Evaluation of the Bio-Evolution Microsporidia generic and typing real-time PCR assays for the diagnosis of intestinal microsporidiosis

Cases of intestinal microsporidiosis infection are underestimated and affect both immunocompromized and immunocompetent patients. Real-time PCR is superseding microscopic examination for its diagnosis in medical analysis laboratories. However, few manufacturers include microsporidia in their PCR panel for the diagnosis of infectious gastroenteritis. Here, we evaluated the performances of the real-time PCR assays microsporidia generic and microsporidia typing (Bio-Evolution, France) on the Rotor-Gene Q real-time PCR cycler (Qiagen, France). We included 45 negative and 44 positive stool samples for Enterocytozoon bieneusi (n = 34, with various genotypes), Encephalitozoon intestinalis (n = 4), Encephalitozoon hellem (n = 4), and Encephalitozoon cuniculi (n = 2). We also studied a four-year survey of an inter-laboratory quality control program including 9 centers that used this commercial assay. Sensitivity and specificity of the microsporidia generic assay were 86.4% and 93.3%, respectively. Encephalitozoon hellem and Encephalitozoon cuniculi were detected by the microsporidia generic PCR assay but not by the microsporidia typing PCR assay. These results were consistent with the results of the inter-laboratory quality control program. In conclusion, Bio-Evolution Real-time PCR assays are useful tools for intestinal microsporidiosis, but negative results for microsporidia typing assays require supplementary analyses to confirm E. hellem or E. cuniculi infections.


Introduction
Intestinal microsporidiosis cases are associated with severe and chronic diarrhea in immunocompromized patients [13]. They were first reported as opportunistic infections in HIVinfected patients in the 1980s and 1990s, but nowadays they are mostly diagnosed in other immunocompromized populations, such as patients with solid organ transplantation [3,5,6,8,11,15]. Enterocytozoon bieneusi accounts for more than 90% of intestinal microsporidiosis cases; the remaining cases involve Encephalitozoon intestinalis, Encephalitozoon hellem and Encephalitozoon cuniculi [13]. For a long time, the diagnosis of intestinal microsporidiosis was exclusively based on microscopy. In recent years, microscopy was superseded by real-time PCR for these diagnoses. Molecular tools allow for more sensitive diagnosis and give rapid species identification which is essential for therapeutic care, since treatment depends on the species [9,13,27]. Several in-house PCR methods have been published, targeting E. bieneusi and/or E. intestinalis [7, 17-19, 23, 29, 30, 32]. These targets are also included in ready-to-use commercial real-time PCR panels, some of which also enable detection of other gastrointestinal pathogens: Para-GENIE Cryptosporidia Microsporidia real-time PCR kit (Ademtech, France), Allplex GI-Helminth real-time PCR assay (Seegene, Germany), Novodiag Ò Stool Parasites (Mobidiag, Finland), and the combination of two real-time PCRs: microsporidia generic assay (MGa) plus microsporidia typing assay (MTa) (Bio-Evolution, France). Only the Seegene and Ademtech panels have been evaluated to date, showing good and comparable performances to in-house real-time PCR assays for microsporidiosis diagnosis [2,16,20]. Regarding the Bio-Evolution assay, it has the particularity of proceeding in 2 steps. An initial assay (MGa) allows for the screening of stool samples and detects simultaneously E. bieneusi and Encephalitozoon species, but does not discriminate between them. The positive samples are then tested with the second assay (MTa) in order to specify whether they are E. bieneusi or E. intestinalis.
The main objective of this study was to assess performances of both Bio-Evolution Microsporidia generic and typing PCR assays with (i) microsporidia-negative and -positive stools, and (ii) a national interlaboratory quality control program.

Clinical samples
A total of 89 DNA extracts from fecal specimen were used in this study. For E. cuniculi, spores purified from an in vitro culture were spiked into a negative stool. Among them, 44 positive samples for microsporidia were obtained from the French Microsporidiosis Network, i.e., initial diagnosis was made in each collaborative center by microscopy and/or real-time PCR and further confirmed by in-house real-time PCR according to Moniot [19]. The different species and genotypes included are listed in Table 1. No coinfection with other pathogens was reported. For E. bieneusipositive samples, genotype was identified by sequencing the Internal Transcribed Spacer (ITS) region, as previously described [14,28]. Encephalitozoon species were also characterized by ITS sequencing. In all, 45 samples negative for microsporidia (checked by in-house real-time PCR according to Moniot et al., [19]) were collected in the Clermont-Ferrand University Hospital, among which microscopic examination was negative for 20 samples and positive with other eukaryotic microorganisms for 25 samples (for more details see Table 1).

