Today we cover

  • electrophoresis — agarose, polyacrylamide, DGGE, Southern blotting;
  • allozymes — the first molecular markers, and why they lost;
  • PCR — what it needs, melting temperature, qPCR, RT-qPCR, ddPCR;
  • Sanger sequencing — terminators, capillaries, reading a gel;
  • fragment length polymorphisms — RFLP, AFLP, RAPD, microsatellites, ISSR.

Electrophoresis

Electrophoresis principle

  • Gel electrophoresis is a simple, rapid, and sensitive analytical technique for separating charged particles.
  • Gels are porous; the size of the pores relative to the molecule determines whether — and how easily — it passes.
  • Separation therefore depends on charge, size, and shape.
  • Most biomolecules carry a net charge at any pH other than their isoelectric point.

Two gel materials, two apparatus types

Horizontal (submarine) — agarose, run under buffer; the standard for nucleic acids.

Vertical — polyacrylamide, used mostly for proteins and for high-resolution sizing.

Agarose electrophoresis

  • Agarose is a natural colloid isolated from seaweed — a linear polysaccharide of repeating agarobiose units (alternating 3,6-anhydrogalactose and galactose).
  • Larger pore size, so it suits molecules above ~200 kDa.
  • Used for both proteins and nucleic acids.
  • Typical working concentration: 1–3%.

Longer molecules stay near the well; shorter ones run further towards the anode.

Seeing the DNA — gel dyes

DNA dyes are what make the separation visible, during or after the run. They differ in sensitivity, toxicity, and compatible detection (UV or blue light).

Dye Notes
Ethidium bromide (EtBr) classic, very sensitive, intercalating — mutagenic, UV
SYBR Green high sensitivity, blue-light compatible
GelGreen designed as a safer EtBr replacement
SYBR Safe blue light, lower toxicity, routine teaching-lab choice
Methylene blue insensitive but non-toxic and visible in daylight

Hoiatus

“Safer” is not “safe”. Every intercalating dye binds DNA — including yours.

Polyacrylamide gels

  • Chains of acrylamide monomers cross-linked with bisacrylamide (N,N′-methylenebisacrylamide).
  • Pore size and resolving power depend entirely on the acrylamide : bisacrylamide concentration — normally 5–25%.
  • Separates proteins of 5–250 kDa and polynucleotides from <5 to ~3000 bp.
  • SDS-PAGE coats proteins with a uniform negative charge, so they separate by mass alone.

Denaturing gradient gel electrophoresis (DGGE)

Polyacrylamide gels containing a gradient of chemical denaturants — urea and formamide.

Fragments of identical length but different sequence melt at different positions in the gradient and stop migrating there. A GC-rich clamp on one primer keeps the fragment from melting completely.

Southern blotting

  • An RFLP method that involves no PCR amplification.
  • Developed by Edward M. Southern (1975).
  • DNA fragments are separated by electrophoresis, transferred to a nylon membrane, and identified by labelled probe hybridisation.
  • The probe is complementary to the target and carries a marker for detection.

Codominant vs dominant markers

Codominant markers (a) discriminate homozygotes from heterozygotes — you see both alleles.

Dominant markers (b) do not: presence of a band means “at least one copy”, so AA and Aa are indistinguishable.

Tähtis

This single property decides which population-genetic statistics you are allowed to calculate.

Allozymes

The allozyme era

Allozymes — proteins differing in electrophoretic mobility as a result of allelic differences in the protein sequence, which ultimately reflect variation in the corresponding DNA sequence.

Population genetics entered the molecular age with two 1966 papers describing electrophoretically detectable variation in:

  1. Drosophila pseudoobscura — Lewontin and Hubby 1966;
  2. humans — Harris 1966.

Genetic diversity was measured in two ways:

  1. the average proportion of loci heterozygous in an individual — heterozygosity or gene diversity, H;
  2. the average proportion of loci polymorphic in the population — gene polymorphism, P.

Allozymes and isozymes

  • Allozymes (alloenzymes) — variant forms of an enzyme that differ structurally but not functionally, coded by different alleles at the same locus.
  • Isozymes — enzymes performing the same function but coded by genes at different loci.
  • Alloenzymes are highly conserved in function across phyla and kingdoms, so phylogeneticists used them as molecular markers for evolutionary relationships.
  • Because they differ in structure, they can be separated by electrophoresis.

Losing ground to DNA markers

Since the 1960s allozymes played an ongoing role in animal and plant population studies — and then lost the field to DNA markers.

By 2008, microsatellites carried over 80% of published studies and allozymes under 10%.

Genome sizes compared

The human genome contains 1000× as many nucleotide pairs as a typical bacterium, 20× as many genes, and about 10,000× as much noncoding DNA.

~98.5% of the human genome is noncoding, against 11% for E. coli.

Genome size and number of genes

Genome size is not gene number

Ten thousand-fold more DNA, the same order of magnitude of genes.

PCR

DNA replication — the reason PCR works

  • DNA polymerases copy each of the two strands independently, always 5′ → 3′.
  • The strands must be unwound at a replication fork.
  • Polymerases cannot initiate synthesis without a primer — in cells, a short RNA stretch laid down by a primase.

