Development and comparative evaluation of TaqMan real-time PCR assays for detecting Enterocytozoon hepatopenaei (EHP) in penaeid shrimp

Abstract

Enterocytozoon hepatopenaei (EHP) is a microsporidian that retards shrimp growth and threatens global aquaculture, making reliable, sensitive detection essential for surveillance and control. Herein, we compared six TaqMan real-time PCR assays targeting distinct EHP loci, including three published assays (SSU rDNA, β-tubulin, and cysteine desulfurase/NFS1), two unpublished internal Aquaculture Pathology Laboratory assays (DNA polymerase and ATP/ADP translocase), and a newly designed spore wall protein (SWP) assay. Assay development and validation followed the World Organisation for Animal Health (WOAH) framework, assessing analytical specificity, analytical sensitivity/limit of detection (LOD), repeatability, and reproducibility using defined plasmid DNA standards. All assays were specific for EHP, with no cross-reactivity to other crustacean pathogens causing enteric disease (e.g., NHP, AHPND, TPD) or systemic infections (e.g., IHHNV, WSSV nor with the crustacean microsporidian Perezia sp. or a novel intranuclear Enterocytozoon sp. The only exception was the SSU rDNA assay, which also amplified the novel intranuclear Enterocytozoon sp. All assays achieved an LOD of 10 copies/µL. Repeatability was strong (%CV < 1.65%) and reproducibility good (%CV < 5.70%); however, qPCR efficiency fell outside the 90–110% range for the β-tubulin and ATP/ADP translocase assays in repeatability tests, and for the β-tubulin, NFS1, and ATP/ADP translocase assays in reproducibility tests. These three assays also showed relatively lower reproducibility, with occasional reductions in linearity (R² < 0.99). Screening of experimentally challenged shrimp with a wide dynamic infection range yielded 100% diagnostic specificity (95% CI 69.15–100) for all assays. Diagnostic sensitivity was lower for the β-tubulin (90.00%; 95% CI 73.47–97.89) and ATP/ADP translocase (80.00%; 95% CI 61.43–92.29) assays, whereas the SSU rDNA, NFS1, DNA polymerase, and SWP assays each reached 100% sensitivity (95% CI 88.43–100). The SSU rDNA and SWP assays were the most sensitive at low infection levels but SSU rDNA may cross-react with related microsporidia. By targeting an EHP-unique region, the SWP assay currently provides higher specificity and strengthen EHP surveillance to help diminish its global spread, as EHP is listed by WOAH as an emerging pathogen.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-52052-2.

Introduction

Enterocytozoon hepatopenaei (EHP) is a microsporidian parasite that has become a major constraint on penaeid shrimp aquaculture by causing chronic growth retardation and cumulative production losses rather than acute mortality¹. The parasite was first officially reported from farmed Penaeus monodon in Thailand and was later formally described and named Enterocytozoon hepatopenaei^2,3. Since its description, EHP has disseminated to P. vannamei production systems across Asia and, more recently, has been detected in the Americas, underscoring its transboundary significance^4,5. Infection with EHP results stunted growth, poor feed conversion, prolonged culture cycles, and reduced harvest yields^3,6,7. Moreover, several studies suggest that prior EHP infection can predispose P. vannamei to opportunistic or enteric pathogens, further intensifying production losses^5,8. Given clinical signs are subtle and non-specific, and co-infections with bacterial or viral agents are common, reliable laboratory diagnosis is essential to support evidence-based farm management, movement control, and surveillance at hatchery, farm, and regional levels⁹.

EHP detection in shrimp relies on molecular and histopathologic approaches¹. Histologically, EHP infects hepatopancreatic tubular epithelial cells, where intracytoplasmic spores and developmental stages can be observed^3,6,9. Although these lesions are characteristic, they can be focal or patchy, so histopathology alone may fail to detect light or early infections. A wide range of molecular assays for EHP detection has therefore been developed, including conventional PCR^10,11, nested PCR^12,13, quantitative PCR (qPCR)^14,15, loop-mediated isothermal amplification (LAMP)^16,17, and real-time recombinase polymerase amplification (RPA)^18,19. These methods provide faster turnaround than histopathology and are now widely used for EHP surveillance and hatchery screening. Molecular targets include the multicopy SSU rDNA gene^10,11,15, single- or low-copy protein-coding loci such as β-tubulin^13,20 and cysteine desulfurase (NFS1)²¹, as well as spore wall protein (SWP) genes^12,16,18. Several of these methods, such as conventional PCR, nested PCR, LAMP, RPA yield qualitative rather than quantitative outputs²². In contrast, TaqMan real-time PCR (TaqMan qPCR) has emerged as the workhorse for EHP detection, offering high analytical sensitivity, specificity, and quantitative results that guide interventions and biosecurity^14,20.

