Tn Smu1 is a functional integrative and conjugative element in Streptococcus mutans that when expressed causes growth arrest of host bacteria

Abstract

Integrative and conjugative elements (ICEs) are major drivers of horizontal gene transfer in bacteria. They mediate their own transfer from host cells (donors) to recipients and allow bacteria to acquire new phenotypes, including pathogenic and metabolic capabilities and drug resistances. Streptococcus mutans, a major causative agent of dental caries, contains a putative ICE, TnSmu1, integrated at the 3' end of a leucyl tRNA gene. We found that TnSmu1 is a functional ICE, containing all the genes necessary for ICE function. It excised from the chromosome and excision was stimulated by DNA damage. We identified the DNA junctions generated by excision of TnSmu1, defined the ends of the element, and detected the extrachromosomal circle. We found that TnSmu1 can transfer from S. mutans donors to recipients when co-cultured on solid medium. The presence of TnSmu1 in recipients inhibited successful acquisition of another copy and this inhibition was mediated, at least in part, by the likely transcriptional repressor encoded by the element. Using microscopy to track individual cells, we found that activation of TnSmu1 caused an arrest of cell growth. Our results demonstrate that TnSmu1 is a functional ICE that affects the biology of its host cells.

Full text

652
|
Molecular Microbiology. 2022;118:652–669.wileyonlinelibrary.com/journal/mmi
1 | INTRODUCTION
Horizontal gene transfer (HGT) is a driving force in microbial evolu-
tion, allowing bacteria to acquire new traits and phenot ypes from
other bac terial lineages. Biofilms, including dental plaque, are hot
spots for HGT, and HGT is well documented in the oral microbiome
(Jones et al., 2021; Lunde et al., 2021; Olsen et al., 2013; Roberts
et al., 1999, 2001). Fur ther, oral bacteria can cause major health
issues. For example, Streptococcus mutans, a major causative agent
of dental caries, acts as a reservoir for antibiotic resistance genes
and mobile genetic elements within the oral microbiome and can
be a causative agent of infective endocarditis (Lunde et al., 2021;
Nomura et al., 2020; Olsen et al., 2013).
Much is known about quorum sensing and HGT through natural
competence in S. mutans (reviewed in Shanker and Federle [2017]).
Other types of HGT are mediated by mobile genetic elements that
are often found integrated in the genome of the host organism.
These can have a broad host range and some can mediate genetic
Received: 23 August 2022
|
Revised: 14 October 2022
|
Accepted: 17 October 2022
DOI : 10.1111/mmi.14992
RESEARCH ARTICLE
TnSmu1 is a functional integrative and conjugative element
in Streptococcus mutans that when expressed causes growth
arrest of host bacteria
Lisa K. McLellan | Mary E. Anderson | Alan D. Grossman
This is an op en acces s article unde r the terms of the Creative Commons Attribution- NonCommercial License, which permits use, distribution and reproduction
in any medium, provided the origina l work is properly cited and is not used for comme rcial purposes.
© 2022 The Authors. Molecular Microbiology published by John Wiley & So ns Ltd.
Department of Biology, Massachusetts
Instit ute of Technology, Cambridge,
Massachusetts, USA
Correspondence
Alan D. Grossman, Depar tment of Biolog y,
Building 68- 530, Massachusetts Institute
of Technolog y, Cambridge, MA 02139,
USA.
Email: adg@mit.edu
Funding information
National Institute of General Medical
Science s, Gra nt/Award Number: R35
GM122 538
Abstract
Integrative and conjugative elements (ICEs) are major drivers of horizontal gene trans-
fer in bacteria. They mediate their own transfer from host cells (donors) to recipients
and allow bacteria to acquire new phenotypes, including pathogenic and metabolic
capabilities and drug resistances. Streptococcus mutans, a major causative agent of
dental caries, contains a putative ICE, TnSmu1, integrated at the 3′ end of a leucyl
tRNA gene. We found that TnSmu1 is a functional ICE, containing all the genes neces-
sary for ICE function. It excised from the chromosome and excision was stimulated by
DNA damage. We identified the DNA junctions generated by excision of TnSmu1, de-
fined the ends of the element, and detected the extrachromosomal circle. We found
th a t TnSmu1 can transfer from S. mutans donors to recipients when co- cultured on
solid medium. The presence of TnSmu1 in recipients inhibited successful acquisition
of another copy and this inhibition was mediated, at least in part, by the likely tran-
scriptional repressor encoded by the element. Using microscopy to track individual
cells, we found that activation of TnSmu1 caused an arrest of cell growth. Our results
demonstrate that TnSmu1 is a functional ICE that affects the biology of its host cells.
KEYWORDS
conjugation, horizontal gene transfer, integrative and conjugative element, mobile genetic
element, Streptococcus mutans

|
653
MCLELLAN et al.
exchange between distantly related organisms that may be unable to
exchange DNA through transformation. Conjugative elements and
bacteriophage are mobile genetic elements that can mediate their
own transfer from a host (donor) bacterium to a recipient.
Integrative and conjugative elements (ICEs) represent the most
prevalent type of conjugative element and are found in all major
clades of bacteria, including those that cause dental caries and oth-
ers found in the mouth (Guglielmini et al., 2011; Lunde et al., 2021;
Roberts et al., 2001; Roberts & Mullany, 2009, 2011). Integrative
and conjugative elements mediate their own transfer from a host
cell to a recipient and often contain cargo genes that confer ben-
eficial phenotypes to the host cells. Phenotypes conferred by
ICEs include: virulence, symbiosis, metabolic functions, and drug
resistances (reviewed in Johnson and Grossman [2015]). ICEs are
major contributors to the spread of antibiotic resistance deter-
minants among multidrug- resistant pathogens and are notable in
the human microbiome and other microbial communities (Botelho
& Schulenburg, 2021; Jiang et al., 2019). However, fundamental
aspects of ICE biology within the complex communities of dental
plaque remain unknown.
Although their size, regulation, and interactions with host cells
can var y considerably, ICEs generally follow a st andard lifecycle
(reviewed in Wozniak and Waldor [2010]; Auchtung et al. [2016];
Botelho and Schulenburg [2021]). Briefly, ICEs t ypically reside inte-
grated in the chromosome of a bacterial host. Under certain condi-
tions or perhaps stochastically, ICE gene expression becomes active
and the ICE excises from the host chromosome to form an extra-
chromosomal circular dsDNA molecule. This is then processed by
ICE- and host- encoded proteins to generate a nucleoprotein com-
plex containing linear single- stranded ICE DNA. Element DNA (as
linear ssDNA) can then be transferred to a new host through the
element- encoded conjugation machinery, a type IV secretion system
(T4SS). Once in the recipient, the linear ssDNA circularizes and is
converted into dsDNA . The circular dsDNA can then be integrated
into the recipient genome through the action of an ICE- encoded
recombinase, generating a stable transconjugant (recipient that has
stably acquired a copy of an ICE).
While integrated in the host chromosome, expression of most
of the genes required for the ICE lifecycle is repressed (Auchtung
et al., 2005, 2007; Robert s & Mullany, 2009). In many ICEs, this is
controlled by a transcriptional repressor. Upon activating signals, the
repressor is inactivated, sometimes via proteolytic cleavage by an
element- encoded anti- repressor (Bose et al., 2008). This allows the
genes required for ICE transfer to be expressed and the ICE lifecycle
to continue.
Throughout its lifecycle, the interactions between an ICE and
its host cell are complex. ICEs often benefit their host cells through
associated cargo genes that confer specific phenotypes (men-
tioned above). However, ICEs can also manipulate host develop-
ment, growth, and viability (Beaber et al., 20 04; Bean, McLellan, &
Grossman, 2022b; Jones et al., 2021; Pembroke & Stevens, 198 4;
Reinhard et al., 2013).
TnS mu1 is a putative ICE found in some isolates of the prototypic
oral pathogen S. mutans. (Ajdic et al., 2002, p. 159; Bi et al., 2012).
TnS mu1 was recognized bioinformatically based on comparisons
to the conjugation machinery and DNA processing machinery en-
coded by two well- studied ICEs: Tn916, the first described ICE dis-
covered through its ability to spread tetracycline resistance through
clinical isolates of Enterococcus (Franke & Clewell, 1981a; Franke &
Clewell, 1981b, p. 916), and ICEB s1, a well- studied ICE in Bacillus sub-
tilis (Auchtung et al., 2007, 2016; Jones et al., 2021; Lee et al., 20 07;
Lee & Grossman, 20 07).
We found that TnSm u1 is a functional ICE that is capable of un-
dergoing a complete ICE life cycle. We demonstrate that TnSmu1
can excise from its host chromosome to form a circular dsDNA mol-
ecule, transfer to recipient cells, and integrate into the chromosome
to generate stable transconjugants. Using PCR- based assays, we
identified the integration site, defined the ends of the element, and
detected the extrachromosomal circle. We found that DNA damage
modulates the excision of TnSmu1. Further, by co- culturing donors
and recipients on a solid medium, we found that TnSmu1 can transfer
from S. mutans donors to S. mutans recipients. However, the pres-
ence of TnSmu1 in recipients prevented successful acquisition of
another copy of the element, at least in par t, through a repressor-
mediated immunity mechanism analogous to that of ICEBs1
(Auchtung et al., 2007; Bose et al., 2008). Using a fluorescent re-
porter in TnSmu1, we were able to visualize individual cells in which
TnS mu1 had become active. Most of the cells with an active element
stopped growing. Together, our results illuminate the functionality
of TnSmu1 and demonstrate that TnSm u1 affects the physiology of
its host cells by halting host cell growth. As our work was nearing
completion, an independent study identified the putative repressor
of TnSmu1 and reported that the element can be activated to excise
from the chromosome by conditional inac tivation of the repressor
and also by DNA damage (King et al., 2022).
2 | RESULTS
2.1 | Comparison of TnSmu1 to Tn916 and ICEBs1
TnS mu1 (Figure 1) is a putative ICE found in the protot ypic oral path-
ogen S. mutans UA159 (Ajdic et al., 2002, p. 159; Bi et al., 2012).
TnS mu1 was recognized bioinformatically based on comparisons
to genes from the ICE Tn916 that are needed for formation of
the conjugation machinery (a T4SS) and DNA processing (Table 1)
(Ajdic et al., 2002; Rober ts & Mullany, 2009, p. 916; Roberts &
Mullany, 2011, p. 916).
S. mutans appears to be the primary host of TnSmu1, although
remnants or fragments appear to be present in other Streptococcal
species. Using cblaster (Gilchrist et al., 2021) to search the NCBI
BLASTp database, we found that at least 30 sequenced S. mutans
genomes appear to contain an intact TnSm u1, based on the presence
of all the genes needed for a functional ICE and the presence of the

