Owl Monkey Alu Insertion Polymorphisms and Aotus Phylogenetics

Owl monkeys (genus Aotus), or “night monkeys” are platyrrhine primates in the Aotidae family. Early taxonomy only recognized one species, Aotus trivirgatus, until 1983, when Hershkovitz proposed nine unique species designations, classified into red-necked and gray-necked species groups based predominately on pelage coloration. Recent studies questioned this conventional separation of the genus and proposed designations based on the geographical location of wild populations. Alu retrotransposons are a class of mobile element insertion (MEI) widely used to study primate phylogenetics. A scaffold-level genome assembly for one Aotus species, Aotus nancymaae [Anan_2.0], facilitated large-scale ascertainment of nearly 2000 young lineage-specific Alu insertions. This study provides candidate oligonucleotides for locus-specific PCR assays for over 1350 of these elements. For 314 Alu elements across four taxa with multiple specimens, PCR analyses identified 159 insertion polymorphisms, including 21 grouping A. nancymaae and Aotus azarae (red-necked species) as sister taxa, with Aotus vociferans and A. trivirgatus (gray-necked) being more basal. DNA sequencing identified five novel Alu elements from three different taxa. The Alu datasets reported in this study will assist in species identification and provide a valuable resource for Aotus phylogenetics, population genetics and conservation strategies when applied to wild populations.


Introduction
Owl monkeys (Aotus spp.) are small, neotropical platyrrhine primates (monkeys found in the Americas) within the Aotidae family; they are often referred to as night monkeys and are widespread across South America and Panama [1,2]. The nocturnal habits of owl monkeys have made them difficult to study in the field compared to other platyrrhine primates, and the morphological similarities of Aotus species have led to frequent misidentifications [2,3]. Early taxonomic studies of owl monkeys only recognized one species, designated as A. trivirgatus, until 1983, when Hershkovitz designated nine species based predominately on pelage coloration and karyotyping and proposed the separation of "red-necked" and "gray-necked" species groups [4]. Hershkovitz described the "gray-necked" group as having the "side of neck grayish agouti to brownish agouti" in pelage and included species "Aotus lemurinus (with subspecies A. 1. lemurinus (I. Geoffroy) and A. 1. griseimembra (Elliot)), A. vociferans (Spix), A. trivirgatus (Humboldt) and A. brumbacki, a species previously misidentified with A. azarae" and named in honor of Dr. Brumback [4]. By contrast, Hershkovitz described the "red-necked" group as having from other countries [11], displacing them from their native habitats. These issues highlight the need for more accurate molecular methods to differentiate Aotus taxa reliably.
Primate-specific Alu retrotransposons are a type of SINE (short interspersed element), are approximately 300 bp long and consist of two diverged dimers separated by an A-rich region [12] and ending in a 3 A-rich tail [13]. Non-autonomous Alu elements mobilize via a "copy and paste" mechanism through an RNA intermediate, utilizing the enzymatic machinery of autonomous LINE (L1) elements [14], a process termed "target-primed reverse transcription" (TPRT) [15]. The TPRT integration process involves staggard nicks in the host DNA that produce 5 and 3 flanking target site duplications (TSDs) upon insertion. The TPRT mechanism is considered unidirectional such that the absence of insertion is, by default, the ancestral state; conversely, shared insertions with matching TSDs are accepted as being inherited from a common ancestor. These unique attributes have made Alu elements well-established diagnostic molecular markers for the study of primate population genetic and phylogenetic relationships [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31].
Random nucleotide mutations in Alu elements that are subsequently mobilized by TPRT allow for the accumulation of diagnostic substitutions representative of discrete Alu repeat subfamilies [32][33][34][35], such that each primate lineage contains a unique group of Alu subfamilies [36][37][38][39][40]. The oldest known Alu subfamily, AluJ, is found in all primates, whereas AluS was more dominant following the separation of Strepsirrhini and Tarsiiformes from Platyrrhini and Catarrhini [39,40] and the younger AluY subfamily is found only in catarrhines [36]. Therefore, only AluJ and AluS subfamilies and their modern derivatives are present in platyrrhines.