Although EHP qPCR is widely used in EHP diagnosis, a rigorous head-to-head comparison of assays targeting different genes has not been reported. This is very much needed now especially considering that the WOAH has listed EHP as an emerging pathogen and a validated assay for its detection is required for global surveillance¹. Generally, differences in assay performance often arise from interactions among primers and probe designed for a given assay, nucleic-acid extraction method employed, sample matrix tested, and thermocycler platform used for amplification. Collectively, these parameters also affect limit of detection (LOD), amplification efficiency, and quantitative accuracy^23,24. These factors are most consequential when copy-number of the target pathogen is very low and near decision thresholds. It is for these reasons that precision (repeatability within runs and reproducibility across runs, operators, and instruments) and efficiency of a given assay must be controlled to prevent biased quantification and misclassification^25,26. Accordingly, assay selection for routine screening versus confirmatory testing should rely on studies that compare targets under uniform conditions and report effect sizes with uncertainty²⁷. International guidance from the WOAH recommends a structured validation pathway that distinguishes analytical characteristics (specificity, sensitivity/LOD, precision) from diagnostic performance in field specimens across realistic infection gradients, with transparent reporting of confidence intervals, controls, and acceptance criteria to support inter-laboratory comparability²⁸. Adherence to these principles, together with Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) best practices for qPCR design and reporting, strengthens claims about assay fitness for surveillance and regulatory decision-making. This facilitates proficiency testing and procurement across government, industry, and academic laboratories^25,26,28.

Here, we address existing gaps in EHP detection by developing and comparatively evaluating six TaqMan qPCR assays spanning a representative cross-section of EHP genomic targets. Three assays were sourced from the literature targeting three different EHP genes (i.e., SSU rDNA, β-tubulin, NFS1), two were derived from an internal Aquaculture Pathology Laboratory (APL) panel (DNA polymerase and ATP/ADP translocase), and the sixth assay was designed based on SWP gene. Following the WOAH guidelines for assay development²⁸, we: (i) establish analytical specificity against common non-target shrimp pathogens encountered in diagnostic workflows (e.g., NHP, IHHNV, WSSV and AHPND-causing Vibrio parahaemolyticus), (ii) estimate LODs using plasmid DNA standards, and (iii) quantify precision via intra-assay repeatability and inter-assay reproducibility. We then assess diagnostic performance using shrimp specimens spanning a wide dynamic range of EHP loads to reflect field conditions. These comparative data are intended to guide target selection, optimize assay panels, and support inter-laboratory harmonization for EHP surveillance worldwide.

Materials and methods

Assay selection and qPCR procedure

Six TaqMan qPCR assays were evaluated based on: three published targets (SSU rDNA, β-tubulin, and NFS1), two from the APL panel (DNApol and AAC), and one additional assay based on SWP gene. Primers and probes were designed from reference EHP gene sequences deposited in GenBank (Table 1). For the published assays, primer–probe sets were used exactly as originally reported. For the remaining three assays, primers and probes were designed in Primer Express™ Software v3.0.1 (Thermo Fisher Scientific, USA). Oligonucleotide sequences, target loci, amplicon sizes, and qPCR cycling conditions are summarized in Table 1; an overview of primer and probe locations with respect to the EHP genome sequence is shown in Fig. S1.

Table 1.

Primers and probes used in the TaqMan qPCR assays to detect EHP.

	Target genes	GenBank Accession no	Primers/probe	Sequence (5′-3′)	Amplicon (bp)	Cycling condition (temp, time), 40 cycles			References
						Initial denaturation	Denaturation	Annealing/extension
1	SSU rDNA (18S rRNA)	KF362129	157-F	AGTAAACTATGCCGACAA	157	95 °C, 30 s	95 °C, 5 s	60 °C, 30 s	15
			157-R	AATTAAGCAGCACAATCC
			SSU-Probe	FAM-TCCTGGTAGTGTCCTTCCGT-TAMRA
2	β-tubulin	KY593130	bTub-F2	GATTTGAGAAAATTGGCGGTTAA	84	95 °C, 20 s	95 °C, 3 s	60 °C, 30 s	20
			bTub-R2	TTCTGAACACAGAGGCGCATA
			bTub-Probe	FAM-TGATTCCATTTCCACGACTGCACTTTTTC-TAMRA
3	Cysteine desulfurase (NFS1)	MNPJ01000001.1	NFS177F	GCAAGGAACACAACATAG	177	95 °C, 5 min	95 °C, 30 s	58 °C, 30 s /72 °C, 20 s	21
			NFS177R	GCTTGGTTATCCTTACAC
			NFS-Probe	FAM-TTCATACAGACGCCGCACAGG-BHQ1
4	DNA polymerase (DNApol)	MNPJ01000016.1	DNApol-qF	CCAGTTGCGCTTGAGTTTGA	69	95 °C, 20 s	95 °C, 3 s	60 °C, 30 s	APL panel
			DNApol-qR	GCCCGCATATCTTTTTTTATTCA
			DNApol-Probe	FAM-AAGGTATATTCTCCATATTTGCT-TAMRA
5	ATP/ADP translocase (AAC)	MNPJ01000024.1	AACP4-qF	AATGATACCGTATAGAAGAATCAACATCA	101	95 °C, 20 s	95 °C, 1 s	60 °C, 20 s	APL panel
			AACP4-qR	AATTGACAGTGAAGCAAATCAACAA
			AACP4-Probe	FAM-AACACCCAGGAAGTTGATGCTACCGAAAA-TAMRA
6	Spore wall protein (SWP)	KX258197.1	SWP-qF	TTGGCGGCACAATTCTCA	103	95 °C, 20 s	95 °C, 3 s	60 °C, 30 s	New (this study)
			SWP-qR	GCTGTGTCTGTGTAAATATCGTCTCTT
			SWP-Probe	FAM-ACATTTTCACCATTGGTC-TAMRA