654
|
MCLELLAN et al.
regulatory genes found in TnSmu1 (Figure S1). We did not detect
in t act Tn Sm u1 in other sequenced genomes outside of S. mutans.
We also searched the whole genome shotgun (WGS) nucleotide
collection for DNA sequences similar to TnSmu1. Parts of TnSmu1
were found in 114 of 364 sequences of S. mutans isolates. Of these,
30 of these sequences had >90% identity and query coverage, con-
sistent with the analysis using cblaster. Of note, matches between
TnS mu1 and DNA from bacteria outside of the Streptococci were
rare and of poor quer y coverage (≤5%), indicating that TnSmu1 is a
Streptococcus- specific element. Despite the apparently limited nat-
ural host range, we suspect that TnSmu1 might readily function in
other bacterial species (see Discussion).
In addition to similarity to genes in Tn916, many genes in TnSmu1
are similar to those in ICEBs1 from B. subtilis (Table 1) (Auchtung
et al., 2016). We analyzed sequence similarities bet ween genes in
TnS mu1, ICEBs1, and Tn916 and found that TnSmu1 is predicted to
contain most, if not all of the components that are needed for the
ICE lifecycle (Figure 1, Table 1).
2.1.1 | Type IV secretion system (T4SS)
We identified all the known components of the T4SS typical of
Gram- positive bacteria. The membrane channel of the T4SS is likely
composed of ConB, ConC, ConD, and ConG (ConBICEBs1, ConC
ICEBs1, ConD ICEB s1, and ConG ICEB s1 and ORF13Tn916, OR F19Tn 916,
ORF17 Tn9 16, and ORF15Tn 916, for ICEBs1 a n d Tn916, respectively)
(Auchtung et al., 2016; Ciric et al., 2013; Leonetti et al., 2015;
Roberts & Mullany, 2009). All of these proteins are predicted to en-
code a specific number of transmembrane domains, and the approxi-
mate size of the proteins is conserved throughout T4SSs (Auchtung
et al., 2016; Berkmen et al., 2 010; Leonetti et al., 2015). Based on
FIGURE 1 Genetic map of TnSmu1. Open reading frames are indicated by horizontal arrows. Gene names are abbreviated to include only
the number designation (i.e., 191c indicates smu _191c). The name of the homologous ICEBs1 gene is written below. Thick ver tical black lines
indicate attachment sites attL and attR at the ends of TnSm u1. The putative origin of transfer (oriT ) is indicated with a vertical arrow. Putative
gene function is indicated by pattern and color: Genes of unknown function (white), genes encoding components of the type 4 secretion
system (T4SS) (gray), DNA processing (diagonal stripes), and regulation (black). There is a toxin (grid pattern) and small RNA antitoxin (wavy
horizontal arrow) encoded in the region between s m u_217c and smu_218c (Koyanagi & Lévesque, 2013).
TABLE 1 Similarity between ICEs TnSmu1, ICEB s1, an d Tn916
TnS mu1 gene1Proposed name2
% similarity (% identity) 3
Vir homolog/function/comments4
ICEBs1 Tn916
smu_191c int int 35 . 8 (19. 0) int 37.6 ( 21. 5) tyrosine recombinase
smu_193c xis xis 31.6 (14.5) xis 34.2 (15.1) excisionase/recombination directionality
factor
smu_196c cwlT cwlT 25. 6 (17.5) or f14 3 2 .7 (19. 3) VirB1- like, cell wall hydrolase
sm u_197c conG conG 26. 0 (15.1) or f15 29.2 (15. 8) VirB6- like, transmembrane protein
smu_198c conE conE 46.2 (28.1) or f16 38.5 (22.7) VirB4- like, AA A+ ATPa se
smu_19 9c conD conD 33.2 (16 .8 ) orf 17 24.7 (15.9) VirB3- like, transmembrane protein
smu_200c conC conC 26. 0 (12 .5 ) o rf19 29.2 (1 9.8) transmembrane protein
smu_201c conB conB 37.0 (21.4) or f13 36.8 (18.4) VirB8- like, transmembrane protein
sm u_2 07c nicK nicK 40.4 (24.2) orf20 46.5 (25.6) relaxase
smu_208c conQ conQ 44.2 (26.7) or f21 42.6 (26.7) VirD4- like, AA A+ ATPase; coupling protein
smu_ 209c helP helP 40.8 (24.3) orf22 or f23 38.3 (22.7) 30.8 (22.6) helicase processivity factor
smu _218c immR immR 51.9 (31.3) n/a repressor; DNA binding protein
smu_ 219c immA immA 30 .2 (16.1) n/a anti- repressor; metalloprotease, cleaves ImmR
1TnSmu1 gene names from (Ajdic et al., 20 02).
2Proposed name based on similarity to known genes. Uses ICEBs1 gene names where appropriate based on conservation (Auchtung et al., 2016).
3Calculated by EMBOSS Needle pairwise global sequence alignment (Needleman & Wunsch, 1970).
4Function based on similarity to characterized proteins (Auchtung et al., 2016; Fronzes et al., 2009).

|
655
MCLELLAN et al.
homology or similarity, size, and predicted transmembrane localiza-
tion, we predict that smu _201c, smu_200c, smu _199c, and smu _197c
encode ConBSmu, ConCSmu, ConDSmu, and ConGSmu, respectively
(Table 1).
ConE (ConEICEBs1 and ORF16Tn916) is one of two conserved
ATPases found in T4SSs of Gram- positive bacteria (Auchtung
et al., 2016; Fronzes et al., 2009) that are required for conjuga-
tion (Auchtung et al., 2007; Berkmen et al., 2010; Iyer et al., 2004;
Leonet ti et al., 2015). smu _198 c enco des a gene p r o d u c t th a t is 46. 2 %
similar to ConEICEBs1 and 38.5% similar to ORF16Tn916 (Table 1), and
thus we infer SMU_198c is ConESmu.
The coupling protein, ConQICEB s1 and ORF21Tn916, binds the
DNA protein complex (relaxasome) and “delivers” or “couples” it
to the T4SS for transfer into a recipient cell. Key features of cou-
pling proteins from T4SSs of Gram- positive bacteria include two
transmembrane helices in the N- terminal domain and a C- terminal
cytoplasmic ATPase domain (Alvarez- Martinez & Christie, 2009;
Auchtung et al., 2016). ConQSmu (SMU_208c) shares 44.2% and
42.6% similarity with ConQICEBs1 and ORF21Tn916, respectively, and
is predicted to have the same structural features found in other cou-
pling proteins (Table 1).
2.1.2 | Cell wall hydrolase
Cell wall hydrolases are critical components of conjugative ele-
ments from Gram- positive bacteria (Auchtung et al., 2016; Bhatty
et al., 2013; DeWitt & Grossman, 2014). SMU_196c contains a phage
tail lysozyme domain (E- value 2.20 e- 30 BL ASTp) used by phage to
degrade the cell wall peptidoglycan layer (Xiang et al., 2008). It also
includes an amidase domain (E- value 1.23 e- 48 BLASTp). Structural
predictions through Phyre2 matched the structure of SMU_196c
to a cell wall hydrolase produced by Staphylococcus aureus (N-
acetylmuramoyl- L- alanine amidase 2 from S. aureus, 100% confi-
dence, 76% coverage). Therefore, we determined this is likely the
cell wall hydrolase for TnSm u1 and we infer SMU_196c is CwlTSmu.
2.1.3 | DNA relaxase, origin of transfer (oriT), and
helicase processivity factor
After excision from the chromosome, the circular dsDNA is
nicked and unwound for a single strand of DNA to be transferred
through the T4SS. Nicking occurs at the origin of transfer (oriT)
through the action of a DNA relaxase (or nickase) (NicKICEBs1 and
ORF20Tn916 ) (Auchtung et al., 2016; Lee & Grossman, 2007; Rocco
& Churchward, 2006). At least several ICEs undergo autono-
mous rolling- circle replication (Johnson & Grossman, 2015; Lee
et al., 2010; Wright & Grossman, 2016). Like conjugation, replication
initiates from oriT by the relaxase and also requires unwinding of the
duplex DNA. In B. subtilis, DNA unwinding is mediated by the host-
encoded DNA translocase PcrA (Lee et al., 2010; Petit et al., 19 98;
Thomas et al., 2013). For ICEBs1 and Tn916, a helicase processivity
factor (HelPICEBs1 and ORF22 and ORF23 for Tn916) is required to
facilitate DNA unwinding (Thomas et al., 2013). The genes encoding
the processivity factors are immediately upstream of and in a cluster
with conQ (coupling protein), oriT, and nicK (relaxase), in that order
(Thomas et al., 2013) (Figure 1).
(i) nicKSmu. smu _ 207c i n TnSmu1 encodes a product that is 40.4%
and 46.5% similar to NicKICEBs1 and ORF20Tn916, respectively
(Table 1). Similar to the location in ICEBs1 a n d Tn916, it is imme-
diately downstream from the gene encoding the coupling pro-
tein (Figure 1). Based on the sequence similarit y and location, we
infer that SMU_207c is NicKSmu.
(ii) oriTSmu. We identified a sequence in TnSmu1 that is identical
in 20 of 24 bp to oriT from ICEBs1 an d Tn916 (Figure 2a) (Lee
& Grossman, 2007). oriTSmu contains an inverted repeat that is
characteristic of many oriTs and is located just upstream of nicK
(Figure 1). Based on the sequence and location, we infer that this
is oriTSmu.
(iii) helPSmu. smu _209c in Tn Sm u1 encodes a product that is 40.8%
similar to the helicase processivity factor HelPICEBs1 and 38.3%
and 30.8% similar to ORF22 and ORF23, the two HelP homologs
fr o m Tn 916 (Table 1). Additionally, it is located immediately up-
stream from the gene (conQ) that encodes the coupling protein
(Figure 1). Based on the location and similarities, we infer that
SMU_209c is HelPSmu.
2.1.4 | Integrase and excisionase
Integrating elements typically utilize a recombinase, often called an
integrase (Int), that is required for integration into and excision from
a host chromosome. Integrating elements also utilize a recombina-
tion directionality factor, also called an excisonase (Xis), in combi-
nation with Int, for excision from the host chromosome (Grindley
et al., 2006; Hirano et al., 2011).
smu_191c encodes a product that is 35.8% and 37.6% similar to
the tyrosine recombinases (Int) encoded by ICEB s1 a n d Tn 916, re-
spectively (Table 1). Integrases (recombinases) are often encoded at
or near an end of an ICE (Cur y et al., 2017). Based on the location and
protein similarities, we infer that SMU_191c is IntSmu.
smu_193c encodes a gene product that is 31.6% similar and
34.2% similar to XisICEBs1 and XisTn916 ( Table 1). Excisionases (Xis) are
typically small, highly charged proteins often with lit tle sequence
similarity to other excisionases {e.g., (Auchtung et al., 2016; Lee
et al., 20 07)}. smu_193c encodes is a small, basic protein and thus we
infer that SMU_193c is likely XisSmu .
2.1.5 | Regulatory proteins (ImmA/ImmR)
TnS mu1 has two genes, smu_ 218c and smu _219c, that encode prod-
ucts similar to the repressor (ImmR) and anti- repressor and metal-
loprotease (ImmA) encoded by ICEBs1. SMU_218c and SMU_219c

656
|
MCLELLAN et al.
are 51.9% and 30.2% similar to ImmRICEBs1 and ImmAICEBs1, respec-
tively (Figure 2b,c), and we infer that the TnSmu1 products are
ImmRSmu (SMU_218c) and ImmASmu (SMU_219c). For ICEBs1, when
activated, the protease ImmA cleaves ImmR to inactivate it, thereby
causing de- repression of the element (Auchtung et al., 2007; Bose
et al., 2008). Homologs of ImmR and ImmA are also encoded by many
phages (Bose et al., 2008; Lucchini et al., 1999). ImmRSmu contains
a conserved a helix- turn- helix motif, a typical DNA- binding domain
for phage- like repressors (Bose et al., 2008). It also has a conserved
phenylalanine that may be the putative cleavage site (Figure 2b) (Bose
et al., 2008). That ImmRSmu acts as the repressor of TnSm u1 is fur-
ther supported by finding that it appears essential in Tn- seq studies
(Shields et al., 2018), as would be expected if deletion of the gene
caused de- repression and excision of the element, thereby deleting
the entire element from the chromosome. Further, ImmASmu contains
a characteristic HEX XH motif found in many zinc- dependent metal-
loproteases (Figure 2c) (Fujimura- Kamada et al., 19 97).
2.1.6 | Genes in TnSmu1 with unknown function
There are 15 genes in TnSm u1 (smu_194c, smu_195c, smu_202c,
smu_204c, smu_205c, smu_206c, smu_210c, smu_211c, smu_212c,
smu_213c, smu_214c, smu_215c, smu_216c, smu_217c, smu_220c)
that lack similarit y to any genes known to be needed for the ICE
lifecycle. Of these, sm u_194c is predicted to encode a protein with
a DUF3850 domain (e.g. S. mutans DUF3850 domain- containing
protein, Sequence ID: WP_019319579.1, E- value 1 e- 53) and has
some similarity to proteins predicted to be involved in chromosome
partitioning (e.g. Myoviridae sp. MAG TPA: chromosome partition-
ing protein, E- value 2 e- 20). Proteins with these properties are found
in a wide range of bacterial species. smu _194 c and sm u_195c gene
products are also similar to proteins encoded by various bac te-
riophage. In contrast to sm u_194c and smu_195c , the other genes
with unknown function in TnSm u1 are similar to genes found only
in Streptococcal species. Based on the lack of similarity to genes or
FIGURE 2 Alignments of oriT, ImmR,
and ImmA from TnSmu1, ICEBs1, and
Tn916. (a) Alignment of the putative
TnS mu1 origin of transfer (oriT) to ICEBs1
an d Tn916 . The nic site of ICEBs1 and
Tn916 is indicated by a vertical arrow.
Inverted repeats are indic ated by lines
under the sequence. (b and c) Global
alignments of ImmR (b) and ImmA (c)
from ICEBs1 a nd Tn Sm u1 as calculated
by EMBOSS Needle pair wise alignment
(Needleman & Wunsch, 1970). “|”
represents a matching amino acid; “:”
represents amino acids with strongly
similar properties; “.” represents amino
acids with weakly similar properties; “- ”
represents a gap (Rice et al., 2000). (b)
For ImmR, amino acids boxed in green
represent a conserved helix- turn- helix
motif (Bose et al., 2008; Oppenheim
et al., 2005) (predicted by GYM 2.0;
Narasimhan et al., 2002). Amino acids
boxed in orange represent the cleavage
site where ImmRICEBs1 is cleaved by
ImmAICEBs1 (Bose et al., 20 08). (c) For
ImmA, amino acids boxed in blue indicate
the characteristic HEXXH motif found in
many zinc- dependent metalloproteases
(Fujimura- Kamada et al., 1997).

