Apidologie Original article

* INRAE, DIB and Springer-Verlag France SAS, part of Springer Nature, 2020
DOI: 10.1007/s13592-020-00740-x

Evolutionary perspectives on bee mtDNA from mito-

OMICS analyses of a solitary species

Elaine FRANÇOSO1 , Natalia de SOUZA ARAUJO1,2 , Paulo Cseri RICARDO1 ,

Priscila Karla Ferreira SANTOS1 , Alexandre Rizzo ZUNTINI3 , Maria Cristina ARIAS1
1
Instituto de Biociências, Universidade de São Paulo, Rua do Matão 277, sala 320, Sao Paulo, SP 05508-090, Brazil
2
Department of Evolutionary Biology & Ecology, Université Libre de Bruxelles, Avenue F.D. Roosevelt 50, 1050,
Brussels, Belgium
3
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, UK

Received 3 July 2019 – Revised 29 November 2019 – Accepted 13 January 2020

Abstract – The analysis of mitochondrial DNA polymorphism has been applied in multiple organisms to obtain
information about species biology, ecology, population dynamics, and evolution. In this manuscript, the complete
sequencing and characterization of the mitochondrial genome (mtGenome) of Tetrapedia diversipes are reported
and discussed from comparative and evolutionary perspectives among all mtGenomes available for bees so far. The
T. diversipes mtGenome is 15,358 bp long and exhibits the typical set of genes and an A+T-rich region of 588 bp.
The overall base composition is biased towards A/T (84.3%), with 42.6% A, 41.7% T, 9.8% C, and 5.9% G
nucleotides. The obtained data also comprise the mitochondrial DNA methylation and single-nucleotide polymor-
phic sites of one T. diversipes population. Transcription follows the “tRNA punctuation” model, with at least three
primary polycistronic transcripts that are posteriorly processed. Additionally, higher expression rates of the 16S gene
suggest the existence of an exclusive transcription site in this region, and the differential expression of the 12S gene
between larvae and adults reveals different isoforms for this gene. The sequence order of protein-coding and rRNA
genes is conserved across different bee lineages, and differences are restricted to tRNA gene positions. The present
results characterize numerous understudied aspects of bee mtGenomes, and a major evolutionary review of this
molecule within the group is provided. Therefore, this work is a valuable resource for studying mitochondrial
molecular biology and evolution in bees.

mtGenome organization / mitochondrial methylation / mitochondrial transcription / oil-collecting bees /

Tetrapedia diversipes

1. INTRODUCTION molecules have been sequenced. At present, ac-

cording to the NCBI database, the genome of this
Since the completion of the first mitochondrial organelle is the most sequenced eukaryotic ge-
genome (mtGenome) from humans in 1981 nome, and approximately 80 complete or nearly
(Anderson et al. 1981), thousands of these complete mtGenomes of bees have been made
publicly available. In contrast to this trend, other
characteristics of mitochondrial DNA (mtDNA)
Electronic supplementary material The online version of
this article (https://doi.org/10.1007/s13592-020-00740-x)
molecules have rarely been described (Smith
contains supplementary material, which is available to 2015; Tian and Smith 2016). Therefore, data on
authorized users. other mitochondrial characteristics, such as mito-
chondrial transcription (Stewart and Beckenbach
Corresponding author: E. Françoso,
[email protected] 2009; Margam et al. 2011) and methylation
Natalia Souza Araujo contributed equally to this work. (Iacobazzi et al. 2013; Mawlood et al. 2016), are
Responsible editor: Marina Meixner missing for most organisms.
 E. Françoso et al.

Nevertheless, a great number of studies rely on some DNA extracts were enriched for mtGenome
the analysis of mtDNA sequence polymorphisms sequences using the protocol described by
to obtain information on species biology, ecology, Françoso et al. (2015). Library preparation and
population dynamics, and evolution (Avise et al. sample sequencing on both Illumina® and Sanger
1987; Beheregaray 2008; Hickerson et al. 2010). platforms were performed by Macrogen (South
In Hymenoptera, mtGenomes have been especial- Korea). For DNA methylation analyses, whole-
ly useful for evolutionary and phylogenetic anal- bisulfite sequencing was conducted on the
yses because rearrangements in this molecule are Illumina® NextSeq 500 platform at the University
frequent in this group (Mao et al. 2015). of Georgia following the protocol described in
The vast majority of bee species are solitary (Urich et al. 2015).
(Michener 2007), and although they are top polli-
nators of crops and wild plants, solitary bees are 2.2. The complete mitochondrial genome
extremely underrepresented in genetic studies sequence, assembly and annotation
(Neumann and Seidelmann 2006). Tetrapedia
diversipes Klug (1810) is a solitary oil- To guarantee assembly quality and sequence
collecting bee of the tribe Tetrapediini (Michener completeness, different strategies were combined
2007) distributed from Costa Rica to Argentina to obtain the entire mtGenome of T. diversipes :
(Moure 2012) that nests in pre-existing holes in
wood, including trap nests (Aguiar and Martins 1. Mitochondrial enrichment followed by
2002; Alves-dos-Santos 2003; Gazola and NGS—the first partial assembly. Under this
Garófalo 2009). Given its wide distribution and approach, Illumina® sequencing of
nesting aggregation behavior, this bee species is 4,000,000 single reads (approximate size of
easy to manage and sample, which makes 100 bp), representing more than 2,000-fold
T. diversipes a useful Neotropical solitary bee coverage, was performed in DNA extracts
model. enriched for mtDNA molecules from one
In the present manuscript, the sequencing and female bee. Library preparation was per-
characterization of the T. diversipes mtGenome is formed according to the instructions of the
described and discussed from an evolutionary manufacturer as adapted by Monica Carlsen
perspective. The obtained data comprise the com- (personal communication). The quality of the
plete sequencing and annotation (including the Illumina reads was evaluated using the
A+T region) results for the molecule along with FastQC v0.11.2 program (Andrews 2010).
its transcriptomic and DNA methylation profiles This dataset was then independently assem-
and the identification of single-nucleotide poly- bled in two ways. [1A] First, Geneious Pro
morphism sites (SNPs) from one population. In v5.6.3 software (Kearse et al. 2012) and the
addition to the identified genomic characteristics, raw reads were used. Reference assemblies
inferences about the evolutionary dynamics of the (based on the mtGenomes of Apis mellifera
mtDNA in bees were made based on a major [NC_001566] and Bombus ignitus
comparative review with other bee mtGenome [NC_010967]) and denovo assembly were
sequences available at GenBank. performed. Contigs congruent in all three as-
semblies were then extended, realigning all
2. MATERIALS AND METHODS reads to them iteratively. After each iteration,
the extended sequence was manually curated.
2.1. Sampling and DNA extraction [1B] The second assembly strategy relied on
the MITObim v1.8 program (Hahn et al.
All samples were obtained from the same trap 2013). Under this approach, reads were ini-
nest aggregation in São Paulo, Brazil (23° 33′ S), tially cleaned by removing the first two nu-
thus representing one population of T. diversipes . cleotides with the FASTX v0.0.14 toolkit
To increase the sequencing coverage and avoid (Gordon 2009), and low-quality bases (phred
numts (nuclear copies of mitochondrial origin), score below 20) and small reads (less than 20
 Evolutionary perspectives on bee mtDNA from mito-OMICS analyses of a solitary species

