How Does DNA Cloning Work? The Step-by-Step Process

DNA cloning works by copying a specific piece of DNA, usually a gene, and inserting it into a small, self-replicating DNA molecule called a vector, which is then placed inside a living cell (most often a bacterium) that duplicates the gene every time the cell divides. The result is millions of identical copies of that DNA fragment, ready to be studied, modified, or used to produce a protein. The basic workflow, first demonstrated in 1973, has not changed dramatically in its logic, though the tools have become far more precise and varied.

Where the Idea Came From

In November 1973, Stanley Cohen and colleagues reported that individual genes could be cloned by cutting DNA molecules into fragments with enzymes, linking those fragments to a self-replicating plasmid, and introducing the resulting molecules into bacteria. Shortly afterward, they showed that genes from entirely unrelated bacterial species could be combined the same way, and that the foreign gene could still produce a functional protein inside its new host. Within months, Herbert Boyer’s lab and Cohen’s demonstrated that even animal genes could be cloned in bacteria.

What made this revolutionary was not any single new invention. The restriction enzymes, the plasmids, and the bacterial transformation techniques all existed before 1973. The breakthrough was in combining known tools in a new way that suddenly made it possible to isolate, copy, and study any gene from any organism.¹ That combination is still the backbone of DNA cloning today, even as newer methods have layered on top of it.

Getting the DNA You Want

Before anything else, you need the piece of DNA you plan to clone. Sometimes this is a gene you want to study, sometimes it is a stretch of regulatory DNA, and sometimes it is a synthetic sequence designed from scratch. The most common way to obtain it is through PCR, the polymerase chain reaction, which uses short synthetic DNA primers to selectively amplify a target sequence from a larger genome. After about 30 cycles of heating and cooling, you go from a few molecules of template DNA to billions of copies of just the fragment you care about.

The primers used in PCR can be designed with extra sequences tacked onto their ends, such as recognition sites for restriction enzymes that will later be used to cut and paste the fragment into a vector.² This means you can tailor the ends of your DNA insert during amplification so they are ready for the joining step. The gene does not need to come from PCR, though. It can be cut directly from genomic DNA, synthesized chemically by a commercial service, or even assembled from several overlapping pieces.

Choosing a Vector

A vector is the vehicle that carries your DNA insert into a host cell and ensures it gets copied. The workhorse vectors for most routine cloning are plasmids, small circular DNA molecules that bacteria naturally replicate alongside their own chromosome. A good cloning vector needs a few features: an origin of replication so the cell copies it, a selectable marker (usually an antibiotic resistance gene) so you can tell which cells received the vector, and a multiple cloning site where you can insert foreign DNA without disrupting anything essential.

Vectors in the pUC family, among the most widely used, include a multiple cloning site with recognition sequences for dozens of restriction enzymes, an ampicillin or kanamycin resistance gene, and a fragment of the lacZ gene that enables a visual screening trick described below.³ The choice of vector depends on the project. High-copy-number plasmids generate lots of DNA but can burden the host cell. Lower-copy vectors are more stable for certain inserts. For very large DNA fragments, researchers turn to bacterial artificial chromosomes (BACs), which can carry hundreds of thousands of base pairs and were originally developed for genome sequencing projects.⁴

Key properties like copy number, cargo capacity, selection markers, and stability systems all shape how the plasmid behaves inside the host, which matters more as projects grow in complexity.⁵ Picking the wrong vector can mean unstable clones, low yields, or toxic effects on the host bacterium.

Cutting and Joining DNA

The classic approach uses restriction enzymes, proteins that recognize specific short DNA sequences and cut the double helix at those points. Many restriction enzymes make staggered cuts, leaving short single-stranded overhangs called “sticky ends.” When the insert DNA and the vector have been cut with the same enzyme, their sticky ends are complementary and can base-pair with each other. DNA ligase, the molecular glue, then seals the backbone, producing a continuous circular molecule that contains the insert nestled inside the vector.

T4 DNA ligase, isolated from a bacteriophage, is the standard enzyme for this step. It can join both sticky-ended and blunt-ended DNA fragments.⁶ Sticky-end ligations are more efficient because the complementary overhangs hold the pieces in place while ligase works. Blunt-end ligations can join any two fragments regardless of sequence, but the efficiency drops because there is nothing aligning the pieces.

A persistent headache in this step is self-ligation: the vector closing back up on itself without taking in any insert. The standard way to prevent this is to remove the phosphate groups from the vector’s cut ends with a phosphatase enzyme, so ligase cannot seal the vector alone. An alternative approach replaces the sugar at the vector’s cut end with a modified version that blocks self-ligation while still allowing the insert to join.⁷ Either way, the goal is to stack the odds so that most of the circular molecules that form contain the insert rather than being empty vectors.

