Staphylococcus saprophyticus surface-associated protein (Ssp) is associated with lifespan reduction in Caenorhabditis elegans

Abstract

Staphylococcal lipases have been proposed as pathogenicity factors. In Staphylococcus saprophyticus the surface-associated protein (Ssp) has been previously characterized as a cell wall-associated true lipase. A S. saprophyticus Δssp::ermB mutant has been described as less virulent in an in vivo model of urinary tract infection compared with its wild-type. This is the first report showing that S. saprophyticus induced a lifespan reduction in Caenorhabditis elegans similar to that of S. aureus RN4220. In two S. saprophyticus Δssp::ermB mutants lifespan reduction in C. elegans was partly abolished.

In order to attribute virulence to the lipase activity itself and distinguish this phenomenon from the presence of the Ssp-protein, the conserved active site of the lipase was modified by site-directed ligase-independent mutagenesis and lipase activity-deficient mutants were constructed. These results indicate that the Ssp is associated with pathogenicity in C. elegans and one could speculate that the lipase activity itself is responsible for this virulence.

Introduction

Staphylococcus saprophyticus causes mainly urinary tract infections, especially in young female outpatients.¹ Severe systemic infections also occur rarely.^2,3 The abundant surface associated protein of S. saprophyticus (Ssp) has been characterized as a non-covalently bound, surface-associated, true lipase.^4-6 Even though some reports have been published with regard to other staphylococcal lipases and their contribution to virulence,^7-12 no data concerning a function of a lipase in the urinary tract with regard to pathogenicity has been published yet. In addition, lipids have not been described in human urine. Lipases hydrolyze triacylglycerols into free fatty acids and glycerol.

Staphylococcal lipases have been proposed as pathogenicity factors in addition to having a processing role in the lipid metabolism. True lipases have been described for Staphylococcus aureus and Staphylococcus epidermidis isolated from human facial sebaceous skin.¹¹ In previous experiments it has been shown that the lipase of S. saprophyticus (Ssp) is a surface-bound, true lipase,⁴ with similarities to lipases of S. aureus and S. epidermidis.⁵ In an in vivo model of urinary tract infection an S. saprophyticus Δssp::ermB mutant is less virulent compared with its wild-type,⁷ indicating an involvement of Ssp in the virulence of urinary tract infections due to S. saprophyticus.

Therefore the question was raised whether the lipase activity itself or the presence of the lipase protein is responsible for this previously described virulence. In uropathogenic Escherichia coli (UPEC), a lipase activity has not been described. UPECs adhere to uroepithelial cells via type 1 pili and interact with the uroplakin receptor, which is located within lipid rafts.^13-15 It has been shown that UPECs invade into the human bladder epithelial cells and also into the human bladder carcinoma cell line 5637.¹³ Interestingly, S. saprophyticus also invades this cell line.¹⁶ Nevertheless, the pathways of the interaction of the gram-positive S. saprophyticus with the eukaryotic bladder host cells is widely unknown but likely different from that of the gram-negative UPECs since structures similar to that of type 1 pili of UPECs have not been described in S. saprophyticus.

We hypothesized that Ssp might modify the composition of lipid rafts or might be part of a phagosomal/phagolysosomal escape mechanism.

Whether this virulence is associated in staphylococci to the lipase activity or to the lipase protein has not been yet described in literature. In order to attribute the virulence to the lipase function, function-deficient mutants of Ssp were constructed.

In order to analyze if the lipolytic activity of Ssp is needed for virulence or if the presence of the protein is sufficient we constructed S. saprophyticus expressing an active protein. A Caenorhabditis elegans killing model was chosen, which has been extensively described as a virulence model for a variety of pathogens and also in staphylococci other than S. saprophyticus.^17-19

In addition the suspected virulence of the Ssp shall be retested in an independent second model.

