A draft of the genome of the Gulf Coast tick, Amblyomma maculatum

Abstract

The Gulf Coast tick, Amblyomma maculatum, inhabits the Southeastern states of the USA bordering the Gulf of Mexico, Mexico, and other Central and South American countries. More recently, its U.S. range has extended West to Arizona and Northeast to New York state and Connecticut. It is a vector of Rickettsia parkeri and Hepatozoon americanum. This tick species has become a model to study tick/Rickettsia interactions. To increase our knowledge of the basic biology of A. maculatum we report here a draft genome of this tick and an extensive functional classification of its proteome. The DNA from a single male tick was used as a genomic source, and a 10X genomics protocol determined 28,460 scaffolds having equal or more than 10 Kb, totaling 1.98 Gb. The N50 scaffold size was 19,849 Kb. The BRAKER pipeline was used to find the protein-coding gene boundaries on the assembled A. maculatum genome, discovering 237,921 CDS. After trimming and classifying the transposable elements, bacterial contaminants, and truncated genes, a set of 25,702 were annotated and classified as the core gene products. A BUSCO analysis revealed 83.4% complete BUSCOs. A hyperlinked spreadsheet is provided, allowing browsing of the individual gene products and their matches to several databases.

Introduction

The Gulf Coast tick, Amblyomma maculatum (Koch, 1844) is a vector of Rickettsia parkeri Luckman (Rickettsiales: Rickettsiaceae), which causes a febrile infection in humans [1–3], and also of Hepatozoon americanum, a pathogen of dogs [4–7]. The distribution of A. maculatum extends from the Southeastern states of the USA bordering the Gulf of Mexico, into Mexico and several other Central and South American countries. In the past decades it has extended northwards and to the West in the United States, including the states of Arkansas, Oklahoma, Kansas, and Southwestern Tennessee [8]. The Northernmost range of this tick species includes Delaware, Connecticut, and New York [9–11]. Current work with this tick aims to understand its relationship with its symbionts and pathogens in general, particularly to understand the tick’s immunity pathways [12–15]. The availability of the genome sequence of A. maculatum would foster the pace of these research goals.

To a researcher interested in the biochemistry and physiology of ticks, the main advantage of having the organism’s genome resides in the availability of an annotated set of coding sequences (CDS) and their protein translations, which allows the building of hypotheses on the roles of these gene products and, for example, planning experiments using RNAi and genome editing to test these hypotheses. The availability of genome will also facilitate to build technologies through realizing the full potential of exploiting small RNAs, including microRNA (miRNA) and PIWI-interactacting RNA (piRNA) biology in ticks.

In this work, we used the 10X Genomics platform to sequence the genome of a single male of the Gulf Coast tick, A. maculatum. To obtain the genome’s coding genes coordinates, we used available RNASeq data to train the BRAKER pipeline [16]. The derived CDS translations were compared to several databases and mapped to a hyperlinked spreadsheet that should allow researchers to search for their genes of interest and plan their experiments. The genome of A. maculatum will provide opportunities for comparative evolutionary analysis with other tick species and arthropod vectors, and allow researchers to explore the tick-pathogen interactions and ways tick parasitize vertebrate hosts.

Material and Methods

Sample origin and DNA extraction and quality

Amblyomma maculatum ticks were maintained at the University of Southern Mississippi according to our modified methods [17]. A. maculatum uninfected and Rickettsia parkeri-infected colonies were established in our laboratory in 2013. Questing unfed adult ticks were collected from Mississippi Sandhill Crane National Wildlife Refuge, Gautier, Mississippi (using the drag cloth method on 28th July 2013. A total of 42 females and 62 males collected from the field were blood-fed on sheep and allowed to engorge and drop off. Each fully engorged female adult tick was kept separately in a snap vial for egg-laying. Individual uninfected and Rickettsia parkeri-infected egg clutches from individual gravid females were selected and allowed to hatch into unfed larva. The unfed larval ticks were blood-fed, allowing them to infest golden Syrian hamsters until they dropped off. Fully engorged larvae were allowed to molt into nymphs and then blood-fed on hamsters. Fully engorged nymphs were molted as male or female adult ticks. Closed colonies from the 6th generation of original wild-caught ticks were used in this study, from which five adult male ticks were selected. For each, the whole adult live tick was cut in four quarters and digested in 500 μL of buffer (10–100 mM Tris, 10–100 mM EDTA, 100–200 mM NaCl, 0.5–1% SDS) with 5 μL of proteinase K 10 mg/mL (QIAGEN, Hilden, Germany). The ticks and digestion mix were incubated in a dry bath overnight at 55°C, mixed by vortex ten times during that period. Then, for each sample, 5 μL of RNAse A (ThermoFisher Scientific, MA, USA) was added, vortexed, and incubated at room temperature for 30 minutes. A 400 μL aliquot was transferred to a new microcentrifuge tube and a Phenol-Chloroform-Isoamyl Alcohol (PCI) DNA extraction protocol was followed. The five extracted genomic DNA (gDNA) samples were individually hydrated in 200 μL of TE 1X buffer (Integrated DNA Technologies, Inc., IA, USA) at room temperature overnight. For verification and visualization of the products, 5 μL of each hydrated DNA sample were run in a 0.8% agarose gel.

The sample that showed the best banding pattern in the agarose gel (brightest, high-molecular weight band), a male adult and therefore with XO sexual chromosome make up, was further processed at the Georgia Genomics and Bioinformatics Core, where gDNA concentration was estimated to be 23.3 ng/μl with a Qubit^® Fluorometer using the High Sensitivity protocol, and also was assessed in a Fragment Analyzer^™ (FA) Automated CE System (Advanced Analytical Technologies, CA, USA) using the HS Large Fragment 50Kb method and FA version 1.2.0.11. The FA report revealed a peak size of 24,754 bp, 0.7707 ng/μl, ranging from 4,550 to 100,798 bp with an average size of 27,872 bp.

