Yang Liu

← Primer

HSC · Healing Source Code

BCL11A——一个转录因子如何成为血红蛋白病治疗的枢纽 BCL11A: how one transcription factor became the linchpin of hemoglobinopathy therapy

2026-07-12 · BCL11A, 血红蛋白病, 基因编辑, 靶点

每个人在出生之前,血液里流淌的都不是成年后那种血红蛋白。胎儿有自己的一套——胎儿血红蛋白(HbF),它更擅长从母体那里夺取氧气。出生之后不久,身体会悄悄地把这套系统关掉,换上成人版本。对绝大多数人来说,这个切换无关紧要;可对镰刀型贫血和 β-地中海贫血的患者而言,被换上的恰恰是那个坏掉的零件,而那套本可以顶上来的胎儿备份,就这样被锁进了抽屉。过去十几年里,科学家一点一点找到了上锁的那只手,并最终学会了重新把抽屉打开。这只手,叫 BCL11A。

一个统计信号

故事的起点并不是一个漂亮的假说,而是一堆统计数字。人群里有些人天生就能把较多的胎儿血红蛋白保留到成年,而这些人如果同时患有血红蛋白病,病情往往轻得多——这个现象很早就被临床注意到了。到了 2000 年代末,全基因组关联研究(GWAS)开始有能力回答”到底是基因组的哪一处在左右这件事”,答案反复指向 2 号染色体上一个基因附近的常见变异,它就是 BCL11A。

但 GWAS 有一个众所周知的局限:它只能告诉你”这个位点和性状有关联”,却不能告诉你这个基因是不是真正的原因,更不能告诉你它能不能被拿来当药物靶点。一个关联信号,和一个可以下手的开关,之间还隔着好几层证明。

从相关,到因果

把这层窗户纸捅破的,是波士顿儿童医院 Stuart Orkin 实验室。2008 年,Sankaran 等人在 Science 上给出了第一组功能证据:那些”高 HbF”的 BCL11A 基因型,对应的恰恰是更低的 BCL11A 表达;而当他们在成人红系前体细胞里把 BCL11A 敲低,沉睡的胎儿血红蛋白便强劲地回升。更关键的是,BCL11A 蛋白直接结合在 β-珠蛋白基因簇的多个位置上——它不是旁观者,而是亲手执行沉默的那个角色。这篇论文第一次把 BCL11A 明确写成”重新激活 HbF 的治疗靶点”。

细胞里的证据还需要在活体里被验证。次年,同一实验室把完整的人类 β-珠蛋白基因簇搬进小鼠,再敲除小鼠的 BCL11A,结果本该在发育中被关掉的人类胎儿型基因,顽固地继续开着——沉默失败了。这说明 BCL11A 不只是”参与”,而是驱动这场发育切换的核心。到 2010 年,Xu 等人进一步拆开了它的作案手法:BCL11A 通过把染色质拉成特定的空间构象、并联合 SOX6、GATA1 等搭档,共同把胎儿基因压下去。至此,从”相关”到”因果”再到”机制”,这条链已经严丝合缝。

不能直接敲掉它

然而,知道了开关在哪,离能安全地拨动它,还差最关键的一步。BCL11A 并不是一个只管血红蛋白的专职基因——它在造血干细胞的维持、在 B 淋巴细胞的发育里同样不可或缺。如果简单粗暴地把整个 BCL11A 敲掉,固然能放出 HbF,却会顺带破坏正常的造血和免疫。这样的”治疗”没人敢用。

破局点出现在 2013 年。Bauer 等人在 Science 上注意到,那个最初被 GWAS 标记出来的变异,并不落在 BCL11A 的蛋白编码区,而是落在一段红系特异的enhancer上——一段只在红细胞谱系里负责驱动 BCL11A 表达的调控序列。他们通过基因工程证明:这段enhancer在红系细胞里是必需的,在 B 淋巴细胞里却不是。这意味着,只要精确地破坏这一小段enhancer,就能在红细胞里把 BCL11A 调暗、把胎儿血红蛋白放出来,而它在造血干细胞和免疫细胞里的本职工作几乎不受影响。论文把这段序列称为”β-血红蛋白病治疗性基因组工程的一个有吸引力的靶点”——这句带着克制的预言,日后一字一句地成真了。

