© The Author(s) 2023. Published by Oxford University Press on behalf of The Genetics Society of America. All rights
reserved. For permissions, please e
-
mail: journals.permissions@oup.com This article is published and distributed
under the terms of the Oxford University Press, Standard Journals Publication Model
(https://academic.oup .com/pages/standard
-
publicati on
-
reuse
-
rights)
1
Molecular identification of a peroxidase gene controlling body size in the entomopathogenic
1
nematode
Steinernema hermaphroditum
2
3
Hillel T. Schwartz
1,4
, Chieh
-
Hsiang Tan
1
, Jackeline Peraza
2
, Krystal Louise T. Raymundo
3
, and
4
Paul W. Sternberg
1,4
5
6
1
Division of Biology and Biological Engineering, California Institute of Technology, Pasadena,
7
CA 91125, USA
8
2
Department of Biology, Barnard College of Columbia University, New York, NY 10027, USA
9
3
Dept. of Neuroscience, Princeton University, Princeton, NJ 08544, USA
10
4
To whom correspondence should be addressed: hillels@caltech.edu and pws@caltech.edu
11
12
Running head: Cloning a
dpy
gene in an entomopathogenic nematode
13
14
Keywords:
Steinernema
, entomopathogenic nematode, CRISPR, body size, mutagenic
15
spectrum
16
17
Abstract
18
The entomopathogenic nematode
Steinernema hermaphroditum
was recently rediscovered and
19
is being developed as a genetically tractable experimental system for the study of previously
20
unexplored biology, including parasitism of its insect hosts and mutualism with its bacterial
21
endosymbiont
Xenorhabdus griffiniae
. Through whole
-
genome re
-
sequencing and genetic
22
mapping we have for the first time molecularly identified the gene responsible for a mutationally
23
defined phenotypic locus in an entomopathogenic nematode. In the process we observed an
24
unexpected mutational spectrum following EMS mutagenesis in this species. We find that the
25
ACCEPTED MANUSCRIPT
Downloaded from https://academic.oup.com/genetics/advance-article/doi/10.1093/genetics/iyad209/7469079 by California Institute of Technology user on 21 December 2023
Schwartz et al
Page
2
of
33
ortholog of the essential
C.
elegans
peroxidase gene
skpo
-
2
controls body size and shape in
1
S.
hermaphroditum
. We confirmed this identification by generating additional loss
-
of
-
function
2
mutations in the gene using CRISPR
-
Cas9. We propose that the identification of
skpo
-
2
will
3
accelerate gene targeting in other
Steinernema
entomopathogenic nematodes used
4
commercially in pest control, as
skpo
-
2
is X
-
linked and males hemizygous for loss of its
function
5
can mate, making
skpo
-
2
an easily recognized and maintained marker for use in co
-
CRISPR
.
6
7
Introduction
8
Entomopathogenic nematodes of the genera
Steinernema
and
Heterorhabditis
reside in the soil
9
as developmentally arrested dispersal
-
stage infective juvenile (IJ) larvae (Dillman and Sternberg
10
2012; Schwartz 2015). Upon encountering a suitable insect host, an entomopathogenic
11
nematode invades its body and resumes development, releasing endosymbiotic pathogenic
12
bacteria from its intestine into its host
(Dziedziech
et al.
2020)
.The nematode and its bacterial
13
symbiote rapidly kill the insect and convert the carcass into an incubator for the
14
nematode
-
bacterial pair. When the carcass is exhausted of nutrients, a subsequent generation
15
of IJs, each carrying pathogenic bacteria, disperse to begin the process anew. The
16
entomopathogenic nematode lifecycle offers an opportunity to study the development and
17
behavior of parasitic nematodes and their interactions with their bacterial symbiotes and their
18
insect prey, along with other aspects of their biology shared with or differing from those
19
described in other nematodes.
20
The extensively described biology of the free
-
living soil nematode
Caenorhabditis elegans
offers
21
a model for establishing entomopathogenic nematodes as a tool for laboratory research. Work
22
on
C.
elegans
has provided major contributions to our understanding of development and
23
disease (Horvitz 2003; Sulston 2003; Brenner 2003; Fire 2007; Mello 2007) in part because
24
C.
elegans
is a small animal with a rapid generation time and reproduces by selfing
25
ACCEPTED MANUSCRIPT
Downloaded from https://academic.oup.com/genetics/advance-article/doi/10.1093/genetics/iyad209/7469079 by California Institute of Technology user on 21 December 2023
Schwartz et al
Page
3
of
33
hermaphroditism (Apfeld and Alper 2018; Singh 2021). More recently, CRISPR
-
Cas9 genome
1
editing has opened new possibilities for exploring gene function
(Frøkjær
-
Jensen 2013)
.