|
657
MCLELLAN et al.
proteins with known function, we do not have insights into the func-
tions of the unknown genes in TnSmu1.
Based on the presence of the regulatory genes and the essential
components of an ICE, we expec ted that TnSmu1 is a functional ICE.
Additionally, based on the presence of several genes of unknown
function, we suspect that TnSmu1 confers one or more phenotypes
to host cells, although the nature of these phenotypes is currently
not known. Below we describe experiments demonstrating that
TnS mu1 is indeed a functional element: it can excise from the host
chromosome, transfer to a new host, and integrate into the chromo-
some of the new host to generate a stable transconjugant.
2.2 | Excision of TnSmu1 from the
host chromosome
When TnSmu1 is integrated in the host cell chromosome, there are
left and right attachment sites, attL and attR, respectively, that de-
marcate the junctions between TnSm u1 and the host chromosome
(Figure 3a). If capable of and upon excision, there would be a sin-
gle attachment site in the element, attTnS mu1, where the element
recombined to form an extrachromosomal circle. There would also
be an attachment site in the bacterial chromosome, attB, that rep-
resents the fusion of chromosomal sequences that had been inter-
rupted by insertion of the element (Figure 3a).
attL had been predicted based on the increase in A + T content
through out the element t hat is character istic of horizonta lly acquired
DNA and on the presence of nearby genes encoding putative recom-
binases (smu_191c) (Ajdic et al., 2002, p. 159; Bi et al., 2012). We
noticed that TnSmu1 attL was in, although it did not disrupt, smu_t33,
encoding a leucyl tRNA. ICEs are often found integrated into tRNA
genes and t ypically do not disrupt the gene (Burrus et al., 20 02;
Burrus & Waldor, 2004). For example, ICEBs1 and ICEH in10 56 from
B. subtilis and Haemophilus influenzae, respectively, are both found
integrated in a tRNA gene (Dimopoulou et al., 2002; Lee et al., 2007).
The location of attR o f TnSmu1 had been predicted to be either after
smu_209c or smu _226c (Figure 1), based on A + T content, the pres-
ence of genes encoding an integrase, relaxase, and- or type IV secre-
tion system, and- or the presence of a large intergenic region (Ajdic
et al., 2002, p. 159; Bi et al., 2012).
To test TnSmu1 excision, we designed primer set s upstream
and downstream of the predicted ends (attL and attR) that would
only produce a PCR product upon excision if attR of Tn Sm u1 was
indeed af ter smu_209c or smu_ 226 c. Essentially, primers located
outside the element attR and attL cannot produce a PCR product
when TnSmu1 is integrated. This is due to the fact that TnSmu1 is
~20 kb and a PCR product of this size would not be generated in the
PCR conditions used. However, upon excision of TnSmu1, PCR of
genomic DNA using primers that are outside but near the ends of
the element (primers A and D in Figure 3a) should generate a readily
detectable product (Figure 3a). Using genomic DNA from stationar y
phase cultures of S. mutans, we were unable to detect PCR products
using these primers. This result indicates that either TnSmu1 did not
excise (or was below the limit of detection) or that at least one of
the primers was either within TnSmu1 (and not chromosomal DNA)
or was too far from the actual recombination sites (attL and attR). If
TnS mu1 had excised, then the predic ted right end is likely not near
smu_209c or smu _226c .
We also designed primers internal to TnSmu1 but oriented
outwards to detect the circularized element (primers B and C in
Figure 3a) following excision. PCR with this set of primers would
generate a product only if the element excised and both primers are
internal to and near the ends of the element. Using genomic DNA as
a template (as above), we did not detect a product from PCR using
these two primers. This indicates that either TnSmu1 had not ex-
cised (or was below the limit of detection), or that the primers were
not internal to and near the ends of the element. If TnSmu1 had
excised, then the predicted right end is likely not near smu_209c or
smu _226c .
We found that this inability to detect excision of TnSmu1 with
the initial primer sets was because attR was dif ferent from what
was postulated initially and not because the element failed to ex-
cise. We tested other primer pairs, keeping the primer near attL
constant and changing the location of the primers near the putative
attR. We identified PCR primers that produced products that, when
sequenced, allowed us to define attL, attR, attB, and attTnSmu1
FIGURE 3 Cartoon of TnSmu1 inserted in the chromosome
and products and sequences after excision. (a) Cartoon of TnSmu1
inserted in the chromosome and the product s after excision. A set
of four primers (labeled A, B, C, D) are able to detect the junctions
between the host chromosome and lef t (attL; primers A + B) and
right (attR; primers C + D) ends of TnSmu1; the host attachment site
without insertion of TnSmu1 (attB; primers A + D) and the excised
TnS mu1 circle (attTnSmu1; primers B + C). (b) DNA sequence of
attB and attTn Sm u1. The predicted 17 bp recombination site is in
bold. The 3′ end of the leucyl tRNA (smu _t33) that overlaps the
recombination site is highlighted in gray.

658
|
MCLELLAN et al.
(Figure 3). Based on these sites, we conclude that TnSm u1 is an
approximately 20 kb element that extends from open reading
frames smu _191c through smu_220c (Figure 1). The sequences of
the various sites also allowed us to identify a 17 bp sequence pres-
ent in both attB and attTn Smu1 that is likely the recombination site
used by the t yrosine recombinase encoded by int (smu _191c) in
TnS mu1 (Figure 3b). This is consistent with excision noted by King
et al. (2022) when examining read coverage in an immRSmu mutant.
Together, these data demonstrate that TnSmu1 is capable of exci-
sion from the chromosome of host cells and that a small proportion
of cells had “spontaneously” activated TnSmu1 in the absence of
any exogenous treatment.
2.3 | Tn Smu1 excision increases in response to
DNA damage and on solid media
Because TnSmu1 has homologs of immA and immR from ICEBs1
(Figure 2b,c, Table 1) and ICEBs1 is de- repressed by DNA damage,
we postulated that DNA damage might also de- repress TnSm u1.
Therefore, we measured excision of TnSmu1 after addition of mito-
mycin C (MMC) to cells to induce DNA damage. We purified DNA
from cells treated and untreated with MMC and performed qPCR
to detect the presence of attB (generated upon excision of TnSmu1).
We normalized excision to a nearby chromosomal locus (ilvB). We
found that TnSmu1 had excised in ~0.1%– 0.2% of cells (~1 – 2 × 10−3
attB/ilvB) after treatment with MMC for 2– 4 h. This is ~20– 30 - fol d
greater than that in the absence of MMC (~0.006%, ~6 × 10 −5 attB/
ilvB) (Figure 4a). These results indicate that excision of TnSm u1 is
induced following DNA damage. We suspect that other conditions
that cause DNA damage (e.g., oxidative stress that occurs in the oral
cavity) likely cause induction of TnSm u1 in a subset of cells.
It also appeared that TnSmu1 underwent autonomous replica-
tion following excision. Both ICEBs1 a n d Tn916 undergo autonomous
rolling- circle replication following excision (Lee et al., 2010; Wright &
Grossman, 2016). This is mos t easily detected by m easuring the rela tive
copy number of the circular element (attTnS mu1) to that of the empty
chromosomal site (attB). Using qPCR to measure attTnS mu1 and attB,
we found that during exponential growth, there were approximately
2– 5 extrachromosomal copies of TnSmu1 per excision event (attB), in-
dicating that TnSmu1 is capable of autonomous replication, although
the copy number was low. By analogy to ICEBs1 and Tn916, TnSmu1
most likely undergoes rolling- circle replication that initiates from the
origin of transfer (oriT ) using the element- encoded relaxase (NicK).
FIGURE 4 TnSmu1 excision increases in response to DNA
damage and on solid medium. Excision of TnSmu1 was measured
by qPCR to detect attB (primers corresponding to A + D indicated
in Figure 3). The proportion of cells containing excised TnSm u1
was calculated by normalizing attB to a nearby chromosomal locus
(ilvB). Data presented are averages from three or more independent
experiments with error bars depicting standard error of the mean.
Error bars could not always be depicted due to the size of each
data point. (a) S. mutans UA159 cells were grown in liquid (TH)
medium for 7 h. After 3 h of growth, cells were either left untreated
(black circles) or treated with 1 μg/mL mitomycin C (MMC; open
circles; time of addition indicated by black arrow below the x- axis).
Samples were harvested at the indicated times to measure excision.
(b) S. mutans strains UA159 (TnSmu1; black circles) and LKM145
(TnSmu1- tet; gray triangles) were grown in TH liquid medium to
mid- exponential phase, pelleted, and resuspended at a low density.
Cells were then spotted and grown on TH solid medium for 1, 3,
5, or 7 days in anaerobic conditions and samples were taken at
the indicated times (days) to measure excision. (c) S. mutans strain
LKM145 (TnSmu1- tet) was grown in TH liquid medium to mid-
exponential phase, pelleted, and resuspended at a low density.
Cells were then spotted and grown on BHI solid medium for 1, 3,
5, or 7 days in anaerobic conditions and samples were taken at the
indicated times (days) to measure excision.

|
659
MCLELLAN et al.
As the typical lifecycle of S. mutans is within biofilms and on the
solid surface of teeth, we sought to determine if TnSmu1 was capa-
ble of excision when cells were grown on a solid surface. S. mutans
cells were grown to mid- exponential phase in TH medium, pelleted,
and resuspended at a low cell density. Cells were then spot ted onto
solid TH medium for 1, 3, 5, or 7 days in anaerobic conditions. We
found that over time, excision increased ~10 to 100- fold on solid
surfaces compared with liquid culture (Figure 4b).
We also compared wild type S. mutans (UA159) with LKM145
(a derivative of UA159 that contains TnSmu1- tet), and found no dif-
ference in excision of the element between strains in either liquid
or on solid TH media (Figure 4b). We also found that growth in TH
liquid medium and plating on BHI solid medium resulted in a simi-
lar increase in excision of TnSmu1- tet (Figure 4c), indicating that the
change in growth medium did not impact excision.
Together, our results indicate that TnSmu1 is activated when
cells are grown on a solid surface, perhaps analogous to the growth
in plaque on teeth. This increased activation could also be due to
some additional stress experienced during growth on solid sur faces,
perhaps resulting in an increase in DNA damage and thus increased
activation.
We anticipate that the mechanism of TnSm u1 de- repression
is analogous to that for ICEBs1. In the presence of DNA damage-
inducing conditions, the ICEBs1- encoded metalloprotease
ImmAICEBs1 cleaves the repressor ImmRICEB s1, thereby causing de-
repression of element gene expression (Auchtung et al., 2007; Bose
& Grossman, 2011). Based on the similarities bet ween these two
ICEBs1 an d TnSmu1 regulators, the simplest model is that during
DNA damage, ImmASmu becomes activated to cleave and inactivate
ImmRSmu, thereby causing de- repression of TnSmu1 gene expres-
sion. We also note that unlike that for ICEB s1 (Auchtung et al., 20 07;
Bose & Grossman, 2011), there are no indications that TnSm u1 is
regulated by population densit y, peptide signaling, or quorum sens-
ing as culture densit y and grow th phase did not affect excision fre-
quency (Figure 4a) and there are no genes in TnSm u1 with similarity
to known cell– cell signaling systems.
2.4 | TnSmu1 can transfer to S. mutans recipients
that lack a copy of the element
Based on the presence of an apparently intact set of genes for a
T4SS and the ability of TnSm u1 to excise, we tested for the ability of
the element to transfer from one cell to another. To monitor trans-
fer of TnSmu1, we constructed S. mutans donor and recipient strains
that could be distinguished by their unique antibiotic resistances and
growth requirements.
2.4.1 | Donors
We introduced a gene conferring tetracycline resist ance (tet) into
TnS mu1 between smu _210c and smu _211c (Figure 1), generating
TnSmu1- tet (strain LKM145). This insertion is in a region of TnSmu1
with unknown function that we anticipated would not interfere with
the typical ICE lifecycle. Indeed, excision of TnSmu1- tet was similar
to that of wild t ype TnSm u1 (Figure 4b).
We also made a mutant of S. mutans that requires D- alanine for
growth due to a null mut ation in alr (alr::erm), which encodes alanine
racemase needed for production of D- alanine (Wecke et al., 1997).
Use of an alr mutant as a donor allows us to distinguish donors from
recipients and transconjugants after the mating because donors
will be auxotrophic and unable to grow on media without addi-
tion D- alanine. Donors were also defective in genetic competence
(∆comS::kan) (Mashburn- Warren et al., 2010) to prevent transforma-
tion with DNA from recipients.
2.4.2 | Recipients
We thought it important to use a recipient that was cured of
TnS mu1 (TnSmu1 0). Some ICEs and temperate phages have im-
munity and exclusion mechanisms that reduce acquisition of ad-
ditional copies of the cognate element (Auchtung et al., 2007;
Gottesman & Weisberg, 2004; Oppenheim et al., 2005; Serfiotis-
Mitsa et al., 2008). Notably, ICEBs1 has repressor- mediated immu-
nity (Auchtung et al., 20 07). Because TnSm u1 encodes a homolog of
ImmRICEBs1, we were concerned that cells containing TnSm u1 might
also have repressor- mediated immunit y. Therefore, we used two dif-
ferent recipient strains, one without (denoted ∆TnSmu1 or Tn Smu10)
and one with TnSmu1 (LMK85 and LMK87, respectively). Of note, we
found there was no noticeable growth difference in strains with or
without TnSmu1 (Figure S2). Further, both recipients contained a null
mutation in comS (Mashburn- Warren et al., 2010 ) to ensure any DNA
transfer detected was not via transformation into the recipients.
We found that TnSm u1 transferred from S. mutans donors into
S. mutans recipients provided that the recipients lacked TnSmu1
(Figure 5 and Figure S3). Donors and recipients were grown to mid-
exponential phase, pelleted, resuspended at a low cell density, mixed
at a ratio of 1:1 and then spotted onto agar plates and allowed to
grow for 1, 3, 5, or 7 days. The largest number of TnSmu1 transcon-
jugants was detected 3 days post- inoculation of the mating plates:
~104 transconjugants were detected (Figure 5), corresponding to a
mating frequency of ~2 x 10−5 transconjugant s per donor present in
the harvested mating mix. The number of transconjugants increased
between 1 and 3 days post- inoculation and then dropped at later
times (Figure 5). This increase was likely due either to transconju-
gants dividing and producing progeny or transconjugant s becoming
donors and further transferring TnSmu1 to additional recipients.
The drop in the number of transconjugant s detected at later times
was most likely due to overall cell death of donors, recipients, and
transconjugants that occurred throughout the mating (Figure 5b).
We also measured conjugation efficiencies at different donor to
recipient ratios. Transfer was detected at all ratios (1:100, 1:10, 1:1,
10:1, 100:1) and at each of the time points tested (1, 3, 5, 7 days)
(Figure S3). The largest number of TnSm u1 transconjugants was

