Primers and beacons were designed using Beacon Designer 8 (Premier Biosoft, Palo Alto, CA). Alignments of HSP90 gene sequences ( Table 1 ) were used to find universally conserved primer sequences surrounding polymorphic regions in M. chitwoodi, M. fallax, and M. minor (Skantar and Carta, 2004). To ensure the beacon probes and primers were specific to their target organisms, the sequences were queried using the BLASTn search program and the non-redundant database (National Center for Biotechnology Information, NCBI). Primers and the beacons for the HSP90 region were synthesized by Sigma-Aldrich (St. Louis, Mo) and are described in Table 2 . The M. chitwoodi beacon has 6-FAM, M. fallax beacon has HEX, and M. minor beacon has Cyan-5 as the reporters. For the quenchers, the Black Hole Quencher (BHQ-1 for 6-FAM and HEX, and BHQ- 3 for Cyan-5) was used.

Three synthetic oligos complementary to each beacon’s loop sequence and three off-target synthetic oligos containing a single nucleotide difference to the loop sequence were synthesized by Sigma-Aldrich ( Table 2 ) and used for creating thermal denaturation curves. For each beacon, two 50 µL reactions containing 200 nM molecular beacon probe in 3 mM MgCl2 and 10 mM Tris- HCl, pH = 8.0 were prepared with 600 nM of either the complementary or off-target synthetic oligo. A third reaction for each beacon was run containing Milli-Q water in place of the synthetic oligo as a control. Reactions were prepared in a 96-well clear qPCR plate and run on a CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA) starting at 80°C and decreasing to 30°C at a rate of 1°C/minute, with fluorescence measured every minute.

For genomic DNA templates, the DNA was extracted from M. chitwoodi (Race 1, Race 2, and Roza) (Bali et al., 2021b; Mojtahedi et al., 2007), M. incognita, M. javanica, M. arenaria, and M. hapla eggs using a phenol/chloroform extraction protocol as described by Gross and Williamson (2011). All three M. chitwoodi isolates (Race 1, Race 2, and Race 1 Roza) were originally provided by Dr. Charles Brown (USDA-ARS). The M. hapla isolate VW9, M. incognita isolate VW6, M. javanica isolate VW4, and M. arenaria isolate HarA were provided by Dr. Valerie Williamson ( Table 3 , UC-Davis). All Meloidogyne species were maintained on the susceptible tomato Solanum lycopersicum cv. Rutgers under greenhouse conditions. The identities of the M. chitwoodi, M. hapla, M. incognita, M. javanica and M. arenaria populations were confirmed using species-specific PCR (Powers and Harris, 1993; Zijlstra, 2000; Wishart et al., 2002).

To amplify and clone the HSP90 amplicon from M. chitwoodi race 1, M. fallax and M. minor, F- HSP90 and R-HSP90 primers were used in a PCR with approximately 15 ng of M. chitwoodi, M. fallax, or M. minor DNA. Reactions were 50 µL total volume and contained 1x AmpliTaq Gold 360 buffer, 3 mM MgCl2, 200 µM dNTPs, 200 nM F-HSP90 and R-HSP90 primers, and 1.25 U of AmpliTaq Gold. The reactions were run on an Eppendorf Mastercycler Pro Thermal Cycler as follows: 94°C for 10 minutes, then 30 cycles of 94°C for 15 seconds, then 58°C for 30 seconds, and 72°C for 30 seconds, and then finally 72°C for 5 minutes. Reactions were pooled and cleaned up using a ThermoFisher GeneJET PCR purification kit (Thermo Fisher, Waltham, MA) and the purified PCR products were ligated into pGEM-T vector. The 10 µL ligation reaction contained 1x Rapid Ligation buffer, 3 Weiss Units of T4 ligase, purified PCR product, pGEM-T vector (Promega, Madison, WI), and Milli-Q water. Plasmids were cloned into TOP10 E. coli and extracted from the overnight bacterial cultures using a GenElute Plasmid Miniprep kit (Sigma-Aldrich, St. Louis, MO). The HSP90 amplicons from M. chitwoodi, M. minor, and M. fallax in the pGEM-T vector will be referred to as plasmid HSP90.

