A sequencing-free assay for foliose Ulva species identification, hybrid detection and bulk biomass characterisation

Abstract

Sea Lettuce (Ulva spp. Ulvophyceae, Ulvaceae) has tremendous ecological and industrial impacts, from the negative effects of green tide events to the industrial production of food, feed, and value-added products. Due to the morphological similarities between Ulva species, their identification requires the use of “barcoding”, which relies on the sequencing of short fragments of DNA and the comparison of the obtained sequence to that of sequences present in public repositories. However, Sanger sequencing can be costly when hundreds of samples need to be analysed. In addition, “barcoding”, which uses Sanger sequencing, cannot be applied directly on bulk biomass, and requires independent assessment of the individuals within, which is often not possible in commercial products. Here, we describe a novel “sequencing-free” method for species identification of foliose Ulva species that by-passes the drawbacks associated with Sanger sequencing. The assay uses restriction digestion enzymes which target species-specific Single Nucleotide Polymorphisms (SNPs) present within the ITS1 sequence. Digestion of the ITS1 PCR product with two enzyme mixtures allows for the discrimination of the main foliose Ulva species, as well as U. compressa, which can have a foliose morphotype. Of the species tested, only U. pseudorotundata and U. arasakii show the same digestion pattern. Since those two Ulva species are allopatric, we expect this issue to be of limited impact for the users. In addition to species identification, we demonstrate that the assay can be used for hybrid detection, which can have interests for Ulva breeding and species delimitation. Importantly, the assay can be used to detect at once the different Ulva species present within a bulk of Ulva biomass, which could allow for traceability and characterisation of the purity of Ulva commercial products. This study provides a new quick and cost-effective method for foliose Ulva species identification that could readily be extended to other species.

Full text

Algal Research 55 (2021) 102280
2211-9264/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
A sequencing-free assay for foliose Ulva species identication, hybrid
detection and bulk biomass characterisation
Antoine Fort
a
,
*
, Charlene Linderhof
a
, In´
es Coca-Tagarro
a
, Masami Inaba
a
, Marcus McHale
a
,
Kevin Cascella
b
, Philippe Potin
b
, Michael D. Guiry
c
, Ronan Sulpice
a
a
National University of Ireland Galway, Plant Systems Biology Lab, Ryan Institute, SFI MaREI Centre for Climate, Energy and Marine, School of Natural Sciences,
Galway H91 TK33, Ireland
b
CNRS, Sorbonne Universit´
e Sciences, UMR 8227, Integrative Biology of Marine Models, Station Biologique de Roscoff, CS 90074, F-29688 Roscoff, France
c
AlgaeBase, Ryan Institute, National University of Ireland, Galway H91 TK33, Ireland
ARTICLE INFO
Keywords:
Barcoding
Cleaved Amplied Polymorphic Sequences
assay
Sea lettuce
Species identication
ABSTRACT
Sea Lettuce (Ulva spp. Ulvophyceae, Ulvaceae) has tremendous ecological and industrial impacts, from the
negative effects of green tide events to the industrial production of food, feed, and value-added products. Due to
the morphological similarities between Ulva species, their identication requires the use of “barcoding”, which
relies on the sequencing of short fragments of DNA and the comparison of the obtained sequence to that of
sequences present in public repositories. However, Sanger sequencing can be costly when hundreds of samples
need to be analysed. In addition, “barcoding”, which uses Sanger sequencing, cannot be applied directly on bulk
biomass, and requires independent assessment of the individuals within, which is often not possible in com-
mercial products. Here, we describe a novel “sequencing-free” method for species identication of foliose Ulva
species that by-passes the drawbacks associated with Sanger sequencing. The assay uses restriction digestion
enzymes which target species-specic Single Nucleotide Polymorphisms (SNPs) present within the ITS1
sequence. Digestion of the ITS1 PCR product with two enzyme mixtures allows for the discrimination of the main
foliose Ulva species, as well as U. compressa, which can have a foliose morphotype. Of the species tested, only U.
pseudorotundata and U. arasakii show the same digestion pattern. Since those two Ulva species are allopatric, we
expect this issue to be of limited impact for the users. In addition to species identication, we demonstrate that
the assay can be used for hybrid detection, which can have interests for Ulva breeding and species delimitation.
Importantly, the assay can be used to detect at once the different Ulva species present within a bulk of Ulva
biomass, which could allow for traceability and characterisation of the purity of Ulva commercial products. This
study provides a new quick and cost-effective method for foliose Ulva species identication that could readily be
extended to other species.
1. Introduction
Morphology-based identication of the common foliose Ulva species
has proven difcult over the years. While many Ulva species have been
shown to display subtle morphological differences [1,2], the plasticity of
those morphological characteristics in response to environmental
changes render species identication based solely on morphological
characters uncertain. For example, some Ulva species can display both
foliose and tubular morphotypes [3,4]. Hence, molecular identication
is the method of choice for Ulva species determination. However, mo-
lecular identication currently relies on the sequencing of PCR products
of barcode genes (typically rbcL, tufA and ITS in Ulva and other algae
[5–9]). Such sequencing is relatively costly and time-consuming when
used on hundreds of Ulva strains and can also lead to erroneous iden-
tications, as described in [10,11]. In the case of a strain-selection
programme where a single foliose Ulva species may be desirable
[12,13], the cost and labour associated with sequencing can be pro-
hibitive. Indeed, selecting individuals belonging to a given species re-
quires i) PCR amplication, ii) PCR products purication, iii)
sequencing of PCR products and iv) species identication based on the
sequencing results and comparison with database entries, which can
prove challenging when misidentications are present within the
* Corresponding author.