660
|
MCLELLAN et al.
detected at 3 days post- inoculation of the mating plates at a donor
to recipient ratio of 1:1, and we used these conditions for most
experiments.
In contrast to the results with TnSmu1- cured recipients, we
detected few, if any, transconjugants in matings with recipients
that contained TnSmu1 (Figure 5a and Figure S3). Transconjugants
(as detected by tetracycline resistance and grow th without D-
alanine) were rarely detected, and when they were, we observed
only a single colony, indicating that there were ≤20 total transcon-
jugant CFUs per mating (at or below the limit of detection) at all
time points and donor to recipient ratios tested. This represents a
decrease of ~500- fold compared with isogenic recipients without
TnS mu1 (1:1 donors: recipients af ter 3 days of growth on solid me-
dium). Together, these results show that TnSmu1 is a functional con-
jugative element, capable of transfer from donor to recipient cells.
They also indicate that there is at least one mechanism conferred
by TnSmu1 that inhibits acquisition of another copy of the element.
2.5 | Expression of the repressor, ImmRSmu, is
sufficient to reduce acquisition of TnSmu1
Because TnSmu1 encodes a homolog of the immunity repressor
ImmR from ICEBs1, we postulated that ImmRSmu might also confer
some level of immunity that inhibits acquisition of a second copy
of the element, analogous to that of ICEBs1 (Auchtung et al., 20 07;
Bose et al., 2008). Therefore, we sought to determine if expres-
sion of the repressor encoded by TnSmu1, in the absence of other
TnS mu1 genes, would result in inhibition of (immunity to) acquisition
of TnSmu1.
We found that ectopic expression of ImmRSmu in recipient cells
inhibited acquisition of TnSmu1. We used recipients that were
missing TnSmu1 but expressed immRSmu under its predicted endog-
enous promoter and compared these to recipients with and with-
ou t TnSmu1, analogous to the experiment s described above. We
measured acquisition of TnSmu1 after 3 days of mating at a donor
to recipient ratio of 1:1. We found that expression of immRSmu, in
the absence of other genes from the element, reduced acquisition
of TnSmu1 to a similar extent (~500- fold) as that of the intact ele-
ment (Figure 6). Together, these results indicate that expression of
th e TnSmu1 repressor (ImmR) is sufficient to inhibit acquisition of
the element.
2.6 | The preferred integration site of TnSmu1 is in
a leucyl- tRNA gene
TnS mu1 resides in donor cells at the 3′ end of a leucyl tRNA gene
(smu_t33) (Figure 3b). We sought to determine if TnSmu1 inte-
grated at this same location following transfer to new cells, that is,
in transconjugants, or if it was more promiscuous in site selection,
perhaps analogous to Tn916 (Robert s & Mullany, 2009).
We found that following conjugation, TnSmu1 integrated spe-
cifically in the 3′ end of smu_t33. After performing matings into
∆TnS mu1 recipient cells (LKM85) as described above, we isolated 16
independent transconjugants and tested these for integration into
FIGURE 5 TnSmu1 can transfer to recipient cells that lack a
copy of the element. Donors containing TnSmu1- tet (LKM145)
were co- cultured with recipients containing no TnSmu1 ( TnSmu10;
LKM85) or with TnSmu1 (TnSmu1+; LKM87). Cells were grown
to mid- exponential phase in TH liquid medium, pelleted and
resuspended at a low density, and donors and recipients were
mixed at a ratio of 1:1. Mating mixes were spotted onto BHI solid
medium and incubated under anaerobic conditions for 1, 3, 5, or
7 days. Cells were then har vested and the numbers of donors (tet,
alr, kan), recipients (alr+, spc), and transconjugants (tet, alr+, spc) were
enumerated based on unique phenotypes associated with each
cell type. The limit of detection was 20 transconjugant CFUs per
mating (1 CFU per plate) as we plated a maximum of one twentieth
of the resuspended mating mix. (a) The number of transconjugant
CFUs per mating formed with TnSmu1- tet donors (LKM145) and
TnS mu1 0 recipients (LKM85, white bars, filled circles are individual
data points) or with TnSmu1+ recipients (LKM87, dark gray bars,
triangles are individual data points). Results with TnSmu1+ recipients
were at or below the limit of detection (LOD ≤20 CFU/mating;
dotted line). (b) The number of CFUs for donors (inverted triangles),
recipients lacking TnSmu1 (squares), and transconjugants (circles)
in a mating mix are shown for the experiment in (a). The number of
transconjugants increased between 1 and 3 days post inoculation
and then dropped. The drop in the number of transconjugants at
later times was most likely due to overall cell death that occurred
throughout the mating as seen with a parallel drop in numbers of
donors and recipients. Data presented are averages from three or
more independent experiments. Error bars represent the standard
error of the mean and are generally not visible as they are too small
relative to the size of each data point. Donor and recipient CFUs are
largely indistinguishable in this graph. The dotted line at the bot tom
represents the limit of detec tion for all cell types.

|
661
MCLELLAN et al.
smu_t33 using PCR primers that would detect attL (Figure 3a; Section
2). We found that TnSmu1 had integrated into smu_t33 in all 16 of
the transconjugants tested. This tRNA gene (smu_t33) and the 17 bp
attB are not found anywhere else in the S. mutans genome (Ajdic
et al., 2002). Together, our results indicate that the identified 17 bp
site referred to as attB (Figure 3b) is the preferred site of integration
of TnSmu1. That said, we cannot rule out the possibility that TnSmu1
might integr ate into other sites in the chromosome at a low frequency.
2.7 | Tn Smu1 causes a growth arrest in host cells
Tn916, an ICE related to TnSmu1, causes growth arrest and death
of host cells following excision and expression of its genes (Bean,
McLellan, & Grossman, 2022b). In contrast, ICEBs1, which encodes
a T4SS and regulatory proteins similar to those encoded by TnSmu1,
does not cause death of its host cells (Babic et al., 2011; Bean,
McLellan, & Grossman, 2022b). Therefore, we wondered whether or
no t TnS mu1 manipulated the growth and or viability of its host cells.
As TnSmu1 only excises in a relatively small fraction of cells (Figure 4),
we decided to examine TnSmu1 ac tivation in single cells using fluores-
cence microscopy and a fluorescent reporter that would be indicative
of TnSmu1 gene expression. We inserted gfpmut2 between smu_210
and sm u_211 in Tn Smu1 (Figure 1) to generate TnSmu1- gfp (LKM137).
Cells should fluoresce green only when TnSmu1 genes are expressed.
Insertion of gfp within TnSmu1 did not have a significant impact on
excision: the frequency of excision of TnSmu1- gfp (~0.004%, ~4x10−5
attB/ilvB) during growth in liquid medium was similar to that of wild
typ e TnSmu1 a n d Tn Smu1- tet (Figure 4b).
We found that cells expressing TnSmu1- gfp had a growth de-
fect. Cells containing TnSmu1- gfp were diluted to early exponential
phase, grown for 3 h, and then spot ted onto an agarose pad on a
microscope slide. Cells were visualized and tracked for 3 h, compar-
ing those that had activated TnSmu1- gfp to those that had not . We
tracked 82 cells in which TnSmu1- gfp was activated (GFP on) and
82 neighboring cells in which TnSmu1- gfp was apparently not acti-
vated (GFP off) (Figure 7). Of the cells ac tivating TnSmu1 (GFP on),
94% (77/82) did not undergo any further cell divisions and 6% (5/82)
divided once (Figure 7). In contrast, in the 82 neighboring cells with-
ou t TnSmu1- GFP activated (GFP off), only 4% (3/82) of cells did not
undergo any further cell divisions, and 96% of cells under went 1 or
more cell divisions (79/82) (77/82 vs. 3/82, χ2 = 133.6 4, p < 0.0001).
FIGURE 7 Cells with an active TnSm u1 stop growing. Cells
containing TnSmu1- gfp- tet (LKM137) were grown in TH liquid
medium to late exponential phase. At time 0 h, cells were spotted
onto agarose pads containing TH medium, 0.1 M propidium iodide,
and 0.35 μg/mL DAPI. Cells were monitored by phase contrast and
fluorescence microscopy for 3 h. GFP (green) was produced in cells
in which TnSmu1 was ac tivated and excised from the chromosome.
Propidium iodide (red) indicates cell death. Images shown are a
merge of phase contrast and fluorescence. Three independent
experiments gave similar results and were quantified (described in
results), in total analyzing 82 cells from ~30 individual microscope
frames. Of the 82 cells obser ved with an active TnSmu1- GFP
(GFP on, green), 94% (77/82) did not undergo any further cell
divisions, and 6% (5/82) divided once. Of 82 neighboring cells
that had not activated TnSmu1- GFP (GFP off), only 4% (3/82) of
cells did not undergo any further cell divisions, and 96% (79/82) of
cells underwent one or more cell divisions. A representative set of
images is shown here. DAPI is not shown for visual clarity.
FIGURE 6 Expression of ImmRSmu in recipients inhibits
acquisition of TnSmu1. Tn Smu1- tet donors (LKM145) were mated
with three different recipients: TnSmu1 0 (LKM85, circles, white
bar); TnSmu1+ (LKM87, triangles, dark gray bar); and TnSmu10
that expressed immRSmu from an ectopic site (LKM233, diamonds,
bar with diagonal stripes). Donors and recipient s were grown and
prepared as described for Figure 5, with donors and recipients
mixed at a ratio of 1:1 and transconjugants measured after
growth on the solid surface (mating) for 3 days. Data presented
are averages from four independent experiments with error bars
depicting standard error of the mean. The limit of detection in
these experiments was two transconjugant CFUs per mating (up
to half of the resuspended mating mix was plated). Results with
TnS mu1+ recipients were at the LOD (2 CFU/mating; dotted line).
*p < 0.05.