A standard curve using M. chitwoodi gDNA as template was prepared using 1:10 serial dilutions. Reactions were run at a volume of 50 µL on a CFX96 Real-Time PCR Detection System in a 96-well clear qPCR plate. Reaction mixes contained 1x AmpliTaq Gold 360 buffer, 3 mM MgCl2, 200 µM dNTPs, 200 nM F-HSP90 and R-HSP90 primers, 200 nM target beacon (C1-Hsp90-FAM-6) and 1.25 U of AmpliTaq Gold. The cycling conditions were as follows: 95°C for 10 minutes, then 55 cycles of 95°C for 15 seconds (sec), 54°C for 30 sec, and 72°C for 15 sec. The increase in fluorescent signal was registered during the annealing step of the reaction. The M. chitwoodi standard curve used 5 points, with template DNA concentrations ranging from 40 ng to 4 pg. Additional standard curves were made for M. chitwoodi, M. fallax (F1-Hsp90-HEX-5 beacon), and M. minor (M2-Hsp90-Cyan-5 beacon) using the same reaction conditions, but in which the plasmid HSP90 templates for the reactions were serial dilutions of 0.4 ng – 4 fg, 2 ng – 2 fg, and 0.4 ng – 0.04 fg respectively. Each concentration was run in triplicate. Non-template controls containing only water were included for all standard curves, and each curve was repeated with similar results. To calculate the PCR efficiency, the following equation was used

The second stage juveniles (J2s) of M. chitwoodi race 1 were hatched from eggs collected from Rutgers tomatoes grown in greenhouses at Washington State University. Meloidogyne fallax and M. minor J2s were provided by the Wageningen Nematode Collection (National Plant Protection Organization, the Netherlands) and stored in DESS at -20°C. Second stage juveniles were digested using the protocol described in Qiu et al. (2006). In brief, J2s were picked under a dissecting scope and transferred to 15 µL droplets of Milli-Q water. The nematodes were crushed using a pipet tip; 10 µL of the crushed nematode was transferred to a 10 µL solution containing 2 µL AmpliTaq Gold 360 buffer, 2 uL of 600 µg/mL Proteinase K, and 6 µL Milli-Q water. The reaction mixes were incubated at -20°C for 20 minutes, heated for 1 hour at 65°C, and then 95°C for 10 min before cooling to room temperature. The nematode samples were spun at 12,000 rpm for 2 min before storing at -20°C until used in PCR. Five µL of these J2 samples were used as template for molecular beacon RT-PCR, and each sample was run in duplicate or triplicate. The 50 µL reaction mixes contained 1x AmpliTaq Gold 360 buffer, 3 mM MgCl2, 200 µM dNTPs, 200 nM F-HSP90 and R-HSP90 primers, 200 nM target beacon (C1-Hsp90-FAM-6, F1-Hsp90-HEX-5, or M2-Hsp90-Cyan-5) and 1.25 U of AmpliTaq Gold. The reaction conditions were the same as those used for the standard curves. All molecular beacon RT-PCR assays with these three species were repeated at least twice with similar results.

To ensure the specificity of this assay 40 ng of gDNA from M. incognita, M. javanica, M. arenaria, or M. hapla was used as template for molecular beacon RT-PCR assay using C1-Hsp90-FAM-6, F1-Hsp90-HEX-5, and M2-Hsp90-Cyan-5 beacons. Non-template controls containing only Milli- Q water were included as well as positive controls using 4 pg of plasmid HSP90 for M. chitwoodi,

M. fallax, or M. minor. The 50 µL reaction mixes contained 1x AmpliTaq Gold 360 buffer, 3 mM MgCl2, 200 µM dNTPs, 200 nM F-HSP90 and R-HSP90 primers, 200 nM target beacon (C1- Hsp90-FAM-6, F1-Hsp90-HEX-5, or M2-Hsp90-Cyan-5) and 1.25 U of AmpliTaq Gold. The reaction conditions were the same as those used for the standard curves. The molecular beacon RT-PCR assays with non-target species were repeated twice with similar results.

All PCR products in this paper were visualized as follows: 10 µL of each PCR product was loaded onto a 1.5% agarose gel and separated for 45 min at 100 V before visualizing with ethidium bromide under UV light. The Invitrogen 1-kb plus DNA ladder was used as reference for size.

