Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
1
Functional analysis of the CpsA protein of Streptococcus agalactiae 2
3
Running title: GBS CpsA functional analysis 4
5
Brett R. Hanson1, Donna L. Runft1, Cale Streeter1, Abhin Kumar1, Thomas W. Carion2, and 6
Melody N. Neely1* 7
8
1Dept. of Immunology and Microbiology 9
Wayne State University School of Medicine 10
Detroit, MI 48201 11
12
2 Department of Biology 13
Kalamazoo College 14
Kalamazoo, MI 49006 15
16
*Corresponding author: 17
Dept. of Immunology & Microbiology 18
540 E. Canfield St. 19
Detroit, MI 48201 20
313-577-1314 21
313-577-1155 (Fax) 22
Copyright © 2012, American Society for Microbiology. All Rights Reserved.J. Bacteriol. doi:10.1128/JB.06373-11 JB Accepts, published online ahead of print on 27 January 2012
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
2
Abstract 24
Streptococcal pathogens such as Streptococcus agalactiae (GBS) are an important cause 25
of systemic disease, which is facilitated in part by the presence of a polysaccharide capsule. The 26
CpsA protein is a putative transcriptional regulator of the capsule locus, but its exact contribution 27
to regulation is unknown. To address the role of CpsA in regulation, full-length GBS CpsA and 28
two truncated forms of the protein were purified and analyzed for DNA binding ability. Assays 29
demonstrated that CpsA is able to bind specifically to two putative promoters within the capsule 30
operon with similar affinity, and full-length protein is required for specificity. Functional 31
characterization of CpsA confirmed that the ΔcpsA strain produced less capsule than wild type, 32
and demonstrated that production of full length CpsA or the DNA-binding region of CpsA 33
resulted in increased capsule levels. In contrast, production of a truncated form of CpsA lacking 34
the extracellular LytR domain (CpsA-245) in the wild type background resulted in a dominant 35
negative decrease in capsule production. GBS expressing CpsA-245, but not the ΔcpsA strain, 36
were attenuated in human whole blood. However, the ΔcpsA strain showed significant 37
attenuation in a zebrafish infection model. Furthermore, chain length was observed to be 38
variable in a CpsA-dependent manner, but could be restored to wild type levels when grown with 39
lysozyme. Taken together, these results suggest that CpsA is a modular protein influencing 40
multiple regulatory functions that may include not only capsule synthesis, but also cell wall 41
associated factors. 42
43
44
45
46
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
3
Introduction 47
Streptococcal pathogens capable of causing systemic disease utilize a number of 48
strategies for survival in the host. The most important of these is the polysaccharide capsule that 49
is produced to shield the pathogens from clearance by components of the immune system, 50
including complement deposition (18) and phagocytosis (15). The production of a 51
polysaccharide capsule has proven to be a successful strategy for a number of human-specific 52
pathogens, including the Group B Streptococcus (GBS) Streptococcus agalactiae, as well as 53
Streptococcus pneumoniae. GBS has long been a significant cause of neonatal mortality (9), and 54
long term sequelae (10), with intrapartum antibiotic prophylaxis the recommended measure to 55
combat incidence of infection (32). GBS infection of neonates remains a disease of significant 56
import in developing countries, and the preemptive use of antibiotics in the colonized mother to 57
circumvent disease is not ideal, as the development of drug resistance is a major concern (32). 58
Though rare in adults, recent work has revealed an alarming trend of increased incidence of GBS 59
infection in the United States in non-pregnant adults (26, 29), particularly in elderly patients with 60
at least one underlying health issue, illustrating that GBS remains an important problem for 61
adults as well. These observations demonstrate the need for further characterization of targets 62
for antimicrobial therapy or vaccine generation, and the unequivocal importance of the 63
polysaccharide capsule during infection makes it a prime candidate for disruption and subsequent 64
alleviation or prevention of disease. 65
The production of a polysaccharide capsule by both GBS and S. pneumoniae rely on a 66
number of shared components, with the first four genes of the capsule operon having the highest 67
conservation between the two species with greater than 60% similarity (7). These genes are 68
annotated as cpsA, cpsB, cpsC, and cpsD for both species. The genes cpsB, cpsC, and cpsD, 69
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
4
constitute a phospho-relay system that regulates polymerization and ligation of the capsular 70
polysaccharide to the cell wall peptidoglycan (3, 5, 21, 22, 34), with the encoded proteins CpsB 71
as a phospho-tyrosine protein phosphatase, CpsD as a tyrosine kinase, and CpsC as a membrane 72
tether and accessory protein for CpsD. The gene cpsA encodes a putative membrane-bound 73
transcriptional regulator of the capsule operon (11), and contains a small intracellular domain 74
and two conserved extracellular protein domains; the DNA Polymerase Processivity Factor 75
(DNA_PPF; Pfam accession no. PF02916) domain and the LytR_cpsA_psr (LytR; Pfam 76
accession no. PF03816) domain. 77
The presence of the DNA_PPF domain is curious for two reasons; first, although this 78
region of the protein is categorized as belonging to a family of sliding clamp proteins that bind 79
directly to DNA, the DNA_PPF domain of CpsA has been shown to reside extracellularly (12). 80
Second, despite this region of the protein being classified with the DNA_PPF designation, the 81
protein sequence of the DNA_PPF region of CpsA proteins diverges a great deal from traditional 82
DNA_PPF sliding clamp proteins, with a BLAST alignment of the GBS CpsA DNA_PPF 83
domain to the bacteriophage RB69 DNA_PPF domain giving no significant similarity. 84
Therefore, we propose that the DNA_PPF designation of this portion of the protein does not 85
correspond to a function consistent with sliding clamps, and that streptococcal species utilize the 86
DNA_PPF domain of CpsA for some other as yet unknown function, which has been suggested 87
previously (13). 88
In contrast to the DNA_PPF designation, the LytR designation of the CpsA protein is 89
much more robust, with a sequence alignment of the GBS CpsA LytR domain to the Bacillus 90
subtilis LytR protein’s LytR domain giving 38% identity and 58% similarity (E-value = 1e-25), 91
indicating possible functional conservation. LytR proteins have been associated with 92
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
5
transcriptional attenuation of the lytRABC divergon (16), which encodes cell-wall modifying 93
enzymes. In contrast to the transcriptional attenuation observed for LytR, CpsA appears to have 94
a transcriptional activating function for the capsule locus (7). The mechanism by which the LytR 95
domain functions remains unclear, and it may act as an environmental sensor that modulates the 96
activity of the protein, thereby controlling the cytoplasmic domain, or in the case of CpsA 97
altering the function of the DNA_PPF domain. In addition to regulation of capsule expression, 98
CpsA may prove to exhibit roles associated with cell wall regulation as well, which would be 99
consistent with the function assigned to LytR proteins. Although GBS lacks the lytABC operon 100
of Bacillus and other species, CpsA may represent a regulatory module at the crossroads of 101
capsule and the cell wall, two of the major surface components of streptococcal cells. 102
The presence of a small N-terminal cytoplasmic domain is common to both GBS CpsA 103
and B. subtilis LytR proteins with 26 and 11 amino acids respectively. We have previously 104
shown that this small cytoplasmic region attached to the transmembrane domains is sufficient for 105
