ABSTRACT
The disease spectrum caused by Streptococcus dysgalactiae subsp. equisimilis resembles that of S. pyogenes (group A streptococcus [GAS]). These two bacterial species are closely related and possess many common virulence characteristics. While some GAS strains express virulence factors called streptococcal inhibitor of complement (SIC) and distantly related to SIC (DRS), some S. dysgalactiae subsp. equisimilis isolates express an orthologue of DRS, which is referred to as DRS-G. We reported previously that seropositivity for either anti-SIC or anti-DRS antibodies (Abs) is associated with poststreptococcal glomerulonephritis (PSGN). However, only seropositivity for anti-SIC Abs is associated with chronic kidney disease (CKD). We now extend the study to test whether seropositivity for anti-DRS-G Abs is also associated with these renal diseases. Stored serum samples collected for our previous study were tested by an enzyme-linked immunosorbent assay (ELISA) for Abs to DRS-G. The samples represented sera from 100 CKD adult patients, 70 adult end-stage renal disease (ESRD) patients, 25 PSGN pediatric patients, and corresponding age-matched control subjects. The proportion of PSGN, CKD, and ESRD patients who showed seroreaction to anti-DRS-G Abs was significantly higher than that of the corresponding age-matched controls, who in general exhibited seropositivity rates commensurate with the isolation rate of drsG-positive S. dysgalactiae subsp. equisimilis in the community during this study period. Since higher rates of seropositivity for anti-DRS-G Abs in the renal disease categories are resultant of previous infections with DRS-G-positive S. dysgalactiae subsp. equisimilis strains, we conclude the seropositivity is an additional risk factor for these renal diseases. In this regard, anti-DRS-G Abs have attributes similar to those of the anti-SIC Abs.
INTRODUCTION
Streptococcus dysgalactiae subsp. equisimilis was often not regarded as a human pathogen in the past, but it is now being recognized as a significant human pathogen, with an increasing frequency of reports of epidemiological observations (1–4). S. pyogenes (group A streptococcus [GAS]) and S. dysgalactiae subsp. equisimilis are associated with similar disease spectra, including the immune system-mediated postinfectious sequela poststreptococcus glomerulonephritis (PSGN) (1, 5). Comparative genomic studies revealed that these two pathogens are genetically related, with many common virulence factors (6, 7).
Among the virulence factors produced by S. dysgalactiae subsp. equisimilis is a secretory protein called DRS-G. This protein has limited primary sequence homology with SIC (streptococcal inhibitor of complement) and DRS (distantly related to SIC) in GAS (6). As might be expected, DRS-G exhibits limited functional overlap with SIC and DRS. For instance, unlike SIC, DRS-G does not inhibit complement-mediated cell lysis. In this regard, DRS-G resembles DRS. However, like DRS and SIC, DRS-G inhibits the antimicrobial peptide LL37 (8).
Several studies (9–11) have shown that positive seroprevalence for anti-SIC or anti-DRS antibodies (Abs) is associated with PSGN. PSGN in turn is an established risk factor for chronic kidney disease (CKD) and end-stage renal disease (ESRD) (12). Our recent studies in the Mumbai, India, population (13) revealed that seroprevalence for anti-SIC Abs, unlike that for anti-DRS Abs, is positively associated with the presence of CKD and ESRD. Moreover, among the anti-SIC antibody-positive patients, the prognosis of CKD is poor.
Associations between DRS-G and PSGN, CKD, and ESRD have not been studied. Here we clearly demonstrate that anti-DRS-G antibody positivity is associated with PSGN, CKD, and ESRD in the population of Mumbai, a city where streptococcal infections and diseases are endemic. We conclude that both anti-DRS-G Abs and anti-SIC Abs are positively associated with chronic renal diseases and are independent risk factors.
MATERIALS AND METHODS
Study subjects.Blood samples were collected from 25 pediatric PSGN patients, 100 CKD patients, and 70 ESRD patients. Independent age-matched healthy control sera were also collected for the PSGN (n = 25) and combined CKD/ESRD (n = 70) groups. These study subjects were the same as those described in our previous reports (9, 13). Sera were stored at −80°C until use. Informed consent was obtained from parents and subjects. Ethics Committee approval was obtained from Seth G.S. Medical College and KEM Hospital (reference EC/GOVT-4/2010). Where appropriate, as part of our previous studies (13), patients had already been assessed for renal impairment (abnormal persistent serum creatinine levels, uremic syndromes, protein urea, urine protein/creatinine ratio, and estimated glomerular filtration rates) and for anti-streptolycin O titers.