Linked-Reads Genomic Library Prep and Sequencing

The gDNA sample was used as input for a library prep with the Chromium^™ Genome Library Kit using the Chromium^™ Genome Reagent Kits v2 (CG00022 Rev C), the Chromium^™ Genome Gel Bead Kit (PN-120216), and the Chromium^™ Genome Chip Kit (PN-120216), all from 10x Genomics (10x Genomics, CA, USA). The protocol followed the manufacturer’s instructions. In brief, we diluted the sample according to the standard for the genome protocol, that is 1 ng/μL, and verified that the concentration range was within acceptable limits. Then, the GEM generation sample mix was prepared and combined with both the Denaturing Agent and the gDNA. The mix was loaded into the Chromium^™ Genome Chip where the Genome Gel Beads and Partitioning Oil were also loaded in the corresponding rows. The chip was placed in the Chromium^™ Controller where the Genome Library program was run to partition and barcode each gDNA fragment. Barcodes were added to allow tracking of each resulting read to its original gDNA fragment. Then, the chip was ejected, and the GEMs were aspirated from the recovery well, transferred to a new tube, and isothermally incubated to generate 10x barcoded amplicons. Then the GEMs were cleaned-up with DynaBeads^™ MyOne^™ Silane (ThermoFisher Scientific, MA, USA), rinsed with 80% ethanol twice, and hydrated in Elution Solution. The library construction was finalized following end repair and A-tailing, adaptor ligation, post-ligation clean-up with SPRIselect, sample-index PCR using set SI-GA-A4 (contains barcodes TATGATTC, CCCACAG, ATGCTGAA, and GGATGCCG), and double-sided size selection SPRIselect. Finally, the library product was analyzed in the Fragment Analyzer^™ Automated CE System (Advanced Analytical Technologies, CA, USA) using the NGS Fragment 1–6000bp method and quantified in a Qubit^® Fluorometer using the High Sensitivity protocol. The FA report showed a peak size of 533 bp, with a 3.8 ng/μL concentration; the graph ranged from 1440 bp to 5,087 bp, with an average size of 705 bp. The qubit showed the library concentration to be 52.4 ng/μL.

The library was sequenced two independent times and the resulting reads were pooled. One of the sequencing runs was performed in an Illumina^™ NovaSeq S4 and the second in an Ilumina^™ HiSeqX, both using PE150 kits at Novogene Co., Ltd (Beijing, China).

Genome Assembly

For each run, samples were demultiplexed using the four barcodes from the 10x sample index set, and the output files were merged together according to reads 1 and 2. Then, both sequencing runs were merged independently for read 1 and read 2. Raw data was used as input in Supernova v. 2.1.1 [18] using the run parameter allowing the use of 1,200 million reads with maxreads in an attempt to reach a 56x raw coverage and allowing the use of 28 cores and 980 Gb of memory. Then the option mkoutput was used to create raw, pseudohap, pseudohap2, and megabubble outputs. The summary files regarding the assembly characteristics can be found in supplemental file 1.

Genome annotation

The BRAKER/Augustus pipeline [16] was used to obtain the putative coding sequences (CDS) from the A. maculatum genome. The program was trained to find the CDS using RNAseq data available from the NCBI (accessions SRR959015 - salivary glands and SRR959016 - ovaries). These reads were concatenated and normalized using the Trinity program insilico_read_normalization.pl [19]. The normalized reads were mapped to the unmasked genome using the program Star [20]. The mapped reads were used to train the gene-discovery pipeline BRAKER [16], which discovered a total of 380,129 coding sequences (CDS). The BUSCO program (version 5.0.0) [21] was run with the BRAKER predicted protein sequences against the lineage dataset arachnida_odb10, created on 2020-08-05, from 10 species and 2,934 BUSCOs. The program RepeatMasker version 4.1.2-p1 was used to identify transposable elements and repeat sequences. It was run in sensitive mode with rmblastn version 2.11.0+. The query species was assumed to be Arthropoda. The databases used were FamDB: CONS-Dfam_withRBRM_3.2. Transposable elements (TE) were identified using the Hmmer tool [22] against a subset of the Dfam database [23] containing transposable element models, excluding repeats.The CDS were also compared to the RepBase [24] protein database toidentify and classify TE. To classify genes accordiing to their functional class, the deducted protein sequences were compared using blastp to a subset of the GenBank database containing sequences from the AracHnidae, to the UniprotKB [25] database, to the Expasy Enzyme (EC) [26] database and to the MEROPS [27] database. Rpsblast was used to search the protein sequences against conserved motifs from the PFAM [28], SMART [29] , KOG [30] and CDD [31] databases. To identify genes associated with a salivary function, the CDS were compared by Rpsblast to the TickSialoFam (TSF) database [32]. Matches that had a model coverage of > 66.6 % and an e-value smaller than 1e-4 were considered as related to salivary function. General functional classification was achieved by using a set of ~ 400 key words that were searched in the definition line of the matches above. Each key word was associated with a functional class. A sequence functional class was determined by the first key word found in the definition line of the match if the product of % identity and % coverage were larger > than 0.25. If no keyword was found, the sequence was assigned to a “Unknown” function.All sequences were also searched for existence of a signal peptide indicative of secretion using the SignalP v. 3.0 program [33], for transmembrane domains using the tmhmm program [34] and for O-glycosylation sites indicative of mucins using the program NetOglyc [35]. Glycosyl-phosphate-inositol membrane anchors were identified by the DGPI program [36].

