Apolipoprotein epsilon 3 alleles are associated with indicators of neuronal resilience

Background Epilepsy is associated with precocious development of Alzheimer-type neuropathological changes, including appearance of senile plaques, neuronal loss and glial activation. As inheritance of APOE ε4 allele(s) is reported to favor this outcome, we sought to investigate neuronal and glial responses that differ according to APOE genotype. With an eye toward defining ways in which APOE ε3 alleles may foster neuronal well-being in epilepsy and/or APOE ε4 alleles exacerbate neuronal decline, neuronal and glial characteristics were studied in temporal lobectomy specimens from epilepsy patients of either APOE ε4,4 or APOE ε3,3 genotype. Methods Tissue and/or cellular expressions of interleukin-1 alpha (IL-1α), apolipoprotein E (ApoE), amyloid β (Aβ) precursor protein (βAPP), synaptophysin, phosphorylated tau, and Aβ were determined in frozen and paraffin-embedded tissues from 52 APOE ε3,3 and 7 APOE ε4,4 (0.25 to 71 years) epilepsy patients, and 5 neurologically normal patients using Western blot, RT-PCR, and fluorescence immunohistochemistry. Results Tissue levels of IL-1α were elevated in patients of both APOE ε3,3 and APOE ε4,4 genotypes, and this elevation was apparent as an increase in the number of activated microglia per neuron (APOE ε3,3 vs APOE ε4,4 = 3.7 ± 1.2 vs 1.5 ± 0.4; P < 0.05). This, together with increases in βAPP and ApoE, was associated with apparent neuronal sparing in that APOE ε4,4 genotype was associated with smaller neuron size (APOE ε4,4 vs APOE ε3,3 = 173 ± 27 vs 356 ± 45; P ≤ 0.01) and greater DNA damage (APOE ε4,4 vs APOE ε3,3 = 67 ± 10 vs 39 ± 2; P = 0.01). 3) Aβ plaques were noted at early ages in our epilepsy patients, regardless of APOE genotype (APOE ε4,4 age 10; APOE ε3,3 age 17). Conclusions Our findings of neuronal and glial events, which correlate with lesser neuronal DNA damage and larger, more robust neurons in epilepsy patients of APOE ε3,3 genotype compared to APOE ε4,4 genotype carriers, are consistent with the idea that the APOE ε3,3 genotype better protects neurons subjected to the hyperexcitability of epilepsy and thus confers less risk of AD (Alzheimer's disease). Please see related article: http://www.biomedcentral.com/1741-7015/10/36


Background
Epilepsy is associated with precocious development of Alzheimer-type neuropathological changes, and the APOE ε4 genotype has been associated with further risk of development of such changes [1,2]. A role for glial activation with excess expression of cytokines in epilepsy pathogenesis was first recognized as enlargement of microglia and astrocytes with overexpression of IL-1 and S100B, respectively [3][4][5]. Such findings gave rise to a new understanding of the role of glial activation and overexpression of cytokines as potential precursors of neurodegenerative change, including Aβ plaques and neurofibrillary tangles [6]. These findings are consistent with the idea that glia-related neuroinflammatory events are early contributors to epilepsy pathogenesis.
Neuronal stress, such as the hyperexcitability induced by glutamate in epilepsy, elevates neuronal expression of βAPP and release of sAPP, which activates microglia and induces excess IL-1 production. This elevation in IL-1 production is attenuated by ApoE 3, but not ApoE 4 [7]. In turn, IL-1 induces further neuronal expression of βAPP and sAPP leading to further microglial activation and further release of IL-1 [8]. IL-1 also induces neuronal expression of ApoE [9], which in turn induces further expression of βAPP in an ApoE isoform-dependent manner; with ApoE3 more effective than ApoE4 [10].
A great deal of research has been dedicated to understanding how and why the presence of an APOE ε4 allele(s) is so strongly associated with negative outcomes in neurological conditions, such as head injury [11]. Here, rather than taking this tack, we chose to investigate the potential for beneficial effects conferred by APOE ε3 alleles due to their neuroprotective potential. Tissue samples from temporal lobes resected from epilepsy patients carrying two APOE ε3 alleles were examined regarding an association between inheritance of these alleles and determinants of neuronal resilience. These determinants included the ability of neurons to mount appropriate acute phase responses, including increases in βAPP and ApoE, as well as management of DNA damage, maintenance of morphological integrity and glial activation. Our findings indicate that the APOE ε3,3 genotype confers a neuroprotective advantage over the APOE ε4,4 genotype, in the setting of intractable epilepsy with its accompanying hyperexcitability-induced neuronal damage, glial activation and excessive expression of the proinflammatory cytokine IL-1α.

