Chapter 2

Question 1

Where are dopaminergic neurons found in the human brain and where do they project to?

Answer 1

Substantia nigra (most affected in PD): projects to caudate-putamen and striatum (“nigrostriatal pathway”, regulating motor function)

Ventral tegmental area and retrorubal field: project to striatum, limbic system, and PFC

Note that the above areas also project to the hippocampus in rats, and that DAn are also found in other areas throughout the CNS

Question 2

Which brain region connections comprise the nigrostriatal pathway, and what is its function?

Answer 2

SN and its projections to the caudate-putamen and striatum; regulation of motor function

Question 3

Describe 5 challenges associated with modelling sporadic PD in cell lines/animals.

Answer 3

1. Few animals develop NDDs (except baboons and dolphins)
2. Epigenetic patterns are species-specific
3. Transgenic cell lines or rodent models may reflect familial PD etiology more closely than sporadic PD etiology
4. Traditionally, living human neurons/patient-derived neurons were difficult to obtain
5. Cell culture models of PD don’t represent the context of aging (e.g., fetal LUHMES, iPSC embryonic reprogramming)

Question 4

Summarize key advances in PD pathophysiology discovered using SH-SY5Y cells and iPSC-derived dopaminergic neurons.

Answer 4

SH-SY5Y: Mitochondrial dysfunction, impaired autophagy
iPSC DAn: cell-cell transfer of aSyn, HDAC4 mislocalization, deregulated gene expression

Question 5

Discuss the benefits and drawbacks of using LUHMES to model human PD epigenetics, comparing and contrasting to iPSCs.

Answer 5

Advantages:
- Midbrain origin
- Require differentiation, but not full reprogramming through embryonic state followed by differentiation
- Can expand easily
- Can synchronize cell cycle
- Express SNCA, NeuN, and DA-specific genes when differentiated; uptake and release DA

Limitations:
- Derived from a “normal” 8-week female embryo (vs patient)
- Clonal (benefit and drawback)
- Differentiation and culturing procedures still affect the epigenome, as with iPSCs

Question 6

How is the LUHMES cell line proliferated and then differentiated to post-mitotic neurons?

Answer 6

1. Neural precursors transformed w/Myc oncogene and tetracycline response element (proliferation and immortalization)
2. Add tetracycline to stop proliferation, and GDNF, cAMP to begin differentiation to DAn
3. Wait 2 days to halt proliferation, then re-seed cells into a plate, to achieve synchronous differentiation and consistent cell density

Question 7

Name 3-5 indicators that LUHMES cells have differentiated to mature dopaminergic neurons.

Answer 7

1. Dendritic branching & neurite growth
2. Maximal NeuN expression
3. Na+/K+ voltage-gated channels, generation of action potentials
4. DAn-specific expression of TH, DAT, DRD2
5. DA uptake and release

Question 8

What happens to the SNCA locus during LUHMES differentiation, and what does this tell us about potential mechanisms for SNCA SNP variants in sporadic PD?

Answer 8

Increased H3K27ac at SNCA gene, in areas enriched for sPD-associated SNCA SNPs, associated with increased SNCA mRNA expression

Question 9

Why are dopaminergic neurons particularly vulnerable to neurodegeneration?

Answer 9

No mitotic division (less opportunity to recover from toxins/aging); millions of axons far from cell body; generation of ROS from DA metabolism

Question 10

What level of aSyn overexpression is seen in SNCA duplication/triplication carriers and REP1 allele carriers, and how does this compare to the level of aSyn overexpression in the LUHMES model used in Chapter 2 and the mouse model used in Chapter 3?

Answer 10

Triplication: ~2-2.5-fold upregulation of SNCA mRNA and protein in frontal cortex.
REP1: Longest allele has 2.5-fold overexpression of SNCA mRNA (reporter assay in SH-SY5Y). Shorter variants have 1.5-fold. Note that effect on expression is tissue-specific, and effect on PD risk is observed in White but not Japanese populations.

LUHMES: 4-fold mRNA, 6-fold protein (much higher than human!)

Mouse: 6-fold mRNA

Question 11

What are the physiological consequences of WT aSyn overexpression in human PD patients and in cellular models of PD?

Answer 11

In SNCA triplication patients, PD has an earlier onset (30s-40s), progresses faster, and is more likely to have cognitive decline/dementia associated. LBs are found in the hypothalamus, nucleus basalis, and cortex, and neurodegeneration occurs in the nucleus basalis, cortex, and hippocampus.