The published genomes of Rhipicephalus microplus and R. sanguineus [37] where used as input to the BRAKER/Augustus pipeline [16] trained with publicly available protein sequences from these organisms.

Transcriptome mapping

Amblyomma maculatum transcriptome reads from the salivary glands and ovaries of adult ticks (NCBI accessions SRR13797277, SRR13797276, SRR13797275, SRR13797274, SRR13797296, SRR13797295, SRR13797294, SRR13797293, SRR13797292, SRR13797290, SRR13797289, SRR13797288, SRR13797287, SRR13797286, SRR13797285, SRR13797284, SRR13797283, SRR13797282, SRR13797305, SRR959015, SRR959016, SRR13797281, SRR13797280, SRR13797279, SRR13797278, SRR13797303, SRR13797302, SRR13797291, SRR13797304, SRR13797273, SRR13797272, SRR13797271, SRR13797270, SRR13797269, SRR13797268, SRR13797301, SRR13797300, SRR13797299, SRR13797298, SRR13797297) were mapped to the predicted CDS using Bowtie2 [38]. Read coverage was measured using samtools coverage program [39].

Phylogenetic Analysis

Protein sequences were aligned with Muscle [40]. Phylogenetic trees were built with the program IQ-tree [41]. The best amino acid evolutionary model was determined by ModelFinder [42]. The tree was bootstrapped using UFBoot2 [43] with the bnni correction. The resulting Newick trees were annotated with Mega X [44],

Data availability

This Whole Genome Shotgun project has been deposited at DDBJ/ENA/GenBank under the accession JAJIZL000000000, BioProject accession PRJNA773936 and BioSample accession SAMN22546173. The reads used to assemble the genome can be found in the Sequence Read Archives (SRA) of the National Center for Biotechnology Information (NCBI) under the accession SRR16911356. The metagenome assembled Rickettsia parkeri genome was deposited in GenBank under the accession CP101541.

Results

We obtained a total of 942,809,836 paired-end reads from both sequencing runs. The genome assembly of A. maculatum resulted in 28,460 scaffolds having equal or more than 10 Kb, totaling 1.98 Gb. The N50 scaffold size was 19,849 Kb. If we add the contigs equal or larger than 1,000 bp, the total assembly size reaches 2.27 Gb. The number of contigs ranging from 1,000 – 9,999 bp is 101,575 and the contig N50 was of 29.12 Kb length. Following search of the assembled genome for bacterial contaminants and duplicated contigs, 14 contigs, summing 1.332 Mbp were found to match known bacterial genomes, including a contig of 1.296 Mbp that matched, with 99% identity, the genome of Rickettsia parkeri, a known endosymbiont of A. maculatum [45]. 4,558 contigs were found to be exactly duplicated, adding to a total of 6.8 Mbp. These contaminants and duplicated contigs were removed from the final assembly. Table 1 lists the current available tick genomes and their characteristics. Although the N50 for the A. maculatum assembly was on the low range when comparing to other tick genome assemblies (Table 1), a BUSCO analysis of the predicted 25,631 CDS from the A. maculatum genome indicated 83.4% complete BUSCOs, 66.8% complete and single-copy BUSCOs, 16.6% complete and duplicated BUSCOs, 1.7% fragmented BUSCOS and 14.9% missing BUSCOs. These results are above the average of those shown on Table 1 which lists other tick genomes so far published.

Table 1:

Published tick genomes characteristics

Scientific name	Annotation	Size (Gbp)	Level	ContigN50 (kb)	Protein-coding	BioProject	Compl ete BUSCO s %	Reference
Amblyomma maculatum		1.9 8	Contig	198	25,70 4	PRJNA773 936	83.4	This work
Dermacentor silvarum	TIGMIC Group	2.4 7	Chromos ome	340	26,69 6	PRJNA633 311	62.4	[37]
Haemaphysalis longicornis	TIGMIC Group	2.5 6	Chromos ome	740	27,14 4	PRJNA633 311	60.6	[37]
Hyalomma asiaticum	TIGMIC Group	1.7 1	Chromos ome	555	29,64 4	PRJNA633 311	65.0	[37]
Ixodes persulcatus	TIGMIC Group	1.9 0	Scaffold	533	28,57 4	PRJNA633 311	76.1	[37]
Ixodes scapularis	Vectorbase	2.0 8	Contig	836	20,48 8	PRJNA345 486	86.8	[55]
Ixodes scapularis	Vectorbase	2.3 0	Contig	1,73 5	19,06 2	PRJNA678 334	81.7	[54]
Rhipicephalus annulatus		2.7 6	Contig	437	N/A	PRJNA593 711	N/A	[Unpublished]
Rhipicephalus microplus	TIGMIC Group	2.0 1	Scaffold	16	24,21 1	PRJNA312 025	55.4	[37]
Rhipicephalus microplus	TIGMIC Group	2.5 3	Chromosome	1,79 1	29,85 7	PRJNA633 311	95.9	[37]
Rhipicephalus sanguineus	TIGMIC Group	2.3 7	Chromosome	542	25,71 8	PRJNA633 311	60.4	[37]

The BRAKER pipeline [16] was used to find the protein coding gene boundaries on the assembled A. maculatum genome, discovering 237,921 CDS. These were compared by blast and rpsblast [46] to several databases, including those at the NCBI (non-redundant and TSA protein sequences) deriving from Arachnida organisms and from Rickettsial bacteria, and the Uniprot database. After removing the sequences matching bacterial phages as well as those that represented fragments with less than 67% coverage to known proteins from the Uniprot and NCBI Arachnida sets, a set of 88,754 sequences were identified as TE (see below), and an additional set of 25,702 were annotated as the core gene products of A. maculatum (Supplemental spreadsheets 1 with all CDS, 2 with TEs and 3, with the core genome set).

