Differences in virulence of pneumolysin and autolysin mutants constructed by insertion duplication mutagenesis and in-frame deletion in Streptococcus pneumoniae

Abstract

Background: Insertion duplication mutagenesis (IDM) and in-frame deletion (IFD) are common techniques for studying gene function, and have been applied to pneumolysin (ply), a virulence gene in Streptococcus pneumoniae (D39). Discrepancies in virulence between the two techniques were observed in both the previous and present studies. This phenomenon was also observed during mutation analysis of autolysin (lytA).

Results: Our data showed that target gene restoration (TGR) occurred in IDM mutants, even in the presence of antibiotics, while the IFD mutants were stable. In PCR result, TGR occurred later in IDM-ply and -lytA mutants cultured in non-supplemented medium (4-5 h) compared with those grown in medium supplemented with erythromycin (erm)/chloramphenicol (cat) (3-4 h), but plateaued faster. Real-time PCR for detecting TGR had been performed. When compared with 8-h culture, TGR detection increased from Day 1 and Day 2 of IDM mutant's culture. erm-sensitive clones from IDM mutant were found. Southern blot hybridization and Western blotting also confirmed the phenomenon of TGR. The median survival of mice following intraperitoneal (IP) injection with a 3-h culture of IDM-mutants was significantly longer than that with an 8-h culture, irrespective of antibiotic usage. The median survival time of mice following IP injection of a 3-h culture versus an 8-h culture of IDM-ply in the absence of antibiotics was 10 days versus 2 days (p = 0.031), respectively, while in the presence of erm, the median survival was 5 days versus 2.5 days (p = 0.037), respectively. For an IDM-lytA mutant, the corresponding values were 8.5 days versus 2 days (p = 0.019), respectively, for non-supplemented medium, and 2.5 versus 2 days (p = 0.021), respectively, in the presence of cat. A comparable survival rate was observed between WT D39 and an 8-h IDM culture.

Conclusion: TGR in IDM mutants should be monitored to avoid inconsistent results, and misinterpretation of data due to TGR could lead to important biological meaning being overlooked. Therefore, based on these results, IFD is preferable to IDM for disruption of target genes.

Figure 1

Construction of the ply and lytA knockdown mutants by IDM. Partial sequences of pneumolysin (ply-s) (A) and autolysin (lytA-s) (B) genes were cloned into plasmids pVA891 (A) and pEVP3 (B), containing an erythromycin (erm)- and a chloramphenicol (cat)-resistance marker, respectively. Cloned plasmids were then used to generate homologous recombinants from their WT ply or lytA genes, respectively. Two primer sets, ply-P1/pVA891-R and ply-P2/pVA891-F or lytA-P1/pEVP3-R and lytA-P2/pEVP3-F, were used to verify the insertion of pVA891 and pEVP3 in IDM-ply(A) and -lytA(B), respectively. C, ClaI.

Figure 2

Verification of D39 and the IDM-ply and -lytAmutants. (A) A product could only be amplified from D39 using primer set ply-P1/ply-P2. (B) The IDM-ply mutant was confirmed by the lack of an amplification product using the ply-P1 and ply-P2 primer set, and was shown to contain an inserted pVA891 and truncated ply using primers pairs ply-P2/pVA891-F and ply-P1/pVA891-R, respectively. (C) A product was only amplified from D39 using primer pair lytA-P1/lytA-P2. (D) The IDM-lytA mutant was verified by no amplification product using primers lytA-P1 and lytA-P2 and shown to have an inserted pEVP3 and truncated lytA using primer sets lytA-P2/pEVP3-F and lytA-P1/pEVP3-R, respectively. The locations of primers and predicted molecular weight of the PCR products for D39, IDM-ply, and IDM-lytA are given in Figure 1 and Table 1. M, molecular DNA marker.

Figure 3

Construction of the ply and lytA knockout mutants by IFD. In step 1, linear DNA fragments containing flanking regions of the ply(A) and lytA(B) genes, generated by PCR, were transformed into IDM-ply(A) and -lytA(B), respectively, for a second round of homologous recombination. The second recombination event could result in either restoration of the original gene, or mutants of IFD-ply(A) and -lytA(B) with a double-crossover. IFD-ply(A) and -lytA(B) mutants were selected by decreasing concentration of erythromycin (erm) for ply or chloramphenicol (cat) for lytA, and increasing concentrations of ampicillin. Primer sets ply-p1/ply-P2 or lytA-P1/lytA-P2 were used to verify the IFD mutants.

