An scFv intrabody against the nonamyloid component of alpha-synuclein reduces intracellular aggregation and toxicity

Abstract

Prevention of abnormal misfolding and aggregation of alpha synuclein (syn) protein in vulnerable neurons should be viable therapeutic strategies for reducing pathogenesis in Parkinson's disease. The nonamyloid component (NAC) region of alpha-syn shows strong tendencies to form beta-sheet structures, and deletion of this region has been shown to reduce aggregation and toxicity in vitro and in vivo. The binding of a molecular species to this region may mimic the effects of such deletions. Single-chain variable fragment (scFv) antibodies retain the binding specificity of antibodies and, when genetically manipulated to create high-diversity libraries, allow in vitro selection against peptides. Accordingly, we used a yeast surface display library of an entire naive repertoire of human scFv antibodies to select for binding to a NAC peptide. Candidate scFv antibodies (after transfer to mammalian expression vectors) were screened for viability in a neuronal cell line by transient cotransfection with A53T mutant alpha-syn. This provided a ranking of the protective efficacies of the initial panel of intracellular antibodies (intrabodies). High steady-state expression levels and apparent conformational epitope binding appeared more important than in vitro affinity in these assays. None of the scFv antibodies selected matched the sequences of previously reported anti-alpha-syn scFv antibodies. A stable cell line expressing the most effective intrabody, NAC32, showed highly significant reductions in abnormal aggregation in two separate models. Recently, intrabodies have shown promising antiaggregation and neuroprotective effects against misfolded mutant huntingtin protein. The NAC32 study extends such work significantly by utilizing information about the pathogenic capacity of a specific alpha-syn region to offer a new generation of in vitro-derived antibody fragments, both for further engineering as direct therapeutics and as a tool for rational drug design for Parkinson's disease.

Figure 1

(a). Amino acid sequence of α-syn. The region comprising the non amyloid component (NAC) of α-syn originally identified in Alzheimer’s plaques is underlined. Syn-2–16 sequence is in blue and syn-53–87 sequence is in red. (b) Clones of yeast-displayed scFvs are specific for biotin-labeled syn-53–87, rather than biotin-labeled syn-2–16. The pool of scFvs isolated from a yeast non-immune scFv library by sequential magnetic enrichment and FACS sorting against syn-53–87, with syn-2–16 as negative control. Yeast cells were labeled by both antigen/SAPE (FL-2) and anti-myc/Alexa 633 (FL-4). The FL-4 channel shows the expression level of C-terminal c-myc tagged scFv on the yeast surface, labeled with mouse 9E10 anti-c-myc monoclonal antibody and Alexa 633-conjugated goat anti-mouse polyclonal antibody. The FL-2 channel shows antigen binding, detected by strepavidin conjugated to PE. A gating box to sort yeast cells that express scFv and bind antigen is shown. In the control syn-2-16, there were no yeast clones in the gating box. (c) The percentage of yeast cells binding to 1 mM syn-53–87 peptide. Six yeast clones showed, via surface displayed scFv, a capability of binding syn-53–87. Data represent the average ± standard error of three separate experiments. Epitope mapping. The three peptides, syn-53–63, syn-61–78, and syn-74–87, were labeled with biotin and used to detect the binding capabilities of the six yeast clones. Data represent the average ± standard error of three separate experiments. (d). Germline usage and sequences of complementarity determining regions (CDRs) of identified scFvs selected against syn-53–87. Germline sequences shown were determined using IgBlast and Vbase. The Germline sequence of scFv D10 is shown for reference. CDR H1 refers to the first CDR of the heavy chain, etc, and CDR L1 denotes the first CDR of the light chain etc.

Figure 2

Comparison of intrabody efficacies when screened by co-transfection with A53T α-syn in ST14A rat neural progenitor cells. (a). Effect of various intrabodies on proliferation and viability of ST14A cells when transiently co-transfected with mutant A53T α-syn, after 48 h of transfection (MTT assay; LDH cytotoxicity assay). Cells were cultured in 10% FBS medium, at 10 % CO2 and 37°C. Comparison of intrabodies against the empty vector control was performed by ANOVA with Fisher’s probability of least significant differences; *, P< 0.05. A control experiment (MTT assay) against toxic mutant huntingtin protein (htt-72Q-GFP) fails to show any efficacy of the NAC intrabodies against this amyloidogenic species. C4 intrabody specific for htt was included for comparison. (b) Western blots of transient co-transfections of A53T α-syn plus the scFv clones. Exposure time was 30 sec, except as noted. Anti-c-myc identifies intrabody steady-state expression levels, showing stability only for NAC3 and NAC32, and occasionally for NAC24 as shown. (Heavy chain clones NAC1 and NAC14 were omitted from further screening).

Figure 3

Immunocytochemistry of transiently transfected species into ST14A cells showing the cellular localization. All cells were fixed 48 h post-transfection. (a)Cellular localization of overexpressed proteins, mutant α-syn A53T, red channel; mutant htt-72Q-GFP, green channel; NAC32, green channel, NAC32-NLS, green channel. Blue nuclear DAPI stain indicates that NAC32 localization is extranuclear, whereas NAC32 NLS re-locates to the nucleus. (b) Co-transfection of mutant α-syn A53T (red channel) with NAC32 (green channel) shows cytoplasmic co-localization, whereas co-transfection of mutant α-syn A53T (red channel) with NAC32-NLS (green channel) shows intrabody relocalization of mutant α-syn to the nucleus. (c) Control co-transfections with toxic high poly-Q mutant huntingtin Exon1 (htt-72Q) does not display similar nuclear relocalization.

Figure 4

Effect of NAC32 stably-transfected intrabody on aggregation in ST14A cells transfected with mutant A53T α-syn, and synT. (a) Immunochemistry was performed 48 h post-transfection. Visual comparison of ST14A cells with ST14A/NAC32 stable cells transfected with A53T mutant, and comparison of ST14A cells with ST14A/NAC32 stable cells, transfected with synT mutant, alike show diminished aggregation behaviour. (b) Cell number counts with aggregates in both A53T and synT transfections are significantly reduced in NAC32 cell lines.