Results and Discussion

Construction of putative lipase function-deficient mutant

In order to construct lipase activity-deficient mutants, the putative active site of the lipase, the Ser₄₈₂-, Asp₆₇₃-, and His₇₁₂-catalytic triad was individually mutated. In addition a homologous ssp-knockout mutant was constructed (MB119). All three amino acids putatively involved in the active site were mutated. Using site-directed, ligase-independent mutagenesis, three lipase derivatives with mutations S₄₈₂C (MB120), D₆₇₃S (MB121), and H₇₁₂P (MB122) were cloned. The mutated ssp-genes were introduced into a lipase-deficient knockout mutant (S. saprophyticus 9325 Δssp::ermB) by protoplast transformation (Table 1).

Table 1. Bacteria and plasmids used.

Species	Strain	Properties	Source
S. aureus	Cowan I	ATCC 12589	30
	RN4220	ATCC 35556
E. coli	DH5a	F- endA1 glnV44 thi-1 recA1 relA1 gyrA96 deoR nupG Φ80dlacZΔM15 Δ(lacZYA-argF)U169, hsdR17(rK- mK+), λ–
	XL1-Blue	endA1 gyrA96(nalR) thi-1 recA1 relA1 lac glnV44 F'[::Tn10 proAB+ lacIq Δ(lacZ)M15] hsdR17(rK- mK+)	Agilent
	OP50	C. elegans feeding strain
S. saprophyticus	9325	Wild-type, clinical isolate	29
	MB119	9325 Δssp::ermB	This study
	MB120	9325 Δssp::ermB+pMB1121	This study
	MB121	9325 Δssp::ermB+pMB1122	This study
	MB122	9325 Δssp::ermB+pMB1123	This study
	7108	Wild-type, clinical isolate	29
	MB123	7108 Δssp::ermB	4
	MB124	7108 Δssp::ermB + pMB1108	4
	Plasmids
	pMB1103	Ssp gene with its own promoter	4
	pMB1106	Ssp gene interrupted by ermB used for homologous recombination	4
	pMB1108	Complete ssp gene (insert from pMB1103 cloned into vector part of pPS44)	4
	pUC18	General cloning-vector
	pRB473	General shuttle-vector	26
	pMB1109	pUC18 + sspS482C and its own promoter	This study
	pMB1110	pUC18 + sspD673S and its own promoter	This study
	pMB1111	pUC18 + sspH712P and its own promoter	This study
	pMB1121	pRB473 + sspS482C and its own promoter	This study
	pMB1122	pRB473 + sspD673S and its own promoter	This study
	pMB1123	pRB473 + sspH712P and its own promoter	This study

The S. saprophyticus wild-type strain and the mutants were tested with regard to their lipase activity by use of a previously described lipase agar assay⁴ enhanced by supplementing gum arabic. The lipase activity of the wild-type S. saprophyticus 9325 was similar to that of S. saprophyticus 7108.⁴

The lipase activity was present in the S. saprophyticus 9325 wild-type strain, abrogated in the S. saprophyticus 9325 Δssp::ermB mutant (MB119) and was also abrogated in this mutant harbouring plasmids with the ssp_S482C (MB120), ssp_D673S (MB121), or ssp_H712P (MB122) containing mutations in the catalytic triad (Fig. 1). The lipase activity was restored in the S. saprophyticus 7108 Δssp::ermB (pMB1108) complemented by the spp gene as previously shown.⁴

Figure 1. Lipase activity of the strains S. saprophyticus 7108 and 9325 and their mutants were determined using an enhanced lipase agar. S. saprophyticus 7108 has a true lipase activity as previously described. The lipase activity was lost in its Δssp::ermB mutant and restored in the complemented mutant. In the S. saprophyticus 9325 Δssp::ermB mutant the lipase activity was abolished. In this mutant harboring plasmids with the ssp_S482C, ssp_D673S, or ssp_H712P genes containing mutations in the lipase catalytic triad also did not show lipase activity. Experiment was repeated three times on different days. The opaque zone was determined in mm + SD.

Lifespan reduction of C. elegans

S. saprophyticus strain 9325, a clinical isolate of human urinary tract infection, induced a lifespan reduction in C. elegans compared with the non-pathogenic feeding strain E. coli OP50 and heat-inactivated S. aureus RN4220. The lifespan reduction was similar to that of the live S. aureus RN4220 strain (Fig. 2).