SNCA triplication iPSCs have impaired differentiation capacity, neurite growth, and action potentials.

Overexpression of WT aSyn increases cytoplasmic inclusions in yeast and 293T cells, increases aSyn polymerization in yeast, results in Golgi fragmentation, miROS, and DNA damage in LUHMES.

Question 12

How do the physiological effects of overexpressing WT and A30P aSyn differ in cellular models?

Answer 12

In 293Ts, A30P aSyn is less likely than WT to form cytoplasmic inclusions and more likely to oligomerize.

In H4 glioma cells, A30P aSyn is more likely than WT to be secreted.

In LUHMES, A30P aSyn does not induce DNA damage and miROS as much as WT aSyn, but does induce similar Golgi fragmentation.

Question 13

Where is the SNCA gene located, how many amino acids are in the protein, and what is its weight?

Answer 13

On chromosome 4 (4q22.1). 140 AAs, 14 kDa.

Question 14

Name 9 functions of aSyn, in the following categories: membrane-related, signalling, metabolic, and survival.

Answer 14

Membrane-related:
1. SNARE complex assembly & vesicle trafficking
2. Maintains PUFA levels at cell membranes
3. Antioxidant (monomeric aSyn binds lipid membranes)
4. Binds membrane-bound GPCRs, initiating cell signalling

Signalling:
5. Modulates calmodulin activity (2nd messenger for GRK5 kinase)
6. Chaperone with homology to 14-3-3 (phosphorylation of ERK, BAD, others…affecting neuronal differentiation)

Metabolic:
7. Regulation of glucose levels (glucose uptake, insulin secretion)
8. Dopamine biosynthesis regulation (inhibits TH, resulting in less dopamine; aSyn aggregation = LoF and too much dopamine)

Survival:
9. Suppression of apoptosis in DAn

Question 15

Name two functions of aSyn which would likely be disrupted by the A30P mutation, and link these concepts to the results of Chapter 2 and [[Paiva et al. 2017]].

Answer 15

1. Membrane-binding –> could affect PUFA levels, lipid antioxidant effect, vesicle transport, membrane-bound GPCR signalling. Increased oxidative stress levels, reduced cell-cell communication, and reduced signalling cascades could indirectly affect DNAm/expr
2. Increased nuclear localization –> could bind DNA and histone modifiers, relating to altered gene expression seen to a greater extent in A30P than in WT cells.
   Overall LoF at membranes/cytosol and GoF in nucleus would alter molecular impact of A30P aSyn as compared with WT aSyn.

Question 16

How do LRRK2 and PRKN initiate cellular disturbances that could result in aSyn aggregation?

Answer 16

LRRK2 interaction results in aSyn S129P, increased oligomerization
PRKN interaction attracts tubulin deposition, leading to cytoskeletal dysfunction

Question 17

How can aSyn both increase and decrease dopamine levels in neurons?

Answer 17

Decrease: inhibits TH
Increase: when aggregated, LoF for TH inhibition

Question 18

How does aSyn affect cell signalling (4 ways)?

Answer 18

1. Binds membrane-bound GPCRs
2. Modulates activity of calmodulin (2nd messenger)
3. Has homology to 14-3-3 and can bind its targets, influencing neuronal differentiation (ERK/MAPK signalling, growth factors)
4. Inhibits phospholipase D, responsible for producing PA (phosphatidic acid) 2nd messenger

Question 19

Compare and contrast mechanisms of cytosolic and nuclear aSyn toxicity.

Answer 19

Cytosolic: Lewy bodies = LoF of normal aSyn activity; toxic GoF, overwhelming Ub-protease system. Associated with Golgi fragmentation and DNA damage.
Nuclear: deregulation of transcription and epigenetic marks (i.e., binding DNA, H3, HATs)

Question 20

Describe 5 ways by which A30P aSyn differs from WT aSyn (structure/function).

Answer 20

1. Cannot bind membranes
2. Increased nuclear localization (through RA shuttling)
3. Less likely to misfold/aggregate (slower fibrillization rate)
4. Interferes with actin polymerization more than WT
5. Interferes with autophagy

Question 21

What is colour correction and how is it conducted in GenomeStudio?

Answer 21

Normalizes the intensities of red/green colour channels for all probes based on intensities of control probes for each colour.

Question 22

What is background subtraction and how is it conducted in GenomeStudio?

Answer 22

Background signal = 5th percentile of negative control probes, calculated separately in red and green channels. Subtracted from all probes in a channel after detection p-value calculation (probe intensity is constrained to be 0 or higher).