Figure 4

Verification of D39 and the IFD-ply and -lytA mutants. PCR using different primer sets specific to ply and lytA were used to confirm parental D39 and the IFD-ply and -lytA mutants. The locations of primers and predicted molecular weights of the PCR products for D39, IFD-ply, and IFD-lytA are given in Figure 3 and Table 1. M, molecular DNA marker.

Figure 5

Target gene restoration (TGR) of wild-type ply (A) and lytA (B) was only observed in IDM mutants. (I) Full-length target genes were amplified by PCR of samples taken at 3–10 h post-inoculation. Restoration of ply(A, I) and lytA(B, I) was confirmed by PCR using the primer sets ply-P1/ply-P2 and lytA-P1/lytA-P2, respectively, in 3–10-h cultures of IDM-ply/-lytA with and without antibiotic selection. No target gene restoration was observed in IFD-ply/-lytA mutants. (II) Housekeeping gene gki, used as a loading control to evaluate the level of gene expression, was amplified by primer pair gki-up/gki-dn. (III) Fold change refers to ratio of the target gene restoration (ply or lytA) to gki in 3–10-h cultures over the aforementioned ratio found at the earliest time of target gene detection.

Figure 6

Validation of TGR of WT-ply from IDM-ply mutant. (A) Expression of WT-PLY in the 8 h, Day 1 and Day 2 IDM-ply culture by real-time PCR (B)ClaI- digested D39, IDM-ply and AG2, which was sensitive to erm and was obtained from Day 1 culture of IDM-ply mutant, were probed with a digoxigenin-labeled ply-pb in southern blot hybridization. (C) AG2 would produce PLY whose size was the same as PLY of D39 when incubating with a rabbit polyclonal PLY antibody.

Figure 7

Detection of TGR of WTply and lytA gene from blood and liver of BALB/c mice with 3h-IDM culture of IDM mutants. (A-D) Cultural media used for IDM mutants growth. IDM-ply mutant was cultivated in plain BHI broth (A) and BHI supplemented with erm (B) whilst IDM-lytA mutant was grown in plain BHI broth (C) and BHI supplemented with cat (D). (I) Detection of the restored WT ply/lytA gene from blood and liver samples following IP injection of 3-h culture of IDM-ply/-lytA mutants at 10min, 1 day and 2 days. WT ply gene was amplified from samples by using the ply-P1/ply-P2 primer set for IDM-ply mutant grown from plain BHI (A,I) and BHI with erm (B,I). Full length of lytA gene was detected by the lytA-P1/lytA-P2 primer set for IDM-lytA mutant grown from BHI (C,I) and BHI with cat (D,I). (II) PCR analysis of IDM-ply/lytA gene from the same samples used in (I). Detection the genes of IDM-ply from IDM-ply mutant grown from BHI (A,II) and BHI with erm (B,II), and IDM–lytA gene from IDM-lytA mutant grown from BHI (C,II) and BHI with cat (D,II) was performed by ply-P1/pVA891-R and lytA-P1/pEVP3-R respectively. The lowest inocula of the 3-h IDM-ply cultures in BHI alone and BHI supplemented with erm for positive TGR of ply were approximately 10⁴ CFU (A) and 10³CFU (B) respectively whilst that of the IDM-lytA mutant grown in BHI and BHI supplemented with cat for positive TGR of lytA were 10² CFU (C) and 10³ CFU (D) respectively. B, blood; L, liver; erm, erythromycin; cat, chloramphenicol.

Figure 8

Survival of BALB/c mice injected intraperitoneally with WT D39 and a 3-h and an 8-h culture of IDM-ply (A) and -lytA (B). All of the inocula were approximately 10² CFU. The median survival time is indicated by the dashed line. Significant differences in the median survival time of mice injected with a 3-h IDM-ply(A) or -lytA(B) culture versus those for mice injected with an 8-h culture of the IDM- mutants are indicated by an asterisk (*) (*p = 0.031 (without antibiotic) versus **p = 0.037 (erm usage) in IDM-ply; ***p = 0.019 (without antibiotic) and ****p = 0.021(cat usage) in IDM-lytA). Significant differences between mice injected with 8-h WT D39 and those injected with 3-h IDM clones are indicated by a pilcrow (^¶). (^¶p < 0.001(without antibiotic) and ^¶¶p = 0.006 (erm usage) when compared to the 3-h IDM-ply; ^¶¶¶p = 0.045 (without antibiotic) when compared to the 3 h IDM-lytA).