Improved genetic transformation by disarmament of type II Restriction–Modification system in Streptococcus zooepidemicus

Abstract

Streptococcus zooepidemicus, group C Streptococci, is currently used for the industrial production of hyaluronic acid (HA). However, genetic manipulation of S. zooepidemicus is severely limited by its low transformation efficiency, which might be in part due to the Restriction–Modification (R–M) systems. The complete genome sequence of S. zooepidemicus ATCC39920 revealed the presence of two putative R–M systems, type I and type II. The putative type I R–M system is encoded by three closely linked genes: hsdR (SeseC_01315), hsdS, hsdM (SeseC_01318), and the putative type II R–M system consists of two closely linked genes: SeseC_02360 and yhdJ (SeseC_02362). Inactivation of hsdR, encoding the restriction endonuclease (REase) of the type I R–M system, showed no apparent effects on transformation efficiency, implying that disarmament of the type I R–M system alone is not sufficient for increasing transformation efficiency. However, inactivation of SeseC_02360, encoding the REase of the type II R–M system, improved transformation efficiency by 4.97 folds, indicating that type II R–M system is the major barrier that restricts genetic transformation in S. zooepidemicus. Furthermore, S. zooepidemicus strains lacking either of the two R–M systems are phenotypically indistinguishable from the wild-type in terms of cell growth and HA production. In summary, our study revealed that the type II R–M system is the main barrier to genetic transformation in S. zooepidemicus ATCC39920, and that the deletion of the type II R–M system renders S. zooepidemicus more transformable, thus facilitating metabolic engineering of this industrially important microorganism.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-022-03227-x.

Introduction

Streptococci, grouping in chains, are cocci-shaped and non-sporulating Gram-positive bacteria, of which the distinguishing feature is the thick capsule surrounding the cells. The major component of capsular polysaccharide is hyaluronic acid, a multiple-function biopolymer widely used in food, cosmetic, and biomedical fields. S. zooepidemicus, a group C Streptococci and a nonpathogenic microorganism of humans (Vázquez et al. 2010) was first used to produce HA in the early 1980s (Matsubara et al. 1991) and remains one of the major source of industrial HA production strains (de Oliveira et al. 2016).

In the last two decades, the complete genome sequences of several S. zooepidemicus strains have been reported, which greatly facilitates the prediction of gene functions and manipulation of metabolic pathways according to various objectives (Ma et al. 2011; Beres et al. 2008). However, low gene transformation efficiency of Streptococci is a major limiting factor for gene replacement through homologous recombination (Takamatsu et al. 2001). Several thermosensitive (Ts) replicons have been used to achieve high transformation efficiency for Streptococci (Maguin et al. 1992; Bhowmik et al. 1993; Perez-Casal et al. 1993). For instance, Takamatsu et al. (2001) demonstrated gene replacement in Streptococcus suis using a thermosensitive suicide vector pSET4s. Based on pSET4s, we developed a marker-less gene deletion system for S. zooepidemicus using sacB expression cassette as a counterselection marker (Sun et al. 2013).

Although thermosensitive vector can be used to remedy the disadvantage of low transformation efficiency for gene knocking out in S. zooepidemicus, significant improvements in its transformation efficiency will be valuable for metabolic engineering or construction of synthetic elements libraries. R–M systems are ubiquitous and regarded as the primitive immune systems in bacteria. The systems are also the major barrier to bacterial transformation by restricting foreign DNA and they can also influence the transformation efficiency (Murray 2002; Zhou et al. 2012; Rego et al. 2011).

The R–M systems usually comprise pairs of opposing enzymes, restriction endonuclease enzymes (REases), and the corresponding DNA methyltransferases (MTases). REases usually recognize and cleave foreign DNA sequences at the specific sites. MTases can transfer the methyl groups to the same specific sites, which can ensure the protection of host DNA from restriction activity. However, functions of the hypothetical R–M system genes in S. zooepidemicus, especially their influences on transformation efficiency as well as on growth and HA yield remain to be determined.