The first platyrrhine-specific Alu subfamily discovered was given the new name AluT because it was created by a fusion event between AluSc and AluSp elements [21]. This platyrrhine-specific AluTa-lineage includes Ta7, Ta10 and Ta15, with AluTa15 thought to be limited to the Cebidae family [21]. Other platyrrhine Alu subfamilies have been characterized in marmoset [41], squirrel monkey [42], capuchin [43,44] and owl monkeys [43,45]. Some studies utilized Alu elements in platyrrhine phylogeny [16,28,29,46,47]; however, an analysis of recently integrated, young Alu insertions in owl monkeys has not been conducted. A scaffold-level genome assembly is currently available for one species, A. nancymaae [Anan_2.0]. The purpose of this study was to computationally ascertain a dataset of the youngest Alu insertions specific to the owl monkey lineage from the [Anan_2.0] genome and to perform locus-specific PCR on a DNA panel of 23 captive Aotus specimens to identify Alu insertion polymorphisms. The Alu datasets reported in this study provide a valuable resource to assist in species identification and to facilitate future studies of Aotus phylogeny, population genetics and conservation strategies.

Lineage-Specific Alu Elements
Ascertainment of lineage-specific Alu insertions from the A. nancymaae genome [Anan_2.0] was performed as described previously [43,45]. Briefly, the scaffold level genome assembly (GCA_000952055.2 Anan_2.0) was obtained from the National Center for Biotechnology Information and analyzed for full-length Alu elements with RepeatMasker [48] (RepeatMasker-Open-4.0). Full-length elements were defined as being 267 bp or longer and starting no more than 4 bp from the 5 start of the Alu consensus sequence (RepeatMasker-Open-4.0). Successive implementations of BLAT [49] were performed with the following genomes in the following order: marmoset (Callithrix jacchus; calJac3), squirrel monkey (Saimiri boliviensis; SaiBol1) and capuchin (Cebus imitator; Cebus imitator_1.0). After each BLAT was completed, the output was searched for shared or unique elements by looking for specific gap sizes between the input Alu sequences and the target genome. The unique elements were computationally extracted (described at link https://github.com/t-beck; accessed on 24 October 2022). A combination of cross_match (http://www.phrap.org/phredphrapconsed.html; accessed on 6 November 2022) and COSEG (www.repeatmasker.org/COSEGDownload.html; accessed on 6 November 2022) analyses were completed to determine if any of the lineage-specific Alu elements extracted from the Aotus genome represented new subfamilies [45]. The putative lineage-specific Alu elements were aligned with the recently discovered Saimiri-specific Alu subfamilies, and exact matches were eliminated.
A local installation of RepeatMasker was used to determine the percent divergence of each lineage-specific Alu element compared to the consensus sequence. Young elements, defined as being less than 2% diverged, were retained. FASTA files of these young, lineagespecific elements were extracted by a custom Python script (described at link https://github. com/t-beck; accessed on 24 October 2022) and then sorted by subfamily. Representative elements from each Alu subfamily were proportionally selected for experimental validation based on the number of subfamily members. If an Alu subfamily contained less than five elements, then all elements were selected for experimental validation. BLAT was used to obtain orthologous sequences, including 600 bp of flanking sequences from the marmoset, squirrel monkey and capuchin genomes. BioEdit was used to create multiple sequence alignments for each locus [50].

Oligonucleotide Primer Design
For polymerase chain reactions (PCRs), some of the oligonucleotide primers were designed manually using Primer3 [51] with the following specifications: Tm range = 57-62 • C, Max TmDifference = 2, Max Poly X = 3, Min GC Content = 40 [52]. Other primers were designed by using an in-house primer design pipeline consisting of a series of custom Python scripts in conjunction with Primer3 (available on link https://github.com/t-beck; accessed on 24 October 2022). Following each method, all primers were analyzed for primer specificity and predicted PCR amplicon length using NCBI Primer Blast [53]. The oligonucleotide primers were obtained from Sigma Aldrich (Woodlands, TX, USA).