Patients and specimens
Resected temporal lobe tissues were obtained from 95 epilepsy patients; of those 59 were included in this study (39 males and 20 females; 52 APOE ε3,3 and 7 APOE ε4,4) with an age at surgery ranging from 0.25 to 71 years. Analyses of surgical waste remains from temporal lobectomy surgeries to treat intractable, drug-resistant epilepsy were compared to those of autopsy samples from neurologically normal individuals brought to autopsy for reasons other than this study. Both surgical waste and autopsy tissue are exempt under 46.101 5(b) and approved by our University of Arkansas Institutional Review Board.
All patients underwent anterior temporal lobectomy for treatment of medication-resistant intractable epilepsy. Tissue was sectioned at 4 mm intervals and alternate sections were preserved by flash freezing for molecular analyses and by formalin fixation for histological evaluation. Preliminary immunohistochemical analysis was performed on all epilepsy cases, and a smaller group was selected for further investigation. Six APOE ε3,3 cases (five males and one female, ages 18, 24, 38, 44, 67 and 57 years, respectively) and four APOE ε4,4 cases (three males and one female, ages 10, 22, 50 and 34, respectively) were selected for more extensive analyses, based on age in the case of APOE ε3,3 patients, and with regard to availability of sufficient frozen tissue for molecular analyses among APOE ε4,4 patients. Sufficient frozen tissue and fixed tissue was available for both immunohistochemical and molecular analyses of four APOE ε4,4 patients (three males and one female, ages as above). For uniformity, immunohistochemical examination was restricted to cortical layers III, IV, V and VI of the superior temporal lobe. For comparison of results from our APOE ε3,3 and APOE ε4,4 genotype patients, analogous temporal lobe tissues from neurologically normal individuals of varying APOE genotype and at older ages (four males and one female, ages 71, 97, 59, 50 and 93 years) were assessed. This selection was based on the premise that individuals with pre-AD (Alzheimer's disease) or with AD at these ages would have plaques.

Immunohistochemistry
Paraffin-embedded tissue was sectioned at 7 μm, deparaffinized in xylene and rehydrated in graduated ethanol solutions to deionized water. Sections for IL-1α immunoreaction were placed in boiling sodium citrate buffer (0.01 M, pH 6.0) for 20 minutes; sections for βAPP and ApoE were placed in trypsin solution for 10 minutes at 37°C, and all were blocked using protein block (DAKO), and incubated overnight at room temperature. Secondary antibodies, Alexa Fluor donkey anti-goat and donkey anti-rabbit were diluted in antibody diluent (DAKO) and sections were incubated for 60 minutes, washed three times for 5 minutes each in distilled H 2 O, and coverslipped with prolong Gold with DAPI.

Plaque analysis
Plaques were identified by the simultaneous presence of ApoE and Aβ immunoreactivity. The number of plaques in 10 consecutive 20X images (0.37 mm 2 ) from sections of tissue from each patient was enumerated. Plaque phase was based on Braak and Braak staging of Aβ plaques [12] and estimated with regard to our experience with such estimation in Alzheimer tissue.

