Molecular Microbiology (2006)
(1), 84–98 doi:10.1111/j.1365-2958.2005.04930.x
First published online 27 October 2005
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd
Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382X© 2005 The Authors; Journal compilation © 2005 Blackwell Publishing Ltd
Muralytic activity of M. luteus RpfG. V. Mukamolova
Accepted 23 September, 2005. *For correspondence. E-mail
firstname.lastname@example.org; Tel. (
44) 1970 622 348; Fax (
44) 1970 622 354.
Muralytic activity of
relationship to physiological activity in promoting
bacterial growth and resuscitation
Micrococcus luteus Rpf and its
Galina V. Mukamolova,
Elena G. Salina,
Arseny S. Kaprelyants
Institute of Biological Sciences, University of Wales,
Aberystwyth, Ceredigion SY23 3DD, UK.
Bakh Institute of Biochemistry, Russian Academy of
Sciences, Leninsky pr.33, 119071 Moscow, Russia.
MRC Centre for Protein Engineering, Hills Road,
Cambridge CB2 2QH, UK.
School of Chemistry, University of Manchester, Faraday
Building, Sackville Street, Manchester M60 1QD, UK.
Alexey G. Murzin,
Galina R. Demina,
and Michael Young
Douglas B. Kell,
The culturability of several actinobacteria is con-
trolled by resuscitation-promoting factors (Rpfs).
These are proteins containing a
that adopts a lysozyme-like fold. The invariant cata-
lytic glutamate residue found in lysozyme and various
bacterial lytic transglycosylases is also conserved in
the Rpf proteins. Rpf from
founder member of this protein family, is indeed a
muralytic enzyme, as revealed by its activity in zymo-
grams containing M. luteus
(i) cause lysis of Escherichia coli
secreted into the periplasm; (ii) release fluorescent
material from fluorescamine-labelled cell walls of
luteus ; and (iii) hydrolyse the artificial lysozyme sub-
totrioside. Rpf activity was reduced but not
completely abolished when the invariant glutamate
residue was altered. Moreover, none of the other
acidic residues in the Rpf domain was absolutely
required for muralytic activity. Replacement of one or
both of the cysteine residues that probably form a
disulphide bridge within Rpf impaired but did not
completely abolish muralytic activity. The muralytic
activities of the Rpf mutants were correlated with their
abilities to stimulate bacterial culturability and resus-
citation, consistent with the view that the biological
c . 70-residue domain
Micrococcus luteus, the
cell walls and its ability to
when expressed and
β β β β
′ ′ ′ ′
′ ′ ′ ′′ ′ ′ ′
activity of Rpf results directly or indirectly from its
ability to cleave bonds in bacterial peptidoglycan.
Some bacteria such as
form dormant endospores, which represent an intermedi-
ate state between life and death. They show no obvious
signs of vitality, but they nevertheless retain the ability to
germinate and to resume growth and division when they
detect specific (usually chemical) signals in their en-
vironment. Although the actinobacteria do not make
endospores, several of them can enter a state in which
they have much reduced metabolic activity and lose cul-
turability. They require resuscitation in a nutrient-poor liquid
medium before they can resume active growth on an agar-
solidified medium. The adoption of this ‘non-culturable’
state may represent a survival strategy that helps these
organisms to persist during nutrient-limited conditions un-
favourable for growth (Mukamolova
the mechanisms by which ‘non-culturable’ cells are pro-
duced remain unknown, their very existence has important
implications for microbial population ecology and for the
control of infectious disease (Kell
ple, some (at least) of the persistent cells of
responsible for reactivation tuberculosis (TB)
in humans may be in a ‘non-culturable’ state (Mukamolova
et al ., 2003). Evidence to support this comes from exper-
iments with a chronic mouse model of TB, in which the
viable count of M. tuberculosis
was two orders of magnitude higher when measured in a
liquid medium that permits resuscitation (MPN) than when
measured by conventional plating (cfu) on a solid medium
that does not (Dhillon et al., 2004). The presence of
tuberculosis DNA in culture-negative and superficially nor-
mal human lung tissue during latent infection is also con-
sistent with this view (Hernández-Pando
‘Non-culturable’ forms of
resuscitated using the supernatant from late logarithmic
phase laboratory cultures (Kaprelyants
active component is a secreted protein called Rpf – resus-
citation-promoting factor (Mukamolova
homologues are widespread throughout the actinobacte-
ria (Kell and Young, 2000) and biological activity (resusci-
Bacillus and Clostridium spp. can
et al ., 2003). Although
et al., 1998). For exam-
10 months post-infection
et al., 2000).