S. iniae CpsA to bind to the promoter upstream of cpsA (12), and it may be that LytR proteins 106
function similarly. Both CpsA and LytR proteins contain a high density of positively charged 107
amino acids in the intracellular domains with 11/26 amino acids for CpsA and 7/11 amino acids 108
for LytR which may help facilitate interaction with specific DNA sequences. Additionally, the 109
CpsA proteins of GBS and S. pneumoniae have putative leucine zipper domains extending from 110
the cytoplasmic region into the first transmembrane region, which could also help facilitate DNA 111
binding through dimerization. LytR proteins lack this property. Although the leucine repeat is 112
present, there is no predicted coiled-coil sequence for CpsA, so the presence of a functional 113
leucine zipper domain requires validation. In addition to the operon promoter upstream of the 114
cpsA gene, a second promoter element may also exist upstream of the cpsE gene in GBS, as a 115
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
6
secondary transcription initiation site was identified in this region with upstream A-T rich 116
repeats (33). 117
This study focuses on the function of each of the CpsA protein domains as they relate to 118
the capsule locus promoters, actual capsule levels, and preliminarily, to cell wall stability. We 119
demonstrate that the GBS CpsA protein is able to bind specifically to both the GBS cpsA and 120
cpsE promoter elements, as well as define the regions of CpsA that facilitate binding to DNA 121
and contribute to the specificity of the interaction. We also demonstrate that expression of these 122
different domains in either the WT or a ΔcpsA strain of GBS alters capsule level and the capacity 123
of the bacteria to survive in human whole blood. Additionally, we present data implicating 124
CpsA in modulating the cell wall. 125
126
Materials and Methods 127
Bacterial strains and growth conditions: Plasmids were maintained in Escherichia coli 128
electro-competent Top 10 cells (Invitrogen). Luria-Bertani (LB) medium (BD) was used to 129
culture E. coli strains. Antibiotics were added as necessary to LB medium at the following 130
concentrations: chloramphenicol, 20 μg/ml, and ampicillin, 100 μg/ml for E. coli strains. E. coli 131
cultures were grown at 37°C with shaking. When growing E. coli cultures for protein 132
purification, LB media was supplemented with 0.2% glucose (w/v). Solid media was generated 133
by supplementing the liquid media with 1.4% agar (Acumedia). The streptococcal strain 134
Streptococcus agalactiae Group B Strep (GBS) 515, a human clinical isolate from the blood of a 135
patient with neonatal septicemia and GBS 515 ΔcpsA were generously provided by M. R. 136
Wessels (7). GBS 515 was cultured in Todd-Hewitt medium (BD) supplemented with 0.2% yeast 137
extract (THYB) (BD) in airtight conical tubes without agitation at 37°C. When transforming 138
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
7
GBS 515 by electroporation, bacteria were grown on solid media supplemented with 1.4% agar 139
(BD) and incubated in GasPak jars (BBL) with GasPak anaerobic system envelopes (BD). 140
Transformation of GBS 515. GBS 515 cultures were grown statically in THYB 141
supplemented with 80 mM glycine overnight, diluted 1:20 in 25 mL of THYB supplemented 142
with 80 mM glycine the following day, and grown with shaking to an OD600 of 0.4. The cells 143
were then harvested by centrifugation, washed 3 times with 10 mL of ice-cold 10% glycerol, and 144
resuspended in 1 mL of ice-cold 10% glycerol. Plasmid DNA was mixed with 200 μl of cells, 145
placed into an electroporation cuvette (DOT Scientific), and electroporated with a BIO-RAD 146
Gene Pulser II at 25 μF, 2.0 kV, and 400 Ω. Cells were immediately transferred to 10 mL of 147
fresh THYB medium, and allowed to recover for 90 minutes at 37°C prior to plating on selective 148
media. 149
Cloning of maltose-binding-protein (MBP)-CpsA fusions. The full-length cpsA gene 150
was amplified from GBS 515 genomic DNA using the primers 5’ GBS-cpsA-SmaI and 3’ GBS-151
cpsA-full-stop-PstI. Truncations of the 3’ end of cpsA were amplified from GBS 515 genomic 152
DNA using the primer 5’GBS-cpsA-SmaI in conjunction with the primers 3’GBS-cpsA-245-stop-153
PstI, 3’GBS-cpsA-117-stop-PstI or 3’GBS-cpsA-39-stop-PstI. These products were digested with 154
SmaI and PstI and cloned into the corresponding SmaI and PstI sites of pMAL-c2x (NEB) 155
leading to in-frame fusions of cpsA fragments downstream of the malE gene. This generated the 156
following plasmids: pMAL-GBS-cpsA-full, pMAL-GBS-cpsA-245, pMAL-GBS-cpsA-117, and 157
pMAL-GBS-cpsA-39. These constructs were transformed into E. coli Top10 cells (Invitrogen). 158
Protein purification: Overnight pMAL-cpsA expressing E. coli strains were sub-159
cultured 1:40 into 300 mL new LB medium supplemented with 0.2% glucose and grown at 37°C 160
with shaking until reaching an OD600 of approximately 0.5, and protein expression was induced 161
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
8
by addition of 0.3 mM isopropyl-β-D-1-thiogalactopyranoside, followed by incubation for 3 162
hours. Cells were then harvested by centrifugation at 6500 × g for 10 minutes, the supernatant 163
discarded, and the cells resuspended in 30 ml of Column Buffer (20 mM Tris-HCl, 200 mM 164
NaCl, and 1mM EDTA) and stored at -20°C overnight. The frozen cultures were thawed on ice, 165
and 10 mL of lysis buffer was added (50 mM Tris-HCl, 150 mM NaCl, 1% Sarkosyl (w/v), 1% 166
Triton-X 100 (v/v), 10 mM CHAPS, pH 7.4) along with 40 μl of 100X ProteoBlock protease 167
inhibitor cocktail (Fermentas). The mixture was then sonicated in 30 second bursts to lyse cells. 168
The lysate was centrifuged at 10,000 x g for 30 minutes and the clarified lysate diluted to a total 169
volume of 100 ml using Column Buffer. The lysate was added to the top of a glass column 170
containing amylose beads (NEB) and eluted according to the manufacturer’s specifications. 171
Purified protein concentration was determined using the BCA protein assay kit (Thermo 172
Scientific) according to the manufacturer’s instructions. Protein purity was assessed by SDS-173
PAGE. 174
Generating digoxigenin-labeled DNA probe and competitor DNA probes: Probes 175
consisting of the GBS 515 cpsA promoter (217 bp) and GBS 515 cpsE promoter (221 bp) were 176
amplified from GBS 515 genomic DNA using the primers 5’ GBS-cpsA-pro with 3’ GBS-cpsA-177
pro, and 5’ GBS-cpsE-pro with 3’ GBS-cpsE-pro, respectively. The S. iniae cpsA promoter (182 178
bp) was amplified from S. iniae 9117 genomic DNA using the primers 5’iniae-cpsA-pro and 179
3’iniae-cpsA-pro. The 515 GBS promoter products were then labeled with digoxigenin using the 180
DIG Gel Shift Kit, 2nd Generation (Roche) according to manufacturer’s instructions. 181
Electromobility Shift Assays: To conduct the EMSA, constant amounts of the MBP–182
CpsA protein fusions were incubated with a constant amount of labeled probe (12 fmol) in a 183
binding buffer containing 100 mM HEPES pH 7.2, 1 mM EDTA, 50 mM KCl, 50 mM MgCl2, 1 184
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
9
mM DTT, and 30% (v/v) glycerol for 30 minutes at room temperature. For reactions with 185
competitor DNA, an excess of unlabeled GBS 515 probe DNA (either cpsA-pro or cpsE-pro) was 186
used as a specific competitor, and unlabeled S. iniae cpsA-pro was used as a non-specific 187
competitor. The samples were loaded onto a 6% polyacrylamide native gel consisting of 6% 188
(v/v) polyacrylamide, 44.5 mM tris base, 44.5 mM boric acid, and 1mM EDTA. Electrophoresis 189
was performed at 4˚C. The gel was then transferred to a nylon membrane (Santa Cruz) using a 190
semi-dry transfer apparatus (Hoefer). Chemiluminescent detection of DIG-labeled DNA on 191
membranes was accomplished with CDP-Star (Roche) according to manufacturer instructions, 192
followed by exposure to X-ray film. Each EMSA was repeated at least twice, and was also 193
repeated by using sheared salmon sperm DNA as a non-specific competitor to confirm results 194
with the S. iniae cpsA-pro non-specific competitor. Only EMSAs using the S. iniae cpsA-pro as 195
a non-specific competitor are reported in the results. 196
Cloning of MBP-cpsA fusions for complementation. The MBP-cpsA fusions MBP-197
cpsA-full, MBP-cpsA-246, and MBP-cpsA-117 were amplified from the plasmids generated 198
above using the primer 5’ MBP-RBS-BamHI in conjunction with the primers 3’ GBS-cpsA-full-199
stop-PstI, 3’ GBS-cpsA-245-stop-PstI, and 3’GBS-cpsA-117-stop-PstI respectively. These 200
fragments were then digested with BamHI and PstI and ligated into the corresponding BamHI 201
and PstI sites on the plasmid pLZ12-rofA-pro (23) behind the rofA promoter (12) creating the 202
following plasmids: pGBS-cpsA-full, pGBS-cpsA-245, and pGBS-cpsA-117. These constructs 203
were transformed into E. coli Top10 cells, propagated, and then transformed into GBS 515 as 204
described above. These GBS 515 strains are listed in Table 2. 205
Measurement of GBS 515 capsule levels using buoyant density centrifugation. 206
Buoyant density centrifugation was performed similarly to previous work (8, 17), but with 207
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
10
modification. Linear density gradients of Percoll (GE Healthcare) were generated by diluting 208
Percoll to a high density limit (1.120 g/mL) and low density limit (1.085 g/mL) with a final 209
concentration of 0.15 M NaCl according to the manufacturer’s instructions, and carefully 210
layering 2 mL of the low density solution on top of 2 mL of the high density solution in a 5 mL 211
Falcon tube (BD). These tubes were then set horizontally at a 15° angle and left overnight. The 212
next morning tubes were set upright and allowed to settle for 30 minutes prior to use. Bacterial 213
cultures were grown overnight as described above, and cultures were normalized to an OD600 of 214
0.6 in 1 mL of THYB medium, pelleted by centrifugation, resuspended in 50 μl of PBS, and 215
added directly to the top of the Percoll gradients. Tubes were then centrifuged for 30 minutes at 216
5000 rpm in a swinging bucket Eppendorf Centrifuge 5403 with Rotor 16A4-44 at room 217
temperature. Measurements were then taken from the meniscus to the bottom of the cell band in 218
each tube, and compared to a set of colored beads of known density (GE Healthcare) to 219
determine bacterial cell density. Results are reported as an average of three separate 220
experiments. When appropriate, the Student’s t-test was used to evaluate the difference in means 221
between groups. 222
Incubation of GBS 515 in human whole blood. Human blood was obtained from 223
healthy volunteers and collected in heparinized vacuum tubes. Bacteria were grown overnight 224
and normalized to an OD600 of 0.3 in PBS, corresponding to ~1 x 108 CFU/ml. Bacteria were 225
serially diluted in PBS, and 2.5 x 105 CFU were added to 1 mL of blood and incubated with 226
rotation at 37°C for 3 hours. After incubation, 100 μl of the inoculated blood was plated onto 227
selective media, and incubated overnight at 37°C in a CO2 incubator to determine colony-228
forming units (CFU). Initial inoculum was serially diluted and plated as above to determine 229
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
11
actual starting CFU. Whole blood assays were repeated a minimum of three times with results 230
reported using a representative experiment. 231
Zebrafish survival assays with GBS 515. Bacteria were grown overnight, diluted 1:50 232
the following day in fresh medium, and grown to an OD600 of approximately 0.3. Cultures were 233
then normalized to 1 x 108 CFU/mL and 10 μl of culture or media alone was injected into 234
zebrafish intramuscularly, resulting in an infectious dose of 1 x 106 CFU. A total of 25 fish were 235
infected with each strain over three experiments, and zebrafish survival monitored over a 6-day 236
period, after which all surviving fish were euthanized. 237
Quantification of GBS 515 chain length. Cultures were grown overnight and 6 μl of 238
culture was directly placed onto a glass slide with a coverslip (Fisher) and viewed at 1000X 239
magnification on a Zeiss AxioSkop 40. Images were captured of randomly selected visual fields 240
using an attached Zeiss AxioCam MRc. Images were captured from two separate experiments 241
and at least 250 chains were counted from each set, for a total of 500 or more chains counted for 242
each strain. Chain length values were distributed between arbitrarily set numerical categories 243
and calculated as a percentage of all counted chains. For analysis of lysozyme treated strains, 244
cultures were grown overnight in the presence of a sub-inhibitory concentration of lysozyme 245
(200 μg/mL) and images captured as above. 246
247
Results 248
GBS CpsA binds to two separate putative promoters located in the capsule operon. 249
All members of the LytR_cpsA_psr protein family have been connected to transcriptional 250
regulation, but the precise method in which these proteins contribute to regulation has not yet 251
been elucidated. As a member of this protein family, CpsA has been identified as a potential 252
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
12
transcriptional activator of the capsule operon (7). Previous work in our lab demonstrated that 253
purified CpsA protein from S. iniae was capable of binding to the promoter region of the S. iniae 254
capsule operon upstream of the cpsA gene with specificity (12), suggesting that S. iniae CpsA 255
may modulate transcription by directly binding to promoter sequences. An alignment of the 256
cpsA promoter DNA sequences of S. iniae and GBS using BLAST reveals that these promoters 257
share no significant similarity. Additionally, despite the CpsA proteins of S. iniae and GBS 258
sharing 57 % amino acid identity and 76 % amino acid similarity, the intracellular regions of 259
CpsA responsible for binding to DNA in S. iniae share no significant similarity with the same 260
regions on the GBS form of the protein. These differences in sequence at the DNA and protein 261
level necessitate confirmation of DNA binding by the GBS form of CpsA. Therefore, 262
electromobility shift assays (EMSA) were performed using labeled probes with the sequence of 263
two putative promoter regions within the GBS capsule operon, upstream of the transcriptional 264
start sites of the cpsA gene and cpsE gene (Fig. 1A) as described previously (33). To determine 265
the importance of different regions of the GBS CpsA protein in binding to DNA, multiple 266
maltose binding protein (MBP) fusions of the CpsA protein were constructed and purified 267
including the full-length protein (MBP-CpsA-full) as well as two truncated forms of the protein, 268
MBP-CpsA-117 and MBP-CpsA-39 (Fig. 1C). 269
Purified MBP-CpsA-full was incubated with either the DIG-labeled GBS cpsA promoter 270
or the DIG-labeled GBS cpsE promoter. MBP-CpsA-full demonstrated the ability to bind to the 271
labeled cpsA promoter with specificity (Fig. 2A). Lane 1 demonstrates the migration of unbound 272
labeled cpsA-pro probe, lanes 2 and 3 demonstrate labeled probe bound by the protein in the 273
presence of 25 fold specific or non-specific unlabeled competitor DNA, lanes 4 and 5 274
demonstrate labeled probe bound by the protein in the presence of 50 fold specific or non-275
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
13
specific unlabeled competitor DNA, and lanes 6 and 7 demonstrate labeled probe bound by the 276
protein in the presence of 75 fold specific or non-specific unlabeled competitor DNA. In the 277
same manner, using the same concentration of protein and probe, the full length GBS CpsA 278
protein demonstrated specific binding to the labeled cpsE promoter as well (Fig. 2B). 279
After observing specific binding of MBP-CpsA-full to both the labeled GBS cpsA and 280
GBS cpsE promoters, determination of binding preference to one of these promoters over the 281
other was analyzed by cross-competing each labeled probe with unlabeled probe (Fig 3A). As 282
shown above, full-length CpsA protein binds to the cspA-pro and cpsE pro probes (Fig 3A, lanes 283
2 and 7). Competition with an excess of unlabeled probe of either cpsA-pro (lanes 3 and 9) or 284
cpsE-pro (lanes 4 and 8), revealed that the full length CpsA protein was able to bind both labeled 285
promoters and did not demonstrate a clear preference for either cpsA or cpsE probe to the 286
exclusion of the other. Again, these interactions were specific as a 50-fold excess of unlabeled S. 287
iniae cpsA-pro (nonspecific competitor) showed no competition (Fig 3A, lanes 6 and 10). 288
To determine what regions of the protein were required for DNA binding, EMSAs were 289
performed using both truncated forms of the GBS CpsA protein, MBP-CpsA-117 and MBP-290
CpsA-39. MBP-CpsA-117 is a truncation of CpsA after the third transmembrane domain (see 291
Fig 1C), thereby removing the large extracellular region to assess its contribution to binding or 292
specificity. When full length GBS CpsA was replaced with the truncation MBP-CpsA-117 (Fig. 293
3B), the protein was still able to bind both labeled probes, and no clear preference for either the 294
labeled cpsA-pro or cpsE-pro probes was observed when cross-competed, as seen for the full 295
length CpsA. However, a reduction in specificity was observed for both the cpsA and cpsE 296
labeled probes when comparing competition with unlabeled specific and non-specific DNA, 297
though some level of specificity was still present. This indicated that the large extracellular 298
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
14
portion of CpsA is not required for binding to DNA, but does affect the specificity with which 299
CpsA is able to interact with a specific DNA sequence. 300
The MBP-CpsA-39 construct produces the CpsA protein truncated at the end of the first 301
transmembrane domain, leaving the putative leucine zipper domain intact, but removing the 302
cytoplasmic loop between the second and third transmembrane domains (see Fig 1C). Thus, an 303
EMSA using this protein fusion assessed the contribution of the cytoplasmic loop to binding 304
ability and specificity. When the MBP-CpsA-39 truncation was used with the same parameters 305
(Fig. 3C), the protein retained the ability to bind to both labeled probes, but lost all semblance of 306
specificity for either the cpsA-pro or cpsE-pro when comparing competition for unlabeled 307
specific and non-specific DNA. This confirmed that the cytoplasmic loop contributes to 308
specificity, but is not required for binding ability. Importantly, previous work with the CpsA 309
protein from S. iniae demonstrated that the MBP domain does not contribute to DNA binding 310
(12). Taken together, these results demonstrate that only the cytoplasmic N-terminus of GBS 311
CpsA is required for binding to DNA, but that the cytoplasmic loop and extracellular region of 312
the protein both contribute to specificity of the interaction. 313
Ectopic expression of CpsA affects GBS capsule level. Deletion of the cpsA gene has 314
previously been associated with decreased capsule production (7). To assess the contribution of 315
each domain of the CpsA protein to capsule production, full length and truncated forms of CpsA 316
(Fig. 1C) were constructed and placed on the plasmid pLZ12-rofA-pro (23), providing 317
constitutive expression. These plasmids were then transformed into either the WT GBS 515 318
strain or a GBS 515 cpsA in-frame deletion strain (ΔcpsA) (7). A growth curve of all strains 319
grown in THYB demonstrated that all strains grew similarly with no significant differences (S1). 320
Buoyant density centrifugation was used to determine relative differences in the level of capsule 321
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
15
produced by the strains created as described above. Measurement of capsule demonstrated that 322
the ΔcpsA strain produced less capsule than the WT strain when both strains harbored the vector 323
alone (Fig. 4). Ectopic expression of MBP-CpsA-full in both the WT and ΔcpsA background led 324
to an increase in capsule over that produced normally by the WT strain, (Fig. 4), demonstrating 325
that expression of the full length form of CpsA was able to complement the ΔcpsA strain. When 326
a truncated form of the protein lacking the LytR domain (MBP-CpsA-245) was expressed in 327
either background a loss of capsule was experienced to levels below that of the ΔcpsA strain 328
(Fig. 4), indicative of a possible dominant negative or repression mechanism. The addition of 329
just the N-terminal DNA-binding domain of CpsA (MBP-CpsA-117) resulted in an increase in 330
capsule for the WT strain to the same level as expression of MBP-CpsA-full and an increase in 331
capsule for the ΔcpsA strain to a slightly lesser degree (Fig. 4), showing complementation of the 332
ΔcpsA mutant with a truncated form of CpsA missing the entire extracellular domain. Overall, 333
these data suggest that specific domains of CpsA contribute to regulation of capsule production 334
in different ways. 335
GBS survival in whole blood is altered by ectopic expression of CpsA. GBS 336
virulence entails dissemination through the bloodstream, an ability that relies on inhibiting 337
phagocytic clearance, which is primarily dependent on the presence of capsule (27). The 338
variations in capsule production observed when different domains of CpsA are ectopically 339
expressed in GBS led us to question whether these variations corresponded to changes in 340
survival in human blood. Surprisingly, when incubated in human whole blood, the ΔcpsA strain 341
of GBS shows no major difference in the number of bacteria killed compared to the WT strain of 342
GBS under the conditions tested (Fig. 5), despite the presence of less capsule as measured by 343
buoyant density. However, the amount of capsule production observed when grown in laboratory 344
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
16
medium may not correlate to that produced in whole blood. Expression of MBP-CpsA-full or 345
MBP-CpsA-117 in the WT background did not significantly alter the number of bacteria killed 346
(data not shown), despite the presence of more capsule than the WT strain as measured by 347
buoyant density (Fig. 4). In contrast, expression of MBP-CpsA-full or MBP-CpsA-117 in the 348
ΔcpsA background caused an approximately 0.4 log increase in bacterial killing compared to the 349
parent strain (Fig. 5), despite the production of more capsule (Fig. 4). Expression of the LytR 350
deletion (MBP-CpsA-245) in both the WT and ΔcpsA background resulted in an approximately 351
0.6 log increase in the number of bacteria killed compared to the respective parent strain (Fig. 5), 352
which may correspond to the loss of capsule these strains exhibit as measured by buoyant density 353
shown above. 354
In the absence of CpsA, GBS virulence is attenuated in a zebrafish model of 355
infectious disease. The unexpected result that the GBS ΔcpsA strain was not attenuated in 356
human whole blood, despite the production of less capsule, led to an in-vivo assessment of 357
virulence for the GBS 515 strains using a zebrafish model of infectious disease. When zebrafish 358
were inoculated with the WT strain of GBS, only 8% of zebrafish survived to day 6 (Fig. 6). In 359
contrast, when zebrafish were inoculated with a ΔcpsA strain of GBS, 68% of fish were viable at 360
day 6 (Fig. 6). The observed decrease in virulence of the ΔcpsA strain in an in-vivo model of 361
pathogenesis, when compared to a lack of attenuation in human whole blood (Fig. 5), suggests 362
that disruption of cpsA may lead to deficiencies that are only observable in the context of 363
systemic disease. These deficiencies could be associated with synthesis of capsular 364
polysaccharide, cell wall integrity, or a combination of both. 365
CpsA affects chain length. The chain length of cocci adopted by streptococcal species 366
depends on a number of conditions, not all of which have been determined. Chain length is often 367
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
17
controlled by production of cell wall amidases or other autolysins (6), but the presence of capsule 368
has also been shown to affect chain length as well (1, 2, 19). CpsA is a member of the same 369
family of proteins as LytR, a group of proteins associated with attenuating expression of cell wall 370
modifying enzymes (16), and various domains of CpsA also influence capsule levels (Fig. 4). 371
Therefore, the effect of these associations on chain length in GBS was analyzed using 372
microscopy. Despite a small relative difference in capsule level, the ΔcpsA mutant produced 373
considerably longer chains than the WT GBS strain (Fig. 7). Complementation of the ΔcpsA 374
strain with either MBP-CpsA-full or MBP-CpsA-117, both of which increased capsule, showed a 375
shift to shorter chains, but did not fully restore the WT shorter chain phenotype (Fig. 7). 376
Furthermore, the expression of the LytR deletion construct, MBP-CpsA-245, in the WT strain, 377
which greatly decreased capsule, led to markedly longer chains than the WT strain (Fig. 7), while 378
the ΔcpsA strain expressing MBP-CpsA-245 maintained the long chain phenotype (Fig. 7). 379
To confirm the microscopy observations, the number of cells per chain for each strain 380
was calculated as described in the Materials and Methods. The results verified the prevalence of 381
short chains in the WT/vector strain with 1-2 cells per chain predominating at 70% of chains 382
(Fig. 8A) as well as a preponderance of long chains for the ΔcpsA/vector strain with greater than 383
10 cells per chain making up 45% of chains (Fig. 8B). Addition of MBP-CpsA-full to the WT 384
background did not substantially alter chain length (Fig. 8A). Addition of MBP-CpsA-full to the 385
ΔcpsA background did not fully alleviate the long chain phenotype, with greater than 10 cells and 386
3-4 cells per chain representing the largest populations at 27% and 26% respectively (Fig. 8B). 387
When the LytR deletion strain (MBP-CpsA-245) was expressed in the WT background, the long 388
chain phenotype predominated with the majority of chains showing greater than 10 cells per 389
chain (Fig. 8A), while the presence of MBP-CpsA-245 in the ΔcpsA parent strain seemed to 390
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
18
exacerbate the long chain phenotype with greater than 10 cells per chain predominating (Fig. 391
8B). Expression of MBP-CpsA-117 in the WT strain did not change the WT chain phenotype 392
(Fig. 8A). Similar to MBP-CpsA-full, expression of MBP-CpsA-117 did not fully alleviate the 393
long chain phenotype of the ΔcpsA strain, with approximately equal chains at greater than 10 394
cells per chain (27%) and 3-4 cells per chain (28%) (Fig. 8B). These results suggest that CpsA 395
either directly or indirectly influences chain length and that capsule level alone is not sufficient 396
to explain chain length variance in these circumstances. 397
To determine if cell wall related factors were primarily responsible for the observed 398
variances in chain lengths, each GBS strain above was cultured in sub-inhibitory concentrations 399
of lysozyme. Lysozyme has muramidase activity and cleaves N-acetyl-D-glucosamine residues 400
of the peptidoglycan cell wall. When grown in the presence of lysozyme, all strains existed 401
almost exclusively as diplococci or single cells (Fig. 9), indicating that CpsA-dependent changes 402
to the cell wall may be responsible for the observed chain length variances. 403
404
Discussion 405
Streptococcal pathogens capable of causing systemic disease remain a major health 406
concern worldwide, and new strategies are currently being utilized to identify and exploit novel 407
vaccine and antimicrobial targets (25, 30). The streptococcal CpsA protein is part of the 408
LytR_cpsA_psr family of proteins associated with regulatory control over cell surface 409
physiology, including polysaccharide synthesis (7), cell wall processing (6, 14, 16), and response 410
to antimicrobial stress (31). The involvement of this protein family with these important 411
virulence determinants highlights its potential as a possible target for virulence reduction, 412
increased clearance by host immune function or antimicrobial therapy. Therefore, our aim was 413
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
19
to characterize the functional properties of CpsA to better understand the role it may play during 414
initiation and perpetuation of disease. 415
The streptococcal CpsA protein has been identified as a putative regulatory activator of 416
capsule, with an in-frame deletion of cpsA resulting in reduced levels of transcript from the 417
capsule operon as well as a concomitant loss in capsule level in GBS (7). Work on a strain of the 418
aquatic pathogen S. iniae in which the cpsA gene was insertionally inactivated demonstrated that 419
interruption of cpsA led to significant attenuation in a zebrafish model of infectious disease when 420
compared to the WT S. iniae strain, similar to what was demonstrated here with the GBS 515 421
strains (17). This evidence suggests that CpsA is required for full virulence as a positive 422
regulator of capsule, and that this may occur through direct interaction with DNA to facilitate 423
transcriptional changes. 424
CpsA proteins contain three discrete domains: a cytoplasmic N-terminal DNA-binding 425
domain, an extracellular DNA_PPF domain, and an extracellular C-terminal LytR domain (12). 426
The N-terminal DNA-binding domain is conserved in other streptococcal CpsA proteins and 427
contains a possible lecuine zipper motif, which may facilitate homo- or hetero-dimerization and 428
DNA binding ability (4). The DNA_PPF domain function is canonically ascribed to sliding 429
clamp structures that enhance the rate of DNA replication through association with DNA 430
polymerases (20, 28). However, the protein sequence of the DNA_PPF domain of CpsA 431
diverges considerably from that of traditional sliding clamp DNA_PPF domains, suggesting a 432
different function for the DNA_PPF domain in the CpsA protein. This is also supported by the 433
observation that the DNA_PPF domain of CpsA resides extracellularly where it would be unable 434
to participate in DNA replication (6). Sliding clamps that contain the DNA_PPF domain 435
typically participate in a number of protein-protein interactions that contribute to their function 436
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
20
(28), and although it appears that the DNA_PPF domain of CpsA has been functionally 437
redirected from traditional sliding clamps, it may be that the ability for facilitating protein-438
protein interactions has been retained. 439
The LytR domain of CpsA demonstrates a relatively high level of homology to traditional 440
LytR proteins of Gram-positive species with a comparison of the GBS CpsA LytR domain and 441
the B. subtilis LytR protein showing 38% identity and 58% similarity at the amino acid level. 442
LytR proteins of Gram-positive species are generally associated with regulation of cell wall 443
maintenance through transcriptional attenuation of autolysin genes, as well as their own 444
promoter (16). Analysis of the LytR protein from Streptococcus pneumoniae demonstrated that 445
LytR is required for normal septum formation during cell division (7). lytR null mutants divide 446
non-symmetrically and have highly variable cell shape and size, and with lytR mutants 447
sometimes demonstrating much larger cell size (14). Similar results were reported for the LytR 448
protein of Streptococcus mutans, with a lytR null mutant exhibiting cell division defects, 449
including the production of significantly longer chains of bacteria (6). These reports are 450
consistent with our results with the CpsA truncation in which the LytR domain has been 451
removed (MBP-CpsA-245). When this construct is expressed alone (in the ΔcpsA strain) or 452
along with the WT CpsA, consistently longer chain lengths are observed. In addition, this 453
construct produces cocci that are noticeably larger in size than the WT strain expressing full 454
length CpsA (data not shown). Increased autolysin production was also observed for the S. 455
mutans lytR mutant, suggestive of a transcriptional attenuator role for LytR over autolysin genes 456
(6). The homology between CpsA and LytR proteins indicates that some functional overlap may 457
be present, and that in addition to regulation of capsule, CpsA may also contribute to regulation 458
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
21
of cell wall maintenance. Evidence to support this hypothesis is found in Staphylococcus aureus 459
where LytR_cpsA_psR family members have been shown to exhibit functional redundancy (24). 460
Members of the LytR_cpsA_psr family of proteins have generally been associated with a 461
regulatory role at the transcriptional level. Using a series of truncated CpsA proteins to analyze 462
the contribution of different domains of the protein to DNA binding ability and specificity, we 463
found that the full length GBS CpsA protein was capable of binding to both the cpsA and cpsE 464
promoters with specificity. Furthermore, we determined that the CpsA protein had the same 465
apparent affinity for both promoters. These results suggest that direct binding of GBS CpsA to 466
both the cpsA and cpsE promoters of the capsule operon may provide a mechanism for the 467
transcriptional changes associated with deletion or interruption of the cpsA gene. The 468
observation that specificity of DNA binding decreases with sequential truncation from the C-469
terminus of the protein suggests that both the large extracellular region of the protein and the 470
cytoplasmic loop between transmembrane domains 2 and 3 likely contribute to specificity either 471
structurally or through direct interaction, much like what has been demonstrated with S. iniae 472
CpsA (12). The observed DNA binding ability for GBS MBP-CpsA-39 is in contrast to what 473
was found with S. iniae CpsA, in which truncation to the cytoplasmic N-terminus abolished 474
DNA binding ability (12). This discrepancy could be explained by the inclusion of the first 475
transmembrane domain for GBS MBP-CpsA-39 (including the putative leucine zipper), which 476
was removed from the comparable S. iniae form of the protein. 477
Functional analysis of different CpsA domains when expressed in either a WT or ΔcpsA 478
GBS background was assayed by determining capsule level with percoll buoyant density 479
gradients. While the ΔcpsA strain produced less capsule than WT, both parent strains could be 480
induced to produce higher levels of capsule when either full length CpsA (MBP-CpsA-full) or 481
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
22
just the DNA-binding domain of CpsA (MBP-CpsA-117) was ectopically expressed. These 482
results support the hypothesis that CpsA is an activator of capsule, and that it does this by 483
binding to the capsule operon promoter, as the DNA-binding domain of CpsA was sufficient for 484
complementation of capsule levels in a ΔcpsA background. In contrast, the production of CpsA 485
lacking the LytR domain (MBP-CpsA-245) resulted in a decrease in capsule for both the WT and 486
the ΔcpsA parent strains. The presence of the DNA-binding domain was not the cause of 487
decreased capsule, because as mentioned above, expression of this domain alone had an 488
activating effect. This indicated that the DNA_PPF domain was responsible for the decrease in 489
capsule levels, which could be due to a dominant negative function that is adopted by CpsA in 490
the absence of the LytR domain, perhaps through inappropriate protein-protein interactions. 491
Another possibility is that the DNA_PPF normally induces a repressing effect on capsule 492
through the DNA-binding domain by a change in protein conformation, and that the LytR 493
domain controls this repressive mechanism. With either option, because of the extracellular 494
location of the DNA_PPF domain, the repression of capsule expression is likely facilitated 495
through either a protein-protein interaction or an induced conformational change. This negative 496
regulation appears to be enacted at the transcriptional level as preliminary analysis of GBS cpsD 497
and cpsG via real time qRT-PCR revealed a ~2.5 fold decrease in transcript for the WT strain 498
ectopically expressing MBP-CpsA-245 compared to WT with vector alone (data not shown). 499
However, further analysis of these strains is necessary to fully characterize the association 500
between CpsA truncation and transcription from the capsule operon. 501
The production of a polysaccharide capsule allows GBS to evade immune clearance upon 502
introduction to the host bloodstream during infection. Therefore, the above strains were 503
incubated in human whole blood to assess how differing levels of capsule, due to manipulation 504
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
23
of CpsA, affect the ability of the bacteria to survive. Despite the production of less capsule in 505
the ΔcpsA background, we observed no major difference in survival when compared to the WT 506
strain. While production of the full length CpsA (MBP-CpsA-full) or the DNA-binding domain 507
(MBP-CpsA-117) in the WT background did not alter the ability of the WT parent to survive in 508
human blood (data not shown), when these constructs were expressed in the ΔcpsA background, 509
a decrease in survival was observed compared to the parent strain alone, despite the presence of 510
more capsule as measured by buoyant density centrifugation. Production of CpsA lacking the 511
LytR domain (MBP-CpsA-245) also led to a decrease in survival in both parent strains, which 512
may be due to reduced capsule levels or some other property of the bacteria, perhaps associated 513
with the cell wall, or a combination thereof. These results demonstrate that survival in human 514
whole blood does not always correlate with levels of capsule, and suggest that there is another 515
CpsA-dependent mechanism. Supporting this hypothesis is the observation that the ΔcpsA strain 516
was attenuated for virulence in the zebrafish model of infectious disease when compared to the 517
WT strain. Although the ΔcpsA strain was not attenuated in human whole blood, a regulatory 518
role for CpsA appears to exist in the context of systemic disease. The discrepancy between these 519
two observations may be due to an amplification of the capsule defect phenotype during systemic 520
disease and its absolute requirement, or it may be due to pleiotropic effects that are not apparent 521
during incubation in blood, but affect the ability of GBS to survive or disseminate within a host 522
organism. 523
Streptococcal LytR proteins have been associated with regulating cell division, including 524
septum formation, cell size, and chain length (6, 14). The homology of CpsA to LytR proteins 525
led us to investigate if similar regulatory events could be observed with the expression of 526
different domains of CpsA. Results from these assays indicate that capsule level does not 527
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
24
necessarily correlate with chain length, at least in GBS, as ΔcpsA strains harboring MBP-CpsA-528
full and MBP-CpsA-117 produce more capsule than WT, but retain a higher proportion of long 529
chains than the WT strain. Additionally, the observation that encapsulation leads to longer 530
chains in S. pneumoniae (1, 2), is not consistent with our observation that the ΔcpsA strain, as 531
well as both parent strains producing MBP-CpsA-245, produce less capsule than WT, yet 532
produce primarily much longer chains than the WT strain. This phenomenon was not due to 533
differences in growth, as all strains grew similarly (see S1). Therefore, CpsA appears to exert 534
direct or indirect control over chain length, suggesting a possible dual role for this protein that is 535
separate from the control of capsule expression. Consistent with this hypothesis, is the 536
observation that the WT strain expressing MBP-CpsA-245 (LytR deletion) appeared to produce 537
larger sized cells compared to other strains. However, further characterization of these strains at 538
higher magnification levels using electron microscopy would be required to better discern 539
differences in septum placement and actual cell size and shape. We propose that CpsA-540
dependent changes in cell wall maintenance are responsible for the observed differences in chain 541
length, as growth in medium with a sub-inhibitory concentration of lysozyme, which selectively 542
cleaves peptidoglycan of the cell wall, results in eradication of long chains for all strains and a 543
switch of nearly all bacteria to single cells or diplococci. 544
Taken together, the results of this study suggest that GBS CpsA is a modular protein 545
containing three functionally distinct domains. The N-terminal region of GBS CpsA is able to 546
bind to promoters within the capsule operon upstream of both the cpsA and cpsE genes, and 547
expression of the N-terminal region alone (MBP-CpsA-117) in both the WT and ΔcpsA strains is 548
sufficient to increase capsule levels, and to partially complement the ΔcpsA strain chain length 549
distribution. This seems to indicate that despite the reduction in DNA-binding specificity of 550
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
25
MBP-CpsA-117 when used alone in EMSAs, a tangible effect can still be incurred in-vivo, 551
possibly through binding of the capsule operon promoters and potentially other gene targets that 552
regulate chain length. The second module of CpsA, the DNA_PPF domain, may be responsible 553
for facilitating protein-protein interactions, a function that could be regulated by the LytR 554
domain as removal of the LytR domain results in a dominant-negative or repressive function for 555
both capsule level and chain length. To date, no mechanistic function has been applied to the 556
LytR domain in either CpsA or LytR proteins, but it appears that the LytR domain may play a 557
similar role in both proteins as similar phenotypes of increased chain length and size are 558
observed when it is removed from CpsA, both of which are seemingly independent of capsule 559
level. This suggests that CpsA may also have a role in regulation of autolysin genes, which 560
would give it a unique position at the interface of regulating the two major components of the 561
bacterial cell surface, capsular polysaccharide and the cell wall peptidoglycan. 562
Acknowledgements: 563
B.R.H. was supported by Wayne State University’s UGRD Fellowship. 564
The authors wish to thank Jennifer Dittmer for phlebotomy services and Daniel Ortiz for 565
technical assistance. 566
567
568
569
570
571
572
573
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
26
References: 574
1. Barendt, S. M., A. D. Land, L.-T. Sham, W.-L. Ng, H.-C. T. Tsui, R. J. Arnold, and 575
M. E. Winkler. 2009. Influences of Capsule on Cell Shape and Chain Formation of 576
Wild-Type and pcsB Mutants of Serotype 2 Streptococcus pneumoniae. J. 577
Bacteriol.191:3024-3040. 578
2. Barendt, S. M., L. T. Sham, and M. E. Winkler. 2011. Characterization of mutants 579
deficient in the L,D-carboxypeptidase (DacB) and WalRK (VicRK) regulon, involved in 580
peptidoglycan maturation of Streptococcus pneumoniae serotype 2 strain D39. J. 581
Bacteriol. 193:2290-2300. 582
3. Bender, M. H., R. T. Cartee, and J. Yother. 2003. Positive correlation between 583
tyrosine phosphorylation of CpsD and capsular polysaccharide production in 584
Streptococcus pneumoniae. J Bacteriol 185:6057-66. 585
4. Busch, S. J., and P. Sassone-Corsi. 1990. Dimers, leucine zippers and DNA-binding 586
domains. Trends in Genetics 6:36-40. 587
5. Byrne, J. P., J. K. Morona, J. C. Paton, and R. Morona. 2011. Identification of 588
Streptococcus pneumoniae Cps2C residues that affect capsular polysaccharide 589
polymerisation, cell wall ligation and Cps2D phosphorylation. J. Bacteriol.193:2341-6. 590
6. Chatfield, C. H., H. Koo, and R. G. Quivey, Jr. 2005. The putative autolysin regulator 591
LytR in Streptococcus mutans plays a role in cell division and is growth-phase regulated. 592
Microbiology 151:625-31. 593
7. Cieslewicz, M. J., D. L. Kasper, Y. Wang, and M. R. Wessels. 2001. Functional 594
analysis in type Ia group B Streptococcus of a cluster of genes involved in extracellular 595
polysaccharide production by diverse species of streptococci. J Biol Chem 276:139-46. 596
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
27
8. DeAngelis, P. L., and P. H. Weigel. 1994. Rapid detection of hyaluronic acid capsules 597
on group A streptococci by buoyant density centrifugation. Diagn Microbiol Infect Dis 598
20:77-80. 599
9. Dermer, P., C. Lee, J. Eggert, and B. Few. 2004. A history of neonatal group B 600
streptococcus with its related morbidity and mortality rates in the United States. Journal 601
of Pediatric Nursing 19:357. 602
10. Edwards, M. S., M. A. Rench, A. A. Haffar, M. A. Murphy, M. M. Desmond, and C. 603
J. Baker. 1985. Long-term sequelae of group B streptococcal meningitis in infants. J. 604
Pediatr. 106:717-722. 605
11. Guidolin, A., J. K. Morona, R. Morona, D. Hansman, and J. C. Paton. 1994. 606
Nucleotide sequence analysis of genes essential for capsular polysaccharide biosynthesis 607
in Streptococcus pneumoniae type 19F. Infect Immun 62:5384-96. 608
12. Hanson, B. R., B. A. Lowe, and M. N. Neely. 2011. Membrane Topology and DNA-609
Binding Ability of the Streptococcal CpsA Protein. J. Bacteriol.193:411. 610
13. Hubscher, J., L. Luthy, B. Berger-Bachi, and P. Stutzmann Meier. 2008. 611
Phylogenetic distribution and membrane topology of the LytR-CpsA-Psr protein family. 612
BMC Genomics 9:617. 613
14. Johnsborg, O., and L. S. Havarstein. 2009. Pneumococcal LytR, a Protein from the 614
LytR-CpsA-Psr Family, Is Essential for Normal Septum Formation in Streptococcus 615
pneumoniae. J. Bacteriol. 191:5859-5864. 616
15. Källman, J., J. Schollin, C. Schalèn, A. Erlandsson, and E. Kihlström. 1998. 617
Impaired phagocytosis and opsonisation towards group B streptococci in preterm 618
neonates. Archives of Disease in Childhood - Fetal and Neonatal Edition 78:F46-F50. 619
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
28
16. Lazarevic, V., P. Margot, B. Soldo, and D. Karamata. 1992. Sequencing and analysis 620
of the Bacillus subtilis lytRABC divergon: a regulatory unit encompassing the structural 621
genes of the N-acetylmuramoyl-L-alanine amidase and its modifier. J Gen Microbiol 622
138:1949-61. 623
17. Lowe, B. A., J. D. Miller, and M. N. Neely. 2007. Analysis of the polysaccharide 624
capsule of the systemic pathogen Streptococcus iniae and its implications in virulence. 625
Infect Immun 75:1255-64. 626
18. Marques, M. B., D. L. Kasper, M. K. Pangburn, and M. R. Wessels. 1992. Prevention 627
of C3 deposition by capsular polysaccharide is a virulence mechanism of type III group B 628
streptococci. Infect Immun 60:3986-93. 629
19. Miller, J. and M. N. Neely. 2005. Large Scale Screen Highlights the Importance of 630
Capsule for Virulence in the Zoonotic Pathogen, Streptococcus iniae. Infection and 631
Immunity. 632
20. Moarefi, I., D. Jeruzalmi, J. Turner, M. O'Donnell, and J. Kuriyan. 2000. Crystal 633
structure of the DNA polymerase processivity factor of T4 bacteriophage. J Mol Biol 634
296:1215-23. 635
21. Morona, J. K., R. Morona, D. C. Miller, and J. C. Paton. 2002. Streptococcus 636
pneumoniae capsule biosynthesis protein CpsB is a novel manganese-dependent 637
phosphotyrosine-protein phosphatase. J Bacteriol 184:577-83. 638
22. Morona, J. K., R. Morona, and J. C. Paton. 2006. Attachment of capsular 639
polysaccharide to the cell wall of Streptococcus pneumoniae type 2 is required for 640
invasive disease. Proceedings of the National Academy of Sciences 103:8505. 641
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
29
23. Neely, M. N., W. R. Lyon, D. L. Runft, and M. Caparon. 2003. Role of RopB in 642
growth phase expression of the SpeB cysteine protease of Streptococcus pyogenes. J 643
Bacteriol 185:5166-74. 644
24. Over, B., R. Heusser, N. McCallum, B. Schulthess, P. Kupferschmied, J. M. Gaiani, 645
C. D. Sifri, B. Berger-Bächi, and P. Stutzmann Meier. 2011. LytR-CpsA-Psr proteins 646
in Staphylococcus aureus display partial functional redundancy and the deletion of all 647
three severely impairs septum placement and cell separation. FEMS Microbiology Letters 648
320:142-151. 649
25. Parsons, J. B., and C. O. Rock. Is bacterial fatty acid synthesis a valid target for 650
antibacterial drug discovery? Curr Opin Microbiol. 651
26. Phares, C. R., R. Lynfield, M. M. Farley, J. Mohle-Boetani, L. H. Harrison, S. Petit, 652
A. S. Craig, W. Schaffner, S. M. Zansky, K. Gershman, K. R. Stefonek, B. A. 653
Albanese, E. R. Zell, A. Schuchat, and S. J. Schrag. 2008. Epidemiology of invasive 654
group B streptococcal disease in the United States, 1999-2005. JAMA 299:2056-65. 655
27. Rubens, C. E., M. R. Wessels, L. M. Heggen, and D. L. Kasper. 1987. Transposon 656
mutagenesis of type III group B Streptococcus: correlation of capsule expression with 657
virulence. Proc Natl Acad Sci U S A 84:7208-12. 658
28. Shamoo, Y., and T. A. Steitz. 1999. Building a replisome from interacting pieces: 659
sliding clamp complexed to a peptide from DNA polymerase and a polymerase editing 660
complex. Cell 99:155-66. 661
29. Skoff, T. H., M. M. Farley, S. Petit, A. S. Craig, W. Schaffner, K. Gershman, L. H. 662
Harrison, R. Lynfield, J. Mohle-Boetani, S. Zansky, B. A. Albanese, K. Stefonek, E. 663
R. Zell, D. Jackson, T. Thompson, and S. J. Schrag. 2009. Increasing Burden of 664
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
30
Invasive Group B Streptococcal Disease in Nonpregnant Adults, 1990-2007. Clin. Infect 665
Dis 49:85-92. 666
30. Tettelin, H., D. Medini, C. Donati, and V. Masignani. 2006. Towards a universal 667
group B Streptococcus vaccine using multistrain genome analysis. Expert Rev Vaccines 668
5:687-94. 669
31. Utaida, S., P. M. Dunman, D. Macapagal, E. Murphy, S. J. Projan, V. K. Singh, R. 670
K. Jayaswal, and B. J. Wilkinson. 2003. Genome-wide transcriptional profiling of the 671
response of Staphylococcus aureus to cell-wall-active antibiotics reveals a cell-wall-672
stress stimulon. Microbiology 149:2719-2732. 673
32. Verani, J. R., L. McGee, and S. J. Schrag. 2010. Prevention of perinatal group B 674
streptococcal disease--revised guidelines from CDC, 2010. MMWR Recomm Rep 59:1-675
36. 676
33. Yamamoto, S., K. Miyake, Y. Koike, M. Watanabe, Y. Machida, M. Ohta, and S. 677
Iijima. 1999. Molecular characterization of type-specific capsular polysaccharide 678
biosynthesis genes of Streptococcus agalactiae type Ia. J Bacteriol 181:5176-84. 679
34. Yother, J. 2011. Capsules of Streptococcus pneumoniae and Other Bacteria: Paradigms 680
for Polysaccharide Biosynthesis and Regulation. Annu. Rev Microbiol. 65:563-81. 681
682
683
684
685
686
687
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
31
688
689
690
691
692
693
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
Table 1. Primers used in this study.
Primer Sequence (5’ - 3’) EMSA: 5’ GBS-cpsA-pro CGC GGA TCC GTT GAA TTC TCA TAA CTC TAG 3’ GBS-cpsA-pro CCG GAA TTC GCG AAT GAT TAG ACA TTG 5’ GBS-cpsE-pro GAA AAA GGA AGT AAG GGG CTC TTG 3’ GBS-cpsE-pro GCC ACG ACT CCA AAA GTC TC 5’ iniae-cpsA-pro CTC ATA ATG ACA GTC TAT C 3’ iniae-cpsA-pro CCA TCA ATA TCA TTT AAG TC Protein purification: 5’ GBS-cpsA-SmaI TCC CCC GGG TCT AAT CAT TCG CGC CGT C 3’ GBS-cpsA-full-stop-PstI AAA ACT GCA GTT ATT CCT CCA TTG TGT TC 3’ GBS-cpsA-117-stop-PstI AAA ACT GCA GTT ACT CAA TTT CAG AGT ATG AAG C 3’ GBS-cpsA-39-stop-PstI AAA ACT GCA GTT ACA TAA GAA ATA ATG AGA CTA C 3’ GBS-cpsA-245-stop-PstI AAA ACT GCA GTT ATG TTG ATA TAG AGC CAA AAG 5’ MBP-RBS-BamHI CGC GGA TCC GCG GAT AAC AAT TTC ACA CAG G
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
Table 2: GBS strains used in this study.
Parent strain Plasmid Protein Construct 515 WT Vector None 515 WT pGBS-cpsA-full Full length CpsA 515 WT pGBS-cpsA-245 CpsA with LytR domain removed 515 WT pGBS-cpsA-117 CpsA with LytR and DNA_PPF domains removed 515 ΔcpsA Vector None 515 ΔcpsA pGBS-cpsA-full Full length CpsA 515 ΔcpsA pGBS-cpsA-245 CpsA with LytR domain removed 515 ΔcpsA pGBS-cpsA-117 CpsA with LytR and DNA_PPF domains removed
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
Figure Legends 1
Figure 1. GBS 515 CpsA protein. (A) Genes in the capsule operon of GBS 515. Putative 2
promoter sequences within the capsule operon are indicated by bent arrows. (B) Membrane 3
topology of CpsA. (C) Arrangement of the CpsA protein where domains are shown as TM 4
(Transmembrane domains), DNA_PPF (DNA Polymerase Processivity Factor domain), and 5
LytR (LytR_cpsA_psr family domain). Below are truncations made to MBP fusions of CpsA 6
representing the full protein, a truncation at amino acid 245, a truncation at amino acid 117, and 7
a truncation at amino acid 39. 8
Figure 2. Electromobility shift assay demonstrating binding of GBS MBP-CpsA-full (10 pmol) 9
to either (A) the labeled GBS cpsA-pro probe or (B) the labeled GBS cpsE-pro probe. 10
Figure 3. Electromobility shift assays showing binding of (A) MBP-CpsA-full (10 pmol), or (B) 11
MBP-CpsA-117 (52 pmol), or (C) MBP-CpsA-39 (7 pmol), to either the labeled GBS cpsA-pro 12
or labeled GBS cpsE-pro probes in the absence or presence of competitor DNA representing 13
unlabeled GBS cpsA-pro, GBS cpsE-pro, or S. iniae cpsA-pro (unlabeled nonspecific). Unbound 14
labeled probe is indicated by “U” and bound labeled probe is indicated by “B.” 15
Figure 4. Percoll buoyant density assay reflecting capsule levels of the GBS WT or ΔcpsA 16
strains harboring the vector plasmid, or a plasmid containing MBP-CpsA-full, MBP-CpsA-245, 17
or MBP-CpsA-117. Error bars represent SEM. Significance between GBS WT/vector and 18
ΔcpsA/vector is * p < 0.01. 19
Figure 5. Whole blood assay measuring the Log10 level of CFU killed for bacterial strains 20
incubated in human whole blood for 3 hours. Error bars represent the standard deviation. 21
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from
Figure 6. Zebrafish infection study tracking survival over time of zebrafish infected 22
intramuscularly with either the GBS 515 WT or ΔcpsA strain, compared to a media only mock 23
infection. 24
Figure 7. Visualization of chain length at 1000X magnification for GBS 515 WT and ΔcpsA 25
strains, carrying the plasmid vector, MBP-CpsA-full, MBP-CpsA-245, or MBP-CpsA-117 as 26
indicated on the top of the panel. 27
Figure 8. Quantification of chain length for parent strains of GBS 515, (A) WT and (B) ΔcpsA, 28
carrying the plasmid vector, MBP-CpsA-full, MBP-CpsA-245, or MBP-CpsA-117 as indicated 29
on the bottom of the panel. Error bars represent the standard deviation. 30
Figure 9. Visualization of chain length at 1000X magnification for GBS 515 WT and ΔcpsA 31
strains, carrying the plasmid vector, MBP-CpsA-full, MBP-CpsA-245, or MBP-CpsA-117 as 32
indicated on the top of the panel when grown in the presence of a sub-inhibitory concentration of 33
lysozyme. 34
35
36
on July 19, 2020 by guesthttp://jb.asm
.org/D
ownloaded from