35.08.04 · health-medicine / future-medicine

CRISPR therapeutics: Casgevy, the first FDA-approved gene-editing drug, and the clinical translation of genome editing

shipped3 tiersLean: none

Anchor (Master): Jinek, Chylinski, Fonfara, Hauer, Doudna & Charpentier 2012 Science 337:816 (reconstituted Cas9); Cong, Ran, Cox et al. 2013 Science 339:819 (first mammalian editing); Frangoul et al. 2021 N. Engl. J. Med. 384:252 (CLIMB-121 sickle cell); Locatelli et al. 2023 N. Engl. J. Med. (CLIMB-111 beta-thalassemia); Gillmore et al. 2021 N. Engl. J. Med. 385:493 (NTLA-2001 first in-vivo CRISPR in humans); Musunuru et al. 2021 Nature (base editing of PCSK9); Anzalone et al. 2019 Nature 576:149 (prime editing); FDA BLA 125787 Casgevy label, 8 Dec 2023; 2020 Nobel Prize Charpentier–Doudna

Intuition Beginner

In 2012 scientists learned to reprogram a bacterial defence system into a precise genome-editing tool. Just eleven years later, in December 2023, that tool became an approved drug. That drug, Casgevy, treats sickle cell disease, an inherited blood disorder in which red blood cells deform into rigid crescents that block vessels, cause crushing pain, and damage organs year after year.

Casgevy does not edit the gene that makes sickle haemoglobin. It edits a different gene, called BCL11A, that switches off the production of fetal haemoglobin shortly after birth. Fetal haemoglobin does not sickle. By cutting the switch, the edit turns fetal haemoglobin production back on, and the new blood cells carry a healthy form of the molecule.

The procedure is a transplant of the patient's own cells. Stem cells are drawn from the blood, edited in a laboratory dish with CRISPR, and returned by infusion after a round of conditioning chemotherapy clears space in the bone marrow. In the trial, 94% of sickle cell patients had no pain crises for at least a year afterward. Why this exists: CRISPR turned a lifelong disease into a one-time, potentially curative treatment.

Visual Beginner

The figure has three panels reading left to right as a treatment timeline. The left panel shows the patient connected to an apheresis machine drawing blood stem cells into a collection bag. The middle panel shows those stem cells in a laboratory dish being transfected with the CRISPR ribonucleoprotein (a cartoon scalpel labelled "Cas9 + sgRNA") cutting a regulatory switch labelled "BCL11A enhancer". The right panel shows edited cells dripping back into the patient through an IV, with the bone marrow highlighted, and red blood cells above shifting from crescent-shaped sickle cells to round healthy discs carrying fetal haemoglobin.

The three panels capture the ex vivo workflow that distinguishes Casgevy from gene therapies delivered directly into the body: the editing happens outside the patient, in a controlled dish, and only edited cells go back in.

Worked example Beginner

Walk through the CLIMB-121 trial of Casgevy for sickle cell disease, ages 12 to 35, with severe disease defined as at least two vaso-occlusive crises per year.

Step 1. Collect stem cells. The patient receives mobilisation drugs, then connects to an apheresis machine that filters haematopoietic stem cells out of the bloodstream into a bag, roughly a few hundred million cells.

Step 2. Edit ex vivo. In the laboratory, the cells are electroporated with the Cas9 protein plus a guide RNA targeting the erythroid-specific enhancer of the BCL11A gene. The Cas9 cuts the enhancer; the cell's own repair scrambles it. With the enhancer broken, BCL11A stops silencing fetal haemoglobin when those cells mature into red blood cells.

Step 3. Condition. The patient is admitted to hospital and receives busulfan, a chemotherapy drug that destroys the remaining bone marrow to make room for the edited cells. This is the hardest part of the procedure and the source of most side effects.

Step 4. Re-infuse. The edited cells are returned through a vein. They home to the bone marrow, engraft, and begin producing red blood cells.

Step 5. Read the outcome. Among the 31 patients evaluated in the primary analysis, 29 had no vaso-occlusive crises for at least twelve consecutive months, and all responders sustained fetal haemoglobin above 40% of total haemoglobin. The FDA approved the drug on 8 December 2023.

What this tells us: editing one regulatory switch, outside the body, is enough to re-route red blood cell production from the disease-causing adult form to the asymptomatic fetal form, and the effect persists for years after a single treatment.

Check your understanding Beginner

Exercise (easy, multiple choice).

Casgevy treats sickle cell disease. Which gene does Casgevy's CRISPR edit actually target inside the patient's cells?

A. The beta-globin gene that encodes the sickle mutation itself B. The BCL11A enhancer that switches off fetal haemoglobin production C. The haemoglobin alpha-chain gene D. The gene encoding the sickle-cell parasite receptor

Hint

Casgevy does not directly fix the disease-causing mutation. It edits a regulatory switch that controls which form of haemoglobin the cell produces.

Answer

B. Casgevy edits the erythroid-specific enhancer of BCL11A, the transcription factor that silences fetal haemoglobin after birth. Disrupting this enhancer in the patient's own stem cells de-represses fetal haemoglobin, which does not polymerise and does not sickle. Feedback-correct: this is the indirect-control strategy that makes a single edit curative. Feedback-wrong: option A is the intuitive but wrong target — direct repair of the sickle point mutation would require precise homology-directed repair in every cell, which is inefficient; option B achieves the clinical goal via a more reliable regulatory detour.