Image analysis
Similar to a previous study [10], a quantitative approach was used to examine the number of glia and neurons. Three images per slide (40X magnification) were captured at identical exposure settings, using a Nikon Eclipse E600 microscope (Melville, NY, USA) equipped with a Coolsnap ES monochrome camera (Photometrics, Tucson, AZ, USA). Each of the three images, spanning 37,241.5 μm 2 , was acquired and analyzed using NIS-Elements BR3 software http://Nikon.com and thresholded. Only microglia immediately adjacent to neuron somas were counted. Data were analyzed by ANOVA to assess difference among groups. Significance was provided by P ≤ 0.05.

Reverse transcription (RT) reaction and polymerase chain reaction (PCR) amplification
Total RNA was extracted from brain tissue using TriRea-gent™ RNA (Molecular Research Center, Cincinnati, OH, USA) according to the manufacturer's instructions. RT-PCR was performed as previously described [9]. Briefly, for comparisons of mRNA levels among different RNA samples, RT reactions were performed simultaneously using reagents from a single master mix. PCR was performed using reagents from Clontech (Mountain View, CA, USA). The sequences of primers for human IL-1α and GAPDH, amplification cycles and annealing temperature are provided in Table 1. PCR reactions were stopped by incubation for 10 minutes at 72°C. Equal volumes of reaction mixture from each sample were loaded onto 1.2% agarose gels, and fluorescent images were digitally captured for analysis of intensity with NIH Image software 1.60 version http://rsbweb.nih.gov/nih-image/. Levels of IL-1α were normalized relative to GAPDH in the same sample.

TUNEL staining procedure
For terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) (NeuroTacs Kit, 4823-30-K, Trevigen, Gaithersburg, MD, USA) reactions, rehydrated sections were permeablized with NeuroPore ® for 30 minutes at room temperature, washed in PBS buffer, placed in TdT labeling buffer for 5 minutes, treated with the labeling reaction mix (TdT dNTP, 50 × Mn +2 , and TdT Enzyme) for 60 minutes at 37°C followed by stop buffer for 5 minutes, then streptavidin AF 594 conjugate (Invitrogen, S32356) for 10 minutes at room temperature. The sections were then treated with 0.1% Sudan black B in 70% ethanol for two minutes to block lipofuscin autofluorescence, washed in three changes of distilled H 2 O, five minutes each; and coverslipped with Prolong Gold with DAPI.

Statistical analysis
Data were analyzed using an unpaired t-test, and values were considered significantly different when the P-value was ≤ 0.05. Results are expressed as mean ± SD.

Results
Numbers of IL-1a-immunoreactive microglia per neuron are APOE genotype-dependent An initial screening of 59 epilepsy patients (52 APOE ε3,3; 7 APOE ε4,4) and 5 neurologically normal controls revealed an observable elevation in numbers of microglia per neuron soma in brain tissue from epilepsy patients compared to that in brain tissue from controls (data not shown). A smaller group of age-matched patients with either APOE ε3,3 (n = 6) or APOE ε4,4 (n = 4) genotype was chosen for further analysis to determine if this increase in numbers of IL-1α immunoreactive microglia (Iba-1-immunoreactive) per ApoE-immunoreactive neuron segregated according to APOE genotype. Numbers as high as 8 IL-1α/Iba-1-immunoreactive microglia were noted clustered immediately adjacent to individual ApoE-immunoreactive neurons in APOE ε3,3 patients ( Figure 1A) compared to a maximum of three microglia per neuron in APOE ε4,4 patients ( Figure 1B). Overall, APOE ε3,3 neurons had twice as many associated microglia as did APOE ε4,4 neurons (3.7 ± 1.2 vs 1.5 ± 0.4; P < 0.01) ( Figure 1C). As to the total number of microglia per 0.037 mm 2 in tissue from epilepsy patients with APOE ε3,3 vs APOE ε4,4 genotype (37.2 ± 10.6 vs 23 ± 5.7; P < 0.05).