nucleotides) were removed with Seqyclean 3. Sanger sequencing—mtGenome final assem-

v1.9.10 (Zhbannikov et al. 2017). To main- bly. PCR extension followed by Sanger se-
tain a maximum coverage of 50-fold, as indi- quencing was performed to solve low-support
cated for the MITObim pipeline, the cleaned regions (i.e., regions with low coverage) and
reads were digitally normalized before as- ambiguities between assemblies and to recov-
sembly following the protocol of Brown er missing regions. Several primers were de-
et al. (2012). Paired reads were interleaved signed using Primer3 (Rozen and Skaletsky
using khmer v2.0 (Crusoe et al. 2015). The 2000) (Table S1; Figure S1) and tested using
final MITObim assembly was conducted with all the possible combinations for the L and H
a combination of three different approaches strands. The PCR conditions followed
using denovo and reference-based exten- Françoso and Arias (2013), with the anneal-
sions, referred to here named ASB0, ASB1, ing temperatures ranging from 38 to 56 °C.
and ASB2. For the de novo assembly The amplification of the A+T-rich region
(ASB0), a fragment of the Cytochrome C failed in direct sequencing attempts; thus, this
Oxidase I (CO1 ) gene from T. diversipes region was cloned into the pGEM plasmid
was used as the seed. In the reference assem- vector (PROMEGA), which was then used
bly (ASB1), the complete mitochondrial ge- to transform competent Escherichia coli
nome from B. ignitus was used as a reference. DH5-α cells prior to sequencing. Genome
For the ASB2 assembly, the contigs resulting annotation was performed using the MITOS
from ASB1 were used as extension seeds for web server (Bernt et al. 2013), which employs
a new denovo assembly. Posteriorly, ASB0, a specialized algorithm that uses similarities
ASB1, and ASB2 were aligned using and structure-based searches to improve mi-
Geneious to obtain a manually curated con- tochondrial genome annotation.
sensus sequence combining all three assem-
blies. This consensus sequence was used as a
reference in a new assembly with MITObim, 2.3. Transcript assembly
which returned the final 1B assembly. Final-
ly, assemblies 1A and 1B were compared to The RNA-Seq data for T. diversipes adults and
generate a consensus sequence. larvae were obtained from Araujo et al. (2018).
2. NGS of nonenriched DNA—the second par- Mitochondrial transcripts from both developmen-
tial assembly. Independent sequencing of the tal stages were assembled using the reference and
total DNA from one T. diversipes male was denovo assembly methods. The reference assem-
performed using the TruSeq DNA PCR-Free bly of the transcripts was generated using the
kit for paired-end library preparation and HISAT2 v2.0.5 (Kim et al. 2015) and StringTie
the Illumina® HiSeq 2500 platform, gener- v1.2.2 (Pertea et al. 2015) programs based on the
ating 293,600,062 paired reads. Based on final mtGenome. In the denovo assembly, mito-
the first partial genome as a reference, these chondrial transcripts were identified through a
new sequencing data were assembled under blastn search (minimum e-value 1e-5) from the
a reference-guided approach with Geneious complete transcriptome of T. diversipes (Araujo
v9.1.6. All parameters used were the de- et al. 2018) against the complete mtGenome. Re-
faults, and the Bowtie2 aligner was set to dundant transcripts, i.e., transcripts overlapping
“fast accurate read mapper and end to end the same mitochondrial region in the denovo as-
alignment”. Due to the presence of paired- sembly, were manually curated so that only the
end reads, the coverage of the aligned reads largest transcript was retained.
from this new dataset was more homoge-
neous than that generated previously, 2.4. SNP identification
allowing the extension and completion of
some missing regions compared with the Transcriptomic data were also used to identify
first partial mtGenome. SNPs in the mitochondrial genome of
 E. Françoso et al.

T. diversipes . Therefore, the SNP information representative. Partial genomes were also used
represents the mitochondrial diversity of 36 indi- when no others were available to represent a ge-
viduals from the studied population (Araujo et al. nus, but only mtGenomes verified at GenBank
2018). The cleaned read alignments (bam files) were employed.
used previously for transcriptome reference as-
sembly were analyzed via the variant-calling pipe- 3. RESULTS
line combining the SAMtools mpileup (v0.1.19)
and BCFtools view (v0.1.19) utilities (Li et al. 3.1. Genomic characterization
2009). Only SNPs with a minimum quality of 30
and 30-fold coverage were selected. SNPs were 3.1.1. mtGenome sequence
manually curated through alignment checking of
the reads with the IGV tool (Robinson et al. 2011). The complete mtGenome of T. diversipes was
15,358 bp in length and exhibited the typical set of
2.5. DNA methylation analyses genes (Table I), including 13 PCGs, 22 transfer
RNAs (tRNA), two ribosomal RNAs (rRNA) and
DNA methylation data were obtained from the the A+T-rich noncoding region (GenBank acces-
whole-body DNA extract of one founder female. sion number: MN732885). The overall base com-
Reads obtained from bisulfite sequencing were position was biased towards A/T (84.3%), with
cleaned using the Trim Galore v0.4.3 (Krueger 42.6% A, 41.7% T, 9.8% C, and 5.9% G. The
2012) wrapper script, with default parameters. PCGs CO1 , CO2 , and CytB presented the lowest
The alignment of the reads to the mitochondrial A/T content among the other PCGs (Table S2),
genome and methylation calling were executed and, in general, a lower A/T content is correlated
following the Bismark v0.17.0 pipeline (Krueger with higher average coverage of short reads
and Andrews 2011). Alignment quality was eval- (Figures S2 and S3).
uated using Qualimap v2.2 (García-Alcalde et al.
2012). 3.1.2. Transcription analyses

