Rapid sex determination of a wild passerine species using loop‐mediated isothermal amplification (LAMP)

Abstract Many bird species are sexually monomorphic and cannot be sexed based on phenotypic traits. Rapid sex determination is often a necessary component of avian studies focusing on behavior, ecology, evolution, and conservation. While PCR‐based methods are the most common technique for molecularly sexing birds in the laboratory, a simpler, faster, and cheaper method has emerged, which can be used in the laboratory, but importantly also in the field. Herein, we used loop‐mediated isothermal amplification (LAMP) for rapid sex determination of blood samples from juvenile European blackcaps, Sylvia atricapilla, sampled in the wild. We designed LAMP primers unique to S. atricapilla based on the sex chromosome‐specific gene, chromo‐helicase‐DNA‐binding protein (CHD), optimized the primers for laboratory and field application, and then used them to test a subset of wild‐caught juvenile blackcaps of unknown gender at the time of capture. Sex determination results were fast and accurate. The advantages of this technique are that it allows researchers to identify the sex of individual birds within hours of sampling and eliminates the need for direct access to a laboratory if implemented at a remote field site. This work adds to the increasing list of available LAMP primers for different bird species and is a new addition within the Passeriformes order.

A variety of molecular methods have been developed for using CHD in sex determination, including restriction fragment length polymorphism (RFLP), and multiple PCR-based methods, with the latter being more common (Ellegren, 1996;Fridolfsson & Ellegren, 1999;Griffiths et al., 1998;Ito et al., 2003;Kahn, St John, & Quinn, 1998;Morinha, Cabral, & Bastos, 2012;Vucicevic et al., 2013). While PCR has become the standard for amplification of specific DNA sequences, this technique requires high-precision instruments for amplification, such as a thermal cycler, and/or an extended protocol, gel electrophoresis for example, for visualization of the amplified product. This can lead to needing more time, money, and supplies for carrying out DNA amplification. However, relatively recently a new amplification technique has been developed, called loop-mediated isothermal amplification (LAMP), which amplifies DNA with high specificity, efficiency, and speed under isothermal conditions (Notomi et al., 2000), and without the need for direct access to a laboratory (Lee, 2017).
Logistically, this new method offers a simpler protocol with the LAMP reaction requiring a single enzyme (DNA polymerase with strand displacement activity that can greatly amplify from a minimal number of DNA copies) and a single temperature, which removes the need for an expensive, high-precision thermal cycler. Moreover, product detection can be achieved directly within the reaction tube by a diagnostic color change and fluorescence, eliminating the need for electrophoretic techniques and related equipment (Centeno-Cuadros, Abbasi, & Nathan, 2017;Lee, 2017;Mori, Nagamine, Tomita, & Notomi, 2001). The principle of this method, based on autocycling strand displacement, is however more complex than traditional PCR and requires four specially designed primers that recognize six distinct regions on the target DNA ( Figure 1). The process by which LAMP recognizes the target allows for amplification of the target sequence with high selectivity, and the final products are stem-loop DNAs with several inverted repeats of the target in the same strand (for a detailed description of the LAMP mechanism, please see Notomi et al., 2000 andTomita, Mori, Kanda, &Notomi, 2008).
The blackcap is a common European passerine and serves as a model system for studying migration ecology as well as the underlying evolutionary genomic machinery controlling this complex behavioral phenomenon. Blackcaps exhibit the entire spectrum of F I G U R E 1 Schematic illustration of LAMP primer design for CHD-Z. (a) Four different regions (forward: F2, F3; backward: B2, B3) and two complementary regions (forward: F1c; backward: B1c) are located on target DNA. (b) LAMP primers: two outer primers (forward: F3; backward: B3) and two inner primer pairs (forward: FIP; backward: BIP) are used in each LAMP reaction. The FIP (BIP) primer is generated by connecting the F1c (B1c) sequence and F2 (B2) sequence with a T-linker. (c) Primer sequences located in CHD-Z (5'-3') showing the six different target regions (forward: F3, F2, F1c; backward: B3, B2, B1c). Primer sequences for B3, F1c, and B2 are ordered as the reverse complements to the sequences above (see Tomita et al., 2008 andCenteno-Cuadros et al., 2017;see also (Berthold, Helbig, Mohr, & Querner, 1992;Cramp, 1992;Helbig, 1991;Perez-Tris, Bensch, Carbonell, Helbig, & Telleria, 2004). For example, blackcap populations may comprise short/medium/long-distance migrants, partial migrants, or nonmigrants (i.e., Herein, we used a subset of naïve, wild-caught juvenile blackcaps, which were sampled from southern Spain, as part of a larger evolution and behavioral study on migratory behavior, to test the applicability of our optimized LAMP primer sets for sex determination in this species. For the behavioral study, the target gender was male as it is the homogametic sex in birds (ZZ), and thus, whole-genome sequencing (a component of the behavioral study) results in equal coverage of the Z chromosome and autosomes. The behavioral study is an example of a study that would benefit greatly from a LAMP-based sexing protocol, where in situ sex determination would lead to rapid identification and collection of male samples, along with instant identification and subsequent release of female birds, thereby reducing stress on them.
Once birds were captured in the field, a blood sample was taken from each individual and samples were then transported back to a laboratory for sex determination and further analyses. We selected a random subset of samples to be tested using a standard PCR-based protocol (Fridolfsson & Ellegren, 1999;Griffiths et al., 1998) and LAMP-based protocol to confirm gender and test reliable TA B L E 1 LAMP primers, including sequences and optimal amplification conditions, designed to amplify species-specific CHD-W and CHD-Z genes for Sylvia atricapilla Target Primer Name 5′-Sequence-3′

