Identification of Upregulated Genes in Scrapie-Infected Brain Tissue

Abstract

The pathogenesis of scrapie, and of neurodegenerative diseases in general, is still insufficiently understood and is therefore being intensely researched. There is abundant evidence that the activation of glial cells precedes neurodegeneration and may thus play an important role in disease development and progression. The identification of genes with altered expression patterns in the diseased brain may provide insight on the molecular level into the process which ultimately leads to neuronal loss. Differentially expressed genes in scrapie-infected brain tissue were enriched by the suppression subtractive hybridization technique, molecularly cloned, and further characterized. Northern blotting and nucleotide sequencing confirmed the identities of 19 upregulated genes, 11 of which were unknown to be affected by scrapie. A considerable number of these 19 genes, namely those encoding interferon-inducible protein 10 (IP-10), 2′,5′-oligo(A) synthetase, Mx protein, IIGP protein, major histocompatibility complex classes I and II, complement, and β₂-microglobulin, were inducible by interferons (IFNs), suggesting that an IFN response is a possible mechanism of gene activation in scrapie. Among the newly found genes, that coding for 2′,5′-oligo(A) synthetase is of special interest because it could contribute to the apoptotic loss of neuronal cells via RNase L activation. In addition, upregulation of the chemokine IP-10 and B-lymphocyte chemoattractant mRNAs was seen at relatively early stages of the disease and was sustained throughout disease development.

Scrapie, a naturally occurring disease of sheep and goats, and the corresponding pathologies in humans (Creutzfeldt-Jakob disease, Gerstmann-Sträussler-Scheinker syndrome, fatal familial insomnia, and Kuru), cattle (bovine spongiform encephalopathy), and other animals are transmissible progressive neurodegenerative disorders (1, 4, 7, 11, 28). The neuropathology of these transmissible spongiform encephalopathies (TSEs) is usually linked to the appearance of an abnormal insoluble and protease-resistant form of a normal host-encoded protein, the prion protein (PrP). Structurally, the disease-associated abnormal form of PrP, termed PrP^res or PrP^Sc, is characterized by a high beta-sheet content in contrast to the predominantly alpha-helical fold of normal PrP (10, 23, 33).

TSEs in general are characterized by a reactive gliosis and the subsequent degeneration of neuronal tissue. The activation of glial cells, which precedes neuronal death, is likely to be caused by the deposition of large amounts of PrP^Sc in the brain (21, 36, 37). Experimental evidence suggests that PrP^Sc participates in initiation of the gliosis and subsequent neuronal loss, since a PrP-derived peptide (PrP_106–126) activates microglial cells in vitro (6, 20). Furthermore, cell culture supernatants from these stimulated cells induce the proliferation of astrocytes and are toxic to neuronal cells (5). Thus, cytokines released by PrP^Sc-activated microglial cells may contribute to scrapie pathogenesis by enhancement and generalization of the gliosis and via cytotoxicity to neurons. However, the mechanisms of PrP^Sc-triggered microglial activation and many of the factors involved are still unknown. A more complete understanding of the disease may help researchers to define possible therapeutic targets and to develop new means of diagnosis.

To identify upregulated genes in the scrapie-infected hamster brain which may be associated with or even cause the neurodegenerative changes, a strategy using the suppression subtractive hybridization (SSH) technique in combination with a differential screening approach was chosen (13). Briefly, hamsters were inoculated intraperitoneally (i.p.) with scrapie strain 263K (16). Control animals were inoculated with the same volume of normal-brain homogenate. Following total-brain RNA isolation at the terminal stage of the infection, synthesis and subtraction of the cDNA pools were carried out with an SSH-based PCR-select cDNA subtraction kit (Clontech, Palo Alto, Calif.) according to the manufacturer's suggestions. Remaining cDNAs were randomly subcloned into a T/A vector (Invitrogen, Carlsbad, Calif.).

In total, 1,200 clones generated by the SSH procedure were analyzed by dot blotting with forward and reverse subtracted cDNA probes from infected and uninfected hamster brain tissue, respectively. One hundred clones displayed signal intensity differences and were further characterized by nucleotide sequencing. Sequence database searches identified eight genes previously described as being upregulated in the scrapie-infected brain (GFAP, transferrin, apolipoprotein J, metallothionein, β₂-microglobulin, major histocompatibility complex [MHC] class I, MHC class II, and MHC class II-associated invariant chain). Northern blot analysis of the remaining clones confirmed that 11 genes which had previously been unknown to be affected by the scrapie infection were differentially expressed: IP-10, BLC, Mx protein, 2′,5′-oligo(A) synthetase, IIGP protein, glycoprotein 39 precursor (gp39), vimentin, aquaporin 4 (AQP-4), lysosome-associated multitransmembrane protein (LAPTm5), the LIM homeodomain protein 7 (Lhx7), and the C1q C chain of complement (Table 1; Fig. 1).