can beMicrococcus luteus
et al., 1994). The
et al., 1998). Rpf
Muralytic activity of M. luteus Rpf85
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd,
Molecular Microbiology ,
tation or growth promotion) has been demonstrated for
several representatives, including all five proteins found in
M. tuberculosis and its close relatives (Mukamolova
2002a; Zhu et al ., 2003). The gene encoding Rpf could
not be inactivated in M. luteus
a second functional copy (Mukamolova
whereas the individual rpf-like genes of
were dispensable for growth
et al ., 2004; Tufariello et al ., 2004). Triple mutants of
tuberculosis showed significantly attenuated virulence in
mice and were unable to resuscitate spontaneously from
the ‘non-culturable’ state in vitro
rpf-like genes grew as well as did the wild type, but
regrowth after a period of storage was significantly
impaired (Hartmann et al., 2004).
The extreme potency of Rpf (activity at picomolar con-
centrations), and the strict requirement for Rpf to permit
growth under certain defined conditions, suggested that
this protein should be regarded as a bacterial growth
factor (Mukamolova et al ., 1998). A mode of action based
on specific enzymatic activity was also considered, and
this was supported by the prediction that Rpf has a
lysozyme-like fold (Finan, 2003; Kazarian
et al ., 2004; http://predictioncenter2.
gc.ucdavis.edu/). However, attempts to mimic Rpf activity
by exogenous addition of traces of lysozyme were unsuc-
cessful (G.V. Mukamolova, unpublished). In this paper we
demonstrate that Rpf possesses muralytic activity. A mod-
elling approach was employed to generate a predicted
structure for the Rpf fold and this was used to inform site-
directed mutagenesis experiments to identify residues
that are important for muralytic and physiological activity.
The results are consistent with the view that the muralytic
activity of Rpf is probably responsible for its observed
activity both in resuscitating dormant cells and in stimu-
lating growth when bacteria are inoculated at low density
into a minimal medium or held in a prolonged stationary
phase. One possible interpretation is that Rpf remodels
the cell wall throughout bacterial growth and that its activ-
ity is especially important during the transition from a ‘non-
culturable’ to an actively growing state.
While this work was in progress, the solution structure
of the Rpf-like domain of M. tuberculosis
was published, confirming the lysozyme-like fold and also
providing some data showing the importance of a con-
served glutamate residue for growth stimulatory activity
(Cohen-Gonsaud et al., 2005).
except in the presence of
in vitro and in vivo
mutant lacking both of its
et al., 2005). A
et al., 2003;
Structure prediction of Rpf
Wanted’ list (TMW), the initiative to predict by a commu-
M. luteus Rpf was included in the ‘Ten Most
nity-wide effort the structures of the 10 most wanted pro-
teins of unknown structure, suggested by biologists on the
basis of their biological importance (Abbott, 2001). Pre-
dictions for the 10 selected proteins were collected and
analysed, and the results were presented at the world
structure predictors’ community meeting, CASP5, in
December 2002 (Tramontano, 2003). There were 54 pre-
dictions for Rpf (target TMW0001) resulting in a consen-
sus prediction of a lysozyme-like fold for the Rpf domain.
One of us (A.G.M.) participated in both CASP5 and TMW
and communicated to the others this consensus predic-
tion together with the 3-D co-ordinates of a detailed model
of the Rpf domain structure.
The Rpf domain was predicted to have a minimal
lysozyme fold, common to all known members of the
lysozyme super-family in the SCOP database (Murzin
et al., 1995). The selected model of the Rpf domain was
constructed on the basis of not one, but several structures
from different lysozyme families, aiming to provide a struc-
tural explanation of every conserved feature in the multiple
sequence alignment of the Rpf domain family. In particu-
lar, the conserved glutamate residue (E54) aligned with
the catalytically essential invariant glutamate residue,
suggesting that the Rpf domain might possess an enzy-
The recent determination of the
domain (PDB entry 1xsf; Cohen-Gonsaud
which has a 50% sequence identity to the
domain, allowed a direct comparison of the predicted
and experimental structures (Fig. 1). The predicted and
observed folds of the Rpf domains are very similar indeed.