All assays were run in 10 µL reactions containing 2.5 µL TaqMan Fast Virus 1-Step Master Mix (Thermo Fisher Scientific, USA), 1.0 µL of primers (0.5 µM each), 0.5 µL of probe (0.2 µM), 1.0 µL of template (plasmid or DNA), and nuclease-free water brought to 10 µL. This composition was used for all assays throughout the study. Amplification, detection, and data analysis were performed on a StepOnePlus Real-Time PCR System (Applied Biosystems™).

In silico oligo structure and thermodynamics

Predicted secondary structures and thermodynamic parameters for all primers and probes were assessed in Geneious Prime® (Oligo Analysis). Hairpins, self-/hetero-dimers (including 3′-end interactions), melting temperature (Tₘ), GC%, and Gibbs free energy (ΔG) were calculated under reaction-matched settings.

DNA extraction

Genomic DNA from shrimp tissues was extracted using the DNeasy Blood & Tissue Kit (QIAGEN, Germany) according to the manufacturer’s instructions. DNA concentration was measured with a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific, USA), and aliquots were stored at − 20 °C. All genomic DNA samples were normalized to 10 ng/µL prior to use.

Analytical specificity comparison

Analytical specificity was assessed by screening each assay against genomic DNA extracted from shrimp tissues representing target and non-target infections. The template panel included DNA from Specific Pathogen-Free (SPF) P. vannamei (negative control) and from P. vannamei experimentally infected with EHP (positive target). To assess specificity, we also screened DNA from shrimp infected with common enteric and systemic pathogens, including Hepatobacter penaei (NHP), decapod penstylhamaparvovirus 1 (IHHNV), white spot syndrome virus (WSSV), and V. parahaemolyticus strains causing acute hepatopancreatic necrosis disease (V_pAHPND). In addition, the panel included shrimp infected with a V. parahaemolyticus strain causing translucent post-larvae disease (V_pTPD)²⁹, the crustacean microsporidian Perezia sp. (cotton shrimp disease)³⁰, and a novel intranuclear Enterocytozoon sp³¹. A no-template control (nuclease-free water) was included in each run. PCR compatibility and DNA quality were verified by a TaqMan assay targeting the shrimp EF-1α gene³².

Construction of plasmid standards

Synthetic DNA fragments (gBlocks; Integrated DNA Technologies, USA) flanking the binding regions of each primer–probe set were custom-synthesized. Target fragments were amplified using the corresponding primer pairs for each assay (Table 1), ligated into the pCR®2.1-TOPO® vector (Thermo Fisher Scientific, USA), and transformed into TOP10 Escherichia coli competent cells. Plasmid DNA was purified from positive clones; concentrations were measured and copy number (copies/µL) was calculated using the following formula:

Ten-fold serial dilutions of each recombinant plasmid, from 10⁶ to 1 copy/µL, were prepared and used as qPCR templates.

Comparison of analytical sensitivity and stability

Analytical sensitivity was assessed using ten-fold serial dilutions of plasmid standards prepared in section above (10⁶ to 1 copy/µL). For each assay, standard curves were generated by plotting mean cycle threshold (C_t) values against the base-10 logarithm of input copy number on a StepOnePlus Real-Time PCR System (Applied Biosystems™). The coefficient of determination (R²) and amplification efficiency (Eff, %) were derived from the linear-regression slope. Sensitivity testing was performed in triplicate at each dilution, and the LOD was defined as the lowest dilution at which all three replicates yielded positive amplification.

Assay stability was evaluated as precision, and partitioned into repeatability (intra-assay) and reproducibility (inter-assay). Repeatability was estimated from triplicate reactions at each dilution within a single run; reproducibility was estimated from three independent runs performed on three separate days, with triplicate reactions at each dilution in each run. Precision was summarized as the coefficient of variation (CV, %) of C_t values at each dilution for both repeatability and reproducibility.

