FS 160 - Lecture 9 - Capillary Electrophoresis PDF

Preview

Summary

ONLINE CLASS.
Instructor: Steven Lee...

Description

FS 160 – Lecture 9 – Capillary Electrophoresis I.

Overview of STR Analysis by Capillary Electrophoresis (CE) a. Humans have 23 pairs of chromosomes b. Early methods i. RFLP ii. Silver stained slab gel iii. Labs were like dark rooms c. Advantages i. Automated ii. Does not require large amounts of sample iii. Can run multiplex 1. Separate by size or color d. Injection i. Electrokinetic injection process 1. Requires formamide and water 2. Most critical process 3. Sample stacking a. Use voltage to inject sample ii. Improved by diluting sample iii. Methods 1. Hydrodynamic 2. Electrokinetic iv. Electrophoresis Theory 1. Increase voltage  increase power  creates heat  viscosity of polymers drops  separation changes a. Keep temperature constant in air-conditioned room! 2. Electrophoretic velocity a. Ions move through gel faster at higher voltage b. Small ions with high charge move fastest 3. As size increases, so does the charge

v. What are sieving buffers?

1. Very similar to polyacrylamide 2. Not gels since they flow 3. Known as entangled linear polymers 4. Many common applications vi. Electric field strength and shape of DNA molecule

e. Separation i. Capillary 1. 50 micrometer fused silica ii. POP-4 polymer 1. Composed of polydimethyl acrylamide, which is linear 2. Different from gels! iii. Buffer 1. TAPS pH 8.0 iv. Denaturants 1. Urea 2. Pyrolidinone 3. Important because DNA needs to be kept single-stranded v. Issues 1. Electrophoresis buffer a. Urea for denaturing and viscosity b. Buffer for consistent pH c. Pyrolidinone for denaturing DNA d. EDTA for DNA stability and chelating metals 2. Polymer solution a. Entangled to separate DNA b. High molecular weight for good resolution c. Minimum concentration/viscosity for easy refilling i. POP-4 ii. POP-6 iii. Etc. 3. Run temperature a. 60° C helps reduce secondary structure, which improves precision

i. Temperature control affects DNA sizing 4. Electric field a. Affects orientation and diffusion of DNA vi. Stacking effects

vii. Ion mobility effects 1. [DNAinj] = E(πr2)[DNAsam](μep + μeof) a. Assuming no interfering ions are present b. Cl- ions and other interferents will compete with DNA 2. [DNAinj] = [DNAinj]/[other ionsinj] a. Ions such as Cl- have higher charge per mass ratio b. μep is higher 3. Golden Gate Effect a. Attributed to poor formamide b. Results will see many “spikes” i. Broad peaks and extra bands c. Problem can be solved by dissolving sample in pure water instead of formamide and denatured i. Indicates degraded formamide 4. Shadow peaks a. Results from bad formamide, incomplete denaturation, or rehybridization b. dsDNA migrates faster than ssDNA c. Extra peaks appear ahead of main peaks d. Most visible in size standard, but can appear in other dye lanes viii. Measuring formamide conductivity 1. Measure bottle when it arrives 2. Purchase good quality formamide and immediately pipette it out into small tubes with or without ROX already added

a. Freeze tubes 3. Do not ever open cold bottle of formamide a. Water will condense inside and aid in formation of conductive formic acid 4. Run heat denatured sample in distilled water a. Looks different? Formamide is contaminated f.

Detection i. Done with laser induced fluorescence ii. Provides exquisite sensitivity and specificity iii. Fluorescent dyes with excitation and emission traits 1. Dyes are added to DNA during synthesis, but can also fall off, yielding dye blobs 2. Can interfere with each other, a major problem in spectroanalysis a. Solution  CCD with defined virtual filters produced by assigning certain pixels iv. Matrix calculations 1. Solve xyzw for each dye individually to determine dye contribution for any mixture

v. Issues with optical system 1. Argon ion lasers outgas and eventually lose intensity a. Must take note of laser power and monitor over time 2. Fluorescence expression a. If = I0kεdCφ where I0 = changes in input intensity [sic] = changes in capillary diameter k = cleanliness of capillary optics b. Baseline noise is more affected by detector c. Monitor ladder intensity as a good quality control step II.