Formal definition Intermediate+

CRISPR therapeutics are a class of genome-editing interventions that use a CRISPR-Cas ribonucleoprotein to alter a defined genomic locus in a patient's cells for therapeutic benefit. The class subdivides by the site of editing and by the chemistry of the edit. Ex vivo editing removes cells from the patient, edits them in culture, and re-infuses them; Casgevy (exagamglogene autotemcel, exa-cel) is the canonical example. In vivo editing delivers the ribonucleoprotein (typically encapsulated in a lipid nanoparticle or adeno-associated virus) directly into the patient; Intellia's NTLA-2001 for transthyretin amyloidosis is the first-in-human example ^{[Gillmore2021]}.

Definition (exa-cel and the BCL11A strategy). Casgevy is an autologous haematopoietic stem cell therapy in which the patient's own CD34-positive cells are edited ex vivo by electroporation of the Cas9 ribonucleoprotein complexed with a single guide RNA targeting the erythroid-specific enhancer of the BCL11A gene. Non-homologous end joining disrupts the enhancer; on maturation of the edited clones into erythroid lineage cells, BCL11A is no longer expressed at the level required to silence the gamma-globin genes; fetal haemoglobin (HbF, $α_{2} γ_{2}$ ) accumulates in the red cell and replaces the function of the defective adult haemoglobin (HbS, $α_{2} β_{2}^{S}$ in sickle cell disease, or impaired beta-chain output in beta-thalassemia). HbF does not polymerise under deoxygenated conditions and thereby blocks the sickling cascade at its root.

Definition (modality taxonomy). Four modalities span the current clinical pipeline. (i) Double-strand break + NHEJ uses Cas9 to introduce a break whose error-prone repair disables a target gene (Casgevy, NTLA-2001). (ii) Base editing fuses a deaminase to a Cas9 nickase to install a single C-to-T, G-to-A, A-to-G, or T-to-C transition without a break, used by Verve Therapeutics to knock out hepatic PCSK9 for hypercholesterolemia and by Beam Therapeutics for sickle cell ^{[Musunuru2018]}. (iii) Prime editing appends a reverse transcriptase to a Cas9 nickase and a prime editing guide RNA that both targets the locus and templates the edit, installing all twelve base substitutions and small indels without a double-strand break ^{[Anzalone2019]}. (iv) Epigenome editing deploys catalytically dead Cas9 fused to chromatin-modifying domains to silence or activate a locus without altering the DNA sequence, enabling durable and reversible regulation.

Counterexamples to common slips Intermediate+

Slip: "Casgevy corrects the sickle mutation." It does not. The point mutation in the beta-globin gene (HBB, the E6V substitution) is untouched. The edit lands in an erythroid enhancer of BCL11A on a different chromosome and works by rerouting the haemoglobin programme to fetal gamma-globin.

A patient treated with Casgevy still carries the sickle mutation in every non-erythroid cell; the cure is functional, not genetic.
Slip: "The list price equals the cost of the disease." Casgevy's $2.2 mi l l i o n l i s t i s t h e a c q u i s i t i o n cos tp er p a t i e n t f or t h e d r ug p r o d u c t a l o n e . T h e f u l l cos t — a p h er es i s, co n d i t i o nin g c h e m o t h er a p y, w ee k so f h os p i t a l i s a t i o n f or t h e m y e l o ab l a t i v e b u s u l f an r e g im e n, an d mana g e m e n t o f co m pl i c a t i o n s — c ana pp r o a c h$ 3 million, and this does not include the infrastructure investment required at authorised treatment centres. The lifetime cost of untreated sickle cell disease, by comparison, is dominated by repeated hospitalisations, transfusions, and lost productivity; the cost-effectiveness comparison depends entirely on the time horizon chosen.
Slip: "In-vivo editing is just ex vivo editing without the transplant." The two workflows differ in kind, not only in location. Ex vivo editing allows quality control: the manufacturer sequences the edited product, measures off-target rates, and releases only lots meeting specification. In-vivo editing has no such release gate; delivery, editing efficiency, and off-target rates are determined inside the patient and cannot be checked before the edit happens. This is why in-vivo editing entered the clinic for a non-curability-barred disease (transthyretin amyloidosis, where even partial knockdown helps) rather than for sickle cell.

Key result: the Casgevy CLIMB-121 trial and the first approved CRISPR therapy Intermediate+

Theorem (CLIMB-121 primary efficacy; Frangoul et al. 2021). In a phase 1/2/3 open-label trial of exa-cel in patients aged 12 to 35 years with severe sickle cell disease (at least two vaso-occlusive crises per year during each of the two prior years), a single course of ex vivo BCL11A-enhancer editing followed by myeloablative busulfan conditioning and autologous re-infusion produced sustained fetal haemoglobin levels exceeding 40% of total haemoglobin and eliminated vaso-occlusive crises for at least 12 consecutive months in 29 of the 31 evaluable patients (93.5%, primary endpoint met at the prespecified threshold).

Proof. The trial design isolates the editing effect from natural history. The argument proceeds through four quantitative checkpoints.

(1) Editing succeeded in the product. Of the harvested CD34-positive cells, on the order of 80% carried the intended indels at the BCL11A erythroid enhancer after electroporation, measured by targeted deep sequencing of the product before re-infusion. Off-target editing at pre-specified candidate sites was below the assay detection floor (under 0.1% by targeted sequencing).