Question 23

What are the 17 types of EPIC array control probes (7 categories) and how is their performance monitored via GenomeStudio?

Answer 23

In GenomeStudio, assessed visually in Controls Dashboard using scatterplots.
Staining (red/green), extension (red/green), restoration (bcDNA quality), hybridization (of synthetic target), target removal (stripping after base extension), bisulfite conversion (I and II, red and green), specificity (I and II), nonpolymorphic (overall assay performance, for each of the 4 nucleotides)

Question 24

Summarize the 6 probe filtering checks applied to the LUHMES dataset and which filters were applied/skipped.

Answer 24

Applied separately in BS and oxBS data.
1. Remove SNP probes (59)
2. Remove X-hybridizing probes (~43K, >47bp homology to off-target site)
3. Kept polymorphic probes (clonal cell line)
4. Kept sex probes (clonal cell line)
5. Probes with >1% of samples with detection p-value > 0.05 (wateRmelon default)
6. Probes with >5% of samples with bead count < 3 (wateRmelon default)

Question 25

Approximately how many cross-hybridizing probes are present on the EPIC array and how are they defined?

Answer 25

~43K, >47bp homology to an off-target site

Question 26

Which categories of probes were not removed in the analysis of the LUHMES data that typically would be removed in other human EWAS, and why?

Answer 26

Kept polymorphic and sex probes due to clonal cell line

Question 27

How were the cutoffs for probe detection p-value and bead count determined in the LUHMES dataset?

Answer 27

wateRmelon defaults (detP: probes with >1% of samples with detP > 0.05 [1 sample in LUHMES]; bead count: probes with >5% of samples with bead count < 3 [2 samples in LUHMES])

Question 28

What criteria was used to identify outlier samples for removal in the LUHMES dataset?

Answer 28

wateRmelon pfilter: samples where >5% of probes have detP > 0.05 (up from default 1% because of oxBS conversion; none flagged)

Question 29

What is dasen normalization (d/s/n?), how does it work, and why was this approach used in the LUHMES dataset?

Answer 29

Was used to avoid BMIQ, which assumes a 3-state beta-mixture model (unmethylated, methylated, hemimethylated) that is not true when considering mixed mC/hmC signal in BS data from neurons.
dasen: correct for difference in type I/II intensities, within M/U. d = type I/II intensity difference bg adjustment, s = quantile normalization of M/U intensities applied separately to type I/II probes, n = no dye bias correction
Applied separately to BS/oxBS data in LUHMES dataset.

Question 30

Describe how PCs are calculated and how PCA can be used in DNAm data QC.

Answer 30

PCA = a form of dimension reduction, transforming data into PCs that explain the overall variance.
1. beta values for each CpG are centered and scaled (transformed so they are on the same range)
2. covariance matrix constructed to ID correlations between CpGs
3. linear combinations of uncorrelated CpGs (PCs) are calculated from covariance matrix (eigenvector = linear combination, eigenvalue = eigenvector’s coefficient (% explained variance))
In DNAm QC, can cluster samples along PCs to identify patterns and outliers, and can correlate PCs with known variables to identify biological/technical signal in the data.

Question 31

How does ComBat work?

Answer 31

empirical Bayes framework used to adjust data for batch effects. Used M-values in LUHMES since they have a more constant stdev across a range of values than betas do (homoskedastic), so L/S estimate may be more accurate.
1. standardize M-value mean/variance across CpGs
2. assume a model for mean and variance of M-values within each batch
3. adjusts each batch to meet model specifications such that mean and variance are standardized across the batches
4. assuming batch effect is similar across CpGs, mean/variance estimates within a batch are shrunk toward the mean across all batches

Question 32

How did I decide which batch variables in the LUHMES dataset to correct for with ComBat, and in what order to do so?

Answer 32

PCs 1-3 correlated with chip, chip position, collection date, and passage. Collection date and passage were related to each other so just chose one (passage) to correct. Sequentially correction from the most to least correlation with DNAm PCs, separately in BS/oxBS (chip or row first, then passage).

Question 33

At what stage was DNAhm computed in LUHMES data preprocessing, and what was the detection cutoff?

Answer 33

After separately probe filtering, normalization, and batch correction of BS/oxBS data. 95% quantile of negative hmC values was 0.036.

Question 34

How is empirical Bayes estimation incorporated into linear modelling with limma?

