The most widely accepted typing methods for the study of C. difficile are currently
restriction endonuclease analysis (REA), pulsed field gel electrophoresis (PFGE), PCR
ribotyping, toxinotyping and multilocus sequence typing (MLST). REA and PFGE type
isolates based on the DNA fingerprint after restriction digestion [83,84].
PCR ribotyping is based on the fact that the intergenic spacer region between the 16S
and 23S rRNA differs between the multiple pairs of alleles in an individual cell. As most
bacteria possess multiple copies of these genes, resulting in products of varying sizes in
a PCR reaction using universal primers. Related isolates will have identical patterns of
PCR products whereas more distantly related isolates will have variations which allows
for grouping of isolates based on these variations [85].
Toxinotyping is a method that amplifies several fragments of the PaLoc in PCR
reactions which are restricted with one to four different restriction enzymes in order
to detect restriction fragment length polymorphisms of the PaLoc amplicons [86]. For
MLST, six to seven housekeeping genes are sequenced to determine the genetic
relationship between strains [87].
Recent studies of MLST data and whole genome sequence comparison have
discovered that all strains can be divided into six major clades which were shown to
have diverged millions of years ago [88]. However, sequence data on specific
housekeeping genes and the PaLoc suggest that some recombination has occurred
between members of these six clades [89,90]. Isolates from the epidemic outbreaks
All of the different typing schemes have been developed in order to assign phenotypic
differences, such as hypervirulence (section 1.1.3.2), to specific types. Overlap is seen
between the results of the typing methods, for instance with the toxinotyping and
ribotyping methods [92]. Over the past decade most literature has used ribotyping or a
combination of methods to type isolates, therefore, in this thesis only the ribotypes of
strains are indicated.
1.1.3.2. Hypervirulence
Since approximately 2000, reports of epidemic outbreaks of C. difficile in hospitals in
the UK and other western countries increased [9,10]. The isolates from these
outbreaks were reported to be more virulent, producing higher levels of toxin, were
more often resistant to fluoroquinolones and were responsible for a higher relapse
rate in patients [11]. The majority of these isolates were ribotype 027 and were
reported to be hypervirulent [93].
As described above, TcdD is an alternative sigma factor that is required for the
transcription of tcdA and tcdB, whereas TcdC is an anti-sigma factor that negatively
regulates toxin production (Figure 1.3) [52,53]. It was shown that strains of ribotype
027 contain an 18-bp deletion in tcdC which results in a truncated protein being
produced that is not capable of regulating toxin production [9]. In 2010 in the UK,
ribotype 027 strains were reported to be isolated most frequently from patients [94],
however, a report from the Netherlands showed the emergence of ribotype 078, of
which the isolates contain a 39-bp deletion in tcdC as well as an early stop codon [95].
Recent studies of complementation of the tcdC mutation give inconclusive results
leaving the role of tcdC in toxin regulation unclear [54,55](section 1.1.2.1) and the
The emergence of ribotype 002 has been reported in China [96], ribotype 001 in
Germany [97] and ribotype 018 in Italy [98]. Hypervirulence in C. difficile is therefore
no longer considered to be limited to ribotype 027. A study of samples from the past
decade from the UK and US revealed that some isolates that were designated ribotype
027 were actually ribotypes 176, 198 and 244 [99]. These ribotypes are highly similar
and are predicted to have evolved from ribotype 027 recently, however it is not
reported if samples from these ribotypes contain the mutated tcdC allele.
Although the in vitro production of TcdA and TcdB was shown to be higher in ribotype
027 strains [9,72], concern over the biological relevance of toxin production in in vitro
batch culture experiments has been raised [100]. Furthermore, a human gut model
shows the production of toxins by ribotype 027 strains is not higher per unit of time
but has a longer duration of toxin production, which may account for the increased
severity of symptoms for patients infected with a ribotype 027 isolate [101].
It was also reported that ribotype 027 strains had a higher level of sporulation [72]. A
recent study which included a large number of strains showed that the sporulation
rate is very variable between strains within a ribotype, and high and low level
sporulation is seen for both epidemic ribotypes and non-epidemic ribotypes, which
suggests the amount of sporulation has no effect on severity of disease [74]. Although
the epidemiology of C. difficile has certainly changed over the past two decades, the
basis for the claim that certain ribotypes are hypervirulent is not solid and therefore
these are now often referred to as epidemic isolates.
Due to the increase in C. difficile infections it became mandatory in the UK to report all
cases of infection for epidemiological surveillance [31]. Between 2007 and 2010 the
to <4000 per year (Figure 1.6). In this period, the number of C. difficile infections
decreased from >55,000 to <22,000 in the UK [102]. The decrease in these numbers is
most likely due to the improvements in hospital procedures. Although the prevalence
of C. difficile is reducing, the occurrence of epidemic strains is still not understood and
further study is necessary in order to improve treatment and prevention strategies.
Figure 1.6 Number of deaths attributed to C. difficile.
The blue bar shows the number of patients where C. difficile is mentioned as the cause of death on the death certificate whereas the grey bar shows the number of patients where C. difficile is mentioned as a possible contributing factor. Figure reproduced from the Office for National Statistics website, last accessed 5th September 2012 [103].