Setting Thresholds a. Thresholds for ABI 310/3100 (determining true allele)

i. Set analytical threshold, stochastic threshold, limit of linearity, and minimum peak height threshold b. Analytical figures of merit i. detection limit (analytical threshold): three times the standard deviation of noise 1. estimated using two times the peak to peak noise a. approximately 35-50 RFUs b. peaks below that level may be random noise ii. stochastic threshold: level of DNA below which significant change of allele dropout can occur 1. set high enough that heterozygous peak will produce companion allele in gray zone between stochastic and analytical threshold a. 150-200 RFUs iii. limit of linearity: level of DNA above which enhanced pull-up, flat top peaks, and elevated stutter occurs 1. Determined by examining relationship between input DNA 2. Fluorescence signal varies a. Approximately 4500 RFUs for ABI 310, 3500 for ABI 3100, more than 20,000 for ABI 3500 iv. heterozygous peak ratio: minimum peak height ratio expected for clean, single source DNA sample at particular concentration 1. Typically 60-70% v. Visual representations

vi. Scientific reasoning behind concept of analytical threshold/limit of detection 1. Issue of reliability 2. For peak intensity below LOD, there is a very real chance that such a signal is result of random fluctuation 3. LOD = 2Npp –or- LOD = 2SDn 4. Levels typically set high to avoid constantly resetting thresholds

vii. Current thresholds 1. Analytical  three times baseline noise plus additional amount to be conservative 2. Stochastic  set to avoid presence of peak height ratios below 60% 3. Gray zone  between two threshold permits detection of second homozygous peak c. Sensitivity studies i. Profiler Plus 1. Peak height variation increases with concentration 2. Difficult to assess quantity of DNA solely by peak height ii. Scientific reasoning behind LOQ/stoichastic threshold 1. Peak intensity below LOQ, means significant variation in height from one sample to next 2. Interpreting data below stochastic threshold presents problem of allele dropout due to variation 3. Rely on peak heights to detect major and minor profiles 4. Be careful when calculating statistics that no heterozygous alleles are dropped 5. Gray zone predicated on minimum peak height ratio (PHR) a. Single peak in gay zone is considered unreliable i. Can be heterozygous and partner allele dropped in noise iii. Alternative procedure 1. Choose low level 2. Amplify two or more samples at range of concentrations (1.0 nanograms – 0.005 nanograms) multiple times a. Score intensity 3. Stochastic limit is intensity at which half alleles have intensity above this value and half are below iv. Sensitivity issues

1. Improved STR multiplexes have better buffers, more mini STRs and increased sensitivity a. Problem with this is stochastic amplification i. Exists regardless of sensitivity of detection and manifests as peak imbalance, enhanced stutter, and peak dropout ii. New kits and instruments detect better, but PCR has not changed 1. Therefore, new thresholds need to be set iii. Set analytical threshold based on method, not instrument 2. Most validation studies are performed on pristine samples derived from clean sources 3. DNA degradation can result in dropped alleles from larger sized amplicons 4. DNA inhibition will result in dropped alleles from any location and effects are difficult to predict 5. Inhibition and degradation can produce stochastic effects (peak balance issues and allele dropout) 6. Bottom line: a. Low signal levels are bad b. Relying on signal level to determine DNA quantity can be misleading d. Interpretation of low level DNA i. Fuzzy logic 1. Capillary electrophoresis is dynamic process 2. Sensitivity varies with: a. Allele size b. Injection solvent c. Input DNA d. Instrument factors e. Presence of PCR inhibitors f. Gel matrix 3. Therefore, interpretation must be conservative and data from studies yields guidelines, not rules 4. Interpretation and significance cannot be dissociated from overall facts of case ii. Why examine low level data at all? 1. Touch DNA can be powerful lead in criminal investigation 2. Detection of presence of low level mixtures 3. Clues to presence of inhibited samples or poor injections 4. Aids in determination if suspect is excluded as contributor iii. Guidelines 1. Low copy DNA is not just more cycles a. Stochastic effects can occur anytime levels of input DNA are low

2. Allele should not be scored (considered real) unless present at least twice in replicate samples a. Usually three are performed 3. Extremely sterile environments are required for PCR setup to avoid contamination from lab personnel or other sources a. Personnel must be typed for contamination events 4. Potential for contamination from DNA not related to events in the case must always be considered 5. Guidelines for minimum number of heterozygous alleles should be considered 6. If one is going to look at low level samples, at least measure them by multicopy qPCR III.

Conclusions a. Capillary electrophoresis based DNA analyses are complex i. Separations affected by polymer length, concentration, and field strength ii. Injections vary greatly with salt content and PCR product quality iii. Detectors require careful monitoring for pull-up and intensity b. Multiplicity of instrument thresholds i. Analytical thresholds are based on standard deviation of noise ii. Stochastic thresholds are based on fundamental aspects of PCR reaction resulting in peak imbalances iii. Limits of linearity affect ultimate peak height, stutter, and pullup iv. Remember that thresholds are guidelines 1. Not rules, so be smart and keep things within context of case...