Genetic Markers for Species Conservation and Timber Tracking: Development of Microsatellite Primers for the Tropical African Tree Species Prioria balsamifera and Prioria oxyphylla

: Research Highlights: Two novel sets of polymorphic microsatellite markers were developed for Prioria balsamifera and Prioria oxyphylla through high-throughput sequencing. Validation in two populations of each species proved the utility of the developed primers to estimate genetic diversity at population level. Background and Objectives: Prioria balsamifera and Prioria oxyphylla are tropical tree species from Central Africa. They produce a high-quality, multi-purpose timber that is of great interest to the international market. Prioria balsamifera has been included as ‘endangered’ on the IUCN Red List of Threatened Species. In order to set up adequate management plans and facilitate timber tracking, knowledge on the genetic diversity at population level is needed. Therefore, we aim to develop microsatellite markers that can be used for species conservation, forensics, plant breeding and population genetics studies. Materials and Methods: Genomic DNA of P. balsamifera and P. oxyphylla was sequenced on an Illumina NextSeq platform (Illumina Inc., San Diego, CA, USA), generating 829,421 and 772,018 paired-end reads that contained 7148 and 7004 microsatellite sequences, respectively. The QDD-pipeline was used to design primers, which were tested for amplification in two populations of each species. Cross-species amplification was tested in all seven African Prioria species. Results: For P. balsamifera , 16 polymorphic microsatellite markers were developed and combined in three multiplexes. Inbreeding appeared to be absent but genetic diversity was low in both populations. For P. oxyphylla , 15 polymorphic microsatellite markers were developed and combined in three multiplexes. Genetic diversity was low in both populations and estimated null allele frequencies were high for multiple loci. Cross-species amplification tests demonstrated the occurrence of conserved loci that amplified for most of the African Prioria species. Conclusions: The microsatellite markers prove to be useful for estimating genetic diversity at population level. These novel markers can be used to study gene flow and spatial genetic structure in Prioria species, which is needed to set up proper conservation guidelines and to prevent genetic erosion.

P. oxyphylla may be suitable for commercial plantations. For this, information about the intraspecific diversity is of great importance, as it optimises breeding programmes by enabling the selection of superior parent trees. Properly managed plantations or production forests can reduce the pressure on wild stands and increase the timber availability for the international market. Additionally, microsatellites provide a useful tool for timber tracking based on DNA fingerprinting [10,14,15]. Since illegal logging causes degradation and loss of forests worldwide, and trade in illegal timber and wood products distorts global markets, unfalsifiable methods to identify the origin of timber, such as DNA fingerprinting, are extremely valuable. Moreover, since microsatellite loci are usually short fragments (<500 bp), amplification from degraded DNA, often found in logs or processed wood, is possible.
To address the listed issues, we aim to develop a novel set of microsatellite markers for Prioria balsamifera and Prioria oxyphylla using high-throughput sequencing data. These markers will be validated in two populations of each species in order to assess their applicability. Additionally, crossspecies amplification will be tested in all seven African Prioria species.

