Distribution of the Serine-Aspartate Repeat Protein-Encoding sdr Genes among Nasal-Carriage and Invasive Staphylococcus aureus Strains

Abstract

The sdr locus was found in all 497 investigated Staphylococcus aureus strains, although in 29 strains it contained only the sdrC gene (sdrD negative, sdrE negative). The sdrC-positive, sdrD-negative, sdrE-negative gene profile was exclusive to methicillin-sensitive S. aureus (MSSA) strains (Fisher's exact test; P = 0.0005) and was not found in the strains collected from bone infections (P = 0.0019). We also found a strong association between the presence of the sdrD gene and methicillin-resistant S. aureus strains (P < 0.0001). Our findings suggest that MSSA strains with the newly uncovered sdrC-positive, sdrD-negative, sdrE-negative gene profile have a substantially decreased potential to establish bone infection.

Staphylococcus aureus is an extremely versatile and frequent pathogen of humans both in the community and in hospitals. Infections caused by this organism involve bacterial adhesion to the host extracellular matrix. S. aureus adhesins are mostly cell wall-anchored proteins and are grouped into a single family named microbial surface components recognizing adhesive matrix molecules (MSCRAMM).

The Sdr proteins are members of the MSCRAMM family that are encoded by the tandemly arrayed sdrC, sdrD, and sdrE genes, of approximately 2.8, 3.9, and 3.5 kbp, respectively, located in the sdr locus (4). The Sdr proteins are members of a family of surface proteins which are characterized by the presence of an R region containing various numbers of the Ser-Asp dipeptides encoded by DNA repeats in the 3′ region of the sdr genes. The Sdr proteins have a comparable structural organization. A signal peptide is followed by an A domain which is similar in size among the different members of the Sdr family. However, they are not closely related, with only 20 to 30% identical amino acid residues. This suggests that different Sdr proteins have different roles in S. aureus pathogenicity. However, a ligand was defined only for Bbp (bone sialo-binding protein), which is an allelic variant of SdrE (12). The Sdr proteins have two, three, or five additional 110- to 113-residue sequences (B motifs) that are tandemly repeated in SdrC, SdrE, and SdrD, respectively. The B repeats bind Ca²⁺ with high affinity, and their structure unfolds when calcium ions are removed. The function of the B domains remains unknown. The B motifs are followed by segments composed of the SD repeats (R region). The C-terminal end (region M) of the Sdr proteins is involved in anchoring the proteins to the bacterial cell wall.

At least two sdr genes are present in all tested S. aureus strains (4) and always include sdrC (8). Therefore, the lack of sdr genes must be explained by the absence of sdrD or sdrE. Peacock and colleagues (8) demonstrated a strong correlation between S. aureus invasiveness and the presence of one of the allelic variants of the sdrE gene. Moreover, Trad et al. (10) reported a significantly higher prevalence of the sdrD gene in S. aureus strains responsible for bone infections.

S. aureus multiple-locus variable-number tandem-repeat analysis (MLVA) (7, 9) utilizes polymorphism of seven individual genes (sspA, spa, sdrC, sdrD, sdrE, clfA, and clfB). During MLVA characterization of S. aureus strains, we found several strains possessing only five bands instead of six or seven in a pattern. After analysis using simplex PCRs with primer pairs designed for the MLVA method, we determined that the subset of S. aureus strains had only a single gene in the sdr locus. Similarities in the DNA sequence flanking the SD repeats of the sdr genes allowed for the selection of a single pair of primers for amplification of all three individual genes in the sdr locus. It was not possible to determine which of the sdr genes is present in the sdr locus. Therefore, we designed a novel triplex PCR procedure to examine the distribution of the sdr genes among nasal-carriage and invasive S. aureus strains as well as methicillin-sensitive S. aureus (MSSA) and methicillin-resistant S. aureus (MRSA) (Table 1).

TABLE 1.

Primers used in this study^a

Gene	Primer	Nucleotide sequence	Cycling conditionsb
sdrC	sdrCF	5′-AAAAGGCATGATACCAAATCGA	Den, 94°C/30 s; Ann, 53°C/30 s; Elon, 72°C/15 s
	sdrCR	5′-AATTCTCCATTCGTATGTTCTG
sdrD	sdrDF	5′-AGTGGGAACAGCATCAATTTTA
	sdrDR	5′-GTGGTAGATTGTACACTTTCTT
sdrE	sdrEF	5′-AGAAAGTATACTGTAGGAACTG
	sdrER	5′-GATGGTTTTGTAGTTACATCGT
Azoreductase	azoF	5′-GTGAATAATAG(C/T)CCATATGAAC	Den, 94°C/30 s; Ann, 50°C/30 s;
Putative glycosyltransferase	glyR	5′-CTACACCTGTAATTTTACTTTC	Elon, 72°C/1 min

^aThe primers were designed based on S. aureus genomes available from the Institute for Genomic Research (http://www.tigr.org) (COL strain), the University of Oklahoma (http://www.genome.ou.edu) (strain 8325), the Sanger Institute (http://www.sanger.ac.uk) (strains MRSA252 and MSSA476), the National Institute of Technology and Evaluation (http://www.bio.nite.go.jp) (strain MW2), and Juntendo University (http://www.juntendo.ac.jp) (strains Mu50 and N315).

