (e.g. p15A and pSC101) or different directionality of the inserts could be used in case toxicity issues occur, which result in genetic instabilities even on medium copy number vectors.

Furthermore, the tetracycline resistance marker could not confer resistance in B. subtilis W168 when the amyE locus was targeted. However, when using a different integration locus (lacA), the transformation was successful. Due to the dependence of expression levels on the chromosomal region, we suspect this to be causing the transformation issues.

The Bacillus SEVA siblings toolbox was developed to provide a versatile starting point for the efficient construction of personalized, yet standardized vectors. Both, the entry parts as well as final vectors can be customized to meet personal needs. The MCS or cargo and ori can easily be exchanged according to the SEVA standard and new or modified resistance markers can be inserted in markerless cargo vectors via AscI and MluI, e.g. markers flanked with target sites for recombinases. These systems would allow directed, recombinase-mediated removal of the marker after chromosomal integration (for reviews on recombinase-mediated cassette exchange, e.g. by Flp and Cre/loxP, see32,33).

It should also be pointed out that the free choice of chromosomal integration loci provided by our vector toolbox allows for replacing even larger (non-essential) chromosomal areas directly in the process of plasmid integration. We have demonstrated this possibility by integrating an mkate2-expressing plasmid into the ypqP locus of the B. subtilis chromosome, thereby deleting the 130 kb prophage SPβ (Table 4). This combined integration/deletion at any desired chromosomal position will greatly enhance the possibilities in genetically manipulating B. subtilis.

The current resistance markers located on cargo vectors were tested for their functionality only in B. subtilis W168. Since they originate from broad host range vectors, they are expected to be functionally expressed in many low G + C Gram-positive bacteria (phylum Firmicutes). The Bacillus SEVA siblings might therefore be suitable for e.g. other bacilli, increasing the number of species where customized vectors can be used for genetic manipulation. Indeed, preliminary results from an ongoing study using our vectors reported similar or even better vector assembly efficiencies as those described in Table 2. Choramphenicol- or erythromycin-resistant mutants were readily obtained in Paenibacillus polymyxa (order: Bacillales, family: Paenibacillaceae) (Christoph Engl, personal communication).

If transformation efficiencies are too low, conjugation can be performed using the oriT already included in the SEVA standard, which provides efficient transfer into various Gram-negative and -positive hosts7,34,35. It can be exchanged with a conjugation marker of choice in the destination or final vector to meet the needs of the target organism. Recombination efficiency of homologous sequences into the chromosome varies more than 30-fold along the B. subtilis chromosome36, and even more between species. In general, integration rates can be improved by using longer stretches of DNA or a vector with a temperature-sensitive replicon (e.g. based on pMAD3,37). Vector replication not only increases the number of vectors per cell and number of cells carrying a vector (by passing it on to the next generation), but also supports the second recombination event during which the vector backbone is excised from the chromosome3. A temperature-sensitive origin of replication can therefore be added to the destination or final vector to improve recombination with the chromosome.

Moreover, the construction logic described for the Bacillus SEVA siblings – that is, the restriction enzymes used for the assembly of the final integrative plasmids – can of course also be applied for developing other SEVA-compatible integration vectors for completely unrelated microorganisms in a similar fashion. But this would require adjusting the corresponding resistance cassettes applicable to these bacteria.

Here, we present the first fully modular, yet standardized vector toolkit for integrative vectors in B. subtilis and beyond. We hope that the Bacillus SEVA siblings vector toolbox will prove to be useful for projects throughout the Bacillus world. In case personalized entry vectors (pBSc, pBSd) are created, we would like to encourage sharing them with the Bacillus community, e.g. via the SEVA or BGSC collections, to further improve the tools available for genetically manipulating these powerful organisms.

Material and Methods

Bacterial strains and growth conditions

All strains used in this study are listed in Table S1. B. subtilis and E. coli were routinely grown in lysogeny broth (LB) medium (1% (w/v) tryptone, 0.5% (w/v) yeast extract, 1% (w/v) NaCl) at 37 °C with agitation (220 rpm). Solid media additionally contained 1.5% (w/v) agar. Selective media contained appropriate antibiotics, as provided in Table 3.


E. coli (XL1 blue, Agilent Technologies, Santa Clara, CA, USA) competent cells were prepared and transformed according to the rubidium chloride method38, achieving ~5*106 colony forming units (CFU) per µg pUC18 DNA. Transformations of B. subtilis were carried out as described previously17,22. The integration of plasmids into the B. subtilis genome was checked on starch plates (amyE), with minimal medium lacking threonine (thrC) or colony PCR (lacA, ypqP, ykoS, ndk). Detailed protocols were published previously17,39.