Streptococcal strains and emm typing.Community S. dysgalactiae subsp. equisimilis isolates (n = 48) were recovered from throat swabs during this study period (2012 to 2014). These isolates were emm typed per the CDC protocol (www.cdc.gov/ncidod/biotech/strep/strepblast.htm).
Purification of recombinant DRS-G.Recombinant DRS-G was prepared essentially as described by Smyth et al. (8). Briefly, DNA encoding mature DRS-G was cloned into expression vector pJ404 (DNA2.0, Menlo Park, CA). Plasmid was transformed into Escherichia coli Top10 cells and expression of recombinant protein induced by 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside). The cell pellets were lysed under nondenaturing conditions, and the recombinant DRS-G was purified using His-Trap HP 1-ml columns (GE Healthcare Life Sciences). Protein concentrations were determined using bicinchoninic acid (Thermo Scientific 23225).
Screening for the drsG gene among the S. dysgalactiae subsp. equisimilis isolates.Sicn2F and Sicc3R primers (5′-GGAGGTCACAAACTAAGCAA-3′ and 5′-TGCCTATAGAAGGCACAACT-3′, respectively) were used in PCR to amplify the drsG gene as described by Smyth et al. (8). PCR products were sent for sequencing to Macrogen, South Korea, to confirm the identity of the product.
Measurement and specificity of anti-DRS-G Abs in human serum samples.Anti-DRS-G antibody titers in serum samples (1-in-300 dilution in phosphate-buffered saline [PBS], pH 7.4) were determined by enzyme-linked immunosorbent assay (ELISA) as described by Karmarkar et al. (13). To determine seropositivity, we used a conservative cutoff value of optical density (OD) determined from the data from the corresponding healthy control group. The cutoff values were the means + 2× the standard deviations (SD) of the OD values. This approach was necessitated because of the common occurrence of streptococcal infections, particularly in regions of endemicity, possibly resulting in increased antibody titers against streptococcal antigens in the general population. Subjects showing test values greater than or equal to the cutoff value were scored as seropositive. Statistical analyses were performed as reported earlier (13). For statistical analysis, we used the unpaired t test for comparison of the disease cohort data with the data from the corresponding controls. Fisher's test was used for calculating probability.
Competitive ELISA was performed as described before (13). Briefly, after preincubation of serum samples with competitor antigens, the samples were added to plates coated with SIC, DRS, or DRS-G. The competing antigen (SIC, DRS, or DRS-G) was used with 0 μg, 10 μg, and 50 μg of the antigens. A control with no competitor was also included.
RESULTS
The rate of the drsG distribution in S. dysgalactiae subsp. equisimilis isolates recovered from the Mumbai population was similar to that of the sic or drs distribution in GAS isolates from the same population in this study period.Examination of draft genome sequences of S. dysgalactiae subsp. equisimilis, screening by PCR, and Southern blot analyses of genomic DNA of isolates from different regions (8, 14, 15) showed that the rates of drsG gene distribution could range between 12% and 17% of the rate seen with S. dysgalactiae subsp. equisimilis isolates of different emm types. In the current study, we tested for the presence of the drsG gene in 48 isolates recovered from the Mumbai region during 2 years of this study period (Table 1). These isolates belong to 24 emm types, with a recovery rate of 1 to 6 isolates per type. Of these, we found that only 4 isolates belonging to 2 emm types were positive for drsG.
emm type distribution of S. dysgalactiae subsp. equisimilis isolates and occurrence of drsG gene determined by PCR screening
We found earlier that less than 5% of GAS isolates recovered from the same community possessed the sic gene and that a similar proportion possessed the drs gene (13). The overall seroprevalence of anti-SIC or anti-DRS Abs in healthy subjects in the Mumbai community is commensurate with this isolation rate. Since the recovery rate of drsG-positive S. dysgalactiae subsp. equisimilis isolates (∼8%) is not significantly different from that of sic- or drs-positive GAS isolates in this population, we expect that a similar proportion of healthy subjects living in this community is likely to be seropositive for Abs for each of these three antigens. Indeed, only 3 to 4% of the control groups whose members had no evidence of renal diseases were seropositive for anti-DRS-G Abs (see below; Fig. 1).