The transposable element landscape within the genome of Amblyomma maculatum

The genome-coding DNA contains the information to determine the sequence of a peptide possibly containing 20 different amino acids and one stop codon. There are 64 possible codons, and 3 of them code for stops. So, a stop codon should arise on average once every 21–22 codons, or 63–66 bp. Accordingly, stretches of ORFs longer than 200 nucleotides are expected to indicate a region coding for polypeptides. However, transposable elements challenge the annotation of sequenced genomes, as they “contaminate” these longer ORFs with their coding sequences [47]. Transposable elements are virus -like organisms that parasitize the majority of eukaryotic genomes, frequently loading more than half of the full genomes with their sequences. Thus, to obtain an accurate transcriptome and proteome prediction of a genome, the TE coding sequences have to be filtered out. The transposable element (TE) landscape of the A. maculatum genome was explored by annotating the predicted coding sequences identified as TE based on blastp matches to sequences annotated as TE from the Swissprot database as well as to coding sequences deducted from the Repbase database [24].

Among the 88,734 transcripts identified as TE’s, we were able to classify 86,752; 80.1 % of which were of the Class I type, 19.8% were of the Class II type and 0.17 % were from endogenous retroviruses (ERV) (Table 2 and Supplemental spreadsheet 2) .Within the Class I elements, 57.7% were Long Terminal Repeat (LTR) retrotransposons and 25.4% were NON-LTR retrotransposons. Within the LTRs, Gypsy elements were most abundant, consisting 97% of the total LTR. I elements were the most abundant within the NON-LTR, reaching 40% of the 19,431 transcripts found within Non-LTR elements. Among these NON-LTR elements the BovB LINE element was identified. This element is widespread in vertebrates and it was proposed that horizontal transfer of these elements among vertebrates was vectored by ticks [48, 49]. Among the Class II elements, the P/Tigger family was most abundant, with 11,303 elements, or 65.7% of all class 2 elements found. The Mariner/TC1 family was the second most abundant, abundant, reaching 21.2% of the 17,195 Class II elements identified.

Table 2:

Coding sequences from transposable elements found on the Amblyomma maculatum genome

Class	Type	Family	Transcripts found	Percent total elements	Percent of class
CLASS I	LTR RETROTRANSPOSON	Gypsy	48,274	55.65	96.42
CLASS I	LTR RETROTRANSPOSON	Bel-Pao	742
CLASS I	LTR RETROTRANSPOSON	CER6-I	637
CLASS I	LTR RETROTRANSPOSON	Copia	116
CLASS I	LTR RETROTRANSPOSON	Ngaro	99
CLASS I	LTR RETROTRANSPOSON	CER2-I	55
CLASS I	LTR RETROTRANSPOSON	CIRCE	37
CLASS I	LTR RETROTRANSPOSON	CER3-I	26
CLASS I	LTR RETROTRANSPOSON	CER13-I	23
CLASS I	LTR RETROTRANSPOSON	SKIPPER	11
CLASS I	LTR RETROTRANSPOSON	Tf2	11
CLASS I	LTR RETROTRANSPOSON	CER15-I	10
CLASS I	LTR RETROTRANSPOSON	DIRS	9
CLASS I	LTR RETROTRANSPOSON	CER11-I	6
CLASS I	LTR RETROTRANSPOSON	CER10-I	4
CLASS I	LTR RETROTRANSPOSON	SACI	3
CLASS I	LTR RETROTRANSPOSON	TC1	3
CLASS I	LTR RETROTRANSPOSON	HERV	2
Total			50,068	57.71	100.00
CLASS I	NON-LTR RETROTRANSPOSON	I	6,870	9.89	39.95
CLASS I	NON-LTR RETROTRANSPOSON	RTE	4,128		24.01
CLASS I	NON-LTR RETROTRANSPOSON	REP2	1,804
CLASS I	NON-LTR RETROTRANSPOSON	Loa	1,083
CLASS I	NON-LTR RETROTRANSPOSON	Tad1	814
CLASS I	NON-LTR RETROTRANSPOSON	Outcast	759
CLASS I	NON-LTR RETROTRANSPOSON	Ingi	605
CLASS I	NON-LTR RETROTRANSPOSON	Jockey	591
CLASS I	NON-LTR RETROTRANSPOSON	Nimb	517
CLASS I	NON-LTR RETROTRANSPOSON	LINE	502
CLASS I	NON-LTR RETROTRANSPOSON	R1	432
CLASS I	NON-LTR RETROTRANSPOSON	Tx1	303
CLASS I	NON-LTR RETROTRANSPOSON	L1	232
CLASS I	NON-LTR RETROTRANSPOSON	RTEX	159
CLASS I	NON-LTR RETROTRANSPOSON	Penelope	147
CLASS I	NON-LTR RETROTRANSPOSON	R2	143
CLASS I	NON-LTR RETROTRANSPOSON	ORTE	94
CLASS I	NON-LTR RETROTRANSPOSON	Crack	66
CLASS I	NON-LTR RETROTRANSPOSON	CRE	51
CLASS I	NON-LTR RETROTRANSPOSON	NeSL	42
CLASS I	NON-LTR RETROTRANSPOSON	RandI	28
CLASS I	NON-LTR RETROTRANSPOSON	CR1	18
CLASS I	NON-LTR RETROTRANSPOSON	SR2B	12
CLASS I	NON-LTR RETROTRANSPOSON	LIN10	9
CLASS I	NON-LTR RETROTRANSPOSON	Proto1	6
CLASS I	NON-LTR RETROTRANSPOSON	R4	5
CLASS I	NON-LTR RETROTRANSPOSON	GENIE1	4
CLASS I	NON-LTR RETROTRANSPOSON	Vingi	3
CLASS I	NON-LTR RETROTRANSPOSON	Daphne	2
CLASS I	NON-LTR RETROTRANSPOSON	Ambal	1
CLASS I	NON-LTR RETROTRANSPOSON	Hero	1
Total			19,431	22.40	100.00
CLASS I total			69,499	80.11	100.00
CLASS II	DNA TRANSPOSON	P/Tigger	11,303	13.03	65.73
CLASS II	DNA TRANSPOSON	Mariner/Tc1	3,659	4.22	21.28
CLASS II	DNA TRANSPOSON	Ginger	793	0.91	4.61
CLASS II	DNA TRANSPOSON	Harbinger	520	0.60	3.02
CLASS II	DNA TRANSPOSON	Kolobok	318	0.37	1.85
CLASS II	DNA TRANSPOSoN	UNKNOWN	130
CLASS II	DNA TRANSPOSON	ISL2EU	129
CLASS II	DNA TRANSPOSON	EnSpm	110
CLASS II	DNA TRANSPOSON	hAT	64
CLASS II	DNA TRANSPOSON	piggyBac	33
CLASS II	DNA TRANSPOSON	Zisupton	27
CLASS II	DNA TRANSPOSON	CACTA	24
CLASS II	DNA TRANSPOSON	Helitron	20
CLASS II	DNA TRANSPOSON	Merlin	18
CLASS II	DNA TRANSPOSON	MuDR	16
CLASS II	DNA TRANSPOSON	MiniSatellite	9
CLASS II	DNA TRANSPOSON	mule	8
CLASS II	DNA TRANSPOSON	THAP9	6
CLASS II	DNA TRANSPOSON	LOOPER	4
CLASS II	DNA TRANSPOSON	Academ	2
CLASS II	DNA TRANSPOSON	PIF-Harbinger	2
CLASS II total			17,195	19.82
ERV	ENDOGENOUS RETROVIRUS	ERV3	38
ERV	ENDOGENOUS RETROVIRUS	ERV1	15
ERV	ENDOGENOUS RETROVIRUS	Endogenous Retrovirus	5
Endogenous retrovirus total			58	0.07
Grand total			86,752	100.00