Microsatellite Primer Development
DNA was isolated from silica-dried leaves of Prioria balsamifera (corresponding voucher: BR0000013003845 [16]) and Prioria oxyphylla (corresponding voucher: BR0000013007126 [17]), using a cetyltrimethylammonium bromide (CTAB) protocol [18] with an additional sorbitol washing step. Subsequently, the extracted DNA was used to develop a non-enriched genomic library for each species using a modified version of the protocol in [19], described in full by Tosso et al. [20]. Genomic libraries were sent for sequencing on an Illumina NextSeq platform at GIGA (Liège, Belgium). The sequencing run generated 829,421 paired-end 150 bp reads for P. balsamifera and 772,018 paired-end 150 bp reads for P. oxyphylla.
Reads were merged with FLASH v1.2.11 (Center for Computational Biology, Baltimore, MD, USA) [21] and the QDD pipeline [22] was used to identify microsatellite loci, for which suitable primers were developed with the implemented Primer3 algorithm [23] using the parameters described in [24]. For P. balsamifera, 7148 microsatellite sequences with primers were generated, from which the 24 best primer pairs, forward and reverse, were selected following the criteria described in [22]. All chosen microsatellite loci contained 8 to 15 di-or trinucleotide repeats. For P. oxyphylla, 7004 microsatellite sequences with primers were generated and as for P. balsamifera, the 24 best primer pairs were selected. Here, all chosen microsatellite loci contained 8 to 23 di-or trinucleotide repeats.
First, the two sets of the 24 selected primer pairs were tested individually on three individuals of the corresponding species. The products used for the amplification reaction and the PCR conditions were the same as in [25]. PCR products were visualized with a QIAxcel Advanced system (QIAGEN, Venlo, Netherlands) to check whether amplification was successful.
Second, the primers that showed amplification in at least one of the three samples were reamplified and genotyped to assess their level of polymorphism and to check the readability of the peak pattern. The amplification protocol with low-cost M13-like fluorescent labelling and the PCR conditions are described in [25]. In this study, four universal sequences (Q1-Q4, Table 1) and fluorescent dyes (6-FAM, NED, VIC and PET respectively) (Applied Biosystems, Foster City, CA, USA) were used; the annealing temperature during the first 25 repeated cycles was set to 55 °C. Genotyping was done on an ABI 3730 DNA Analyzer (Applied Biosystems, Foster City, CA, USA), with 0.8 µL PCR product (3 to 4 samples were pooled), 10 µL Hi-Di Formamide (Applied Biosystems) and 0.3 µL MapMarker 500 labelled with DY-632 (Eurogentec, Liège, Belgium). The resulting electropherograms were analysed using the Microsatellite Plugin 1.4.6 in Geneious 9.1.8 (Biomatters Ltd., Auckland, New Zealand).  1 The universal linkers (Q1-Q4) attached to the sequence of the forward primers are underlined. 2 The number of repeats found in the sample used for the primer development. Corresponds to the GenBank accession number.
Third, unreadable and monomorphic loci were excluded, while the remaining loci were combined in various multiplexes per species using Multiplex Manager 1.2 [26] to reduce costs and hands-on time in the lab. Amplification of the different primer combinations was tested in seven samples per species with the Type-it Microsatellite PCR kit (QIAGEN) using the amplification volumes and PCR conditions described in [25], except that the annealing temperature during the first 25 repeated cycles was at 56 °C instead of 57 °C.
Cross-species amplification was tested with the final P. balsamifera and P. oxyphylla multiplexes on the other five African Prioria species: P. buchholzii, P. msoo, P. gilbertii, P. mannii and P. joveri.
The microsatellite containing sequences obtained through high-throughput sequencing and used for the primer development were deposited on GenBank (Table 1).
Extractions were made from leaves (silica-dried or herbarium) and cambium using a cetyltrimethylammonium bromide (CTAB) protocol [18] with an additional sorbitol washing step. The PCR and genotyping conditions were the same as those in the third step of the primer development (see Section 2.1). The resulting electropherograms were analysed using the Microsatellite Plugin 1.4.6 in Geneious 9.1.8 (Biomatters Ltd.).
The number of alleles per locus (NA), observed (Ho) and expected (He) heterozygosity, Weir and Cockerham's fixation index (FIS) and the deviation from Hardy-Weinberg Equilibrium (HWE) were calculated with SPAGeDi 1.5d [27]. The Jackknife method in INEST v2.2 [28] was used to estimate the expected frequency of null alleles. Polymorphism Information Content (PIC) was calculated using CERVUS 3.0.7 (Field Genetics Ltd., London, UK) [29].

Microsatellite Primer Development
For Prioria balsamifera, 23 out of the 24 primer pairs selected for testing in the laboratory could be amplified in at least one of three samples. These 23 pairs were then reamplified and genotyped, resulting in 16 useful polymorphic primer pairs, which were combined in three multiplexes (Table 1) and used for all subsequent analyses. Two out of the seven removed primer pairs (Table S1) appeared monomorphic in the seven samples used for testing, while another four pairs failed to amplify, probably due to problems with the incorporation of the fluorescently labelled Q-tails. A last primer pair showed polymorphism and successfully amplified in simplex, but amplification was unsuccessful in all multiplex combinations that were tested.
For Prioria oxyphylla, 16 out of 24 primer pairs amplified in at least one of three samples. Reamplification and genotyping of these 16 pairs yielded 15 readable polymorphic primer pairs, which were combined in three multiplexes ( Table 1). The pair that was removed showed multiple peaks and stutters, which made it impossible to infer allele sizes and to define bins.