Seropositivity of anti-DRS-G Abs in PSGN, CKD, and ESRD patients. Recombinant DRS-G protein was coated for plating in ELISA. Panel A shows OD values for PSGN and control 1 (n = 25 each). Panel C shows OD values for CKD (n = 100), ESRD (n = 70), and control 2 (n = 70). The dot plots show medians (the longer cross lines), the first and the third quartiles (the two shorter lines), and calculated cutoff values (mean + 2× the standard deviation [SD]) (dotted lines). Samples with values equal to or above the cutoff values were taken as positive. Panels B and D show percent seropositive samples for anti-DRS-G Abs within each group and compared with respective controls. Lines with asterisks (**, 0.04 ≤ P ≥ 0.0003; ***, P < 0.0003) indicate statistically significant differences between the means (A and C) or between the proportions of seropositives (B and D).
Anti-DRS-G Ab seroprevalence is significantly higher in renal disease patients than in the age-matched control subjects in Mumbai.Panels A and C of Fig. 1 show the spread of raw data (OD values) for all the disease categories and the corresponding controls. Statistically significant differences between the OD values for PSGN and the corresponding age-matched control and between CKD or ESRD cohorts and the corresponding age-matched control were found. In both the controls, the values for the third quartile are lower than the respective means + 2× SD, whereas 24 to 36% of serum samples from the kidney disease patients had OD values greater than this cutoff. For instance, among the 25 pediatric PSGN patients, 24% were positive for anti-DRS-G Abs (Fig. 1B). In contrast, only 4% were positive among the age-matched controls (P = 0.0380). Likewise, 27% and 36% were positive for anti-DRS-G Abs among the CKD (n = 100) and ESRD (n = 70) cohorts, respectively (Fig. 1D), whereas the proportion for the corresponding control subjects is only 3% (P = 0.0114 and 0.0002 for CKD and ESRD, respectively).
In the chronic renal disease cohorts, there are subjects who are positive for anti-SIC and anti-DRS-G Abs or anti-DRS and anti-DRS-G Abs. We have carried out competitive ELISA to show that the double seropositivity was not due to cross-reactivity of anti-DRS-G Abs with the other antigens (Fig. 2), as only homologous competitors effectively competed with the antibody binding. Moreover, as shown in Table 2, 31%, 26%, and 7% of the combined CKD-plus-ESRD cohorts were positive for Abs to a single antigen, i.e., DRS-G, SIC, and DRS, respectively. Based on these values, the expected rate of double seropositivity for anti-SIC plus anti-DRS-G Abs is estimated to be 8% (31% × 26%) and that for anti-DRS plus anti-DRS-G Abs is estimated to be 2% (31% × 7%). These expected values are similar to the corresponding observed rates (6.5% and 0.6%, respectively) (Table 2). Together, these results suggest that rate of seroprevalence for anti-DRS-G Abs is independent of past exposure to SIC or DRS antigen and also is an independent risk factor for CKD/ESRD.
Competitive ELISAs. Serum samples positive for antibodies to SIC only, DRS only, and DRS-G only were preincubated with homologous (A) or heterologous (B) antigens at 0 μg, 10 μg, and 50 μg prior to incubation. The plates were coated with the corresponding antigen for the respective seropositive sera.