Evaluation of practical detection

Biological sample preparation

An experimental bioassay was conducted to generate a panel of clinical specimens for evaluating the practical performance of the assays. SPF P. vannamei (5.0 ± 0.5 g) were obtained from Oceanic Institute (Oahu, Hawaii, USA) and used in the trial conducted at the West Campus Aquaculture Center, Aquaculture Pathology Laboratory, The University of Arizona. All shrimp were certified SPF and had consistently tested negative for WOAH-listed and non-listed pathogens, including EHP. As P. vannamei are invertebrates, this experimental work did not require Institutional Animal Care and Use Committee (IACUC) approval, and all procedures were performed in accordance with institutional guidelines for the care and use of aquatic invertebrates. Shrimp were maintained in aerated 200-L artificial seawater tanks at 30–32 ppt salinity and 26–28 °C and were subsequently allocated to control and challenge groups. Hepatopancreas (HP) tissue from shrimp previously confirmed EHP-positive was pooled and homogenized under sterile conditions to produce the parasite inoculum. Control animals received only standard commercial feed, whereas challenged animals received feed mixed with the EHP-positive HP homogenate. Ten shrimp were randomly sampled at three post-exposure time points, including Day 15, Day 30, and Day 45 (n = 10 per time point; total n = 30). Negative controls comprised ten SPF shrimp sampled at Day 45 (n = 10). In total, 40 specimens were obtained, spanning a broad range of anticipated infection loads and statuses.

Diagnostic performance

Genomic DNA was extracted from the HP of each shrimp, and every specimen was assigned a unique code at collection. Coded specimens from all-time points and groups were combined to form a single blind panel of 40 samples. This blind panel, comprising varying infection statuses and loads arising from time-staggered exposure was used to compare the diagnostic accuracy of the candidate qPCR assays, including diagnostic sensitivity (DSe) and diagnostic specificity (DSp), under uniform nucleic-acid extraction conditions.

Statistical analysis

Data were acquired on a StepOnePlus Real-Time PCR System (Applied Biosystems™). Data were managed in Microsoft Excel (Microsoft, USA), and graphs were generated in GraphPad Prism (GraphPad Software LLC, UK). Results are reported as mean ± standard deviation (SD). Intra- and inter-assay variability were calculated from C_t values and expressed as coefficients of variation (CV, %). DSe and DSp for each primer–probe set were estimated in MedCalc v23.1.7 (MedCalc Software Ltd., Belgium) with 95% confidence intervals (CI).

Results

Predicted secondary structures and thermodynamic parameters

Predicted secondary structures and Gibbs free energies (ΔG, kcal·mol⁻¹) for all forward/reverse primers and probes are shown in Fig. 1. Hairpin structures were present in the 18S rRNA probe; β-tubulin forward (F) and reverse (R); NFS1 reverse; and AAC forward, reverse, and probe. All remaining oligonucleotides appeared unstructured (no discernible hairpins). This visual assessment accords with ΔG magnitudes: hairpin-bearing oligos showed more negative ΔG (e.g., β-tubulin R =  − 2.38; AAC ≈ − 1.10 to − 1.27; 18S rRNA probe =  − 0.98; NFS1 R =  − 0.67), whereas unstructured conformers were near zero (e.g., 18S rRNA R =  − 0.08; DNApol R =  − 0.05; NFS1 F = 0.00). The most stable predicted self-structure was the β-tubulin reverse primer (ΔG =  − 2.38), followed by the AAC reverse primer (ΔG =  − 1.27), the β-tubulin forward primer (ΔG =  − 1.25), the AAC probe (ΔG =  − 1.23), and the AAC forward primer (ΔG =  − 1.12).

Fig. 1.

Predicted secondary structures and thermodynamic parameters of primers and probes used in TaqMan qPCR assays to detect EHP. The Gibbs free energy (ΔG) values are shown next to each oligonucleotide.

Evaluation of specificity

All candidate TaqMan qPCR assays amplified EHP-positive controls as expected. No amplification was observed in SPF shrimp or in the non-target pathogen panel (i.e., NHP, IHHNV, WSSV, AHPND, TPD, Perezia sp., and the novel intranuclear Enterocytozoon sp.; Fig. S2). The only exception was the 18S rRNA assay, which also amplified the novel intranuclear Enterocytozoon sp.

Comparison of sensitivity, repeatability, and reproducibility

All assays achieved an LOD of 10 copies per reaction; single-copy reactions were not detected (Table 2). Intra-assay standard curves were highly linear (R² = 0.990–0.998; Fig. S3). Amplification efficiencies were within the 90–110% acceptance range for 18S rRNA (102.66%), NFS1 (96.60%), DNApol (101.53%), and SWP (104.27%), but were elevated for β-tubulin (123.69%) and AAC (113.09%) (Table 2). Inter-assay analyses showed occasional reductions in linearity (R² < 0.99) for β-tubulin, NFS1, and AAC, while preserving the intra-assay efficiency pattern (Table 3): β-tubulin remained > 120% (123.69%, 123.67%, 128.89%), AAC varied from acceptable to high (97.71%–118.92%), and 18S rRNA, NFS1, DNApol, and SWP stayed near target (~ 95–106%).

Table 2.

Evaluation of detection repeatability (intra-assay variation) of six TaqMan qPCR assays in detecting EHP.