Initial multiplex molecular beacon RT-PCR assays were performed using equal template concentrations of each species. Each reaction tube contained the same amount of M. chitwoodi, M. minor and M. fallax template HSP90 plasmid DNA (20 pg of each species). To the template DNAs, the reaction mix was added. The reaction mix contained 1x AmpliTaq Gold 360 buffer, 3 mM MgCl2, 200 µM dNTPs, 200 nM F-HSP90 and R-HSP90 primers, 200 nM of each of the three target beacons (C1-Hsp90-FAM-6, F1-Hsp90-HEX-5, and M2-Hsp90-Cyan-5) and 1.25 U of AmpliTaq Gold polymerase. The amount of fluorescence for each reaction at different template concentrations with the three different probes was measured at each cycle on the CFX96 Real- Time PCR Detection System.

Varying ratios of plasmid HSP90 templates were used to test the validity of multiplexing. One template was kept at 20 pg while the other two were loaded ten times more concentrated at 200 pg each. Reaction mixes contained 1x AmpliTaq Gold 360 buffer, 3 mM MgCl2, 200 µM dNTPs, 200 nM F-HSP90 and R-HSP90 primers, 200 nM of each beacon (C1-Hsp90-FAM-6, F1-Hsp90-HEX-5, M2-Hsp90-Cyan-5) and 1.25 U of AmpliTaq Gold. The reaction conditions were the same as those used for the standard curves.

Multiplex with proteinase K digested J2s were tested with 1 J2 or 5 J2 of each species present in the reaction mix. A single J2 from M. chitwoodi, M. fallax, and M. minor were picked under a dissecting microscope and placed into a single 15 µL droplet of Milli-Q water and digested as described above and used as template for qPCR. Similarly, for the 5 J2 multiplexing, 5 J2s from M. chitwoodi, M. fallax, and M. minor were picked under a dissecting microscope and placed into a single 15 µL droplet of Milli-Q water and digested and used as template for qPCR. Reaction mixes contained 1x AmpliTaq Gold 360 buffer, 3 mM MgCl2, 200 µM dNTPs, 200 nM F-HSP90 and R-HSP90 primers, 200 nM of each beacon (C1-Hsp90-FAM-6, F1-Hsp90-HEX-5, M2- Hsp90-Cyan-5) and 1.25 U of AmpliTaq Gold. The reaction conditions were the same as those used for the standard curves.

All molecular beacon RT-PCR multiplex samples were run in duplicate, or triplicate and the assays were repeated at least once with similar results.

The molecular beacon probes were based on the heat shock protein 90 (HSP90) sequence information in Genbank for M. chitwoodi, M. fallax, and M. minor ( Table 1 ). This gene was chosen because the sequence is polymorphic between the three species, and the gene had been previously used in a PCR-based nematode identification assay (Nischwitz et al., 2013; Skantar and Carta, 2004). Primers were designed that were specific for the conserved HSP90 sequence in M. chitwoodi, M. fallax, and M. minor, but span the polymorphic regions between these species within the gene ( Figure 1 ). The HSP90 gene was amplified using the genomic DNA (gDNA) of M. chitwoodi race 1, M. fallax, and M. minor as the PCR templates. Sequencing the amplicons confirmed that the predicted polymorphic regions were present in the HSP90 sequence of each species.

Molecular beacons were designed to target the polymorphic regions in each species, with the beacons hybridizing to the complementary sequences shown in Table 2 . To determine the optimal annealing temperature for each molecular beacon probe, a melting curve analysis was performed using oligonucleotides complementary to the molecular beacon probe sequence and oligonucleotides with a single nucleotide substitution ( Table 2 , Figure 2 ). The optimal annealing temperature for M. chitwoodi, M. fallax, and M. minor molecular beacon probes was 54°C.

To measure the amplification efficiency of the molecular beacon RT-PCRs, the cycle threshold (Ct) values were obtained over a range of template concentrations for all three nematode species. This information was used to generate standard curves at an annealing temperature of 54°C. First, M. chitwoodi race 1 genomic DNA was used as the PCR template with DNA concentrations ranging from 40 ng to 4 pg. When using M. chitwoodi race 1 DNA template, the molecular beacon RT-PCR assay had an efficiency of 89% and an R² = 0.9984 ( Figure 3A ). The commonly occurring M. chitwoodi Race 1 (Brown et al., 2009) was used as the template for subsequent PCRs and molecular beacon RT-PCR assays ( Figure 3A ). DNA from Race 2 and the pathotype Race 1 Roza of M. chitwoodi generated similar standard curves ( Figure S1 ), indicating that the molecular beacon RT-PCR assay could detect M. chitwoodi, including isolates of the species common in Washington (Bali et al., 2021a).