662
|
MCLELLAN et al.
Although the cells expressing TnSmu1 had a growth arrest, we
found no evidence of cell death. Using propidium iodide (PI) to mon-
itor cell viability, we found that of the 82 cells that had activated
TnS mu1- GFP, only 12% (10/82) became or had progeny that became
PI- positive during the course of the experiment (Figure 7). This was
not significantly different from that of cells without TnSmu1 acti-
vated: 13% of cells became PI- positive or had progeny that became
PI- positive (11/82) throughout the experiment. These numbers are
consistent with S. mutans viability of randomly selected wild type
cells (UA159, 10/82) and cells without TnSm u1 (LKM68, 9/82). This
is also consistent with what is seen in S. mutans biofilms and vari-
ous growth conditions via fluorescence microscopy (Decker, 2001;
Decker et al., 2014; Zhang et al., 2009). Together, these results indi-
cate that cells in which TnSmu1 becomes activated are unable to di-
vide; however, they do not lose viability, at least over a period of 3 h.
3 | DISCUSSION
Our work demonstrates that TnSmu1 is a functional ICE, capable of
excision from the host cell chromosome and transfer to recipient cells.
We found that ac tivation of TnSmu1 is stim ulated by DNA damage and
by growth on a solid sur face. Fur ther, activation of TnSmu1 caused a
growth arrest, indicating that it can have a profound effect on host
physiolog y. Our findings provide a basis for future work examining
the impacts of TnSm u1 on the physiology of S. mutans. Further, as
ICEs can be utilized for genetic engineering purposes (Bean, Herman,
et al., 2022a; Brophy et al., 2018; Miyazaki & van der Meer, 2013;
Peters et al., 2019), Tn Smu1 may perhaps be developed as an engi-
neering tool for manipulation of organisms in the oral microbiome.
3.1 | The TnSmu1 repressor
immRSmu (smu _218c) almost certainly encodes the repressor of
TnS mu1, based on three lines of evidence. First, ImmRSmu is similar to
the repressor of ICEBs1 and others in this family. Second, like other
‘immunity’ repressors, expression in recipient cells (in the absence of
other element genes) caused a reduction in acquisition of a copy of
the element, similar to immunity to superinfection exhibited by vari-
ous temperate phages (Gottesman & Weisberg, 2004; Oppenheim
et al., 2005) and ICEBs1 (Auchtung et al., 2007; Bose et al., 2008).
La stly, smu_218c (immR) appears to be an essential gene based on
Tn- seq experiments with S. mutans (Shields et al., 2018). Repressors
of mobile genetic elements that integrate and excise (e.g., ICEs and
temperate phages) can appear to be ‘essential’ because their loss can
result in excision and loss of the element in which they are encoded.
For example, if loss of a repressor leads to excision of the element,
that element will likely be lost from the population of cells. This loss
makes it very difficult to establish a null mutation in the gene for
the repressor as the loss of function mutation in the repressor gene
will be lost along with the resulting excised element. In this way,
genes potentially encoding element repressors can be identified in
genome- wide screens and appear as “essential” even though they re-
side in an element that itself is not essential. This point is highlighted
by the fact that we were able to delete TnSmu1 and cells are viable,
but the repressor appears to be essential (Shields et al., 2018).
3.2 | Costs and benefits to cells carrying an ICE
ICEs are of ten double- edged swords to t heir host cells: They can pro-
vide both fitness costs and benefits. They can benefit their host cells
through associated cargo genes that confer specific phenot ypes,
such as antibiotic resistances, metabolic traits, and virulence factors.
However, some ICEs can manipulate host development, growth, and
viabilit y (Beaber et al., 2004; Bean, McLellan, & Grossman, 2022b;
Jones et al., 2021; Pembroke & Stevens, 198 4; Reinhard et al., 2013).
This is similar to plasmids, which are known to provide beneficial
phenotypes to their host s but are energetically costly to maintain
(San Millan & MacLean, 2017 ).
This complex interplay between element and host is evident in
Tn916 and Pseudomonas ICEclc. Tn 916 was discovered through its
ability to spread tetracycline resistance through clinical isolates
of Enterococcus (Franke & Clewell, 1981a, 1981b), thus providing a
clear benefit to its host cells. However, activation of Tn916 halt s cell
growth and leads to decreased bacterial viability (Bean, McLellan,
& Grossman, 2022b). Similarly, when activated, ICEclc causes slow
growth and decreased viability (Reinhard et al., 2013). These growth
defective cells are in a “transfer competent state.” Deletion of the
genes required for the decreased cell growth and viabilit y cause a
decrease in conjugation efficiency. This indicates that this state of
decreased cell grow th and viability is important for efficient transfer
of ICEclc (Delavat et al., 2016; Reinhard et al., 2013).
There are certainly parallels between the growth arrest caused
by TnSmu1, Tn 916, and ICEclc. However, unlike Tn916 and ICEclc,
our results indicate that TnSmu1 does not cause host cell death.
Additionally, a CRISPRi knockdown of immRSmu caused an arrest in
growth of the entire population (King et al., 2022). It is possible that
the cells with an activated TnSm u1 die following growth arrest, but
we have not observed this, nor have assays measuring death been
reported. The apparent essentiality of immRSmu (Shields et al., 2018)
could indicate that there is cell death following inac tivation of the
repressor and subsequent activation of TnSmu1. However, as dis-
cussed above, we postulate that this apparent essentiality is due to
excision and loss of TnSmu1 and the consequent loss of the immR
null allele in TnSmu1. This fur ther begs the questions: What is the
mechanism of the growth arrest caused by activation of TnSm u1,
is this growth arrest impor tant for conjugation and transfer, and is
there a benefit that TnSmu1 confers to host cells?
3.3 | Host range of TnSmu1
Bioinformatic searches indicate that TnSmu1 is naturally located
within S. mutans species. However, it is not known if TnSm u1 is

|
663
MCLELLAN et al.
capable of transfer to other bacterial species typically found in
the oral cavity. For many ICEs, there is a difference between natu-
rally occurring host range and experimental host range. For exam-
ple, ICEB s1 is naturally found only in Bacillus sp. but can transfer
to a diverse array of other microbes (Auchtung et al., 20 05; Brophy
et al., 2018). Tn916 is naturally found in Enterococcus, Clostridium,
Streptococcus, and Staphylococcus species. However, it is also func-
tional in Bacillus sp. {e.g., (Rober ts & Mullany, 2009, 2 011; Wright
& Grossman, 2016)}. Understanding experimental host range al-
lows utilization of ICEs for genetic engineering purposes to a wider
diversit y of organisms. Further, many ICEs are able to mediate
transfer of other mobile genetic elements, including plasmids that
replicate by rolling- circle replication (Johnson & Grossman, 2015;
Lee et al., 2012; Santoro et al., 2014). We suspect that TnSmu1 is
able to drive transfer of plasmids and other mobile genetic elements
found within the oral microbiome. As many mobile genetic elements
encode antibiotic resistance, the impact of TnSm u1 on clinically im-
portant phenotypes (virulence trait s and drug resistances) may ex-
tend beyond transfer of TnSmu1 alone.
4 | MATERIALS & METHODS
4.1 | Media and growth conditions
For liquid growth, S. mutans cultures were grown statically in 50%
Todd Hewitt (TH) broth in tightly closed 15 ml conical tubes. For
growth on solid media, S. mutans were grown on Brain Heart Inf usion
(BHI) plates (or TH plates where indicted) with 1.5% agar under an-
aerobic conditions (using Anaerogen Anaerobic Gas Generator,
Hardy Diagnostics). Cultures and plates were incubated at 37°C.
When appropriate, media were supplemented with 1.6 mg/mL D-
alanine. A ntibiotics were used at the following concentrations: 1 mg/
mL kanamycin, 1 mg/mL spectinomycin, 10 μg/mL erythromycin.
4.2 | Strains and alleles
S. mutans strains (Table 2) were derived from S. mutans UA159 (ATCC
700610) (Ajdic et al., 2002) and were made by natural transforma-
tion (Li et al., 2001; Petersen & Scheie, 2010). New alleles were con-
structed as double crossover events using long- flanking homology
PCR by isothermal assembly (Gibson et al., 2009; Xie et al., 2011).
Markers used to select for transformants were spc (spectinomycin
resistance), erm (erythromycin resistance), tet (tetracycline resist-
ance), or kan (kanamycin resistance). All mutants were constructed
in a clean, isogenic background and alleles were confirmed through
Sanger sequencing. Any alleles obtained from other sources were
moved into a clean, isogenic background. Construction of new
strains and alleles is summarized below.
∆TnS mu1::spc (constructed in LKM68 then transferred via nat-
ural transformation to LKM85 and LKM233) was constructed by
replacing TnSmu1 with the spectinomycin resistance cassette from
pUS19 (Benson & Haldenwang, 1993). The attL recombination site
within smu_t33 was left intact (leaving 4 bp downstream of the 3′
end of smu_t33, thus leaving intact the tRNA gene smu_t33), but the
attR recombination site was removed (preserving 67 bp downstream
of the 3′ end of smu _221c). The allele was constructed via isothermal
assembly of the antibiotic resistance cassette and ~750 bp of up-
stream and downstream homology arms.
∆TnS mu1::kan (constructed in LKM139 then transferred via natural
tr ans fo rma tion to LKM167) was made in th e sa me way as ∆TnSmu1::spc,
with the same boarders, except that the kanamycin resistance cassette
from pGK67 (Lemon et al., 20 01) was used instead of spc.
∆(smu_t33- TnSmu1)::kan in LKM141 deletes TnSmu1 and the
tRNA gene smu_t33 and inser ts kan from pGK67. The deletion end-
points ex tend from 14 bp downstream of smu_t32 through attR,
leaving smu_221c (the gene downstream of TnSmu1) intact, and
ending 67 bp downstream from the smu_221c stop codon. The allele
was constructed via isothermal assembly of the antibiotic resistance
cassette and ~750 bp of upstream and downstream homology arms.
Although smu_t33 encodes a unique tRNA , it was not essential: the
deletion was viable, albeit with a growth defect.
∆agaL::spc (constructed in LKM62 and then transferred via
natural transformation to LKM87) was constructed by replacing
the agaL open reading frame (maintaining the first 558 bp and last
458 bp of the agaL open reading frame) with the spec tinomycin re-
sistance casset te from pUS19. The allele was constructed via iso-
thermal assembly of the antibiotic resistance cassette and ~750 bp
of upstream and downstream homology arms. agaL is considered
a non- essential chromosomal location suitable for cloning (Reck &
Wagner- Döbler, 2016).
TnSmu1- tet (initially in LKM76 and then transferred via natural
transformation to make LKM145) was constructed by inserting the
tetracycline resistance cassette (tetM) 3 bp downstream of the 3′
end of sm u_211c. The allele was constructed via isothermal assembly
of the antibiotic resistance cassette and ~750 bp of upstream and
downstream homology arms. tetM fro m Tn916 was used to confer
tetracycline resistance from CMJ253 (including 376 bp upstream of
tetM so as to in clude the tetM promoter) (Johnson & Grossman, 2014).
tetM was co- direc tional with the upstream and downstream genes
in TnSmu1 and did not contain a transcriptional terminator down-
stream of tetM. TnSmu1 gfpmut2- tet in LKM137 was constructed in
an identical manner by inserting gfpmut2 allele and tetM fro m Tn916
from CMJ253 (Johnson & Grossman, 2014) 3 bp downstream of the
3′ end of smu _211c. gfpmut2 was obtained from ELC1458 (Bean,
McLellan, & Grossman, 2022b). gfpmut2 is promoter- less and co-
directional with the upstream and downstream genes in TnSmu1. It
has the B. subtilis spoVG ribosome- binding site to initiate translation.
∆alr::erm (initially in LKM127 and then transferred via natural
transformation to make LKM145) was constructed via isothermal
assembly of the erythromycin antibiotic resistance cassette (erm)
and ~750 bp of upstream and downstream homology arms. The first
4 bp of the alr open reading frame was retained as it overlapped