E-mail address: antoine.fort@nuigalway.ie (A. Fort).
Contents lists available at ScienceDirect
Algal Research
journal homepage: www.elsevier.com/locate/algal
https://doi.org/10.1016/j.algal.2021.102280
Received 18 October 2020; Received in revised form 26 February 2021; Accepted 5 March 2021

Algal Research 55 (2021) 102280
2
database [14]. When considering a breeding programme with thousands
of samples, a faster/cheaper approach is desirable. Moreover, trace-
ability of seaweed based products is of great importance, and the use of
Sanger sequencing does not allow for species identication within a bulk
biomass containing several species at once. An approach to circumvent
these issues is the detection of specic SNPs associated with a given
species by Cleaved Amplied Polymorphic Sequences (CAPS) assay.
CAPS assays use restriction enzymes digestion of PCR products to detect
the presence (or absence) of specic SNPs in the samples. We have
designed a CAPS assay that discriminates foliose Ulva species present in
the North East Atlantic based on species-specic SNPs in the ITS1
dataset (part of the 45S rDNA genes). Such a system is particularly
powerful for foliose Ulva samples as we previously reported a generally
low sequence divergence within species, and high between-species ge-
netic diversity [10]. Hence, individuals from the same species are likely
to share the same sequence when a small DNA locus is examined, such as
the ITS1. Using a set of 295 foliose strains and their associated ITS1
sequences, we have identied several species-specic SNPs that can be
revealed by restriction enzymes following PCR amplication of the ITS1
locus, allowing for routine species identication without barcoding. We
further describe the potential uses for such an assay, from the detection
of F1 hybrids to the qualitative determination of Ulva species present in
bulk biomass for commercial purposes.
2. Materials and methods
2.1. Foliose Ulva sample collection and DNA extraction
The samples used in this study, all foliose Ulva individuals of >100
cm
2
in diameter, originating from Ireland, the Netherlands, Portugal and
Brittany (France) are described in [10], in addition to 185 samples from
the same locations. DNA was extracted from ~5 mg of freeze dried
biomass using the magnetic-beads protocol described in [15].
2.2. ITS1 dataset
All ITS1 sequences originate from [14]. Briey, all sequences from
Ulva species in the NCBI database were downloaded, aligned using
MAFFT [16], trimmed using Trimal [17] and species delimitation per-
formed as described by [10] using a Generalized Mixed Yule Coalescent
(GMYC) model. The full alignment is available here as Supplementary
File 1. After assigning a species name to each sequence and retaining
only foliose species entries, the sequences were manually screened to
detect species-specic SNPs, and NEB cutter online tool (http://nc2.neb.
com/NEBcutter2/) used to select enzymes with restriction sites recog-
nising the species-specic SNPs.
2.3. DNA amplication, sanger sequencing and CAPS assay
The extracted DNA was amplied using a single set of primers
spanning the Internal Transcribed Spacer 1 of the 45S rDNA repeats:
CAPS_ITS1F (5′-TCGTTGAACCCTCCCGTTTA-3′) and CAPS_ITS1R (5′-
CGATGACTCACGGAATTCTGC-3′), rbcL primers SHF1 and SHR4 [18],
and the tufA primers from [7]. PCR amplication was performed in 25
μ
L reaction volume containing 1
μ
L of undiluted DNA, 0.5 pmol forward
and reverse primers, 9.25
μ
L of MilliQ water and 12.5
μ
L of MyTaq Red
mix (Bioline). The PCR programme was as follows: 95 ◦C for 3 min, 35
cycles of 95 ◦C for 15 s, 60 ◦C for 15 s and 72 ◦C for 15 s, and a nal
elongation step of 72 ◦C for 5 min. Amplicons were sent to LGC genomics
GmbH for Sanger sequencing. Species were determined based on
sequence similarity with Ulva species from [10,14]. NCBI accession
numbers are available as Table S1. For enzymatic digestion, 5
μ
L of the
raw PCR product was digested with the enzyme mixtures (BamHI +
HpaI +BfaI +PspOMI and CviQI +BtsCI, all Fisher Scientic, and T7
endonuclease, New England Biolabs), 0.1
μ
L of each enzyme, 0.5
μ
L
Tango Buffer (Fisher Scientic), and 4.4
μ
L H
2
O. Digestion was carried
out for a minimum of 4 h at 37 ◦C. T7 endonuclease +BamHI +CviQI +
BfaI +BtsCI digestion was carried out using buffer 2 (NEB) instead of
Tango buffer (Fisher Scientic). Gel electrophoresis was performed
using a 2% agarose gel in TAE (Tris-Acetate-EDTA) buffer for 40 min at
100 V. The ladder used was hyperladder 50 bp (Bioline). For bulk
biomass trials, the biomass (~1 kg FW) was freeze dried, and ground
into a ne powder. DNA extraction and PCR assay were performed as
above. For sensitivity trials, DNA from U. armoricana and U. australis was
pooled in different ratios, with decreasing concentrations of U. australis
DNA (from 50% to 1.5%), prior to PCR and restriction digestion using
HpaI, BtsCI and T7 endonuclease.