Population Genetics in Prioria balsamifera
The final three microsatellite multiplexes containing 16 primers for P. balsamifera were tested for population genetic analyses in 65 individuals originating from Luki (n = 34) and Yangambi (n = 31). Microsatellite loci showed one to nine alleles per locus in both populations combined, with an average of 4.310 alleles per locus (Table 2). Although loci PriB4 and PriB14 appeared to be monomorphic in individuals from Luki and Yangambi, multiple alleles were detected when testing for polymorphism during primer development. Consequently, both primers could be informative in genetic studies that consider a larger part of the species' distribution area. The observed and expected heterozygosity ranged from 0 to 0.735 (average Ho = 0.412) and 0 to 0.741 (average He = 0.399) in Luki, respectively. In Yangambi, Ho ranged from 0 to 0.655 (average = 0.273) and He from 0 to 0.749 (average = 0.302) ( Table 2). Deviation from HWE occurred at one locus (PriB12) in the Luki population and two loci (PriB02 and PriB18) in the Yangambi population. These loci were associated with higher null allele frequencies (0.141, 0.125 and 0.209, respectively). Although inbreeding appeared to be absent, genetic diversity was low in both P. balsamifera populations. Hence, genetic erosion appears to be a real risk and efforts should be made to protect P. balsamifera genetic resources. PIC values ranged from 0 to 0.737, with an average of 0.364. Thus, most of the developed microsatellite markers could be useful for parentage analyses and setting up breeding programmes for the endangered P. balsamifera.

Population Genetics in Prioria oxyphylla
For P. oxyphylla, the final three microsatellite multiplexes containing 15 primers were validated in 53 individuals collected in Luki (n = 21) and Yangambi (n = 32). Microsatellite loci showed 2 to 10 alleles per locus in both populations combined, with an average of 5.467 alleles per locus (Table 2). Interestingly, locus PriO4 showed only two alleles, with each allele being specific to either Luki or Yangambi. So, based on the allele observed for this locus only, the origin of an individual could be determined when considering trees from Luki and Yangambi.
Ho ranged from 0 to 0.762 (average = 0.261) and He from 0 to 0.788 (average = 0.429) in Luki, and similarly in Yangambi, Ho ranged from 0 to 0.750 (average = 0.289) and He from 0 to 0.810 (average = 0.488) (Table2). Multiple loci deviated from HWE in both Luki and Yangambi (4 and 7 respectively). All these loci showed null alleles. Surprisingly, various loci were characterized by high levels of missing genotypes. Only the sample used for the primer development showed successful amplification for all 15 loci. PIC values ranged from 0.136 to 0.782 with an average of 0.564. Hence, the developed markers could be a useful tool for parentage studies and for setting up breeding programmes.

Microsatellite Cross-Species Amplification in African Prioria
The cross-species amplification tests showed that multiple loci from both microsatellite sets are conserved across species (Table 3), as various loci were successfully amplified in the different African Prioria species. Therefore, the primers developed in this study could be of great value for genetic studies focused on other Prioria species as well. Since P. oxyphylla and P. buchholzii are closely related sister-taxa [30], cross-species amplification appeared to be highly successful. The low amplification success in P. joveri (both primer sets) and P. balsamifera (P. oxyphylla primer set) could result from higher species divergence.

Conclusions
The developed sets of microsatellite primers for Prioria balsamifera and Prioria oxyphylla, containing 16 and 15 polymorphic loci respectively, appear to be useful for estimating genetic diversity at population level. The novel markers can be used to study gene flow and spatial genetic structure in Prioria species, which is needed to prevent genetic erosion and to set up proper conservation guidelines. Additionally, the microsatellite markers can be valuable for timber tracking through DNA fingerprinting.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Figure S1