Assay	Mean Ct ± SD*							LOD (copies)	R2	Efficiency (%)
	106	105	104	103	102	101	100
18S rRNA	18.19 ± 0.04	21.78 ± 0.11	25.36 ± 0.08	28.50 ± 0.36	32.38 ± 0.53	35.25 ± 0.57	ND	10	0.990	102.66
β-tubulin	19.77 ± 0.13	23.23 ± 0.07	25.99 ± 0.08	29.51 ± 0.03	32.17 ± 0.22	33.95 ± 0.20	ND	10	0.992	123.69
NFS1	17.51 ± 0.10	20.97 ± 0.12	24.52 ± 0.18	27.87 ± 0.27	31.91 ± 0.06	34.62 ± 0.51	ND	10	0.996	96.60
DNApol	17.13 ± 0.09	20.41 ± 0.16	24.11 ± 0.16	27.25 ± 0.29	30.86 ± 0.36	33.85 ± 0.35	ND	10	0.996	101.53
AAC	19.27 ± 0.03	22.71 ± 0.06	25.86 ± 0.06	28.87 ± 0.08	32.01 ± 0.05	35.18 ± 0.23	ND	10	0.996	113.09
SWP	18.03 ± 0.07	21.34 ± 0.02	24.58 ± 0.05	27.91 ± 0.06	31.46 ± 0.22	34.42 ± 0.25	ND	10	0.998	104.27

*triple replicates of each dilution. ND, Not detected.

Table 3.

Evaluation of reproducibility (inter-assay variation) of six TaqMan qPCR assays in detecting EHP.

Assay	Mean Ct ± SD*							LOD (copies)	R2a	Efficiencyb (%)
	106	105	104	103	102	101	100
18S rRNA	17.51 ± 0.68	21.07 ± 0.68	24.43 ± 0.84	28.09 ± 0.45	31.78 ± 0.52	35.17 ± 0.47	ND	10	0.990 0.995 0.998	102.66 92.57 94.08
β-tubulin	19.83 ± 1.13	23.30 ± 0.97	26.13 ± 1.16	29.86 ± 1.36	32.63 ± 1.37	34.05 ± 1.07	ND	10	0.992 0.973 0.980	123.69 123.67 128.89
NFS1	18.53 ± 0.88	22.07 ± 0.96	25.83 ± 1.15	29.00 ± 0.98	32.38 ± 0.50	34.87 ± 0.41	ND	10	0.996 0.986 0.987	96.60 106.26 111.42
DNApol	17.79 ± 0.85	21.14 ± 0.81	24.97 ± 0.80	27.96 ± 1.05	31.33 ± 0.67	34.94 ± 0.94	ND	10	0.996 0.998 0.991	101.53 98.79 96.03
AAC	19.23 ± 0.87	22.37 ± 1.26	25.89 ± 0.86	29.15 ± 0.84	32.53 ± 0.99	35.24 ± 0.14	ND	10	0.996 0.992 0.982	113.09 97.71 118.92
SWP	18.03 ± 0.28	21.69 ± 0.46	25.02 ± 0.68	28.30 ± 0.48	31.66 ± 0.43	35.00 ± 0.78	ND	10	0.998 0.993 0.998	104.27 97.29 105.18

*triple replicates of the intra-assay.

^a,b value for each intra-assay (three independent runs). ND, Not detected.

Repeatability (intra-assay) was strong, with SDs of mean C_t values of ~ 0.03–0.57 cycles across inputs and a modest increase at 10 copies (Table 2). Reproducibility (inter-assay) across three runs was likewise high, overall SDs of mean C_t values ranged from ~ 0.14–1.37 cycles, with per-assay ranges of SWP (0.28–0.78), 18S rDNA (0.45–0.84), DNApol (0.67–1.05), NFS1 (0.41–1.15), AAC (0.14–1.26), and β-tubulin (0.97–1.37) (Table 3). Median intra-assay CVs (Table 4) were 0.24% (AAC), 0.30% (SWP), 0.45% (β-tubulin), 0.65% (NFS1), 0.89% (18S rRNA), and 0.93% (DNApol). Median inter-assay CVs were 1.91% (SWP; lowest), 2.43% (18S rRNA), 3.18% (AAC), 3.48% (DNApol), 3.86% (NFS1), and 4.32% (β-tubulin; highest), mirroring the SDs of mean C_t ranking, lowest dispersion for SWP/18S rRNA, intermediate for DNApol/NFS1/AAC, and highest for β-tubulin.

Table 4.

Coefficients of variation (CV, %) of six TaqMan qPCR assays in detecting EHP.