3. Results and discussion
3.1. Design of a sequencing-free assay for rapid foliose Ulva species
determination
Our own dataset contains sequences for 295 foliose Ulva strains
belonging to the seven foliose species we identied during our sampling
in the North East Atlantic [10], i.e. U. armoricana, U. rigida, U. gigantea,
U. pseudorotundata [≡Ulva rotundata], U. australis, U. ohnoi and U. fen-
estrata; nomenclatural authorities follow [19]. We have renamed our
previous U. laetevirens sequences into U. armoricana, given that the
recently sequenced U. laetevirens lectotype revealed that the name is in
fact synonymous to U. australis [20]. The clade containing all previous
“U. laetevirens” sequences is conspecic with the sequenced holotype of
U. armoricana [14,21,22], thus we reassigned all U. laetevirens sequences
into U. armoricana.
Two hundred and seventy-two (272) out of 295 individuals produced
full length ITS1 sequences. Based on these sequences, we identied re-
striction digestion sites containing SNPs present in all of the individuals
for a given species (Fig. 1A). We identied eight SNPs containing re-
striction sites for six enzymes that can be used for the discrimination of
foliose Ulva species. For example, all 100 U. armoricana individuals
possess a unique restriction site for the enzyme BfaI in the amplied
fragment of the ITS1, making it simple to discriminate U. armoricana
amplicons from those of the 6 other species after PCR product digestion
with BfaI.
The BamHI/HpaI/BfaI/PspOMI mixture discriminated U. armor-
icana, U. rigida, U. ohnoi and U. gigantea individuals, with the complete
digestion of the ITS1 PCR product and generation of a specic band
pattern for each species (Fig. 1B). U. pseudorotundata, U. australis and U.
fenestrata samples are not digested using this mixture. For those species,
another enzyme cocktail (CviQI/BtsCI) digested the ITS1 PCR product
and yielded a specic band pattern that distinguished between those
three species (Fig. 1B). Only U. australis showed a restriction site poly-
morphism within the 44 individuals of this species in our dataset, with
the rst CviQI restriction site being present in only 9 out of the 44 in-
dividuals. Nonetheless, this polymorphism creates a specic band
pattern that is different from samples of the 6 other species (Fig. S1).
Hence, this assay represents a sequencing-free, cheap, quick and precise
method to identify common foliose Ulva species, particularly for the
correct identication required for use as food, feed or industrial
applications.
3.2. The CAPS assay allows for F1 hybrid detection
In addition to species discrimination, this assay allows for the un-
equivocal detection of interspecic hybrids. Hybridity is rare between
foliose Ulva species [10], and we found a single occurrence of a F1
hybrid between U. armoricana and U. rigida within a set of 110 in-
dividuals. Using the enzymes described above, hybrids can be readily
detected by the presence of intermediate band patterns indicating a
mixture of two Ulva species in the same sample. We used our assay on
the U. armoricana X U. rigida hybrid (strain U99) detected using next-
generation sequencing in [10] and were able to detect it as a hybrid
A. Fort et al.

Algal Research 55 (2021) 102280
3
Fig. 1. Sequencing-free species determination assay
for foliose Ulva. A) Result of the amplication of the
ITS1 in each species. Restriction sites are indicated
above the lines, with numbers representing the
number of individuals of the species harbouring the
restriction site. Expected band sizes are indicated. B)
Gel results of the Cleaved Amplied Polymorphic
Sequences (CAPS) assay. Top: undigested PCR
amplicon of the ITS1, middle: amplicon digested with
CviQI+BtsCI, bottom: amplicon digested with Bam-
HI+BfaI+HpaI+PspOMI). U. australis sample does
not contain the rst CviQI site.
Fig. 2. Comparison of species attribution based on sequencing data and the CAPS assay. Each column represents a strain, coloured according to the species iden-
tication result based on the CAPS assay (top row) or sanger sequencing data of the ITS1, tufA or rbcL. No sequencing trace indicates strains which failed to produce
satisfactory Sanger sequencing results. F1 hybrids uncovered by the CAPS assay are indicated in black.
A. Fort et al.

Algal Research 55 (2021) 102280
4
(Fig. S2). Indeed, digestion of the ITS1 product of strain U99 with BfaI
(specic to U. armoricana) showed an incomplete digestion of the PCR
product, indicating the presence of U. armoricana DNA and that of
another species. When digested with BamHI (specic to U. armoricana
and U. rigida), complete digestion was observed. Hence, DNA sample
from strain U99 contains a mixture of U. armoricana and U. rigida nu-
clear DNA, which is explained by the fact that U99 is a F1 hybrid be-
tween those two species. We used the CAPS assay on 185 other foliose
Ulva samples in addition to the hybrid U99, to investigate whether other
hybrids were present. We detected one additional F1 hybrid (Fig. 2),
again between U. armoricana and U. rigida.