DNA Samples
DNA samples and their origins are described in Supplementary File S1. Both tissue and DNA samples were obtained from a variety of universities, museums and biomedical research centers. DNA was prepared from tissue samples by proteinase K digestion followed by phenol:chloroform extraction and ethanol precipitation. Extracted DNA was stored in 10 mM Tris/0.1 mM EDTA (TLE) and quantified by an Eppendorf Biophotometer.

PCR Amplification
PCR amplifications were performed in 25 µL reactions that contained 25-50 ng of DNA, 200 nM of each primer, 1.5 mM MgCl 2 , 10× PCR buffer, 0.2 mM deoxyribonucleotide triphosphates and 1 unit of Taq DNA polymerase. PCR cycling parameters were: 95 • C for 1 min, 32 cycles of denaturation at 94 • C for 30 s, 30 s at the primer's respective annealing temperature (typically 57 • C) and extension at 72 • C for 30 s, followed by an extension step at 72 • C for 2 min. Gel electrophoresis was performed on a 2% agarose gel containing 0.2 µg/mL ethidium bromide for 60 min at 175 V. A BioRad ChemiDoc XRS imaging system was used to visualize the DNA fragments by UV fluorescence (Hercules, CA, USA). Gel images were exported for publication as *.tiff files and uploaded to PowerPoint for annotation. PCR results were used to identify Alu insertion polymorphisms based on DNA fragment sizes for Alu presence/absence following gel electrophoresis. Genotypic data for each allele were recorded in an Excel spreadsheet as follows: (1,1) represented individuals homozygous present for the Alu insertion; (0,0) represented individuals homozygous absent for the Alu element; and (1,0) represented individuals heterozygous for the Alu element.

Sanger Chain Termination DNA Sequencing
Due to the many conflicting reports regarding the phylogenetic relationships among Aotus species, all potentially parsimony informative PCR results were subjected to traditional chain-termination DNA sequencing [54]. DNA sequencing was performed as described previously [55]. Briefly, four PCR fragments per locus were gel purified using a Wizard SV gel purification kit (Promega Corporation, Madison, WI, USA, catalog A9282) and 4 µL of the elution were used for chain termination cycle sequencing using BigDye Terminator v3.1. Four separate reactions were conducted for each locus using the original PCR primers (forward vs. reverse) and two internal-Alu primers (IntF1: 5 GGTGGCTCACGCCTGTAATC 3 [55] vs. SIntR1: 5 TCTCGGCTCACCGCAACCTCC 3 [16]). Following capillary electrophoresis, sequence quality was evaluated using the ABI software Sequence Scanner v2.0. Sequence alignments were constructed in BioEdit [50]. Alu elements in the reverse orientation (minus strand) were reverse complemented to be viewed in BioEdit in the forward orientation, and then a consensus sequence for each locus was aligned from the multiple forward and reverse sequences obtained for each locus.

