Differential allelic expression of PrP mRNA in carriers of the E200K mutation

The largest focus of genetic Creutzfeldt-Jakob disease (CJD), an autosomal dominant fatal neurodegenerative disorder, is found among the Libyan Jewish community living in Israel.¹ CJD in Libyan Jews has been shown to be linked to a mutation at codon 200 of the prion protein (PrP) gene, resulting in the substitution of lysine for glutamate.^2-4 Other mutations in the PrP gene have been described as well.^5-9

Many lines of evidence argue that the abnormal isoform of the prion protein (PrP^Sc) is necessary for both the transmission and pathogenesis of CJD and other prion diseases.¹⁰ It was suggested that the fundamental event underlying prion propagation features the loss of part of the α-helical structure present in the normal isoform, PrP^C, to form a large amount of β-sheet in PrP^Sc.^11-13 Pathologic mutations in the PrP gene may result in destabilization of at least oneα-helix in the mutant PrP^C.

CJD linked to the E200K mutation presents with a wide range of ages at disease onset. Calculations with use of life-table analysis suggested a penetrance close to 100% for CJD outbreak among carriers of the E200K mutation, although the penetrance is very much age dependent.^14,15 Although the penetrance can be as low as 1% for mutation carriers at age 40, it approaches 100% for those individuals above age 80. These results suggest that although the presence of the mutation is sufficient for the onset of the disease, other factors may determine the time frame for such an event.

The cluster within the Libyan Jews that carries the E200K mutation seems to be a close genetic group. This notion is supported by findings within the PrP gene itself. Most of the Libyan Jewish patients and carriers are homozygous for methionine (Met) at codon 129 of the PrP gene,⁴ and therefore this polymorphism has been ruled out as a factor that may influence age at disease onset in this community. In other genetic CJD clusters, the same polymorphism has been shown to determine disease phenotype.^8,16-21 Sequencing of the PrP gene from a number of old carriers and young patients also did not reveal additional substitutions within the gene (Rosenmann and Gabizon, unpublished results).

Heterozygous carriers of the E200K mutation are presumed to produce both wild-type (wt) and mutant PrP^C. Recent results suggest wt and mutant PrP^C proteins, while sharing some properties (PK sensitivity and solubility in detergents), differ in others (concentration, glycosylation profile, and degradation rate).^22-24 PrP^Sc in the brains of heterozygous E200K CJD patients is composed primarily of mutant PrP.²⁴ Such is also the case for Gerstmann-Sträussler-Scheinker (GSS) patients, in which PrP plaques are composed mostly of the designated mutant PrP.^19,25 These results suggest that the two PrPs do not play the same role in CJD pathogenesis, although they may compete for a mutual binding site (protein X).²⁶ If this is the case, then the ratio of wt to mutant PrP could be important for the initiation of the pathogenic process.

We investigated the possibility that preferential expression of one of the PrP proteins in heterozygous mutation carriers may correlate with age at disease onset. Such results will be consistent with the evidence obtained from transgenic mice experiments, which suggests endogenous PrP inhibits scrapie infection in transgenic mice expressing human PrP.²⁶ We tested PrP protein, total mRNA as well as differential mutant and wt PrP mRNA expression in Epstein-Barr virus(EBV)-transformed B cells derived from healthy heterozygous E200K carriers and CJD patients of the same age range. These cells were chosen for this project since they are prepared from peripheral blood, which is easy and ethical to collect. EBV-transformed B cells have been used in other systems to investigate protein expression.^27-29

Our results suggest that whereas in CJD patients expression of both PrP proteins was similar, there was a considerable excess of wt PrP protein in most healthy carriers. Comparable results were obtained for the mRNA expression of both PrPs. Testing the levels of the total mRNA suggested that preferential wt PrP mRNA expression in these individuals may be caused by an excess of wt PrP rather than by inhibition of mutant allele expression.