Researchers can also mix and match sticky ends that were generated by different restriction enzymes, using a controlled partial-filling technique to create new compatible overhangs. This gives more flexibility in designing cloning strategies, especially when the insert and vector do not share convenient restriction sites.⁸

Modern Assembly Methods

The restriction-enzyme-and-ligase method still works well for simple projects, but it leaves behind short “scars,” extra sequences at the junctions where the enzyme cut. For projects in synthetic biology that require stitching together many DNA fragments seamlessly, newer methods have largely taken over.

Gibson assembly, developed around 2009, is one of the most popular. It uses three enzymes working together in a single tube at a constant temperature: an exonuclease chews back the ends of overlapping DNA fragments to expose single-stranded regions, a polymerase fills in any gaps, and a ligase seals the nicks. The result is a seamless join with no extra sequences left behind.⁹ It has been used to assemble everything from individual genes to entire genomes.

Other restriction-free methods include Golden Gate assembly, which uses a special class of enzymes that cut outside their recognition sequence to enable seamless multi-fragment joins, and a growing family of overlap-based techniques such as In-Fusion, Circular Polymerase Extension Cloning, and Sequence-Independent Cloning.¹⁰ The general trend is toward seamless methods that can assemble multiple fragments in one step, a capability that traditional restriction-enzyme cloning handles poorly.¹¹

Getting the DNA Into Cells

Once you have a recombinant plasmid, it needs to enter a living host cell. For bacteria, this is called transformation. The two main methods are chemical transformation and electroporation.

In chemical transformation, bacteria are treated with calcium chloride, which makes their membranes more permeable to DNA. The cells are mixed with the plasmid DNA on ice, briefly exposed to heat (usually around 42°C for less than a minute), then returned to ice. The sudden temperature shift encourages the DNA to cross the membrane.¹² Interestingly, some studies have found that the heat shock itself matters less than commonly assumed, and that calcium chloride treatment alone can achieve similar transformation rates.

Electroporation uses a brief electrical pulse to open transient pores in the bacterial membrane, large enough and long-lasting enough for plasmid molecules to slip inside.¹³ Electroporation tends to be more efficient than chemical methods, which matters when you are working with limited amounts of DNA or need the highest possible number of transformants. After either method, the bacteria are given a recovery period in nutrient-rich broth before being spread on selection plates.

Picking the Right Colonies

After transformation, millions of bacteria are spread on agar plates containing an antibiotic. Only cells that received the plasmid, which carries the antibiotic resistance gene, survive. But surviving does not guarantee the plasmid contains your insert. Some cells will have taken up vectors that re-circularized without any insert at all.

The classic way to distinguish insert-carrying clones from empties is blue-white screening. Vectors like pUC contain a portion of the lacZ gene, and when this gene fragment is intact, the resulting protein turns a chemical indicator called X-Gal blue. Inserting foreign DNA into the cloning site disrupts the lacZ fragment, so bacteria carrying recombinant plasmids form white colonies while those with empty vectors form blue ones.¹⁴ White colonies are the ones you pick for further analysis.

Blue-white screening tells you that something was inserted, but not whether it is the right something, or whether it went in facing the correct direction. The standard verification step is colony PCR, where you pick a white colony, lyse the cells, and run a PCR with primers that flank the insertion site. If the product is the expected size, the insert is likely present.¹⁵ For final confirmation, the plasmid DNA is purified and sent for Sanger sequencing to verify that the insert sequence is correct, with no mutations introduced during PCR or ligation.¹⁶

Using Cloned Genes to Make Proteins

Cloning a gene is often just the beginning. Many projects aim to express the cloned gene, turning it into the protein it encodes, in large quantities. Expression vectors include a strong promoter upstream of the cloning site so the host cell’s machinery transcribes the gene at high levels. One of the most common systems uses the T7 promoter, where adding a chemical inducer (IPTG) triggers massive transcription of the cloned gene.

