OUP user menu

A dam mutant of Yersinia pestis is attenuated and induces protection against plague

Victoria L. Robinson, Petra C.F. Oyston, Richard W. Titball
DOI: http://dx.doi.org/10.1016/j.femsle.2005.09.001 251-256 First published online: 1 November 2005


We have constructed a dam mutant of Yersinia pestis GB. In BALB/c mice inoculated subcutaneously, the median lethal dose of the mutant was at least 2000-fold higher than the wild type. Mice inoculated with sub-lethal doses of the mutant were protected against a subsequent challenge with virulent Y. pestis. The effect of dam inactivation on gene expression was examined using a DNA microarray, which revealed increased expression of a number of genes associated with the SOS response. These results confirm the key role of Dam in the regulation of virulence, and its potential role as a target for the generation of attenuated strains of pathogenic bacteria.

  • DNA adenine methylase
  • Plague
  • Vaccine

1 Introduction

DNA adenine methylation plays a key role in a range of bacterial functions including the initiation of replication, the mismatch repair mechanism and the regulation of gene expression. Dam forms heritable DNA methylation patterns that do not alter the sequence of the DNA. The role of Dam in the initiation of replication is to maintain synchronicity of the origins of replication. The repair mechanism requires hemi-methylation to distinguish between the template and the new (nascent) strand [1]. Altering the levels of Dam interferes with the repair mechanism and bacteria with altered Dam activity have increased rates of mutation. Dam mutants also show increased sensitivity to UV radiation, base analogues and bile salts [2]. Dam has been shown to affect the expression of a number of genes including pilus operons in both Escherichia coli and Salmonella enterica var. Typhimurium [3].

Dam-deficient mutants have been constructed in a range of Gram negative bacteria including S. enterica var. Typhimurium [3], Yersinia pseudotuberculosis [4] and Shigella flexneri [5]. In Y. pseudotuberculosis, the ease with which the dam gene can be inactivated appears to be strain dependent, and inactivation of the gene in one strain has been reported to be a lethal event [6]. Isogenic dam mutants of all three species were attenuated and might therefore be used as live vaccines. The degree of attenuation afforded by inactivation of the dam gene also suggests that novel drugs that inhibit methylation may be useful as broad-spectrum antibiotics.

The attenuation of dam mutants has been linked to the global role that Dam plays in gene regulation. In the case of the E. coli pap operon, which directs the synthesis of pili required for adhesion to uroepithelial cells, Dam determines the binding of the regulatory protein Lrp to GATC sites upstream of the promoter [3]. A number of other E. coli fimbrial genes, such as sfa, daa and fae, are also regulated by Lrp binding [3]. In S. enterica var. Typhimurium Dam regulates the expression of at least 20 in vivo-induced genes leading to suggestions that Dam is a global regulator [reviewed in 3]. Dam-deficient mutants of S. enterica var. Typhimurium have a defect in their ability to invade non-phagocytic cells and enterocytes and they are less cytotoxic to M cells of ileal Peyer's patches [7]. Secretion of virulence proteins by the type III secretion system is enhanced in Dam-deficient mutants of S. enterica var. Typhimurium [7] and Dam overexpressing mutants of Y. pseudotuberculosis [8]. The increased secretion of virulence proteins may explain the high level of protective immunity provided by dam mutants.

In this study, we have set out to construct a dam mutant of Yersinia pestis and to determine whether this mutant might be exploited as a live attenuated vaccine against plague.

2 Materials and methods

2.1 Bacterial strains, plasmids, growth conditions and chemicals

Unless stated chemicals were purchased from Sigma–Aldrich (Poole, United Kingdom). DNA isolation and manipulations were performed using standard procedures [9] using enzymes from Promega Ltd. (Southampton, United Kingdom), Roche (United Kingdom) or Amersham Pharmacia (United Kingdom). Plasmids were introduced into E. coli by electroporation and into Y. pestis by conjugation [10] with E. coli 19851 as the donor strain. E. coli was cultured in LB broth or on LB agar at 37 °C. Y. pestis was routinely cultured at 28 °C in Blood Agar Base (BAB) broth, on BAB agar [10], or on Yersinia selective agar (YSA; Oxoid, Basingstoke, United Kingdom). When required, media were supplemented with kanamycin (50 μg ml−1), ampicillin (55 μg ml−1) and 2-aminopurine (2-AP) (200–400 μg ml−1). Y. pestis was grown for 18–22 h at 28 °C in BAB broth, serially diluted in PBS, and plated in duplicate onto BAB-haemin agar or BAB-haemin agar containing the required supplement. After incubation at 28 or 37 °C for 48 h colonies were enumerated.