Furthermore, we found two individuals with an apparent discrep-
ancy in their species identity when using organellar marker genes or the
ITS1 nuclear marker gene (Fig. 2), the organellar genome belonging to
U. armoricana and the homozygous nuclear ITS1 locus belonging to U.
rigida. These discrepancies are most likely due to previous hybridization
events between the two species, which, following sexual reproduction
and crossing over events, have lost the ITS1 allele from one of the
original parental species. Thus, these two individuals are at least two
generations older than the original hybridization event and named F2+.
Interestingly, this indicates that F1 hybrids between U. armoricana and
U. rigida are likely sexually competent and can generate offspring.
Hence, population boundaries between the two species are unlikely to
be solely determined by reproductive barriers. Those results point to-
wards uniparental inheritance of the organellar genome in Ulva, since all
ve hybrids appear homozygous for the chloroplast barcodes, in
agreement with previous reports [10,23,24].
Since the CAPS assay does not use the organellar genome, sequencing
is still required for the determination of the species of the maternal
parent. However, [25] have recently described an assay similar to the
one presented here using structural differences in the chloroplast
genome of some Ulva species, also described in [10] and targeting the
chloroplast genome. Their assay is similarly PCR-based and allows the
discrimination of U. ohnoi, U. lactuca, U. australis and U. fenestrata. In
combination with the CAPS assay described here, researchers have ac-
cess to a powerful screening method for foliose Ulva species determi-
nation and hybrid (F1 and F2+) detection.
3.3. Species determination in bulk biomass
With the rise of seaweed production worldwide for food and fodder
[26], comes the potential issue of traceability of seaweed products.
Indeed, seaweed pellets and seaweed in bulk biomass are difcult to
characterise for the presence of different species. When a mixture of
species is present within a seaweed product, barcoding cannot readily be
used. In effect, amplifying a fragment such as the ITS or other barcodes
from a DNA sample containing several species will not give an adequate
result using Sanger sequencing, due to the presence of different DNA
sequences within the same amplicon. Solutions to overcome this prob-
lem are to i) clone the PCR products and sequence each PCR product
individually, or ii) use next-generation sequencing to quantify the
presence of reads belonging to different species. While these solutions
are working, they are relatively costly and time-consuming. Using the
assay described here, it is possible to detect the presence of different
foliose Ulva species within seaweed bulk biomass. Using a combination
of the enzymes of the CAPS assay, a “ngerprint” of the species con-
tained within the samples can be obtained. We tested the assay on seven
samples of seaweed biomass from green-tide areas in Brittany (four
samples taken near Roscoff [ROS], and three close to Saint Brieuc [StB],
Fig. 3). Two bands were visible following amplication of the ITS1 in
four of the seven samples, potentially indicating the presence of more
than one species. We then digested the PCR product with a sequential
combination of enzymes (see Fig. 3): CviQI/BtsCI, specic to U. australis,
U. fenestrata and U. pseudorotundata, BfaI, specic to U. armoricana;
BamHI, specic to U. armoricana, U. rigida and U. ohnoi; PspOMI, specic
to U. ohnoi.
Digestion with CviQI/BtsCI revealed the presence of U. australis in
four of the seven bulk samples (ROS2, containing the double CviQI U.
australis haplotype, and StB1, 2 and 3 samples, containing the single U.
australis CviQI haplotype). Digestion using CviQI/BtsCI/BfaI then
revealed that all seven samples contained U. armoricana, with a leftover
undigested band probably indicating the presence of U. rigida. This was
conrmed by further digesting with BamHI restriction enzyme, which
lead to the complete digestion of the PCR product in all samples.
Digestion with PspOMI did not cleave the PCR product (Fig. S3), indi-
cating that the leftover DNA band undigested by CviQI/BtsCI/BfaI
indeed belongs to U. rigida.
However, faint leftover bands remained in four samples (Fig. 3, ar-
rows). This band could indicate the presence of another species which
does not belong to foliose Ulva, but it could also be from a PCR artefact.
Indeed, when multiple-template PCRs are used, heteroduplex formation
of PCR products can occur [27], leading to a nal PCR amplicon effec-
tively containing the amplicon from one species in the plus strand, and
that of another species in the minus strand. Those amplicons cannot be
cut with classical restriction enzymes such as the ones used here since
they require both strands at the restriction site to be complementary. We
conrmed that this additional band is a PCR artefact by performing a
PCR on a mixture of U. armoricana and U. australis DNA, which indeed
led to the formation of heteroduplexes, and an HpaI +CviQI (which
should cleave ITS1 PCR products of all foliose Ulva) resistant band. This
does not occur when PCR products from both species are mixed and then
digested (Fig. S3). Removal of PCR artefacts such as heteroduplexes can
be readily achieved using a T7 endonuclease, which will cut PCR
amplicons containing mismatches between the two strands. Adding a T7
endonuclease to the nal digestion with CviQI/BtsCI/BfaI/BamHI
effectively removed the extraneous band (Fig. 3), conrming that all
seven bulk samples only contain DNA from U. australis, U. rigida and U.
armoricana. Hence, in order to remove the heteroduplexes formed dur-
ing PCRs on DNA containing several species, we strongly recommend
Fig. 3. Ulva bulk biomass analysis using the CAPS assay. ITS1 PCR amplicons
and digestion with the addition of different enzymes, demonstrating the pres-
ence of multiple species. Stars indicate samples digested by the enzyme.