Moreover, 59 of 78 (
75%) of the modelled residues
appear aligned correctly and can be superimposed with
an r.m.s. deviation of 2.0 Å. The correctly aligned residues
include many of those modified by mutagenesis in this
M. tuberculosis RpfB
., 2005),et al
M. luteus Rpf
Rpf is a muralytic enzyme
According to the predicted structure of
the recently solved solution structure of the Rpf domain
of M. tuberculosis RpfB (Cohen-Gonsaud
the Rpf domain has a lysozyme-like fold and the Rpf
proteins may therefore show muralytic activity. This was
initially investigated using zymography. Native, secreted
Rpf was partially purified from culture supernatants and
tested following electrophoresis through polyacrylamide
gels containing either cell wall fragments or autoclaved
whole cells as substrate. A clearance band was observed
corresponding to a 25 kDa protein (Fig. 2B) that reacted
with anti-Rpf antibodies. At least two additional superna-
tant proteins substantially larger than Rpf also had mura-
lytic activity, but anti-Rpf antibodies did not detect them
(Fig. 2A). Recombinant Rpf (expressed from pET19b with
M. luteus Rpf, and
et al., 2005),
86G. V. Mukamolova et al.
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd,
Molecular Microbiology, 84–98
a polyhistidine tag at the N-terminus) isolated from
Escherichia coli usually gave negative results in this assay
for muralytic activity. Removal of the polyhistidine tag or
repositioning it at the C-terminus or varying the assay
conditions did not improve activity. This was finally
achieved by Rpf expression and secretion into the
periplasm, providing an environment conducive to disul-
phide bridge formation. The protein thus produced yielded
a distinct clearance band of the expected size in zymo-
grams (Fig. 2C). Sometimes, multiple clearance bands
were observed, probably representing different forms of
Rpf. Nucleic acid strongly inhibited the observed muralytic
activity, in accordance with previous observations on mur-
alytic enzymes from
E. coli (Kusser and Schwarz, 1980).
The recombinant protein was less active than the native
protein and both lost activity after storage at 4
Hyper-expression of Rpf in the
causes bacterial lysis
E. coli periplasm
sequence from the T7 promoter in pET19b, it accumulates
internally as inclusion bodies with no evidence of bacterial
lysis (Mukamolova et al., 1998; 2002b). However, arabi-
nose-induced expression and secretion of Rpf in
from the recombinant pBAD/gIIIb vector resulted in sub-
stantial lysis of the bacteria (Fig. 3). Lysis was confirmed
microscopically and has been described previously for
muralytic enzymes expressed in
1993; Ehlert et al., 1995). Lysis was observed in LB and
other rich media but it was much reduced in the SOB
medium (containing 8.5 mM Mg
centration) that was employed for routine protein expres-
rpf gene is expressed without its signal
E. coli (Fischer et al .,
and a reduced salt con-
sion. Lysis was correlated with the accumulation of Rpf
and was not observed in strains harbouring either the
expression vector alone or a derivative expressing calm-
odulin (Fig. 3). M. luteus Rpf contains an N-terminal Rpf
domain and a C-terminal LysM domain (Bateman and
Bycroft, 2000) separated by a linker region (Mukamolova
et al., 1998). It was previously reported that a truncated
form of Rpf comprising only the Rpf domain retained bio-
logical activity (Mukamolova
when expressed separately, the LysM domain showed no
evidence of muralytic activity on zymograms (data not
shown) nor was there any lysis of the
host (Fig. 3). We were unable to express sufficient quan-
tities of the Rpf domain alone in a recombinant pBAD/gIIIb
vector to test its activity.
et al., 2002b). As expected,
E. coli expression
M. luteus cell walls labelled with
A semi-quantitative estimate of the muralytic activity of
Rpf was obtained using M. luteus
fluorescamine as previously described (Mintz
1975). Under the assay conditions employed here (see
Experimental procedures ), 10
Rpf hydrolysed 16.9
1.3% of the fluorescamine-labelled
wall material per hour. The observed activity was well
above background, which was 0.27
0.02% for a purified extract from
bouring the empty expression vector, pBAD/gIIIb. The
activity of purified recombinant Rpf was between five- and
six-fold lower than that of an equivalent amount of freshly
prepared hen egg white lysozyme. Similar results were
obtained using three different batches of recombinant Rpf.