2.2 Construction of a Y. pestis dam mutant

The mutant was constructed as described previously for the Y. pseudotuberculosis dam mutant [4]. The dam gene, interrupted with a kanamycin resistance cassette, was cloned into the suicide vector pCVD442, electroporated into E. coli 19851, and then transformed into Y. pestis GB by conjugation in a three-way mating with E. coli λpir pNJ5000 as a helper strain [4]. A mixture of 50 μl of each strain was spotted onto an LB agar plate and incubated at 28 °C for 5 h. The bacteria were re-suspended in 1 ml of LB broth and cultured on YSA supplemented with ampicillin and kanamycin for 48 h at 28 °C. Colonies were inoculated into BAB broth, grown overnight at 28 °C, and plated onto BAB agar supplemented with kanamycin and 10% (w/v) sucrose [10]. Colonies were screened by PCR with primers CdamF (5′-TCAGACACCCTGAAT-3′) (Starting from bp 168247 of Accession No. AL590842) and CdamR (5′-GCTTTATCAACCTGGAC-3′) (Accession No. AL590842 – 169487 bp).

2.3 Virulence in mice

The median lethal doses (MLD) of Y. pestis GB or C46 by the subcutaneous route was determined in groups of five female 6-week-old BALB/c mice (Charles River Laboratories, Margate, United Kingdom). Bacteria were grown statically for 18–22 h at 28 °C in BAB broth and mice challenged with 100 μl volumes of appropriate dilutions in PBS. Humane endpoints were strictly observed and animals deemed incapable of survival were humanely killed by cervical dislocation. The MLD was calculated using the method of Reed and Muench [11]. Six weeks after challenge with Y. pestis C46 mice were challenged subcutaneously (s.c.) with 75 or 7500 cfu of Y. pestis GB and survival was recorded for up to 2 weeks. For organ colonisation studies, spleens were removed following challenge with Y. pestis C46 or GB, homogenised in 3 ml LB broth and dilutions plated onto BAB-haemin. Bacterial colonies were enumerated after incubation at 28 °C for 48 h.

2.4 Microarray analysis

Microarray analysis was carried out as described previously [4]. Briefly, 2 ml of Y. pestis GB or C46 grown to optical density of 0.4 at 600 nm was placed into RNA protect (Qiagen) and RNA was extracted using the RNeasy kit (Qiagen). Y. pestis GB RNA was labelled with Cy3-dCTP. Y. pestis C46 RNA was labelled with Cy5-dCTP (Amersham Biosciences). After labelling with Cy3-dCTP or Cy5-dCTP the DNA was purified using a Qiagen mini-elute column. The denatured DNA was applied to pre-hybridised microarray slides and hybridised for 18 h at 65 °C. Washed slides were scanned using an Affymetrix 428 scanner (MWG Biotech), the scanned images were loaded into Imagene 5.0 (Biodiscovery) and empty spots were flagged. Quantified data were loaded into Genespring 6.0 (Silicon Genetics). The data were normalised using LOWESS, a 1.5-fold difference in expression level between the wild type and the mutant was set as the cut-off for significance in Genespring. Genes that demonstrated at least a 1.5-fold difference in expression level were tested by the Student's t test to determine that the difference was statistically significant. The Benjamini and Hochberg false discovery rate was used to correct for multiple testing. The results are a compilation of the gene expression profiles of three biological replicates grown on different days and two arrays of each replicate also performed on different days, resulting in a total of six arrays. Fully annotated microarray data has been deposited in BuGSbase (Accession No. E-BUGS-32; http://bugs.sgul.ac.uk/E-BUGS-32) and also ArrayExpress (Accession No. E-BUGS-32).

3 Results and discussion

3.1 Construction of a Y. pestis dam mutant

Analysis of the Y. pestis genome sequence revealed a coding sequence that showed 80% identity with the S. enterica serovar Typhimurium dam gene. The DNA fragment containing the dam gene was amplified in two sections and cloned into plasmid pUC18. The central portion of the gene was replaced with a kanamycin cassette and the construct was cloned into the suicide vector pCVD442. An isogenic mutant was constructed by allelic replacement using this construct. Several potential dam mutants were identified after growth on agar containing kanamycin and 10% w/v sucrose. Six of these colonies were selected for further analysis. A confirmatory PCR was used to verify the mutation, with primers CdamF and CdamR and one Y. pestis clone (C46) was selected for further study (Fig. 1).

Figure 1

Detection of Dam-defective Y. pestis GB mutants using the polymerase chain reaction. The PCR was used to screen kanamycin resistant, sucrose resistant colonies of Y. pestis GB for the presence of the inactivated dam gene. The inactivated gene is 1900 bp while the functional copy of the gene is 1240 bp. PCR products were analysed by agarose gel electrophoresis. Lanes: 1, DRIgest III (Amersham Pharmacia Biotech); 2–7, Y. pestis clones 41–46; 8, Y. pestis GB (wild type control); 9, pCdamFRK (Inactivation control); 10, DRIgest III.