DNA manipulation

Vectors and plasmids used in this study are listed in Table 1. General cloning procedure, such as endonuclease restriction digest, ligation and PCR, was performed with enzymes and buffers from New England Biolabs® (NEB; Ipswich, MA, USA) or Thermo ScientificTM (Waltham, MA, USA) according to the respective protocols. Phusion® polymerase was used for PCRs if the resulting fragment was further used, otherwise OneTaq® was the polymerase of choice. PCR-purification was performed with the HiYield PCR Gel Extraction/PCR Clean-up Kit (Süd-Laborbedarf GmbH (SLG), Gauting, Germany). Plasmids were prepared using alkaline lysis and subsequent DNA precipitation. All plasmids created during this study are listed in Table 1, their construction is described in supplemental Table S2 and all primer sequences are given in Table S3.

Golden Gate assemblies of final vectors

Golden Gate assemblies were performed using T4 DNA ligase (30 WU) from Thermo ScientificTM with the accompanied buffer. BsaI, BbsI, BsmBI and BtgZI were purchased from NEB, AarI from Thermo ScientificTM. Entry parts were diluted to 20 or 40 nM stock concentrations and 2 or 1 µl were used per 15 µl reaction, respectively. For all enzymes, 0.5 µl were used per reaction, except for AarI where 1.5 µl were necessary because of its lower activity. BsaI and AarI are active in ligase buffer, but for AarI the accompanying oligonucleotides were supplemented for optimal efficiency. For BbsI half of the ligase buffer was replaced by NEBuffer 2.1 and for BsmBI half was replaced by NEBuffer 3.1.

The general assembly protocol was 37 °C, 30 min; 16 °C, 30 min; (37 °C, 3 min; 16 °C, 5 min) × 15; 37 °C, 10 min; 50 °C, 10 min; 80 °C, 10 min. The exception was BsmBI, where 55 °C were required for the first incubation step and ligase was only added afterwards. 7.5 µl of the final reaction were used for E. coli transformation.

Measurement of PxylA activity

Fluorescence intensity of strains carrying a transcriptional fusion of the xylose-inducible promoter PxylA and the fluorescence reporter mkate2 were assayed using a SynergyTM NEOALPHAB multi-mode microplate reader from BioTek® (Winooski, VT, USA). The reader was controlled using the software Gen5TM (version 2.06). Cells were inoculated 1:1000 from fresh overnight cultures and grown to OD600 ~0.2, treated with 0.2% xylose and grown for two hours. Cells were harvested by centrifugation and resuspended in phosphate buffered saline (137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4). 200 µl per well in 96-well plates (black walls, clear bottom; Greiner Bio-One, Frickenhausen, Germany) were measured for their OD600, and mKate-fluorescence using the monochromator with following parameters: endpoint measurement, gain: 100, excitation wavelength: 588 nm, emission wavelength: 633 nm.

Data availability

The vectors generated in this study are available from the Bacillus Genetic Stock Center (BGSC,, accession numbers ECE701-26) and the SEVA collection ( For sequence information, the following accession numbers apply: GenBank, KY995178 to KY995203; ACS Synthetic Biology Registry (, JPUB_008862 to JPUB_008887).

Change history

  • 06 March 2018

    A correction to this article has been published and is linked from the HTML and PDF versions of this paper. The error has been fixed in the paper.

  • 17 January 2018

    A correction to this article has been published and is linked from the HTML version of this paper. The error has been fixed in the paper.


  1. 1.

    Wright, O., Delmans, M., Stan, G. B. & Ellis, T. GeneGuard: A modular plasmid system designed for biosafety. ACS Synth Biol 4, 307–316, (2015).

    Article  CAS  PubMed  Google Scholar 

  2. 2.

    Shimada, T. et al. Classification and strength measurement of stationary-phase promoters by use of a newly developed promoter cloning vector. J Bacteriol 186, 7112–7122, (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. 3.

    Arnaud, M., Chastanet, A. & Debarbouille, M. New vector for efficient allelic replacement in naturally nontransformable, low-GC-content, gram-positive bacteria. Appl Environ Microbiol 70, 6887–6891, (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. 4.

    Sinah, N., Williams, C. A., Piper, R. C. & Shields, S. B. A set of dual promoter vectors for high throughput cloning, screening, and protein expression in eukaryotic and prokaryotic systems from a single plasmid. BMC Biotechnol 12, 54, (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. 5.

    Andrianantoandro, E., Basu, S., Karig, D. K. & Weiss, R. Synthetic biology: new engineering rules for an emerging discipline. Mol Syst Biol 2(2006), 0028, (2006).

    Article  PubMed  Google Scholar 

  6. 6.

    Martinez-Garcia, E., Aparicio, T., Goni-Moreno, A., Fraile, S. & de Lorenzo, V. SEVA 2.0: an update of the Standard European Vector Architecture for de-/re-construction of bacterial functionalities. Nucleic Acids Res 43, D1183–1189, (2015).