Sera from chronic renal disease (CKD plus ESRD) patients seropositive for single and double antibodies (anti-SIC, anti-DRS, and anti-DRS-G Abs)
DISCUSSION
The overlap of the disease spectrum of GAS and S. dysgalactiae subsp. equisimilis and the high frequency of recovery of S. dysgalactiae subsp. equisimilis from the throats of children in some communities makes S. dysgalactiae subsp. equisimilis an important opportunistic pathogen. The two species share many common virulence factors, including M protein, streptolysin O, streptolysin S, streptokinase, C5a peptidase, DNase, fibronectin binding proteins, and NADase (6). M1 and M57 GAS express a major secretory protein called SIC; M12 and M55 GAS express the related DRS. SIC inhibits complement function, binds to various host proteins, and inhibits various antimicrobial peptides. Anti-SIC antibodies are associated with PSGN, CKD, and ESRD (9–11, 13, 16–23). Compared to SIC, DRS seems to possess restricted functions. Although it binds to several complement proteins, DRS does not inhibit their function (17). DRS, however, inhibits the antimicrobial activity of LL37. While seropositivity for anti-DRS Abs is associated with PSGN, no association was found with CKD or ESRD (13). In Table 3, functional attributes of SIC, DRS, and DRS-G and disease associations of their antibodies are summarized. DRS-G has at best only limited sequence similarity to either SIC or DRS. Despite this, like SIC and DRS, DRS-G inhibits LL37 function (8). But like DRS, DRS-G does not act as an inhibitor of complement function. We now show that, like anti-SIC Abs, anti-DRS-G Abs are associated with PSGN, CKD, and ESRD. Furthermore, our results show that these two antibodies are independent risk factors for CKD and ESRD.
Comparison of functional attributes of SIC, DRS, and DRS-G and disease associations of their antibody positivitya
Interestingly, the three antigens elicit distinct antibody responses in humans, as these antibodies do not cross-react with heterologous antigens, as shown by competition ELISA results. Nonetheless, these antibodies show partial overlapping spectra of associations with diseases, namely, pyoderma, PSGN, CKD, and ESRD (Table 3). Hence, we conclude that no single common epitope that elicits a major antibody response in humans is responsible for the common attributes of these antibodies in the renal disease association.
Because S. dysgalactiae subsp. equisimilis is recovered more often than GAS from throat swabs in schoolchildren in Mumbai (24), our current observation emphasizes the need to monitor for seropositivity for anti-DRS-G antibody in this population so that early interventions could be offered to prevent or delay progression to CKD and ESRD. As summarized in Table 2, 26% of CKD-plus-ESRD patients are seropositive for anti-SIC Abs. By inclusion of a test for anti-DRS-G Abs, the proportion of these patients positive for either of the Abs significantly increased to 50% (P = 0.0079). These results clearly suggest that anti-DRS-G Abs are independent risk factors for the chronic renal diseases. We therefore propose that, in regions of streptococcal disease endemicity, regular monitoring of children and young adults for these anti-SIC and anti-DRS-G Abs may identify the at-risk individuals, to whom educational and medical intervention could be offered to delay or avoid the onset of late-stage CKD.
Of note, the mean ages of the members of the two control cohorts in this study (pediatric and adult control subjects) are different. Nevertheless, the results of this cross-sectional study show that the seroprevalence of anti-DRS-G Abs is independent of the mean age of the members of a given population. The results suggest that exposure to these antigens early in life may be a determining factor.
As PSGN is a consequence of streptococcal infection, in our recent study (9) we tested whether anti-SIC and anti-DRS Abs are responsible for a greater predilection for GAS pyoderma. This was found to be so, and we proposed that the increased infection rate is akin to that seen with antibody-dependent enhancement (ADE) of skin infection. We could not test this feature for anti-DRS-G Abs because we did not have an S. dysgalactiae subsp. equisimilis pyoderma cohort. Be that it may, we tested for anti-DRS-G Abs in the same set of sera from pyoderma and control subjects used in the earlier study mentioned above (9). Figure 3 shows that no “cross-species” ADE was attributable to the presence of anti-DRS-G Abs.
Comparison of association of seropositivity with anti-SIC, anti-DRS, and anti-DRS-G Abs with pyoderma and the respective controls.
In summary, this report highlights the importance of monitoring for anti-DRS-G Abs among patients post-S. dysgalactiae subsp. equisimilis infection and of following up those who give positive results to identify subjects at risk for CKD.
ACKNOWLEDGMENTS
We are grateful for support from Seth G.S. Medical College & KEM Hospital, Mumbai, India, the Indian Council of Medical Research, India, and the National Health and Medical Research Council, Australia.
We thank Santosh Kaul for emm typing of S. dysgalactiae subsp. equisimilis strains.
FOOTNOTES
- Received 4 May 2015.
- Returned for modification 27 May 2015.
- Accepted 8 June 2015.
- Accepted manuscript posted online 17 June 2015.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
REFERENCES
