Comparison of fluorescent antibody test, immunohistochemistry, and PCR testing for diagnostic confirmation of neurolisteriosis in 25 goats

Abstract

Neurolisteriosis, a common disease of small ruminants, is most often caused by Listeria monocytogenes. Here we describe 25 cases of caprine neurolisteriosis diagnosed in our laboratory over a 5-y period and compare our fluorescent antibody test (FAT) results with immunohistochemistry (IHC) and polymerase chain reaction (PCR) testing for diagnostic confirmation. Neurohistologic changes consistent with neurolisteriosis affected the pons in all cases, extending rostrally to the mesencephalon in 6 cases, caudally to the medulla oblongata in 6 cases, and/or dorsally to the cerebellum in 4 cases. Acute inflammatory changes were observed in 17 cases, and included neuroparenchymal microabscesses, neuronal necrosis and neuronophagia, axonal swelling, microgliosis and astrogliosis, and perivascular neutrophils with macrophages, lymphocytes, and plasma cells that occasionally extended to the leptomeninges. Subacute-to-chronic changes (8 cases) consisted of neuroparenchymal and perivascular clusters of macrophages with rare neutrophils, lymphocytes, and plasma cells admixed with glial nodules. Bacterial bacilli were observed within neutrophils or macrophages in H&E-stained tissue sections in 4 cases. Gram stain highlighted gram-positive bacilli in 13 cases. Neurolisteriosis was confirmed by FAT in 2 cases, by IHC in 19 cases, and by PCR in 20 cases.

Listeriosis is an infectious and zoonotic disease caused most commonly by Listeria monocytogenes and less often by L. ivanovii. ¹³ Listeriosis is typically associated with systemic infection in aborted or newborn ruminants, horses, pigs, rabbits, and birds, with abortion in ruminants, and with neurologic disease in adult goats, sheep, and less often cattle. ³ Neurolisteriosis is most commonly associated with L. monocytogenes infection. ¹³ Bacteria enter the oronasal or ocular mucosal surfaces and reach the CNS after ascending migration through the cranial nerves, leading to suppurative inflammation in the brainstem that can occasionally extend cranially to the thalamus, dorsally to the cerebellum, and caudally to the cervical spinal cord. ¹² No age or sex predisposition has been observed in affected patients, and the disease is more frequent during the winter and early spring, usually associated with consumption of poor-quality silage, increased animal density, and stress during the cold season. ³ Cases can also occur in the spring and summer, particularly in Australia and South America, in which susceptible individuals have no access to silage and likely acquire infection from the environment and/or close contact with other animal species.^3,14,15

A presumptive diagnosis of neurolisteriosis can be achieved by histology with confirmation by fluorescent antibody test (FAT), bacterial culture (BC), immunohistochemistry (IHC), and/or PCR testing.^4,7,10,14,15 Although BC may be useful for better characterization of the isolated bacteria and for epidemiologic investigations of outbreaks, BC has proven difficult as a routine laboratory test as it requires special procedures that render its use problematic for a timely diagnostic confirmation of cases. ⁷ In our laboratory (Athens Veterinary Diagnostic Laboratory [AVDL], University of Georgia, Athens, GA, USA), FAT is an ancillary test that is offered as part of the autopsy, with no extra charge for the client. For this reason, it is often the test of choice to confirm or rule out Listeria spp. infection in ruminants with neurologic disease. However, the number of false-negative results over the years have led to anecdotal evidence that FAT is not reliable for the diagnostic confirmation of neurolisteriosis in our cases, and that IHC or PCR testing should be used as alternative tests to establish a definitive diagnosis (Rissi DR, pers. comm., 2023 Oct 1).

To test our hypothesis, we compiled cases of neurolisteriosis in goats diagnosed in our laboratory over a 5-y period and compared the FAT results with IHC and PCR test results. We selected caprine cases because goats are the animal species most commonly diagnosed with neurolisteriosis in our laboratory submissions.

Cases of neurolisteriosis in goats were identified retrospectively from the archives of the AVDL between July 2018 and July 2023 using the keywords caprine, listeriosis, and Listeria. Submission forms and autopsy reports were reviewed for cases with inflammatory lesions in the brainstem (specifically in the pons) and with FAT results performed using a fresh or frozen cross-section of pons at the time of diagnosis (the sample is collected at the level of the cerebellar peduncles and attached to the cerebellum, according to AVDL diagnostic routine guidelines). For the FAT, samples were frozen and cut on a cryostat. A 5-μm tissue section representing ~25% of the area of the pons was transferred to a clean microscope slide. The slide was allowed to air dry at room temperature for ≥10 min, fixed in room temperature acetone for 10 min, and allowed to air dry at room temperature for 10 min. A FITC-conjugated fluorescent antibody (FITC labeled goat anti-Listeria genus specific antibody, 02-90-90; KPL/SeraCare), with an Evans blue counterstain, was added to the slide and incubated in a warm humid chamber for 30 min. The slide was then rinsed with 0.85% saline for ≥10 min and subsequently coverslipped and examined under a fluorescence microscope (ET green/orange #1 FISH dual band filter; Chroma). A positive control sample used in each submission was a caprine neurolisteriosis case confirmed by PCR assay; the negative control sample was a caprine brain sample that was confirmed as negative for Listeria spp. by PCR assay. In addition, brainstem sections from 5 cases of idiopathic lymphoplasmacytic rhombencephalitis with gliosis and necrosis of undetermined cause in goats (not consistent with neurolisteriosis) that tested negative for listerial FAT served as additional control cases in our investigation.