(2) Edited cells engrafted and repopulated. After busulfan conditioning created marrow space, the re-infused cells homed to the bone marrow and reconstituted haematopoiesis. Tracking the indel signature in peripheral blood showed that the edited fraction was stably maintained across the follow-up window, indicating that long-term haematopoietic stem cells (not only short-lived progenitors) carried the edit.

(3) Fetal haemoglobin was re-expressed at therapeutic level. In responders, total HbF rose above 40% of haemoglobin and was distributed pancellularly (in the majority of red cells, not only a minority subpopulation), which is the biophysical requirement for blocking HbS polymerisation. Pancellular distribution matters because HbF protects only the cells that contain it; a focal HbF distribution would leave most red cells vulnerable to sickling.

(4) The clinical endpoint followed. With HbF at therapeutic level and pancellular, the polymerisation cascade that drives vaso-occlusion was blocked at its root. Twenty-nine of 31 patients achieved the primary endpoint of at least 12 consecutive months free of severe vaso-occlusive crises; the two non-responders had lower engraftment of edited cells and lower HbF fractions, consistent with the mechanistic model. The FDA approved Casgevy on 8 December 2023 on the strength of this result ^[FDA2023]. $□$

Bridge. The CLIMB-121 result builds toward 35.08.01 future medicine, where genome-editing cures for monogenic disease appear as one prong of a broader precision-medicine programme, and appears again in 17.11.03 CRISPR-Cas9 genome editing, where the molecular mechanism (PAM-licensed cleavage, NHEJ repair) is the load-bearing machinery that the present unit translates into clinical effect. The foundational reason the indirect BCL11A strategy outperforms direct repair of the sickle mutation is that disrupting a regulatory enhancer tolerates any of the dozens of indels NHEJ produces, whereas precise correction requires a single HDR outcome in every edited cell. This is exactly the gap between NHEJ-dominant and HDR-dominant editing that the mechanism unit catalogues, and the bridge is that the entire pipeline of base editing, prime editing, and epigenome editing extends the same insight: choose a modality whose stochastic repair distribution is dominated by the therapeutically useful outcome.

Exercises Intermediate+

Exercise 1 (easy, multiple choice).

Casgevy edits an erythroid-specific enhancer of BCL11A. What is the downstream molecular consequence of disrupting this enhancer?

A. The sickle mutation in beta-globin is repaired B. Fetal haemoglobin production is reactivated in the erythroid lineage C. Adult haemoglobin is overproduced to compensate for sickling D. Red blood cell membranes are thickened to resist sickling

Hint

BCL11A is a transcription factor that silences the gamma-globin genes after birth. Removing its erythroid enhancer removes that silencing.

Answer

B. BCL11A represses the gamma-globin genes in adult erythroid cells. Disrupting the erythroid-specific enhancer lowers BCL11A expression in the erythroid lineage specifically (not in other lineages, where BCL11A has separate roles), de-repressing fetal haemoglobin (HbF, $α_{2} γ_{2}$ ). HbF does not participate in the polymerisation that drives sickling, so its production functionally rescues the red cell. Options A, C, and D describe alternative strategies that Casgevy does not pursue.

Exercise 3 (medium, multiple choice).

A pharmaceutical company is choosing between an ex vivo and an in vivo editing strategy for a liver disease target. Which consideration most strongly favours the ex vivo approach?

A. Lipid nanoparticle delivery to the liver is technically harder than for haematopoietic cells B. Ex vivo editing allows sequencing the product and releasing only lots that meet editing-efficiency and off-target specifications C. Ex vivo editing is always cheaper than in vivo editing D. In vivo editing cannot achieve editing efficiencies above 10% in any tissue

Hint

Think about what quality controls are available when the editing happens in a dish versus inside the patient.

Answer

B. The defining advantage of ex vivo editing is the release gate: the manufacturer can sequence the edited cell product, measure editing efficiency and off-target rates, and discard substandard lots before the patient is exposed. In vivo editing offers no such gate; whatever happens inside the patient is what the patient gets. Option A is incorrect (LNP delivery to liver is one of the most efficient in-vivo targets); option C is generally false (ex vivo requires apheresis, conditioning chemotherapy, and weeks of hospitalisation); option D is false (in-vivo liver editing above 50% has been achieved in humans with NTLA-2001).

Exercise 4 (medium, numeric).

Gillmore et al. 2021 reported that a single intravenous infusion of NTLA-2001 at 0.3 mg/kg produced a 93% mean reduction in serum transthyretin (TTR) protein relative to baseline at the nadir timepoint. If a patient had a baseline serum TTR of $60 mg/dL$ , compute the serum TTR level at the post-infusion nadir. Report the answer in mg/dL.

Hint

A 93% reduction means 7% of the baseline value remains. Multiply 60 by 0.07.

Answer

4.2 mg/dL. A 93% knockdown leaves 7% of the baseline: $60 \times (1 - 0.93) = 60 \times 0.07 = 4.2$ . This near-complete knockdown of a hepatic secreted protein after a single intravenous dose established NTLA-2001 as the proof of concept that in-vivo CRISPR editing can produce pharmacologically meaningful protein-level effects in humans, not just cellular editing in a biopsy.

Exercise 5 (medium, symbolic).

Let $f$ be the fraction of long-term haematopoietic stem cells (HSCs) in the re-infused product that carry a BCL11A-enhancer-disrupting indel, and let $g$ be the fraction of re-infused cells that successfully engraft in the bone marrow after busulfan conditioning. Assume the two are independent. Write the fraction of the patient's reconstituted long-term haematopoietic output that derives from edited cells, in terms of $f$ and $g$ .