这一步之所以是整条故事的枢纽,是因为它把一个”太重要以至于不能碰”的基因,变成了”可以在正确的地方、正确的谱系里安全地碰一下”的靶点。选靶的智慧,不在于找到一个强效的开关,而在于找到一个能被局部、精准、且只在需要处拨动的开关。

走到病床

有了干净的靶点,基因编辑技术只是补上了最后一块拼图。2020 年前后,一套流程被搭了起来:从患者体内采集自身的 CD34⁺ 造血干细胞,在体外用 CRISPR-Cas9 剪断那段红系enhancer,再经过清髓,把编辑后的细胞回输——让它们在骨髓里重新定居,从此源源不断地产出携带胎儿血红蛋白的红细胞。

2021 年,Frangoul 等人在 New England Journal of Medicine 报告了首批两名患者,一名 β-地中海贫血、一名镰刀型贫血。在用健康供者细胞验证这套编辑方案时,该位点约八成的等位基因被成功修改、且未见脱靶;两名患者随后各自回输了以同样方式编辑的自体造血干细胞,一年多以后,他们的骨髓和血液里都维持着高比例的编辑,胎儿血红蛋白广泛地、几乎在每一个红细胞里回升,双双摆脱了输血依赖,镰刀型贫血的那位患者不再发作血管闭塞危象。这套疗法后来以 exa-cel(商品名 Casgevy)之名,成为第一个获批上市的 CRISPR 基因编辑疗法——人类第一次把”改写基因组”从实验室送进了药房。

还没解决的部分

把 Casgevy 摆上药架,并不意味着 BCL11A 这条路已经走到尽头。眼下最刺眼的限制,其实不在靶点本身,而在流程:它仍然是一场完整的移植——需要采集、体外操作、清髓预处理,费用高达百万美元级别,只有少数中心能做。对于全球绝大多数血红蛋白病患者集中的地区,这样的疗法近乎可望而不可即。

更深一层的问题则关乎生物学本身。重新放出的胎儿血红蛋白究竟够不够、在红细胞之间是否分布得足够均匀,直接决定了疗效的上限,尤其对镰刀型贫血,需要足够高的 HbF 才能真正压住镰变——而这受制于个体的生物学,并不是把编辑效率调得更高就能线性解决的。CRISPR 依赖 DNA double-strand break 来工作,这背后潜藏的大片段缺失、p53 反应等风险,也随着这个靶点一同进入了人体(这部分留到讲安全性时再细说)。此外,BCL11A 这只开关远比”一段enhancer”复杂:2020 年 Basak 等人发现,它还受 LIN28B 在蛋白翻译层面的调控——我们目前只不过碰到了这套系统的一部分。

一个可以被复制的范式

回头看,BCL11A 留下的最持久的东西,或许不是某一款药,而是一条被走通的路径:从人类遗传学的一个统计信号出发,锁定一个发育开关,证明它的因果,再找到一段能被谱系特异地拨动的调控序列,最后用基因编辑把它安全地调暗。它示范的,是”绕过病因、重启备份”这一整套思路——不去逐一修复成百上千种致病突变,而是唤醒一套人体本就拥有的替代方案。

沿着这条路往前看,下一程几乎是清晰的:用 base editing 或 epigenetic editing 取代粗放的 double-strand break,让改动更温和;把整套流程从”体外 + 清髓”推向体内,让它有朝一日能压缩成一针。当那一天到来时,今天这个从统计数字里被打捞出来、又被打磨得如此干净的靶点,才算真正走完了它从实验室到世界的旅程。

参考文献

  1. Sankaran VG, et al. Human fetal hemoglobin expression is regulated by the developmental stage-specific repressor BCL11A. Science. 2008;322(5909):1839-42. DOI
  2. Sankaran VG, et al. Developmental and species-divergent globin switching are driven by BCL11A. Nature. 2009;460(7259):1093-7. DOI
  3. Xu J, et al. Transcriptional silencing of γ-globin by BCL11A involves long-range interactions and cooperation with SOX6. Genes Dev. 2010;24(8):783-98. DOI
  4. Bauer DE, et al. An erythroid enhancer of BCL11A subject to genetic variation determines fetal hemoglobin level. Science. 2013;342(6155):253-7. DOI
  5. Basak A, et al. Control of human hemoglobin switching by LIN28B-mediated regulation of BCL11A translation. Nat Genet. 2020;52(2):138-145. DOI
  6. Frangoul H, et al. CRISPR-Cas9 Gene Editing for Sickle Cell Disease and β-Thalassemia. N Engl J Med. 2021;384(3):252-260. DOI
  7. Locatelli F, et al. Autologous gene therapy for hemoglobinopathies: From bench to patient’s bedside. Mol Ther. 2024;32(5):1202-1218. DOI