TABLE 1.

Upregulated genes in the scrapie-infected hamster brain

Identified genea	No. of clones	mRNA size (kb)	Scrapie/control (fold increase)
IP-10	4	1.6	NPb
BLC	5	1.8	4–6
Mx protein (p78)	1	2.1	2–5
2′,5′-oligoadenylate synthetase	1	1.9	2–5
IIGP protein	1	3	1.5–3
gp39 precursor	1	2	1.7–5
Vimentin	2	2.5	2.5–3
AQP-4	3	6	2–4
LAPTm5	1	1.9/2.8	1.5–4.5
Lhx7	1	2	2.2–2.2
Complement C1q C chain	5	1.2	2.5–4.5
GFAP	34	3	4–5
Transferrin	6	NDc	ND
Apolipoprotein J	4	ND	ND
Metallothionein	2	ND	ND
β2-Microglobulin	2	ND	ND
MHC class I	1	ND	ND
MHC class II	6	ND	ND
MHC class II-associated invariant chain	1	ND	ND

^aThe genes in the upper group were previously not known to be upregulated in scrapie. The genes in the lower group were previously reported to be affected by the scrapie infection, and this was confirmed in this study. 

^bNP, not possible to determine. 

^cND, not determined. 

FIG. 1.

Northern blot analysis of upregulated genes in the scrapie-infected hamster brain. RNA (10 μg/lane) was fractionated by agarose gel electrophoresis, transferred to nylon membranes, and hybridized with clones IP-10 (a), BLC (b), Mx protein (c), 2′,5′-oligo(A) synthetase (d), IIGP protein (e), gp-39 precursor (f), vimentin (g) AQP-4 (h), LAPTm5 (i), Lhx 7 (j), C1q C chain of complement (k), GFAP (l), and β-actin control (m). Control, RNA from two mock-infected hamsters; scrapie, RNA from two scrapie-infected hamsters at the terminal stage of the disease.

The levels of mRNA induction in scrapie-infected brain among the 19 different genes ranged from 1.5- to 6-fold compared to those of uninfected controls as determined by Northern blotting and subsequent densitometric quantification. Only the IP-10 mRNA was virtually undetectable in uninfected brain tissue and was highly expressed in the terminal stage of the infection (Fig. 1). In addition, expression of the chemokines IP-10 and BLC in the brains of BALB/c mice i.p. infected with scrapie strain 139A was analyzed over time. By using a semiquantitative reverse transcription (RT)-PCR method, both chemokines were found to be induced at day 114 postinfection, which was about 90 days before the terminal stage of the disease (Fig. 2). Even with the highly sensitive RT-PCR, no IP-10 mRNA was seen in the normal mouse brain tissue.

FIG. 2.

Time point of IP-10 and BLC upregulation. RNA from scrapie-infected BALB/c mice was isolated on the days postinfection indicated at the top of the figure. RNA from mock-infected animals, sacrificed at 202 days postinfection, served as controls (C). Further RT-PCR controls were carried out without template (neg1) and with RNA but no prior cDNA synthesis to check for DNA contamination of the RNA preparation (neg2). RT-PCRs were performed with gene-specific primers for IP-10 (3′ primer, GCT GCA ACT GCA TCC ATA TCG A; 5′ primer, TTG GCT AAA CGC TTT CAT TAA ATT C) (a), for BLC (3′ primer: TCA GCA CAG CAA CGC TGC TTC T; 5′ primer, CTG GAG CTT GGG GAG TTG AAG A) (b), and for glycerol-3-phosphate dehydrogenase (G3PDH; to control the amounts and quality of the RNAs used) (3′ primer, TCC ACC ACC CTG TTG CTG TA; 5′ primer, ACC ACA GTC CAT GCC ATC AC) (c) under standard conditions. Ten microliters of each reaction mixture was loaded on an agarose gel and, after electrophoresis, visualized by ethidium bromide staining.