2
We are developing the entomopathogenic nematode
Steinernema hermaphroditum
into a
3
similarly tractable and powerful platform for laboratory research. This would enable research
4
into aspects of the entomopathogenic nematode life cycle not amenable to study in previously
5
available nematode species, such as interactions between the nematodes and their bacterial
6
symbiotes, or specific to this nematode, such as its unusual mode of reproduction. First reported
7
in 2000 from studies in the Moluccan islands of Indonesia,
S.
hermaphroditum
was
8
subsequently lost until its rediscovery outside New Delhi was reported in 2019
(Griffin
et al.
9
2000; Stock
et al.
2004; Bhat
et al.
2019)
. We recently reported that
S.
hermaphroditum
10
consistently reproduces as a selfing hermaphrodite, established an inbred wild
-
type strain and
11
protocols for its propagation in the laboratory, and used chemical mutagenesis screens to
12
recover mutants that we complementation tested, genetically mapped, and cryopreserved
(Cao
13
et al.
2022)
. No other entomopathogenic nematode species is known to reproduce as
14
hermaphrodites in every generation.
15
The first entomopathogenic nematode mutants described had a short body size (Dumpy, or
16
Dpy) phenotype
(Zioni (Cohen
-
Nissan)
et al.
1992; Rahimi
et al.
1993; Tomalak, M. 1994)
.
17
Continuing our development of
S.
hermaphroditum
as a platform for laboratory exploration, we
18
sought proof
-
of
-
principle for molecular identification of a mutationally defined locus. Through
19
whole
-
genome sequencing of three independent alleles of an X
-
linked gene with a strong Dpy
20
phenotype we identified
Sthm
-
skpo
-
2
, the
S.
hermaphroditum
ortholog of the
C.
elegans
21
peroxidase gene
Cel
-
skpo
-
2
, as the only mutated gene likely to be responsible for this Dpy
22
phenotype.
Sthm
-
skpo
-
2
mutants generated using CRISPR
-
Cas9 phenocopied and failed to
23
complement the existing Dpy mutants.
24
25
26
ACCEPTED MANUSCRIPT
Downloaded from https://academic.oup.com/genetics/advance-article/doi/10.1093/genetics/iyad209/7469079 by California Institute of Technology user on 21 December 2023
Schwartz et al
Page
4
of
33
Materials and Methods
1
Nematode genetics
2
Steinernema hermaphroditum
strains were derived from the inbred wild
-
type strain PS9179 and
3
cultured with the bacterial strains
Xenorhabdus griffiniae
HGB2511 and
Comamonas sp
.
4
DA1877 as food sources
(Cao
et al.
2022)
. Individual HGB2511 or DA1877 colonies were grown
5
overnight at room temperature in 20
g/L Proteose Peptone
No.
3 containing 0.1%
sodium
6
pyruvate and dispensed onto agar media in Petri plates to grow bacterial lawns as a food
7
source for
S.
hermaphroditum
. HGB2511 lawns were grown on NGM agar media as described
8
(Cao
et al.
2022)
. DA1877 lawns were grown on Enriched Peptone Plates except with
9
1.8%
(weight/volume) agar (Evans 2006).
10
Caenorhabditis elegans
were derived from the wild
-
type strain N2 and cultured on
E.
coli
OP50
11
(Brenner 1974). Existing
C.
elegans
mutants obtained for use in this study included
12
skpo
-
1(ok1640)
II
and
mlt
-
7(tm1794)
II
, along with the balancer chromosome
13
tmC6[dpy
-
2(tmIs1189)]
II
(Thein
et al.
2009; Tiller and Garsin 2014; Dejima
et al.
2018)
. Existing
14
S.
hermaphroditum
mutants used included
unc(sy1647)
,
dpy(sy1639)
X
,
dpy(sy1644)
X
,
15
dpy(sy1646)
X
,
dpy(sy1662)
X
, and
unc(sy1636)
X
(Cao
et al.
2022)
.
16
A genetic screen for visible phenotypic mutants of
S.
hermaphroditum
was performed using
17
ethyl methansulfonate (EMS) mutagenesis as described
(Cao
et al.