Figure 2.S. saprophyticus wild-type strain 9325 in the C. elegans killing model. S. saprophyticus wild-type strain 9325 induced a lifespan reduction in C. elegans similar to that of S. aureus RN4220. Heat-inactivated S. aureus RN4220 was largely non-pathogenic, and was relatively similar to the E. coli OP50 feeding strain.

Initially the wild-type strains of S. saprophyticus 7108 and 9325 were tested in the C. elegans model. Ours is the first study describing lifespan reduction of C. elegans by S. saprophyticus similar to that of S. aureus. In the wild-type S. saprophyticus 7108, a lifespan reduction of C. elegans was detected, but partly abolished in its Δssp::ermB mutant (Fig. 3). In addition to the results obtained with the strain 7108, lifespan was also reduced in the wild-type S. saprophyticus 9325, and partly abolished in its Δssp::ermB mutant (Fig. 4). The abolished lifespan reduction in S. saprophyticus 7108 Δssp::ermB was again restored in experiments using this previously described mutant bearing a plasmid with the wild-type ssp gene (Fig. 3).

Figure 3.S. saprophyticus wild-type strain 7108 in the C. elegans killing model. S. saprophyticus wild-type strain 7108 induced a lifespan reduction in C. elegans also similar to that of S. aureus RN4220. In the S. saprophyticus 7108 Δssp::ermB mutant the lifespan-reduction is partly abolished and restored again in the ssp-gene complemented mutant.

Figure 4. Influence of lipase activity on the lifespan of C. elegans. The lifespan in C. elegans was reduced in the S. saprophyticus wild-type strain 9325. In the lipase-deficient mutants harboring plasmids with the ssp_S482C, ssp_D673S, or ssp_H712P genes containing mutations in the lipase catalytic triad the lifespan was similar to that of the Δssp::ermB mutant and different from the wild-type (*tested by a sign test, each P = 0.01), indicating that the lipase activity itself is associated with the lifespan-reduction in C. elegans.

This finding strongly suggests that Ssp is a virulence factor in C. elegans.

Since staphylococcal lipases may possess additional activity,²⁰ such as adhesion to collagen,²¹ we wanted to test whether the lipase activity or the Ssp-protein is responsible for pathogenicity. Therefore, lipase-deficient mutants harboring plasmids with the ssp_S482C, ssp_D673S, or ssp_H712P genes containing mutations in the lipase catalytic triad were tested in a C. elegans killing model.

In all three cases (MB120, MB121, and MB122) the lifespan reduction of C. elegans was similar to that of the Δssp::ermB mutant and different from that of the wild-type (sign test, P = 0.01). One might speculate that the lipase activity of S. saprophyticus itself is responsible for virulence in C. elegans (Fig. 4). Nevertheless, a drawback of this strain set is that a functionally complemented mutant of these lipase function-deficient mutants (MB120, MB121, and MB122) is lacking.

The mechanism of the virulence in C. elegans has not yet been published in staphylococci, even though multiple factors have been suspected similar to other species. Most isolates of S. saprophyticus has been described to produce a fatty acid modifying enzyme (FAME)-activity (esterification) in addition to the lipase activity.⁹ One might speculate that the Ssp lipase activity could be involved in lipid modification, not only in the degradation, but also in the production of toxic lipids. Another hypothesis is an involvement of extracellular matrix in the lifespan reduction in C. elegans. The capsule of S. saprophyticus strain 7108 has been recently characterized and differs from that of S. aureus and S. epidermidis,²² since different cup genes were expressed. In addition, the S. saprophyticus 9325 also produces substantial amounts of extracellular matrix. In S. epidermidis the exopolysacharide biofilm has previously been shown to induce lifespan-reduction in C. elegans.²³ In addition, in S. aureus Hu et al. has shown that biofilm-formation was impaired in a lipase-knockout mutant compared with wild-type S. aureus.¹⁰ Since one might expect that an abundant and non-covalently bound surface protein, such as a lipase is involved in biofilm formation, the reduction of biofilm in the lipase knockout mutant is an expected result. These lipase function-deficient mutants, which produce the lipase protein, could be used to further investigate a putative link between staphylococcal biofilm-formation and lipase activity. Whether the biofilm-production of S. saprophyticus is also involved in lifespan-reduction in C. elegans and whether Ssp is involved in this pathway have to be proven in further experiments.