Among the predicted CDS coding for Mariner/TC1 transposases, there were 365 sequences with predicted peptide length between 400 and 600 aa (The average length of full-length Mariner transposases is near 410 aa [50]), without internal stop codons, and containing Pfam domains coding for the DDE superfamily endonuclease and domain HTH_Tnp_Tc5, coding for theTc5 transposase DNA-binding domain. Mariner/TC1 elements have been domesticated in vertebrates, including the centromere-associated protein B (CENPB) and the genes named Tigger transposable element-derived 2 to 7 (TIGD2–7) so far found only in vertebrates [51, 52]. Representatives of these sequences were submitted for phylogenetic analysis, together with the here deduced Mariner/TC1 sequences from A. maculatum and other similar proteins from other tick species found by blast of A. maculatum sequences against the non-redundant database from NCBI. Interestingly, a clade with high (99% bootstrap) support (Clade XI, supplemental figure 1) contained, in subclade XIb, the mammal sequences orthologous to the human TGD6 protein and tick proteins, in subclade XIa from R. microplus, R. sanguineus, I. scapularis and A. maculatum. Transcription of g129797 was found in ovaries, attaining a FPKM (Fragment Per Kilobase of transcript per Million mapped reads) of 8.78 and linear sequence coverage of 98.9%, while g180094 was found expressed in the salivary glands with a FPKM of 7.09 and linear sequence coverage of 97.9%. It is possible that these transposable elements have been also domesticated in ticks.

To compare the TE identification based on putative coding transcripts which are based on protein sequence identity with the TE predictions done from DNA sequence homologies (that are not disturbed by intruding stop codons), we used the program RepeatMasker which identified 1,323,280 TE and other repetitive elements in the A. maculatum genome, representing 25% of the 2.35 GBases of scanned genome (Table 3). Class I elements covered 12.73% of the genome, totaling 838,798 elements, while class II elements (DNA transposons) represented 0.26% of the genome with 82,533 elements, the majority being from the Mariner/TC1 family (36,962 elements). Table 3 has additional information regarding TE and repetitive elements found in the A. maculatum genome.

Table 3:

Transposable elements identified in by RepeatMasker the Amblyomma maculatum genome. Total genome size scanned = 2,350,858,905 bases.

Element family	Number of elements	Base Pairs	Percentage of Genome
Retroelements	667,681	227,801,712	9.24%
SINEs:	189,995	30,535,656	1.24%
Penelope	9,123	923,733	0.04%
LINEs:	306,569	111,282,227	4.51%
CRE/SLACS	7	415	0.00%
L2/CR1/Rex	92,277	13,151,584	0.53%
R1/LOA/Jockey	44,875	7,279,321	0.30%
R2/R4/NeSL	3,298	305,707	0.01%
RTE/Bov-B	127,837	87,675,663	3.56%
L1/CIN4	1,891	100,605	0.00%
LTR elements:	171,117	85,983,829	3.49%
BEL/Pao	10,454	1,214,535	0.05%
Ty1/Copia	5,007	265,943	0.01%
Gypsy/DIRS1	154,051	84,423,689	3.42%
Retroviral	0	0	0.00%
Total Class I elements	838,798	313,785,541
DNA transposons	82,533	6,360,904	0.26%
hobo-Activator	8,996	630,064	0.03%
Tc1-IS630-Pogo	36,962	3,489,969	0.14%
En-Spm	0	0	0.00%
MuDR-IS905	0	0	0.00%
PiggyBac	922	115,270	0.00%
Tourist/Harbinger	945	118,864	0.00%
Other (Mirage, P-element, Transib)	2,924	179,953	0.01%
Rolling-circles	11,361	683,173	0.03%
Unclassified:	5,696	432,101	0.02%
Total interspersed repeats		234,594,717	9.51%
Small RNA:	190,873	30,766,714	1.25%
Satellites:	1,573	135,324	0.01%
Simple repeats:	0	0	0.00%
Low complexity:	0	0	0.00%
Total	1,323,280	617,660,512	25.07