Young Lineage-Specific Alu Elements and PCR
From a total of 658,009 full-length Alu insertions [43], we found 12,089 lineage-specific to the owl monkey [Anan_2.0] genome. We identified 1,992 of these elements as young (≤2% sequence divergence from their consensus). Local RepeatMasker [48] (RepeatMasker-Open-4.0) output for these 1,992 loci, along with their genome coordinates and Alu subfamily designation, are shown in Supplementary File S1. Wet bench locus-specific PCR analyses were performed on a subset of 332 young insertions, with 305 successful results. An additional nine phylogenetically informative Aotus-specific Alu insertion polymorphisms taken from Storer et al. (2020) [43] with >2% sequence divergence were re-analyzed for inclusion in this study for a total of 314 elements with successful PCR results. Potential oligonucleotide primers for PCR for an additional 1040 Alu elements (52% of the 1992 elements) designed using an in-house primer pipeline were also provided (https: //github.com/t-beck; accessed on 24 October 2022) (Supplementary File S1). Each locus name described in the text below is available in Supplementary File S1, along with a matching "query sequence" column containing the genomic coordinates. Data for the 314 young insertions with successful PCR results are available in Supplementary File S1 (worksheets "PCR this study" and "genotypes") and include representatives from 37 different Alu subfamilies [45], summarized in Table S1. PCR results for these 314 Alu elements segregated into three basic genotype categories: (1) homozygous present in all Aotus samples (N = 155) indicating high allele frequency or possible fixation within the genus; (2) homozygous absent in all DNA samples (N = 10), indicating very low allele frequency or perhaps unique to the [Anan_2.0] genome; or (3) polymorphic pattern for insertion presence/absence (N = 149). These results correspond to a~50% polymorphism rate within the candidate dataset. These 149 insertion polymorphisms were further divided into four general categories. (1) The target Alu insertion was restricted to the reference genome [Anan_2.0] species A. nancymaae and was homozygous present (h.p.) in all A. nancymaae samples (PCR code: A.n. (h.p.)) N = 31. An example is shown in Figure 1A. These are potentially useful molecular markers for A. nancymaae species identification. (2) The target Alu insertion was restricted to A. nancymaae and was polymorphic for insertion presence/absence among the nine A. nancymaae samples on our DNA panel (PCR code: A.n. (poly)) N = 83 ( Figure 1B). These Alu insertion polymorphisms could be applied to the study of population genetics among A. nancymaae wild populations. (3) The target Alu insertion was present in both A. nancymaae and A. azarae samples while being absent in A. trivirgatus and A. vociferans (PCR code: A.n. and A.a.) N = 21 ( Figure 1C). These could be employed to study Aotus phylogeny. (4) The target Alu was homozygous present in A. nancymaae, A. azarae and A. trivirgatus while being absent from A. vociferans, N = 3 ( Figure 1D), suggesting that A. vociferans is most basal among the species represented in this study. Further support of this phylogenetic relationship was found in a single locus that was homozygous present in all Aotus samples, with the exception of two A. vociferans individuals who lacked the insertion (Figure 2A). By contrast, two other Alu insertion polymorphisms (Aotus_127 and Aotus_1602) supported an alternative topology in which the target Alu was present in A. nancymaae, A. azarae and A. vociferans while being absent from A. trivirgatus ( Figure 2B), suggesting that A. trivirgatus is more basal than A. vociferans, among the taxa available in this study. All potential phylogenetically informative results were subjected to DNA sequencing of PCR amplicons to either confirm the initial genotypes or refute them and provide an alternative explanation.

DNA Sequencing
All potential parsimony informative insertion polymorphisms, as well as any ambiguous PCR fragments or evidence of incomplete lineage sorting (ILS), were subjected to DNA sequencing. These data are shown as Supplemental Figures in Supplementary File S2 and described below. DNA sequencing resulted in the discovery of five novel Aotus Alu insertion polymorphisms, two each from A. azarae and A. vociferans and one from A. trivirgatus samples (GenBank Accession Nos. OP549031 to OP549035). Future investigations with expanded species sampling and sample sizes are required to determine the utility of these novel elements. Sequencing also identified traces of non-repeat DNA with the potential for species identification.  Figures S1-S19), while in two other cases, the amplicons were too weak to be successfully sequenced (Locus Owl_2LS_040 and Storer_Aotus_1124). DNA sequencing of the remaining two candidates revealed the existence of "near parallel insertions", a different Alu insertion from the intended target, but still located between the oligonucleotide primers producing a PCR amplicon of the predicted "Alu-present" size. The PCR-based gel image for Locus Aotus_404 is shown in Figure S20a and illustrates how the amplicon sizes are similar for A. nancymaae and A. azarae, while the sequencing alignment ( Figure S20b) reveals that A. azarae has a different Alu about 90 bp 5 of the target insertion and in the opposite orientation. The second novel A. azarae Alu insertion was discovered in Locus Aotus_1196. In this case, A. azarae sample 85818 was the only one of three A. azarae individuals displaying the Alu present genotype, and DNA sequencing revealed that this individual has a different Alu about 260 bp upstream of the target insertion ( Figure S21). These two novel A. azarae-derived Alu insertions are not parsimony-informative in the current dataset but could be applied to Aotus phylogeny, particularly in neighboring populations of A. boliviensis and A. infulatus. An alternative oligonucleotide primer was designed in each case ( Figures S20b and S21) for investigators to potentially genotype these novel insertions separately.