This system works well, but it has quirks. The T7 promoter is somewhat “leaky,” meaning it produces some protein even without induction. In one striking case, a human heat shock protein was expressed on such a large scale spontaneously that it represented more than half of all the bacterial cell’s proteins, without any inducer being added.¹⁷ Temperature also matters in unexpected ways: inducing the T7 promoter at higher temperatures can actually decrease protein yield rather than increase it, because the cells produce less soluble protein.¹⁸

The T7 system was originally limited to E. coli, but broad-host-range versions now allow the same induction approach to work across diverse types of bacteria, which opens up protein production in organisms that fold or modify certain proteins better than E. coli does.¹⁹

The Application That Changed Medicine

The first commercial health care product made by DNA cloning was human insulin. Before recombinant technology, insulin for people with diabetes came from pig and cow pancreases, which worked reasonably well but occasionally caused immune reactions because the animal proteins were not quite identical to the human version. In 1982, human insulin produced in bacteria using cloned genes was approved for medical use, providing for the first time a scalable, reliable source of the exact human protein.²⁰²¹

That proof of concept paved the way for dozens of other recombinant proteins used in medicine: growth hormone, clotting factors for hemophilia, erythropoietin for anemia, and monoclonal antibodies for cancer and autoimmune diseases. In each case, the basic workflow is recognizable. Clone the human gene, put it in an expression vector, grow the host cells at scale, purify the protein. The details vary enormously, but the logic is the same one Cohen and Boyer worked out in the 1970s.

DNA Cloning in Agriculture

The same principles apply to plants, though the delivery method is different. A soil bacterium called Agrobacterium tumefaciens has a natural ability to insert DNA into plant cells, and scientists have co-opted this system to deliver cloned genes into crop species.²² The gene of interest is cloned into a special binary vector that the bacterium can transfer into plant tissue. Infected plant cells are then regenerated into whole plants carrying the new gene.

Since the early 1980s, this approach has been extended from model plants to economically important crops, enabling traits like pest resistance, herbicide tolerance, and disease resistance.²³ In peanuts, for example, researchers achieved a transformation efficiency of about 55% using Agrobacterium carrying a viral coat protein gene, aiming to develop resistance to peanut clump virus.²⁴

DNA Cloning Versus Organism Cloning

One of the most persistent sources of confusion is the word “cloning” itself. When molecular biologists say “cloning,” they almost always mean copying a piece of DNA in a test tube and a bacterial cell. When the public hears “cloning,” they tend to picture Dolly the sheep. These are completely different technologies that happen to share a name.

Organism cloning, more precisely called somatic cell nuclear transfer, involves taking the nucleus from a body cell of one animal and placing it into an egg cell whose own nucleus has been removed. The egg is then stimulated to develop into an embryo that is genetically identical to the donor animal. It took more than 20 years after Dolly the sheep in 1997 to successfully clone the first non-human primate, which happened in 2018, largely because the epigenetic barriers that block reprogramming in primates are harder to overcome.²⁵

DNA cloning, by contrast, does not create organisms. It creates copies of specific DNA sequences inside bacterial or yeast cells. The confusion matters because it colors public perception of the technology. When someone hears that a gene has been “cloned,” the mental image of a duplicate animal is misleading. What actually happened is much more mundane: a piece of DNA was cut, pasted into a plasmid, and grown in bacteria.

Automation and the Future of DNA Assembly

The manual version of DNA cloning, where a researcher pipettes tiny volumes by hand through each step, is still the everyday reality in many labs. But high-throughput facilities called biofoundries are increasingly automating the entire workflow. Robotic platforms handle PCR setup, DNA assembly, transformation, colony picking, and even sequencing verification, allowing thousands of constructs to be built in parallel.²⁶

Artificial intelligence is being layered on top of these automated systems to design optimal assembly strategies, predict which constructs are likely to work, and troubleshoot failures. For large-scale synthetic biology projects, where you might need to build and test hundreds of gene variants to find the best-performing one, automation turns what would be months of bench work into days. The underlying logic of cut, paste, transform, screen, and verify has not changed, but the speed and scale at which it can happen is orders of magnitude beyond what Cohen and Boyer could have imagined when they first stitched two DNA fragments together in a plasmid over fifty years ago.²⁷

Vectors for Very Large DNA

Standard plasmid vectors top out at inserts of roughly 10 to 15 thousand base pairs. Beyond that size, they become unstable and tend to rearrange or lose portions of the insert. Bacterial artificial chromosomes solve this problem by using the replication machinery of a naturally occurring low-copy plasmid (the F factor) in E. coli, which can stably maintain inserts up to about 300,000 base pairs.²⁸

BACs have been put to particularly creative use in virology. The entire genome of mouse cytomegalovirus, roughly 230,000 base pairs, was cloned and maintained as a BAC in E. coli. When that BAC plasmid was put into animal cells, it produced a fully infectious virus.²⁹ Having the entire viral genome on a single plasmid in bacteria makes it straightforward to introduce precise mutations using standard bacterial genetics tools, then test how those mutations affect the virus in cell culture. This approach has been extended to other large DNA viruses and has become a standard method for studying viral gene function.