    Article  CAS  PubMed  Google Scholar 

  7. 7.

    Silva-Rocha, R. et al. The Standard European Vector Architecture (SEVA): a coherent platform for the analysis and deployment of complex prokaryotic phenotypes. Nucleic Acids Res 41, D666–675, (2013).

    Article  CAS  PubMed  Google Scholar 

  8. 8.

    Standard European Vector Architecture, and (accessed March 2017)

  9. 9.

    Kim, S. H., Cavaleiro, A. M., Rennig, M. & Norholm, M. H. SEVA Linkers: A versatile and automatable DNA backbone exchange standard for synthetic biology. ACS Synth Biol 5, 1177–1181, (2016).

    Article  CAS  PubMed  Google Scholar 

  10. 10.

    van Dijl, J. M. & Hecker, M. Bacillus subtilis: from soil bacterium to super-secreting cell factory. Microb Cell Fact 12, 3, (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. 11.

    Hohmann, H.-P., van Dijl, J. M., Krishnappa, L. & Prágai, Z. In Industrial Biotechnology, 221–297 (Wiley-VCH Verlag GmbH & Co. KGaA, 2017).

  12. 12.

    Liu, Y., Li, J., Du, G., Chen, J. & Liu, L. Metabolic engineering of Bacillus subtilis fueled by systems biology: Recent advances and future directions. Biotechnol Adv 35, 20–30, (2017).

    Article  CAS  PubMed  Google Scholar 

  13. 13.

    Mell, J. C. & Redfield, R. J. Natural competence and the evolution of DNA uptake specificity. J Bacteriol 196, 1471–1483, (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. 14.

    Zucca, S., Pasotti, L., Mazzini, G., De Angelis, M. G. & Magni, P. Characterization of an inducible promoter in different DNA copy number conditions. BMC Bioinformatics 13(Suppl 4), S11, (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. 15.

    Adams, C. W. & Hatfield, G. W. Effects of Promoter Strengths and Growth Conditions on Copy Number of Transcription-Fusion Vectors. J Biol Chem 259, 7399–7403 (1984).

    CAS  PubMed  Google Scholar 

  16. 16.

    Guerout-Fleury, A. M., Frandsen, N. & Stragier, P. Plasmids for ectopic integration in Bacillus subtilis. Gene 180, 57–61 (1996).

    Article  CAS  PubMed  Google Scholar 

  17. 17.

    Radeck, J. et al. The Bacillus BioBrick Box: generation and evaluation of essential genetic building blocks for standardized work with Bacillus subtilis. J Biol Eng 7, 29, (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. 18.

    Hartl, B., Wehrl, W., Wiegert, T., Homuth, G. & Schumann, W. Development of a new integration site within the Bacillus subtilis chromosome and construction of compatible expression cassettes. J Bacteriol 183, 2696–2699, (2001).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. 19.

    Schmalisch, M. et al. Small genes under sporulation control in the Bacillus subtilis genome. J Bacteriol 192, 5402–5412, (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. 20.

    Stülke, J. et al. Induction of the Bacillus subtilis ptsGHI operon by glucose is controlled by a novel antiterminator, GlcT. Mol Microbiol 25, 65–78 (1997).

    Article  PubMed  Google Scholar 

  21. 21.

    Wach, A. PCR-synthesis of marker cassettes with long flanking homology regions for gene disruptions in S. cerevisiae. Yeast 12, 259–265 (1996).

    Article  CAS  PubMed  Google Scholar 

  22. 22.

    Harwood, C. R. & Cutting, S. M. Molecular Biological Methods for Bacillus (John Wiley & Sons, 1990).

  23. 23.

    Engler, C., Kandzia, R. & Marillonnet, S. A one pot, one step, precision cloning method with high throughput capability. PLoS One 3, e3647, (2008).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  24. 24.

    Engler, C., Gruetzner, R., Kandzia, R. & Marillonnet, S. Golden gate shuffling: a one-pot DNA shuffling method based on type IIs restriction enzymes. PLoS One 4, e5553, (2009).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  25. 25.

    Khasanov, F. K., Zvingila, D. J., Zainullin, A. A., Prozorov, A. A. & Bashkirov, V. I. Homologous recombination between plasmid and chromosomal DNA in Bacillus subtilis requires approximately 70 bp of homology. Mol Gen Genet 234, 494–497, (1992).

    Article  CAS  PubMed  Google Scholar 

  26. 26.

    Gibson, D. G. Synthesis of DNA fragments in yeast by one-step assembly of overlapping oligonucleotides. Nucleic Acids Res 37, 6984–6990, (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. 27.

    Gibson, D. G., Smith, H. O., Hutchison, C. A. 3rd, Venter, J. C. & Merryman, C. Chemical synthesis of the mouse mitochondrial genome. Nat Methods 7, 901–903, (2010).