Clinical submission forms and autopsy reports from selected cases were retrieved, and archived H&E-stained and Gram-stained tissue sections were reviewed for the characterization of the lesions in the brain. According to the predominant neuroinflammatory changes, histologic changes in the pons (as well as in the available sections of cranial and caudal brainstem and cerebellum) were characterized as acute (microabscesses with neuronal necrosis, neuronophagia, axonal swelling, and vasculitis with fibrinoid change) or subacute-to-chronic (neuroparenchymal and perivascular clusters of macrophages with rare neutrophils),^3,15 and as unilateral or bilateral. Formalin-fixed, paraffin-embedded (FFPE) tissue sections of pons were retrieved and subjected to IHC for Listeria spp. IHC was performed on an automated stainer (IntelliPATH; Biocare Medical) using a rabbit polyclonal antibody for Listeria spp. (L2650-09A; US Biological) at a dilution of 1:100 for 45 min at 37°C. Antigen retrieval was achieved using proteinase K (S302030-2; Agilent) for 5 min. A biotinylated goat anti-rabbit antibody at a dilution of 1:100 for 10 min was utilized to detect the target, and immunoreaction was visualized using 3,3′-diaminobenzidine (DAB; Biocare Medical) substrate for 12 min, counterstained with hematoxylin.

When present, the overall distribution of immunolabeling within the inflammatory foci was classified as 1 (<30% immunolabeling), 2 (30–70% immunolabeling), or 3 (>70% immunolabeling), and a mean immunolabeling score was generated to compare the extent of immunolabeling between acute and subacute-to-chronic cases. IHC scores were analyzed using SPSS v.29.0 software (IBM). The Kolmogorov–Smirnov test was applied to decide whether distributions were parametric. The results between groups were compared using the student t-test given that equal variances were assumed. A 99% CI of the difference was used, with p ≤ 0.05 considered significant.

Tissue scrolls using FFPE were also retrieved and subjected to PCR for Listeria spp. Nucleic acid was extracted (DNA FFPE kit, EZ2 Connect automated nucleic acid extraction instrument; Qiagen) according to the manufacturer’s instructions. All samples were tested with an exogenous internal control (IC; QuantiNova, Qiagen), which ensures the quality of the extraction process and detects naturally occurring inhibitory compounds in the submitted specimen. The IC was added to each sample during the extraction to monitor the quality of both purification and amplification. A PCR assay for GAPDH was also run as a reference gene to assess the quality of the tissue sample. All tested samples passed IC and GAPDH QC. A Listeria real-time PCR (rtPCR) was performed (IndiMixJoe master mix; Indical Bioscience) with final concentrations of 10 µM of each primer, Lall1055 forward (GTTAAAAGCGGTGACACTATTTGG), Lall1163 reverse (TTTGACCTACATAAATAGAAGAAGAAGATAA), and 6 µM probe Lall1118 (ATGTCATGGAATAAT). For each PCR, 5 μL of DNA template was added to the master mix containing 12 µL of IndiMixJoe, 1 µL of forward primer, 1 µL of reverse primer, and 1 µL of probe. The cycling conditions were 95°C for 5 min, followed by 40 cycles of 95°C for 5 s, 55°C for 30 s, and 68°C for 30 s. The PCR was performed on a 0.1-mL block (QuantStudio 5; Applied Biosystems). Tissue sections from the 5 additional control cases (idiopathic lymphoplasmacytic rhombencephalitis) also served as negative controls for the IHC and PCR assays.

We retrieved 25 cases for inclusion in our study (Table 1; Suppl. Table 1). Briefly, affected goats were 3-mo- to 5-y-old (x̄ = 1.6 y). Clinical signs were reported in 23 of 25 cases and consisted mainly of recumbency (11 of 23), drooling (7 of 23), dysphagia (6 of 23), nystagmus (5 of 23), blindness (4 of 23), and circling (4 of 23). Patients were subjected to an autopsy after euthanasia (16 of 25) or spontaneous death (9 of 25). Gross neuropathologic changes were reported in 1 case (case 18) and consisted of bilateral intraventricular hemorrhage with pinpoint hemorrhages in the mesencephalon.