Assay	Repeatability (Intra-assay)							Reproducibility (Inter-assay)
	106	105	104	103	102	101	Median	106	105	104		103	102	101	Median
18S rRNA	0.21	0.51	0.32	1.27	1.65	1.63	0.89	3.88	3.23	3.44		1.60	1.64	1.34	2.43
β-tubulin	0.63	0.29	0.32	0.10	0.67	0.58	0.45	5.70	4.16	4.44		4.55	4.20	3.14	4.32
NFS1	0.56	0.57	0.72	0.98	0.19	1.48	0.65	4.75	4.35	4.45		3.38	1.54	1.18	3.86
DNApol	0.53	0.81	0.68	1.07	1.17	1.05	0.93	4.78	3.83	3.20		3.76	2.14	2.69	3.48
AAC	0.16	0.25	0.23	0.26	0.16	0.66	0.24	4.52	5.63	3.32		2.88	3.04	0.40	3.18
SWP	0.38	0.12	0.21	0.23	0.70	0.73	0.30	1.55	2.12	2.72		1.70	1.36	2.23	1.91

Diagnostic performance

All assays achieved 100% DSp (95% CI 69.15–100; Table 5). Four assays, 18S rRNA, NFS1, DNApol, and SWP, also showed 100% DSe (95% CI 88.43–100) and 100% accuracy (95% CI 91.19–100). In contrast, β-tubulin achieved a DSe of 90.0% (95% CI 73.47–97.89) and an accuracy of 92.5% (95% CI 79.61–98.43), whereas AAC achieved a DSe of 80.0% (95% CI 61.43–92.29) and an accuracy of 85.0% (95% CI 70.16–94.29). Positive predictive value was 100% for all assays (95% CI 88.43–100), reflecting the absence of false positives. Negative predictive value was 100% for 18S rRNA, NFS1, DNApol, and SWP (95% CI 69.15–100) but lower for β-tubulin (76.92%; 95% CI 53.26–90.70) and AAC (62.50%; 95% CI 44.89–77.32), indicating missed low-load infections. The C_t distributions (Fig. 2) were concordant with these indices, SPF controls were uniformly undetected under a 40-cycle cutoff, and C_t values decreased as EHP load increased (Day 15 → Day 30 → Day 45). At low load (Day 15), 18S rRNA and SWP produced earlier C_t values with tighter dispersion (medians 29.97 and 33.84, respectively), whereas β-tubulin and AAC produced higher C_t values and greater variability (medians 34.30 and 40.00/undetected among detected samples); NFS1 and DNApol had medians of 35.09 and 35.21, respectively. Notably, on Day 15, β-tubulin and AAC failed to detect 3/10 and 6/10 samples, respectively. Overall, the indices and C_t profiles indicate that 18S rRNA and SWP provided the strongest practical performance (high DSe/DSp, high NPV, and favorable low-load C_t), followed by DNApol and NFS1, whereas β-tubulin and AAC underperform near the decision threshold and failed to detect EHP when the EHP load low.

Table 5.

Comparison of diagnostic accuracy among six TaqMan qPCR assays in detecting EHP in P. vannamei.

Assay	Diagnostic specificity	Diagnostic sensitivity	Positive predictive value	Negative predictive value	Accuracy
18S rRNA	100 (69.15–100)	100 (88.43–100)	100 (88.43–100)	100 (69.15–100)	100 (91.19–100)
β-tubulin	100 (69.15–100)	90.00 (73.47–97.89)	100 (87.23–100)	76.92 (53.26–90.70)	92.50 (79.61–98.43)
NFS1	100 (69.15–100)	100 (88.43–100)	100 (88.43–100)	100 (69.15–100)	100 (91.19–100)
DNApol	100 (69.15–100)	100 (88.43–100)	100 (88.43–100)	100 (69.15–100)	100 (91.19–100)
AAC	100 (69.15–100)	80 (61.43–92.29)	100 (86.75–100)	62.50 (44.89–77.32)	85.00 (70.16–94.29)
SWP	100 (69.15–100)	100 (88.43–100)	100 (88.43–100)	100 (69.15–100)	100 (91.19–100)

Fig. 2.

Dot-plot distributions of C_t values for six TaqMan qPCR assays in detecting EHP in DNA extracts from P. vannamei tissues. For each infection-status, samples (n = 10) are shown as individual points, with medians and ranges (each dot represents one sample). Undetected reactions were assigned a C_t of 40 for visualization. SPF, specific-pathogen-free.

Discussion

This head-to-head comparison highlights shared strengths and practical differences among six TaqMan qPCR assays for EHP detection. All assays showed high analytical specificity, with no amplification from the non-target pathogen panel (i.e., NHP, IHHNV, WSSV, AHPND, TPD, Perezia sp., and a novel intranuclear Enterocytozoon sp.), supporting exclusivity against these common endemic agents under the tested conditions. The only exception was the 18S rRNA assay, which also amplified the novel intranuclear Enterocytozoon sp., indicating incomplete specificity in the presence of closely related microsporidia. Each assay achieved an analytical LOD of ~ 10 target copies per reaction, comparable to original reports, β-tubulin: ~ 10¹ copies/mg tissue²⁰, NFS1: 1.67 × 10¹ copies/µL²¹, 18S rDNA: ~ 4 × 10¹ copies/reaction¹⁵. Although single-copy inputs were not reliably amplified, all targets were consistently detected at 10 copies. Achieving reliable detection at the 1–5 copy level typically requires very large replicate numbers and probit regression to estimate the concentration yielding 95% detection probability³³. Thus, an empirical 10-copy LOD (100% hit rate in triplicate) represents strong, practice-ready performance for diagnostic qPCR under standardized laboratory conditions.