2.6. mtGenome comparisons From the adult RNA-Seq analysis, 6,408,822

paired reads were aligned to the mtGenome (mean
Twenty-three mtGenomes, comprising the ge- coverage of 37,323-fold, 167,141 s.d.), and from
nomes of bees from thirteen genera and six fam- the larval data, 9,495,025 paired reads were
ilies, were compared. The wasp Philanthus aligned (mean coverage of 55,313-fold, 261,975
triangulum (Apoidea), and Squilla mantis , which s.d.) [alignments available at NCBI—BioProject
represents the ancestral pancrustacean mitochon- ID: PRJNA590962]. The assembly of the tran-
drial genome organization (Cook et al. 2005), scripts resulted in three contigs for both develop-
were used as external groups. All genomes were mental stages when the reference genome was
aligned using the MUSCLE algorithm (Edgar used and six contigs for larvae and ten for adults
2004) implemented in Geneious 9.1.6, and the via the denovo assembly method. The reduction in
rearrangements were visually compared. The coverage was more pronounced in regions
length and non-ambiguous base composition were encoding tRNAs, affecting transcript continuity,
also obtained with Geneious. Only one complete especially when the de novo assembly method is
mtGenome per genus of the Anthophila lineage used (Figure 1). In the same figure, it is possible to
was used in the analyses, except when differences note the alignment of few reads in a small portion
in arrangement were observed. When more than of the A+T-rich region (between 164 and 362 bp
one mtGenome was available per genus, we se- and 432 and 502 bp).
lected the most complete molecule (i.e., the mol- Coverage analyses of the transcripts also indi-
ecule containing the greatest number of protein- cated a bias in the expression of ribosomal RNAs.
coding genes, PCGs) including the A+T-rich re- In both developmental stages, the second-highest-
gion and the longest sequence assembled as the coverage region of the mtGenome was obtained
 Evolutionary perspectives on bee mtDNA from mito-OMICS analyses of a solitary species

Table I. Tetrapedia diversipes mitochondrial genome annotation. tRNA gene codons are shown in brackets. L: light
strand; H: heavy strand. Clusters of tRNAs are defined according to gene junction positions.

Cluster Gene Position Size (bp) Strand

1 tRNA Ala [tgc] 560–620 61 L

tRNA Met [cat] 621–685 65 L
tRNA Ile [gat] 717–780 64 L
tRNA Gln [ttg] 779–844 66 H
– ND2 890–1801 912 L
2 tRNA Cys [gca] 1839–1907 69 H
tRNA Tyr [gta] 1915–1980 66 H
tRNA Trp [tca] 1993–2057 65 L
– CO1 2058–3584 1527 L
3 tRNA Leu2 [taa] 3616–3682 67 L
– CO2 3683–4348 666 L
4 tRNA Asp [gtc] 4366–4431 66 L
tRNA Lys [ttt] 4435–4502 68 L
– ATP8 4504–4656 153 L
– ATP6 4656–5309 654 L
– CO3 5321–6109 789 L
5 tRNA Gly [tcc] 6117–6185 69 L
– ND3 6210–6530 321 L
6 tRNA Asn [gtt] 6544–6609 66 L
tRNA Arg [tcg] 6629–6690 62 H
tRNA Ser1 [tct] 6707–6762 56 L
tRNA Glu [ttc] 6763–6827 65 H
tRNA Phe [gaa] 6828–6892 65 H
– ND5 7044–8468 1425 H
7 tRNA His [gtg] 8532–8595 64 H
– ND4 8625–9854 1230 H
– ND4L 10,059–10,328 270 H
8 tRNA Thr [tgt] 10,336–10,400 65 L
tRNA Pro [tgg] 10,406–10,469 64 H
– ND6 10,519–10,968 450 L
– CytB 11,036–12,100 1065 L
9 tRNA Ser2 [tga] 12,145–12,211 67 L
– ND1 12,250–13,131 882 H
10 tRNA Leu1 [tag] 13,126–13,193 68 H
– rRNAL (16S) 13,151–14,516 1366 H
11 tRNA Val [tac] 14,505–14,571 67 H
– rRNAS (12S) 14,569–15,329 761 H
– Control region 15,330–559 588 –
(A+T-rich region)
 E. Françoso et al.

Figure 1 . Transcription of the mitochondrial genome of Tetrapedia diversipes . From inner to outer circles: 1,
representation of genomic positions; 2, genome annotation. Gray: protein-coding genes; orange: tRNA; purple:
rRNA, and red: A+T-rich region. 3, GC content graph, where the outer peaks represent guanine or cytosine
nucleotide bases; 4 and 7, in pink: mitochondrial transcripts obtained using the reference assembly method (4 from
larvae and 7 from adult data); 5 and 8, in dark purple: mitochondrial transcripts obtained using the denovo assembly
method (5 from larvae and 8 from adult data); 6 and 9: expression coverage of RNA-Seq sequencing, in which blue
areas represent coverage greater than 50-fold. In circle 6, larval sample coverage is represented, and the red areas are
regions with coverage < 50; in circle 9, adult sample coverage is shown, and orange areas represent coverage < 50.