Subjects and methods. Experimental population. The patient population consisted of 12 (2 homozygous and 10 heterozygous) E200K CJD patients between the ages of 44 and 65, as well as 11 healthy heterozygous carriers of the E200K PrP mutations (ages 33 to 65). In addition, we examined 15 unrelated normal controls. The study was approved by the Hadassah Hospital Ethical Board.

Cultured cells. EBV-transformed B cells were prepared according to published protocols.³⁰ Briefly, leukocytes were separated from 5 to 10 mL of heparinized blood on a ficoll gradient(Pharmacia), and subsequently incubated with EBV at 37 °C for 2 hours. Cyclosporin A was added to eliminate the T cells, and the B-cell line was cultured in RPMI medium with 10% fetal calf serum.

Protein studies. For immunoblotting of PrP from B cells, cells were collected from a confluent T75 flask and washed twice with PBS before extraction at 4 °C, in 1 mL lysis buffer containing 10 mM Tris HCl, pH 7.4 100 mM NaCl, 10 mM EDTA, and 0.5% Nonidet P-40 (NP40). After centrifugation at 2,000 g for 15 minutes, the supernatants were concentrated by methanol precipitation and processed for SDS-PAGE withα-PrP mAb 3F4.³¹ The intensity of the bands was evaluated by scanning through the Adobe Photoshop program.

RNA studies. RNA was isolated from EBV-transformed B cells by the Tri-reagent TMLS method (Molecular Research Center, Inc., Cincinnati, OH).

RT-PCR of PrP mRNA. Ten to 20 µg of total RNA from lymphoblastoids were pretreated with DNase. RNA samples were transcribed to cDNA by reverse transcriptase (RT) with oligo dt₂₁ as a primer followed by PCR amplification with PrP-specific primers C and D.¹⁶ The transferrin receptor mRNA was used as a control gene as described elsewhere.²⁷

DNA studies. PCR amplification. DNA from leukocytes and lymphoblastoids was prepared as previously described.³ DNA was utilized for PCR amplification of the PrP open reading frame with use of C and D primers under conditions described elsewhere.¹⁶

Allele-specific oligonucleotide hybridization. The PrP PCR products derived from genomic DNA and the RT-PCR products derived from RNA amplification were dot-blotted into a GeneScreen+ membrane and hybridized with oligonucleotides specific either for mutant (K) or wt (E) PrP under conditions previously described.³ Total PrP RT-PCR product was hybridized with use of probe 178D (5′ TGT-GCA-CGA-CTG-CGT 3′) under 42 °C. A numeric evaluation of the dots area and intensity was performed by analyzing the data from the scanned dots with the use of Sigma Scan Program (Jandel Scientific Software, CA).

Results. CJD linked to the E200K mutation has a wide range of age at disease onset (40 to 75 years).¹⁵ The CJD pedigree depicted in figure 1 presents with such age differences. While two cousins succumbed to CJD at a 20-year interval, all the carrier siblings of patients who were also mutation carriers remain healthy. The mutation carriers in this pedigree were all heterozygous for the E200K mutation.

PrP proteins in heterozygous individuals. As in fibroblasts and brains,²⁴ the molecular weight of wt PrP in B cells was higher than that of PrP E200K. Also, the ratio of expression of both PrP (wt and mutant) proteins was not constant in all individuals. Examples of these results are illustrated in figure 2. A sample from a B-cell extract of a homozygous E200K patient (lane 3) presents with a considerably lower molecular weight than that of a wt control sample in lane 4. Note also the lower total concentration of PrP in the control compared with all the carriers. The samples in lanes 1 and 2 were obtained from two brothers of about the same age, one healthy and the other a CJD patient, both heterozygous for the E200K mutation. The bands representing the two PrP proteins were not present at the same ratio in the cells of the heterozygous healthy carrier (lane 1) compared with the cells of the heterozygous patient(lane 2). Numeric evaluation of the bands suggests that while in the cells obtained from the patient the concentration of both proteins was similar, in the cells of the healthy brother wt PrP was present with an excess of more than threefold over that of mutant PrP. This effect may be due either to unequal degradation rate of the mutant PrPs in different mutation carriers or to differences in the ratio of expression of wt to E200K PrP mRNAs.