Hint

Independence means the joint probability is the product. Every engrafted cell contributes equally to long-term output on average.

Answer

The edited fraction of long-term haematopoietic output is $f \cdot g$ . Independence gives the joint probability that a cell is both edited and engrafted as the product. In CLIMB-121, $f \approx 0.8$ in the product and engraftment efficiency $g$ was high enough that the peripheral-blood edited fraction remained stable across the follow-up window, indicating that $f \cdot g$ in the marrow exceeded the threshold required to push HbF above 40% pancellular. The two non-responders in the primary analysis had lower $f \cdot g$ , consistent with this model.

Exercise 6 (medium, multiple choice).

Base editing (Komor 2016; Verve VERVE-101) and standard Cas9 double-strand-break editing differ in a key mechanistic feature. Which statement correctly contrasts the two?

A. Base editing cuts both DNA strands; standard Cas9 does not cut at all B. Base editing chemically converts a single base without cutting the phosphodiester backbone; standard Cas9 introduces a double-strand break C. Base editing requires a donor template; standard Cas9 does not D. Base editing installs any of the twelve possible base substitutions; standard Cas9 cannot

Hint

Base editing fuses a deaminase (cytidine or adenosine) to a Cas9 nickase. The deaminase chemically modifies one base; the nickase nicks one strand rather than cutting both.

Answer

B. Base editing uses a Cas9 nickase (typically D10A, which cuts only the non-target strand) fused to a deaminase that chemically converts a cytosine to uracil (C-to-T) or an adenine to inosine (A-to-G) within a narrow window of the protospacer, without a double-strand break and without a donor template. Option A reverses the mechanism; option C is wrong (base editing needs no donor); option D describes prime editing, not base editing. Base editing can install only transitions (C-to-T, G-to-A, A-to-G, T-to-C), not transversions.

Exercise 7 (hard, symbolic).

Model the HbF fraction in peripheral blood after Casgevy treatment. Let $f$ be the fraction of engrafted long-term HSCs carrying a BCL11A-disrupting indel, let $h$ be the steady-state HbF fraction produced by a single edited erythroid cell (assume $h \approx 0.9$ , that is, 90% of the haemoglobin in an edited red cell is fetal), and let $u$ be the steady-state HbF fraction in an unedited erythroid cell (assume $u \approx 0.01$ , residual fetal haemoglobin in adult cells).

Derive the patient's total HbF fraction as a function of $f$ , and solve for the minimum editing fraction $f^{*}$ required to exceed the 40% HbF clinical-response threshold.

Hint

Total HbF is the weighted sum of edited and unedited contributions: $f \cdot h + (1 - f) \cdot u$ . Set this equal to 0.40 and solve for $f$ .

Answer

The total HbF fraction is

HbF (f) = f \cdot h + (1 - f) \cdot u = u + f (h - u) .

Setting $HbF (f^{*}) = 0.40$ and solving,

f^{*} = \frac{0.40 - u}{h - u} = \frac{0.40 - 0.01}{0.90 - 0.01} = \frac{0.39}{0.89} \approx 0.438.

Roughly 44% of engrafted long-term HSCs must carry the edit for the patient to clear the 40% pancellular-HbF threshold. The CLIMB-121 product specification targeted $f$ well above this floor, and the two non-responders fell short, consistent with the model. This calculation is the mechanistic basis for the release specification: a manufacturing lot must achieve an editing fraction comfortably above $f^{*}$ before release.

Exercise 8 (hard, numeric).

A payer is evaluating the cost-effectiveness of Casgevy for sickle cell disease at a list price of $$2.2 M$ per patient. Assume the therapy is curative and restores normal life expectancy (a 50-year horizon, undiscounted), and compare against standard of care at $$50, 000$ per year in chronic disease management costs. Compute the break-even year (the year at which cumulative standard-of-care cost equals the Casgevy list price), and report the cumulative undiscounted cost difference at year 50 in favour of Casgevy.

Hint

Divide the Casgevy price by the annual standard-of-care cost to get the break-even year. At year 50, compare $50 \times $50, 000$ (standard of care) against $$2.2 M$ (Casgevy).

Answer

Break-even year: $$2, 200, 000/$50, 000 = 44$ years. At year 50, standard of care has cost $50 \times $50, 000 = $2, 500, 000$ , while Casgevy has cost $$2, 200, 000$ . The undiscounted difference in favour of Casgevy is $$300, 000$ .

This simplified calculation illustrates why cost-effectiveness analyses for one-time curative therapies are sensitive to the discount rate, the actual cost of the conditioning regimen (which can add substantially to the $$2.2 M$ list), and the realistic post-treatment life expectancy. Under typical 3% annual discounting, the break-even year is pushed later, and under realistic full-treatment-cost assumptions (including hospitalisation for busulfan conditioning), Casgevy may not break even on direct medical cost alone within the patient's lifetime, though indirect costs (productivity, avoided transfusions, quality of life) shift the balance.

Advanced results Master

Theorem 1 (CLIMB-121 sickle cell primary efficacy; Frangoul 2021). Exa-cel produced sustained HbF above 40% pancellularly and eliminated vaso-occlusive crises for at least 12 consecutive months in 29 of 31 evaluable severe-sickle-cell patients aged 12 to 35, with off-target editing below the assay detection floor and a stable edited fraction in peripheral blood throughout follow-up ^{[Frangoul2021]}. The result converted a lifelong haemolytic disease with multi-organ damage into a one-time intervention with durable response.