664
|
MCLELLAN et al.
with acpS, but the rest of the alr open reading frame was deleted (to
172 bp upstream of the 5′ end of recG). The ery thromycin resistance
cassette from pC AL215 was used (Auchtung et al., 2007), originally
derived from pDG795 (Guérout- Fleury et al., 1996).
∆comS::kan (originally in LKM69 and then transferred via natu-
ral transformation to make LKM145) was constructed by replacing
the entire comS open reading frame with the kanamycin resistance
cassette from pGK67. The allele was constructed via isothermal as-
sembly of the antibiotic resistance cassette and ~750 bp of upstream
and downstream homology arms. The ∆comS::erm allele was kindly
provided by Stephen J. Hagen and construction was previously
described (Son et al., 2012). This allele was transferred via natural
transformation to LKM85, LKM87, and LKM233.
∆agaL::(immRSmu kan) (smu_218c kan) (initially in LKM218 and
then transferred via natural transformation to make LKM233) was
constructed by replacing the agaL open reading frame with immRSmu
and kan (maintaining the first 558 bp and last 458 bp of the 2,160 bp
agaL open reading frame). immRSmu was cloned in the opposite ori-
entation of agaL to prevent expression from the agaL promoter. im-
mRSmu was amplified from TnSmu1 from S. mutans UA159, amplifying
746 bp upstream of the 5′ end of immRSmu such that it contains the
predicted promoter driving expression of immRSmu, through the end
of the immRSmu open reading frame. The gatB/yerO bidirectional
terminator from B. subtilis was added 6 bp downstream of the stop
codon of immRSmu to prevent possible transcriptional conflic t from
the agaL promoter. The organization of TnSmu1 is reminiscent of
that of ICEB s1 where immR and immA are co- transcribed and diver-
gent from most of the genes in the element (Auchtung et al., 2007;
Bose et al., 2008). Therefore, we predicted the intergenic region
between sm u _217c and immRSmu (smu_ 218c) would contain the pro-
moter of ImmRSmu. The kanamycin resistance cassette from pGK67
was cloned divergent to immRSmu. The allele was constructed via
TABLE 2 S. mutans strains
Strain1Relevant genotype (reference; comment)
UA159 Originally isolated from a child with ac tive caries, GenBank: AE014133.2 (Ajdic et al., 2002)
LKM62 ∆agaL::spc (used to cons truc t LKM87)
LKM68 ∆Tn Smu1::spc2 (used to construct LKM85, LKM167, LKM233, used to determine growth of ∆TnSmu1 compared
to WT UA159)
LKM69 ∆comS::kan (used to construc t LKM145)
LK M76 Tn Sm u1- tet3 (used to construct LKM145)
LKM85 ∆Tn Smu1::spc2 ∆comS::erm4 (used as recipient in matings)
LKM87 ∆agaL::spc ∆comS::erm4 (used as recipient in matings, ∆agaL::spc was used as a selectable marker to confirm
identify of transconjugants)
LKM127 ∆alr::erm (used to cons truc t LKM145, ∆alr::erm was used to provide counterselection to prevent donor growth)
LKM137 Tn Smu1- (gfpmut2 tet)5 (insertion downstream from gene smu _211c, used to monitor cells with TnSmu1
expression)
LKM139 ∆TnS mu1::kan2 (used to construct LKM167); leaves smu_t33 (tRNA) intact
LKM141 ∆(smu_t33- Tn Sm u1)::kan6 (used to determine essentiality of smu _t33)
LKM145 TnSmu1- tet3 ∆alr::erm ∆comS::kan (used as donor during mating assays, ∆alr::erm was used to provide
counterselection to prevent donor growth)
LKM16 4 ∆lacE::(attBSmu spc)7 (used to construct LKM167)
LKM165 ∆lacE::(attTnSmu1 s pc)8 (used as qPCR s tandard curve control)
LKM167 ∆lacE::(attBSmu spc)7 ∆TnSmu1::kan2 (used as qPCR standard curve control)
LKM218 ΔagaL::(immRSmu kan)9 (used to cons truct LKM233)
LKM233 ∆Tn Sm u1::spc2 ∆agaL::(immRSmu kan)9 ∆comS::erm4
1Strains are derived from UA159.
2∆TnS mu1::spc was const ructed by replacing TnSm u1 with t he spec tinomycin resistance cassette. The attL recombination site within smu_t33 was left
intact but the attR recombination site was removed. ∆TnSmu1::kan was constructed in an identical fashion instead using the kanamycin resistance
cassette.
3TnSmu1- tet contains the tetracycline resis tance c asset te tetM inserted 3 bp downstream of the 3' end of s mu _ 211c.
4∆comS::erm was kindly provided by Stephen J. Hagen (Son et al., 2012).
5TnS mu1- (gfpmut2 tet) contains gfpmut2 and tetM 3 bp downstream of the 3' end of sm u_ 211c .
6∆(smu_t33- TnSmu1)::kan replaces smu_t33 a nd Tn Smu1 with the kanamycin resistance cassette. The TnSmu1 attL/attR recombination sites were also
deleted and the tRNA gene smu_t33 is disrupted. These cells are viable, so this gene is not essential.
7∆lacE::(attBSmu spc) contains the genomic attachment site for TnSm u1 cloned, along with a spectinomycin resistance cassette, into lacE.
8∆lacE::(attTnS mu1 s pc) contains the attachment site from the TnSmu1 circle cloned, along with a spectinomycin resistance cassette, into lacE.
9∆agaL::(immRSmu kan) contains immRSmu (including the putative promoter) and kan cloned into agaL.

|
665
MCLELLAN et al.
isothermal assembly of the antibiotic resistance cassette, the im-
mRSmu fragment, and ~750 bp of upstream and downstream homol-
ogy arms. agaL is a non- essential chromosomal location suitable for
cloning (Reck & Wagner- Döbler, 2016).
∆lacE::(attTnS mu1 spc) in LKM165 was constructed by inser ting
attTnS mu1 and spc into lacE, maintaining the first 512 bp and the last
463 bp of the lacE open reading frame. attTn Sm u1 and the surround-
ing regions in TnSmu1 (containing the last 730 bp of the intSmu open
reading frame and the first 157 bp of the smu_220 open reading
frame) were amplified from TnSm u1 that had been spontaneously
excised from liquid cultures of S. mutans UA159. The spectinomycin
resist ance cassette fr om pUS19 was used. T he allele was const ructed
via isothermal assembly of the antibiotic resistance cassette, the
attTnS mu1 fragment, and ~750 bp of upstream and downstream ho-
mology arms. lacE is considered a non- essential chromosomal loca-
tion suitable for cloning (Reck & Wagner- Döbler, 2016).
∆lacE::(attBSmu spc) (initially in LKM164 and then transferred via
natural transformation to make LKM167) was constructed by inser t-
ing attBSmu and spc into lacE (with regions of lacE as described above).
attBSmu was constructed by amplifying the genomic region upstream
of attL (including the last 16 bp of s mu_t24 to 5 bp downstream of
the 3′ end of smu_t33 so that the recombination site attB is pres-
ent) and stitching it together with the genomic region downstream
of attR (from 67 bp downstream of the 3′ end of smu _221 to 124 bp
upstream of the 5′ end of smu_221) and cloned into pCAL1422
(Thomas et al., 2013). The attBSmu was amplified from the resulting
plasmid and the spectinomycin resist ance cassette was amplified
from pUS19. The allele was constructed via isothermal assembly of
the antibiotic resistance cassette, the attBSmu fragment, and ~750 bp
of upstream and downstream homology arms.
4.3 | Homology, bioinformatic analyses, and
TnSmu1 conservation
Global alignments between TnSmu1 were calculated with EMBOSS
Needle pairwise global sequence alignment (Needleman &
Wunsch, 1970). BLASTp and BLASTn publicly available databases
were accessed on June 9, 2022. Structural predictions were done
using Phyre2, accessed June 9, 2022. Helix- turn- helix domains were
predicted using GYM 2.0 (Narasimhan et al., 2002).
We used cblaster (Gilchrist et al., 2021) to look at conservation of
TnS mu1 across sequences of S. mutans publicly available as of June
22, 2022. We used the predicted protein sequences of the genes
within TnSmu1 as the query for cblaster. Subjects were grouped
into clusters based on BLASTp hits to TnSmu1 protein sequence
and ranked by cluster similarity. Cluster similarity score is calcu-
lated by cblaster as S = h + i•s, where h is the number of query se-
quences with BLASTp hits, s is the number of contiguous gene pairs
with conserved synteny, and i is a weighting factor (default value
0.5) determining the weight of synteny in the similarit y score. If a S.
mutans strain appeared twice due to multiple copies of the genome
sequence available in the NCBI database, the genome with the lower
cluster similarit y score was excluded from Figure S1.
4.4 | Determination of TnSmu1 excision and
integration
Genomic DNA was isolated from overnight cultures of S. mutans
UA159 using Qiagen DNeasy kit with 40 mg/mL lysozyme. TnSmu1
excision and the resulting attB sequence was identified with prim-
ers oLM27 (5′- ACACCAGATTGTGGCTCTG) and oLM49 (5′- GGC A
AGTCTTGATTATCGCTTTTAGAAAGAG). The attTnS mu1 junction
formed via site- specific recombination was determined using prim-
er s oLM 3 8 (5′- CATCA AG T TAGCAC AGTC AGATA A AATC G) a n d
oL M107 (5′- CATA ATAGGTTCCAT T TA A ACTACTGCC ). The re sul tin g
products were determined by Sanger sequencing. The location of
integrated TnSmu1 in transconjugants was confirmed using prim-
ers oLM40 (5′ - CGACCAAAGAAAAAATTATCTCAAGAGACAAAG
−3′) and oLM44 (5′- GGCAGACGCGCTGGAC - 3′) that detected attL
within smu_t33.
4.5 | Quantitative PCR to determine excision and
replication of the element
Overnight cultures were diluted to OD = 0.05 in 50% TH medium
and grown at 37°C. After 3 h of growth, the culture was split and
1 μg/mL Mitomycin C (MMC) was added to half of the culture
(where indicated). Samples were taken every hour pre- and post-
MMC addition and into stationary phase (7 h total). Genomic DNA
was isolated using Qiagen DNeasy kit. Excision was measured using
primers oLM166 (5′- TTGGTTCGAATCCAGCTACC) and oLM109
(5′- GACTTATGGTCATTTGGTTGCG) to amplify the vacant inser-
tion site attB. attB amplification was normalized to a control chro-
mosomal region in ilvB, which is ~7 kb downstream of attB. ilvB was
amplified with primers oLM173 (5′- AGGTGGCGGTGTCAATTATG)
and oLM174 (5′- GCATCTCCCACAACTGGAATAG). The copy
number of the TnSmu1 circle was measured with primer pair
oLM224 (5′- AATCTTCTATCCCAAATTTTCTCCC) and oLM226
(5′- TGGGAGAAATTTTGGGAGAGAAAATC) to quantitate the
unique attTnS mu1 junction formed via site- specific recombination.
To determine if TnSm u1 was replicating, we determined the ratio of
the number of copies of circular TnSm u1 (attTn Sm u1) to the number
of copies of the excision site (attB).
qPCR was done using SSoAdvanced SYBR mas ter mix and CFX96
Touch Real- Time PCR system (Bio- Rad). Copy numbers of attP and
attB were determined by the Pfaffl method (Pfaffl, 2001). Standard
curves for attTnSmu1, attB, and the control chromosome locus ilvB
were generated from genomic DNA of LKM165, LKM167, and S.
mutans UA159, respec tively. LKM165 contains an ectopic copy of
attTnS mu1 inserted at lacE. LKM167 does not contain TnSm u1 and
therefore contains a copy of the unoccupied chromosome site attB