A. Fort et al.

Algal Research 55 (2021) 102280
5
the use of the T7 endonuclease at the same time as the restriction
enzyme digestion. It will avoid the detection of restriction-resistant
amplicons due to PCR artefacts and a remaining band will indicate the
presence of another alga or protist.
While this assay is not quantitative because it relies on the ampli-
cation of the ITS1, itself part of the 45S rDNA repeats which can vary in
number between species [28,29], it does provide evidence for the
presence of different Ulva species and other possible contaminants
within bulk biomass. Regarding sensitivity, we used a synthetic bulk
sample containing a mixture of two species (U. armoricana and U. aus-
tralis), with a dynamic range of 50% to 1.5% DNA originating from U.
australis. The assay could detect the presence of U. australis in the syn-
thetic bulk sample when its abundance was in the range of 50 to 6.25%
of the total DNA, by the presence of an undigested band following HpaI
digestion and complete digestion using HpaI/BtsCI (Fig. 4). The band
corresponding to U. australis disappeared when below 6.25% of total
DNA, indicating that the assay has a minimum sensitivity threshold
between 3.12 and 6.25%. A similar detection threshold would require
the sequencing of >30 clones of PCR amplicons. For higher sensitivity,
Next-Generation sequencing would be required, but the cost and time
associated to such methodology can be prohibitive. Overall, our enzy-
matic assay could be used for traceability and detection of the purity of
Ulva products, and this method could be extended to other commercial
macroalgal species.
3.4. Robustness of the CAPS assay based on ITS1 sequences in the
literature
We investigated the robustness of the assay by mining an ITS1
dataset containing all Ulva ITS1 vouchers in the NCBI database (in
addition to those of this study, see [14]) for the presence of restriction
sites for the 6 enzymes. We restricted our comparison to distromatic
foliose Ulva species, which can readily be distinguished from mono-
stromatic tubular Ulva species prior to DNA analysis. We however added
the normally monostromatic and tubular U. compressa [≡Enteromorpha
compressa], as this species can occasionally display a foliose morphotype
[30]. It should be noted that this analysis only counts a restriction site as
present or absent if the sequence does not possess gaps/undetermined
nucleotide at those positions. Of the ~40 species present in the ITS1
NCBI dataset, 11 belong to distromatic foliose species, of which seven
were present in our study, in addition to U. compressa. The assay appears
very robust for 10 of the 12 species (Fig. 5): all of U. armoricana, U.
australis, U. fenestrata, U. ohiohilulu and U. expansa individuals harbour
the same discriminating restriction sites, and only minor discrepancies
were found in U. rigida (1/47 sequences), U. lactuca (4/25 sequences), U.
ohnoi (1/29 sequences) and U. compressa (1/85 sequences). The pres-
ence of rare allelic variants in individuals of those species could be
genetically true but could also be explained by errors during sequencing,
or the presence of F1 hybrids/contaminating DNA, which we cannot
assess without access to the chromatogram trace or repeating the Sanger
sequencing on those entries. The only case where the assay cannot
discriminate between distromatic foliose species is between U. pseu-
dorotundata and U. arasakii, which both contain the same restriction
sites for CviQI, despite considerable genetic variation between the two
species (Fig. 5). However, U. arasakii [31] is currently known only from
the Sea of Japan, while U. pseudorotundata is only known from Ireland,
Brittany and Portugal, an inability to distinguish between them using
this assay is not an issue since both species are unlikely to be present in
the same areas.
3.5. Limitations of the methodology
Our methodology is aimed at performing species identication of
single samples and bulk biomass of foliose Ulva species. Outside of the
possible presence of rare allelic variants potentially producing erroneous
species identication (Fig. 5), the assay is well suited to single foliose
Ulva sample species identication. It can then be used for species se-
lection prior to strain selection and ecological studies. For bulk biomass,
as demonstrated in Fig. 4, low-abundance foliose species within the bulk
biomass can be missed. In addition, the presence of species other than
foliose Ulva within the bulk biomass, for example in Ulva blooms con-
taining a mixture of tubular and foliose morphotypes [32], could pro-
duce unexpected band patterns. Similarly, the presence of contaminants
can only be detected if all species in the bulk biomass can be amplied
by the CAPS primers described here. For the highest precision, sensi-
tivity and resolution, the use of next-generation-sequencing (NGS) is
required, with the caveat that it requires a signicantly higher cost and
results turnaround, as well as expertise in the analysis of metagenomics
data.
For F2+hybrid detection, it should be noted, that this assay requires
an organellar marker and that it can miss F2+hybrids if nuclear DNA
recombinations during the sexual reproduction of the F1 generation
occur outside of the 45S rDNA repeats. In such a case, the individual can
be homozygous for the ITS1 locus but heterozygous in other loci in the
genome. Hence, the presence of F2+hybrids is likely to be under-
estimated using this method. However, sanger sequencing of barcodes
would not produce better results, and only NGS [33,34] or micro-
satellites [35] analysis can conclusively infer the degree of hybridization
Fig. 4. Sensitivity of the CAPS assay in synthetic bulk samples. DNA of U. armoricana and U. australis was mixed with decreasing concentration of U. australis and
detected using digestion of the ITS1 PCR product with HpaI and BtsCI. The presence of an undigested band is indicated by a star.
A. Fort et al.

Algal Research 55 (2021) 102280
6
within a population.