Theorem 2 (CLIMB-111 beta-thalassemia; Locatelli 2023). In patients with beta-thalassemia (non- $β^{0} / β^{0}$ genotypes), the same exa-cel product eliminated the need for chronic transfusion in the majority of evaluable patients for the duration of follow-up, with total haemoglobin sustained in the normal range without exogenous red cells. The FDA approved exa-cel for transfusion-dependent beta-thalassemia in January 2024 on the basis of this cohort.

Theorem 3 (NTLA-2001 first in-vivo CRISPR in humans; Gillmore 2021). A single intravenous infusion of NTLA-2001, a lipid-nanoparticle-encapsulated Cas9 mRNA plus guide RNA targeting the transthyretin (TTR) gene in hepatocytes, produced a mean 93% knockdown of serum TTR at the 0.3 mg/kg dose in patients with hereditary transthyretin amyloidosis with polyneuropathy ^{[Gillmore2021]}. The result established that in-vivo CRISPR editing can achieve pharmacologically meaningful protein-level effects in humans and opened the in-vivo pipeline for hepatic targets including alpha-1 antitrypsin deficiency, glycogen storage diseases, and familial hypercholesterolemia.

Theorem 4 (base editing of PCSK9; Musunuru 2018; Verve VERVE-101). Lipid-nanoparticle delivery of an adenine base editor plus a guide RNA targeting the splice acceptor of PCSK9 in hepatocytes produced durable, potentially permanent knockdown of serum PCSK9 and LDL cholesterol in non-human primates (Musunuru 2018) and entered human trials (VERVE-101, Verve Therapeutics) as a one-time treatment for heterozygous familial hypercholesterolemia. The modality converts a daily statin pill into a single edit by permanently disabling the gene encoding the LDL-receptor-degrading enzyme.

Theorem 5 (prime editing pipeline; Anzalone 2019; Prime Medicine). Prime editing extends the editable sequence space beyond the indels of NHEJ and the transitions of base editing to all twelve base substitutions and small insertions and deletions, without producing a double-strand break ^{[Anzalone2019]}. The modality entered clinical development for chronic granulomatous disease (PM359, correcting a specific NCF1 mutation) and other monogenic disorders where the pathogenic lesion is a specific substitution that base editing cannot install and that HDR cannot reach in the relevant cell type.

Theorem 6 (epigenome editing; Chroma, Tune, Epic Bio). Catalytically inactivated Cas9 (dCas9) fused to chromatin-modifier recruitment domains (DNMT3A for DNA methylation, KRAB for heterochromatin spreading, VP64-p65-RTA for activation) silences or activates a target locus without altering the DNA sequence. Epigenome editors under development target PCSK9 (for durable cholesterol control without a permanent DNA change), hepatitis B integrated DNA, and MYC. The modality trades the irreversibility of a DSB or base edit for the reversibility and tunability of a chromatin state.

Theorem 7 (delivery frontier sets the in-vivo addressable space). The addressable in-vivo target space is set by delivery, not by editing chemistry. Lipid nanoparticles preferentially traffic to hepatocytes (enabling NTLA-2001, VERVE-101), adeno-associated virus serotypes have tissue tropisms (AAV9 for cardiac and central nervous system, AAV8 for liver), and engineered virus-like particles and GalNAc-conjugated LNPs are extending the editable tissue list. Editing of the central nervous system, muscle, and immune cells remains delivery-limited, not mechanism-limited, and is the dominant preclinical frontier.

Synthesis. The Casgevy result builds toward 35.08.01 future medicine, where genome-editing cures are one prong of a broader precision-medicine programme that also includes cell therapies and mRNA therapeutics, and appears again in 17.11.03 CRISPR-Cas9 genome editing, where the molecular mechanism (PAM licensing, blunt DSB, NHEJ repair) is the chassis the present unit translates into clinical effect. The foundational reason the indirect BCL11A strategy outperforms direct correction of the sickle mutation is that disrupting a regulatory enhancer tolerates any of dozens of NHEJ indels, whereas precise HDR requires a single outcome in every cell; this is exactly the gap between NHEJ-dominant and HDR-dominant editing that the mechanism unit catalogues, and the central insight is that modality choice should be matched to the repair-outcome distribution the target tolerates.

The bridge is that the entire pipeline — base editing (Verve), prime editing (Prime Medicine), epigenome editing (Chroma, Tune) — generalises the same insight by re-engineering what happens after target recognition while preserving the search-and-bind half of the Cas9 architecture; the pattern recurs across the modality zoo, where each variant is a different chemistry module on the same search chassis. Putting these together with the delivery frontier identifies the addressable-disease space with the addressable-tissue space: the bridge is between the editing chemistry, which the mechanism unit formalises, and the clinical-disease space, which the present unit maps, and the clinical translation pipeline grows exactly as delivery engineering expands the editable-organ list.

Full proof set Master

Proposition (BCL11A-enhancer editing is a single-point-of-control for HbF silencing). Let the erythroid-specific enhancer of BCL11A be the genomic element required for BCL11A expression in the erythroid lineage, and let BCL11A be the transcription factor required for silencing the gamma-globin genes in adult erythroid cells. Disrupting the erythroid enhancer de-represses fetal haemoglobin in the erythroid lineage specifically, without affecting BCL11A function in non-erythroid lineages.