In this study, the roles of a type I R–M system and a type II R–M system in S. zooepidemicus were investigated. R–M system gene knockout mutants were constructed and their effects on the transformation efficiency were evaluated. Furthermore, the R–M system deficient mutants were subjected to growth and HA fermentation evaluation to assess the effects of gene deletion on general metabolism. This study is the first attempt to carry out genetic manipulation of the R–M systems in S. zooepidemicus. Successful elimination of the R–M systems will provide improved genetic transformation efficiency of S. zooepidemicus, thus accelerating the development of commercial strains for efficient production of HA.

Materials and methods

Strains and growth conditions

The strains used in this study are listed in Table 1. Escherichia coli JM109 was used for plasmid construction and was cultured in Luria–Bertani (LB) medium at 37 ℃. The wild type, S. zooepidemicus ATCC39920 (S12) and its derivatives were cultured in Todd-Hewitt Yeast (THY) medium (Sun et al. 2013) at 37 ℃. Mediums were supplemented with spectinomycin (50 μg/mL for E. coli, and 100 μg/mL for S. zooepidemicus) when required. For HA production, S. zooepidemicus was cultured in HA fermentation medium (FSB) following our previously reported protocols (Zhang et al. 2016).

Table 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant genotype and characteristics	Source
Strains
S12	Wild type, S. zooepidemicus ATCC39920	Laboratory
S1515	S12 derivative, ΔSeseC_02360	This work
S1518	S12 derivative, ΔhsdM	This work
S1812	S12 derivative, ΔhsdR	This work
Plasmids
pLH67	pSET4s::sacB, shuttle vector, temperature-sensitive (ts) replication origin and sacB gene	Laboratory
pLH881	pLH67-derivation with SeseC_02360 deletion fragment	This work
pLH994	pLH67-derivation with hsdM deletion fragment	This work
pLH995	pLH67-derivation with hsdR deletion fragment	This work

Bioinformatic analysis of the S. zooepidemicus S12 R–M systems

The R–M systems of S. zooepidemicus S12 strain was predicted via the NEB website (http://rebase.neb.com/rebase/rebase.html) (Gao et al. 2019). To determine the relatedness of R–M enzymes investigated in this study, the putative R–M enzymes of S. zooepidemicus S12 were compared to the sequences contained in the NCBI database by BLAST analyses (http://www.ncbi.nlm.nih.org).

Construction of individual mutant strains

The target gene was knocked out using a marker-less gene knocking out method as described previously (Sun et al. 2013). Briefly, the obtained up- and down-stream fragments of the target gene (N) via PCR were linked by overlapping PCR. The obtained DNA was ligated into the digested pSET4s::sacB to obtain targeting vector pSET4s::sacB::NLR via recombinant cloning. The plasmids used in this study are listed in Table 1. The primers used are listed in Table S1. The plasmids were transformed into S. zooepidemicus via high osmolarity electroporation method as previously described (Sun et al. 2013). S. zooepidemicus containing pSET4s::sacB::NLR grew at 30 ℃ for 12 h first, and then was cultured at 37 ℃ for 4 h in THY medium supplemented with spectinomycin to select the single-crossover stains. Then, the double crossover stains were selected on the THY plate added with 5% (w/v) sucrose. The sucrose-resistant and spectinomycin-sensitive clones were selected, and the target gene-deletion mutants were detected by PCR using the detective primer pair and further identified by DNA sequencing.

Transformation efficiency assay

Transformation efficiency was assessed by electroporation of competent S. zooepidemicus strains with 0.5 μg plasmid pLH67 conferring spectinomycin resistance isolated from E. coli JM 109. Calculation of the transformation efficiency was carried out according to the previous report (Nye et al. 2019). Briefly, electroporation was performed using a MicroPluser (BioRad, Hercules, CA, USA) under the following settings: Volts at 2.5 kV, Cuvette of 0.2 cm. Half of the transformants were plated onto THY agar supplemented with spectinomycin. In addition, the other half of the sample was dilutedly plated onto THY agar with no antibiotics to determine the total viable cell count. Transformation efficiency was defined as the number of spectinomycin resistant cells per total viable cell count.

