Chapter 19: Variable Number Tandem Repeat Profiling

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to the Deep Dive.

Today, we are really getting into the nuts and bolts, charting a course straight into the history of forensic biology.

Yeah, we're looking at the, foundational techniques.

We're focusing entirely on variable number tandem repeat profiling, VNTRs for short.

Exactly.

This is crucial because it tracks how the technology started really way before PCR was everywhere.

Right.

We're talking about these VNTRs.

They're also called mini satellites.

Think of them as specific DNA units repeated one after another, kind of like a molecular stutter.

And the key difference from later techniques is the length, right?

These repeat units themselves can be quite long and they form these arrays that can stretch for several kilobases.

It's chunky DNA we're talking about.

And what really matters for forensics, for telling people apart, is the variation.

The genotype isn't about the sequence inside the repeat, but the number of times it's repeated at that spot.

That specific number.

That's what makes your profile unique, or at least highly individual.

And that variability is really the engine behind whole technique's power in forensics.

Absolutely.

For this to work forensically,

these VNTR spots, or loci, they have to be super variable in the population that's highly polymorphic.

Lots of different lengths out there.

So many different genotypes possible.

Exactly.

And critically, they also need to be inherited independently.

You don't want loci that are linked together on the same chromosome, otherwise you're not getting independent points of comparison.

Usually they're on different chromosomes.

Okay, so you combine the power of that variation across multiple independent loci, and that gets you to the core metric, doesn't it?

The population match probability, or PM?

That's the goal, yes.

A really, really low PM.

You want the odds of two random, unrelated people matching by chance to be incredibly small.

Vanishingly small.

Right.

With these early VNTR tests, when they analyzed multiple loci successfully, LAMS could actually get PM values down to like 10 to the minus 12.

Wow.

Which is an incredibly powerful statement for either excluding someone or suggesting an inclusion.

Okay, so let's unpack the first major technique that actually made this kind of analysis possible.

Restriction Fragment Length Polymorphism, RFLP.

RFLP, yeah, the original method.

It was complex, took days sometimes.

So there are six main steps involved.

Let's walk through them, but really focus on why each step was necessary.

Okay, so RFLP is basically about cutting DNA at specific points and then measuring the lengths of the pieces that contain the VNTRs.

Step one is pretty straightforward.

You prepare your genomic DNA sample, get it clean.

Simple enough, then step two.

Step two is where the real action starts.

Restriction Endonuclease Digestion.

You use these highly specific enzymes to top up the DNA.

The molecular scissors we hear about, they recognize very specific short sequences, the restriction sites, and SNP.

Exactly, and the choice of enzyme here is absolutely critical.

Why is that?

Well, the enzyme you pick must cut the DNA at sites that flank the VNTR core repeat region, so on either side of the variable.

Okay, not inside the repeats themselves.

Definitely not.

If it cuts inside the repeats, the whole length variation measurement is ruined.

The enzyme has to cut consistently outside that variable region.

Got it.

So once the DNA is chopped into fragments and the length of those fragments depends on how many repeats are between the cutting sites.

What's step three?

Step three is separating them.

That's agarose gel electrophoresis.

You run an electric current through a gel and the DNA fragments move through it.

Smaller ones move faster, larger ones slower.

And since RFLP fragments are pretty big.

They're huge.

Yeah, often kilobases long.

So you needed these large, fairly low resolution agarose gels to get any separation at all.

And then we hit a really famous step.

Step four, the Southern Transfer.

Named after Sir Edwin Southern.

Yeah, it's a crucial step.

After you've separated the fragments on the gel, you basically need to transfer them onto something more stable and accessible.

Like a permanent record.

Kind of.

You denature the DNA on the gel, make it single stranded, and then use capillary action, basically like blotting paper, to transfer that pattern of DNA fragments onto a solid nylon membrane.

Then you lock it there, often using UV light.

Okay, the DNA pattern is now fixed on this durable membrane.

That sets us up for step five.

Hybridization with probes.

How do we actually find the specific VNTR fragments we care about among all the other DNA?

We use a probe.

It's a small piece of synthetic DNA that's been labeled, usually radioactively back then or later with chemicals.

This probe is designed to be complementary to the specific VNTR repeat sequence you're looking for.

So that's like a little magnet sticking only to the VNTR fragments.

Exactly.

It's a molecular homing beacon.

It hybridizes, or binds, only to its matching sequence among all the fragments stuck on the membrane.

And then the final step, step six, detection.

The probe has found its target.

How do they see it?