In conclusion, we first describe C. elegans lifespan reduction induced by two different wild-type strains of S. saprophyticus (9325 and 7108). The lifespan-reduction in C. elegans induced by S. saprophyticus 9325 was similar to that of S. aureus RN4220. Whether these results in C. elegans and in mice are also relevant in human urinary tract infection is speculation, since no data in humans with regard to S. saprophyticus lipase activity has been published yet. Our results indicate that the lipase activity of Ssp was abolished when the putative catalytic triad in S. saprophyticus was modified. In addition, one might speculate that the lipase activity of S. saprophyticus itself and not only the presence of the Ssp protein is responsible for virulence in C. elegans.

Materials and Methods

Bacterial strains

The strains S. saprophyticus 9325 and S. saprophyticus 7108 were collected from human urinary tract infections and have been described previously (Table 1). The species-diagnosis was confirmed using matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry²⁴ and amplification and sequencing of the soda gene,²⁵ as described previously.

Growth conditions

E. coli DH5α used for cloning was grown in Luria Bertani (LB) medium (Invitrogen) at 37 °C. Staphylococci were grown at 37 °C in LB or PY-medium (10 g L⁻¹ pepton (Oxoid), 5 g L⁻¹ yeast extract (Oxoid), 1 g L⁻¹ glucose, 5 g L⁻¹ NaCl, 1.25 g L⁻¹ Na2HPO4) unless described otherwise. Antibiotics were added at the following concentrations when appropriate: ampicillin 100 µg mL⁻¹ (for E. coli), chloramphenicol 20 µg mL⁻¹, or erythromycin 5 µg mL⁻¹.

Manipulation and extraction of DNA

DNA manipulations were done using standard methods. Restriction enzymes were obtained from Fermentas (Fermentas) and used according to the manufacturer’s instructions. Plasmid DNA from E. coli was isolated using the NucleoSpin Plasmid Kit (Macherey Nagel). Chromosomal DNA of S. saprophyticus was isolated using 20 µL of a 5 mg mL stock-solution of lysostaphin (Ambicin L, WAK-Chemie Medical GmbH) and the NucleoSpin Tissue kit (Macherey Nagel). PCR products were purified using the Extract II kit (Macherey Nagel). Ligation into the restricted plasmids pUC18 and pRB473²⁶ were performed with T4-Ligase (Life Technologies).

Construction of function-deficient ssp gene complementant strains

A knockout-mutant of the ssp gene of S. saprophyticus strain 9325 was constructed by transformation of the previously described shuttle-vector pMB1106 into the S. saprophyticus wild-type strain 9325. In this plasmid the ssp gene is interrupted by an ermB resistance gene. After homologous recombination, the temperature-sensitive plasmid pMB1106 was cured as described earlier.⁴ For the detection of the presence of the ssp gene primer ssp_specific_F and ssp_specific_R were used (Table 2).

Table 2. Primer and adaptors used in this study.