The assays diverged more substantially in amplification efficiency and reproducibility. Four assays (i.e., 18S rRNA, NFS1, DNApol, SWP) produced near-ideal standard curves (efficiency 96–105%; R² ≥ 0.99), within the accepted 90–110% range for robust qPCR performance^34,35. In contrast, the β-tubulin and AAC assays consistently yielded efficiency estimates above this range (113–128%). Efficiencies exceeding ~ 110% are generally regarded as undesirable artifacts, often reflecting slope distortion caused by nonspecific amplification, primer-dimer formation, or small deviations in serial dilution preparation and pipetting³⁴. Given that testing was performed on purified plasmid DNA, suboptimal primer design or reaction conditions are the most plausible contributors. In silico analysis supported this interpretation as the β-tubulin and AAC primer–probe sets had more pronounced predicted hairpin loops and self-dimerization, features that can disrupt primer annealing and polymerase extension kinetics. Exceptionally high amplification efficiencies are undesirable because they indicate anomalous amplification (i.e., more than a two-fold increase per cycle), which inflates copy-number estimates and increases the risk of false positives³⁶. Laboratories adopting these assays should calibrate with assay-specific standard curves across the intended dynamic range and consider re-optimization (e.g., primer concentration, amplicon length, annealing temperature). The other four assays met efficiency criteria, showed low intra- and inter-assay variation, and maintained strong linearity (R² ≈ 0.99) across ~ 6 log₁₀ dilutions, performance that supports reliable load estimation over a broad dynamic range and can aid infection stratification and disease management decisions³⁵. In precision testing, β-tubulin showed the highest between-run variability (inter-run CV ~ 4.3%), and AAC efficiency fluctuated between runs (sometimes acceptable ~ 98% and other times over 115%). In contrast, SWP showed the lowest between-run variation (median CV ~ 1.9%), indicating highly consistent C_t values across days/runs, an advantageous trait for routine screening, proficiency testing, and international ring trials³⁷. In line with MIQE guidance, assays with poor efficiency or reproducibility should be redesigned or revalidated to avoid biased quantification^25,26. Despite efficiency issues, β-tubulin and AAC still detected ≥ 10-copy plasmids without false negatives; the trade-off was reduced quantitative accuracy and modestly lower diagnostic sensitivity at very low infection levels.

In shrimp specimens spanning early through advanced infection stages, four assays (i.e., 18S rRNA, NFS1, DNApol, and SWP) achieved 100% DSe and 100% DSp in our blind panel, detecting all panel-classified positives and yielding no false positives among SPF negatives. In contrast, β-tubulin missed 3 of 30 positives (DSe ≈ 90%) and AAC missed 6 of 30 (DSe ≈ 80%), although both assays maintained 100% DSp (no false positives). All missed detections occurred in Day-15 specimens with low parasite burden, consistent with the higher C_t values and greater variance observed for β-tubulin and AAC near the decision threshold. This pattern is biologically plausible, as SSU rDNA is typically multi-copy in microsporidian genomes, enhancing detectability during early infection^38,39. Although SWP is typically single-copy, its transcripts and proteins increase during sporogony and accumulate at the maturing spore wall, yielding a robust signal in mature spores and boosting detectability once sporulation is underway^40,41. In contrast, β-tubulin, NFS1, and DNA polymerase are typically single-copy housekeeping genes that present fewer template molecules per parasite, which can constrain sensitivity at very low burdens. Nevertheless, in our panel NFS1 and DNApol detected all positives, likely reflecting efficient primer–probe design and sufficient template input (≈10 copies per reaction). Operationally, missing 10–20% of early infections can be consequential for broodstock or post-larval screening, where even low-level infections are important to catch. We therefore recommend using an assay targeting a highly expressed, multi-copy gene (SSU rDNA) or a spore-abundant transcript target (SWP) as the primary screening tool for EHP to maximize sensitivity in early or subclinical infections. Positive cases should then be confirmed with an orthogonal assay targeting a different gene (e.g., confirm an 18S rDNA-positive result with an SWP assay) to reduce the risk of false positives and to provide complementary information (e.g., approximate spore burden). This two-step approach aligns with recent recommendations that pairing a highly sensitive screen with a second, more specific test can balance the risks of false negatives and false positives^9,42. Importantly, the performance indices reported in this study were generated using a single controlled laboratory panel and may differ in field samples with different matrices, inhibition profiles, or background microsporidia; therefore, broader multi-site clinical validation remains necessary.