F I G U R E 2
The European blackcap (Sylvia atricapilla). Adult male (left) and female (right) birds with characteristic black (m) and brown (f) "caps" applicability of the LAMP method in this setting. Our optimized LAMP-based protocol proved to be just as accurate, but much more time efficient compared to the PCR-based protocol.

| Sample collection
The collection of birds was designed in a way to support the needs of the larger behavioral study, and the specifics of collection, for example, site, date, and migratory phenotype, are not necessarily relevant for the present study. In summary, 67 birds were collected in August 2017 from two different field sites in southern Spain. Birds were captured using mist nets and tape luring, using audio recordings of the male blackcap territorial song. Blood samples (ca. 50 µl) were taken, using a capillary tube, from the brachial vein and stored in 500 µl SET buffer (0.015 M NaCl, 0.05 M Tris, 0.001 M ethylenediaminetetraacetic acid, pH 8.0) and stored in a −20°C freezer until tested. We chose blood over feathers, for example, because bird blood is a better tissue for obtaining high amounts of DNA; nucleated erythrocytes provide much more DNA compared to relatively low amounts of DNA recovered from feather samples (Harvey, Bonter, Stenzler, & Lovette, 2006). Eighteen samples with a roughly equal representation of males and females were randomly chosen to be tested, using both PCR and LAMP protocols in the laboratory.

| DNA extraction optimization
Prior to testing the samples with LAMP, we optimized DNA extraction using blood samples from adult birds of known gender (positive controls; data not included). Using no more than half of the sample volume of blood in SET buffer (~250 µl), DNA was extracted using a simple and fast HotSHOT DNA extraction method (Truett et al., 2000), which uses sodium hydroxide (NaOH), requires less than 45 min, and can be carried out using a single thermoblock set to 95°C. For a detailed protocol, please see Truett et al., 2000, as we followed it directly. Optimizing incubation timing (10 min, 30 min, 1 hr) resulted in 30 min incubation duration working best. We optimized starting DNA amount by testing a range of concentrations (1.5-3.5 ng/µl, 15-45 ng/µl and ~400 ng/µl) and found <5 ng/µl to be optimal. LAMP is known to be highly sensitive and able to amplify only a few copies of DNA. DNA yield was quantified using a NanoDrop™ spectrophotometer.
We also optimized the DNA extraction protocol to be used as a "quick and dirty" method for field application, where access to a nucleic acid quantification machine, for example, NanoDrop™, is unlikely. While a 50 µl blood sample is recommended, variability/ difficulty in handling live animals (i.e., sampling error) may lead to more or less than that. Therefore, blood samples should be binned into three different categories, based on the relative color of the sample. More concentrated samples will appear dark red while less concentrated samples will appear diluted and lighter in color. We applied this system to our known control samples and designated them as light, medium, or dark red (with two replicates each), carried out the HotSHOT DNA extraction protocol, quantified DNA yield using a NanoDrop™, performed 1:1, 1:5, 1:10, and 1:100 dilutions to determine final DNA concentrations and which would be optimal for the subsequent LAMP protocol. Too much or too little DNA will lead to failed LAMP reactions. We found light and medium samples (30-50 µl blood + 500 µl SET buffer) yielded ~130-150 ng/ µl of isolated DNA; therefore, 1:30 dilutions would be necessary.
Alternatively, dark samples (>50 µl blood + 500 µl SET buffer) yielded ~450-550 ng/µl of isolated DNA, so 1:100 dilutions would be sufficient for more concentrated blood samples. While it is an imprecise method for estimating DNA concentration, we found it to work reliably across the tested samples in the laboratory, and therefore know it will also work robustly in the field.