Neuron number and size with regard to APOE genotype
To assess the consequences of epilepsy and APOE genotype on neurons, we counted neurons in cortical layers III to VI of three specific areas of the temporal lobe. By histological inspection, changes in the appearance of individual neurons were evident, with the distinguishing feature being an APOE ε4,4 genotype-associated decrease in size of both the cytoplasm and nucleus (Figure 2A, B). When quantified, this general impression correlated with a genotype-related difference in neuronal size. The average size of neurons in patients with APOE ε4,4, was smaller than in patients with APOE ε3,3 (173 ± 27 vs 356 ± 45 μm 2 ; P < 0.001) ( Figure 2C, D), but the numbers of neurons were not different in cortical layers of patients with APOE ε4,4 and APOE ε3,3 genotypes (24 ± 2 vs 22 ± 3; P = 0.18) ( Figure 2E, F),.
Neuronal DNA damage is related to APOE genotype DNA damage was present in similar numbers of neurons in analogous areas of the temporal lobe, without regard to APOE genotype ( Figure 3A, C). However, the extent of damage per neuron, as assessed by TUNEL fluorescence intensity, was greater in patients with APOE ε4,4 than with APOE ε3,3 genotype (67 ± 10 vs 39 ± 2; P = 0.01) ( Figure 3B, D).
bAPP tissue levels as a function of IL-1a expression and APOE genotype The increased numbers of microglia per neuron noted in APOE ε3,3 vs APOE ε4,4 patients were reflected at the level of IL-1α synthesis and production. Overall, levels of IL-1α mRNA were elevated within a group of 10 of our epilepsy patients with various combinations of APOE ε2, APOE ε3, or APOE ε4 alleles compared to levels in control patients ( Figure 4A), although one patient with APOE ε4,4 genotype had mRNA levels that were not different from that of control patients ( Figure  4B). Overall, patients with epilepsy had IL-1α mRNA levels that were five-fold higher than those of   neurologically normal controls. In contrast to the epilepsy-associated increase in IL-1α mRNA in patients of both genotypes, IL-1α protein measured by Western blot showed much greater elevation of IL-1α protein in APOE ε3,3 patients than that in APOE ε4,4 patients ( Figure 4C). The increased IL-1α expression in APOE ε3,3 patients was approximately four-fold that of APOE ε4,4 patients (APOE ε3,3 = 51 ± 29 vs APOE ε4,4 = 9 ± 4; control 6 ± 1; P < 0.01) ( Figure 4D). These elevations in the levels of both IL-1α mRNA and protein were accompanied by an APOE ε3,3 genotype-specific increase in βAPP expression in epilepsy patients compared to controls (APOE ε3,3 = 42 ± 11 vs APOE ε4,4 = 20 ± 11; and control = 21 ± 7; P = 0.03) ( Figure 4E, F).

Ab and ApoE deposition relative to APOE genotype
We found Aβ neuritic plaques in 13 of 52 APOE ε3,3 patients, including one who was only 17 years old; one of the 7 APOE ε4,4 patients had plaques; the youngest, 10 years of age ( Figure 5A-C). For this one APOE ε4,4 patient, the number of plaques observed was 17, which was similar to that found in four APOE ε3,3 patients who had Aβ plaques, numbering 11, 15, 14 and 16. Interestingly, the Aβ plaque phase in the one APOE ε4,4 patient was dense core neuritic while the plaques in APOE ε3,3 patients did not have dense cores. With our limited number of patients with APOE ε4,4 genotype, it is not possible to say whether a specific APOE genotype is associated with either a higher probability of having plaques or a genotype-related earlier age of onset for such Aβ plaques. However, epilepsy was associated with an increased probability of Aβ plaque development at a very early age. All Aβ plaques observed contained ApoE deposits as well as ApoE-immunoreactive neurites, without regard to age or gender.