Name	(5′ to 3′)
Ssp_Ser482Cys_F	GTCGGTCATTGTATGGGAGGACAAACTGT
Adaptor 1	CCGTCAATTAG
Ssp_Ser_F	GGACAAACTGTCCGTCAATTAG
Ssp_Ser482Cys_R	TCCCATACAATGACCGACTAGATGTACTTTTT
Adaptor 2	GGCCAGGTTG
Ssp_Ser_R	TAGATGTACTTTTTGGCCAGGTTG
Ssp_Asp673Ser_F	GAGAGAAAACTCAGGACTTGTTTCTGT
Adaptor 3	AATTTCATCTCAACATCC
Ssp_Asp_F	TCTGTAATTTCATCTCAACATCC
Ssp_Asp673Ser_R	AACAAGTCCTGAGTTTTCTCTCCATT
Adaptor 4	CTTTCTCAGGTGCTTTAC
Ssp_Asp_R	CATTCTTTCTCAGGTGCTTTAC
Ssp_His712Pro_F	GGGATCCAGTCGATTTTTAGGACA
Adaptor 5	AGATAGTAAAGATACAA
Ssp_His_F	TAGGACAAGATAGTAAAGATACAA
Ssp_His712Pro_R	CAAAATCGACTGGAGTCGTGTT
Adaptor 6	TTGTTGGTGTTACT
Ssp_His_R	AGTCGTGTTTTGTTGGTGTTACT
ssp_specific_F	AGCATCAACTACAGAAACACATT
ssp_specific _R	AGCACCTTTATTTCCAACAGTTT
ssp_SacI_F	CGC GAGCTC CTCGTATGTAACCTTCTT
ssp_XbaI_R	CGC TCTAGA TAATTGAAAGATATCACCAGA
ssp_SC_F	CGTGTGTACCATTGTGTGGTGTACCT
ssp_SC_R	GGTGTTTATAAAGATTGGCAACCTGGC
ssp_DS_HP_F	AAACAACAACCTGGAGAATCTT
ssp_DS_HP_R	TTTGAGTTTGAGTGAGCGTT

Mutated region and restriction sites are shown in bold.

The S. saprophyticus 9325 Δssp::ermB mutant was verified by PCR (isolate was ssp-negative and ermB-positive) and by sequencing of the region up- and downstream to the ssp gene and then named MB119 (Table 1).

In order to analyze if the lipolytic activity of Ssp is needed for virulence or if the presence of the protein is sufficient, the conserved active sites (catalytic triad) of the lipase^21,27 was modified by site-directed ligase independent mutagenesis as described previously²⁸ using a set of four primers and two adaptors (10 pmol each) for each mutation (Table 2). The amino acids Ser₄₈₂, Asp₆₇₃, and His₇₁₂, putatively comprising the catalytic triad, were targeted and were mutated individually, not simultaneously.

Two short primers were constructed (e.g., for S₄₈₂C mutation: Ssp_Ser_F and Ssp_Ser_R), which were up to 9 base pairs from the region to be mutated. Two (long) primers (e.g., for S₄₈₂C mutation: Ssp_Ser482Cys_F Ssp_Ser482Cys_R) were used together with two adaptors (adaptor 1 and 2) bearing the desired mutation (Table 2).

As a template DNA the plasmid pMB1103 was used, where the ssp gene with its own promoter has been cloned into a pUC18.⁴ In addition, a 1:1 mixture of a Pfx and Taq-Polymerase (GE-Healthcare, Invitrogen) was included into this assay. For the inverse PCR a denaturation temperature of 94 °C for 30 sec, an annealing temperature of 61 °C for 20 sec, and an elongation temperature of 68 °C was used and repeated 25 times. In this reaction 4 PCR products of 5832 bp were obtained, two products each with one with an adaptor at one side, one product with two adaptors and one product without an adaptor (Table 2). The non-mutated DNA was restricted with DpnI, was denaturated and annealed to obtain several heteroduplex forms. The DNAs with the (complementary) adaptors were more stable and were transformed in to XL1-Blue (Stratagene, Agilent). The obtained mutants (Ssp_S482C, Ssp_D673S, and Ssp_H712P) were named pBM1109, pMB1110, and pMB1111 (Table 1) and were verified by sequencing. The active site of the lipase was then modified by replacement with distinct amino acids (Ssp_S482C, Ssp_D673S, and Ssp_H712P). The mutated ssp-genes in pUC18 were amplified with primers ssp_SacI_F and ssp_XbaI_R (Table 2), restricted and cloned with their original promoter region into the pRB473 shuttle-vector, and named pMB1121, pMB1122, and pMB1123. The derived mutants were verified by sequencing using following primers ssp_SC_F, ssp_SC_R, ssp_DS_HP_F, and ssp_DS_HP_R. Preparations of cell wall proteins (including non-covalently bound Ssp) from S. saprophyticus 9325 Δssp::ermB mutants, harboring plasmids with mutated ssp-genes (ssp_S482C, ssp_D673S, or ssp_H712P-genes each) were loaded in an SDS-PAGE gel and stained with Coomassie brilliant blue (Biorad) as previously shown⁴ and compared with purified wild-type Ssp protein. In the mutants a band similar to that of Ssp was detected, indicating that the mutated Ssp-variants are expressed.

