Competing for Iron: Duplication and Amplification of the isd Locus in Staphylococcus lugdunensis HKU09-01 Provides a Competitive Advantage to Overcome Nutritional Limitation

Abstract

Staphylococcus lugdunensis is a coagulase negative bacterial pathogen that is particularly associated with severe cases of infectious endocarditis. Unique amongst the coagulase-negative staphylococci, S. lugdunensis harbors an iron regulated surface determinant locus (isd). This locus facilitates the acquisition of heme as a source of nutrient iron during infection and allows iron limitation caused by "nutritional immunity" to be overcome. The isd locus is duplicated in S. lugdunensis HKU09-01 and we show here that the duplication is intrinsically unstable and undergoes accordion-like amplification and segregation leading to extensive isd copy number variation. Amplification of the locus increased the level of expression of Isd proteins and improved binding of hemoglobin to the cell surface of S. lugdunensis. Furthermore, Isd overexpression provided an advantage when strains were competing for a limited amount of hemoglobin as the sole source of iron. Gene duplications and amplifications (GDA) are events of fundamental importance for bacterial evolution and are frequently associated with antibiotic resistance in many species. As such, GDAs are regarded as evolutionary adaptions to novel selective pressures in hostile environments pointing towards a special importance of isd for S. lugdunensis. For the first time we show an example of a GDA that involves a virulence factor of a Gram-positive pathogen and link the GDA directly to a competitive advantage when the bacteria were struggling with selective pressures mimicking "nutritional immunity".

Fig 1. Duplication of isd in HKU09-01.

(A) Schematic diagram of the ApaLI restriction sites in the single region of N920143 and the duplication in HKU09-01. Restriction sites are indicated by vertical arrows. The binding site of the DIG-labelled probe is indicated by the black dash. Predicted sizes of the fragments recognized by the probe are indicated. Primer binding sites and amplification direction (F and R) for the rapid screening for the duplication are indicated by horizontal filled arrows. (B) Results of the Southern blot. Chromosomal DNA of S. lugdunensis strains N920143 and HKU09-01 were digested with ApaLI and separated by electrophoresis. The DNA fragments were subsequently denatured, blotted onto a nylon membrane and hybridized with the DIG-labelled probe. Hybridization was detected using anti-DIG-Fab fragments conjugated to alkaline phosphatase. (C) HKU09-01 (wild-type and ΔrecA mutant) cultures were plated out and 22 colonies were screened for the presence of the duplication using primer F/R indicated in A. The frequency of loss of the duplication of seven independent cultures is shown.

Fig 2. Amplification of isd.

(A) Schematic diagram of the recombination event (RE) leading to the different numbers of fragments labelled in the Southern blot experiment. ApaLI restriction sites are indicated by vertical arrows. The binding site of the DIG-labelled probe is indicated by the black dash. Predicted sizes of the fragments recognized by the probe are indicated. (B) Results of the Southern blot. Chromosomal DNA of HKU::tetK strains (ΔrecA) with different tetracycline resistance levels were digested with ApaLI and separated by electrophoresis. The DNA fragments were subsequently denatured, blotted onto a nylon membrane and hybridized with the DIG-labelled probe. Hybridization was detected using anti-DIG-Fab fragments conjugated to alkaline phosphatase. Y1, X1, W1, W2 and Z1 designate strains with different colony sizes on 8 μg/ml Tc. (C) qPCR experiment to determine the isd copy number. Known concentrations of N920143 DNA (one copy of isd) were used to create the standard curves for isdJ and ori. Relative amounts of template ori and isdJ for each strain (all ΔrecA) were measured. The value for ori was set to 1 and the template amount of isdJ was expressed in relation to this value, thereby giving the copy number of isdJ in the chromosome of each strain. The mean and SD of three experiments is shown.

Fig 3. Isd protein expression.

Isd copy number variants (ΔrecA) were grown in RPMI with 0,5 mM bipiridyl, adjusted to OD₅₇₈ = 5 and the cell wall was digested with lysostaphin and mutanolysin in the presence of 500 mM sucrose to stabilize the protoplasts. (A,B,C)–Western Blot: Cell wall and membrane fractions were separated by SDS-PAGE and blotted onto a PVDF membrane. Isd proteins were detected using specific rabbit serum followed by goat anti-rabbit IgG DYLight 800. Fluorescent signals were quantified using Li-Core infrared technology. Absolute values measured for isd-1 were set to 100% and values obtained for the other strains were expressed in relation to this. The mean and SD of four independent experiments is shown. Statistical evaluation was performed using a paired two tailed t-test. P-values <0.05 were regarded as significant and are indicated by *. *** indicate P-values of <0.0001. IsdK was measured in the membrane fraction, IsdJ in the cell wall fraction and IsdC in the cell wall fraction. (D) Loading control: Each experiment was controlled by loading a part of the sample used for Western blotting on a second acrylamide gel. Cell wall and membrane fractions were separated by SDS-PAGE and stained with Coomassie blue. A representative gel is shown. Isd proteins predicted according to their size are indicated. MF- membrane fraction, CWF—cell wall fraction.