DNA extraction
DNA extractions were performed within the next few days following sample collection (samples were stored at 4°C until DNA extraction). Two hundred mg of stool were placed in vials containing 800 lL of easyMAG lysis buffer (bioMérieux, France) and 100 lL of 0.5 mm glass beads (Next Advance, France). Then, a bead beating step was performed for 3 min at 3000 Hz (TissueLyser, Qiagen, France), followed by centrifugation for 10 min at 20,000 Âg. Then, DNA from 200 lL of the supernatant was extracted using a QIAamp Fast DNA Stool Mini Kit (Qiagen, France). DNA extracts were aliquoted to prevent several cycles of freezing/thawing and stored at À80°C during the study (i.e., three years maximum).

Real-time PCR
The Bio-Evolution Microsporidia generic assay (MGa) and microsporidia typing assay (MTa) were performed on the Rotor-Gene Q thermocycler (Qiagen), according to the manufacturer's instructions. For both MGa and MTa, the primers and probes target the 18S rRNA encoding gene. For MGa, the amplification consisted in an initial denaturation step of 15 min at 95°C, followed by 40 cycles of denaturation at 95°C for 15 s, and annealing/elongation at 62°C for 1 min. For MTa, the amplification consisted in an initial denaturation step of 15 min at 95°C, followed by 45 cycles of denaturation at 95°C for 15 s, and annealing/elongation at 60°C for 1 min. MGa includes an internal control for assessing the presence of PCR inhibitors, MTa does not. Sensitivity, specificity, and predictive positive and negative values claimed by the manufacturer are 88%, 83%, 96% and 63% for MGa, and 86% (E. bieneusi)/100% (E. intestinalis), 100% (both E. bieneusi and E. intestinalis), 100% (both E. bieneusi and E. intestinalis) and 63% for MTa, respectively. PCR results were given as positive or negative. As recommended by the manufacturer, results were considered positive when the Ct value was less than or equal to 38 for MGa and 40 (or between 40 and 45 twice consecutively) for MTa. For MGa-positive samples, Ct values were also recorded. Positive samples for MGa were submitted to MTa, according to the manufacturer's recommendations. A positive MTa identifies both E. bieneusi and E. intestinalis in a sample. A negative MTa should be interpreted as presence of microsporidia DNA other than E. bieneusi or E. intestinalis.

Interlaboratory quality control program
Each year, DNA samples from positive or negative microsporidia stool samples obtained by the same protocol mentioned before were sent at À20°C to several medical analysis laboratories in France participating in a national external quality assessment, proposed by the Parasitology-Mycology unit of the Clermont-Ferrand University Hospital (France, coordinating center of the French Microsporidiosis Network). Four years of this program (2018-2021) were analyzed with laboratories using the Bio-Evolution PCR assays; this corresponded to nine samples and nine institutions.

Results analysis and statistical considerations
Statistical analyses were performed with Excel software 2010. For PCR assay performances, sensitivity, specificity, and positive and negative predictive values were calculated considering qualitative results of MGa and MTa. For positive samples, MGa quantitative results (Ct values) were analyzed. Each discordant result was discussed regarding the Ct values and the species/genotype implicated.

PCR assays performances
The results of the study are presented in Figure 1. Among the 89 DNA tested, the MGa PCR assay was positive for 41 and negative for 48 samples. Three samples with Blastocystis sp. had a positive result with MGa. However, they were considered false positives because both in-house real-time PCR and end-point multiplex PCR (targeting E. bieneusi and Encephalitozoon spp. ITS), and microscopy examination, performed on  Table 1). Of the 41 MGa-positive samples, 33 were successfully identified with MTa including three E. intestinalis and 29 E. bieneusi belonging to various genotypes (A, C, CAF-1, C-like01, C-like02, C-like03, D, HND-I, IV, Wildboar2, Wild-boar3, WR5-like01). As expected, MTa did not detect one isolate of E. cuniculi and three isolates of E. hellem. However, one E. cuniculi isolate gave a misidentification with a positive result on the E. intestinalis target. The three Blastocystis spp. that were positive with MGa were negative with MTa. The remaining five MTa-negative samples (among microsporidia positive samples) contained E. bieneusi (genotype A, n = 1), E. cuniculi (n = 1), and E. hellem (n = 3).