Finally, this assay is not aimed at providing higher level resolution
than species level given that it uses extremely conserved species-specic
SNPs. For the detection of ecotypes/haplotypes/geographical origin of
individuals, a combination of several barcodes can be used instead
[36,37], or for the highest resolution, next-generation-sequencing
[38–40].
4. Conclusions
Our CAPS assay can thus be used to accurately, cheaply and quickly
select individuals from a species of scientic or commercial interest
during strain selection, or to assess the distribution of foliose Ulva spe-
cies within areas. Importantly, this assay is accessible to any lab with a
minimum of molecular biology equipment, as it only requires a PCR
machine, and an electrophoresis/gel imaging system. Additionally, this
assay is cost-effective as a single reaction containing all enzymes costs
€
0.50 [US$0.59], compared with ~
€
4.00 [US$4.70] per Sanger
sequencing reaction using service providers. This assay could be
extended to tubular Ulva species, and we provide the ITS1 alignment
used for validation purposes (containing all Ulva ITS1 sequences in the
NCBI, excluding those without species information) in Supplementary
File 1. The concept of CAPS assay for species identication could also be
used on any other taxa of interest for fundamental ecological research
and industry traceability purposes.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.algal.2021.102280.
Funding
This work was funded by the European Union Horizon 2020 pro-
gramme (project ID 727892, GenialG - GENetic diversity exploitation for
Innovative Macro-ALGal biorenery, http://genialgproject.eu/), SFI
Frontiers for the Future (Project Pristine Coasts, grant number 19/FFP/
6841) and the European Union Northern Periphery and Arctic Pro-
gramme (project number 366, SW-GROW - Innovations for Seaweed
Producers in the Northern Periphery Area project; http://sw-grow.
interreg-npa.eu/).
CRediT authorship contribution statement
Antoine Fort: conceptualization, methodology, writing,
Fig. 5. Prediction of ITS1 digestion pattern of foliose
Ulva species based on all the ITS1 sequences from the
literature. Top: restriction sites are indicated above
the lines. Numbers represent the number of in-
dividuals of the species harbouring the restriction
site. Rare allelic variants are indicated with a pale
shading. Bottom: predicted gel results obtained from
the digestion of the PCR product with the enzyme
mixtures BamHI+BfaI+HpaI+PspOMI (number 1)
and CviQI+BtsCI (number 2).
A. Fort et al.

Algal Research 55 (2021) 102280
7
visualisation, formal analysis; Charlene Linderhof: formal analysis;
In´
es Coca-Tagarro: formal analysis; Masami Inaba: formal analysis;
Marcus McHale: methodology; Michael D Guiry: writing-review &
editing; Kevin Cascella: project administration; Philippe Potin: project
administration, funding acquisition; Ronan Sulpice: writing, supervi-
sion, funding acquisition.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgments
The authors would like to thank Isabel Azevedo (CIIMAR), Ricardo
Bermejo (NUI Galway), Adrie van der Werf and Annelies Beniers
(Wageningen University), Paolo Ruggeri (Station Biologique de Rosc-
off), Helena Abreu (Alga +), Alexandre Berthelot (France Haliotis),
Marie-Mathilde Perrineau and Phil Anderson (Scottish Association For
Marine Science) for providing some of the strains used in this study.
Statement of informed consent, human/animal rights
No conicts, informed consent, or human or animal rights are
applicable to this study.
References
[1] M.M. Flagella, N. Andreakis, M. Hiraoka, M. Verlaque, M.C. Buia, Identication of
cryptic Ulva species (Chlorophyta, Ulvales) transported by ballast water, J. Biol.
Res. 13 (2010) 47.
[2] H.W. Lee, J.C. Kang, M.S. Kim, Taxonomy of Ulva causing blooms from Jeju Island,
Korea with new species, U. pseudo-ohnoi sp. nov. (Ulvales, Chlorophyta), Algae 34
(2019) 253–266.
[3] I.H. Tan, J. Blomster, G. Hansen, E. Leskinen, C.A. Maggs, D.G. Mann, H.
J. Sluiman, M.J. Stanhope, Molecular phylogenetic evidence for a reversible
morphogenetic switch controlling the gross morphology of two common genera of
green seaweeds Ulva and Enteromorpha, Mol. Biol. Evol. 16 (1999) 1011–1018.
[4] S. Steinhagen, F. Weinberger, R. Karez, Molecular analysis of Ulva compressa
(Chlorophyta, Ulvales) reveals its morphological plasticity, distribution and
potential invasiveness on German North Sea and Baltic Sea coasts, Eur. J. Phycol.
54 (2019) 102–114.
[5] R. Miladi, A. Manghisi, S.A. Minicante, G. Genovese, S. Abdelka, M. Morabito,
A DNA barcoding survey of Ulva (Chlorophyta), Tunisia and Italy reveals the
presence of the overlooked alien U. ohnoi, Cryptogamie, Algologie (2018), https://
doi.org/10.7872/crya/v39.iss1.2018.85.
[6] L. Kirkendale, G.W. Saunders, P. Winberg, A molecular survey of Ulva
(Chlorophyta) in temperate Australia reveals enhanced levels of cosmopolitanism,
J. Phycol. 49 (2013) 69–81.
[7] G.W. Saunders, H. Kucera, An evaluation of rbcL, tufA, UPA, LSU and ITS as DNA
barcode markers for the marine green macroalgae, Cryptogam. Algol. 31 (2010)
487–528.