for the CO1 gene (with approximately 250,000- sudden decrease in coverage in the 5′ portion of
fold maximum coverage). Nevertheless, the cov- 12S in larvae (Figure S4).
erage in the 16S region reached values six-fold
greater than those in adults (≅ 1,450,000-fold 3.1.3. SNPs
maximum coverage) and up to nine-fold greater
in larvae (≅ 2,200,000-fold maximum coverage) Eighty-one SNPs were identified in the studied
(Figure 2). This enormous increase in coverage population (Table S3). Most of the SNPs (64.2%)
was only observed for this gene; therefore, it was were in protein-coding regions, particularly at the
not skewed by a richer GC content and did not third codon position, which led to synonymous
include the 12S ribosomal gene region (200-fold mutations in most cases. Only 18.5% of the SNPs
maximum coverage in adults and 51-fold in lar- resulted in nonsynonymous substitutions. The
vae). Additionally, the comparison of the 12S ND4 gene presented the highest number of
sequences from adults and larvae revealed a nonsynonymous SNPs (five) and the greatest
 Evolutionary perspectives on bee mtDNA from mito-OMICS analyses of a solitary species

2,000,000

adults
larvae

1,500,000
RNASeq Coverage

1,000,000

500,000

0
0.5
% GC

0.25

0
4

3
% mC

0
ND4l
CO2

Atp8
Atp6

ND3

ND6

ND2 CO1 CO3 ND5 ND4 Cytb ND1 16S 12S

0 2,500 5,000 7,500 10,000 12,500 15,000

Genomic position
Figure 2 . Bulk RNA-Seq coverage and cytosine methylation (mC) across the mtGenome. From top to bottom: the
first panel shows the bulk distribution of RNA-Seq reads; the second panel shows the GC percentage for
comparison; the third panel shows the percentage of methylation estimated for all cytosines from a female founder
sample; and in the last panel, PCG positions are indicated along the mtGenome for reference. In the methylation
panel, the red dashed line marks the mean mC level in the whole mtGenome. Below the panels, the order of the main
mitochondrial genes verified for T. diversipes is presented.

number of SNPs (twelve). However, CO3 was the 3.1.4. DNA methylation
PCG with the highest SNP ratio (Figure S5). The
number of SNPs observed in coding regions was Bisulfite sequencing coverage across the
correlated with the region size (Pearson coeffi- mtDNA was uniform, with a mean coverage of
cient = 0.74) (Figure S5). 1,566-fold (279 s.d.). Altogether, 161,789 reads
 E. Françoso et al.

with a mean length of 148 bp were aligned to the 1), in cluster 2 (shuffling of tRNA Trp), and in
mitochondrial genome [alignments available at cluster 4 (shuffling between tRNA Lys and tRNA
Asp
NCBI—BioProjectID: PRJNA590962]. The ) (Figure 3).
mean overall methylation level of the genome
was 0.96%, with 0.6% methylated cytosines in
the CpG context, 0.7% in the CHG context and 4. DISCUSSION
1.0% in the CHH context. Among all the cyto-
sines identified as methylated, 96.6% occurred at 4.1. Genomic characterization
a non-CpG site in the genome. In Figure 2, where
the methylation profile of all cytosines is indicat- Due to the repetitive nature and high A/T con-
ed, it can be noted that that C-methylation gener- tent of the mitochondrial genomes of bees, the
ally showed the opposite pattern to the GC content completion of the T. diversipes mtGenome was
across the genome. Indeed, GC content and C- challenging, even with the use of high-throughput
methylation were negatively correlated in the sequencing. Therefore, multiple approaches for
T. diversipes mtGenome (Pearson coefficient = sequencing and data analyses were necessary. Al-
−0.54). though Illumina sequencing using enriched
mtDNA and total DNA generated a very high
3.1.5. Genomic comparisons across multiple average coverage, this coverage was extremely
bee species heterogeneous and was correlated with the GC
content across the genome (Figures S2 and S3).
The general profiles of the analyzed genomes In general, regions with a lower A/T content, such
are provided in Table S4. Most of the 23 as the CO1 , CO2 , and CytB genes, were well
mtGenomes analyzed presented the typical gene represented, while other areas, such as the genes
content of 13 PCGs, two rRNAs, 22 tRNAs, and encoding tRNAs, ND2 , and 12S and the A+T-
one extra copy of each of tRNA Leu and tRNA Ser . rich region, presented lower coverage and a re-
Apis koschevnikovi nevertheless presented a du- duced mapping quality (Table S2; Figure S2).
plication of tRNA Met rather than tRNA Ser . Among This methodological bias was most likely induced
the 21 bee genomes, Melipona bicolor presented by problems in the alignment of short reads into
the highest A/T ratio (87%), and bees from the repetitive regions, and it greatly impacted the
Andrenidae family exhibited the lowest (78.6% assembly of these areas. Consequently, the effec-
and 79.4%). The longest mtGenome was from tive completion of the A+T-rich region was only
Bombus consobrinus , which consisted of possible by traditional methods of amplification
17,966 bp (Table S4). The order and orientation with cloning and Sanger sequencing.
of the PCGs and rRNAs were conserved in all In addition to the complete sequence of the
species. Differences were restricted to changes in mtGenome, we also report the polymorphic sites
tRNA positions, possibly due to local inversions, of the studied population in this molecule
translocations, and shuffling of adjacent tRNA (Table S3). These data might be useful for popu-
clusters (Table I, Figure 3). lation genetic, phylogenetic and conservation
Most of the events were translocations, espe- studies (revised by Smith 2015) and are especially
cially from tRNA cluster 6 (ND3 -ND5 junction) relevant considering all of the efforts recently
to cluster 1 (A+T-rich region-ND2 junction), and applied to develop conservation strategies for na-
many of the variations represented putative syn- tive bees (Dicks et al. 2016; Potts et al. 2016), due
apomorphies at the family or genus level to their importance as pollinators of native and
(Figure 3). Compared with the ancestral commercial plants (Garibaldi et al. 2014, 2016).
pancrustacean genome, represented here by Squil- In general, the number of SNPs was correlated
la mantis , T. diversipes differed in cluster 6 (in- with gene size (Figure S5), but a higher ratio of
cluding a shuffling of tRNA Met , shuffling and SNPs could be observed in the CO3 and ND4
inversion of tRNA Arg, inversion of tRNA Glu, and genes, suggesting that these genes are good can-
translocation of tRNA Ala from cluster 6 to cluster didates for phylogenetic and taxonomic studies.
 Evolutionary perspectives on bee mtDNA from mito-OMICS analyses of a solitary species

Figure 3 . Putative synapomorphies in bee mitochondrial genomes. The wasp Philanthus triangulum and Squilla
mantis , which represents the ancestral pancrustacean mitochondrial genome organization, were used as external
groups.