The observed activity was completely abolished by boiling
cell walls labelled with
et al .,
g purified recombinant
0.1% for buffer only
E. coli har-
retical model of the
ing this study (A) is compared with the recent
experimental structure of the
RpfB domain (B). Coloured regions have both
similar backbone conformation and correctly
predicted sequence alignment. Depicted with
sticks are the side-chains of selected con-
served residues targeted by site-directed
mutagenesis in A (first number) and their coun-
terparts in B (second number): D48/286, C53/
291, E54/292, Q72/310 and C114/355.
Structure of the Rpf domain. The theo-
M. luteus Rpf domain guid-
Muralytic activity of M. luteus Rpf 97
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 84–98
of soluble and active hen egg white lysozyme in Escheri-
chia coli. Appl Microbiol Biotechnol 39: 537–540.
Girardin, S.E., Boneca, I.G., Carneiro, L.A., Antignac, A.,
Jehanno, M., Viala, J., et al. (2003) Nod1 detects a unique
muropeptide from Gram-negative bacterial peptidoglycan.
Science 300: 1584–1587.
Grutter, M.G., Weaver, L.H., and Matthews, B.W. (1983)
Goose lysozyme structure: an evolutionary link between
hen and bacteriophage lysozymes? Nature 303: 828–831.
Hartmann, M., Barsch, A., Niehaus, K., Puhler, A., Tauch, A.,
and Kalinowski, J. (2004) The glucosylated cell surface
protein, Rpf2, containing a resuscitation-promoting factor
motif, is involved in intercellular communication of Coryne-
bacterium glutamicum. Arch Microbiol 182: 299–312.
Heidrich, C., Ursinus, A., Berger, J., Schwarz, H., and Holtje,
J.V. (2002) Effects of multiple deletions of murein hydro-
lases on viability, septum cleavage, and sensitivity to large
toxic molecules in Escherichia coli. J Bacteriol 184: 6093–
Hernández-Pando, R., Jeyanathan, M., Mengistu, G., Agui-
lar, D., Orozco, H., Harboe, M., et al. (2000) Persistence
of DNA from Mycobacterium tuberculosis in superficially
normal lung tissue during latent infection. Lancet 356:
Holtje, J.V., Mirelman, D., Sharon, N., and Schwarz, U.
(1975) Novel type of murein transglycosylase in Escheri-
chia coli. J Bacteriol 124: 1067–1076.
Hoppner, C., Liu, Z., Domke, N., Binns, A.N., and Baron, C.
(2004) VirB1 orthologs from Brucella suis and pKM101
complement defects of the lytic transglycosylase required
for efficient type IV secretion from Agrobacterium tumefa-
ciens. J Bacteriol 186: 1415–1422.
Huard, C., Miranda, G., Wessner, F., Bolotin, A., Hansen, J.,
Foster, S.J., and Chapot-Chartier, M.P. (2003) Character-
ization of AcmB, an N-acetylglucosaminidase autolysin
from Lactococcus lactis. Microbiology 149: 695–705.
Jacobs, C., Huang, L.J., Bartowsky, E., Normark, S., and
Park, J.T. (1994) Bacterial cell wall recycling provides cyto-
solic muropeptides as effectors for beta-lactamase induc-
tion. EMBO J 13: 4684–4694.
Jacobs, C., Frere, J.M., and Normark, S. (1997) Cytosolic
intermediates for cell wall biosynthesis and degradation
control inducible beta-lactam resistance in Gram-negative
bacteria. Cell 88: 823–832.
Kaprelyants, A.S., Mukamolova, G.V., and Kell, D.B. (1994)
Estimation of dormant Micrococcus luteus cells by penicil-
lin lysis and by resuscitation in cell-free spent medium at
high dilution. FEMS Microbiol Lett 115: 347–352.