2022)
. A single phenotypic
18
mutant, PS9839
dpy(sy1926)
X
, was recovered. Complementation tests were performed using
19
dpy(sy1926)
and the X
-
linked Dpy mutants
sy1646
and
sy1662
, marked with
unc(sy1636)
X
to
20
identify cross progeny.
21
22
DNA sequencing and analysis
23
Genomic DNA was prepared essentially as described, except without grinding of frozen animals
24
(Emmons
et al.
1979)
. Animals were grown on 10
cm Petri plates containing NGM agar with a
25
lawn of HGB2511 bacteria. Animals were washed repeatedly in M9 buffer and digested using
26
ACCEPTED MANUSCRIPT
Downloaded from https://academic.oup.com/genetics/advance-article/doi/10.1093/genetics/iyad209/7469079 by California Institute of Technology user on 21 December 2023
Schwartz et al
Page
5
of
33
proteinase
K in the presence of SDS and beta
-
mercaptoethanol. Lysate was extracted with
1
phenol/chloroform/isoamyl
alcohol followed by chloroform. Nucleic acids were precipitated from
2
the aqueous fraction using ethanol and recovered by spooling. RNA was removed by digestion
3
with RNase
A, after which DNA was recovered by ethanol precipitation. Purified DNA was sent
4
to Novogene (Sacramento, CA) for Illumina sequencing with a target of 26.6 million paired
-
end
5
150
nt reads for each sample.
6
Analysis of high
-
throughput sequencing data was adapted from a published pipeline for
7
C.
elegans
(Smith and Yun 2017). Sequencing reads were filtered using BBTools bbduk
8
(http://sourceforge.net/projects/bbmap/) to remove reads matching an assembly of
X.
griffiniae
9
HGB2511 genome sequence (Jennifer Heppert and Heidi Goodrich
-
Blair, personal
10
communication). Reads were mapped to a draft annotated
S.
hermaphroditum
PS9179 genome
11
(Erich Schwarz, personal communication), reads were sorted, duplicate reads were removed,
12
and reads were indexed using Samtools
(Danecek
et al.
2021)
. Mutations were detected using
13
Freebayes (Garrison and Marth 2012) and were mapped onto gene models and categorized for
14
coding changes using ANNOVAR
(Wang
et al.
2010)
. Annotated changes were sorted,
15
compared, and counted using Excel (Microsoft, Redmond, WA).
16
Individual animals or small groups of animals were lysed and sequences were amplified from
17
them using PCR as described for
C.
elegans
(Wicks
et al.
2001)
using oligonucleotide primers
18
whose sequences are listed in Table
S1. Restriction enzymes were obtained from New England
19
Biolabs (Beverly, MA). For Sanger sequencing, at least two PCR products were combined for
20
each sample; nucleic acid was purified using QiaQuick (QIAGEN, Germantown, Maryland) and
21
sent to Laragen for Sanger sequencing (Laragen, Culver City, CA).
22
Homology searches of additional
Steinernema
nematodes were performed using
23
BLAST
2.2.24+ on a Debian GNU server
(Altschul
et al.
1990)
using genome and transcriptome
24
assemblies downloaded from the NCBI or from WormBase ParaSite
(Howe
et al.
2017)
;
25
accession numbers were
Steinernema carpocapsae
GCA_000757645.3 (DNA),
Steinernema
26
ACCEPTED MANUSCRIPT
Downloaded from https://academic.oup.com/genetics/advance-article/doi/10.1093/genetics/iyad209/7469079 by California Institute of Technology user on 21 December 2023
Schwartz et al
Page
6
of
33
carpocapsae
WBPS16 (mRNA),
Steinernema diaprepesi
GCA_013436035.1,
Steinernema
1
feltiae
GCA_000757705.1,
Steinernema glaseri
GCA_000757755.1,
Steinernema
2
hermaphroditum
GCA_030435675.1 (DNA and mRNA),
Steinernema khuongi
3
GCA_016648015.1,
Steinernema monticolum
GCA_000505645.1, and
Steinernema scapterisci
4
GCA_000757745.1
(Dillman
et al.
2015; Serra
et al.
2019; Baniya
et al.
2019; Baniya and
5
DiGennaro 2021)
. MEGA11 software
(Tamura
et al.