3.2 Dam methylation is inactive in Y. pestis C46

Genomic DNA isolated from wild type Y. pestis GB or from Y. pestis C46 was incubated with the restriction endonucleases Mbo I, Sau 3AI or Dpn I in order to determine the methylation status of the DNA. All three enzymes recognise the sequence GATC and have previously been used to confirm the presence or absence of Dam methylation [12]. Sau 3AI cleaves at the adenine residue regardless of the state of methylation of the DNA. Mbo I only cleaves at the adenine residue if the sequence is non-methylated. Dpn I can only cleave at the adenine residue if the DNA is adenine methylated. The DNA of Y. pestis GB was cleaved by Sau 3AI and Dpn I but was resistant to digestion with Mbo I. The DNA from Y. pestis C46 was cleaved by the restriction endonucleases Mbo I and Sau 3AI but not by Dpn I (Fig. 2). These results indicate that Y. pestis GB possesses a methylation system that modifies DNA at the expected site but this methylation system is inactive in Y. pestis C46.

Figure 2

Differential digests of genomic DNA from Y. pestis GB and Y. pestis C46. Lanes: 1, DRIgest III (Amersham Pharmacia Biotech); 2, GB digested with Mbo I; 3, GB digested with Sau 3AI; 4, GB digested with Dpn I; 5, undigested genomic DNA from GB; 6, C46 digested with Mbo I; 7, C46 digested with Sau 3AI; 8, C46 digested with Dpn I; 9, undigested genomic DNA from C46; 10, DRIgest III.

3.3 The Y. pestis dam mutant is sensitive to 2-AP

The base analogue 2-AP resembles adenine with the amino group at position 2 instead of position 6 and can be converted into a dNTP that pairs with cytosine. This mispairing is poorly corrected in the absence of Dam methylation. Therefore, dam mutants often display an increase in spontaneous mutation rate, which can result in lethality when grown in the presence of 2-AP. The incorporation of the base analogue 2-AP into agar has been used previously to confirm the inactivation of the dam gene, as dam mutants show an increased sensitivity to growth inhibition by 2-AP [1315]. It has previously been demonstrated that Y. pseudotuberculosis dam mutants are not sensitive to 400 μg ml−1 of 2-AP, but Dam-deficient strains of E. coli and Serratia marcescens show sensitivity to 100 μg ml−1 [13,14]. Wild type Y. pestis GB and the dam mutant were grown in the presence of a range of concentrations of 2-AP up to 400 μg ml−1 in solid media. Y. pestis GB grew in the presence of 200 and 400 μg ml−1 2-AP at both 28 and 37 °C. The growth of the Y. pestis dam mutant was restricted in media containing 200 μg ml−1 2-AP; this was more marked at 37 °C (Fig. 3). Our results indicate that Dam has a role in the methyl-directed mismatch repair mechanism (MMR) in Y. pestis. These results are in contrast to those found for a Dam mutant of Y. pseudotuberculosis strain IP32953 [4]. Growth curves in BAB broth, carried out previously, indicated that there was no difference in growth rate between Y. pestis GB and the dam mutant (Data not shown).

Figure 3

Growth of Y. pestis GB and C46 at 28 or 37 °C in the presence of 200 μg ml−1 2-aminopurine. The data represent the mean values of three independent experiments. Error bars indicate standard deviation.

3.4 Yersinia pestis C46 is attenuated in mice

Groups of 5 BALB/c mice were challenged with strain C46 by the s.c. route of infection. In parallel, groups of mice were challenged with Y. pestis GB. The MLD of Y. pestis GB was approximately 1 cfu, while the MLD for Y. pestis C46 was 2.3 × 103 cfu. On day 20 the spleens were isolated from mice that had been dosed with 100 cfu of the Y. pestis dam mutant, homogenised and plated onto BAB-haemin agar. No bacteria were cultured after incubation of these plates.

3.5 Mice dosed with Y. pestis C46 are protected against a subsequent wild type challenge

We determined whether exposure to the Y. pestis dam mutant could induce protective immunity to plague. BALB/c mice given a single dose of approximately 1.5 × 103 cfu of Y. pestis C46 by the s.c. route were challenged six weeks later with 75 or 7500 cfu of Y. pestis GB. Naïve mice given 75 or 7500 cfu of Y. pestis GB died on days 5 and 6. None of the mice that had been immunised with the Y. pestis dam mutant died following challenge with Y. pestis GB (Table 1). However, almost 50% of the mice died at the high immunisation dose used. On day 14 post-challenge with strain GB the mice were killed by cervical dislocation and the spleens were removed and plated onto BAB-haemin agar. The spleens were free from infection with Y. pestis. The initial immunisation dose given in this study is not the suggested vaccinating dose but was used to demonstrate that the mutant is protective.