In addition, PCR-based genotypes for Locus Aotus_364 indicated that the target Alu insertion was shared by A. nancymaae and A. azarae but also appeared to be homozygous present in A. vociferans sample 86218 ( Figure S22a). DNA sequencing confirmed that A. azarae shares the target insertion with A. nancymaae (the 20th confirmed case) but also revealed that A. vociferans 86218 has a different Alu in the opposite orientation about 115 bp upstream of the target ( Figure S22b). This novel A. vociferans-derived Alu insertion polymorphism is perhaps useful for Aotus phylogeny. Therefore, an alternative reverse PCR primer was designed, Aotus_364_vociferans-R: GAGTGTTTACTTTGTGCCAAGC, which could be used in conjunction with the existing forward primer to potentially genotype this locus separately ( Figure S22b). Depending on the allele frequency distribution of this locus, it could also be useful for species identification.
A second novel A. vociferans Alu insertion, was discovered in Locus Owl_2LS_120. PCR-based genotypes indicated that the target Alu element was restricted to the reference species A. nancymaae except for A. vociferans sample 85962, which appeared heterozygous for the insertion. Sequencing alignments revealed that A. vociferans 85962 lacks the target insertion and instead has a different Alu, a near parallel insertion, about 200 bp upstream ( Figure S23) in a genomic region dense with repetitive elements. An alternative reverse primer, Owl_2LS_120-A.voc.-R: CAAGTGAAGTGGAGGTTAGTGT, was designed to use in conjunction with the existing forward primer to potentially genotype this novel A. vociferans Alu element separately. These oligonucleotide primers were confirmed to have only one exact match each, in spite of the repeat dense genomic landscape.
In addition to the novel A. azarae and A. vociferans Alu elements discovered, a fifth novel Alu insertion, this one derived from A. trivirgatus, was revealed in Locus Aotus_68. PCR-based genotypes indicated that the target Alu insertion was polymorphic among A. nancymaae individuals and homozygous present in other Aotus samples, while A. trivirgatus individuals displayed an additional larger amplicon fragment suggesting that they might harbor an additional Alu insertion that is not present in other Aotus species ( Figure S24a). Sequence alignments confirmed that A. trivirgatus has a second Alu insertion located about 113 bp 5 from the target insertion. An alternative forward PCR primer (Aotus_68_A.triv-F: ACTTGCTTTGTCATGGCTTCATC) was designed that could be used in conjunction with the existing reverse primer to potentially genotype this newly discovered A. trivirgatus derived Alu element separately from the [Anan_2.0] ascertained target insertion ( Figure S24b).

Grey-Necked Species Group (A. vociferans and A. trivirgatus)
Alu insertion polymorphisms that grouped the two red-necked species, A. nancymaae and A. azarae, were generally absent from both A. trivirgatus and A. vociferans individuals.
However, as shown in Figures 1D and 2A, there were other Alu insertions that grouped A. trivirgatus, a traditional grey-necked or Northern species, together with A. nancymaae and A. azarae to the general exclusion of A. vociferans. DNA sequencing confirmed this genotype pattern for Locus Owl_2LS_078 ( Figure S25) and Locus Aotus_1518 ( Figure S26). In addition, PCR-based genotypes for Locus Aotus_1546 indicated that the target Alu insertion was polymorphic among A. nancymaae individuals, homozygous present in A. trivirgatus, A. azarae and the A. lemurinus griseimembra sample, while completely absent from all six A. vociferans individuals ( Figure S27a). DNA sequencing confirmed these genotypes ( Figure S27b), placing A. vociferans as the most basal among the Aotus species in this study; further, based on the heterozygous genotypes among some A. nancymaae individuals at this locus, these results suggest that A. nancymaae is more closely related to the grey-necked species than is A. azarae. Alternatively, this could be evidence of ILS or a consequence of captivity, although pedigree records indicate that the two heterozygous A. nancymaae individuals (86334 and 85974) are unrelated males.