Analytical procedures

The OD_600nm of the assay cultures was determined by a Shimadzu UV-1800 spectrophotometer (Kyoto, Japan). HA concentration of the strains was measured using cetyltrimethylammonium bromide (CTAB) (Sigma–Aldrich, United States) turbidimetric method as described previously (Chen and Qiang 2009; Xie et al. 2019). It has been reported that exopolysaccharide synthesized by S. zooepidemicus was identified and confirmed as HA (Vázquez et al. 2015). For exploring the effect of R–M system destruction on the exopolysaccharide synthesis, S. zooepidemicus wild type S12 and the three deletion mutants were cultivated in FSB medium at 37 ℃, 200 rpm for 24 h. Samples were taken at 12 and 24 h, respectively. The capsule of the strains was released via adding 0.1% (w/v) SDS and centrifuged at 10,000 g for 10 min to remove cells. The supernatant was mixed with three volumes of pre-cooled absolute ethanol and placed at 4 ℃ for at least 1 h. The precipitant was collected by centrifugation at 5000 g for 10 min and re-dissolved in one volume distilled water. Two volumes of CTAB buffer (2.5 g/L) were added to the samples and mixed gently. The OD_600nm measurement was carried out after a reaction time of 10 min at 37 ℃.

Results

Bioinformatic analysis of the R–M systems of S. zooepidemicus S12

R–M systems are ubiquitous and regarded as the primitive immune systems in bacteria. Nevertheless, in industrial strains, the systems are the major barrier to bacterial transformation by restricting foreign DNA (Murray 2002; Rego et al. 2011). Prediction of the R–M systems using the NEB website revealed that S. zooepidemicus S12 possesses five MTases and two REases (Fig. S1).

Type I R–M systems were shown to be bi-functional, multi-subunit complexes (Miller et al. 2005), which are composed of Restriction, Specificity, and Methylation subunits (Walder et al. 1981). The quaternary structure of the active type I restriction enzyme is HsdM2HsdR2HsdS (Pingoud et al. 2005). Therefore, we hypothesized that S. zooepidemicus S12 might possess one type I R–M system, which is composed of restriction endonuclease (HsdR), specificity (HsdS), and methyltransferase (HsdM) (Table 2). Most type II RM systems contain both REases and the corresponding MTases, so we hypothesized that S12 strain might possess one type II R–M system, which consists of Sesec_02360 and YhdJ (Table 2).

Table 2.

Predicted R–M system genes of S. zooepidemicus S12

R–M system type	Gene No.	Location	Gene name	Protein
Type I	SeseC_01315	1,103,371–1,106,076	hsdR	Type I restriction endonuclease subunit R
	SeseC_01317	1,107,349–1,108,176	hsdS	Type I restriction endonuclease subunit S
	SeseC_01319	1,108,455–1,110,043	hsdM	DNA methylase
Type II	SeseC_02360	1,953,490–954,101	Sesec_02360	ScaI family restriction endonuclease
	SeseC_02362	1,954,438–955,289	yhdJ	Site-specific DNA-methyltransferase

The protein sequences of the putative type I R, S, and M subunits (HsdR, HsdS, and HsdM) as well as type II REases and MTases (Sesec_02360 and YhdJ) in S. zooepidemicus S12 were aligned with the NCBI database via BLASTp. BLASTp analysis showed that the putative HsdR, HsdS, and HsdM share 99.78%, 99.27%, and 99.59% similarity, respectively, with the known HsdR family type I site-specific deoxyribonuclease, restriction endonuclease subunit S and type I restriction–modification system subunit M of Fusobacterium necrophorum. Sesec_02360, the putative type II REase, showed 96.55% and 73.63% similarity to the ScaI family restriction endonuclease of S. suis and S. oralis, respectively. YhdJ, the putative type II MTase, showed 98.94%, 91.19%, and 79.21% similarity to the site-specific DNA-methyltransferase of S. suis, S. oralis, and Lactococcus cremoris, respectively.