Proof. BCL11A is expressed in multiple haematopoietic lineages, where it serves distinct transcriptional roles. The full BCL11A knockout would disrupt these roles and produce broader toxicity; the clinical strategy targets a discrete erythroid-specific enhancer element (within the intergenic region downstream of the BCL11A coding gene) that drives BCL11A expression specifically in erythroid cells. Disrupting this enhancer by NHEJ-mediated indels abolishes erythroid-lineage BCL11A expression while sparing BCL11A in B cells, dendritic cells, and neurons where the gene uses different regulatory elements.

With BCL11A absent from the erythroid lineage, the gamma-globin genes (HBG1 and HBG2) are no longer repressed. The gamma-globin chains pair with alpha-globin to form fetal haemoglobin (HbF, $α_{2} γ_{2}$ ), which is the dominant haemoglobin during fetal life and is normally silenced shortly after birth. The mechanism is exactly the recapitulation, in a controlled edit, of the natural-genetic phenomenon of hereditary persistence of fetal haemoglobin (HPFH), in which mutations in the same erythroid enhancer produce lifelong HbF elevation and ameliorate sickle cell disease in carriers. The clinical edit copies a known benign human polymorphism. The de-repression is lineage-specific, sustained across erythroid differentiation, and the edited regulatory state is heritable through the long-term haematopoietic stem cell compartment, which is the requirement for a single-intervention cure. $□$

Proposition (busulfan conditioning sets the floor on Casgevy toxicity). The dominant acute toxicity of Casgevy is attributable to the busulfan myeloablative conditioning, not to the CRISPR edit itself; eliminating the conditioning would require a cell-autonomous competitive advantage for edited cells that the current BCL11A strategy does not confer.

Proof. Busulfan is an alkylating agent that ablates the recipient bone marrow to create niche space for the re-infused edited cells. Without conditioning, the edited cells engraft poorly because they must compete with the resident (un-edited, disease-causing) haematopoiesis for niche occupancy, and the edited fraction in peripheral blood remains too low to clear the HbF threshold derived in Exercise 7. The conditioning is therefore load-bearing for efficacy, not a removable accessory.

The cost is that busulfan produces the dominant adverse-event profile of the procedure: prolonged cytopenias, infertility risk, mucositis, infection susceptibility, and a small but nonzero risk of secondary myelodysplasia or leukaemia. The CRISPR edit itself contributes a different and much smaller risk (off-target cleavage, below the assay floor in CLIMB-121, plus the theoretical risk of structural variants at the on-target site). A conditioning-free Casgevy would require either (i) a BCL11A edit that confers a cell-autonomous proliferative advantage (so edited cells outcompete resident cells without conditioning), which the current edit does not, or (ii) an in-vivo delivery strategy that edits resident HSCs in place, which is delivery-limited and not yet clinical. Until either condition is met, busulfan conditioning is the rate-limiting toxicity of the Casgevy workflow and the dominant target for next-generation improvement. $□$

Connections Master

Future medicine survey: genomics, AI, and global health 35.08.01. The present unit is the depth companion to the future-medicine survey: 35.08.01 places genome editing alongside cell therapy and mRNA therapeutics as the three pillars of the coming generation of medicine, and the present unit supplies the worked example of how a single editing modality moves from discovery to approved drug in eleven years. The survey's framing of precision medicine as programmable interventions on the molecular basis of disease is exactly what Casgevy instantiates.
CRISPR-Cas9 genome editing: PAM recognition, sgRNA guidance, and DSB repair outcomes 17.11.03. The mechanistic foundation of every result in the present unit is the PAM-licensed cleavage and repair-outcome distribution formalised in the molecular-biology methods unit. Casgevy is NHEJ-mediated disruption of the BCL11A enhancer; NTLA-2001 is NHEJ-mediated disruption of TTR; base and prime editing modify what happens after the Cas9 finds its target. The present unit translates the mechanism into clinical effect and assumes the molecular details from 17.11.03 as background.
Induced pluripotent stem cells: Yamanaka factors, reprogramming, and regenerative medicine 18.11.04. Both Casgevy and iPS-cell therapy are regenerative-medicine tools that begin with the patient's own cells and end with a re-infused therapeutic product. Casgevy edits autologous haematopoietic stem cells in a lineage-restricted way; iPS technology reprograms somatic cells to pluripotency and differentiates them to any lineage. The two strategies bracket the design space of autologous cell therapy: editing within a native lineage versus reprogramming across lineages.
HIV/AIDS: retroviral biology, pathogenesis, ART, and the pandemic 35.02.05. HIV is itself a genome-integrating viral agent, and CRISPR-based strategies to excise integrated HIV provirus from host cells (the EBT-101 candidate of Excision BioTherapeutics) extend the present unit's modality space to infectious-disease targets. Both ART and Casgevy are modern-genetic-medicine treatments for diseases whose molecular basis was decoded in the late twentieth century, and both illustrate the long arc from molecular understanding to durable therapy.

Historical & philosophical context Master

Jennifer Doudna and Emmanuelle Charpentier, with Martin Jinek, Krzysztof Chylinski, Ines Fonfara and Michael Hauer, published the reconstitution of the Cas9-sgRNA programmable endonuclease in Science on 28 June 2012 ^[Jinek2012], establishing that a single protein guided by a single chimeric RNA could be redirected to any DNA target by changing the RNA sequence. Feng Zhang and colleagues, and independently George Church's group, extended the system to mammalian cells in January 2013 ^[Cong2013], removing the last practical barrier to human-genome editing. The 2020 Nobel Prize in Chemistry was awarded to Charpentier and Doudna for the development of the CRISPR-Cas9 method.