The classic method was autoradiography.

If you use a probe labeled with radioactive phosphorous, like P32, you'd lay the membrane against x -ray film in the dark.

And the radiation exposes the film.

Right where the probe bound.

So you develop the film and see dark bands corresponding to the location, and therefore the size, of your VNTR fragments.

Later, chemiluminescence methods, using enzyme -linked probes, became more common, but the principle was the same.

Visualize where the probe stuck.

Going back to that cutting step, step two, you mentioned the enzyme choice was critical.

The sources say HAVERY was really popular, especially in North America.

Why that specific one?

Well, HAVERY recognizes a really short sequence, just four base pairs, GGCC.

Okay.

Because it's so short, that sequence pops up much more frequently in the human genome, compared to sites for enzymes that need five or six bases, like say, INVI or PSDI.

Right.

So more frequent cutting sites means?

On average, the fragments generated, including the ones flanking the VNTRs, tend to be smaller.

And smaller fragments were just, well, easier to separate effectively on the standard agarose gels they were using.

They resolved better.

Made the analysis a bit cleaner.

Yes.

But perhaps the most important reason, forensically speaking, was its reliability.

HAI -THI's cutting activity isn't affected by a common type of natural DNA modification called methylation, which can sometimes block other enzymes.

Oh.

So it cuts consistently, regardless of those natural chemical tags in the DNA?

Precisely.

You needed results that were dependable and comparable between different samples and different labs.

Consistency was key for building trust, and eventually databases.

HAI -THI delivered that.

That need for consistency and dealing with complexity leads us nicely into the probes.

Step five again.

Initially, Sir Alec Jeffreys developed the multi -locus probe, the MLP.

That's what gave us the term DNA fingerprinting, right?

That's the one, back in 1984.

The MLP was clever.

It had a core sequence that was similar enough to bind to many different VNTR loci across the genome, all at the same time.

So instead of one or two bands?

You get this really complex, barcode -like pattern of maybe 15, 20 or more bands, highly individual -specific, fantastic for things like paternity testing or immigration cases where you had clean single -source samples.

But I can immediately see the problem for crime scenes.

What if you have a mixture of DNA, victim and attacker?

Exactly.

It was a massive disadvantage.

Trying to interpret a mixture with MLPs was, well, basically impossible.

You just got a confusing smear of overlapping bands.

You couldn't tell which band belonged to whom.

So that limitation had to be solved, which led to the single -locus probe, or SLP.

How did that change the game?

SLPs were the breakthrough for practical forensics.

They were designed very specifically to hybridize to the unique sequences, flanking just one VNTR locus at a time.

So targeting just one spot in the genome.

Right.

Suddenly, the complex barcode became a much simpler pattern.

You'd see one band.

If the person inherited the same length allele from both parents, that's a homozygote, or you'd see two distinct bands if they inherited different length alleles ahead of a zygote.

Much easier to interpret.

Much easier.

And crucially, this simplicity meant you could start to deconvolute mixtures, figure out which bands belonged to which contributor, and maybe even more importantly, you could measure the size of those one or two bands and assign them a value.

Which allowed for databasing.

Precisely.

You could create databases frequencies.

This SLP approach was the technology used in that landmark first criminal case solved by DNA.

The Leicestershire murders in the UK in the mid -80s.

That's the one.

The SLP profile developed from the crime scene evidence didn't match the initial suspect, a man named Richard Buckland.

He was exonerated.

And that led to that massive voluntary screening.

Yeah.

Thousands of men gave samples.

And eventually, they caught Colin Pitchfork, partly because he'd coerced a friend to give a sample for him.

When they tested Pitchfork directly, his SLP profile matched the semen from the victims.

A landmark case that really cemented SLP's role in forensic science.

Absolutely.

And SLPs also helped address the problem of limited evidence.

Because those nylon membranes from the southern blot were pretty robust.

You could reuse them.

You could.

Labs developed a process called probe stripping.

After they got the result for the first SLP, they'd wash the membrane at high temperature and with certain chemicals to remove that first probe.

Without destroying the DNA pattern fixed on there.

Mostly, yeah.

Then they could come back and hybridize it with a different SLP, targeting a second VNTR locus, then strip again, probe for a third, and so on.

It allowed them to get multi -locus data from a single, often small, initial sample.

It sounds incredibly painstaking.

You're doing multiple rounds of hybridization and detection on the same membrane?

It was.

Early forensic DNA work required a huge amount of skill, patience, and careful handling.

Okay, so let's dive into that complexity.