Construction of R–M system deficient mutants of S. zooepidemicus

To disarm the R–M systems of S. zooepidemicus, we conducted single gene deletion for hsdR, hsdM, Sesec_02360, and yhdJ (Sun et al. 2013) using the gene marker-less deletion method according to the schematic diagram in Fig. 1. The temperature-sensitive deletion plasmid pSET4s::sacB::NLR, which contains sacB as a counter selectable marker and UP-DN as up- and down-stream homologous arms for the each target gene, was introduced to the wild-type strain S12. Transformants were then cultured at 37 ℃ with spectinomycin to enforce plasmid integration. Primer pairs p1/p3 and p2/p4 were used to identify the single-crossover recombinants. Cultures of the single-crossover recombinants were diluted and plated on THY agar plate containing 5% sucrose. Double-crossover mutants obtained comprised two classes: the wild type or the target gene deleted. Primer pair p1/p4 was used to distinguish the two types based on the fact that target gene-deletion mutant results in PCR amplification products of a DNA fragment shorter than that of the wild-type (Fig. S2).

Fig. 1.

Schematic of marker-less gene deletion method in S. zooepidemicus. UP and DN represent the up- and down-stream homologous arms flanking the target gene to be deleted

As shown in Table 1, hsdR, hsdM, and Sesec_02360 were individually deleted using this method. However, deletion of yhdJ was unsuccessful under our experiment conditions for unknown reasons. Fortunately, deletion of Sesec_02360 is sufficient to destruct the type II R–M system of S12, because type II R–M systems require both REases and the corresponding MTases to perform functions (Vasu & Nagaraja 2013).

Transformation efficiency of R–M defective mutants

To examine whether the putative R–M systems influence the transformation efficiency, approximately 500 ng of the E. coli–S. zooepidemicus shuttle vector pLH67 isolated from E. coli JM109, was transformed into each of the above three mutant strains and the wild-type strain S12. After 24 h, the bacterial colonies were counted and the transformation efficiency was calculated. As shown in Fig. 2, the electroporation efficiency of the wild-type strain S12 was 0.0098%. S1518, for which the MTase encoding gene hsdM of the type I R–M system was knocked out, showed results (0.0080%) similar to that of S12. Mutant S1812, for which the REase encoding gene hsdR of the type I R–M system was deleted, showed a 67% decrease of transformation efficiency (0.0032%) compared with S12. However, the transformation efficiency of S1515, for which the type II R–M system was eliminated, reached 0.0487%, a 4.97-fold increase compared with that of S12.

Fig. 2.

a A plate view of transformants after transformation of pLH67 into the S. zooepidemicus wild type and R–M system defective mutants. THY agar supplemented with spectinomycin means 25 folds dilution for the transformants, while the THY agar with no antibiotics means 10⁶ folds dilution. b Transformation efficiency of the S. zooepidemicus wild type and R–M system defective mutants. The transformation efficiency was calculated after electroporation of the shuttle vector, pLH67. Data represent the mean values for three independent experiments

Growth and HA fermentation analysis of R–M system defective mutants

Given the importance of S. zooepidemicus in industrial HA production (de Oliveira et al. 2016), we evaluated whether destruction of the R–M systems affects cell growth or HA synthesis. The growth curve in THY medium and HA fermentation in the FSB medium for the wild-type strain and the three mutants were monitored. As shown in Fig. S3, the three R–M system gene defective mutants exhibited almost the same growth rate as the wild-type strain in THY medium. Measurements of HA production at 12 h and 24 h of fermentation were conducted. As shown in Fig. 3a, after fermentation for 12 h, hsdM deletion strain S1518 showed 17.8% higher HA titers than the wild-type S12, while SeseC_02360 and hsdR gene deletion strains S1515 and S1812 exhibited similar HA titers to S12. At the end of HA fermentation, all three mutants exhibited similar HA titers compared with S12 (Fig. 3b).

Fig.3.