Analyses of the mitochondrial transcriptome As can be observed in Figure 2, the coverage of

led to some insightful observations about mtDNA the RNA-Seq data across the mtDNA showed a
expression dynamics in this species. These results considerably higher expression rate of 16S.
suggest that mitochondrial genome transcription Higher expression of this gene has also been
in T. diversipes follows the “tRNA punctuation” observed in other insects such as Drosophila
model (Ojala et al. 1981), with the formation of at (Torres et al. 2009), mosquitoes (Neira-Oviedo
least three primary polycistronic transcripts that et al. 2011), and one bee (Araujo and Arias
are posteriorly processed at tRNA positions, lead- 2019). In mammals, increased expression of ribo-
ing to the potential formation of 13 mitochondrial somal genes has been associated with the exis-
mRNAs. This is supported by the reduction in the tence of an exclusive transcription site for the 16S
coverage of the mitochondrial transcriptome in and the 12S genes (Taanman 1999). In
tRNA regions and the reconstruction of three dis- T. diversipes mtDNA, the existence of this differ-
tinct transcripts after mitochondrial transcriptome ential transcription site is also supported by the
assembly. This processing mechanism of mito- increase in the coverage of 16S; however, the
chondrial mRNA through tRNA punctuation is region of higher sequence coverage does not com-
apparently conserved in many organisms prise the smaller ribosomal gene. Reduced tran-
(Taanman 1999; Stewart and Beckenbach 2009), scription coverage of 12S decoupled from 16S has
including bees (Crozier and Crozier 1993). How- also been reported in Drosophila (Torres et al.
ever, mRNA transcription and processing in each 2009) and M. bicolor (Araujo and Arias 2019)
species is variable, and the number of primary and and might be driven by a methodological bias
processed transcripts may differ (Taanman 1999; caused by the reduced polyA tail of the 12S
Stewart and Beckenbach 2009; Neira-Oviedo mRNA, rather than by differences in transcription
et al. 2011; Tian and Smith 2016). initiation itself (Stewart and Beckenbach 2009).
 E. Françoso et al.

Considering that the library preparation method DNMT1, DNMT3A, and DNMT3B methyltrans-
used in the present study for RNA-Seq sequenc- ferases with the mitochondria (as reviewed in
ing relied on the polyA structure to select Iacobazzi et al. 2013), that the importance of DNA
mRNAs, it is possible that the employed method- methylation in mtDNA dynamics began to receive
ology was not appropriate to efficiently capture more attention. Evidence suggests that, similar to
12S transcripts (Neira-Oviedo et al. 2011). There- methylation of nuclear DNA, changes in mtDNA
fore, although some evidence suggests the exis- methylation are driven by environmental elements
tence of an alternative transcription site for 16S in (Iacobazzi et al. 2013). Differential DNA methyla-
T. diversipes and other insects, it is still unclear if tion in mitochondria has been associated with aging
12S is also transcribed. (Mawlood et al. 2016), diseases (Infantino et al.
Another intriguing result from transcription da- 2011; Iacobazzi et al. 2013) and metabolic processes
ta was the great decrease in coverage observed in that play a role in oxidative stress (as discussed in
the 12S 5′ region in larvae compared with that in van der Wijst and Rots 2015). Specifically, for
adults (Figure S4). This is unlikely to be a result of T. diversipes foundresses, changes in mitochondrial
methodological bias because the two samples gene expression are related to differences between
were prepared using the same methods; thus, this individuals from different reproductive generations
difference suggests that T. diversipes presents at (Araujo et al. 2018). Therefore, it would be interest-
least two distinct isoforms of this mitochondrial ing to determine whether the changes in the pattern
gene, one of which is transcribed in the larval of mtDNA methylation documented here are asso-
stage and the other in the adult stage. Evidence ciated with these changes and other mechanisms of
of multiple isoforms of the 12S has been reported expression control in mitochondria.
previously in the bee M. bicolor , in which the
RNASeq coverage in 12S conflicted with the 4.2. Genomic comparisons across multiple
complete annotation of this gene in the 5′ region bee species
(Araujo and Arias 2019) and for the stink bug
Erthesina fullo on the basis of long read sequenc- Although mitochondrial genomes have been
ing (Gao et al. 2016). described as highly conserved (Wolstenholme
Low levels of gene expression in the control 1992), the order in which genes are arranged is
region of mtDNA have been reported in mam- more variable than initially predicted, especially
mals, where they are associated with the produc- for tRNA genes. For example, cluster 6 (ND3 -
tion of an initiation primer or a long non-coding ND5 junction) is a region of frequent rearrange-
RNA that functions in the control of both the ments in Hymenoptera that are rarely described in
transcription and replication of the mitochondrial other groups of Insecta (Dowton et al. 2003).
genome (Taanman 1999; Gao et al. 2018). Ac- Interestingly, tRNA translocations are not recipro-
cordingly, it is possible that this small portion of cal in this region; i.e., this cluster tends to lose
the A+T-rich region in the RNA-Seq read align- genes instead of gaining them from other clusters
ment (between 164 and 362 pb and 432 and in bee mtGenomes (Dowton et al. 2003).
502 pb), refers to a similar region functionally On the basis of the comparison of Apis ,
relevant for the initiation of transcription and/or Melipona , and Bombus , Dowton et al. (2009)
replication of the mtGenome in T. diversipes (and suggested that the translocation of tRNA Ala to
possibly other bees, Araujo and Arias 2019). cluster 1 is an Apidae synapomorphy. In the pres-
However, as discussed previously, the A+T-rich ent work, this hypothesis was corroborated by the
region presents low complexity and is very repet- analyses of the genus Tetrapedia and Nomada
itive; consequently, the alignment of short reads to and extended to the family Megachilidae. It was
this region cannot be trusted without further also suggested that the shuffling between tRNA Asp
evidences. and tRNA Lys in cluster 4 (CO2 -ATP8 junction)
The first studies on DNA methylation in mito- would erroneously phylogenetically group
chondria were performed in the 1970s (Nass 1973), Bombus and Apis because this shuffling was not
but it was not until 2011, after the association of the present in Melipona (Silvestre et al. 2008).
 Evolutionary perspectives on bee mtDNA from mito-OMICS analyses of a solitary species

