Reading without copying: detection by hybridization
The last two guides gave you two superpowers. PCR copies a chosen stretch of DNA a billionfold; sequencing reads the letters one by one. Both are wonderful when you do not know what is there and want the whole answer. But a great deal of real biology asks a narrower question — *is this particular sequence present, and if so, where and how much?* You are not trying to read an unknown book; you are checking whether one specific sentence appears in it. For that you do not need to copy or to read at all. You only need to let a known piece of DNA go looking for its partner.
The whole family of methods in this guide rests on one idea you met long ago on the very first rung: complementary base pairing, A always reaching across to T and G to C. A short single strand of known sequence — a probe — will find and stick to its complement and to nothing else, picking its one partner out of a crowd of millions of unrelated sequences. That sticking-together of two complementary strands is called hybridization, and it is the engine under everything below. Heat a double helix and the two strands let go of each other (denaturation); cool them and a strand re-finds its true complement and zips back up. A probe simply exploits that homing instinct: you supply one half of a helix and let the sample provide the other.
The microarray: thousands of questions on one chip
A single probe answers one question. A DNA microarray answers thousands at once by the simple move of running thousands of probes in parallel. Picture a glass slide the size of a thumbnail, divided into a grid of tiny spots, tens of thousands of them. Each spot has a different known single-stranded probe anchored to it — spot one carries copies of a probe for gene A, spot two for gene B, and so on across the whole grid. The position of each spot is the label: the chip's maker recorded which sequence sits where, so a glow at row 12, column 40 means *that* gene's probe found a partner.
The most famous use is measuring gene expression — which genes are switched on, and how strongly. You collect all the mRNA from your cells, copy it into more stable DNA with reverse transcriptase (information flowing RNA -> DNA, the central dogma running backward, as you learned it can), and tag every copy with a fluorescent dye. Then you pour this glowing soup over the chip. Each tagged strand drifts until it finds the spot whose probe is its complement, hybridizes there, and stays; the rest washes away. Now shine a laser and photograph the grid. A bright spot means that gene's mRNA was abundant; a dim spot means it was scarce; a dark spot means it was off. In one image you have read the activity of every gene on the chip.
Be honest about what a microarray can and cannot do. It is fast and cheap and reads a whole genome's worth of spots in one shot — but it is fundamentally a *closed* question. It can only detect sequences someone already chose to print as probes; a gene nobody anticipated, or a brand-new pathogen variant, simply has no spot to light up and is invisible. Its signal also saturates, so it measures abundance crudely rather than counting molecules. For these reasons RNA-seq — sequencing the transcripts directly — has largely replaced microarrays for expression studies, because sequencing is open-ended and counts. The array remains a clean, economical workhorse when you already know exactly which sequences you want to ask about.
FISH: lighting up a sequence where it lives
A microarray tells you *whether* and *how much*, but it grinds the cell to soup first, so it throws away all information about *where*. Sometimes location is the whole point: is a gene on chromosome 7 or chromosome 9? Are there two copies of it or three? Fluorescence in-situ hybridization, or FISH, answers exactly that. "In situ" is Latin for "in place" — the trick is to run the hybridization *inside* an intact cell, on its chromosomes, without taking anything apart, so the glow appears at the physical spot where the sequence sits.
- Fix the cells onto a glass slide so the chromosomes hold still, then gently denature the DNA in place — heat or chemicals pry the double helix open into single strands while the chromosome keeps its shape.
- Add a fluorescent probe — a single-stranded piece of DNA complementary to the sequence you are hunting, with a glowing dye attached.
- Let it hybridize. The probe roams the exposed chromosomal DNA and base-pairs only where it finds its exact complement, anchoring itself to that one spot.
- Wash away the unbound probe, then look down a fluorescence microscope: each place the probe stuck shows up as a bright dot of colour against the dark chromosome.
Because each dot marks one copy of the target, you can simply *count* them, and that is clinically powerful. Two dots is the normal pair of copies; three dots betrays an extra chromosome, the signature of trisomy 21 (Down syndrome). Probes painted in different colours can show whether two genes that should sit on separate chromosomes have been fused together — the hallmark of certain leukaemias. FISH bridges this rung back to what you learned about chromosomes and ploidy: a karyotype shows the chromosomes as grey shapes, but FISH lets you ask a pointed question — *is this one gene present, and on which chromosome* — and see the answer glow.
The molecular beacon: a probe that glows only when it binds
Microarrays and FISH share one chore: you hybridize, then you *wash away* every probe that did not bind, because a glowing probe floating free looks just like a glowing probe that found its target. That washing step is slow and impossible to do while a reaction is running. The molecular beacon is a beautiful piece of design that abolishes it — a probe that is dark until the moment it binds, so the bound and unbound versions look different and no wash is needed.
The mechanism is a small marvel of using base pairing against itself. The beacon is a single strand folded into a hairpin: a probe sequence in the middle, flanked by a few extra bases at each end that are complementary to *each other*. Those self-complementary ends pair up and pull the strand into a closed loop, which holds a fluorescent dye on one end right next to a quencher molecule on the other. While they touch, the quencher absorbs the dye's light and the beacon stays dark. But when the loop's probe region meets its true target, base pairing to that target is longer and stronger than the little hairpin stem; the target wins the tug-of-war, the hairpin springs open, and the dye is yanked away from the quencher. Freed, it glows. Fluorescence appears *only* when the probe has found its sequence.
Where this lives: diagnostics and forensics
These hybridization tools are not laboratory curiosities; they are the quiet machinery of modern molecular diagnostics. When a clinic tests a swab for a virus, the core of the test is almost always a probe that lights up if and only if the pathogen's signature sequence is present — often a beacon reporting inside a qPCR run, which is why such a test can give an answer in an hour. Pathogen detection by sequence is fast and ferociously specific: a well-chosen probe can tell one strain of bacterium from a close cousin by a handful of bases, where growing the microbe in a dish might take days or fail entirely.
Forensics leans on the same base-pairing logic in a different way. DNA fingerprinting does not read whole genes; it measures a few dozen spots in the genome where a short motif repeats a variable number of times — at one spot you might have inherited 8 repeats, someone else 13. The exact set of repeat counts across many such spots is, for practical purposes, unique to you (identical twins aside). PCR amplifies just those spots, and their lengths are read out and compared; a match between crime-scene DNA and a suspect, or between a child and a claimed parent, is a match of these length profiles. The same probe-and-hybridize toolkit that diagnoses disease also identifies people.
Two honest cautions ride along with all this power. First, the very sensitivity that makes these methods marvellous makes them treacherous: PCR amplifies *any* matching DNA, so a few stray molecules of contamination — a technician's skin cell, carryover from a previous tube — can be copied into a false positive. Scrupulous controls and physically separated workspaces are not bureaucratic fuss; they are what keeps a forensic or diagnostic result trustworthy. Second, a probe only ever finds what it was designed to find: it is blind to a new variant whose target sequence has mutated, which is exactly why pathogen tests must be redesigned as a virus evolves. The strength of detection-by-hybridization — its perfect specificity — is the same edge that cuts the other way.