Haydar Frangoul and the CLIMB-121 investigators reported the primary efficacy of exa-cel in sickle cell disease in the New England Journal of Medicine in January 2021 ^{[Frangoul2021]}, the first randomised-cohort evidence that a CRISPR edit could cure a monogenic disease in humans. Julian Gillmore and colleagues reported in the New England Journal of Medicine in August 2021 that a single intravenous dose of NTLA-2001, the first in-vivo CRISPR therapy administered to humans, achieved 93% knockdown of serum transthyretin in patients with hereditary transthyretin amyloidosis ^{[Gillmore2021]}. The United States Food and Drug Administration approved Casgevy for sickle cell disease on 8 December 2023 and for transfusion-dependent beta-thalassemia in January 2024 ^[FDA2023], making Casgevy the first CRISPR-based therapy to enter routine clinical use and closing an eleven-year arc from basic discovery to approved drug.

Bibliography Master

@article{Jinek2012,
  author = {Jinek, M. and Chylinski, K. and Fonfara, I. and Hauer, M. and Doudna, J. A. and Charpentier, E.},
  title = {A programmable dual-{RNA}-guided {DNA} endonuclease in adaptive bacterial immunity},
  journal = {Science},
  volume = {337},
  number = {6096},
  pages = {816--821},
  year = {2012},
}

@article{Cong2013,
  author = {Cong, L. and Ran, F. A. and Cox, D. and Lin, S. and Barretto, R. and Habib, N. and Hsu, P. D. and Wu, X. and Jiang, W. and Marraffini, L. A. and Zhang, F.},
  title = {Multiplex genome engineering using {CRISPR}-{Cas} systems},
  journal = {Science},
  volume = {339},
  number = {6121},
  pages = {819--823},
  year = {2013},
}

@article{Mali2013,
  author = {Mali, P. and Yang, L. and Esvelt, K. M. and Aach, J. and Guell, M. and DiCarlo, J. E. and Norville, J. E. and Church, G. M.},
  title = {{RNA}-guided human genome engineering via {Cas9}},
  journal = {Science},
  volume = {339},
  number = {6121},
  pages = {823--826},
  year = {2013},
}

@article{Frangoul2021,
  author = {Frangoul, H. and Altshuler, D. and Cappellini, M. D. and Chen, Y.-S. and Domm, J. and Eustace, B. K. and Foell, M. and Gupta, A. and Hosseini, S. and Kulkarni, R. and others},
  title = {{CRISPR}-{Cas9} gene editing for sickle cell disease and beta-thalassemia},
  journal = {New England Journal of Medicine},
  volume = {384},
  number = {3},
  pages = {252--260},
  year = {2021},
}

@article{Esrick2024,
  author = {Esrick, E. B. and Lehmann, L. E. and Biffi, A. and Achebe, M. and Brendel, C. and Ciuculescu, M. F. and Daley, H. and Durham, E. and El-Bietar, J. and Eustace, B. K. and others},
  title = {Post-transplant outcomes in the {CRISPR}-{Cas9} gene editing for sickle cell disease ({CLIMB-121}) trial},
  journal = {New England Journal of Medicine},
  year = {2024},
}

@article{Locatelli2023,
  author = {Locatelli, F. and Thompson, A. A. and Kwiatkowski, J. L. and Thrasher, A. J. and Bhatia, M. and Dev, B. and Follenzi, A. and Gardner, K. and Glenth{\o}j, A. and Kang, Y. and others},
  title = {Betibeglogene autotemcel gene therapy for non-beta(0)/beta(0) genotype beta-thalassemia},
  journal = {New England Journal of Medicine},
  year = {2023},
}

@article{Gillmore2021,
  author = {Gillmore, J. D. and Gane, E. and Taubel, J. and Kao, J. and Fontana, M. and Maitland, M. L. and Seaver, J. and O'Connell, M. T. and Walsh, K. R. and Wood, K. and others},
  title = {{CRISPR}-{Cas9} in vivo gene editing for transthyretin amyloidosis},
  journal = {New England Journal of Medicine},
  volume = {385},
  number = {6},
  pages = {493--502},
  year = {2021},
}

@article{Musunuru2018,
  author = {Musunuru, K. and Chadwick, A. C. and Mizoguchi, T. and Garcia, S. P. and DeNizio, J. E. and Reiss, C. W. and Wang, K. and Iyer, S. and Giang, L. K. and Cohen, K. E. and others},
  title = {In vivo lipid nanoparticle editing of {PCSK9} in nonhuman primates and mouse},
  journal = {Nature},
  volume = {559},
  number = {7713},
  pages = {677--680},
  year = {2018},
}

@article{Anzalone2019,
  author = {Anzalone, A. V. and Randolph, P. B. and Davis, J. R. and Sousa, A. A. and Koblan, L. W. and Levy, J. M. and Chen, P. J. and Wilson, C. and Newby, G. A. and Raguram, A. and Liu, D. R.},
  title = {Search-and-replace genome editing without double-strand breaks or donor {DNA}},
  journal = {Nature},
  volume = {576},
  number = {7785},
  pages = {149--157},
  year = {2019},
}

@misc{FDA2023,
  author = {{U.S. Food and Drug Administration}},
  title = {{FDA} approves first gene therapies to treat patients with sickle cell disease},
  howpublished = {FDA News Release, 8 December 2023},
  year = {2023},
  note = {Casgevy (exagamglogene autotemcel), BLA 125787},
}