Kazarian, K.A., Yeremeev, V.V., Kondratieva, T.K., Telkov,
M.V., Kaprelyants, A.S., and Apt, A.S. (2003) Proteins of
Rpf Family as Novel TB Vaccine Candidates. First Interna-
tional Conference on TB Vaccines for the World, Montreal,
Kell, D.B., and Young, M. (2000) Bacterial dormancy and
culturability: the role of autocrine growth factors. Curr Opin
Microbiol 3: 238–243.
Kell, D.B., Kaprelyants, A.S., Weichart, D.H., Harwood, C.L.,
and Barer, M.R. (1998) Viability and activity in readily cul-
turable bacteria: a review and discussion of the practical
issues. Ant van Leeuwenhoek 73: 169–187.
Koraimann, G. (2003) Lytic transglycosylases in macromo-
lecular transport systems of Gram-negative bacteria. Cell
Mol Life Sci 60: 2371–2388.
Korsak, D., Liebscher, S., and Vollmer, W. (2005) Suscepti-
bility to antibiotics and beta-lactamase induction in murein
hydrolase mutants of Escherichia coli. Antimicrob Agents
Chemother 49: 1404–1409.
Kusser, W., and Schwarz, U. (1980) Escherichia coli murein
transglycosylase. Purification by affinity chromatography
and interaction with polynucleotides. Eur J Biochem 103:
de Man, J.C. (1975) The probability of most probable num-
bers. Eur J Appl Microbiol 1: 67–78.
Mintz, G., Herbold, D.R., and Glaser, L. (1975) A fluorescent
assay for bacterial cell wall lytic enzymes. Anal Biochem
Moak, M., and Molineux, I.J. (2000) Role of the Gp16 lytic
transglycosylase motif in bacteriophage T7 virions at the
initiation of infection. Mol Microbiol 37: 345–355.
Mukamolova, G.V., Kaprelyants, A.S., Young, D.I., Young, M.,
and Kell, D.B. (1998) A bacterial cytokine. Proc Natl Acad
Sci USA 95: 8916–8921.
Mukamolova, G.V., Kormer, S.S., Kell, D.B., and Kaprely-
ants, A.S. (1999) Stimulation of the multiplication of Micro-
coccus luteus by an autocrine growth factor. Arch Microbiol
Mukamolova, G.V., Turapov, O.A., Young, D.I., Kaprelyants,
A.S., Kell, D.B., and Young, M. (2002a) A family of auto-
crine growth factors in Mycobacterium tuberculosis. Mol
Microbiol 46: 623–635.
Mukamolova, G.V., Turapov, O.A., Kazaryan, K., Telkov, M.,
Kaprelyants, A.S., Kell, D.B., and Young, M. (2002b) The
rpf gene of Micrococcus luteus encodes an essential
secreted growth factor. Mol Microbiol 46: 611–621.
Mukamolova, G.V., Kaprelyants, A.S., Kell, D.B., and Young,
M. (2003) Adoption of the transiently non-culturable state
– a bacterial survival strategy? Adv Microb Physiol 47: 65–
Murzin, A.G., and Bateman, A. (2001) CASP2 knowledge-
based approach to distant homology recognition and fold
prediction in CASP4. Proteins (Suppl. 5): 76–85.
Murzin, A.G., Brenner, S.E., Hubbard, T., and Chothia, C.
(1995) SCOP: a structural classification of proteins data-
base for the investigation of sequences and structures. J
Mol Biol 247: 536–540.
Mushegian, A.R., Fullner, K.J., Koonin, E.V., and Nester,
E.W. (1996) A family of lysozyme-like virulence factors in
bacterial pathogens of plants and animals. Proc Natl Acad
Sci USA 93: 7321–7326.
Nambu, T., Minamino, T., Macnab, R.M., and Kutsukake, K.
(1999) Peptidoglycan-hydrolyzing activity of the FlgJ pro-
tein, essential for flagellar rod formation in Salmonella
typhimurium. J Bacteriol 181: 1555–1561.