Protoplast transformation

S. saprophyticus strain 9325 Δssp::ermB (MB119) was seeded from cryoculture on sheep blood agar and incubated for 18 h at 37 °C. Bacteria from this plate were inoculated into 5 mL PY-medium containing 5 µg mL⁻¹ erythromycin and incubated for 6 h at 39 °C. One-hundred microliters of this culture was inoculated into 20 mL PY-medium containing 5 µg mL⁻¹ erythromycin, and incubated for 18 h at 40 °C. The bacteria were harvested at 5400 × g for 10 min and then transferred into 20 mL of SOMMP²⁹ with the addition of 1 mL of 5% BSA (Invitrogen). Fifty microliters of lysostaphin of the 5 mg mL⁻¹ stock-solution was added. The cultures were incubated at 30°C and 90 rpm until the OD₆₀₀ had dropped to one-third of the initial value (between 30 and 60 min). Cells were pelleted at room temperature at 5400 × g and washed in SOMMP before they were transferred into 2 mL of SOMMP and aliquots of 300 µL were prepared. For transformation a mixture of 300 µL protoplast suspension, 3–6 µg plasmid DNA and 2 mL of PEG 6000 40% (Merck) was prepared and incubated for 2 min at room temperature. Six milliliters of SOMMP was added and the cells were harvested at room temperature at 5400 × g. The supernatant was discarded and cells were taken up in 2 mL SOMMP for plating on DM3-agar. The plates were incubated for 4 to 6 h at 30 °C, covered with CY-overlay agar containing chloramphenicol to select for transformants and incubated at 30 °C for up to 14 d.

Caenorhabditis elegans killing model

The C. elegans Bristol N2 was used in this killing model. C. elegans were propagated at 24 °C on NGM-agar with the feeding strain E. coli OP50. Ten microliters of a 1:10 diluted overnight-culture of the bacteria was plated on antibiotic containing TSB-agar plates (killing plates) and incubated at 30 °C for 12 h. The heat-inactivated and non-viable Staphylococcus aureus RN4220 was plated on the agar in a 200-fold concentration (200 µL of undiluted overnight-culture of bacteria) compared with the strains tested. The worms were synchronized and an estimated 30–50 worms in the L4 larval stage were plated on antibiotic and bacteria containing killing plates, as previously described for S. aureus,^17,18 but with the addition of 5-fluoro-2'-desoxyuridin (FUDR) (25 mg L⁻¹) as an inhibitor of worm replication. Experiments were performed with and without the addition of nalidixic acid (5 µg mL⁻¹) where appropriate in order to kill remnant feeding E. coli OP50. The non-pathogenic feeding strain E. coli OP50 was used without nalidixic acid. S. aureus RN4220 was used as a control for lifespan-reduction. Non-viable, heat-inactivated (2 h at 95 °C) S. aureus RN4220 was used in addition to E. coli OP50 to evaluate background levels of worm death.

The plates were then incubated at 24 °C and scored for living and dead worms at the beginning and daily for up to 144 h (6 d). A worm was considered to be dead when it failed to respond to plate tapping or gentle touch with a hair. Worms that died as a result of getting stuck to the wall of the plate were excluded from the analysis. The experiments was performed in duplicate and repeated seven times.

Lipase activity assay

Lipase activity was determined by an agar plate assay using 20 g L⁻¹ tributyrin agar base (Merck) and a Ca²⁺-containing 1% tributyrylglycerol solution (Merck) according to the manufacturer’s instructions. The agar plate assay was enhanced by adding 5% (w v⁻¹) gum arabic (Roth) to the tributyrylglycerol solution (Merck). Bacteria were plated onto the media and the size of the opaque zone resulting from lipase activity was determined.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.