Before any of us is born, the blood coursing through us is not the hemoglobin of adulthood. The fetus has its own version—fetal hemoglobin (HbF), which is better at wresting oxygen away from the mother. Soon after birth, the body quietly shuts this system down and switches over to the adult version. For the vast majority of people, the changeover is of no consequence; but for patients with sickle cell disease and β-thalassemia, the part that gets swapped in is precisely the broken one, while the fetal backup that could have stepped in is locked away in a drawer. Over the past decade and more, scientists have, piece by piece, found the hand that turned the lock—and eventually learned how to open the drawer again. That hand is called BCL11A.

A statistical signal

The story does not begin with an elegant hypothesis but with a pile of statistics. Some people are naturally able to retain more fetal hemoglobin into adulthood, and among those who also have a hemoglobinopathy, the disease tends to be far milder—a phenomenon clinicians had noticed long ago. By the late 2000s, genome-wide association studies (GWAS) had become capable of answering the question of “which part of the genome is actually driving this,” and the answer pointed, again and again, to a common variant near a gene on chromosome 2: BCL11A.

But GWAS has a well-known limitation: it can tell you that “this locus is associated with the trait,” yet it cannot tell you whether the gene is the true cause, still less whether it can be turned into a drug target. Between an association signal and an actionable switch lie several more layers of proof.

From correlation to causation

The group that broke through that final layer was Stuart Orkin’s laboratory at Boston Children’s Hospital. In 2008, Sankaran and colleagues published in Science the first set of functional evidence: the “high-HbF” BCL11A genotypes corresponded precisely to lower BCL11A expression; and when they knocked down BCL11A in adult erythroid precursor cells, the dormant fetal hemoglobin came surging back. More decisively, the BCL11A protein binds directly at multiple sites across the β-globin gene cluster—it is not a bystander but the very agent that carries out the silencing. This paper was the first to name BCL11A explicitly as a “therapeutic target for reactivating HbF.”

Evidence in cells still needed to be confirmed in a living organism. The following year, the same laboratory moved the entire human β-globin gene cluster into mice and then knocked out the mouse’s BCL11A; the result was that the human fetal-type genes, which should have been switched off during development, stubbornly stayed on—the silencing had failed. This showed that BCL11A is not merely “involved” but is the core driver of this developmental switch. By 2010, Xu and colleagues took the mechanism further apart: BCL11A represses the fetal genes by pulling chromatin into a specific spatial conformation and by joining forces with partners such as SOX6 and GATA1. From “correlation” to “causation” to “mechanism,” the chain was now airtight.

You can’t just knock it out

Knowing where the switch is, however, is still one crucial step short of being able to flip it safely. BCL11A is not a single-purpose gene that governs only hemoglobin—it is equally indispensable in the maintenance of hematopoietic stem cells and in the development of B lymphocytes. Crudely knocking out BCL11A in its entirety would indeed release HbF, but it would also wreck normal hematopoiesis and immunity along the way. No one would dare use such a “cure.”

The breakthrough came in 2013. Bauer and colleagues noted in Science that the variant originally flagged by GWAS does not fall within the protein-coding region of BCL11A but within an erythroid-specific enhancer—a regulatory sequence that drives BCL11A expression only in the red-cell lineage. Through genetic engineering, they demonstrated that this enhancer is required in erythroid cells but not in B lymphocytes. This meant that precisely disrupting this small enhancer would dim BCL11A in red cells and release fetal hemoglobin, while its day job in hematopoietic stem cells and immune cells would be left almost untouched. The paper called this sequence “an attractive target for therapeutic genome engineering of the β-hemoglobinopathies”—a restrained prediction that, in the years to come, would come true word for word.

The reason this step is the linchpin of the whole story is that it turned a gene that was “too important to touch” into a target that “can be touched safely, in the right place and in the right lineage.” The wisdom of target selection lies not in finding a potent switch but in finding one that can be flipped locally, precisely, and only where it is needed.