| LAMP primer design and optimization
Sex determination using LAMP requires two sets of primers and thus two reactions per sample, based on CHD-Z and CHD-W genes.
CHD-Z primers target the Z chromosome which is found in both males (ZZ) and females (WZ) and will therefore show a positive LAMP reaction for both sexes. A CHD-Z reaction that produces a negative result indicates a problem, and a repeat test should be con- To optimize primers, we used the same set of blood samples of for CHD-Z or CHD-W) that yielded no amplification for any combination of temperature and time was discarded, and a new primer set was randomly selected and tested. Final primer sets, along with sequences and optimal amplification conditions, are described in Table 1.

| Sex determination using LAMP
For setting up LAMP reactions, we used a protocol that was developed and tested in the laboratory but could be easily modified for application in the field. We followed Hamburger et al. (2013)  (1:100 dilution or 1:30 dilution, depending on the initial blood concentration) was added to each reaction tube, mixed well, and then incubated in the same thermoblock that was utilized for DNA extractions, at 65°C for 60 min (CHD-W reactions) or 80 min (CHD-Z  (Parida, Sannarangaiah, Dash, Rao, & Morita, 2008). A negative LAMP reaction will not produce a color change. The color change can also be easily visualized by irradiating the LAMP reactions with UV light (320 nm). Positive reactions will fluoresce (yellow), negative reactions will not.

| Sex determination using PCR
To corroborate LAMP results, we tested the same 18 samples using a standard PCR-based molecular sexing protocol and a single set of primers, P2 and P8, which are also designed from the CHD genes (for a detailed description of the protocol and primers, please see Griffiths et al., 1998). PCR amplification was carried out using a thermal cycler and lasted roughly two hours. Afterward, PCR products were visualized by running a 3% gel agarose electrophoresis (100 V, 2.5 hr). A single band (CHD-Z) indicates a male, whereas the presence of a second band (CHD-W) indicates a female. For this test, the number of bands serves as the diagnostic indicator, not fragment size.

| LAMP optimization and reactions
After testing different combinations of incubation temperature and time, we found that males and females were clearly and consistently distinguished when LAMP reactions were run at 65°C for 60 min (CHD-W primers) and 80 min (CHD-Z primers; Table 1

| Comparison of LAMP versus PCR
We tested the same 18 samples using a standard molecular sexing protocol (PCR-based) to confirm LAMP results and compare the efficiency of the protocols in the laboratory. We visualized amplified products from both protocols by running an agarose gel electrophoresis and found that results corroborated each other, with samples 2-3, 7-9, and 13-15 determined to be female while the rest were identified as male (Figure 4). For the LAMP reactions, CHD-W and CHD-Z products were run in parallel for each sample, and a ladder-like pattern, indicating a positive LAMP reaction, was observed for all CHD-Z reactions ( Figure 4a), but only eight CHD-W reactions (samples 2-3, 7-9, 13-15; Figure 4b). For the PCR amplicons, each sample was run in a single lane and results revealed the same pattern of sex determination, eight females (two bands) and ten males (1 band) (Figure 4c).

