Yang Liu

← Primer

HSC · Healing Source Code

造血干细胞的后代就是整个免疫系统——同一个性质，撑起了两条相反的治疗路径 One Cell's Progeny Is the Whole Immune System — How a Single Property Supports Two Opposite Therapies

2026-09-27 · 造血干细胞, CAR-T, 免疫耐受, 混合嵌合, 胸腺, 基因治疗

一个细胞的后代，是一整套自我定义

绝大多数细胞治疗讲的是「给病人一批细胞」。造血干细胞不一样：它的后代不是一批细胞，而是一整个免疫系统——T 细胞、B 细胞、NK 细胞、单核细胞和树突状细胞，都从它分化而来。更要紧的是，这套系统在发育过程中顺带完成了一件事：决定什么算「自己」。胸腺里的阴性选择、外周的调节性 T 细胞，都是这套供应链的产物。

一旦把这两层含义摊开，一个奇怪的对称就出现了。你可以往上游走，给造血干细胞装上工程化受体，让骨髓在余生里源源不断地送出带武器的免疫细胞；你也可以反过来走，利用它重写「自己」的边界，让免疫系统对某个原本会被攻击的抗原彻底沉默。前者为了让免疫系统更凶，后者为了让它更安静，用的却是同一个细胞、同一条发育通路。

往上游装受体：持久性从哪里来

今天临床上的 CAR-T，取的是病人外周血里已经成熟的 T 细胞。这条路走得通，但有两个反复出现的短板：细胞数量有限、易于耗竭；以及一旦肿瘤在数月甚至数年后重新出现，输进去的那批细胞往往已经不在了。把受体装到造血干/祖细胞上，逻辑上可以绕开这两点。

最早的可行性验证停在体外。2013 年，De Oliveira 等在《Human Gene Therapy》上把针对 CD19 的第一代与第二代 CAR 用慢病毒载体导入人造血干/祖细胞，然后在体外把它们分别向髓系和 NK 细胞方向诱导分化。在每细胞一到两个拷贝的条件下，基因导入没有损害分化和增殖，而分化出来的髓系细胞和 NK 细胞都能特异性地裂解 CD19 阳性靶细胞——含 CD28 结构域的第二代 CAR 在 NK 细胞里更有效。这是一份漂亮的体外结果，但它没有回答最关键的那个问题：经过胸腺发育出来的 T 细胞会是什么样子。

这个问题在同一年前后由另一组实验回答了。2011 年，Vatakis 等在《PNAS》上把一个 HLA-A*0201 限制性的黑色素瘤特异性 T 细胞受体（针对 MART-1）导入人造血干细胞，再移植进带有人胸腺植入物的 BLT 人源化小鼠。结果不只是长出了相当规模的肿瘤特异性初始 CD8⁺ T 细胞群，而且在肿瘤攻击之后，这些细胞在没有任何额外操作的情况下限制并清除了人黑色素瘤。作者特别强调了一件事：这些细胞经历了恰当的胸腺选择——对 HLA 不匹配的肿瘤没有反应，而在缺少人胸腺的动物里，什么杀伤都不发生。

胸腺在这里不是可有可无的中转站，而是这条路径的品质检验环节。它还带来一个别处见不到的好处：Seet 等 2017 年在《Nature Methods》上建立的人工胸腺类器官（artificial thymic organoid）系统显示，用经 TCR 改造的造血干/祖细胞起始培养，长出的 T 细胞几乎完全不表达内源性 TCR Vβ，与 Vβ 基因座的等位排斥一致——走胸腺这条路，天然地压住了外源受体与内源受体错配的隐患。

把这套思路装进 T 细胞受体之外的谱系，效果同样成立。2019 年，Zhu、Yang 等在《Cell Stem Cell》上把人恒定型自然杀伤 T 细胞（invariant natural killer T，iNKT）的 TCR 基因导入造血干细胞，在人造血干细胞移植的小鼠模型里高效、长期地生成了 HSC-iNKT 细胞；这些细胞在表型和功能上都接近内源人 iNKT，能用多种机制攻击肿瘤，在多个人源肿瘤移植模型里抑制了肿瘤生长，临床前安全性研究未见毒性或致瘤性。仍然是小鼠。