Gene target choice remains an important nuance in EHP molecular diagnostics. The SSU rDNA gene is highly conserved across microsporidia and may therefore amplify non-EHP microsporidian DNA in complex matrices^12,31. Consistent with this limitation, EHP SSU rDNA shares high sequence identity with other microsporidia, which can reduce exclusivity in mixed microsporidian backgrounds¹⁴. In our dataset, the 18S rRNA assay showed high analytical sensitivity but reduced analytical specificity, as it also amplified the novel intranuclear Enterocytozoon sp. Because this target includes conserved sequences, it may also detect environmental or commensal microsporidia in settings such as polyculture ponds or natural waters, increasing the risk of misclassifying non-EHP microsporidia as EHP-positive. Moreover, because our exclusivity testing did not include a comprehensive microsporidia-focused panel, cross-reactivity with additional non-EHP microsporidia cannot be excluded. Accordingly, more divergent targets such as spore wall protein (SWP) genes are well suited for confirmatory testing and may provide a specificity advantage where other microsporidia co-occur^12,16,18,43. . In this study, the newly designed SWP TaqMan assay matched the 18S rRNA assay in analytical sensitivity and detected low parasite burdens (e.g., Day 15 post-exposure). Prior SWP-based assays have also shown improved exclusivity within the panels tested¹², and recent reviews recommend prioritizing SWP or other highly divergent loci to strengthen diagnostic specificity⁴². Taken together, our results support SWP as a strong complementary target; however, its field specificity should still be empirically verified against diverse non-EHP microsporidia where feasible.

Beyond the technical comparisons, standardization remains a broader priority. The WOAH recommends that assays undergo thorough validation (e.g., covering analytical sensitivity/specificity, precision, and diagnostic performance), before informing regulatory actions²⁸. Following a structured workflow (plasmid-based LOD, pathogen exclusivity panel, repeatability, reproducibility, and diagnostic testing on a blinded experimental specimens), we delineated each assay’s operating characteristics and limits, providing evidence-based guidance for assay selection. Assay choice should be context-dependent, where EHP is the only microsporidian of concern and rapid detection is prioritized, SSU rDNA-based assays be advantageous for screening; in settings where other microsporidia may co-occur (e.g., polyculture systems or farm-adjacent waters), an SWP-targeted assay or sequencing-based confirmation can improve interpretability by reducing the risk of cross-amplification¹². Importantly, the present exclusivity testing was limited to common shrimp pathogens available at our facility and did not include a comprehensive panel of microsporidia; therefore, the absence of amplification in our panel should not be construed as evidence of microsporidia-wide specificity. Similarly, the diagnostic indices were derived from a single controlled laboratory panel. Positive and negative predictive values are prevalence-dependent and therefore reflect this study panel rather than field prevalence. To strengthen field-relevant specificity claims, future work should test against a broader range of crustacean microsporidia (e.g., Perezia, Agmasoma, Astathelohania), newly described microsporidia species of penaeids³¹ and non-shrimp microsporidia that may occur in aquaculture environments (e.g., Enterocytozoon bieneusi, Vavraia spp.)⁴⁴. Such testing would empirically quantify the risk of SSU rDNA cross-reactivity. At present, the SWP assay retains an edge, having been derived from a sequence region unique to EHP.

In summary, not all EHP qPCR assays are equivalent in performance, and careful selection or combined use of assays can enhance detection and accuracy. All six assays reliably detected overt infections, whereas 18S rRNA and SWP provided the strongest performance at low parasite loads in this experimental panel. The NFS1 and DNApol also performed robustly and are suitable as confirmatory targets or for multiplexing (e.g., relating genome equivalents to spore load). In contrast, β-tubulin and AAC exhibited less optimal amplification kinetics and reduced sensitivity near the decision threshold; they may benefit from further optimization before adoption as primary diagnostics. Our findings support continued optimization and harmonization of molecular diagnostics for hepatopancreatic microsporidiosis caused by EHP. Careful target selection and/or paired use of complementary assays can improve early detection, reduce false results, and strengthen surveillance. As EHP continues to spread globally, WOAH-aligned assay validation and inter-laboratory harmonization are essential to ensure credible, comparable results across laboratories. The data presented here can inform inter-laboratory evaluations (proficiency panels and ring tests) and the development of standard operating procedures. Ultimately, consistent and reliable EHP diagnostics across regions will support better disease management and help safeguard shrimp aquaculture productivity.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

N.D.H conceptualized the study, performed the assay comparisons, curated the data, conducted the analyses, and drafted the manuscript. H.N.M conceptualized the study, designed the primers, performed the in silico analyses, and edited the manuscript. A.K.D conceptualized the study, conceived and secured the funding for the project, provided supervision, reviewed and edited the manuscript. All authors read and approved the final version.

Funding

The research work was funded by the USDA National Animal Disease Preparedness and Response Program (NADPRP) (Grant No. USDA-APHIS-10025-VSSP0000-24-0001) to A.K.D.

Data availability

The datasets analyzed in the present study are available in the GenBank repository under the following accession numbers: KF362129, KY593130, MNPJ01000001.1, MNPJ01000016.1, MNPJ01000024.1, and KX258197.1. Additional data supporting the findings of this study are provided within the manuscript and its supplementary information files.

Declarations

Competing interests

The authors declare no competing interests.