Payie, K.G., Rather, P.N., and Clarke, A.J. (1995) Contribu-
tion of gentamicin 2′-N-acetyltransferase to the O acetyla-
tion of peptidoglycan in Providencia stuartii. J Bacteriol
Pedulla, M.L., Ford, M.E., Houtz, J.M., Karthikeyan, T., Wad-
sworth, C., Lewis, J.A., et al. (2003) Origins of highly
mosaic mycobacteriophage genomes. Cell 113: 171–182.
Pisabarro, A.G., de Pedro, M.A., and Vazquez, D. (1985)
Structural modifications in the peptidoglycan of Escheri-
98G. V. Mukamolova et al.
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 84–98
chia coli associated with changes in the state of growth of
the culture. J Bacteriol 161: 238–242.
Potvin, C., Leclerc, D., Tremblay, G., Asselin, A., and
Bellemare, G. (1988) Cloning, sequencing and expression
of a Bacillus bacteriolytic enzyme in Escherichia coli. Mol
Gen Genet 214: 241–248.
Ravagnani, A., Finan, C.L., and Young, M. (2005) A novel
firmicute protein family related to the actinobacterial resus-
citation-promoting factors by non-orthologous domain dis-
placement. BMC Genomics 6: 39.
Rosenthal, R.S., Nogami, W., Cookson, B.T., Goldman,
W.E., and Folkening, W.J. (1987) Major fragment of soluble
peptidoglycan released from growing Bordetella pertussis
is tracheal cytotoxin. Infect Immun 55: 2117–2120.
Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecu-
lar Cloning, A Laboratory Manual. New York: Cold Spring
Harbor Laboratory Press.
Shleeva, M., Mukamolova, G.V., Young, M., Williams, H.D.,
and Kaprelyants, A.S. (2004) Formation of non-culturable
cells of Mycobacterium smegmatis in stationary phase in
response to growth under sub-optimal conditions and their
Rpf-mediated resuscitation. Microbiology 150: 1687–1697.
Signoretto, C., Lleo, M.M., Tafi, M.C., and Canepari, P.
(2000) Cell wall chemical composition of Enterococcus
faecalis in the viable but nonculturable state. Appl Environ
Microbiol 66: 1953–1959.
Signoretto, C., del Mar Lleo, M., and Canepari, P. (2002)
Modification of the peptidoglycan of Escherichia coli in the
viable but nonculturable state. Curr Microbiol 44: 125–131.
Sun, Z., and Zhang, Y. (1999) Spent culture supernatant of
Mycobacterium tuberculosis H37Ra improves viability of
aged cultures of this strain and allows small inocula to
initiate growth. J Bacteriol 181: 7626–7628.
Thunnissen, A.M.W.H., Dijkstra, A.J., Kalk, K.H., Rozeboom,
H.J., Engel, H., Keck, W., and Dijkstra, B.W. (1994) Dough-
nut-shaped structure of a bacterial muramidase revealed
by X-ray crystallography. Nature 367: 750–754.
Thunnissen, A.M., Isaacs, N.W., and Dijkstra, B.W. (1995)
The catalytic domain of a bacterial lytic transglycosylase
defines a novel class of lysozymes. Proteins 22: 245–258.
Tramontano, A. (2003) Of men and machines. Nat Struct Biol
Tufariello, J.M., Jacobs, W.R., Jr, and Chan, J. (2004) Indi-
vidual Mycobacterium tuberculosis resuscitation-promot-
ing factor homologues are dispensable for growth in vitro
and in vivo. Infect Immun 72: 515–526.
Weaver, L.H., Grutter, M.G., and Matthews, B.W. (1995) The
refined structures of goose lysozyme and its complex with
a bound trisaccharide show that the ‘goose-type’ lysozymes
lack a catalytic aspartate residue. J Mol Biol 245: 54–68.
Zhang, Y., Yang, Y., Woods, A., Cotter, R.J., and Sun, Z.
(2001) Resuscitation of dormant Mycobacterium tubercu-
losis by phospholipids or specific peptides. Biochem Bio-
phys Res Commun 284: 542–547.
Zhu, W., Plikaytis, B.B., and Shinnick, T.M. (2003) Resusci-
tation factors from mycobacteria: homologs of Micrococcus
luteus proteins. Tuberculosis (Edinb) 83: 261–269.
The following supplementary material is available for this
Table S1. Mutagenic primers.
This material is available as part of the online article from