Answer 34

The same model is fit to each CpG, but information about mean/variance across all CpGs is “borrowed” such that mean/variance estimates for individuals CpGs are shrunk toward the average. This results in a more robust estimation of variance when sample size is small.
The variance estimations for each CpG are weighted, so more variable and less variable CpGs can be tested separately.

Question 35

Why should limma NOT be used with variability-filtered DNAm data?

Answer 35

Filtering the CpGs prior to limma could result in a biased/artifically high estimation of variance across CpGs, affecting the eBayes prior/shrinkage.

Question 36

How is significance at each CpG determined in limma, and how is multiple test correction accounted for?

Answer 36

A moderated t-test is applied at each CpG, which borrows variance estimation across CpGs.
BH adjustment applied.

Question 37

Describe the Benjamini-Hochberg procedure for controlling FDR.

Answer 37

1. P-values ranked in ascending order (smallest to largest)
2. p-value rank/# of tests x FDR = BH critical value
3. the largest p-value smaller than the corresponding critical value is considered significant, and all p-values smaller than this

Question 38

Why was a delta beta cutoff of 0.05 selected for the LUHMES study?

Answer 38

In a DNAm titration experiment on the 27K array (Du et al. 2010), 0.05 beta value cutoff achieved good tradeoff between TP/FP rate. It is also larger than technical noise on the array (typically ~2-3%).

Question 39

How is significance determined in ermineR over-representation analysis?

Answer 39

Hypergeometric test (one-tailed Fisher’s exact): probability of drawing k or more successes without replacement from a population

Question 40

How are multifunctionality and multiple testing corrected for in ermineR over-representation analysis?

Answer 40

Algorithm removes the most multifunctional genes and recalculates GO enrichments, iteratively until either enrichments are maximally sensitive to gene removal or until half of genes have been removed.
FDR correction is applied to results.

Question 41

Describe the 6 main steps in LUHMES H3K4me1 ChIP-seq library preparation.

Answer 41

1. Cross-link chromatin and DNA with formaldehyde in culture dishes (create covalent bonds)
2. Shear DNA
3. IP with antibody to H3K4me1
4. Use magnetic beads to extract DNA fragments
5. Reverse crosslinking
6. Library preparation (adapter ligation, PCR amplification) and sequencing

Question 42

How was the LUHMES H3K4me1 ChIP-seq data pre-processed? Describe quality assessment, alignment, and read filtering.

Answer 42

1. Quality check with FastQC
2. Aligned to hg38 (not hg19) using Bowtie2
3. Filtered for MAPQ > 30? This is typically done and was done for hippEE project

Question 43

How was MACS2 used to call LUHMES H3K4me1 peaks and how were consensus peaks determined?

Answer 43

Broad peaks called using input (10% of sample after IP) as control. MACS models the peak locations based on chromatin-DNA contact sites (will be in between the tag locations of forward/reverse strands), and finds enriched tags relative to input by sampling 1000X and averaging distance between tags.

Consensus peaks present in 2 or more replicates

Question 44

Describe 5 general steps involved in RNA-seq library preparation.

Answer 44

1. RNA extraction (3 bio reps used for LUHMES)
2. Reverse transcription into cDNA
3. Adapter ligation
4. PCR amplification
5. Sequencing

Question 45

Compare the QC steps and differential expression thresholds used for RNA-seq data in the LUHMES and mouse experiments.

Answer 45

• FastQC used to check sequencing quality (PCR bias, read quality, etc)
• Both data were aligned to the reference genome using STAR, with default parameters and allowing gapped alignment to account for differential splicing.
• Quality after alignment was checked with samtools and visualization in a genome browser tool in LUHMES and mice
• DESeq2 was used to normalize read counts to library size in both experiments
• LUHMES data was filtered for genes with mean read count > 10, mouse was filtered for mean read count > 50
• LUHMES DEGs were considered at padj < 0.05 and log2FC > 0.5, while mouse DEGs were considered at padj < 0.15 and log2FC > 0.3 (citing variability and small effect sizes in mouse brain RNA studies…)

Question 46

Explain how differential expression analysis using DESeq2 is conducted.

Answer 46

1. Input matrix of library size-normalized, filtered read counts
2. Bayesian estimation of dispersion (variation in read counts between replicates) for each gene, shrunk toward the mean dispersion across the experiment
3. GLM + Wald test for differential expression at each gene (does including the covariate significantly contribute to the model?), accounting for experiment-wide dispersion factor
4. BH multiple test correction