^bDen, denaturation temperature and time; Ann, annealing temperature and time; Elon, elongation temperature and time.

A total of 497 strains (Table 2) were obtained during this study between 1990 and 2005, and these were selected from previously characterized collections (2, 6, 14). Only one isolate per person was tested. In addition, none of these strains were part of an outbreak, and none were determined to be from patients or personnel with obvious cross-infection. All strains in the collections were characterized by at least one genome typing method, including pulsed-field gel electrophoresis, multilocus sequence typing, and MLVA. The strains showed a large degree of diversity, and not a single clonal type was overrepresented (data not shown).

TABLE 2.

Distribution of sdrD and sdrE among S. aureus strain collections used in this study

Source of isolation	No. of strains	No. of strains carrying:						Geographic origin(s)	Reference(s)
		sdrD	sdrE	sdrD and sdrE	sdrD but not sdrE	sdrE but not sdrD	Neither sdrD nor sdrE
Community-acquired strains from nasal swabs of asymptomatic carriers	116 MSSA	53	99	51	2	48	15	United Kingdom	2
	1 MRSA	1	1	1	0	0	0
	Community-acquired strains from nasal swabs of asymptomatic carriers	111 MSSA	43	104	40	3	64	4	Holland	14
Hospital-acquired strains collected from blood infections	52 MSSA	23	44	21	2	23	6	Holland	14
Hospital-acquired strains derived from a variety of noninvasive human infections	12 MSSA	3	12	3	0	9	0	Bulgaria, Czech, Republic, Croatia, England, Germany, Lithuania, Poland, Russia, Slovenia, Turkey	6 and this study
	85 MRSA	85	79	79	6	0	0
	Community-acquired strains from nasal swabs of asymptomatic carriers	31 MSSA	15	27	15	0	12		4	Poland	This study
Community-acquired strains collected from osteomyelitis	60 MSSA	46	58	44	2	14	0		Poland	This study
	29 MRSA	28	21	20	8	1	0
Summary	382 MSSA	183	344	174	9	170	29
	115 MRSA	114	101	100	14	1	0
Total	497 MSSA and MRSA	297	445	274	23	171	29

DNA was extracted with the Genomic DNA Prep Plus kit (A&A Biotechnology, Gdynia, Poland) or the MagnaPure LC DNA system (DNA isolation kit III; Roche, Almere, The Netherlands) as previously described (9, 13). The triplex PCR products specific in size for sdrC (144 bp), for sdrD (272 bp), and for sdrE (433 bp) were analyzed by electrophoresis in 2% agarose SeaKem LE gels (FMC). The two-tailed Fisher exact test was used to analyze the distribution of the sdr genes among S. aureus strains originating from different hosts. P values less than 0.05 were considered a statistically significant difference.

The sdrC gene was present in all investigated strains (n = 497), which was concordant with a previous report (8). However, in 29 MSSA strains (of the total 382 MSSA strains), only the sdrC gene (sdrD negative, sdrE negative) was found in the sdr locus. The result was confirmed by PCR amplification of a sequence covering the sdr locus, using primer pair azoF and glyR (Table 1), targeting the sequences in flanking genes encoding azoreductase (assigned as SACOL0607 in the COL genome) and putative glycosyltransferase (SACOL0611). The amplicon sizes around 3.5 kbp confirmed the presence of only a single sdr gene in the sdr locus (data not shown). Two or three sdr genes were always detected in all MRSA strains (n = 115). A significant association between the sdrC-positive, sdrD-negative, sdrE-negative gene profile and MSSA strains was found (29/353 versus 0/115; Fisher's exact test; P = 0.0005). The strains with only the sdrC gene in the sdr locus represented different sequence types (STs) defined by multilocus sequence typing (Table 3). In the same STs we found strains with different combinations of the genes in the sdr locus, suggesting a high degree of variability. In contrast, the sdrD and sdrE genes were heterogeneously distributed. Among the tested strains, sdrD was significantly associated with MRSA strains (183/199 versus 114/1; Fisher's exact test; P < 0.0001), whereas the sdrE distribution did not differ between the MSSA and MRSA strains (344/38 versus 101/14; Fisher's exact test; P = 0.4898).

TABLE 3.