| D ISCUSS I ON
Herein, we designed species-specific LAMP primers for rapid sex determination of wild-caught juvenile blackcaps, S. atricapilla.
We optimized the entire protocol, including DNA extraction and LAMP, in a way that would allow for fast and easy application in F I G U R E 3 LAMP results for rapid sex determination of wild-caught juvenile blackcaps, S. atricapilla. Two LAMP reactions were run for each sample (n = 18), using primer sets amplifying fragments of CHD-Z (a, c) and CHD-W (b, d) genes. (a-b) Positive LAMP reactions were visualized by an immediate color change, orange to yellow, after staining with SYBR Green I. Negative reactions remained orange. Across samples, eight were reliably identified as female ("f", positive CHD-Z and CHD-W reactions; both yellow) and ten were male ("m", positive CHD-Z reactions (yellow) but negative CHD-W reactions (orange)). Negative controls (−) were included to test for spurious amplification.
(c-d) Positive LAMP reactions were confirmed by irradiating tubes with UV light and detecting fluorescence (here, color is inverted to see the contrast). Negative reactions did not fluoresce. The same samples, 2-3, 7-9, and 13-15 reliably revealed to be female ("f") while the rest were male ("m"; 1, 4-6, 10-12, and 16-18) both the laboratory and a field setting. Consequently, one of our goals was to find one single incubation temperature that would work accurately for both CHD-W and CHD-Z primer sets so that separate reactions could be run simultaneously on the same thermoblock, requiring less equipment and time overall. We found that when incubated at 65°C, LAMP could clearly and reliably distinguish male and female samples, with only a 20-min difference in incubation times for CHD-W and CHD-Z primers. Therefore, The most commonly reported drawback of LAMP is high rates of false positives. The powerful amplification mechanism of LAMP, which is beneficial for having highly sensitive assays, also renders it highly susceptible to carry over contamination, where amplified DNA products from previous LAMP reactions become templates for reamplification (Hsieh, Mage, Csordas, Eisenstein, & Soh, 2014;Tomita et al., 2008). Because this assay produces a large amount of DNA, which can easily spread into the open air, and because the protocol requires opening of tubes, these attributes ultimately lead to increased risk of carry-over contamination and false-positive results. High temperature, humidity, and inadequate volume of reagents are known risk factors contributing to carry over contamination (Nagai et al., 2016). False-positive results may also be obtained through primer-dimer formation, which is more likely to occur when high concentrations of primers are used, as are in a LAMP assay (Meagher, Priye, Light, Huang, & Wang, 2018 (Ball et al., 2016;Hsieh et al., 2014;Wang, Brewster, Paul, & Tomasula, 2015).
In terms of time and efficiency in the laboratory, the LAMP protocol was completed in less than two and half hours while the standard molecular sexing protocol required twice as long (~5 hr), along with the need for access to laboratory infrastructure. The LAMP protocol can be further simplified by only carrying out two of the three visualization methods, the staining and UV irradiation, which offer immediate observable diagnostic results. Running a gel serves as an additional visual confirmation and was used here for further validation, but it requires more time and supplies to carry out and is not essential for detecting successful LAMP reactions. Furthermore, running a gel at a remote field site would not be possible, so dropping this step lends the LAMP protocol to be more field applicable, unlike the standard PCR-based molecular sexing protocol, which relies solely on gel electrophoresis for visualizing results.
The key objective of this study was to develop species-specific LAMP primers that could be used to easily and quickly identify the gender of juvenile blackcaps both in the laboratory and in the wild.
While our protocol was developed and carried out in the laboratory, the CHD-Z and CHD-W primers designed herein can be easily implemented in a remote field setting (for a field-based LAMP protocol, please see Centeno-Cuadros et al., 2017). Having prepared ready-mixes that can be used and stored at room temperature, a simple DNA extraction protocol that can be run in a single tube per sample, isothermal amplification requiring a single piece of equipment operated from a car, and an immediate visual diagnosis of results, together, allow for a quick and streamlined process of sex determination. However, for field application we strongly recommend preparing primer mixes for each set and vacuum-drying them, as Centeno-Cuadros et al., 2017 describe, to preserve their integrity during transportation to the field.
Finally, for studies that are looking to collect birds or samples from one focal gender, the amount of stress on captured birds with the nontarget gender can be reduced by means of rapid in situ sex determination and subsequent release. Or for studies needing to collect a certain number of birds or samples for both sexes, rapid sex determination can be done on the spot. This eliminates the need to take birds or samples back to a laboratory after every collection, employ traditional molecular sexing techniques, and then return to the field to either release birds of the nonfocal gender or to collect more of the underrepresented sex, respectively.
In conclusion, this work adds to the increasing list of available LAMP primers for various avian species and is a new addition within the Passeriformes order.

ACK N OWLED G M ENTS
We thank Javier Pérez-Tris, Juan Carlos Illera, Kira Delmore, Gillian Durieux, Juan S. Lugo Ramos, Hannah C. Justen, Susanne Reinsch for providing samples from the field to test this protocol on juvenile birds. Juan Carlos Illera triggered the idea to develop a LAMP protocol for the blackcap, so he deserves special thanks here.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest. wrote the manuscript. All authors contributed critically to the drafts and gave final approval for publication.

DATA ACCE SS I B I LIT Y
All relevant data for this study are included in and accessible through this manuscript.