Differential expression of PrP alleles. In order to determine whether there is a preferential mRNA expression of one of the two PrP alleles in heterozygous CJD patients, we used the RT-PCR method to amplify mRNA samples from all the members of the pedigree described above as well as from other carriers and patients. The products were subsequently hybridized with a probe specific for either the E (wt) or the K (mutant) allele. These probes have been shown to be very specific at our hybridization conditions since DNA from individuals homozygous for either the wt or the mutant allele reacted only with the appropriate E or K probe, and not with the other.³ As reference for each RT-PCR reaction, we used the same PCR primers and probes to amplify and hybridize DNA from the same individual. Since there is only one copy of the PrP gene in each DNA allele, the amplified DNA in heterozygous E200K carriers will be 50% E and 50% K. The intensity of the amplified RNAs from both alleles was compared with that of the amplified DNA. Differences of at least twofold in the ratio of RNA expression from both alleles compared with DNA was defined as preferential PrP mRNA expression of one of the alleles. A similar method was used in other systems as well.^27,32,33

The results of these experiments are described infigure 3. In most of the samples shown in the "patients" panel, the dots from both alleles were equally intense (8/10). One sample(number 1) showed clear preferential expression of the E allele, while another (number 8) showed a borderline effect. In the "carriers" panel, the majority (7/11) of the samples showed various levels of preferential expression for the E allele. Five of the seven samples present with more than 50 times for E/K. It is worth noting that among the few carriers that show similar expression of PrP mRNA from both alleles, one sample (number 21) was typed as heterozygous for the Met/Val polymorphism at codon 129. This polymorphism, which is rare in our population,⁴ was shown to induce late disease onset in other CJD clusters.^16,18,21 Another patient (number 18), who was a carrier when the B-cell line was prepared, has since then (1 year later) succumbed to CJD.

Table 1 summarizes these results: while the great majority of CJD patients expressed an equal amount of wt and mutant PrP mRNA, most of the carriers express an excess of the wt PrP. The difference between the ratio of expression of wt to mutant PrP between carriers and patients is significant at the level of 10% (p = 0.08, Fisher's exact test). This result is statistically significant considering the small number of samples.

The percentage of the E > K cases in patients and healthy carriers altogether was 43%. This is a meaningful result since carriers might become patients at an unknown point in time and influence the statistical significance of the results presented above. For example, if sample number 18 is calculated as a patient instead of as a carrier, the p value of a Fisher test is now 0.03, suggesting the results are even more statistically significant.

Total PrP expression. In an attempt to determine whether the preferential expression of wt PrP mRNA in the carriers of the E200K mutation results from overexpression of the wt allele or from inhibition of the mutant allele expression, we compared total PrP mRNA concentrations in controls, patients, and carriers. If carriers that have an E > K pattern infigure 3 express an excess of total PrP mRNA when compared with carriers or patients with an E = K pattern, it would suggest that the first possibility is more likely-wt PrP mRNA is expressed in excess in these healthy individuals.

Since EBV-transformed B cells express low levels of PrP mRNA (Rosenmann and Gabizon, unpublished results), we used the RT-PCR method also to compare total expression levels of PrP mRNA in these cells. The southern blot of PCR-amplified reverse-transcribed PrP mRNA samples from carriers (healthy and patients) as well as controls can be seen in figure 4. Human transferrin receptor was used as the control gene. The results of these experiments suggest a great variability in the levels of PrP mRNA expressed in lines of different individuals. It is also noteworthy that while most CJD patients express low levels of PrP mRNA (7/7), the majority of healthy carriers (6/7) express (relatively) larger amounts of total PrP mRNA, suggesting the possibility that the preferential expression of wt PrP is the result of its increased expression in healthy carriers. This conclusion, however, is only tentative because of the general low expression of PrP in all the EBV-transformed B cells.