Distribution of the sdrC-positive, sdrD-negative, sdrE-negative gene profile among sequence types defined by multilocus sequence typing

ST (allelic profile)	No. of strains	Source of isolation
1 (1, 1, 1, 1, 1, 1, 1)	3	Nasal carriage
12 (1, 3, 1, 8, 11, 5, 11)	2	Nasal carriage (n = 1)
		Blood infection (n = 1)
25 (4, 1, 4, 1, 5, 5, 4)	2	Nasal carriage
30 (2, 2, 2, 2, 6, 3, 2)	6	Nasal carriage (n = 2)
		Blood infection (n = 4)
34 (8, 2, 2, 2, 6, 3, 2)	3	Nasal carriage
39 (2, 2, 2, 2, 2, 2, 2)	2	Nasal carriage (n = 1)
		Blood infection (n = 1)
45 (10, 14, 8, 6, 10, 3, 2)	2	Nasal carriage (n = 1)
		Blood infection (n = 1)
47 (10, 11, 8, 6, 10, 3, 2)	3	Nasal carriage
182 (18, 18, 6, 2, 13, 15, 18)	2	Nasal carriage
New-1 (3, 1, 31, 1, 29, 5, 3)	1	Nasal carriage
New-2 (3, 51, 1, 1, 4, 4, 3)	3	Nasal carriage

We checked the distribution of the sdr genes between the strains collected from nose swabs of asymptomatic carriers (n = 259) and invasive infections, osteomyelitis (n = 89), and blood infections (n = 52). The strains possessing only the sdrC gene in the sdr locus were not found in the strains collected from bone infections (23/236 versus 0/89; Fisher's exact test; P = 0.0019). This indicates strongly that the carriage strains with only the sdrC gene have abolished or substantially decreased the potential to establish bone infection. In contrast to osteomyelitis, the sdrC-positive, sdrD-negative, sdrE-negative gene profile was found in the strains isolated from blood infections (23/236 versus 6/46; Fisher's exact test; P = 0.6001). There was no significant difference between the strains associated with nasal colonization and those associated with both types of infections (P > 0.05). However, the sdrD gene was significantly associated with osteomyelitis (112/147 versus 74/15; Fisher's exact test; P < 0.0001) but not with blood infections (112/147 versus 23/29; Fisher's exact test; P = 1.0). There was no significant correlation of sdrE with blood infections (231/28 versus 44/8; Fisher's exact test; P = 0.3457) and osteomyelitis (231/28 versus 79/10; Fisher's exact test; P = 1.0).

In previous investigations a contribution to the pathogenic process of the allelic variants (sdrE/bbp) of the sdrE gene was explored. Peacock and colleagues (8) have shown a possibility that one (sdrE) of the allelic variants of the sdrE gene is associated with invasive disease while another (bbp) is evenly distributed among isolates recovered from healthy individuals and from patients with invasive S. aureus disease. Tristan and colleagues (11) have investigated only one of the allelic variants of the sdrE gene, and they have revealed that the distribution of bbp is significantly associated with osteomyelitis/arthritis. There was no difference in the incidence of bbp when carriage-associated strains were compared to invasive isolates. During our investigations we designed the primers which had sequences corresponding to regions of both allelic variants of the sdrE gene. Therefore, our results strengthen the observation of Peacock et al. that the allelic variants at a given locus may have different contributions to the pathogenic process. Our data are also concordant with the studies by Trad et al. (10) in which a significantly higher prevalence of only the sdrD gene and not the sdrE gene in bone infections has been found. The sequence alignments of the sdrC and sdrD genes (sequences obtained from the S. aureus genomes for which websites are given in Table 1) show that their polymorphism level is comparable to that of the sdrE gene (data not shown). It is very important to address future studies to determine sequence variability of the sdr genes and on the basis of their polymorphism to investigate the pathogenic potential of allelic variants of the sdrC, sdrD, and sdrE genes.

It is not clear why sdrC alone seems to be limited to MSSA strains only. This might be a reflection of the fact that MRSA primarily consists of a limited number of highly successful pandemic clones. Katayama and colleagues (5) have demonstrated that genetic background profoundly influences the stability of mecA in S. aureus. We can only hypothesize now that the same genetic mechanisms could play a role in acquisition and stability of SCCmec and the sdr genes as well as other genetic elements.

Most infections caused by S. aureus result from the combined action of a variety of factors. However, the contribution of particular virulence factors to S. aureus pathogenicity in humans is poorly understood. The results obtained in experimental models suggest that S. aureus strains producing receptors for bone sialoprotein, collagen, and fibronectin are associated with osteomyelitis and arthritis (1, 3, 12). However, our results show that strains lacking the sdrD and sdrE genes have decreased potential to infect bones. Studies in which strains with knockouts of sdrD and/or sdrE are used in experimental models should be performed to cast light on the role of these genes in osteomyelitis. Furthermore, studies to investigate the differential distribution of genes encoding virulence factors in a larger number of sdrD- and/or sdrE-positive strains and strains lacking both sdrD and sdrE are needed. The understanding of specific pathogenetic mechanisms may have an important prophylactic and therapeutic impact. The results obtained during this study suggest that both sdrD and sdrE may play comparable and important roles in bone infections. Strains lacking the sdrD gene had also potential to give rise to osteomyelitis, but then they always possessed the sdrE gene. These findings show that most strains are capable of bone infections. For adequate and cost-effective infection prevention, it is important to distinguish nasal colonizers which may be more aggressive from those with abolished or substantially decreased potential to establish an invasive infection.