However, in Figure 3, it is possible to see that this

event is actually distributed among all bees. Thus, ACKNOWLEDGMENTS
the translocation of tRNA Lys is instead a synapo-
morphy in Melipona . The authors would like to thank Dr. Lucia Lohmann
Although frequent tRNA rearrangements are and Dr. Monica Carlsen for the discipline with which
common, some genomic positions are highly con- the first mtGenome was generated; Dr. Isabel Alves dos
served. For example, tRNA Phe in cluster 6 (ND3 - Santos and Dr. Guaraci Duran Cordeiro for their support
ND5 junction) and tRNAPro in cluster 8 (ND4L - in specimen sampling; Susy Coelho for laboratory
ND6 junction) exhibit the same position and orien- maintenance; Dr. Bob Schmitz (University of Georgia)
for the whole-bisulfite sequencing and library prepara-
tation in all bees. It has been suggested that the
tion of the methylation data; and Dr. Denis Jacob
tRNA Phe position might be under selective con- Machado for useful advice about assembly and annota-
straint because this gene is located at a site where tion methods for mitochondrial genomes.
transcription polarity changes; therefore, it could be
a putative signal for endonucleolytic cleavage dur-
AUTHOR CONTRIBUTIONS
ing the maturation of the primary polycistronic tran-
script (Ojala et al. 1981; Dowton et al. 2003). The
EF: work idealization, genome assembly; NSA:
same reasoning can be used to explain the position
genome assembly, transcription, methylation, and
of tRNA Pro in cluster 8 (Table I; Figure 1), since
SNP analyses; ARZ: assistance in data analysis,
both genes are phylogenetically conserved and lo-
comparative analyses /rearrangements; PCR:
cated at sites of polarity changes. Additionally, these
PCR amplification and cloning, assistance in data
tRNAs are positioned adjacent to cleavage sites of
analyses; PKFS: genome assembly, comparative
the polycistronic transcripts assembled for
analyses /rearrangements; and MCA: project ad-
T. diversipes (Figure 1), reinforcing the hypothesis
vising. All authors contributed to the preparation
of their functional role as maturation signals.
of the manuscript.
5. CONCLUSIONS
FUNDING INFORMATION
Here, we provide a valuable dataset for the mito-
chondrial genome of T. diversipes , including its Financial support was provided by the Fundação
complete sequence and annotation, transcription pat- de Amparo à Pesquisa do Estado de São Paulo
terns in two life stages, methylated sites in females (Proc. 10/50597-5, 12/18531-0 and 13/12530-4;
during nest foundation and population genomic di- scholarship to EF 2009/07124-1, 2010/20548-2,
versity determined through SNP identification. Ad- 2013/03961-1 and 201425023-6),
ditionally, we combined the present sequencing data CAPES—Coordenação de Aperfeiçoamento de
with database sequences to understand the molecu- Pessoal de Nível Superior, Brazil (Finance Code
lar mechanisms underlying mitochondrial genomic 001 and scholarship to PKFS), CNPq—Conselho
evolution in bees through a comparative review of Nacional de Desenvolvimento Científico e
all available bee mtGenomes. The results highlight Tecnológico (research sponsorship to MCA, Pro-
the importance of tRNA rearrangement events in the cess number 306932/2016-4), and the Research
evolution of this molecule in bees and the existence Center on Biodiversity and Computing
of DNA methylation in the T. diversipes mitochon- (BioComp) of the Universidade São Paulo
drial genome in a predominant non-CG context, and (USP), supported by the USP Provost’s Office
they show some intricate mechanisms involved in for Research.
gene expression regulation. The reported analyses
and datasets may be used to address important evo- COMPLIANCE WITH ETHICAL
lutionary questions not only concerning STANDARDS
T. diversipes but also for other bee species, especial-
ly regarding the underrepresented group of solitary Conflict of interest The authors declare that they have no
bees. conflict of interest.
 E. Françoso et al.