Given all these steps, enzymes, gels, probes, transfers, what could actually go wrong?

What were some major factors that could mess up an RFLP result?

Oh, 20.

Probably the biggest headache, especially with older or poorly preserved samples, was DNA degradation.

The DNA breaking down over time.

Exactly.

RFLP needs relatively large, intact DNA fragments.

Remember, sometimes kilobases long.

If the DNA starts to degrade, the longer pieces break down first.

So, thinking about a heterozygote profile, someone with one long VNTR allele and one short one.

If the DNA is degraded, that larger allele might break down so much that it falls below the limit of detection.

Your probe just doesn't find enough intact large fragments to bind to?

And the result?

Your two -band heterozygous profile suddenly looks like a single band from the shorter, more robust allele.

It looks like a homozygote.

Which is a disaster, forensically.

It's a false exclusion.

You'd incorrectly rule out a suspect who actually is the source of the DNA.

To try and catch this, labs would often run a quick yield gel first, just to get a rough idea of how intact the DNA was before starting the whole RFLP process.

Okay, so degradation was a major sample quality issue.

What about problems with the RFLP chemistry itself?

The enzymes?

Enzyme problems were common, too.

One issue was partial restriction digestion.

Maybe the enzyme wasn't active enough, or the incubation time was too short, or there were inhibitors in the sample.

So it doesn't cut everywhere it's supposed to.

Right.

You end up with a mix of correctly cut fragments and some larger fragments where one or more restriction sites were missed.

This could create extra unexpected bands, often appearing fainter, potentially confusing the interpretation, maybe making a single source look like a mixture.

Labs usually check for this with a test gel using a known DNA standard.

Any other enzyme issues?

Another one was star activity.

This happens if the digestion conditions are off, maybe the wrong buffer,

or too much enzyme, or incorrect salt concentration.

The enzyme starts to lose its specificity.

It gets sloppy and cuts where it shouldn't.

Kind of.

It might start recognizing and cutting sequences that are slightly different from its proper restriction site.

If one of these star activity sites happens to be inside the VNTR region or in the flanking sequence between the proper site and the VNTR, it would create extra smaller fragments.

Exactly.

You'd get unexpected bands showing up.

Again, muddying the interpretation.

And even if the chemistry worked perfectly, there were physical issues with the electrophoresis stuff, right?

Running the gel.

Absolutely.

A big one was simply the limit of separation resolution.

Because these RFLP fragments were so large and the agarose gels weren't super high resolution,

two alleles that only differed by one or maybe two repeat units might not separate clearly.

They'd run so close together they looked like one band.

Yeah, they could merge into a single thicker band.

Another potential false homozygote.

You'd miss the fact that there were actually two different alleles present.

And then there was the opposite problem almost.

Band shifting.

Right.

This was really tricky.

Sometimes, for various reasons,

maybe slight differences in salt concentration or DNA load across the gel,

identical DNA fragments in different lanes would migrate at exactly the same speed.

So the band for the suspect sample might appear slightly higher or lower on the gel than the band for the cram scene sample, even if they were genetically identical.

Precisely.

And that could lead you to mistakenly conclude they don't match.

Another potential false exclusion, when in fact they do.

Labs had to use internal size standards in every lane and complex statistical windows to account for this.

And one last simple one.

Bands just running off the end of the gel.

Yeah.

Especially the smaller alleles.

If you ran the gel for too long trying to separate the big fragments, the smallest fragments could literally run right off the bottom edge and be lost.

Again, potentially making a heterozygote look like a homozygote if the smaller allele was lost.

Wow.

So RFLP, while groundbreaking, was clearly fraught with potential pitfalls.

It needed large amounts of good quality DNA and very careful, time -consuming work.

Incredibly high stakes, yes.

Which is exactly why the field was pushing hard for something better, leading to the next major revolution.

Which brings us to Amplified Fragment Length Polymorphism, or AFLP.

The big change here was incorporating PCR, the plimmer's chain reaction.

That was the game changer.

PCR lets you make millions or billions of copies of a specific DNA target region.

How did that immediately help with the problems RFLP faced?

Two main ways.

First, sample quantity.

Because you could amplify the target, you needed way less starting DNA.

Picograms sometimes, instead of the micrograms needed for RFLP.

Second, sample quality.

AFLP targeted VNTRs with shorter repeat units and overall smaller allele sizes.

Usually less than a thousand base pairs or one kilobase.

And smaller fragments are more likely to survive degradation.

Much more likely.