Table 1.

Diagnostic testing performance (number of positive samples) in 25 cases of neurolisteriosis in goats.

Cases (n)	FAT (%)	IHC (%)	PCR (%)
All (25)	2 (8)	19 (76)	20 (80)
Acute (17)	1 (6)	16 (94)	15 (88)
Subacute-to-chronic (8)	1 (12)	3 (37)	5 (62)

FAT = fluorescent antibody test; IHC = immunohistochemistry.

Neurohistologic changes affected the pons in all cases. Lesions occasionally extended rostrally to the mesencephalon (6 of 25), caudally to the medulla oblongata (6 of 25), and/or dorsally to the cerebellum (4 of 25). Acute inflammatory changes were observed in 17 of 25 cases, and subacute-to-chronic changes were observed in 8 of 25 cases. Lesions were bilateral in 19 cases and predominantly unilateral in 6 cases. Neurolisteriosis was histologically suspected by the diagnostic pathologists in all cases. Infection was confirmed by FAT (Fig. 1) in 2 of 25 cases and suspected by FAT in 1 of 25 cases.

Figures 1–6.

Neurolisteriosis in goats. Figure 1. Strong antigenic labeling using fluorescent antibody test for Listeria spp. in case 2. Figure 2. Acute inflammatory changes consisting of neuroparenchymal clusters of neutrophils (microabscesses) in the pons in case 3. H&E. Figure 3. Acute neuronal necrosis and neuronophagia in case 3. H&E. Figure 4. Acute inflammation consisting of perivascular clusters of neutrophils and foamy macrophages surrounding a blood vessel with fibrinoid change in case 3. H&E. Figure 5. Occasional gram-positive bacilli within the cytoplasm of neutrophils. Gram stain. Figure 6. Subacute-to-chronic inflammation consisting of neuroparenchymal and perivascular clusters of macrophages with rare neutrophils, lymphocytes, and plasma cells in case 19. H&E.

Acute inflammatory changes (Figs. 2–4) were neuroparenchymal clusters of neutrophils (microabscesses) that were surrounded by areas of neuronal necrosis and neuronophagia and scattered axonal swelling (spheroids). Affected areas of the neuroparenchyma had microgliosis and astrogliosis, with perivascular clusters of neutrophils and fewer macrophages, lymphocytes, and plasma cells that occasionally extended to the leptomeninges. Vascular fibrinoid change (3 of 17) or thrombosis (1 of 17) were also observed. Bacterial bacilli were observed within neutrophils or macrophages on H&E-stained tissue sections in 4 acute cases. Gram stain highlighted scattered gram-positive bacilli in 8 of 17 acute cases (Fig. 5).

Subacute-to-chronic inflammatory changes (Fig. 6) consisted of neuroparenchymal and perivascular clusters of macrophages with rare neutrophils, lymphocytes, and plasma cells admixed with glial nodules. No bacteria were observed in H&E-stained tissue sections of subacute-to-chronic cases. Rare gram-positive bacilli were observed in 5 of 8 subacute-to-chronic cases.

Infection was confirmed by IHC in 19 of 25 cases. Immunolabeling was observed in the cytoplasm of neutrophils and macrophages, and highlighted clusters of bacterial antigen or individual bacteria within these inflammatory cells (Figs. 7–12). The immunolabeling scores were statistically higher (p = 0.048) in the acute (2.25/3 ± 0.86) compared to subacute-to-chronic cases (1.33/3 ± 0.58). PCR was positive in 20 of 25 cases. IHC and PCR results were concomitantly confirmatory in 17 of 25 cases. FAT, IHC, and PCR results were concomitantly confirmatory in 1 of 25 cases. Single cases were positive only on FAT (case 2), IHC (case 5), or PCR (case 1). Diagnostic confirmation of neurolisteriosis could not be achieved in 2 of 25 cases (cases 7, 16). All negative control tissues tested negative on FAT, IHC, and PCR.

Figures 7–12.

Neurolisteriosis in goats. Figure 7. Immunohistochemistry (IHC) for Listeria monocytogenes highlights numerous clusters of bacteria and bacterial antigen within neutrophils and macrophages in a microabscess in an acute case (immunolabeling score = 3; case 3). Figure 8. Closer view of case 3 (Fig. 8) with abundant cytoplasmic clusters of bacterial antigen within neutrophils and macrophages. Figure 9. IHC for L. monocytogenes highlights scattered clusters of bacteria and bacterial antigen within macrophages in a subacute-to-chronic case (immunolabeling score = 2; case 19). Figure 10. Closer view of case 19 (Fig. 9) with occasional cytoplasmic clusters of bacterial antigen within macrophages and neutrophils. Figure 11. IHC for L. monocytogenes highlights rare clusters of bacteria within macrophages (circle) in a subacute-to-chronic case (immunolabeling score = 1; case 24). Figure 12. Closer view of case 24 (Fig. 11) with rare cytoplasmic bacteria within macrophages (circle).

