There were 16 males and 14 females in the RE group, with a mean age of seizure onset of 5.6 years and a mean age at surgery of 6.8 years. The 16 trauma patients in the control group were generally matched with the RE patient group. There were 10 males and 6 females with a mean age at surgery of 8.2 years.
Expression levels of EBV and HHV6 in the brain tissues of patients with RE were analyzed by IHC and recorded according to a scoring methodology described previously (Wang et al. 2017). As shown in Table 1, in the RE group, EBV was detected in 56.7% (17/30) of patients with RE. Strongly and moderately positive staining levels were each found in two cases and thirteen cases showed weakly positive staining. The positive rate and staining intensity were similar to those of our previous report (Wang et al. 2017). For HHV6, 50% (15/30) of patients with RE showed positive staining. Three and four cases exhibited strongly and moderately positive staining, respectively, and eight cases showed weakly positive staining (Table 2, Fig. 1A). In contrast, both EBV and HHV6 were negative in the brain tissues of 16 control individuals. The co-expression of EBV and HHV6 was detected in 20.0% (6/30) of patients. Both EBV and HHV6 antigens were observed in the cytoplasm and/or nuclei of neuron-like cells in lesion areas of RE brains (Fig. 1A, Wang et al. 2017).
Table 1. EBV antigen expression in brain tissues of 30 patients with RE and 16 control individuals.
Table 2. HHV6 antigen expression in brain tissues of 30 patients with RE and 16 control individuals.
Figure 1. HHV6 expression in brain tissues of patients with RE and controls, as well as characteristics of magnetic resonance imaging (MRI) and histopathological changes in the RE case. A Representative images of strong, moderate, and weak positive staining and negative staining for the HHV6 antigen under low (scale bar = 100 μm) and high (scale bar = 50 μm) magnification. Neuron-like cells were stained. B Atrophy of the left hemisphere cortex (white arrows) and widening of the caudate nucleus (black arrow) were observed in the patient by T2 (B-a) and FLAIR (B-b) images. HE staining shows microglial nodule formation (B-c), lymphocyte infiltration, and perivascular cuff (B-d) in the temporal lobe cortex of the brain. Activation of CD8+ T cells was detected in the RE (B-e) and (B-f) but not in the brain tissue from the trauma patient (B-g) and (B-h) by IHC under low (scale bar = 100 μm) and high (scale bar = 50 μm) magnification.
According to the MRI results for patients with RE, various grades of brain atrophy were observed. The association between EBV, HHV6, and co-infection with EBV and HHV6 and brain atrophy was analyzed. According to a Fisher's exact test, RE patients with positive EBV and HHV6 staining exhibited more severe brain atrophy (Tables 3, 4) than that of EBV- and HHV6-negative cases (χ2 = 10.47, P = 0.002; χ2 = 8.89, P = 0.008, respectively). Moreover, all 6 patients with EBV and HHV6 co-infection exhibited a high grade of brain atrophy (3 patients each for AG 2 and 3), indicating a link between co-infection and the severity of brain atrophy. This result suggests that EBV and HHV6 co-infection may lead to more severe neuronal damage and may be involved in disease progression.
Table 3. Association between brain atrophy and EBV antigen expression in brain tissues of patients with RE.
Table 4. Association between brain atrophy and HHV6 antigen expression in brain tissues of patients with RE.
Among 6 cases with EBV and HHV6 co-expression, we focused on a typical RE case. A 4-year-old boy who presented with falling down while walking or nodding more than 20 times a day was admitted to Beijing Sanbo Brain Hospital (Capital Medical University, China) on March 5, 2015. According to his mother, the boy experienced repeated fever and convulsions several times and was diagnosed with "viral encephalitis" at the local hospital during the spring and summer in 2014. After treatment with sodium valproate (20 mg, tid) and clonazepam (0.5 mg, bid) for about 3 months, the clinical signs improved significantly and the frequency of focal seizures was reduced. However, he manifested nodding and limb shaking again in the winter of 2014. Further treatment with valproate, clonazepam, and levetiracetam (0.75 mg, bid) did not result in the expected therapeutic effect and the patient could no longer stand beginning at the end of 2014, mainly due to gradually increased muscle tension. On March 5, 2015, the boy was admitted to our hospital for further treatment. MRI of the brain showed atrophy of the left temporal lobe, a swollen sulcus, and caudate nuclei. Grey and white matter hyperintensities of the left cerebral hemisphere were detected on FLAIR images (Fig. 1B-a, 1B-b). According to the European diagnostic criteria (Varadkar et al. 2014), the boy was diagnosed with RE. Finally, he had a hemispherectomy on March 16, 2015 and the seizures disappeared after the operation.
During the operation, samples from different brain areas were collected from the patient. The brain sections were subjected to a histopathological analysis and HSV1, HSV2, HCMV, EBV, and HHV6 viral antigen detection. By HE staining, pyknosis, loss of neurons, diffuse lymphocyte infiltration, perivascular lymphocyte cuffing, and glial cell proliferation were observed in the temporal lobe cortex of the brain (Fig. 1B-c, 1B-d). Moreover, using a rabbit anti-CD8a polyclonal antibody (cat: ab17335-1-AP, 1:200; Proteintech, Chicago, IL, USA), a large number of CD8+ T cells were detected in the brain tissue of this patient and were located close to neurons (Fig. 1B-e). This elevation in CD8+ T cells was not observed in the control brain samples from a trauma patient (Fig. 1B-g). These results indicate that cytotoxic T cells are activated in the brains of patients with RE.
Notably, strongly EBV- and HHV6-positive immune reactions were detected in the cortex of the patient brain using a mouse anti-EBV-LMP1 antibody and anti-HHV6 antibody (Fig. 2). EBV and HHV6 antigens were detected in both the 'nidus' and 'non-nidus' areas of the brain, with varying staining levels, but not in the control brain tissue. Double immunofluorescence staining showed that the EBV and HHV6 antigens not only co-existed with MAP-2, a neuronal biomarker, but also co-localized with glial fibrillary acidic protein GFAP, an astrocyte biomarker, indicating that EBV and HHV6 could infect both neurons and astrocytes (Fig. 3). However, no HSV1, HSV2, or HCMV was detected in RE brain tissues by IHC. In combination with the detection of a large number of CD8+ T cells in the RE brain, our results suggest that EBV and HHV6 infection may lead to neuronal damage and brain atrophy by activating CD8+ T cells.
Figure 2. Representative images showing positive staining for EBV and HHV6 and negative staining for HSV1, HSV2, and HCMV in brain tissue of the RE case. Negative staining was obtained for HSV1, HSV2, EBV, HCMV, and HHV6 in the brain tissue from a trauma patient. Bar = 100 μm or 50 μm.