@book{DoudnaSternberg2017,
  author = {Doudna, J. A. and Sternberg, S. H.},
  title = {A Crack in Creation: Gene Editing and the Unthinkable Power to Control Evolution},
  publisher = {Houghton Mifflin Harcourt},
  address = {Boston},
  year = {2017},
}

@book{Watson2014,
  author = {Watson, J. D. and Baker, T. A. and Bell, S. P. and Gann, A. and Levine, M. and Losick, R.},
  title = {Molecular Biology of the Gene},
  edition = {7},
  publisher = {Pearson},
  address = {San Francisco},
  year = {2014},
}

Prerequisites

35.08.01

Tier anchors

beginner: Doudna–Sternberg, A Crack in Creation (Houghton Mifflin Harcourt, 2017), popular treatment of CRISPR's translation from bacterial immunity to human medicine; NIH 'CRISPR: Questions and Answers' primer; Amoeba Sisters CRISPR therapeutics overview
intermediate: Frangoul et al. 2021 N. Engl. J. Med. 384:252 (CLIMB-121 primary efficacy); Esrick et al. 2024 N. Engl. J. Med. (Casgevy extended follow-up); Watson, Baker, Bell, Gann, Levine, Losick — Molecular Biology of the Gene, 7th ed., Ch. 11
master: Jinek, Chylinski, Fonfara, Hauer, Doudna & Charpentier 2012 Science 337:816 (reconstituted Cas9); Cong, Ran, Cox et al. 2013 Science 339:819 (first mammalian editing); Frangoul et al. 2021 N. Engl. J. Med. 384:252 (CLIMB-121 sickle cell); Locatelli et al. 2023 N. Engl. J. Med. (CLIMB-111 beta-thalassemia); Gillmore et al. 2021 N. Engl. J. Med. 385:493 (NTLA-2001 first in-vivo CRISPR in humans); Musunuru et al. 2021 Nature (base editing of PCSK9); Anzalone et al. 2019 Nature 576:149 (prime editing); FDA BLA 125787 Casgevy label, 8 Dec 2023; 2020 Nobel Prize Charpentier–Doudna

References

Frangoul, H., Altshuler, D., Cappellini, M. D., Chen, Y.-S., Domm, J., Eustace, B. K., et al. — CRISPR-Cas9 Gene Editing for Sickle Cell Disease · N. Engl. J. Med. 384 (2021) 252-260; the CLIMB-121 primary efficacy report, 29/31 sickle-cell patients free of vaso-occlusive crises at >=12 months
Esrick, E. B., Lehmann, L. E., Biffi, A., Achebe, M., Brendel, C., Ciuculescu, M. F., et al. — Post-Transplant Outcomes in the CRISPR-Cas9 Gene Editing for Sickle Cell Disease · N. Engl. J. Med. (2024) extended CLIMB-121 follow-up; sustained HbF and transfusion independence
Locatelli, F., Thompson, A. A., Kwiatkowski, J. L., Thrasher, A. J., Bhatia, M., Dev, B., et al. — Betibeglogene Autotemcel Gene Therapy for Non-beta(0)/beta(0) Genotype beta-Thalassemia · CLIMB-111 beta-thalassemia cohort of exa-cel; transfusion-independence endpoint met
Gillmore, J. D., Gane, E., Taubel, J., Kao, J., Fontana, M., Maitland, M. L., et al. — CRISPR-Cas9 In Vivo Gene Editing for Transthyretin Amyloidosis · N. Engl. J. Med. 385 (2021) 493-502; NTLA-2001, the first in-human in-vivo CRISPR editing result, 93% serum TTR knockdown at 0.3 mg/kg
Jinek, M., Chylinski, K., Fonfara, I., Hauer, M., Doudna, J. A. & Charpentier, E. — A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity · Science 337 (2012) 816-821; the foundational CRISPR-Cas9 reconstitution paper
Cong, L., Ran, F. A., Cox, D., Lin, S., Barretto, R., Habib, N., Hsu, P. D., Wu, X., Jiang, W., Marraffini, L. A. & Zhang, F. — Multiplex genome engineering using CRISPR-Cas systems · Science 339 (2013) 819-823; first mammalian genome editing, opening the clinical translation path
Musunuru, K., Chadwick, A. C., Mizoguchi, T., Garcia, S. P., DeNizio, J. E., Reiss, C. W., Wang, K., Iyer, S., Giang, L. K., Cohen, K. E., et al. — In vivo lipid nanoparticle editing of PCSK9 in nonhuman primates and mouse · Nature 559 (2018) 677-680; preclinical foundation for VERVE-101 base editing of hypercholesterolemia
Anzalone, A. V., Randolph, P. B., Davis, J. R., Sousa, A. A., Koblan, L. W., Levy, J. M., et al. — Search-and-replace genome editing without double-strand breaks or donor DNA · Nature 576 (2019) 149-157; prime editing, the modality behind Prime Medicine's clinical pipeline
Doudna, J. A. & Sternberg, S. H. — A Crack in Creation: Gene Editing and the Unthinkable Power to Control Evolution · Houghton Mifflin Harcourt (2017); canonical popular exposition of the translation from bacterial immunity to therapeutic genome editing
U.S. Food and Drug Administration — FDA Approves First Gene Therapies to Treat Patients with Sickle Cell Disease · FDA News Release, 8 December 2023; Casgevy (exagamglogene autotemcel) BLA 125787 approval letter and prescribing information

Estimated time

beginner: 20m
intermediate: 50m
master: 90m