To the bedside

With a clean target in hand, gene-editing technology merely supplied the last piece of the puzzle. Around 2020, a workflow was assembled: harvest the patient’s own CD34⁺ hematopoietic stem cells, use CRISPR-Cas9 in vitro to cut that erythroid enhancer, then, after myeloablation, reinfuse the edited cells—letting them resettle in the bone marrow and, from then on, continuously produce red cells carrying fetal hemoglobin.

In 2021, Frangoul and colleagues reported the first two patients in the New England Journal of Medicine, one with β-thalassemia and one with sickle cell disease. When the editing approach was validated using healthy-donor cells, about eighty percent of the alleles at this locus were successfully modified, with no off-target effects observed; the two patients then each received a reinfusion of their own hematopoietic stem cells edited in the same way. More than a year later, both bone marrow and blood maintained a high proportion of editing, fetal hemoglobin had risen broadly—in nearly every red cell—and both patients became free of transfusion dependence, with the sickle cell patient no longer experiencing vaso-occlusive crises. This therapy would later become known as exa-cel (brand name Casgevy) and the first CRISPR gene-editing therapy approved for market—the first time humanity had carried “rewriting the genome” out of the laboratory and into the pharmacy.

What remains unsolved

Putting Casgevy on the pharmacy shelf does not mean the BCL11A path has reached its end. The most glaring limitation right now lies not in the target itself but in the process: it is still a full transplant—requiring harvest, in vitro manipulation, and myeloablative conditioning, at a cost on the order of a million dollars, and only a handful of centers can perform it. For the regions where the vast majority of the world’s hemoglobinopathy patients are concentrated, such a therapy is all but out of reach.

A deeper problem concerns the biology itself. Whether the newly released fetal hemoglobin is sufficient, and whether it is distributed evenly enough among red cells, directly sets the ceiling on efficacy—especially for sickle cell disease, where high enough HbF is needed to truly suppress sickling—and this is constrained by an individual’s biology, not something that can be resolved linearly just by pushing editing efficiency higher. CRISPR relies on a DNA double-strand break to work, and the risks lurking behind it—large deletions, the p53 response—have entered the human body along with this target (a matter left for the discussion of safety). Moreover, the BCL11A switch is far more complex than “a single enhancer”: in 2020, Basak and colleagues found that it is also regulated by LIN28B at the level of protein translation—we have so far touched only a part of this system.

A paradigm that can be replicated

Looking back, the most enduring thing BCL11A leaves behind is perhaps not a particular drug but a path proven walkable: starting from a single statistical signal in human genetics, pinning down a developmental switch, proving its causality, then finding a regulatory sequence that can be flipped in a lineage-specific way, and finally using gene editing to safely dim it. What it demonstrates is the whole idea of “bypassing the cause and rebooting the backup”—not repairing hundreds or thousands of disease-causing mutations one by one, but awakening an alternative that the human body already possesses.

Looking further down this path, the next leg is almost clear: replace the blunt double-strand break with base editing or epigenetic editing, so the change is gentler; push the entire workflow from “in vitro + myeloablation” toward in vivo, so that one day it might be compressed into a single injection. When that day comes, this target—fished out of a set of statistics and polished so clean—will finally have completed its journey from the laboratory to the world.

References

  1. Sankaran VG, et al. Human fetal hemoglobin expression is regulated by the developmental stage-specific repressor BCL11A. Science. 2008;322(5909):1839-42. DOI
  2. Sankaran VG, et al. Developmental and species-divergent globin switching are driven by BCL11A. Nature. 2009;460(7259):1093-7. DOI
  3. Xu J, et al. Transcriptional silencing of γ-globin by BCL11A involves long-range interactions and cooperation with SOX6. Genes Dev. 2010;24(8):783-98. DOI
  4. Bauer DE, et al. An erythroid enhancer of BCL11A subject to genetic variation determines fetal hemoglobin level. Science. 2013;342(6155):253-7. DOI
  5. Basak A, et al. Control of human hemoglobin switching by LIN28B-mediated regulation of BCL11A translation. Nat Genet. 2020;52(2):138-145. DOI
  6. Frangoul H, et al. CRISPR-Cas9 Gene Editing for Sickle Cell Disease and β-Thalassemia. N Engl J Med. 2021;384(3):252-260. DOI
  7. Locatelli F, et al. Autologous gene therapy for hemoglobinopathies: From bench to patient’s bedside. Mol Ther. 2024;32(5):1202-1218. DOI