Single-neuron transcriptome and methylome sequencing for epigenomic analysis of aging

Abstract

Enormous heterogeneity in transcription and signaling is the feature that slows down progress in our understanding of the mechanisms of normal aging and age-related diseases. This is critical for neurobiology of aging where the enormous diversity of neuronal populations presents a significant challenge in experimental design. Here, we introduce Aplysia as a model for genomic analysis of aging at the single-cell level and provide protocols for integrated transcriptome and methylome profiling of individually identified neurons during the aging process. These single-cell RNA-seq and DNA methylation assays (methyl-capture/methyl enrichment) are compatible with all major next generation sequencing platforms (we used Roche/454 and SOLiD/Life Technologies as illustrative examples) and can be used to integrate an epigenetic signature with transcriptional output. The described sequencing library construction protocol provides both quantitative and directional information from transcriptional profiling of individual cells. Our results also confirm that different copies of DNA in polyploid Aplysia neurons behave similarly with respect to their DNA methylation.

Fig 1

Aplysia is an emerging model for cell biology and epigenomics of aging. The life expectancy of Aplysia is approximately a year, and the animal can grow very rapidly from a small microscopic swimming larva, less than 0.5 mm in diameter to up to a 1–3 kg mature adult comparable in its size with a rabbit. Yet, the Aplysia central nervous system consists of only 10,000 neurons of different coloration. Some of these neurons are the largest in the animal kingdom [57] and are visible to the naked eye. As a result of their surface location and unique morphological and functional properties, these neurons can be reliably identified and mechanically isolated for microchemical and genomic analysis. The isolated cell on a penny is one of the large cholinergic motoneurons (modified from [24, 59, 60])

Fig 2

Different neurons age differently. Comparison of age-related changes in gene expression between two identified cholinergic neurons LPl1 and R2 (modified from [1]). (a) The schematic representation of the reference design microarray experiments to compare two different cell types R2 and LPL1 during the aging process in Aplysia. In these microarray tests, individual neurons were compared to the same reference CNS sample. The individual circles represent single neurons (LPl1—blue tones; R2—orange tones) from young or old animals. (b) Photograph of the freshly dissected right abdominal semi-ganglion with the position of R2 and R14 neurons marked (connective tissues from the ganglionic surface were removed and the natural coloration of cell somata were preserved). This R2 cell is the largest neuron ever photographed reaching 1.1 mm in diameter. When this cell was isolated (insert) and fixed in 100 % ethanol it lost its pigmentation. 1.9 μg of total RNA was obtained from this neuron. (c) We found that only 58 neuronal transcripts (~0.1 %) are differentially expressed when LPl1 and R2 neurons are compared from young animals. In contrast, when the same cells were directly compared from old animals, we identified 2,508 differentially expressed transcripts (~4.5 %). This suggests that identified cholinergic motoneurons are more similar to each other in younger animals than the same neuronal types in older animals. (d) The landscape diagram is modified from Waddington, C. H., 1956 (Principles of Embryology, op. cit., p. 412; [61]). Following Waddington’s visual schematics, the ball represents a neuronal fate. The valleys are the different fates a given neuron might roll into. At the beginning of its journey, development is plastic, and a cell can have many fates. However, as development and aging or memory proceeds, certain molecular events occur randomly and this can lead to different underlying molecular phenotypes and decisions that cannot be reversed

Fig 3

Isolation of single identified neurons and their RNA/DNA content. (a) Panel of a series of photographs showing the isolation of the L7 neuron with a glass suction pipette (modified from [41]). Below is a schematic depiction of a single neuron and its neurites; the diagram shows the amounts of RNA isolated from neuronal somata and axodendritic processes. (b) The amount of RNA isolated from a single Aplysia neuron is directly proportional to the volume of that cell. The gigantic R2 neuron yielded 1.9 μg of total RNA and 250 ng of gDNA

Fig 4

Single-cell RNA-seq and its validation. (a) The diagram presents the workflow of the RNA-seq protocol outlined in the text. Single Aplysia neurons were isolated, RNA extracted, and 454 sequencing libraries constructed. (b) Absolute RT-PCR was used to generate the intracellular copy numbers for four transcripts of interest. This copy number showed a linear correlation to transcript abundance (expression) in the sequence data set. (c) In situ hybridization [42] was performed on one of the neuron-specific and quite abundant transcripts (the neuropeptide FMRFamide) in the R2 neuron. The photo had been captured in 100 % ethanol. Note, the white nuclei (blue asterisks) can be seen in many neurons. (d) Quantitative RT-PCR of transcripts of interest (FMRFamide) displayed a correlation between the digital profile in the different single cells and their corresponding quantitative RT-PCR expression. Expression profiles for both the frequency of sequencing reads and the QRT-PCR displayed similar patterns. However, caution should be taken for interpretation of the RNA data, since abundant transcripts known to be selectively transported to synaptic terminals and located on neuronal somata can also be captured by RNA-seq (e.g., in contrast to R2, L7 does not express FMRFamide). Thus, to test cell specificity of expression, a complementary in situ hybridization should be performed (c)

Fig 5

Neuron-specific expression of sense and antisense transcripts in identified neurons. The library construction protocol reported here preserves directionality and allows quantification of sense (red) and antisense (blue) as revealed by RNA-seq for two indentified neurons: the gill motoneuron L7 and feeding interneuron MCC. Differential expression is demonstrated by the abundance of reads for L7-specific secretory peptide (confirmed by in situ hybridization for the same transcript in L7, see insert) compared to the interneuron MCC which does not express this gene. Remarkable transcriptional complexity is evident even from a small region of the genome shown here

Fig 6

Single-cell Enriched Methylated Genomic DNA Library constructed using the MethylMiner™ approach. The diagram presents the workflow of the reported MethylMiner™ enriched sequencing library protocol. Genomic DNA is first isolated from single neurons, then fragmented to 150 bp. Methylated DNA is enriched from fragmented genomic DNA via binding to the methyl-CpG binding domain of human MBD2 protein, which is coupled to paramagnetic Dynabeads^® M-280 Streptavidin via a biotin linker from the MethylMiner™ kit. The enriched gDNA is eluted with one high-salt elution step. Resultant gDNA is ligated with the appropriate P1 and P2 SOLiD adaptors and amplified. At this point this product can be directly sequenced for regional mapping, bisulfite sequenced for single-nucleotide resolution or used for absolute RT-PCR for copy number determination of genes of interest

Fig 7

Data analysis of both expression and methylation profiling. RNA-seq and methylation sequence data are loaded as tracks on the UCSC browser http://genome.ucsc.edu/. For a specific gene, the correlation between expression and methylation can be viewed

Fig 8

Methylation in polyploidal neurons occurs in all DNA copies. In this illustrated example, after bisulfite sequencing, bands from the amplified PCR reactions were sequenced for DNMT1 gene. The starred cytosines in the sample chromatogram show no background from other nucleotides, thus indicating all the genome copies in Aplysia polyploidal neurons are equally methylated. If some of the cytosines were not methylated, there would be multiple peaks in the chromatogram indicating a mixed population. However, this is not the case here because there are no background peaks in any of the cytosines marked