Perspectives d’évolution de l’ADNmt des abeilles à Bernt M, Donath A, Jühling F, et al (2013) MITOS: Im-
partir d’analyses “mito-OMIQUES” d’une espèce proved de novo metazoan mitochondrial genome an-
solitaire. notation. Mol Phylogenet Evol 69 :313–319.
https://doi.org/10.1016/j.ympev.2012.08.023
organisation du génome mt / méthylation Brown CT, Howe A, Zhang Q, et al (2012) A Reference-
mitochondriale / transcription mitochondriale / abeilles free algorithm for computational normalization of
collectrices d’huile / Tetrapedia diversipes. shotgun sequencing data. arXiv 1203.4802 [q-bio.GN]
Cook CE, Yue Q, Akam M (2005) Mitochondrial genomes
suggest that hexapods and crustaceans are mutually
Evolutionäre Perspektiven zur mtDNA von Bienen aus paraphyletic. Proc R Soc B Biol Sci 272 :1295–1304.
der Analyse von “mito-OMICS” Daten einer solitären https://doi.org/10.1098/rspb.2004.3042
Art.
Crozier RH, Crozier YC (1993) The mitochondrial genome
of the honeybee Apis mellifera : complete sequence
Organisation des mt Genoms / mitochondriale and genome organization. Genetics 133 :97–117
Methylierung / mitochondriale Transkription / Crusoe MR, Alameldin HF, Awad S, et al (2015) The
Ölsammelnde Bienen / Tetrapedia diversipes. khmer software package: enabling efficient nucleotide
sequence analysis. F1000Research. 4 :900. 1–10.
https://doi.org/10.12688/f1000research.6924.1
Dicks L V., Viana B, Bommarco R, et al (2016) Ten policies
for pollinators. Science (80) 354 :975–976. https://doi.
org/10.1126/science.aai9226
REFERENCES Dowton M, Cameron SL, Dowavic JI, et al (2009) Charac-
terization of 67 mitochondrial tRNA gene rearrange-
ments in the hymenoptera suggests that mitochondrial
Aguiar AJC, Martins CF (2002) Abelhas e vespas solitárias tRNA gene position is selectively neutral. Mol Biol
em ninhos-armadilha na Reserva Biológica Guaribas Evol 26 :1607–1617. https://doi.org/10.1093
(Mamanguape, Paraíba, Brasil). Rev Bras Zool /molbev/msp072
19 :101–116. https://doi.org/10.1590/S0101-
81752002000500005 Dowton M, Castro LR, Campbell SL, et al (2003) Frequent
mitochondrial gene rearrangements at the hymenopter-
Alves-dos-Santos I (2003) Trap-nesting bees and wasps on an nad3-nad5 junction. J Mol Evol 56 :517–526.
the university campus in São Paulo, Southeastern Bra- https://doi.org/10.1007/s00239-002-2420-3
zil (Hymenoptera: Aculeata). J Kansas Entomol Soc
76 :328–334 Edgar RC (2004) MUSCLE: multiple sequence alignment
with high accuracy and high throughput. Nucleic
Anderson S, Bankier AT, Barrell BG, et al (1981) Sequence Acids Res 32 :1792–1797. https://doi.org/10.1093
and organization of the human mitochondrial genome. /nar/gkh340
Nature 290 :457–465. https://doi.org/10.1038/290457
a0 Françoso E, Arias MC (2013) Cytochrome c oxidase I
primers for corbiculate bees: DNA barcode and mini-
Andrews S (2010) FastQC: a quality control tool for high barcode. Mol Ecol Resour 13 :844–850. https://doi.
throughput sequence data. http://www.bioinformatics. org/10.1111/1755-0998.12135
babraham.ac.uk/projects/fastqc/
Françoso E, Gomes F, Arias MC (2015) A protocol for
Araujo NS, Arias MC (2019) Mitochondrial genome char- isolating insect mitochondrial genomes: a case study of
acterization of Melipona bicolor : Insights from the NUMT in Melipona flavolineata (Hymenoptera:
control region and gene expression data. Gene Apidae). Mitochondrial DNA 27 :1–4. https://doi.
705 :55–59. https://doi.org/10.1016/j. org/10.3109/19401736.2015.1028049
gene.2019.04.042
Gao S, Ren Y, Sun Y, et al (2016) PacBio full-length
Araujo NS, Santos PKF, Arias MC (2018) RNA-Seq re- transcriptome profiling of insect mitochondrial gene
veals that mitochondrial genes and long non-coding expression. RNA Biol 13 :820–825. https://doi.
RNAs may play important roles in the bivoltine gen- org/10.1080/15476286.2016.1197481
erations of the non-social Neotropical bee Tetrapedia
diversipes . Apidologie 49 :3–12. https://doi. Gao S, Tian X, Chang H, et al (2018) Two novel lncRNAs
org/10.1007/s13592-017-0542-2 discovered in human mitochondrial DNA using
PacBio full-length transcriptome data. Mitochondrion
Avise JC, Arnold J, Ball RM, et al (1987) Intraspecific 38 :41–47. https://doi.org/10.1016/j.mito.2017.08.002
phylogeography: the mitochondrial DNA bridge be-
tween population genetics and systematics. Annu Rev García-Alcalde F, Okonechnikov K, Carbonell J, et al
Ecol Syst 18 :489–522. https://doi.org/10.1146 (2012) Qualimap: evaluating next-generation sequenc-
/annurev.es.18.110187.002421 ing alignment data. Bioinformatics 28:2678–2679.
https://doi.org/10.1093/bioinformatics/bts503
Beheregaray LB (2008) Twenty years of phylogeography:
the state of the field and the challenges for the Southern Garibaldi LA, Carvalheiro LG, Leonhardt SD, et al (2014)
Hemisphere. Mol Ecol 17 :3754–3774. https://doi. From research to action: enhancing crop yield through
org/10.1111/j.1365-294X.2008.03857.x
 Evolutionary perspectives on bee mtDNA from mito-OMICS analyses of a solitary species

wild pollinators. Front Ecol Environ 12:439–447. Mawlood SK, Dennany L, Watson N, et al (2016) Quanti-
https://doi.org/10.1890/130330 fication of global mitochondrial DNA methylation
Garibaldi LA, Carvalheiro LG, Vaissiere BE, et al (2016) levels and inverse correlation with age at two CpG
Mutually beneficial pollinator diversity and crop yield sites. Aging (Albany NY) 8 :636–641. https://doi.
outcomes in small and large farms. Science 351:388– org/10.18632/aging.100892
391. https://doi.org/10.1126/science.aac7287 Michener CD (2007) The bees of the world, 2nd edn. The
Gazola AL, Garófalo CA (2009) Trap-nesting bees (Hyme- Johns Hopkins University Press, Baltimore
noptera: Apoidea) in forest fragments of the State of Moure JS (2012) Tetrapediini Michener & Moure, 1957.
São Paulo, Brazil. Genet Mol Res 8:607–622. In: Moure JS, Urban D, Melo GAR (Orgs). Catalogue
https://doi.org/10.4238/vol8-2kerr016 of Bees (Hymenoptera, Apoidea) in the Neotropical
G o r d o n A ( 2 0 0 9 ) FA S T X - To o l k i t . h t t p : / / Region - online version. http://www.moure.cria.org.
hannonlab.cshl.edu/fastx_toolkit/ br/catalogue. Accessed 8 Nov 2013