For Locus Aotus_1288V2 (Figure 2A), the target Alu element is homozygous present in all Aotus subjects except for two A. vociferans individuals, 86100 and 86230, which displayed precise pre-integration sites with sequence traces spanning the target region, confirming the genotypes ( Figure S28). This locus also places A. vociferans as the most basal among the Aotus species in this study. However, it should be noted that A. vociferans individuals 86100 and 86230 are known to be full siblings (sister and brother, respectively; pedigree records provided by the source institution).
Locus Owl_2LS_042 was originally suspected of having a second Alu element, in addition to the target insertion, that was shared by A. trivirgatus and A. vociferans to the exclusion of A. nancymaae and A. azarae ( Figure S29a); if confirmed, that would represent a novel Alu element that grouped A. trivirgatus and A. vociferans together. However, DNA sequencing revealed that this shared larger PCR amplicon is not an Alu element, but rather a~224 bp insertion of extra non-repeat sequence about 80 bp 5 of the reverse primer and about 100 bp downstream of the target reference Alu insertion, which was confirmed to be shared by both A. trivirgatus and A. vociferans ( Figure S29b). The target Alu is likely homozygous present within Aotus. This extra shared sequence in both A. trivirgatus and A. vociferans was most likely inherited from a common ancestor, as it seems unlikely that such a relatively large insertion occurred independently in each lineage. Thus, it represents an informative serendipitous finding even though not a homoplasy-free structural variant (GenBank Accession No. OP549036).
Although the above results placed A. vociferans as the most basal Aotus species, we also recovered three other Alu insertions ( Figure 2B) that grouped A. vociferans together with A. nancymaae and A. azarae to the general exclusion of A. trivirgatus, suggesting that A. trivirgatus is the most basal Aotus species. Gel images for Locus Aotus_127 ( Figure S30a) and Locus Aotus_1602 ( Figure 2B) illustrate this genotype distribution. DNA sequencing results confirmed these insertion presence/absence patterns for both Aotus_127 and Ao-tus_1602 ( Figures S30b and S31, respectively). PCR-based genotypes for a third Alu element, Locus Storer_Aotus_622 indicated that the target Alu insertion was homozygous present in A. nancymaae and A. azarae, polymorphic among A. vociferans samples while being absent in all four A. trivirgatus individuals. DNA sequencing confirmed that both A. vociferans MVZ-155159 and A. azarae 85818 share the target Alu insertion with A. nancymaae, while both A. trivirgatus crl1556 and A. vociferans 86230 have a clean pre-integration site spanning the target region, thus confirming the genotypes ( Figure S32).
The remaining seven Alu insertion polymorphisms did not meet any of the general classifications described above and instead displayed erratic insertion patterns indicative of ILS or artifactual origin. PCR-based genotypes for Locus Owl_2LS_018 indicated that the target Alu insertion was polymorphic among A. nancymaae and A. vociferans individuals, homozygous present in A. trivirgatus while being absent in A. azarae. DNA sequencing confirmed this strange genotype distribution ( Figure S33), indicative of extensive ILS. PCR-based genotypes for Locus Aotus_1582 indicated that the target Alu insertion was ho-mozygous absent from all Aotus individuals except for A. vociferans MVZ-155159 and 86218 ( Figure S34a), perhaps suggestive of a de novo insertion unique to A. vociferans. However, DNA sequencing confirmed that both A. vociferans individuals share the target Alu insertion ( Figure S34b), unveiling an interesting case of ILS, especially given that no A. nancymaae samples in our DNA panel possess the reference target insertion. Another suspected case of extensive ILS occurred for Locus Owl_2LS_046 in which PCR-based genotypes indicated that the target Alu insertion was polymorphic among A. nancymaae, absent from A. azarae and A. vociferans while being homozygous present in all four A. trivirgatus individuals ( Figure S35a), giving the appearance that A. trivirgatus is most related to A. nancymaae. However, DNA sequencing revealed that the A. trivirgatus amplicon does not contain an Alu element but rather is comprised of 306 bp of non-Alu sequence, some of which is an L1, and therefore the target Alu insertion is restricted to A. nancymaae ( Figure S35b).