Discussion. While PrP-ablated mice are immune to scrapie infection,^34,35 transgenic mice carrying a large number of hamster PrP gene copies succumbed to the disease significantly earlier than those with only a few copies.³⁶ These results suggest a certain dependence between the incubation time for transmissible prion diseases and the total dose of PrP expressed in the brain of the inoculated animals. Other experiments with mice expressing human transgenes on PrP-ablated background compared with wt mice expressing human transgenes had suggested that endogenous mouse PrP gene can prevent the transmission of human CJD by the conversion of human PrP^C to human PrP^Sc in those mice. These results are explained by the presence of a hypothetical species-specific protein X, which participates in the conversion process and for which different PrP molecules may compete. Protein X is postulated to assist in the binding of a PrP^C molecule to a PrP^Sc molecule.²⁶

During the course of CJD in heterozygous carriers of prion mutations, while mutant PrP transforms from PrP^C to PrP^Sc, wt PrP presents only with partial properties of the pathologic isoform.²⁴ Prion propagation in genetic prion diseases, therefore, favors the mutant PrP^C for conversion into PrP^Sc.

If the hypothetical protein X has a similar affinity to both PrP^Cs(wt and mutant), but only mutant PrP^C is a candidate for transformation into PrP^Sc, then the ratio of these two proteins as well as the actual dosage of mutant PrP^C may have an important effect on the age at which a mutation carrier has accumulated enough PrP^Sc to develop disease.

The results presented in this manuscript support this hypothesis because they suggest that the ratio of wt PrP to mutant PrP protein may influence the onset of CJD in carriers of the E200K mutation. Data regarding the results for two brothers, one who died of CJD and the other a healthy carrier, are summarized in table 2. In the patient's cells, the two PrP proteins, as judged by their different molecular weights, were present at a similar concentration, while there was a considerable excess of wt PrP in the unaffected carrier. Wt PrP mRNA was expressed 50 times more than the E200K PrP mRNA. Total mRNA concentration was considerably higher for the healthy carrier compared with that of his deceased brother. Thus, the increased concentration of wt PrP over E200K PrP appears to be caused by overexpression of the wt allele.

Since imprinting has been ruled out (see table 2: brothers of the same CJD mother differ in their E/K ratio) and experiments using actinomycin D suggest no differences in allele-specific PrP mRNA stability (Rosenmann and Gabizon, unpublished data), two other mechanisms can be considered to explain the differential expression of PrP mRNAs: (1) polymorphisms in the PrP gene regulatory regions may result in low expression of mutant PrP mRNA; and (2) the ratio of the two PrPs may not be stable over the life span of a mutation carrier, but rather most individuals are born with an excess of wt to mutant PrP. This situation changes at some point in time, causing the termination of protection from disease that results from an excess of wt PrP. It is noteworthy that the presence of wt PrP by itself is insufficient to prevent disease, since the only few individuals (five so far) homozygous for the E200K mutation identified in this community were well and healthy at least until middle age. This suggests that an additional age-dependent trigger is required for the manifestation of the disease.

Another open question is whether E200K PrP and other mutant PrPs are functionally inactive. The investigation of this question awaits the elucidation of the normal function of PrP. The study of various PrP null-mice models may facilitate this task.^34,37,38

Elucidating the mechanism by which wt PrP is overexpressed in heterozygous individuals, apparently protecting carriers of some PrP mutations from contracting the disease at young age, may help identify factors that are important for the pathogenesis of prion diseases.

Similar differences in allelic expression have been described also in myotonic dystrophy, another inherited disorder with phenotypic variation.²⁷ It is possible that also in other inherited diseases preferential allelic expression may contribute to the appearance of variable phenotypes.