What a PCR reaction needs

  1. Template DNA — contains the target region the primers bind and the polymerase copies.
  2. Primers — short single-stranded DNA, usually 18–30 nt, complementary to the sequences flanking the target. Forward primer binds the 5′ end of the target; reverse primer binds the 3′ end of the complementary strand.
  3. DNA polymerase — usually Taq, from Thermus aquaticus: heat-stable, so it survives the denaturation steps.
  4. dNTPs — dATP, dTTP, dCTP, dGTP: the building blocks.
  5. Buffer and Mg²⁺ — the cofactor the polymerase cannot work without.

What a PCR reaction needs

8. Thermocycler — precisely controls the temperature changes, cycling through three steps:

Step Temperature What happens
Denaturation 94–98 °C double-stranded DNA separates into single strands
Annealing 50–65 °C primers bind their complementary sequences
Extension ~72 °C polymerase adds dNTPs to the primed 3′ ends

9. Additives (optional) — DMSO, glycerol, or betaine help with GC-rich templates and secondary structure, improving efficiency and specificity under difficult conditions.

One cycle, drawn out

Melting temperature (Tm)

  • The temperature at which 50% of the DNA is denatured.
  • G–C pairs have three hydrogen bonds; A–T pairs only two.
  • Matters for primers (PCR) and for probes (hybridisation).

Rule of thumb: set the annealing temperature Ta ≈ Tm − 3…5 °C.

Primer arithmetic

primers <- c(`16S-27F`  = "AGAGTTTGATCCTGGCTCAG",
             `16S-1492R`= "TACGGCTACCTTGTTACGACTT",
             `ITS1`     = "TCCGTAGGTGAACCTGCGG")

tibble::tibble(
  primer = names(primers),
  length = nchar(primers),
  gc     = sapply(primers, gc_pct),
  tm_wallace = sapply(primers, tm_wallace),   # 2(A+T) + 4(G+C)
  tm_gc      = sapply(primers, tm_gc),        # 64.9 + 41(GC - 16.4)/length
  ta_suggested = sapply(primers, tm_gc) - 5
) |> knitr::kable(digits = 1)
primer length gc tm_wallace tm_gc ta_suggested
16S-27F 20 50.0 60 51.8 46.8
16S-1492R 22 45.5 64 53.0 48.0
ITS1 19 63.2 62 55.4 50.4

Two rules of thumb, two answers — which is why every new primer pair still needs a gradient PCR.

PCR is exponential

tibble::tibble(cycle = c(10, 20, 30, 40),
               copies = 2^c(10, 20, 30, 40)) |>
  knitr::kable(format.args = list(big.mark = " "))
cycle copies
10 1.024000e+03
20 1.048576e+06
30 1.073742e+09
40 1.099512e+12

2⁴⁰ ≈ 10¹² copies — in theory. In practice reagents run out and the curve plateaus.

Evolution of thermal cyclers

An early PCR machine: three water baths.

1987 — the first commercial thermal cycler, the TC1.

A lab-on-chip cassette for PCR.

qPCR

Quantitative PCR measures DNA during amplification.

A DNA dye (e.g. SYBR Green) or a sequence-specific probe is added and fluorescence is measured every cycle.

The cycle at which signal crosses the threshold (Cq) is proportional to the log of starting template.

RT-qPCR

Quantitative reverse transcription PCR is used when the starting material is RNA: reverse transcriptase makes cDNA first, and the qPCR runs on that.

Märkus

“RT” is reverse transcription, not real time. A real-time PCR from DNA is qPCR; from RNA it is RT-qPCR.

qPCR: multiplexing

Combining different probes and fluorophores, with different emission filters and detectors, lets several targets be quantified in one well.

Droplet digital PCR (ddPCR)

The reaction is split into ~20,000 droplets, each an independent nanoreactor. After endpoint PCR, droplets are counted as positive or negative and the starting concentration follows from the Poisson distribution — no standard curve.

Right now: considerably more expensive than qPCR.

ddPCR: multiplexing

Two targets, two dyes (FAM and HEX), read as 2D amplitude clusters — each droplet is assigned to one of four classes.

Sanger sequencing

Sanger sequencing (first generation)

“Terminator” nucleotides: a ddNTP has no 3′-OH, so the chain stops the moment one is incorporated. A population of chains of every possible length is the result.

Capillary electrophoresis

Capillary gel electrophoresis with a different fluorescent dye linked to each terminator nucleotide.

One capillary, one laser, one detector — and the four-colour trace we still read today.

Decoding the gel

Four lanes, one per terminator. Read the ladder from the bottom (shortest fragment) upwards — that gives the newly synthesised strand 5′ → 3′:

Your sequence: 5′-TCGAATGTCAGG-3′

Template sequence: 3′-AGCTTACAGTCC-5′

Nõuanne

The gel gives you the synthesised strand. The template is its reverse complement — a step people routinely forget under exam conditions.

Decode this gel

ddATP  -
ddTTP  -
ddGTP  -
ddCTP  -

Write out the sequence 5′ → 3′, then give the template strand.