[8] C.J. O’Kelly, A. Kurihara, T.C. Shipley, A.R. Sherwood, Molecular assessment of
Ulva spp. (Ulvophyceae, Chlorophyta) in the Hawaiian islands, J. Phycol. 46 (2010)
728–735.
[9] S. Steinhagen, R. Karez, F. Weinberger, Cryptic, alien and lost species: molecular
diversity of Ulva sensu lato along the German coasts of the North and Baltic Seas,
Eur. J. Phycol. 54 (2019) 466–483.
[10] A. Fort, M. McHale, K. Cascella, P. Potin, B. Usadel, M.D. Guiry, R. Sulpice, Foliose
Ulva species show considerable inter-specic genetic diversity, low intra-specic
genetic variation, and the rare occurrence of inter-specic hybrids in the wild,
J. Phycol. 57 (2020) 219–233.
[11] J.R. Hughey, C.A. Maggs, F. Mineur, C. Jarvis, K.A. Miller, S.H. Shabaka, P.
W. Gabrielson, Genetic analysis of the Linnaean Ulva lactuca (Ulvales, Chlorophyta)
holotype and related type specimens reveals name misapplications, unexpected
origins, and new synonymies, J. Phycol. 55 (2019) 503–508.
[12] A. Fort, M. Lebrault, M. Allaire, A.A. Esteves-Ferreira, M. McHale, F. Lopez, J.
M. Fari˜
nas-Franco, S. Alseekh, A.R. Fernie, R. Sulpice, Extensive variations in
diurnal growth patterns and metabolism among Ulva spp. strains, Plant Physiol.
180 (2019) 109–123.
[13] A. Fort, C. Mannion, J.M. Fari˜
nas-Franco, R. Sulpice, Green tides select for fast
expanding Ulva strains, Sci. Total Environ. 698 (2020), 134337.
[14] A. Fort, M. McHale, K. Cascella, P. Potin, M.-M. Perrineau, P. Kerrison, E. da Costa,
R. Calado, M. Domingues, I.C. Azevedo, Exhaustive reanalysis of barcode
sequences from public repositories highlights ongoing misidentications and
impacts taxa diversity and distribution: a case study of the Sea Lettuce, Authorea
Preprints, (2020).
[15] A. Fort, M.D. Guiry, R. Sulpice, Magnetic beads, a particularly effective novel
method for extraction of NGS-ready DNA from macroalgae, Algal Res. 32 (2018)
308–313.
[16] K. Katoh, J. Rozewicki, K.D. Yamada, MAFFT online service: multiple sequence
alignment, interactive sequence choice and visualization, Brief. Bioinform. 20
(2019) 1160–1166.
[17] S. Capella-Guti´
errez, J.M. Silla-Martínez, T. Gabald´
on, trimAl: a tool for automated
alignment trimming in large-scale phylogenetic analyses, Bioinformatics 25 (2009)
1972–1973.
[18] S. Heesch, J.E.S. Broom, K.F. Neill, T.J. Farr, J.L. Dalen, W.A. Nelson, Ulva,
Umbraulva and Gemina: genetic survey of New Zealand taxa reveals diversity and
introduced species, Eur. J. Phycol. 44 (2009) 143–154.
[19] M. Guiry, G. Guiry, AlgaeBase. World-Wide Electronic Publication, National
University of Ireland, Galway. 2020, URL: http://www. algaebase. org, (2020).
[20] J.R. Hughey, P.W. Gabrielson, C.A. Maggs, F. Mineur, K.A. Miller, Taxonomic
revisions based on genetic analysis of type specimens of Ulva conglobata,
laetevirens U., pertusa U., spathulata U., (Ulvales, Chlorophyta), Phyco. Res.
(2020), https://doi.org/10.1111/pre.12450.
[21] G. Coat, P. Dion, M.-C. Noailles, B. De Reviers, J.-M. Fontaine, Y. Berger-Perrot, S.
Loiseaux-De Go¨
er, Ulva armoricana (Ulvales, Chlorophyta) from the coasts of
Brittany (France). II Nuclear rDNA ITS sequence analysis, Eur. J. Phycol. 33 (1998)
81–86.
[22] P. Dion, B. De Reviers, G. Coat, Ulva armoricana sp. nov.(Ulvales, Chlorophyta)
from the coasts of Brittany (France). I. Morphological identication, Eur. J. Phycol.
33 (1998) 73–80.
[23] T. Bråten, Autoradiographic evidence for the rapid disintegration of one
chloroplast in the zygote of the green alga Ulva mutabilis, J. Cell Sci. 12 (1973)
385–389.
[24] Y. Kagami, Y. Mogi, T. Arai, M. Yamamoto, K. Kuwano, S. Kawano, Sexuality and
uniparental inheritance of chloroplast DNA in the isogamous green alga Ulva
compressa (Ulvophyceae), J. Phycol. 44 (2008) 691–702.
[25] C. Mitsuhashi, H. Teramura, H. Shimada, Construction of genomic marker sets
based on the chloroplast genome of a green alga, Ulva pertusa (syn. Ulva australis),
leads to simple detection of Ulva species, Genes Genetic Sys. (2020) 19-00054,
https://doi.org/10.1266/ggs.19-00054.