Hahn C, Bachmann L, Chevreux B (2013) Reconstructing Nass MMK (1973) Differential methylation of mitochon-
mitochondrial genomes directly from genomic next- drial and nuclear DNA in cultured mouse, hamster and
generation sequencing reads - A baiting and iterative virus-transformed hamster cells In vivo and in vitro
mapping approach. Nucleic Acids Res 41 :e129. methylation. J Mol Biol 80 :155–175. https://doi.
https://doi.org/10.1093/nar/gkt371 org/10.1016/0022-2836(73)90239-8
Hickerson MJ, Carstens BC, Cavender-Bares J, et al (2010) Neira-Oviedo M, Tsyganov-Bodounov A, Lycett GJ, et al
Phylogeography’s past, present, and future: 10 years (2011) The RNA-Seq approach to studying the expres-
after Avise, 2000. Mol Phylogenet Evol 54 :291–301. sion of mosquito mitochondrial genes. Insect Mol Biol
https://doi.org/10.1016/j.ympev.2009.09.016 20 :141–152. https://doi.org/10.1111/j.1365-
2583.2010.01053.x
Iacobazzi V, Castegna A, Infantino V, Andria G (2013) Mito-
chondrial DNA methylation as a next-generation bio- Neumann K, Seidelmann K (2006) Original article
marker and diagnostic tool. Mol Genet Metab 110 :25– Microsatellites for the inference of population struc-
34. https://doi.org/10.1016/j.ymgme.2013.07.012 tures in the Red Mason bee Osmia rufa (Hymenoptera
, Megachilidae). Apidologie 37 :75–83. https://doi.
Infantino V, Castegna A, Iacobazzi F, et al (2011) Impair- org/10.1051/apido
ment of methyl cycle affects mitochondrial methyl
availability and glutathione level in Down’s syndrome. Ojala D, Montoya J, Attardi G (1981) tRNA punctuation
Mol Genet Metab 102 :378–382. https://doi. model of RNA processing in human mitochondria.
org/10.1016/j.ymgme.2010.11.166 Nature 290 :470–474. https://doi.org/10.1038/290470
a0
Kearse M, Moir R, Wilson A, et al (2012) Geneious Basic:
An integrated and extendable desktop software plat- Pertea M, Pertea GM, Antonescu CM, et al (2015)
form for the organization and analysis of sequence StringTie enables improved reconstruction of a tran-
data. Bioinformatics 28 :1647–1649. https://doi. scriptome from RNA-Seq reads. Nat Biotechnol
org/10.1093/bioinformatics/bts199 33 :290–295. https://doi.org/10.1038/nbt.3122

Kim D, Langmead B, Salzberg SL (2015) HISAT: a fast spliced Potts SG, Imperatriz-Fonseca V, Ngo HT, et al (2016)
Safeguarding pollinators and their values to human
aligner with low memory requirements. Nat Methods well-being. Nature 540 :220–229. https://doi.
12 :357–360. https://doi.org/10.1038/nmeth.3317 org/10.1038/nature20588
Krueger F (2012) Trim Galore. http://www.bioinformatics. Robinson JT, Thorvaldsdóttir H, Winckler W, et al (2011)
babraham.ac.uk/projects/trim_galore/ Integrative genomics viewer. Nat Biotechnol 29 :24–
Krueger F, Andrews SR (2011) Bismark: a flexible aligner 26. https://doi.org/10.1038/nbt.1754
and methylation caller for Bisulfite-Seq applications. Rozen S, Skaletsky H (2000) Primer3 on the WWW for
Bioinformatics 27:1571–1572. https://doi.org/10.1093 General Users and for Biologist Programmers. In: Bio-
/bioinformatics/btr167 informatics Methods and Protocols. Humana Press,
Li H, Handsaker B, Wysoker A, et al (2009) The Sequence New Jersey, pp 365–386
Alignment/Map format and SAMtools. Bioinformatics Silvestre D, Downton M, Arias MC (2008) The mitochon-
25 :2078–2079. https://doi.org/10.1093 drial genome of the stingless bee Melipona bicolor
/bioinformatics/btp352 (Hymenoptera, Apidae, Meliponini): Sequence, gene
Mao M, Gibson T, Dowton M (2015) Higher-level phylog- organization and a unique tRNA translocation event
eny of the Hymenoptera inferred from mitochondrial conserved across the tribe Meliponini. Genet Mol Biol
genomes. Mol Phylogenet Evol 84 :34–43. https://doi. 31 :451–460. https://doi.org/10.1590/S1415-
org/10.1016/j.ympev.2014.12.009 47572008000300010
Margam VM, Coates BS, Hellmich RL, et al (2011) Mito- Smith DR (2015) The past, present and future of mitochon-
chondrial genome sequence and expression profiling drial genomics: have we sequenced enough mtDNAs?
for the legume pod borer Maruca vitrata (Lepidoptera: Brief Funct Genomics 15 :47–54. https://doi.
Crambidae). PLoS One 6 :e16444. https://doi. org/10.1093/bfgp/elv027
org/10.1371/journal.pone.0016444 Stewart JB, Beckenbach AT (2009) Characterization of
mature mitochondrial transcripts in Drosophila , and
 E. Françoso et al.

the implications for the tRNA punctuation model in bisulfite sequencing. Nat Protoc 10:475–483.
arthropods. Gene 445 :49–57. https://doi.org/10.1016 https://doi.org/10.1038/nprot.2014.114
/j.gene.2009.06.006 van der Wijst MGP, Rots MG (2015) Mitochondrial epige-
Taanman J-W (1999) The mitochondrial genome: structure, netics: an overlooked layer of regulation? Trends Gen-
transcription, translation and replication. Biochim e t 3 1 : 3 5 3 – 3 5 6 . h t t p s : / / d o i . o rg / 1 0 . 1 0 1 6 / j .
Biophys Acta Bioenerg 1410 :103–123. https://doi. tig.2015.03.009
org/10.1016/S0005-2728(98)00161-3 Wolstenholme DR (1992) Animal mitochondrial DNA:
Tian Y, Smith DR (2016) Recovering complete mitochon- structure and evolution. Int Rev Cytol 141 :173–216.
drial genome sequences from RNA-Seq: A case study https://doi.org/10.1016/S0074-7696(08)62066-5
of Polytomella non-photosynthetic green algae. Mol Zhbannikov IY, Hunter SS, Foster JA, Settles ML (2017)
Phylogenet Evol 98 :57–62. https://doi.org/10.1016/j. SeqyClean: a pipeline for high-throughput sequence
ympev.2016.01.017 data preprocessing. In: Proceedings of the 8th ACM
Torres TT, Dolezal M, Schlötterer C, Ottenwälder B (2009) International Conference on Bioinformatics, Compu-
Expression profiling of Drosophila mitochondrial tational Biology,and Health Informatics - ACM-BCB
genes via deep mRNA sequencing. Nucleic Acids ‘17. ACM Press, New York, pp 407–416
Res 37 :7509–7518. https://doi.org/10.1093
/nar/gkp856
Publisher’s note Springer Nature remains neutral
Urich MA, Nery JR, Lister R, et al (2015) MethylC-seq with regard to jurisdictional claims in published maps
library preparation for base-resolution whole-genome
and institutional affiliations.