Although all 25 of our cases had histologic lesions in the brainstem that were typical of neurolisteriosis, the diagnosis could be confirmed by routine FAT in only 2 cases. The stage (acute vs. subacute-to-chronic) or symmetry (bilateral vs. unilateral) of the neurohistologic changes had no apparent impact on the FAT results, as most cases were negative regardless of the lesion. FAT is regarded as a reliable and fast method to identify L. monocytogenes in animal autopsy samples and food specimens,^1,2 but there are only rare investigations documenting its effectiveness in the routine diagnosis of neurolisteriosis in ruminants.^4,5,9 Our results confirm that FAT was not reliable in the diagnosis of caprine neurolisteriosis using fresh-frozen tissue samples from autopsy. If performed, FAT should be interpreted in combination with the neurohistologic changes and aided by listerial IHC and/or PCR for diagnostic confirmation.

The reasons for the high rate of false-negative FAT results in our cases remain undetermined. Although there is no guarantee that tissue sections with lesions and bacterial organisms were sampled during autopsy and submitted for FAT, our histologic findings indicate that inflammatory changes and bacterial organisms were present not only in the pons but often extended to the cranial and caudal portions of the brainstem or to the overlying cerebellum, as is typical for neurolisteriosis in ruminants. ³ Further, tissue sampling for FAT during autopsy always results in a cross-section of pons and cerebellum, indicating that lesions and bacterial antigen would have likely been present in the samples submitted for FAT. Issues with the FAT antibody or technique are also unlikely to have occurred, as positive and negative control tissues stained appropriately.

Issues associated with sample collection from the affected pons and antigen localization within the lesions are suspected to have played a role in these discrepancies. Only one lateral section of the submitted pons (~25% of the tissue area) was tested for listerial FAT. Because neurolisteriosis may cause unilateral lesions in the brainstem (as observed in 6 of our cases), ¹³ there is a possibility that the tested sample was collected from the less affected side and/or it did not have a cluster of inflammation with bacterial antigen. However, although this may be considered as a potential problem, lesions were bilateral in 19 of our cases, which may not entirely support this assertion. Future investigations testing unilateral versus bilateral listerial FAT are necessary to resolve this hypothesis. Further, differences in the amount and intensity of FAT staining were observed when using different sections of the same sample, even when evaluating the same tissue area (Fishburn J, pers. comm., 2024 Jan 10). Our findings suggest that testing more than one tissue section in a case could be considered as a means to increase the chances of diagnostic confirmation. However, this would increase the turnaround time and the cost of the FAT. Another apparent issue with our FAT is that the conjugate only stains free bacteria and not intracellular antigen. Our IHC revealed that most bacteria and bacterial antigen labeling within the lesions was intracellular, which may be the most likely explanation for the false-negative FAT results, as intracellular antigen would not have been stained by the FAT conjugate.

IHC has been successfully used in the diagnosis of neurolisteriosis in ruminants,^{6–8,10,14–16} and it has been shown to be a superior detection method compared to Gram stain and BC. ⁷ Bacteria are more easily observed on routine H&E and Gram stains in acute neurolisteriosis cases, in which microabscesses and necrotic lesions predominate over macrophagic infiltration.^7,15 This feature was confirmed in our cases and was further highlighted by the higher immunolabeling scores in acute lesions than in subacute-to-chronic cases. Similarly, increased listerial immunolabeling has been correlated with positive bacterial isolation from affected animals.^7,10 BC was not performed in our cases, and its assessment was out of the scope of our study. IHC was confirmatory in 19 of our cases, corroborating the results of previous investigations on caprine, ovine, and bovine neurolisteriosis.^7,10,15 Further, our PCR assay was confirmatory in 20 cases, highlighting its reliability in the diagnostic confirmation of neurolisteriosis, particularly when subacute-to-chronic lesions are observed. Our results support our assumption that IHC and PCR tests are significantly more reliable than FAT to confirm neurolisterial infection.

Although histologic changes suggestive of neurolisteriosis were present in 2 cases, no diagnostic confirmation could be achieved with FAT, IHC, and/or PCR. The reason for these previously reported discrepancies remains unknown. ⁷ There is a possibility that these 2 cases could reflect infection with other species of listerial organisms, such as L. ivanovii or L. innocua.^11,13 In addition, to rule out other possible bacterial infections, we tested FFPE tissue scrolls from these 2 goats with a Eubacteria Universal DNA PCR assay; results were negative (results not shown), even though bacterial bacilli were observed in these cases.