    Article  CAS  PubMed  Google Scholar 

  28. 28.

    Weber, E., Engler, C., Gruetzner, R., Werner, S. & Marillonnet, S. A modular cloning system for standardized assembly of multigene constructs. PLoS ONE 6, e16765, (2011).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  29. 29.

    Sarrion-Perdigones, A. et al. GoldenBraid: An iterative cloning system for standardized assembly of reusable genetic modules. PLoS ONE 6, e21622, (2011).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  30. 30.

    Sarrion-Perdigones, A. et al. GoldenBraid 2.0: a comprehensive DNA assembly framework for plant synthetic biology. Plant Physiol 162, 1618–1631, (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. 31.

    Sauer, C. et al. Effect of genome position on heterologous gene expression in Bacillus subtilis: An unbiased analysis. ACS Synthetic Biology 5, 942–947, (2016).

    Article  CAS  PubMed  Google Scholar 

  32. 32.

    Turan, S., Zehe, C., Kuehle, J., Qiao, J. & Bode, J. Recombinase-mediated cassette exchange (RMCE) - a rapidly-expanding toolbox for targeted genomic modifications. Gene 515, 1–27, (2013).

    Article  CAS  PubMed  Google Scholar 

  33. 33.

    Dong, H. & Zhang, D. Current development in genetic engineering strategies of Bacillus species. Microb Cell Fact 13, 63, (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. 34.

    Martinez-Garcia, E., Calles, B., Arevalo-Rodriguez, M. & de Lorenzo, V. pBAM1: an all-synthetic genetic tool for analysis and construction of complex bacterial phenotypes. BMC Microbiol 11, 38, (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. 35.

    Trieu-Cuot, P., Carlier, C., Martin, P. & Courvalin, P. Plasmid transfer by conjugation from Escherichia coli to Gram-positive bacteria. FEMS Microbiol Lett 48, 289–294, (1987).

    Article  CAS  Google Scholar 

  36. 36.

    Vagner, V. & Ehrlich, S. D. Efficiency of homologous DNA recombination varies along the Bacillus subtilis chromosome. J Bacteriol 170, 3978–3982 (1988).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. 37.

    Rachinger, M. et al. Size unlimited markerless deletions by a transconjugative plasmid-system in Bacillus licheniformis. J Biotechnol 167, 365–369, (2013).

    Article  CAS  PubMed  Google Scholar 

  38. 38.

    Mülhardt, C. Der Experimentator. Molekularbiologie/Genomics. 6 edn, (Spektrum Akademischer Verlag, 2009).

  39. 39.

    Sambrook, J. & Russell, D. W. Molecular Cloning - a laboratory manual (Cold Spring Harbor Laboratory Press, 2001).

  40. 40.

    Joseph, P., Fantino, J. R., Herbaud, M. L. & Denizot, F. Rapid orientated cloning in a shuttle vector allowing modulated gene expression in Bacillus subtilis. FEMS Microbiol Lett 205, 91–97 (2001).

    Article  CAS  PubMed  Google Scholar 

  41. 41.

    Guerout-Fleury, A. M., Shazand, K., Frandsen, N. & Stragier, P. Antibiotic-resistance cassettes for Bacillus subtilis. Gene 167, 335–336 (1995).

    Article  CAS  PubMed  Google Scholar 

  42. 42.

    Balbás, P. & Bolívar, F. In Recombinant Gene Expression: Reviews and Protocols (eds Paulina Balbás & Argelia Lorence) 77–90 (Humana Press, 2004).

Download references


We thank Wilfried Meijer for fruitful initial design discussions and Beate Lüdke for assistance with B. subtilis transformations. This work was supported by the Deutsche Forschungsgemeinschaft (DFG) priority program SPP1617 ‘Phenotypic Heterogeneity and Sociobiology of Bacterial Populations’ [grant number MA 2837/3-2 to TM]. Funding for open access charge: dito.

Author information




J.R. and T.M. conceptualized Bacillus SEVA siblings. J.R. detailed and planned the experiments, analyzed the data and created the figures. J.R., D.M. and N.L. cloned the vectors and J.R. performed Golden Gate assemblies. J.R. and T.M. wrote the manuscript. All authors reviewed the manuscript.

Corresponding author

Correspondence to Thorsten Mascher.

Ethics declarations

Competing Interests

The authors declare that they have no competing interests.

Additional information

Change History: A correction to this article has been published and is linked from the HTML version of this paper. The error has been fixed in the paper.

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Radeck, J., Meyer, D., Lautenschläger, N. et al. Bacillus SEVA siblings: A Golden Gate-based toolbox to create personalized integrative vectors for Bacillus subtilis. Sci Rep 7, 14134 (2017).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing