Summary of Research (PDF)

Investigating the Role of Prophage in Phenotypic
Heterogeneity associated with Pathogenesis and Virulence
in Streptococcus pyogenes

Alex Remmington
Supervised by Dr. Claire Turner

Alex Remmington



Dr. Claire Turner

Chapter One
Streptococcus pyogenes
The Lancefield Group A Streptococcus (GAS) or Streptococcus pyogenes, is a humanrestricted, Gram-positive pathogen responsible for a diversity of clinical manifestations and
considerable global disease burden, over 700 million infections each year by conservative
estimates. Infectious states include, (but are not limited to) superficial infections such as
pharyngitis, non-bullous impetigo and scarlet fever to potentially lethal invasive manifestations
such as streptococcal toxic shock syndrome (STSS), necrotising fasciitis and puerperal sepsis
(Walker, et al. 2014). In addition, GAS is associated with serious post-infectious sequelae,
notably post-streptococcal glomerulonephritis and rheumatic fever (Cunningham, 2000).
Despite the progress made with global programmes targeting other bacterial pathogens such as
Streptococcus pneumoniae, Neisseria meningitidis and Haemophilus influenzae type b –
attempts to control GAS disease have hitherto served only to highlight the many disparities in
our understanding of this dynamic organism and no vaccine presently exists (Trotter, et al.
2008; Steer, et al. 2016).
Epidemiology and Disease Burden
It is clear that the impact of GAS disease is greatest in countries of low economic status;
however, there have been numerous reports of disease upsurges occurring sporadically across
the globe in recent years (Carapetis et al. 2016; Efstratiou and Lamagni, 2016). Such increases
range from transient periods of intensification to long-term, sustained changes in epidemiology
(Tse, et al. 2012; Nasser, et al. 2014; Davies, et al. 2015; Turner, et al. 2015). Acute rheumatic
fever and rheumatic heart disease are responsible for the majority of deaths in countries of low
economic status (Carapetis et al. 2016). In more affluent countries, the highest mortality is
attributed to invasive GAS disease (iGAS) which, while relatively rare, carries significant case
fatality rates of up to 40% (Efstratiou and Lamagni, 2016). The United Kingdom has the highest
incidence of invasive disease by notification in Europe (3.33 per 100,000) (Turner, et al. 2012)
Recent years have seen the dramatic recrudescence of scarlet fever in Vietnam, China, and
other parts of Asia as well as in the United Kingdom, which in the ongoing 2016/2017 season,
is experiencing its fourth year of dramatically elevated activity (fig. 1) (Efstratiou and Lamagni,
2016; Public Health England, 2017). The precise molecular determinates of disease upsurges
are poorly understood and vary geographically (Turner et al. 2015). Sharp increases in disease
such as those occurring in the UK are cause for concern, in part due to the anticipation of a

Alex Remmington



Dr. Claire Turner

corresponding nationwide increase in iGAS (Stockmann et al. 2014; Efstratiou and Lamagni,
2017;. Guy et al. 2014).

Figure 1: Rates of Scarlet Fever in the United Kingdom from the 2011 to 2017. Increased incidence began with an
unprecedented rise in the 2013/14 season, peaking in the following 2015/16 season. Increased activity has been sustained
in subsequent seasons, including the ongoing 2016/17 season. (Public Health England, 2017).

Emm typing
GAS are subdivided into genotypes using a system based on sequencing of the hypervariable 5’ end of the surface protein encoding emm gene, which encodes the M protein surface
protein. The M protein is ubiquitous within GAS and is therefore highly important in
epidemiological reporting and notifications (Beall, et al. 1996; Bao, et al. 2014). Over 200
emm-types have been described, but the most prominent and globally prevalent are emm-types
1, 3, 12, 28 and 89 (O’Loughlin, et al. 2007; Steer, et al. 2012; Sanderson-Smith, et al. 2014)
(fig. 2).

Alex Remmington



Dr. Claire Turner

Figure 2: The 25 most common emm types, expressed as a proportions of all isolates in high-income countries
(Efstratuiou and Lamagni, 2016).

It has long been understood that there is considerable phenotypic heterogeneity within emm
genotypes and epidemiological data has demonstrated that there are a number of non-random
associations between certain genotypes and their propensity to cause specific clinical
manifestations (Commons, et al. 2008; Bessen, et al. 2011).
A system developed by Bessen, et al. also goes some way to explain these trends, using the
sequence and arrangement of emm genes and the emm-like paralogues that lay both upstream
and downstream of the emm locus to place strains within four groups: A-C, pharyngeal
specialists; D, skin specialists and E, generalist strains that readily cause disease at both sites.
The lattermost group contains the majority of clinical isolates recovered worldwide (Bessen
and Lizano, 2010). Though there are exceptions, for the majority of emm-types, this system
provides a strong inference of their tissue tropism and infection character, which in turn, can
enhance surveillance by serving as an indicator of potential outbreaks.
Pathogenesis
The ability of GAS to cause disease is determined by the coordinated expression of
virulence factors at each stage of disease. GAS produces a vast range of different virulence
factors, with the majority of these targeting host immunity, particularly neutrophils and
complement (fig.3) (Walker et al. 2014).

Alex Remmington



Dr. Claire Turner

Figure 3: GAS virulence factors are largely targeted at host immunity, shown above is a general summary. Some virulence
factors such as SpyCEP and M protein, are common to all genotypes, whereas others such as antimicrobial resistance
elements and capsule, but most notably those associated with prophage such as DNases are more varied in their abundance.
(Walker, et al. 2014).

Group A streptococcal Superantigens
A total of eleven streptococcal superantigens have been described hitherto, varying in
their potency, although the constellation of these toxins present in individual strains also varies
(Commons, et al .2008). Three of the streptococcal superantigens are chromosomally encoded,
notably smeZ, speG and speJ, yet, however despite being components within the core genome,
that is not to say that all three are present or produced uniformly in all strains (Sriskandan and
Reglinski, 2014; Turner, et al. 2012). The remaining members of the streptococcal
superantigen family are associated with prophage elements, these include the scarlet fever
associated superantigens, speC, speA and ssa in addition to speM, speL, speK, speH and speI
(Kasper et al. 2014). As such, the constellation of streptococcal superantigens present in each
isolate varies, and goes some way to explaining the breadth of pathologies associated with GAS
and exhibit a non-random association with individual emm genotypes (Commons, et al. 2008).
Streptococcal superantigens non-specifically cross-link the variable beta chain of human Tcells (TcRVβ) with MHC class II molecules expressed by CD28, bypassing ordinary epitope
processing (Sriskandan and Reglinski, 2017). The consequence is mass non-specific T cell
proliferation which subsequently release copious quantities of pro-inflammatory cytokines
(particularly IFN-γ, and TNF-α) into the circulation (Sriskandan and Reglinski, 2017) (fig. 4).

Alex Remmington



Dr. Claire Turner

Figure 4: (A) Typically, epitope (blue) presentation to T cells does now allow for cross-linkage of the antigen binding domain
with the T cell receptor (orange) and MHC class II molecules (red) on the surface of antigen presenting immune cells. (B)
Conversely, streptococcal superantigens such as SPEA (green) and SPEC (purple) are capable of overstimulating a larger
population of T cells through binding with a specific subset of T cell receptor. In this way, they are capable of binding nonspecifically to either the α or β chain of MHC class 2 molecules on CD28 T cells. (C) An additional zinc-dependent mechanism
of activity has been described whereby the streptococcal superantigen binds to the β chain and irreversibly stimulates the T
cell with the epitope being presented. (Modified from Sriskandan and Reglinski, 2017).

In this way, certain streptococcal superantigens are thought to be key contributories to the
development of STSS, a severe systemic complication of invasive infection occurring in
approximately 10-16% cases (Kasper et al. 2014). Streptococcal superantigens may also be
associated with the clinical presentation of scarlet fever which is classically considered to be a
toxin-mediated disease (Walker, et al. 2014; Commons, et al. 2014). The benefit of
superantigen activity to GAS is unclear, however it is possible that these toxins serve as a
means of immune distraction and may also enhance transmission, particularly in pharyngitis
and scarlet fever, by causing localised inflammation, tissue destruction and the production of
highly infectious aerosols (Kasper, et al. 2014; Proft and Fraser, 2016; Afshar et al. 2017).
Group A streptococcal DNases
A number of bacterial pathogens produce DNases as virulence factors. In GAS, they
are for the most part extracellular, with the exception of the cell wall bound nuclease spna,
identified relatively recently in the genotype emm1 SF370 strain (Hasegawa, et al. 2010).
Although their broad function is known, their precise role in infection is not. It is thought that
streptococcal DNases facilitate dissemination of GAS cells in the human host by liquefying
purulent exudate (Broudy, et al. 2002). It has also been speculated that nucleases facilitate
dissemination and transmission of progeny virions between bacterial hosts, potentially
conferring a selection advantage to both the bacteriophage as well as the bacterium (Broudy,
et al. 2002). A prevailing hypothesis regarding the function of these enzymes is a role in
prevention of neutrophil activation and production of neutrophil extracellular traps (NET’s)

Alex Remmington



Dr. Claire Turner

(Buchanan, et al. 2006). These fibrous extracellular networks are composed of chromatin,
proteolytic enzymes and other peptides, binding to invading bacteria and prevent their spread
from the original focus of infection (Brinkmann, et al. 2004). Once entrapped, secreted cationic
antimicrobial peptides attack the offending organism and degrade virulence factors. However,
by secreting DNases, GAS are able to escape these traps and disseminate. GAS DNases do not
only degrade host nucleic acids, however. Depolymerisation of bacterial DNA by these
enzymes has been shown to prevent immune killing of GAS by reducing host TLR-9 signalling
and subsequent recognition of un-methylated CpG-rich DNA by macrophages (Uchiyama, et
al. 2012). Similar to the streptococcal superantigens, only two GAS DNases are associated
with the core genome, specifically sdaB and spnA, the remaining five (sda2, spd1, spd3, spd4
and sdn1) are prophage borne (Sumby, et al. 2005; Hasegawa, et al. 2010). As such, individual
strains may possess numerous DNase genes, the combination thereof may influence the
phenotype of the bacterial host and it seems likely that these enzymes have a hitherto
unidentified role in infection and pathogenesis.
GAS Genomics
The GAS genome is on average ~1.8Mb in size and is regulated by three key
mechanisms; two-component systems (TCS), stand-alone transcriptional regulators and small,
noncoding sRNA molecules (Kreikemeyer, et al. 2003; Cao, et al. 2014). The corollary of these
events contribute to the ability of GAS to colonise, infect, persist and transmit, modulating the
phenotype to exploit and invade the diversity of micro-environments in the human host and
ultimately causing disease. Regulation of the GAS genome is therefore of pivotal importance
for not only maintenance of the core genome and host cell homeostasis, but may also represent
a significant nexus in the regulation of prophage elements by the bacterium (Kreikemeyer, et
al. 2003; Anbalagan, et al. 2013).

Alex Remmington



Dr. Claire Turner

Outbreaks, epidemics and upsurges have highlighted that emm-typing, the most widely used
tool implemented in surveillance, may fail to provide a true representation of GAS population
dynamics (Beres, et al. 2004; Davies, et al. 2015; Al-Shahib, et al. 2016). Through the rapid
increased use of whole genome sequencing in recent years, it has become evident that the
molecular ecology of this organism is more complex and intra-specific, intra-genotype
competition is now well recognised (Beres, et al. 2006; Turner, et al. 2015). Crucially,
however, large scale genomic studies of a range of different emm-types have indicated that the
use of the emm sequence for genotyping well defines lineages of GAS isolates (fig. 5).

e
Figure 5: A collection of 480 GAS isolates representing 60 different emm-types were whole genome sequenced. Resulting
sequence reads were mapped to the core genome (no prophages) of a reference strain and SNPs were used to generate a
maximum-likelihood phylogenetic tree. Colours indicate emm-types, except where unique emm-types are represented by a
single isolate. With few exceptions, all emm-types are defined by a single lineage. Scale bar represents single nucleotide
polymorphism. (Turner, et al. unpublished data)

GAS are not considered constitutively genetically competent under most conditions and
transformation as a means of horizontal gene transfer (HGT) in GAS and other pyogenic
streptococci is poorly understood, and thus the uptake of naked DNA by transformation was
not thought to be a significant mechanism by which the chromosome could undergo change
(Everitt, et al. 2014; Turner et al. 2015). As such, for the bacterial chromosome to undergo
change, GAS is largely reliant on other means of HGT. Notably, transduction by temperate
bacteriophage and acquisition of integrative and conjugative elements (ICE) (Tse, et al. 2012;
Nasser, et al. 2014; Davies, et al. 2015; Turner, et al. 2015). It has been shown recently that

Alex Remmington



Dr. Claire Turner

when GAS are associated with biofilms, uptake of extracellular DNA may occur (Marks et al.
2014). Of the entire GAS chromosome, 10-12% of this is comprised by prophage and
integrative and conjugative elements (Holden et al. 2007; Euler et al. 2016; Ferretti et al. 2004;
Banks et al. 2002).
Prophage Genomics
Retrospective GAS genomic studies suggest that the most dramatic changes in epidemiology
and infection character are often, but not exclusively, due to integrative and conjugative
elements (ICE) or transduction by prophage (Beres, et al. 2002; Tse, et al. 2012; Davies, et al.
2015; Al-Shahib, et al. 2016; Afshar, et al. 2017). A greater understanding of circulating
strains, and prophage genotype could serve as an important facet to existing surveillance tools
enabling the prediction of epidemic waves and the emergence of new strains, as opposed to
relying on retrospective longitudinal analyses.
GAS has a long-standing evolutionary relationship with bacteriophage, and the ability of these
mobile genetic elements to carry virulence factors between bacteria, can thereby influence the
phenotype of the lysogen (Wilkening and Federle, 2017). The application of whole genome
sequencing has dramatically expanded our understanding of these elements and have confirmed
that these viruses often constitute substantial portions of the host chromosome. Indeed, most
strains are poly-lysogenic, with up to seven prophage integrated throughout the genome (fig.
6) (Beres, et al. 2004; McShan and Nguyen, 2016).