Effects of individual R–M system gene deficiency on HA biosynthesis. a HA titer of the indicated strains grown in liquid FSB medium flasks for 12 h. b HA titer of the indicated strains grown in liquid FSB medium flasks for 24 h

In summary, the deficiency in two R–M systems had no obvious effects on the cell growth or HA production for S. zooepidemicus ATCC39920. Therefore, the strain engineered with targeted R–M system modifications in this study is suitable for constructing more efficient strains for industrial HA production.

Discussion

To expand the application of S. zooepidemicus in industry, development of efficient and stable transformation/recombination tools is critical. In this study, the roles of R–M systems in the genetic transformation of foreign DNA were explored by constructing R–M deficient mutants and evaluating the phenotypes of individual mutants.

The predicted seven R–M genes in S. zooepidemicus S12 (REBASE database: http://tools.neb.com/genomes/view.php?view_id=17133) were assigned to type I and II R–M systems (Table 2). Generally, the gene encoding a given REase is linked to the gene of its cognate MTase, and together, they form an R–M complex (Vasu and Nagaraja 2013). However, three genes were predicted to encode methylases in the genome of S12, of which the corresponding REases were not found. Further study should be pursued to uncover the additional roles of methylases in S. zooepidemicus S12.

Inactivation of the R–M system genes resulted in differential effects on genetic transformation efficiency of S. zooepidemicus with a replicative plasmid. Deletion of SeseC_02360, which encodes the REase of the type II R–M system, was shown to significantly increase its electroporation efficiency. Conversely, defects in REase of the type I R–M system decreased the efficiency. It is uncertain why deficient mutants of type I R–M system and type II R–M system showed opposite effects on transformation efficiency of S. zooepidemicus S12.

Nye et al (2019) reported that deletion of the R, S, and M subunits of the type I restriction–modification system lost all detectable m6A at the recognition sites and resulted in transformation of foreign-methylated DNA in Streptococcus pyogenes. However, in S. zooepidemicus, lack of hsdM function showed no apparent effects on the transformation efficiency of plasmids isolated from E. coli JM109. Furthermore, loss of hsdR function resulted in significant decrease in transformation efficiency compared with S12. It was speculated that pLH67, the plasmid used to assay the transformation efficiency might not possess the corresponding m6A recognition sites, which may account for hsdM deficiency did not improve the transformation efficiency. However, it could be rewarding to further explore why deletion of hsdR caused apparently decreased transformation efficiency.

Besides, based on pLH67, many gene targeting vectors of different sizes for gene manipulations were constructed. It was shown that the transformation efficiency of these vectors was comparable (data not shown). On the contrary, the transformation efficiency of pLH67 and pLH421 was significantly different, though they are similar in size (Fig. S4). pLH421 is high copy plasmid and used for gene overexpression. Transformation efficiency of pLH421 is much lower (data not shown). Therefore, we speculated that the copy number of plasmid might exhibit stronger effects for transformation efficiency than the size.

In view of that S. zooepidemicus is used as industrial cell factories, disarming the restriction–modification systems might increases the sensitivity of bacteria to bacteriophages, which can increase the risk of contamination and the associated economic losses. Fortunately, only the REase of the type II R–M system was destroyed, while the MTase could not be knocked out. This might help the cells resist most of the bacteriophages. More attention should be payed when the mutant was used for industrial production.

Prior to this study, no attempt was undertaken to activate the R–M systems of S. zooepidemicus. Our current work focusing on dissecting the roles of the type I R and M subunit and type II ScaI family restriction endonuclease by in-frame deletion resulted in the creation of type II ScaI family restriction endonuclease deficient mutant with significantly increased transformation efficiency, which can be used as a valuable tool to accelerate the development of more efficient and robust strains for industrial production HA and other fine chemicals.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

WG: designed all the experiments, analyzed the data, and wrote the manuscript. YX, GZ, and MZ: generated all the genetic constructs and conducted experiments. HL: contributed with scientific discussions and commented on the manuscript. All of the authors read and approved the final manuscript.

Data availability

The datasets generated during the current study are available from the corresponding author on reasonable request.

Code availability

Not applicable.

Declarations

Conflict of interest

Authors declare that they have no conflict of interest.

Ethical approval

This article does not contain any studies with human participants or animals.