666
|
MCLELLAN et al.
inserted at lacE. S. mutans UA159 was considered the wild type and
was used for measuring the nearby chromosomal locus, ilvB.
4.6 | Growth curves
Strains were grown in 50% TH broth overnight. Cultures were di-
luted to an OD600 of 0.05 and grown statically in closed 15 ml conical
tubes. The number of colony forming units (CFUs) and OD60 0 was
determined every hour. Cells were monitored for 7 h of growth.
4.7 | Mating assays
Donor and recipient cells were diluted to early exponential phase
and grown to mid- exponential phase over 4 h. Cells were then pel-
leted at 5000 rpm for 5 mins and resuspended to an OD of 0.01 in
1x Spizizen's salts (Harwood & Cutting, 199 0). Donor and recipient
cells were combined at various ratios and 50 μl of each mix was spot-
ted onto a BHI plate (or TH plate) supplemented with D- alanine.
Mating plates were then incubated at 37°C in anaerobic conditions
(using Anaerogen Anaerobic Gas Generator, Hardy Diagnostics) for
up to 7 days. Spots were then har vested by scrapping the spot into
1x Spizizen's salts and vortexed. Cells were then plated on selective
media to detect TnSmu1 transfer. The number of donor (tetracycline-
resistant, D- alanine auxotrophs), recipient (tetracycline sensitive,
D- alanine prototrophs), and transconjugant (tetracycline- resistant,
D- alanine prototrophs) CFUs were enumerated both pre- and post-
mating. Conjugation efficiency was calculated as the percentage of
transconjugant CFUs formed normalized to the number of donor
cells harvested at the end of the mating to account for cell growth.
4.8 | Time- lapse microscopy and analysis
S. mutans cells were diluted to early exponential phase in 50% TH
media. Af ter 3 h of growth to late exponential phase, cells were
transferred to an agarose pad (1.5% UltraPure agarose, Invitrogen;
dissolved in growth medium) containing 0.1 M propidium iodide
and 0.35 μg/mL DAPI. The agarose pad was created in an incu-
bation chamber, which was made by stacking two Frame- Seal
Slide Chambers (Bio- Rad) on a standard microscope slide (VWR).
Cells were then grown at 37°C for 3 h while monitoring growth.
Fluorescence was generated using a Nikon Intensilight mercury il-
luminator through appropriate sets of excitation and emission fil-
ters (Chroma; filter sets 49,000, 49,002, and 49,008). Time- lapse
images were c aptured on a Nikon Ti- E inver ted microscope using a
CoolSnap HQ camera (Photometrics). ImageJ software (Schindelin
et al., 2012) was used for image processing and analysis.
AUTHOR CONTRIBUTIONS
Lisa K. McLellan: Conceptualized, investigated, validated, visual-
ized, wrote (original and editing); Mary E. Anderson: Investigated,
validated, visualized, wrote (editing); Alan D. Grossman:
Conceptualized, funding acquisition, super vision, wrote (original
manuscript and editing).
ACKNOWLEDGMENTS
We thank Katharina Ribbeck (MIT, Cambridge, MA), Dan Smith
(Forsyth Institute, Cambridge, MA) for sharing strains. We thank
Stephen J. Hagen (University of Florida, Gainesville, FL) for kindly
sharing strains and the ∆comS::erm allele, James S. Weagley
(Washing ton University, St. Louis, MO) for thoughtful conversions on
bioinformatic analyses of TnSmu1 distribution among bacterial spe-
cies and Sa ng- Joon Ahn (Un iversity of Florida, Gainesville, FL) for use-
ful conversations and Robert Shields (Arkansas State University) for
discussions regarding results and appropriate nomenclature for genes
in TnSmu1. Research repor ted here is based upon work supported,
in part, by the National Institute of General Medical Sciences of the
National Institutes of Health under award number R35 GM122538 to
ADG. Any opinions, findings, and conclusions or recommendations
expressed in this report are those of the authors and do not necessar-
ily reflect the views of the National Institutes of Health.
CONFLICT OF INTEREST
The authors have declared that no competing interests exist.
DATA AVAIL ABI LIT Y STAT EME NT
The data that support the findings of this study are available from
the corresponding author upon reasonable request.
ETHICS STATEMENT
No human or animal subjects were used in this study.
ORCID
Lisa K. McLellan https://orcid.org/0000-0001-6418-8039
Mary E. Anderson https://orcid.org/0000-0002-6930-3830
Alan D. Grossman https://orcid.org/0000-0002-8235-7227
REFERENCES
Ajdic, D., McShan, W.M., McL aughlin, R.E., Savic, G., Chang, J., Carson,
M.B. et a l. (2002) Ge nome sequen ce of Streptococcus mutans UA159,
a cariogenic dental pathogen. Proceedings of the National Academy
of Sciences of the United States of America, 99, 144 34 – 14439.
Alvarez- Martinez, C .E. & Christie, P.J. (2009) Biological diversity of
prokar yotic t ype IV secretion systems. Microbiology and Molecular
Biology Reviews, 73, 775– 808.
Auchtung, J.M., Lee, C.A., Monson, R.E ., Lehman, A. P. & Grossman,
A.D. (2005) Regulation of a Bacillus subtilis mobile genetic element
by intercellular signaling and the global DNA damage response.
Proceedings of the National Academy of Sciences of the United States
of America, 102, 12554– 12559.
Auchtung, J.M., Lee, C.A., Garrison, K.L. & Grossman, A.D. (2007)
Identification and characterization of the immunity repressor
(ImmR) that controls the mobile genetic element ICEB s1 of Bacillus
subtilis. Molecular Microbiology, 64, 1515– 1528.
Auchtung, J.M., Aleksanyan, N ., Bulku, A. & Berkmen, M.B. (2016)
Biolog y of ICE Bs1, an integrative and conjugative element in
Bacillus subtilis. Plasmid, 86, 14– 25.

|
667
MCLELLAN et al.
Babic, A ., Berkmen, M.B., Lee, C. A. & Grossman, A. D. (2011) Efficient
gene transfer in bacterial cell chains. mBio, 2, e00027- 11.
Beaber, J.W., Hochhut, B. & Waldor, M.K. (2004) SOS response promotes
horizontal dissemination of antibiotic resistance genes. Nature, 427,
7 2 – 7 4 .
Bean, E .L., Herma n, C., Ander son, M.E. & Gr ossman, A. D. (2022a) Biolog y
and engineering of integrative and conjugative elements: construc-
tion and analyses of hybrid ICEs reveal element functions that af-
fect species- specific efficiencies. PLoS Genetics, 18 , e1009998.
Bean, E.L., McLellan , L.K . & Grossman, A .D. (2022b) Activation of the in-
tegrative and conjugative element Tn916 causes growth arrest and
death of host bacteria. PLoS Genetics, 18, e10104 67.
Benson, A.K . & Haldenwang, W.G. (1993) Regulation of sigmaB levels
and activity in Bacillus subtilis. Journal of Bacteriology, 175,
2347– 2356.
Berkmen, M.B ., Lee, C .A ., Loveday, E.- K. & Grossman, A.D. (2010) Polar
positioning of a conjugation protein from the integrative and con-
jugative element ICEBs1 of Bacillus subtilis. Journal of Bacteriology,
192, 38– 45.
Bhatt y, M., Laverde Gomez, J.A . & Christie, P.J. (2013) The expanding
bacter ial type IV secretion lexicon. Research in Microbiology, 164,
62 0– 6 39.
Bi, D., Xu, Z., Harrison, E.M ., Tai, C., Wei, Y., He, X . et al. (2012) ICEberg:
a web- based resource for integrative and conjugative elements
found in bacteria. Nucleic Acids Research, 40, D621– D626.
Bose, B. & G rossman, A.D. (2011) Regulation of horizontal gene transfer
in Bacillus subtilis by activation of a conser ved site- specific prote-
ase. Journal of Bacteriology, 193, 22– 2 9.
Bose, B., Auchtung, J.M., Lee, C.A. & Grossman, A.D. (200 8) A conserved
anti- repressor controls horizontal gene transfer by proteolysis.
Molecular Microbiology, 70, 570– 582.
Botelho, J. & Schulenburg, H. (2021) The role of integrative and con-
jugative elements in antibiotic resistance evolution. Trends in
Microbiology, 29, 8– 18.
Brophy, J.A.N., Triassi, A.J., Adams, B.L., Renberg, R .L., Stratis- Cullum,
D.N., Grossman, A.D. et al. (2018) Engineered integrative and con-
jugative elements for efficient and inducible DNA transfer to undo-
mesticated bacteria. Nature Microbiology, 3, 1043– 1053.
Burrus, V. & Waldor, M.K . (2004) Shaping bacterial genomes with in-
tegrative and conjugative elements. Research in Microbiology, 155,
376– 386.
Burrus, V., Pavlovic, G., Decaris, B . & Guédon, G. (2002) Conjugative
transposons: the tip of the iceberg. Molecular Microbiology, 46,
60 1– 610 .
Ciric, L ., Jasni, A., de Vries, L .E., A gersø, Y., Mullany, P. & Roberts, A .P.
(2013) The Tn916/ Tn1545 family of conjugative transposons. In:
Madame Curie Bioscience Database [Internet]. Landes Bioscience:
Austin ( TX), pp. 200 0– 2013. Available from: https://www.ncbi.nlm.
nih.gov/books/ NBK63 531/
Cury, J., Touchon, M. & Rocha, E.P.C. (2017) Integrative and conjugative
elements and their hosts: composition, distribution and organiza-
tion. Nucleic Acids Research, 45, 8943– 8956.
Decker, E.- M. (2001) The ability of direct fluorescence- based, t wo-
colour assays to detect different physiological st ates of oral strep-
tococci. Letters in Applied Microbiology, 33, 188– 192.
Decke r, E.- M. , Kl ei n, C., Schwi nd t, D. & vo n Oh le , C. (20 14) Met ab ol ic ac-
tivity of Streptococcus mutans biofilms and gene expression during
exposure to xylitol and sucrose. International Journal of Oral Science,
6, 195– 204.
Delavat, F., Mitri, S., Pelet, S. & van der Meer, J.R. (2016) Highly vari-
able individual donor cell fates characterize robust horizontal gene
transfer of an integrative and conjugative element. Proceedings of
the Nation al Academy of Scienc es of the United States of Ame rica, 113,
E3375– E3383.
DeWitt, T. & Grossman, A .D. (2014) The bifunctional cell wall hy-
drolase CwlT is needed for conjugation of the integrative and
conjugative element ICEBs1 in Bacillus subtilis and B. anthracis.
Journal of Bacteriology, 196, 1588– 1596.
Dimopoulou, I.D., Russell, J.E., Mohd- Zain, Z., Herber t, R. & Crook,
D.W. (2002) Site- specific recombination with the chromosomal tR-
NA(leu) gene by the large conjugative Haemophilus resistance plas-
mid. Antimicrobial Agents and Chemotherapy, 46, 160 2– 1603.
Franke, A .E. & Clewell, D.B . (1981a) Evidence for conjugal transfer of
a Streptococcus faecalis transposon (Tn916) from a chromosomal
site in the absence of plasmid DNA. Cold Spring Harbor Symposia on
Quantitative Biology, 45(Pt 1), 77– 80.
Franke, A .E. & Clewell, D.B . (1981b) Evidence for a chromosome- borne
resistance transposon (Tn916) in Streptococcus faecalis that is capa-
ble of “conjugal” transfer in the absence of a conjugative plasmid.
Journal of Bacteriology, 145, 494– 502.
Fronzes, R ., Christie, P.J. & Waksman, G. (2009) The struc tural biology of
type I V secretion sys tems. Nature Reviews. Microbiology, 7, 7 0 3 – 7 1 4 .
Fujimura- Kamada, K., Nouvet, F.J. & Michaelis, S. (1997) A novel
membrane- associated metalloprotease, Ste24p, is required for the
first s tep of NH2- terminal processing of the yeast a- factor precur-
so r. The Journal of Cell Biology, 136, 2 71– 2 8 5.
Gibson, D.G., Young, L ., Chuang, R.- Y., Venter, J.C., Hutchison, C.A ., III
& Smith, H .O. (2009) Enzymatic assembly of DNA molecules up to
several hundred kilobases. Nature Methods, 6(343), 345.
Gilchrist, C.L.M., Booth, T.J., van Wersch, B., van Grieken, L., Medema,
M.H. & Chooi, Y.- H. (2021) Cblaster: a remote search tool for
rapid identification and visualization of homologous gene clusters.
Bioinformatics Advances, 1, vbab016.
Gottesman, M.E. & Weisberg, R.A. (200 4) Little lambda, who made thee?
Microbiology and Molecular Biology Reviews, 68, 796– 813.
Grindley, N.D.F., Whiteson, K. L. & Rice, P.A. (20 06) Mechanisms of site-
specific recombination. Annual Review of Biochemis try, 75, 567– 605.
Guérout- Fleury, A.M ., Frandsen, N. & Stragier, P. (1996) Plasmids for ec-
topic integration in Bacillus subtilis. Gene, 180, 5 7– 61 .
Guglielmini, J., Quintais, L., Garcillán- Barcia, M.P., de la Cruz, F. & Rocha,
E.P.C. (2011) The repertoire of ICE in prokaryotes underscore s
the unit y, diversit y, and ubiquity of conjugation. PLoS Genetics, 7,
e1002222.
Harwood, C.R. & Cut ting, S.M. (1990) Molecular biological methods for
bacillus. John Wiley & Sons. UK: Chichester.
Hirano, N., Muroi, T., Takahashi, H. & Haruki, M . (2011) Site- specific
recombinases as tools for heterologous gene integration. Applied
Microbiology and Biotechnology, 92, 2 27– 2 39.
Iyer, L.M., Makarova, K.S., Koonin, E.V. & Aravind, L. (200 4) Comparative
genomic s of the FtsK– HerA superfamily of pumping ATPases: im-
plications for the origins of chromosome segregation, cell division
and viral capsid packaging. Nucleic Acids Research, 32, 5260– 5279.
Jiang, X ., Hall, A.B., Xavier, R. J. & Alm, E.J. (2019) Comprehensive anal-
ysis of chromosomal mobile genetic elements in the gut microbi-
ome reveals phylum- level niche- adaptive gene pools. PLoS One, 14 ,
e0223680.
Johnson, C.M. & Grossman, A.D. (2014) Identification of host genes
that affect acquisition of an integrative and conjugative element in
Bacillus subtilis. Molecular Microbiology, 93, 1284– 1301.
Johnson, C.M. & Grossman, A.D. (2015) Integrative and conjugative el-
ements (ICEs): what they do and how they work. Annual Review of
Genetics, 49, 57 7– 601 .
Jones, J.M., Grinberg, I., Eldar, A. & Grossman, A .D. (2021) A mobile ge-
netic element increases bacterial host fitness by manipulating de-
velopment. eLife, 10, e65924.
King, S., Quick, A., King, K., Walker, A.R . & Shields, R.C . (2022) Activation
of TnS mu1, an integrative and conjugative element, by an ImmR-
like transcriptional regulator in Streptococcus mutans. Microbiology
(Reading), 168, 001254. Available from: ht tps://doi.o rg/10.1099/
mic.0.001254
Koyanagi, S . & Lévesque, C.M. (2013) Characterization of a
Streptococcus mutans intergenic region containing a small toxic