View this table:
Table 1

Survival data of mice challenged with Y. pestis GB following immunisation with 1.5 × 103 cfu of Y. pestis C46 by the s.c. route

Number of survivors/challenged
75 cfu7500 cfu
Naïve mice0/40/4
Dam immunised mice4/44/4

3.6 Microarray analysis

Analysis of the microarrays indicated that 129 genes had altered gene expression in Y. pestis compared to the dam mutant when both strains were grown under the same conditions at 37 °C (Table 2). In total 76 genes were upregulated in the dam mutant, of these 28 had an unknown function. Genes upregulated that have a known function include a number of genes encoding the SOS response, such as sulA, uvrB, recA and dinP. There is also evidence to suggest that a number of genes encoding the heat shock response, such as clpB, lon, hslV, gapA and the principal heat shock sigma factor encoded by rpoH, are also upregulated. The upregulation of the SOS response has been seen previously in Dam-deficient E. coli mutants but not in Y. pseudotuberculosis dam mutants, while upregulation of the heat shock response has been seen in both E. coli [16] and Y. pseudotuberculosis dam mutants [4]. The increased expression of the SOS response in both E. coli and Y. pestis but not in Y. pseudotuberculosis could be a reflection of the damage sustained to the DNA in absence of a fully functioning mismatch repair mechanism. In total 53 genes showed reduced expression in the dam mutant, of these 21 had an unknown function. Genes downregulated that have a known function include a number of genes involved in the metabolism and transport of galactose. The expression of the pPCP1 plasmid replication regulatory protein, rop, was also downregulated in the dam mutant, this could indicate a reduction in the copy number of pPCP1. One of the genes encoded for on pPCP1 is the plasminogen activator, Pla, a putative invasin that is required for virulence by the s.c. route of infection [17].

View this table:
Table 2

Summary of genes that demonstrated altered expression in the Y. pestis dam mutant when compared to Y. pestis GB following growth in BAB broth at 37 °C to an OD600 of 0.4

Number upregulatedNumber downregulated
Energy metabolism23
Information transfer63
Surface related312
Degradation of large molecules20
Degradation of small molecules22
Central/intermediary/miscellaneous metabolism85
Conserved hypothetical63
Pseudogenes and partial genes22
Phage and insertion elements110

There were no known virulence determinants that demonstrated reduced expression levels in the Dam-deficient Y. pestis. Therefore, the attenuation and immunity demonstrated by the Y. pestis dam mutant may be attributed to an increased visibility to the immune system due to dam-mediated alterations in gene expression as has been suggested by Heithoff et al. [2].

3.7 Conclusions

The level of attenuation seen in the Y. pestis dam mutant is similar to those seen for the intraperitoneal route in S. enterica var. Typhimurium but slightly lower than the 10,000-fold level of attenuation seen when the S. enterica var. Typhimurium dam mutant is given by the oral route [3]. The S. flexneri dam mutant was only slightly attenuated [5] while the Y. pseudotuberculosis dam mutant was attenuated a million-fold by both the oral and the intravenous routes of infection [4]. The higher level of attenuation seen with the Y. pseudotuberculosis dam mutant can be accounted for by the instability of the virulence plasmid, pYV. It is worth noting that a similar instability in the virulence plasmid did not exist in the Y. pestis dam mutant. It is possible that replication of the virulence plasmids is regulated differently in the two pathogens.

In the absence of Dam methylation Y. pestis and Y. pseudotuberculosis responded differently to the presence of 2-AP in the growth media; Y. pestis was sensitive to the presence of 2-AP while Y. pseudotuberculosis was not, this could suggest differences in the MMR between the two species. In E. coli the MMR does not function correctly in the absence of Dam methylation [18] resulting in the upregulation of the SOS response in order to repair the damage caused to the DNA [19]. In Dam-deficient Y. pestis a number of genes of the SOS response are upregulated suggesting damage to the DNA. However, in Dam-deficient Y. pseudotuberculosis there was no increase in the level of expression of genes of the SOS response suggesting that a Dam-independent repair mechanism may exist.

The level of attenuation seen in the Y. pestis dam mutant does not make it a suitable vaccine candidate. However, it may form the basis of a vaccine candidate if other deletions could be incorporated that increased the level of attenuation but maintained the level of protection.


The authors thank the Bacterial Microarray Group, based at St. Georges Medical Hospital and Dr. Karen Isherwood, Dstl, for technical assistance with the microarrays. The work was funded by the UK Ministry of Defence.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
View Abstract