2021)
was used to generate a
6
neighbor
-
joining phylogeny of predicted peroxidases identified by a BLAST search of the
7
C.
elegans
proteome as having significant homology to
Sthm
-
SKPO
-
2, and the predicted
8
S.
hermaphroditum
proteins closely related to them, using protein sequences from
C.
elegans
9
version WS290
(Davis
et al.
2022)
and from
S.
hermaphroditum
GCA_030435675.1.
10
11
Identification of candidate genes from whole
-
genome sequencing
12
To identify the mutations responsible for the Dpy phenotypes of our three sequenced strains, we
13
first searched for genes on the X
chromosome that had coding mutations in all three strains,
14
ideally distinct mutations (each of the
four alleles was descended from an independently
15
mutagenized P
0
animal, and
sy1926
was recovered in a separate screen from the first three
16
alleles). Mutations were expected to be single
-
nucleotide C
-
to
-
T changes consistent with EMS
17
mutagenesis
(Anderson 1995; Volkova
et al.
2020)
. These criteria resulted in four candidates,
18
encoding the hypothetical proteins QR680_001060, QR680_001389, QR680_001390, and
19
QR680_002483. Further inspection suggested the latter three candidates were likely the result
20
of sequencing and software issues: the mutations associated with these three candidates were
21
defined by low read counts that had low quality scores. Proteins 001389 and 001390 are
22
encoded by neighboring genes and include nearly identical sequence; these two genes have 14
23
different mutations annotated between them among the three strains, which did not seem
24
consistent with the mutations having arisen after mutagenesis and being causative for the Dpy
25
phenotype. Protein 002483 has 14 mutations annotated, of which three were annotated in more
26
ACCEPTED MANUSCRIPT
Downloaded from https://academic.oup.com/genetics/advance-article/doi/10.1093/genetics/iyad209/7469079 by California Institute of Technology user on 21 December 2023
Schwartz et al
Page
7
of
33
than one strain; this also is not consistent with the gene having been mutated to cause the Dpy
1
phenotype. By contrast, the gene encoding protein 001060 has only three mutations annotated
2
among the three strains, one in each strain; all three annotations have high read counts and
3
quality scores.
4
5
Homology analysis of candidate genes
6
We examined the four multiply mutated X
-
linked genes’ homology to assess them as
7
candidates. Predicted proteins 001389 and 001890 lack identifiable homologs by BLAST
8
searches, with none found even in the other available
Steinernema
genomes or the
9
Steinernema carpocapsae
transcriptome, and lack conserved domains identifiable by SMART or
10
by Pfam
(Letunic
et al.
2021; Mistry
et al.
2021)
.
11
The closest characterized homolog of protein 002483 is in
C.
elegans
, Cel
-
HGRS
-
1; protein
12
002483 is also the predicted
S.
hermaphroditum
protein most closely related to Cel
-
HGRS
-
1.
13
RNAi
-
mediated inactivation of
Cel
-
hgrs
-
1
causes a Dpy phenotype and other defects
(Kamath
14
et al.
2003)
. This reported Dpy phenotype nominated it as a viable candidate for the Dpy
15
phenotypes of our
S.
hermaphroditum
mutants, despite the low read count and the poor quality
16
scores of the sequence data implicating this gene.
17
The last of the four candidates is the gene encoding protein 001060, orthologous to
C.
elegans
18
Cel
-
SKPO
-
2, predicted to encode a peroxidase (see Figure
1B).
Cel
-
skpo
-
2
does not have a
19
reported abnormal mutant phenotype, but it is closely related to
Cel
-
mlt
-
7
, loss of which causes
20
defects in cuticle formation and molting along with nearly fully penetrant lethality and a Dpy
21
phenotype in the survivors (Figure
1C). Protein 001060 is more distantly related to the product
22
of
Cel
-
bli
-
3
, which mutates to cause a blistered cuticle defect;
bli
-
3
functions with
mlt
-
7
to
23
regulate cuticle structure, and other blister mutants genetically interact with cuticular Dumpy
24
phenotypes
(Higgins and Hirsh 1977; Cox
et al.
1980; Simmer
et al.
2003; Thein
et al.
2009)
.
25
This homology implicated the gene encoding protein 001060 as a strong candidate.
26
ACCEPTED MANUSCRIPT
Downloaded from https://academic.oup.com/genetics/advance-article/doi/10.1093/genetics/iyad209/7469079 by California Institute of Technology user on 21 December 2023