668
|
MCLELLAN et al.
peptide and its cis- encoded antisense small RNA antitoxin. PLoS
One, 8, e54291.
Lee, C. A. & Grossman, A.D. (2007) Identification of the origin of transfer
(oriT) and DNA relaxase required for conjugation of the integra-
tive and conjugative element ICEBs1 of Bacillus subtilis. Journal of
Bacteriology, 189, 7254– 7261.
Lee, C. A., Auchtung, J.M., Monson, R.E . & Grossman, A.D. (2007)
Identification and characterization of int (integrase), xis (excision-
ase) and chromosomal attachment sites of the integrative and con-
jugative element ICEBs1 of Bacillus subtilis. Molecular Microbiology,
66, 1356– 1369.
Lee, C. A., Babic, A. & Grossman, A.D. (2010) Autonomous plasmid- like
replication of a conjugative transposon. Molecular Microbiology, 75,
26 8– 2 79.
Lee, C. A., Thomas, J. & Grossman, A .D. (2012) The Bacillus subtilis con-
jugative transposon ICEBs1 mobilizes plasmids lacking dedicated
mobilization functions. Journal of Bacteriology, 194, 3165– 3172.
Lemon, K .P., Kurtser, I. & Grossman, A.D. (2001) Effects of replication
termination mutants on chromosome partitioning in Bacillus sub-
tilis. Proceedings of the National Academy of Sciences of the United
States of America, 98, 212– 217.
Leonet ti, C.T., Hamada, M.A., Laurer, S.J., Broulidakis, M.P., Swerdlow,
K.J., Lee, C.A. et al. (2015) Critical components of the conjugation
machinery of the integrative and conjugative element ICEBs1 of
Bacillus subtilis. Journal of Bacteriology, 197, 2558– 2567.
Li, Y.- H., L au, P.C.Y., Lee, J.H., Ellen, R. P. & Cvitkovitch, D.G. (2001)
Natural genetic transformation of Streptococcus mutans growing in
biofilms. Journal of Bacteriology, 183, 8 97– 9 08.
Lucchini, S., Desiere, F. & Brüssow, H. (1999) Similarly organized lysog-
eny modules in temperate Siphoviridae from low GC content gram-
positive bacteria. Virology, 263, 427– 435.
Lunde, T.M., Hjerde, E. & Al- Haroni, M. (2021) Prevalence, diversity and
transferability of the Tn916 - Tn1545 family ICE in oral streptococci.
Journal of Oral Microbiology, 13, 1896874.
Mashbur n- Warren, L., Morrison, D.A. & Federle, M.J. (2010) A novel
double- tryptophan peptide pheromone is conserved in mutans and
pyogenic streptococci and controls competence in Streptococcus
mutans via an Rgg regulator. Molecular Microbiology, 78, 589– 606.
Miyaza ki, R. & van der Mee r, J.R. (2013) A n ew large DNA fragm ent delivery
system based on integrase activity from an integrative and conjuga-
tive element. Applied and Environmental Microbiology, 79, 4440– 4447.
Narasimhan, G., Bu, C., Gao, Y., Wang, X., Xu, N. & Mathee, K . (2002)
Mining protein sequences for motifs . Journal of Computational
Biology, 9, 707– 720.
Needleman, S.B. & Wunsch , C.D. (1970) A general method applicable to
the search for similarities in the amino acid sequence of two pro-
teins. Journal of Molecular Biology, 48, 443– 453.
Nomura , R., Matayoshi, S ., Otsugu, M., Kitamura, T., Teramoto, N . &
Nakano, K. (2020) Contribution of severe dental caries induced by
Streptococcus mutans to the pathogenicity of infective endocarditis.
Infection and Immunity, 88, e 0 089 7- 19.
Olsen, I ., Tribble, G .D., Fiehn, N.- E. & Wang, B.- Y. (2013) Bacterial sex
in dental plaque. Journal of Oral Microbiology, 5, 20736. Available
from: https://doi.org/10.3402/jom.v5i0.20736
Oppenheim, A .B., Kobiler, O., Stavans, J., Court, D.L. & Adhya, S. (20 05)
Switches in bacteriophage lambda development. Annual Review of
Genetics, 39, 409– 429.
Pembroke, J.T. & Stevens, E. (1984) The effect of plasmid R391 and other
IncJ plasmids on the survival of Escherichia coli after UV irradiation.
Journal of General Microbiology, 130, 1839– 1844.
Peters, J.M., Koo, B .- M., Patino, R., Heussler, G.E ., Hearne, C.C ., Qu,
J. et al. (2019) Enabling genetic analysis of diver se bacteria with
Mobile- CRISPRi. Nature Microbiology, 4, 244– 250.
Peterse n, F.C . & Scheie, A. A. (2010) Natural transformation of oral s trep-
tococci. In: Seymour, G.J., Cullinan, M.P. & Heng, N.C .K. (Eds.) Oral
biology: molecular techniques and applications. Totowa, NJ: Humana
Press, pp. 167– 180. Available from: ht tps://doi.o rg/10.1007/978-
1- 6 0761 - 8 20 - 1 _12
Petit, M.- A., Derv yn, E., Rose, M., Entian, K.- D., McGover n, S., Ehrlich,
S.D. et al. (1998) PcrA is an essential DNA helicase of Bacillus sub-
tilis fulfilling functions both in repair and rolling- circle replication.
Molecular Microbiology, 29, 261– 2 73 .
Pfaf fl, M.W. (2001) A new mathem atical model for relative q uantificatio n
in real- time RT- PCR. Nucleic Acids Research, 29, e45.
Reck, M. & Wagner- Döbler, I. (2016) Carolacton treatment causes
delocalization of the cell division proteins PknB and DivIVa in
Streptococcus mutans in vivo. Frontiers in Microbiology, 7, 684.
Reinhard, F., Miyazaki, R., Pradervand, N. & van der Meer, J.R. (2013)
Cell dif ferentiation to “mating bodies” induced by an integrating
and conjugative element in free- living bacteria. Current Biology, 23,
255– 2 59.
Rice, P., Longden, I. & Bleasby, A. (2000) EMBOSS: the European molec-
ular biology open soft ware suite. Trends in Genetics, 16 , 276– 277.
Roberts, A .P. & Mullany, P. (2009) A modular master on the move: the
Tn916 family of mobile genetic elements. Trends in Microbiology, 17,
251– 258.
Roberts, A .P. & Mullany, P. (2011) Tn916- like genetic elements: a diverse
group of modular mobile elements conferring antibiotic resis tance.
FEMS Microbiology Reviews, 35, 856– 871.
Roberts, A .P., Pratten , J., Wilson, M. & Mullany, P. (1999) Transfer of
a conjugative transposon , Tn5397 in a model oral biofilm. FEMS
Microbiology Letters, 177, 63– 66.
Roberts, A .P., Cheah, G., Ready, D., Pratten, J., Wilson, M. & Mullany,
P. (2001) Transfer of Tn916- like elements in microcosm dental
plaques. Antimicrobial Agents and Chemotherapy, 45, 2943– 2946.
Rocco, J.M. & Churchward, G. (2006) The integrase of the conjugative
transposon Tn916 directs strand- and sequence- specific cleavage
of the origin of conjugal transfer, oriT, by the endonuclease Orf20.
Journal of Bacteriology, 188, 2207– 2213.
San Millan, A. & MacLean, R.C. (2017) Fitness costs of plasmids: a limit to
plasmid transmission. Microbiol Spectr, 5, 5.5.02.
Santoro, F., Vianna, M. E. & Roberts, A.P. (2014) Variation on a theme; an
overview of the Tn916/Tn1545 family of mobile genetic elements in
the oral and nasophary ngeal streptococci. Frontiers in Microbiology,
5, 535.
Schindelin, J., A rganda-Carreras, I., Frise, E., Kaynig, V., Longair, M.,
Pietzsch, T. et al. (2012) Fiji: an open-source platform for biologi-
cal-image analysis. Nature Methods, 9, 676 – 6 82 .
Serfiotis- Mitsa, D., Robert s, G. A., Cooper, L.P., White, J.H., Nutley, M.,
Cooper, A. et al. (20 08) The Orf18 gene product from conjugative
rransposon Tn916 is an ArdA antirestriction protein that inhibits
type I DNA restriction– modification systems. Journal of Molecular
Biology, 383, 970– 981.
Shanker, E. & Federle, M.J. (2017) Quorum sensing regulation of com-
petence and bacteriocins in Streptococcus pneumoniae and mutans.
Genes, 8, 15.
Shields, R.C., Zeng, L ., Culp, D. J. & Burne, R. A. (2018) Genomewide
identification of essential genes and fitness determinants of
Streptococcus mutans UA159. mSphere, 3, e00031- 18.
Son, M., Ahn, S.- J., Guo, Q., Burne, R .A . & Hagen, S.J. (2012) Microfluidic
study of competence regulation in Streptococcus mutans: environ-
mental inputs modulate bimodal and unimodal expression of comX.
Molecular Microbiology, 86, 258– 272.
Thomas, J., Lee, C .A . & Grossman, A .D. (2013) A conserved helicase
processivity f actor is needed for conjugation and replic ation of an
integrative and conjugative element. PLoS Genetics, 9, e1003198 .
Wecke, J., Madela, K. & Fischer, W. (1997) The absence of D- alanine
from lipoteichoic acid and wall teichoic acid alters surface charge,
enhances autolysis and increases susceptibility to methicillin in
Bacillus subtilis. Microbiology, 143, 2953– 2960.

|
669
MCLELLAN et al.
Wozniak, R.A.F. & Waldor, M.K. (2010) Integrative and conjugative el-
ements: mosaic mobile genetic elements enabling dynamic lateral
gene flow. Nature Reviews. Microbiology, 8, 552– 563.
Wright, L .D. & Grossman , A.D. (2016) Autono mous replica tion of the con-
jugative transposon Tn916. Journal of Bacteriology, 198, 3355– 3366.
Xiang, Y., Morais, M.C ., Cohen, D.N., B owman, V.D., Anderson, D.L. &
Rossmann, M.G. (2008) Cryst al and cr yoEM structural studies
of a cell wall degrading enzyme in the bacteriophage phi29 tail.
Proceedings of the National Academy of Sciences of the United States
of America, 105, 9552– 9557.
Xie, Z., Okinaga, T., Qi, F., Zhang, Z. & Merrit t, J. (2011) Cloning-
independent and counterselectable markerless mutagenesis sys-
tem in Streptococcus mutans. Applied and Environmental Microbiology,
77, 8025– 8033.
Zhang, K ., Ou, M., Wang , W. & Ling, J. (2009) Ef fects of qu orum sensing o n
cell viability in Streptococcus mutans biofilm formation. Biochemical
and Biophysical Research Communications, 379, 933– 938.
SUPPORTING INFORMATION
Additional supporting information can be found online in the
Suppor ting Information section at the end of this article.
How to cite this article: McLellan, L. K., Anderson, M. E. &
Grossman, A. D. (2022). TnSmu1 is a functional integrative and
conjugative element in Streptococcus mutans that when
expressed causes growth arrest of host bac teria. Molecular
Microbiology, 118, 652–669. htt ps://doi.or g/10.1111/
mmi.14992