So AFLP worked far better on the kind of degraded samples that were often found at crime scenes.

Samples that would have failed completely with RFLP.

Okay, so the key locus they often used for AFLP was called D1S80.

It has a 16 base pair repeat unit.

What else changed in the process besides adding PCR?

The separation method changed significantly.

Instead of those large, low resolution agarose gels used for RFLP.

They used something with higher resolution.

Yes, they switched to polyacrylamide gel electrophoresis or PAGE.

PAGE gels have a tighter matrix structure and can resolve DNA fragments that differ in size by just a few base pairs.

Sometimes even a single base pair.

And that higher resolution enabled a really critical improvement in how results were interpreted, didn't it?

It absolutely did.

Remember how with RFLP the size measurement was kind of fuzzy, leading to this idea of binning where you grouped fragments into size ranges?

Because you couldn't precisely tell if two bands that were close were really the same size or just almost the same size.

Exactly.

But with the much higher resolution of PAGE used in AFLP, you could often resolve discrete alleles.

You could say this band corresponds to exactly 18 repeats or this one is 24 repeats.

And compare it directly to a standard.

Right.

They developed things called allelic ladders.

This was basically a cocktail mix of the most common known alleles for that locus run in a separate lane on the same gel.

You could directly compare your samples bands to the latter and precisely determine the allele designation.

Much more accurate and objective than RFLP binning.

Okay.

So, AFLP sounds like a huge leap forward.

Needs less DNA, works better on degraded samples, higher resolution, more precise allele calling.

Why wasn't that the final answer?

Why did we move on again?

Well, it had its own limitations.

A big one was that typically AFLP systems focused on analyzing just one locus at a time, usually D1S80.

Whereas RFLP,

using probe stripping, could eventually analyze multiple loci, maybe five or six.

Right.

So while D1S80 was highly variable, the overall discriminating power of a single locus profile wasn't nearly as high as a multi -locus RFLP profile.

The PM value wasn't as impressively low.

Okay.

Less power.

Any technical issues?

There were still some PCR related artifacts.

One common one was preferential amplification in a heterozygous sample with one large allele and one small allele.

The PCR process might copy the smaller one more easily.

Often, yes.

The smaller template can amplify more efficiently.

So on the final gel, the band for the smaller allele might look much stronger, much brighter than the band for the larger allele.

Which could cause interpretation issues, maybe even make the larger allele appear so faint it's missed.

It could, especially if the sample was borderline low quantity to begin with.

So you might misinterpret heterozygote as a homozygote for the smaller allele.

Not ideal.

So despite being a major improvement, AFLP still has some drawbacks in terms of power and potential PCR bias.

Exactly.

It was a really important stepping stone, bridging the gap from RFLP.

But these limitations meant the field kept innovating.

And that led, in the late 1990s, to the widespread adoption of the multiplex STR systems we rely on today.

Okay.

Let's quickly recap that journey for you.

We started with the basic idea of VNTRs, those variable length repeats in our DNA.

Then we walked through the complex six step RFLP process, the first method to measure them, using restriction enzymes, gels, that critical southern blot transfer, and probes.

And we saw how tricky RFLP was.

Grappling with issues like DNA degradation, leading to false exclusions, enzymes not working perfectly, causing partial digests or star activity, and the physical limits of gel resolution causing band shifting or false homozygotes.

Those challenges directly drove the shift towards PCR -based methods like AFLP.

AFLP needed less DNA, handled degraded samples better, and offered higher resolution with discrete allele calling using allelic ladders.

But even AFLP had limitations, mainly being single locus, which reduced its discriminating power, and suffering from PCR issues like preferential amplification.

The big takeaway here is really understanding that technological evolution,

seeing how scientists identified the problems with one technique, the degradation, the interpretation challenges, and specifically designed the next technique to overcome those hurdles.

The sheer technical ingenuity required back then was incredible.

Definitely.

Moving from, you know, a tiny biological stain to a reliable DNA profile that could stand up in court using those early methods required immense skill and troubleshooting.

So here's a final thought to leave you with.

We've seen how RFLP and AFLP, while mostly historical now, provided crucial lessons.

Modern STR analysis is built specifically to avoid the pitfalls of RFLP and AFLP.

It makes you wonder, what widely used scientific technique today has hidden flaws that we're currently overlooking?

And what future breakthrough will those flaws eventually force us to invent?

Something to think about.

Thank you for joining us for this deep dive into the foundations of forensic biology.

We really hope walking through these early techniques helps solidify your understanding as you continue your studies.