Locus Aotus_1153 displayed a PCR pattern in which only the A. vociferans samples lacked what was suspected to be a second Alu element (a larger amplicon) in addition to the intended target insertion ( Figure S36a). However, DNA sequencing revealed that this larger amplicon observed in some A. trivirgatus, A. nancymaae and A. azarae samples was not an Alu element but rather an uninformative artifact sequence ( Figure S36b). Similarly, the target Alu insertion in Locus Aotus_1422 appeared to be restricted to A. nancymaae except for a possible truncated form observed only among A. vociferans; however, DNA sequencing revealed A. vociferans contained only a non-repetitive sequence in this region ( Figure S37). Lastly, two Alu insertion polymorphisms seemed homozygous present in all Aotus samples except for a single heterozygous individual, indicative of ILS. DNA sequencing confirmed the heterozygous genotype in both cases, Locus Owl_2LS_110 and Locus Aotus_1765 (Figures S38 and S39, respectively). Post-sequencing adjusted genotypes and comments are available in Supplementary File S1. Sequence alignments constructed in BioEdit [50] are available as *.GB files in Supplementary File S3.

Maximum Parsimony
Post-sequencing adjusted genotype data were used to construct a data matrix of 26 taxa (23 Aotus samples and 3 outgroups) and 314 characters to perform a maximum parsimony heuristic search. Of the 314 Alu insertion loci, 165 were constant, and another 114 were restricted to A. nancymaae, leaving 35 potentially parsimony informative among Aotus species (Figure 3).
The debate over the phylogenetic relationships among Aotus species has generally centered around the placement of A. nancymaae, either as a red-necked species within the southern group or as a member of a northern geographical group more closely related to the traditional grey-neck species. The topology of this tree strongly supports A. nancymaae and A. azarae as sister taxa with nearly 100% bootstrap support, as well as the monophyly of both A. trivirgatus and A. vociferans as separate taxa. A. vociferans is basal on the tree with only 59% bootstrap support, while three Alu elements support an alternative topology, placing A. trivirgatus as the most basal. The tree generally supports the division of traditional red and grey-necked species across these five taxa.

Discussion
The Alu data presented in this study are in general agreement with the research of Hershkovitz [4] and studies using other nuclear genetic markers [7,8] by providing support for the separation of Aotus taxa along traditional red and grey-necked groups. Among the studies that utilized mtDNA markers, this Alu study is more in agreement with Ruiz-Garcia et al. (2011) [3], who concluded that A. nancymaae is the red-necked species most closely related to the grey-necked group and that the ancestors of the grey-necked A. vociferans likely represent the original species that led to the formation of the current Aotus genus. The Alu evidence in this study firmly places A. nancymaae together with A. azarae, in contradiction to other studies [1,2] that placed A. nancymaae within the northern species group alongside A. vociferans.
However, a recognized limitation of the current Alu study is that all the specimens were primarily from captive breeding colonies, a zoo or from museum archives and therefore are unlikely to harbor widespread geographic origins that would accurately simulate the modern dynamics of wild populations [1]. Further, this study only included five Aotus taxa, and only four of them contain multiple specimens per species, some of which were obtained from at least two different independent sources, while others are known relatives based on pedigree records. Moreover, this study did not include representative samples from several recognized species (including A. infulatus or A. nigriceps, which are geographic neighbors to A. azarae), nor did it include A. miconax, A. zonalis or A. brumbacki from northern species neighboring A. vociferans [1]. Therefore, the resolution of Aotus phylogeny was not a reasonable expectation of this study. However, the observation that the Alu genotypes in this study are often segregated by species group, more so than by sample source, indicates that this Alu dataset could be broadly applied to wild specimens to track population genetics and infer phylogenetic relationships.