^*most repeats fragmented by insertions or deletions have been counted as one element

The query species was assumed to be arthropoda

RepeatMasker version 4.1.2-p1, sensitive mode

Endogenous viral sequences

The CDS g178917.t1 codes for a nucleocapsid protein from a rhabdovirus [53] which appears to have been incorporated into the genomes of various tick species, as represented by the similar sequences found in the genomes of R. sanguineus (XP_037519053.1), R. microplus (XP_037281023.1), Dermacentor silvarum (XP_037579436.1), I. ricinus (ASY03265.1), I. persulcatus (KAG0426363.1) and I. scapularis (XP_040355436.1).

Annotation of the core genome of Amblyomma maculatum

By comparing the predicted gene products with several databases (see methods), 25,702 gene products were annotated in 29 classes, including 7,976 that were classified as “Unknown” (Table 4 and supplemental spreadsheet 3).

Table 4:

Classification and number of core gene products identified in the Amblyomma maculatum genome

Class	Number of gene products
Putative salivary secreted	2,277
Cytoskeletal proteins	638
Detoxification	233
Oxidant metabolism/Detoxification	157
Extracellular matrix	383
Immunity	187
Amino acid metabolism	358
Carbohydrate metabolism	334
Energy metabolism	515
Intermediary metabolism	139
Lipid metabolism	682
Nucleotide metabolism	206
Nuclear export	33
Nuclear regulation	558
Protein export	2,361
Protein modification	691
Proteasome machinery	641
Protein synthesis machinery	606
Secreted protein	914
Signal transduction	3,293
Storage	39
Transcription factor	50
Transcription machinery	1,392
Transporters and channels	1,040
Unknown conserved	349
Unkown conserved membrane protein	211
Unknown product	7,065
Unkown membrane protein	351
Viral product	1
Total	25,704
Total - Unknown	17,728

Salivary proteins

The search for genes associated with secreted salivary proteins was done by matches of the predicted proteins against the TSF database revealing 2,277 gene products possibly coding for salivary proteins (Table 5 and supplemental spreadsheet 3). Among these, 170 lipocalins, 38 members of the anti-complement/8.9 kDa protein family and 17 evasins were found. Comparisons of the number of members of these protein families found in the proteome annotation of published tick species [37, 54, 55] revealed a much-increased diversity of these protein families in A. maculatum (Table 6A). A possible reason for this discrepancy could be the failure of annotating the salivary-coding transcripts in tick genomes, possibly due to their unique sequences. In support of this this hypothesis, we found larger number of these sequences in the ab initio predicted proteins of the genomes of R. microplus and R. sanguineus. Additionally, we searched the published salivary transcriptomes of R. microplus [56] and R. sanguineus [57], where we found larger number of these protein family members than in the annotated genomes (Table 6B).

Table 5:

Classification and abundance of putative salivary expressed genes from Amblyomma maculatum predicted by the TicskSialoFam database