[26] A.H. Buschmann, C. Camus, J. Infante, A. Neori, ´
A. Israel, M.C. Hern´
andez-
Gonz´
alez, S.V. Pereda, J.L. Gomez-Pinchetti, A. Golberg, N. Tadmor-Shalev,
Seaweed production: overview of the global state of exploitation, farming and
emerging research activity, Eur. J. Phycol. 52 (2017) 391–406.
[27] J.R. Thompson, L.A. Marcelino, M.F. Polz, Heteroduplexes in mixed-template
amplications: formation, consequence and elimination by ‘reconditioning PCR’,
Nucleic Acids Res. 30 (2002) 2083–2088.
[28] J.A. Klappenbach, P.R. Saxman, J.R. Cole, T.M. Schmidt, rrndb: the ribosomal RNA
operon copy number database, Nucleic Acids Res. 29 (2001) 181–184.
[29] S.O. Rogers, A.J. Bendich, Ribosomal RNA genes in plants: variability in copy
number and in the intergenic spacer, Plant Mol. Biol. 9 (1987) 509–520.
[30] O. De Clerck, S.M. Kao, K.A. Bogaert, J. Blomme, F. Foonker, M. Kwantes,
E. Vancaester, L. Vanderstraeten, E. Aydogdu, J. Boesger, G. Califano, B. Charrier,
R. Clewes, A. Del Cortona, S. D’Hondt, N. Fernandez-Pozo, C.M. Gachon,
M. Hanikenne, L. Lattermann, F. Leliaert, X. Liu, C.A. Maggs, Z.A. Popper, J.
A. Raven, M. Van Bel, P.K.I. Wilhelmsson, D. Bhattacharya, J.C. Coates, S.
A. Rensing, D. Van Der Straeten, A. Vardi, L. Sterck, K. Vandepoele, Y. Van de Peer,
T. Wichard, J.H. Bothwell, Insights into the evolution of multicellularity from the
sea lettuce genome, Curr. Biol. 28 (2018) 2921–2933, e2925.
[31] M. Chihara, Ulva arasakii, a new species of green algae: its life history and
taxonomy, Bull, Nat. Sci. Mus. Tokyo 12 (1969) 849–861.
[32] M. Guidone, C.S. Thornber, Examination of Ulva bloom species richness and
relative abundance reveals two cryptically co-occurring bloom species in
Narragansett Bay, Rhode Island, Harmful Algae 24 (2013) 1–9.
[33] P. Zhao, H.-J. Zhou, D. Potter, Y.-H. Hu, X.-J. Feng, M. Dang, L. Feng, S. Zulqar,
W.-Z. Liu, G.-F. Zhao, Population genetics, phylogenomics and hybrid speciation of
Juglans in China determined from whole chloroplast genomes, transcriptomes, and
genotyping-by-sequencing (GBS), Mol. Phylogenet. Evol. 126 (2018) 250–265.
[34] P.A. Hohenlohe, S.J. Amish, J.M. Catchen, F.W. Allendorf, G. Luikart, Next-
generation RAD sequencing identies thousands of SNPs for assessing
hybridization between rainbow and westslope cutthroat trout, Mol. Ecol. Resour.
11 (2011) 117–122.
[35] J. Zhang, N. Yotsukura, A. Jueterbock, Z.-M. Hu, J. Assis, C. Nagasato, J. Yao,
D. Duan, Detecting no natural hybridization and predicting range overlap in
Saccharina angustata and Saccharina japonica, J. Appl. Phycol. 33 (2021) 693–702.
[36] P.-G. Sauriau, M. Dartois, V. Becquet, F. Aubert, V. Huet, M. Br´
eret, A. Viricel,
E. Pante, Multiple genetic marker analysis challenges the introduction history of
Ulva australis (Ulvales, Chlorophyta) on French coasts, Eur. J. Phycol. (2021)
1–13.
[37] T.T. Bringloe, H. Verbruggen, G.W. Saunders, Unique biodiversity in Arctic marine
forests is shaped by diverse recolonization pathways and far northern glacial
refugia, Proc. Natl. Acad. Sci. 117 (2020) 22590–22596.
[38] M.W. Horton, A.M. Hancock, Y.S. Huang, C. Toomajian, S. Atwell, A. Auton, N.
W. Muliyati, A. Platt, F.G. Sperone, B.J. Vilhj´
almsson, Genome-wide patterns of
A. Fort et al.

Algal Research 55 (2021) 102280
8
genetic variation in worldwide Arabidopsis thaliana accessions from the RegMap
panel, Nat. Genet. 44 (2012) 212–216.
[39] K. Xu, X. Yu, X. Tang, F. Kong, Y. Mao, Organellar genome variation and genetic
diversity of Chinese Pyropia yezoensis, Front. Mar. Sci. 6 (2019).
[40] J. Guzinski, M. Ballenghien, C. Daguin-Thi´
ebaut, L. L´
evˆ
eque, F. Viard, Population
genomics of the introduced and cultivated Pacic kelp Undaria pinnatida:
marinas—not farms—drive regional connectivity and establishment in natural
rocky reefs, Evol. Appl. 11 (2018) 1582–1597.
A. Fort et al.