The discrepancy in the placement of A. nancymaae observed between nuclear and mtDNA-based phylogenies implies that the maternally derived mitochondrial genome is ancestrally distant from that of the nuclear genome, as has been reported, for instance, in baboons [57]. This can be caused by male migration and introgression into more stable female populations. An alternative explanation provided by Martins-Junior et al. (2022) [1] suggests that successive periods of rapid divergence among owl monkeys allowed the faster-evolving mitochondrial genome to keep pace, compared to the slower rate of nuclear markers. Another explanation discussed is possible ancestral hybridization. Although there is evidence of inter-species Aotus hybridization in captivity [58], there is no evidence of hybrids among the samples in the current study and very little evidence of successful inter-species hybridization in wild Aotus populations in sympatric regions [1]. The Alu data in this study, however, confirmed the existence of extensive ILS among the Aotus taxa investigated, warranting further investigation. Implementation of identical-by-descent retrotransposons could help resolve some of these complex issues.
The number of Aotus lineage-specific Alu insertions (~12,000) is considerably higher than the~5000 reported for humans [59] and somewhat higher than the~9000 reported for capuchin [43,44], a fellow member of the related Cebidae lineage. A high rate of Alu retrotransposition increases the likelihood of detecting "near parallel insertions", which are independent retrotransposons within the span of the same PCR amplicon. This situation can complicate PCR interpretation unless resolved by sequencing, as they were in this study. However, a high retrotransposition rate also provides an opportunity for serendipitous findings of novel Alu elements from taxa currently lacking whole genome sequence assemblies. The five novel Alu elements reported in this study were not ascertained from the A. nancymaae reference genome [Anan_2.0] and are, therefore, not subject to the same level of directional ascertainment bias. Until more genomes for other Aotus species become available, the five novel Alu elements reported here provide alternate evolutionary perspectives to [Anan_2.0].
Alu insertion polymorphisms were identified from 37 different platyrrhine Alu subfamilies, all of which were AluTa15 derived [45]. The highest polymorphism rate occurred among subfamilies discovered in the owl monkey genome [Anan_2.0], meaning that those Alu subfamilies do not have detectable members in the platyrrhine genomes characterized prior to Aotus, those being marmoset, [calJac3], squirrel monkey, [SaiBol1] and capuchin, [Cebus imitator_1.0]. This strategic approach facilitates the discovery of the youngest Alu subfamilies with the ongoing proliferation of progeny elements.

Conclusions
This study provides a comprehensive dataset of Aotus lineage-specific Alu elements and demonstrates their utility as phylogenetically diagnostic insertion polymorphisms. It shows that multiple Alu subfamilies have evidence of recent mobilization within owl monkey genomes. Alu elements restricted to A. nancymaae, as well as the discovery of novel Alu elements from three additional owl monkey taxa, will assist in species identification. Application of this dataset of Alu insertion polymorphisms to a large sample size of wild populations could potentially resolve some of the complex issues influencing the current taxonomy. The Alu insertions reported in this study will provide researchers with an additional resource for the study of Aotus phylogenetic and population genetic relationships, bringing more insight into the elusive "night monkeys" to further facilitate conservation strategies.
Supplementary Materials: The following supporting information can be downloaded at https:// www.mdpi.com/article/10.3390/genes13112069/s1: Supplementary File S1 is an Excel file containing the RepeatMasker output for the Alu elements ascertained from [Anan_2.0] and separate worksheets for Table S1, PCR this study, DNA samples and Genotypes. Supplementary File S2 is a PDF containing the supplemental figures, Figures S1 to S39, including DNA sequencing alignments and the associated gel image if required for illustration. Supplementary File S3 is a compressed zipped folder containing the DNA sequencing alignments as *.GB file format.