Fragment length polymorphisms

Restriction nucleases

Enzymes that cut the double helix at sites defined by the local nucleotide sequence. HpaI leaves blunt ends; EcoRI, HindIII, and PstI cut in a staggered way, leaving cohesive ends. Because the recognition site is only 4–6 bp, any given DNA yields a consistent set of fragments.

RFLP — restriction fragment length polymorphism

  • Uses the properties of restriction nucleases; requires separation and sizing of fragments by electrophoresis.
  • Mutations destroy or create cutting sites — that is the polymorphism.
  • RFLP can start from:
    1. total DNA extracted from samples;
    2. specific sequences first amplified by PCR.

PCR-RFLP

Amplifying first and digesting afterwards adds:

  • sensitivity — tiny amounts of starting material suffice;
  • simplicity — no blotting, no probes, no radioactivity.

A single restriction digest of a PCR product remains one of the cheapest species-diagnostic assays available.

AFLP — amplified fragment length polymorphism

Requires no sequence information from the study organism.

  1. Digest sample DNA with two restriction enzymes giving different overlapping sticky ends.
  2. Ligate double-stranded linkers with compatible sticky ends.
  3. PCR with primers matching the linker sequences plus 1–3 arbitrary selective nucleotides at the 3′ end.

The whole procedure is a selection in which a small, essentially random proportion of the starting DNA is amplified.

RAPD — random amplification of polymorphic DNA

A PCR in which the amplified segments are random: several arbitrary short primers (8–12 nt) are run against a large template of genomic DNA, in the hope that fragments amplify.

Resolving the pattern gives a semi-unique profile.

Hoiatus

Dominant, and notoriously sensitive to template quality and cycler model — reproducibility between laboratories is the weak point.

Microsatellite markers

  • Microsatellites (1–10 nt motifs) and minisatellites (>10 nt) are subcategories of tandem repeats, which together with interspersed repeats make up the genome’s repetitive fraction.
  • They matter because they are unstable: mutation rates of 10⁻³–10⁻⁶ per cell generation, up to ten orders of magnitude above point mutations.
  • SSR, STR, SSLP — the same thing under different names — occur in prokaryotes and eukaryotes, especially in eukaryotic euchromatin, in coding and noncoding, nuclear and organellar DNA.
  • They are not randomly distributed along the genome.
  • Mutations that escape mismatch repair create new alleles, so an SSR locus carries many alleles — making it more informative than a SNP.

Microsatellite markers

For more than a decade microsatellite loci were the marker of choice.

Traditionally, alleles at one locus were:

  • PCR amplified with locus-specific primers in the flanking regions;
  • sized by electrophoresis — the allele is the length.

Replication slippage

Replication slippage is the mutational mechanism behind STRs: the nascent strand dissociates and re-anneals out of register, giving contraction or expansion by whole repeat units.

The high mutation rate is exactly why many microsatellite loci show high heterozygosity — and why they make useful markers.

ISSR — inter-simple sequence repeats

  • Highly polymorphic, relatively easy to develop, and inexpensive.
  • A PCR-based system using microsatellite loci across the genome — it amplifies the variation found between two proximate microsatellite loci with the same repeat.
  • No sequence information needed for the study species.
  • Weaknesses: poor reproducibility and time-consuming data analysis.

RADseq — where this is all heading

Restriction-site associated DNA sequencing keeps the restriction enzyme but replaces the gel with a sequencer: digest, size-select, sequence the fragments flanking each cut site, and call SNPs across thousands of loci at once.

The logic of RFLP, with the throughput of NGS — which is Lecture 3.

Choosing a marker

Marker Sequence info needed Dominance Loci Main weakness
Allozymes no codominant tens sees only charge-changing variation
RFLP some codominant few needs a lot of clean DNA
RAPD no dominant many reproducibility
AFLP no dominant hundreds dominant, complex profiles
Microsatellites yes (flanks) codominant tens development cost, homoplasy
SNP / RADseq no (with RADseq) codominant thousands computation, cost per sample

Check yourself

Why does agarose separate large molecules better than polyacrylamide?

Its pore size is much larger, so molecules above ~200 kDa (or above a few kb of DNA) still migrate instead of being trapped at the well.

Why must Ta be set below Tm?

At Tm half the duplex is melted; primers need a temperature where they stay annealed long enough for the polymerase to extend them. Too far below and they anneal to near-matches, giving non-specific product.

Why are AFLP and RAPD dominant, but microsatellites codominant?

AFLP and RAPD score presence or absence of a band, so a heterozygote looks like a dominant homozygote. A microsatellite locus is amplified with locus-specific primers and each allele has its own length, so both alleles are visible.

Why does ddPCR not need a standard curve?

It counts positive and negative droplets at endpoint and converts that fraction to a concentration through the Poisson distribution — an absolute measurement rather than a relative one.

Thank you

Tähtis

peeter.laas@ut.ee · peeter.laas@taltech.ee

Next lecture: next-generation sequencing — second generation, third generation, and what they made possible in molecular ecology.