Discussion
Diagnosis of intestinal microsporidiosis in immunocompromized patient is increasing, especially in case of solid organ transplantation [3,4,8,11,15,19]. Nowadays, real-time PCR assay is the best method to detect microsporidiosis in stool samples [9,13,27]. Numerous in-house PCR methods have been described and some commercial kits are available [1, 2, 7, 10, 16-20, 23, 29, 30, 32]. However, some of these have not been evaluated with clinical samples until now. In the Morio et al. (2019) study, the ParaGENIE Cryptosporidia Microsporidia real-time PCR kit showed sensitivity and specificity of 97.3% and 98.7% for E. bieneusi diagnosis, respectively [20]. Hence, the main objective of this study was to evaluate the performances of the two PCR assays commercialized by Bio-Evolution (MGa and MTa) with a panel of 44 microsporidia-positive DNA from clinical stool samples collected through the French Microsporidiosis Network. Of note, this study evaluates only the amplification step and not the extraction step which is known to be fundamental [33]. More generally, performances of molecular diagnostic techniques for intestinal parasites are highly dependent on stool sample pretreatment and the method used for DNA extraction [22]. Nevertheless, the manufacturer's instructions did not include recommendations for these steps. One limitation when interpreting MGa and MTa performances for Encephalitozoon species was the difficulty in collecting enough positive samples for that genus. Because microsporidia are closely related to fungi, we recognize that it would have been relevant to test more fungal species during the cross-reactivity experiments. This will have to be addressed more deeply in future studies on molecular tools for the diagnosis of intestinal microsporidiosis. Overall sensitivity and specificity of MGa were 86.4% and 93.3%, respectively (Table 2 and Fig. 1). The absence of detection of some positive samples by MGa can be explained by a low load of microsporidia-DNA, as shown by the national interlaboratory quality control program. The "2018-DNA3" E. intestinalis-positive and the "2021-DNA1" E. bieneusi-positive samples had a low target amount, which could explain a poorer detection rate among participating laboratories (Fig. 2). Concerning the DNA load, Ct values varied greatly between laboratories for some samples, which can not only be explained by inter-operator variability. We hypothesized that the variability could be due to the thermocycler used, but this was not evaluated in our work because some centers did not specify the thermocycler used. For E. bieneusi-positive samples, it can also be hypothesized that sensitivity could depend on ITS genotypes. Of note, MGa targets the small subunit ribosomal RNA just before the ITS region. However, E. bieneusi discordant samples showed different genotypes (A, IV, C and Wild-boar3). Moreover, other positive samples with these genotypes were successfully detected in our study (Supplementary Table 1). The lack of sensitivity seems to be correlated with low microsporidia-DNA quantity rather than a particular genotype of E. bieneusi, but the number of discordant samples is not sufficient to confirm this hypothesis. Interestingly, in some cases we observed discordances between MGa and MTa results. MGa detected E. cuniculi and E. hellem samples, but also some Blastocystis sp. samples. On the one hand, the PCR assay handbook clearly mentions that Blastocystis spp. may result in a flat curve with MGa, but are associated with an MTa-negative result, allowing microsporidiosis exclusion. On the other hand, E. cuniculi and E. hellem were not detected by MTa, which does not target these two species. These findings were confirmed by the qualitative results of the national interlaboratory quality control program. However, E. cuniculi and E. hellem are also responsible for intestinal microsporidiosis, even more frequently than E. intestinalis in some parts of the world [12,21,[24][25][26]31]. So, a positive MGa with an exponential curve associated with a negative MTa requires another confirmation assay to rule out E. cuniculi or E. hellem infections. This issue may also exist with other commercial kits: the ParaGENIE Cryptosporidia Microsporidia real-time PCR kit is able to differentiate between E. bieneusi and E. intestinalis, but there are no data about other Encephalitozoon species, and the Allplex GI-Helminth real-time PCR assay detects E. bieneusi and Encephalitozoon species without differentiation, which is inadequate for species-specific treatment (i. e., fumagillin for E. bieneusi and albendazole for Encephalitozoon spp. infections) [2,13,16,20]. Detection of and differentiation between E. bieneusi and Encephalitozoon spp. are possible with Novodiag Ò Stool Parasites, although no evaluation has yet been published.
In conclusion, the Bio-Evolution Microsporidia generic and typing assays show good performances for E. bieneusi or E. intestinalis intestinal microsporidiosis diagnosis, as shown by our study in Clermont-Ferrand and the inter-laboratory quality control program. The two-step process is time consuming, but enables us to differentiate between E. bieneusi and E. intestinalis for specific treatment. Nevertheless, in case of discordant MGa/MTa results, additional analyses should be considered to detect/exclude other Encephalitozoon species.