Previous research efforts employed methods like immunohistochemistry, in situ hybridization, subtractive hybridization, and differential-display RT-PCR to analyze gene or protein expression alterations in the scrapie-infected brain (12, 15, 17, 18). Most importantly, proinflammatory cytokines (e.g., tumor necrosis factor alpha, interleukin-1 [IL-1], and IL-6), glial cell activation markers (e.g., GFAP, cathepsin S, and complement C1q), and indicators of cellular stress (e.g., HSP70 and metallothionein II) were found to be upregulated in the scrapie-infected brain (8, 12, 14, 25, 35, 36). The more recently developed SSH technique, a powerful tool for the analysis of differential gene expression, was used in this study to increase our current understanding of the pathological changes in the scrapie-infected brain on the molecular level.

In total, we found 19 upregulated genes (Table 1), 8 of which were previously described by other groups as being differentially expressed. Among the 11 genes newly found to be affected by scrapie infection, IP-10, BLC, vimentin, gp39, LAPTm5, AQP-4, and the C1q C chain of complement are most likely expressed by activated glial cells. Increased vimentin and AQP-4 levels are indicative of an ongoing astrocytosis (27, 32). Among the microgliosis markers, gp39 is seen in late stages of macrophage differentiation and is thus probably expressed by activated and differentiating microglial cells. Upregulation of the C chain of complement C1q in stimulated microglial cells is in agreement with a similar observation for the C1q B chain (12). High LAPTm5 expression levels are associated with an increased lysosomal activity of microglial cells in the vicinity of spongiform histological changes (37). This observation further substantiates the possible involvement of an aberrant activation of the microglial lysosomal system in neurodegeneration (26).

2′,5′-Oligo(A) synthetase is most likely expressed by glial as well as neuronal cells (2). 2′,5′-Oligo(A) synthetase causes activation of RNase L, which has been demonstrated to participate in apoptotic cell death via its nonspecific rRNA-degrading activity (9, 39). Hence, activation of RNase L may directly contribute to neuronal loss in scrapie. In contrast to the upregulation of IP-10 and BLC at day 114 postinfection, increasing 2′,5′-oligo(A) synthetase mRNA levels were first seen at day 141 by RT-PCR (data not shown).

It is less clear which cell types express increased levels of the GTPase IIGP, the putative transcription factor Lhx7 mRNA, and the Mx protein. Mx protein and 2′,5′-oligo(A) synthetase are part of the antiviral response induced by interferons (IFNs). Moreover, given that the expression of IIGP, MHC classes I and II, complement, β₂-microglobulin, and IP-10 is also IFN inducible, gene activation in scrapie bears all of the hallmarks of a typical IFN response. Interestingly, gamma IFN (IFN-γ) was previously shown to activate microglia after stimulation with β-amyloid (31). Furthermore, low doses of IFN-γ, but not IL-1β, IL-2, IL-4, IL-6, or IL-12, synergistically enhance CD40 expression on microglial cells upon stimulation with β-amyloid. Thus, IFN-γ may contribute to the CD40-CD40L interaction-dependent activation of microglia in Alzheimer's disease as well as in scrapie (34). We are presently investigating the role of IFN-γ in the pathogenesis of scrapie by looking at scrapie infections of IFN-γ receptor knockout mice in comparison to wild-type controls.

The chemokines IP-10 and BLC, so far only known as chemoattractants for T and B cells (19, 22, 29), respectively, are likely to be produced by activated glial cells (30). However, since infiltrating lymphoid cells are rarely observed in scrapie, IP-10 and BLC may bind to other target cells in the brain rather than to cells in lymphoid organs. Evidence for chemokine receptor expression on neurons surrounding amyloid plaques in Alzheimer's disease suggests that neuronal cells, in addition to glial cells, may well be potential targets for IP-10 and BLC (24, 38). Studies of chemokine expression and function in the normal and the diseased brain are still in their early stages, and we currently do not know the effects of IP-10 and BLC on receptor-expressing neurons. The induction of both chemokines is seen at early stages of the scrapie infection and is sustained at high levels until the end, which indicates an involvement in disease progression. Increased levels of IP-10 and BLC mRNAs were seen in two species infected with different scrapie strains. Thus, this observation is likely to be of general relevance in TSEs and less variable than the levels of transforming growth factor β1 and cathepsin S seen in different TSE model systems (3). Finally, given that chemokine mRNA levels are early markers for infection, determination of chemokine levels in the brain and in cerebrospinal fluid is potentially useful for the differential diagnosis of TSEs.