Fig 4. Isd surface expression.

Whole cell immunoblot: Overnight cultures of isd copy number variants (ΔrecA) were grown in RPMI, adjusted to an OD₅₇₈ = 5 and doubling dilutions were spotted on the membrane. Isd proteins were detected with specific rabbit serum followed by goat anti-rabbit IgG DYLight 800. Fluorescence intensity was measured quantitatively using Li-Core infrared detection. (A) Upper and lower panels show representative blots for IsdC and IsdB, respectively. (B) Statistical evaluation: Absolute values measured for isd-1 were set to 100% and values obtained for the other strains were expressed in relation to this. The mean and SD of five independent experiments is shown. Statistical evaluation was performed using a paired two tailed t-test. P-values of <0.05 were regarded as significant and are indicated by*. ** indicate P-values of <0.01 and *** indicate P-values of <0.0001.

Fig 5. Binding of hemoglobin to the cell surface.

Isd copy number variants (ΔrecA) were grown in RPMI with 0,5mM bipiridyl, adjusted to OD₅₇₈ = 2 and incubated with 10 μg/ml human hemoglobin (hb). After washing cell-surface bound hb was released by boiling and supernatants were separated by SDS-PAGE. Proteins were blotted onto a PVDF membrane. (A) Human hb was detected using specific rabbit serum followed by goat anti-rabbit IgG DYLight 800. Fluorescence intensity was measured quantitatively using Li-Core infrared detection. The upper panel shows a representative blot. The lower panel shows the statistical analysis. Absolute values measured for isd-1 were set to 100% and values obtained for the other strains were expressed in relation to this. The mean and SD of six Independent experiments is shown. Statistical evaluation was performed using a paired two tailed t-test. P-values of <0.05 were regarded as significant and are indicated by*. *** indicate P-values of <0.0001. (B) Each experiment was controlled by loading a part of the sample used for Western blotting on a second acrylamide gel. SDS gels were strained with Coomassie blue and an apparent non-hb band (130kDa) was chosen as a loading control. A representative gel is shown.

Fig 6. Competitive growth.

RPMI (with 10μM EDDHA) cultures were supplemented with (A) 20 μM FeSO₄, (B) 2,5 μg/ml hemoglobin or (C) 200 nM heme and inoculated with isd-1::pIPI03ery^R and one of the isd copy number variants: isd-0::pIPI03Kan^R, isd-1:: pIPI03Kan^R, isd-2:: pIPI03Kan^R, isd-3:: pIPI03Kan^R, isd-4:: pIPI03Kan^R. All strains used were RecA positive. The CFU of each strain was enumerated by plating on erythromycin and kanamycin containing agar plates. The change in the ratio of the strains was used to calculate the growth rate of each copy number variant in comparison to the single copy control strain. The percentage difference between the growth rate of the isd-1 and copy number variants is shown. The mean and SD of (A) N = 4, (B,C) N = 6 independent experiments is shown. Statistical analysis was performed using One Way Anova followed by Bonferroni’s correction. P-values of <0.05 were regarded as significant and are indicated by*. *** indicate P-values of <0.0001.

Fig 7. Hemoglobin/heme dependent growth in individual cultures.

(A-D) representative growth curves of all copy number variants (RecA positive) grown individually in RPMI (10 μM EDDHA) without additional iron (A) or in the presence of presence of 20 μM FeSO₄, 2,5 μg/ml hb or 200 nM heme (B-D). The experiment was performed in 48 well microtiter plates using an Epoch2 reader with 300 rpm shaking at 37°C. OD_600nm was determined every 15 minutes. (E) Generation times of strains calculated form growth experiments describe in (B-D). Shown is the mean and SD of six independent experiments. Statistical evaluation was performed using One Way Anova followed by Bonferroni’s correction. P-values of <0.05 were regarded as significant and are indicated by*. *** indicate P-values of <0.0001.

Fig 8. Growth in Human Serum.

Human serum was inoculated for competitive growth. (A,B) isd-1::pIPI03ery^R vs. isd-0::pIPI03Kan^R in 25% and 50% serum (C,D) isd-1::pIPI03ery^R vs. isd-2::pIPI03Kan^R in 25% and 50% serum (E,F) isd-1::pIPI03ery^R vs. isd-4::pIPI03Kan^R in 25% and 50% serum. All strains used were RecA positive. After 24 h of growth new serum was inoculated 1:10 with previous sample. The process was repeated up to three times. The CFU of each strain after each cycle of growth was enumerated by plating on erythromycin and kanamycin containing agar plates. The ratio of the two strains in each culture is given. Shown is the mean and SD of six experiments (D0-D2) and three experiments (D3-D4) respectively. For Isd-1 vs Isd-2 mean and SD of four experiments is given. The proportion of the Isd-0, isd-1 and isd-4 after each culture (D1-D4) was compared to the respective proportion at the beginning of the experiment (D0). Statistical evaluation was performed using a paired two tailed t-test. P-values of <0.05 were regarded as significant and are indicated by*. ** indicate P-values < 0,005 and *** indicate P-values of <0.0001.