Family  Subfamily	Number of CDS
12kDa family
Generic	5
Metastriate	11
pk4/12kDa	3
12kDaBasic	1
13–14kDa
13kDa	10
13kDa-Basic	1
23–24kDa
23kDa	15
15kDaBasic	3
18kDa	19
19kDa	1
23–24kDa family
23kDa	1
24kDa	36
28kDa	8
8.9kDa	38
8kDa
8 kDa metastriate	2
AlaRich	63
Amb-25–357	4
Antigen-5	8
BSMAP	1
Basic tail family
Generic	22
TSGP1	5
CalreticulinCalnexin	3
Cell adhesion molecule	7
Coiled-coil domain-containing	4
Complement receptor	1
Complement-binding protein	9
CUTA1	2
CystineKnotToxin	3
Cytotoxin	44
DAP-36	2
Down syndrome family of cell adhesion molecules
Generic	2
Ig_3	28
IG_like	8
Ig-domain	6
EFh_CREC_Calumenin_like	4
Evasin
EvasinA	16
EvasinB	1
Fasciclin-1	1
Ficolin/Ixoderin	11
Fukutin
Glycine Rich protein family
Generic	7
Cement	1
GRP_cement_450	1
GRP_cement_833	1
Collagen-like	1
Chitin binding	48
Dystroglycan	1
GGY	6
GRP21	6
Grp7_allergen	16
Large GYY	8
Large_GRP_II	2
Hematopoietic stem/progenitor cells -like	1
HVA22/Cytokine	1
Hyp_94	1
Hyp2009	2
Insulin_growth_factor	11
Integrin
Alpha subunit	1
Beta subunit	3
Interleukin17-like	14
Ixodegrin	25
Ixodegrin-like	1
Kielin/chordin-like	1
Laminin	9
Lipocalin	170
His binding	98
Generic	47
lipocal-1 1	1
Metastriate IgG-binding lipocalin	17
94	7
Low-density lipoprotein receptor	21
ML_domain	20
Mucin
Generic	78
HRP	6
Peritrophin	8
Sialomucin	1
MYS-2	1
Mys-25–289	1
Mys-25–299	8
Mys-30–170	2
Mys-30–60	6
Mys-30–94	5
Niemann-Pick	6
OneOfEach	38
OSTMP1	1
Papa	2
Peptidoglycan_recognition_protein	5
Phosphatidylethanolamine binding	16
Prich	15
Prich	3
Rapp-25–325	1
Salp15/Ixostatin
Ixostatin	7
Saposin	1
Selenoprotein	1
Serum amyloid A	4
Synaptotagmin 1	1
TGF-beta propeptide	7
TGF-beta propeptide	1
Tick Hirudin	1
Tick-MYS1	1
TMEM9	1
Toll4_associated	1
Toll-like	57
Tolloid-like	2
translocon-associated protein subunit alpha	1
Vitellogenin	4
Vitelogenin-VWF	4
YRP	2
HVA22/Cytokine	1
Hyp669	2
Malectin	1
Toll-like	5
Antimicrobial
5.3kDa
Metastriate_5	2
DAE-2	1
Defensin	6
Is4	6
Lysozyme	4
Microplusin	15
Microplusin_2	20
Enzymes
5’nucleotidase/Apyrase	9
Coesterase	99
Phospholipase A2	6
Cysteinyl_peptidase	26
Dehydrogenase	82
Angiotensin converting enzyme	5
Ectonucleotide pyrophosphatase/phosphodiesterase	3
Endonuclease	14
Epoxide hydrolase	18
IPPase	4
Multiple inositol polyphosphate phosphatase	2
M13_peptidase	377
Metalloprotease	63
Sphingomyelinase	13
Serine carboxypeptidase	22
Zinc carboxypeptidase	2
Serine protease	62
Peroxidase	12
Selenium dependent glutathione peroxidase	4
Superoxide dismutase, cu2+/zn2+ superoxide dismutase sod1	8
Tyrosine sulfotransferase	1
Catalytically inactive chitinase-like lectin	71
Proteinase inhibitors
Longistatin	4
Carboxypeptidase inhibitor	1
Cystatin	14
Thyropin	2
Kunitz	90
Serpin	68
Kazal	1
SPARC/Kazal	24
TIL	20
Total	2,278

Table 6A:

Number of gene products coding for typical salivary proteins in tick genomes.

Species	Lipocalins	8.9 kDA	Evasins	Reference
A. maculatum	170	38	17	This work
D. silvarum	27	3	0	[37]
H. asiaticum	48	7	0	[37]
H. longicornis	1	0	0	[37]
I. persulcatus	12	2	0	[37]
I. scapularis SE6	37	12	2	[54]
I. scapularis Wikel	41	15	2	[55]
R. microplus	20	1	0	[37]
R. sanguineus	29	1	0	[37]

Table 6B:

Number of gene products or coding sequences coding for typical salivary proteins in published tick genomes, ab-initio genomes or transcriptomes.

Species	Lipocalins	8.9 kDA	Evasins	Reference
R. sanguineus genome	29	1	0
R. sanguineus ab initio	48	8	1
R. sanguineus transcriptome	141	34	17	[57]
R. microplus genome	20	1	0
R. microplus ab initio	48	3	0
R. microplus transcriptome	140	22	12	[56]

Digestive enzymes

The sole food of ticks is blood, which is digested intracellularly with the aid of lysosomal cathepsins [58]. Serine proteases may be involved in the late phase of tick engorgement [59]. We have annotated 370 protease genes in the A. maculatum genome, including metalloproteases, calpains, legumains, serine and cysteinyl cathepsins, serine proteases, dipeptidyl peptidases, amino and carboxy peptidases, and protein modification enzymes (Table 7 and supplemental spreadsheet 3, worksheet “Proteases”). Of notice is the expansion of the M13 metalloproteases, with 447 genes, compared to 255 found in R. microplus and 41 on the I. scapularis annotated proteomes (Table 8). Other peptidases are listed on the worksheet “Protein modification” of supplemental spreadsheet 3.

Table 7:

Annotated proteases found in the Amblyomma maculatum genome

Class	Number of genes
M10 metalloproteases	2
M12B metalloproteases	17
M13 metalloproteases	185
Calpains	5
Cathepsin B	4
Cathepsin D (Pepsin)	7
Cathepsin K (Papain)	2
Cathepsin L (Papain)	10
Cathepsin O (Papain)	2
Serine proteases	35
Dipeptidyl peptidase	2
Legumains	22
Amino and Carboxypeptidases	59
Other peptidases	13
Protein modification enzymes	5
Total	370

Table 8:

Number of gene products coding for M13 proteases within tick genomes

Species	Number of genes	Reference
A. maculatum	447	This work
D. silvarum	174	[37]
H. asiaticum	165	[37]
H. longicornis	120	[37]
I. persulcatus	59	[37]
R. sanguineus	129	[37]
R. microplus	115	[37]
I. scapularis Wikel	41	[55]
I. scapularis SE6	41	[54]

Protein modification enzymes

Within the “protein modification enzymes” we highlight the finding of a putative tyrosine sulfotransferase, an enzyme that adds a sulfate group to a tyrosine residue, an important protein modification in tick hormones [60] and some tick salivary peptides [61, 62].

Among other protein modification enzymes, we found several genes coding for members of the prolyl hydroxylase complex, which are important in the production of mature collagen proteins [63]. These can be browsed in the worksheet “Protein modification” from supplemental spreadsheet 3.

