CRISPR, short for clustered regularly interspaced short palindromic repeats, denotes a family of DNA sequences found in many bacteria and archaea. These sequences contain repeated segments separated by variable “spacers,” some derived from invading genetic material. Together with CRISPR-associated (Cas) proteins, they form sequence-specific defense systems. In biotechnology, “CRISPR” also commonly refers to tools adapted from this machinery, especially RNA-guided systems used for genome editing. The repeat array, the natural defense system, and the engineered tools are related but distinct concepts. (pubmed.ncbi.nlm.nih.gov)
Natural structure and function
A CRISPR array is a stretch of DNA in which similar short repeats alternate with distinct spacer sequences. Nearby cas genes often encode proteins involved in acquiring spacers, processing RNA, or recognizing foreign genetic material. The repeats are relatively conserved within an array, whereas the spacers provide much of its recognition specificity. (pubmed.ncbi.nlm.nih.gov)
Natural CRISPR defense is commonly described in three stages. During adaptation, fragments of invading genetic material become new spacers. During expression and processing, transcription produces an RNA precursor that is processed into CRISPR RNAs, or crRNAs. During interference, these RNAs guide Cas machinery toward matching foreign sequences, enabling their cleavage or inactivation. Experiments have shown that adding or removing particular spacers changes resistance to corresponding bacteriophages, viruses that infect bacteria. (pubmed.ncbi.nlm.nih.gov)
This constitutes a form of adaptive immunity: previous encounters provide sequence information for subsequent defense. Its molecular machinery differs from the antibody- and lymphocyte-based systems of vertebrates. CRISPR systems are diverse; Cas9-based DNA targeting is one example rather than a universal mechanism. (pubmed.ncbi.nlm.nih.gov)
Development as a research tool
The term CRISPR was introduced in a 2002 study that also identified associated cas genes. In 2007, experiments in Streptococcus thermophilus directly demonstrated that acquiring virus-derived spacers could confer resistance to those viruses. These findings established a biological function for the repeat-and-spacer architecture. (pubmed.ncbi.nlm.nih.gov)
In 2012, Martin Jinek and colleagues, including Jennifer Doudna and Emmanuelle Charpentier, demonstrated programmable DNA cleavage by Cas9 in a purified system. They also combined its two natural RNA components into an engineered single-guide RNA. In 2013, researchers demonstrated targeted Cas9-mediated editing in human cells, helping establish CRISPR as an experimental technology. Doudna and Charpentier received the 2020 Nobel Prize in Chemistry for developing a method for genome editing. (pubmed.ncbi.nlm.nih.gov)
How CRISPR–Cas9 editing works
In a widely used configuration, Cas9 binds a guide RNA containing a targeting sequence complementary to the intended DNA site. Recognition also requires a nearby protospacer-adjacent motif, or PAM. Cas9 examines DNA for an appropriate PAM and then tests complementarity with the guide. Its two cleavage domains cut opposite DNA strands, producing a double-strand break. Thus, targeting depends on both guide–DNA pairing and the enzyme’s recognition requirements. (nature.com)
The resulting edit depends on cellular DNA repair, not simply on the cut. End-joining pathways can introduce small insertions or deletions, sometimes disrupting gene function. When a suitable repair template is supplied, homology-directed repair can introduce a specified replacement or insertion. Neither pathway guarantees a single desired outcome; editing efficiency and the distribution of products vary with the target and cellular context. (nature.com)
An important practical difference from earlier programmable nucleases is that changing the target generally requires changing the guide RNA rather than constructing a new DNA-binding protein. This makes CRISPR convenient for investigating many genomic sites, including experiments that perturb multiple genes. (genome.gov)
Related CRISPR technologies
Other Cas enzymes extend the available mechanisms. Cas12a, originally called Cpf1, is a single-RNA-guided DNA-cleaving enzyme with targeting requirements and cleavage properties different from Cas9. Cas13a targets RNA rather than DNA and, after recognizing its target, can cleave additional RNA molecules—a property exploited in molecular detection. (pmc.ncbi.nlm.nih.gov)
CRISPR-derived tools need not produce double-strand breaks:
- CRISPR interference uses catalytically inactive Cas9 to bind DNA and obstruct transcription without cutting the target sequence. (pubmed.ncbi.nlm.nih.gov)
- Base editing couples a targeting protein to chemical modification machinery. Established cytosine and adenine editors can convert particular base pairs without requiring a double-strand break. (nature.com)
- Prime editing combines a Cas9 nickase, a reverse transcriptase, and an extended guide RNA encoding the desired change. It can introduce substitutions and small insertions or deletions without requiring a donor DNA template. (nature.com)
These techniques have different targeting constraints and possible byproducts; they are not interchangeable versions of one editing reaction. (nature.com)
Applications and limitations
CRISPR enables researchers to disrupt genes, alter particular sequences, and reduce gene expression to investigate biological function. Its RNA-targeting machinery also supports detection: the SHERLOCK platform combines Cas13a’s target-triggered RNA cleavage with nucleic-acid amplification to generate a measurable signal. Such detection does not require editing a patient’s genome. (nature.com)
A clinical milestone occurred on December 8, 2023, when the United States Food and Drug Administration approved Casgevy for eligible patients aged 12 or older with sickle cell disease and recurrent vaso-occlusive crises. It was the first FDA-approved therapy using CRISPR–Cas9. The treatment edits a patient’s blood-forming stem and progenitor cells outside the body, disrupting a regulatory region associated with BCL11A to increase fetal hemoglobin production. It does not directly correct the sickle-cell mutation. (fda.gov)
Limitations include unintended changes at other genomic sites, incomplete editing, and delivery to the required cells. Mosaicism occurs when only some cells carry an edit. Human applications also distinguish somatic editing, affecting treated body cells, from germline editing, which can produce heritable changes. The latter raises additional questions concerning safety, informed consent, and effects on future generations. (genome.gov)