Figure 4
Neural cell acute phase responses in epilepsy relative to APOE genotype IL-1a mRNA levels in temporal lobe tissue from epilepsy patients were elevated compared to levels in control patients (A). With the exception of one patient with APOE ε4,4 genotype, patients with epilepsy had mRNA levels that were greater than those of control patients (B). By Western blot (C), the relative levels of IL-1α protein from epilepsy patients with APOE ε3,3 were four-fold higher (51 ± 29) than those from either epilepsy patients with APOE ε4,4 genotype Synaptophysin and phosphorylated tau expression relative to APOE genotype The synaptic density in photomicrographs from each of the patient groups appeared to be similar, with somewhat higher levels in those with APOE ε3,3 genotype, as measured by relative levels of synaptophysin (APOE ε3,3 = 3.57 ± 0.76 vs APOE ε4,4 = 2.62 ± 0.80 vs controls = 2.41 ± 0.36; P = 0.11) (Figure 6A. B). As previously reported [14], phosphorylated tau immunofluorescent labeling was not elevated in any of our patients regardless of age or ApoE genotype. This was confirmed by Western blot analysis; the bands were faint and the measured levels were low ( Figure 6C, D).

Discussion
Tissues from patients undergoing temporal lobectomies for drug-resistant epilepsy reveal APOE genotype-specific links between glial and neuronal stress responses. This influence of APOE genotype in epilepsy appears to occur without regard to gender or age at the time of surgery. Glial activation with overexpression of IL-1 is well known to induce neuronal expression of two ADassociated, stress-related proteins ApoE and βAPP [10,15]. Connections among APOE genotype, epilepsy and AD have been drawn, but mechanisms by which the APOE ε4,4 genotype heightens intensity of neuronal damage or, conversely, how the APOE ε3,3 genotype may act to promote neuronal resilience remains unclear. The numbers of neurons in temporal lobe tissue of our epilepsy patients who were either APOE ε3,3 or APOE ε4,4 genotype were similar, but there were striking differences in the indicators of degeneration in neurons, as neurons from patients with APOE ε3,3 were larger, appeared more normal morphologically, and had less DNA damage. These findings suggest that neurons from individuals with the APOE ε3,3 genotype are better able to mount appropriate and more liberal repair responses to the damaging hyperexcitability of epilepsy than are their APOE ε4,4 counterparts, suggesting that APOE ε3, but not APOE ε4, alleles confer resilience to host neurons no matter the type of injury. This might be inferred from studies reporting earlier onset of epilepsy, especially following traumatic brain injury in patients with APOE ε4 alleles [16,17].
Our finding of elevated synthesis of IL-1α in the temporal lobe of epilepsy patients compared to that in neurologically normal controls confirms an earlier report [4] of elevated IL-1α protein and accompanying glial activation and other neuroinflammatory changes. However, the association made here between this overexpression of IL-1α and beneficial effects toward enhancing neuronal resilience may help to explain, at least in part, why IL-1α elevation is necessary for neuronal survival in dorsal root ganglion cell cultures [18]. Moreover, evidence of greater neuron sparing in epilepsy patients with APOE ε3,3 than APOE ε4,4 genotype may be a case in point for genetic variation favoring typical, evolutionarily old, acute phase responses [19] of neurons to adverse stimuli, which includes elevation of IL-1α, βAPP and ApoE expression [10] and protection against DNA fragmentation.
The original report of a role for IL-1α in induction, maintenance and propagation of axonal sprouting in an experimental model of neurodegeneration [20] and an association between glial activation and sprouting of mossy fibers in epilepsy [21] is supported by our finding of somewhat elevated synaptophysin levels in combination with high numbers of neuron-associated, IL-1α immunoreactive microglia and elevation of IL-α mRNA and protein levels. In addition, the apparent elevation of synaptophysin expression noted here in immunoblots of neural tissue proteins from our epilepsy patients compared to that from our neurologically and neuropathologically normal controls may be explained if, as previously noted in animal models of epilepsy, [21,22] there is neuronal sprouting in epilepsy patients.