Next, a fragment of the M. chitwoodi HSP90 gene was amplified using the F-HSP90 and R-HSP90 primers. This amplicon was cloned to generate the “M. chitwoodi HSP90 plasmid,” which was used as the PCR template for further reactions. The efficiency of the molecular beacon RT-PCR using the M. chitwoodi HSP90 plasmid as the PCR template was compared to using genomic DNA as the reaction template. By calculating the slope of the standard curve, the PCR efficiency for M. chitwoodi HSP90 plasmid was 101% with an R² = 0.9980 ( Figure 3B ). This was better, but comparable to the reaction efficiencies using M. chitwoodi genomic DNA as the reaction template (efficiency of 89% and an R² = 0.9984).

Cloned HSP90 fragments from M. fallax and M. minor were also used as template for the molecular beacon RT-PCR assays. The M. minor molecular beacon RT-PCR had an efficiency of 90% and an R2 = 0.9994 ( Figure 3D ), and the M. fallax molecular beacon RT-PCR had an efficiency of 83% with an R² = 0.9984 ( Figure 3C ).

The M. chitwoodi molecular beacon probe could detect the M. chitwoodi amplicon, but not the amplicons from M. fallax and M. minor in a molecular beacon RT-PCR ( Figure S2A ). The same was true for the other molecular beacon probes; the M. fallax probe could only hybridize to the M. fallax amplicon ( Figure S2B ) and the M. minor probe could only hybridize to the M. minor amplicon ( Figure S2C ).

Meloidogyne hapla, M. incognita, M. javanica, and M. arenaria are the four major species of root- knot nematodes found worldwide. When the HSP90 sequences of M. hapla, M. incognita, M. javanica, or M. arenaria were aligned with the HSP90 sequences from M. chitwoodi, M. fallax and M. minor, the probes were specific to their respective nematode ( Figure S3 ). When qPCR was run using gDNA from M. hapla, M. incognita, M. javanica, or M. arenaria none of the molecular beacon probes produced fluorescence ( Figure S4 ), indicating that there is no off-target binding to these other root-knot nematode species.

To validate whether the molecular beacon RT-PCR assays could detect M. chitwoodi, M. fallax, or M. minor using DNA isolated directly from juveniles, crude extracts from 1 or 5 J2s of each species was used as the template for molecular beacon RT-PCR. The M. minor assay was the most sensitive with an average Ct of 35.7 (standard error of means (SEM) = 0.22, n = 7) for a single J2, and an average Ct of 32.7 for 5 J2s (SEM = 0.46, n = 2) ( Figure 4C ). The M. fallax assay was the least sensitive, with a single J2 sample having an average Ct of 43.6 (SEM = 1.16, n = 8) and 5 J2s having an average Ct of 39.8 (SEM = 1.04, n = 2) ( Figure 4B ). Finally, the molecular beacon RT- PCR for M. chitwoodi produced an average Ct for a single J2 of 38.9 (SEM = 0.64, n = 10) and a Ct of 34.7 for 5 J2s (SEM = 0.83, n = 2) ( Figure 4A ). All J2s digests were positive for all three species satisfying the 95% positive replicate criterion suggested by Forootan et al. (2017) for qPCR limit of detection. In the negative controls, no significant fluorescence was observed. These results demonstrate the sensitivity of the molecular beacon probes and their ability to detect the presence of M. chitwoodi, M. fallax, or M. minor from the crude extract of a single juvenile.

Further investigation was carried out to see if the molecular beacon RT-PCR assay could detect the nematode of interest when there is a mixture of nematodes in the reaction. When 20 pg of M. chitwoodi, M. fallax, and M. minor HSP90 templates were mixed in a single reaction containing all three beacon probes, it was possible to detect each species ( Figure 5A ). Detection of all three species was also possible using crude extracts from digested J2s when all 3 species were mixed, either as single J2s or as 5 J2s each ( Figure 6 ).

To investigate how the assay performed when templates were not present in equal parts, mixes of M. chitwoodi, M. fallax, and M. minor HSP90 plasmid templates were made that had one species present at 20 pg, and the other two species present at 200 pg. The templates present at 200 pg were reliably detected, but the template at 20 pg could not be detected in these reactions ( Figures 5B–D ).

Overall, the data suggests that the molecular beacon RT-PCR assay can detect each species in a multiplexed reaction. However, an important caveat is that the sample detection is inhibited for the target species if the other two species are in relative excess in concentration.