Protein glycosyl transferases adds carbohydrate residues to proteins. In ticks, these enzymes have received recent attention due to the epidemics of alpha-gal allergies, which are thought to be triggered by alpha-galactosyl residues decorating the salivary proteins of some tick species, including Amblyomma americanum and Ixodes scapularis, but not in Dermacentor variabilis or A. maculatum [64]. In I. scapularis, typical α-Gal transferases (GALT) were absent in the genome, but enzymes of the α1–4 and β−14 GALT families were able to generate protein α-Galactosylation [65]. These enzymes can be recognized by the “Lactosylceramide 4-alpha-galactosyltransferase” TSFam motif [32]. No enzymes matching this motif or other α-GALT enzymes were found in the A. maculatum genome. The worksheet named “glycosyltransferases” of supplemental spreadsheet 2 presents data on 192 glycosyltransferases.

Cytoskeletal and extracellular proteins

On supplemental spreadsheet 3, worksheet “Cytoskeletal”, annotations can be found for myosins, actins, tubulins, their interacting proteins, and diverse collagen proteins, proteoglycans, and their related enzymes, cuticles and other chitin binding proteins, and gap-junction Innexin proteins.

Immunity-related products

Annotation of genes coding for products associated with immunity revealed proteins coding for: (1) the antimicrobial peptides Defensins, microplusins, lysozymes, Is4 and DAE-2 (Supplemental spreadsheet 3, see results on both Salivary and Immunity worksheets), (2) the RNAi/antiviral response, including Argonaute, Armitage, Aubergine, Tudor, RM62 and Serrate, (3) several members of the alpha-macroglobulin family of complement-like thio-ester esterases (4), several proteins associated with the interferon response (5), three products with similarities to Interleukin-16 and IL-17 (6), Chemokine-like products (7). Several proteins associated with the Tumor Necrosis Factor (TNF) response, including the TNF receptor protein (8), members of the IMD pathway such as Bendless, Caspar, Caudal, Effete, IKK famma - protein kinase, TAB2, TAK1, Uev1a, IAP2 and akirins (9), several products associated with pathogen-recognition motifs (10), members of the SOCS-JAK Stat pathway such as JAK Hopscotch Tyrosine protein kinase, JAK Receptor (Domeless), PIAS Sumo ligase, SOCS box SH2-domain-containing protein and Stat3 (10), members of the TOLL pathway Cactus, Dorsal, MYD88, Pelle, Tube, Spaetzle and several Toll-like receptors.

Epigenetic control and transcription factors

Products affecting epigenetic control, such as histone lysine methyl transferases, histone acetylases and acetyltransferases, histone deacetylases, sirtuins and several members of the chromatin remodeling complex are identified in the supplemental spreadsheet 3 under the row named “Epigenetic control. Transcription factors (47 sequences) are also annotated in Supplemental spreadsheet 2.

Oxidative and detoxification metabolisms

Catalases, peroxidases, superoxide dismutases, Cytochrome −450, Cytoglobins, Selenoproteins, Thioredoxins, Sulfotransferases, Aryl and Glycosyl sulfatases and Glutathione transferases are listed on the worksheet named “Detoxification” on supplemental spreadsheet 3.

Signal transduction

Worksheet “Signal transduction” of supplemental spreadsheet 3 lists several transcripts giving best matches to proteins annotated as 7 transmembrane receptors, G protein-coupled receptors, alpha-1a adrenergic receptor, and receptors for acetylcholine, dopamine, adenosine, serotonin. histamine, adiponectin, rmrfamide, ecdysone, allatostatins, leucokinin atrial natriuretic factor, calcitonin, cholecystokinin, corticotropin, gaba, glycine, octopamine, gonadotropin-releasing hormone, melanocortin neuropeptide y receptor, pyrokinin, relaxin, sifamide and vasopressin. These receptors can be targets of novel acaricides. Several hormonal precursors are also listed, including for the crustacean chh/mih/gih neurohormone family, neurohypophysial hormones and several prohormones.

Additional annotations

Supplemental spreadsheet 3 also details genes coding for proteins implicated on nuclear regulation and nuclear export, transcription and translation machineries, protein export, amino acid, carbohydrate, lipid, and energy metabolisms and proteasome machinery,

Discussion

Using the Chromium Genome Library Kit and the 10X Genomics platform, we obtained a draft genome sequence of the tick Amblyomma maculatum, the first genome for this tick genus, using the DNA extracted from a single male tick. A total of 237,921 putative coding sequences were discovered by the Augustus/BRAKER pipeline trained with public RNAseq data. After excluding transposable elements and truncated sequences, we arrived at a core set of 25,702 coding genes that were functionally annotated and available for browsing in hyperlinked spreadsheets, which we hope will be valuable for further research with this tick species and contributing to the understanding of tick phylogeny.

Analysis of the expanded salivary gland expressed families (such as lipocalin) from the genome of A. maculatum and 3 other tick species show a considerable absence of sequences predicted by transcriptome assembly. It is possible that the “missing “ salivary-coding genes could derive from a higher polymorphism of these genes. Indeed, variable mutation rates are known to occur among different genes [66], associated with those having high transcription [67] or associated with adaptation to variable environments, such as those caused by the host immune response [68], conditions that are found for the highly expressed salivary-coding genes, such as those coding for the lipocalins or metalloproteases. Additionally, increased recombination rates within salivary-coding genes, as observed in some organisms [69, 70], could cause large sequence variation among the individual tick genomes, causing the repertoire of genes at the level of the population being much larger than at the individual level. This hypothesis could be tested by comparing the abundance and similarities of salivary-coding genes from genomes assembled from different individuals.