Amyloid-β plaques are obligatory for the diagnosis of AD and are most prominent in the elderly. In contrast, Aβ plaques in epilepsy, as shown here and as reported in about 10% of cases [2], are evident at young ages. For instance among our patients, a 10-year-old patient had Aβ/ApoE immunoreactive plaques in a distribution similar to that noted in temporal lobes of Alzheimer patients. The presence of plaques at such early ages suggests that they are harbingers of impending neurodegeneration and AD. Although the number of plaques was similar in tissue from our patients without regard to APOE genotype, in our one APOE ε4,4 patient the developmental phase of Aβ plaques appeared to be advanced relative to those observed in our APOE ε3,3 patientsour APOE ε4,4 Figure 6 Synaptophysin, phosphorylated tau and actin in epilepsy relative to APOE genotype An impression of similar synaptic densities in photomicrographs from each of the patient groups was confirmed by western blot analysis for the relative levels of synaptophysin compared to actin (APOE ε3,3 = 3.57 ± 0.76 vs APOE ε4,4 = 2.62 ± 0.8 vs controls = 2.41 ± 0.36; P = 0.11 for all comparisons) (A and B). Western blot analysis of phosphorylated tau (P-tau) (C) showed no differences with regard to genotype (D). Western blot analysis of actin (E) was used to assess specific protein levels of IL-1α, βAPP, ApoE, synaptophysin, and P-tau relative to actin. Actin expression was not different with regard to APOE genotype or levels of actin in tissue from controls (F). patient had dense core neuritic Aβ plaques, while such dense core plaques were not found among the plaques observed in our APOE ε3,3 patients. This observation is consistent with the possibility that the phase of Aβ plaque progression is accelerated in those with APOE ε4,4 genotype and supports the findings of Marz et al., regarding the role of APOE genotype in the onset of Aβ plaque pathology and the presence of dense core plaques [23].
Alzheimer's patients are more likely to have seizures than are those in the general population [24]. This, together with our findings and the previously reported preferential occurrence of seizures in younger Alzheimer patients [25], supports a suggested relationship between the high levels of Aβ in the brains of epilepsy patients [26] and increased risk for development of AD. These findings are consistent with the idea that AD-related neuronal stress and its sequelae, including excess neuronal βAPP and ApoE expression and glial activation with elevated cytokine expression, combined with known IL-1-driven elevation of neuronal and glial glutamate production contribute to the hyperexcitability of epilepsy [9]. Moreover, these findings, together with evidence from our epilepsy patients, suggest that ApoE genotype, in particular APOE ε4,4 may favor rapidity of disease progression as well as risk for associated memory disturbances. Conversely, a better understanding of mechanisms by which APOE ε3 alleles confer the neuronal protection shown here may facilitate development of therapeutic strategies toward improving outcomes for epilepsy patients, as well as patients with other neuronal distresses.

Conclusion
The most striking aspect of this work is that our findings illuminate the "other" side of the APOE genotypic equation in showing ways in which APOE ε3 alleles may act to preserve important aspects of neuronal abilities to mount appropriate, beneficial stress responses to hyperexcitability, neuroinflammation and neuronal DNA damage. In addition, our findings are consistent with the idea that as neurons with APOE ε4 alleles are less resilient to the chronic excitation of epilepsy and more susceptible to DNA damage, patients who carry APOE ε4 alleles are at greater risk of developing AD than are those with APOE ε3 alleles. Moreover, our findings are in accord with the possibility that epilepsy-related neuropathological changes, such as increases in the levels of Aβ peptides, contribute to propagation of epileptiform activity in adjacent neurons and furtherance of neuropathological changes and the risk of AD [26].