两年，在猴子身上

真正把证据推上一层的，是非人灵长类。2017 年，Zhen 等在《PLOS Pathogens》上给猪尾猴（pigtail macaque）移植了携带一种保护性 CD4 嵌合抗原受体（C46CD4CAR）的造血干/祖细胞，用来对付 SHIV 感染。携带 CAR 的细胞持续了两年以上，未见可测量的毒性，并且实现了多谱系植入——T 细胞、NK 细胞、B 细胞、单核与巨噬细胞里都能找到 CAR。抗病毒治疗撤药后，CAR 组动物的病毒反弹低于对照，表现出类似免疫记忆的反应；组织中 SHIV RNA 水平下降，肠道 CD4/CD8 比值显著高于对照。

2021 年，Barber-Axthelm 等在《JCI Insight》上用免疫组化把这件事定位到了具体部位：造血干细胞来源的 CAR 细胞归巢到组织内的病毒储存库，在淋巴组织生发中心、脑和胃肠道中以多谱系形式存活了接近两年。

这条线目前的诚实描述是：细胞层面已证可行，小鼠层面已证有效并经过胸腺教育，非人灵长类层面已证长期存活与组织归巢，而在人体，还没有看到造血干细胞来源 CAR 细胞的疗效数据。横在中间除了常规的安全性门槛，还有一个生物学上的硬约束——成人胸腺是退化的。让一个五十岁的患者从骨髓重新生成一整批胸腺教育过的工程化 T 细胞，和在幼年小鼠或幼年猴身上做同一件事，并不是一回事。

把同一个性质反过来用

胸腺既然能决定什么值得攻击，当然也能决定什么不值得攻击。这条反向路径的源头比 CAR 早了半个多世纪。1945 年，Owen 在《Science》上报告了共用胎盘血管吻合的孪生牛犊体内长期混存彼此的红细胞；1953 年，Billingham、Brent 和 Medawar 在《Nature》上把这件事变成了一个可操作的原理——「主动获得性耐受」。这个原理的现代形式是混合嵌合（mixed chimerism）：让受体的造血系统里稳定地混入供体来源的细胞，供体抗原便会作为「自己」参与胸腺教育，针对供体的新生 T 细胞在成熟前就被删除。1989 年，Sharabi 与 Sachs 在《Journal of Experimental Medicine》上证明，用一个非致死性的预处理方案就能在小鼠中诱导混合嵌合与永久的、供体特异的移植耐受。

把它做到人身上，是一次漫长的、结果参差的努力。2008 年，Kawai 等与 Scandling 等在同一期《New England Journal of Medicine》上分别报告了联合肾与造血细胞移植后撤除免疫抑制的病例。麻省总医院那一组的长期结果发表于 2014 年的《American Journal of Transplantation》，读起来克制得近乎冷峻：十名接受 HLA 不匹配联合肾与骨髓移植的受者，全部只出现了短暂嵌合；其中七人成功停用免疫抑制剂四年以上，至报告时四人已无免疫抑制地维持了 4.5 到 11.4 年，另三人在五到八年后因原病复发或慢性抗体介导的排斥而重新用药；十个移植肾里有三个因血栓性微血管病或排斥而失功。与免疫学条件相近、按常规免疫抑制治疗的活体供肾受者相比，长期停药者的移植后并发症显著更少。同一年代另一条路线，即 Leventhal 等 2012 年在《Science Translational Medicine》上报告的 HLA 不匹配联合肾与造血干细胞移植，把重点放在了避开移植物抗宿主病与植入综合征。

耐受究竟是怎么维持的？2015 年，Morris、Sykes 等在《Science Translational Medicine》上追踪了这些耐受患者体内供体反应性 T 细胞克隆的去向，给出了支持克隆删除的证据。这篇文章随后收到了同刊的一篇 Comment，对其分析方法提出了异议——把它算作一条已经定论的机制，为时尚早。

而当基础疾病本身就带着免疫缺陷时，这条路可以走得更干净。2022 年，Bertaina 等在《NEJM》上报告了三例 Schimke 免疫－骨骼发育不全（Schimke immuno-osseous dysplasia）患儿：先接受经 αβ T 细胞与 CD19 B 细胞去除的单倍体亲代移植，获得完全供体嵌合，随后接受同一供体的肾移植，并在术后第 30 天前撤除免疫抑制；三人在肾移植后 22 至 34 个月时肾功能正常且无需免疫抑制，其中一例出现二级皮肤急性移植物抗宿主病，对糖皮质激素有反应，无慢性 GVHD，无临床排斥。

把一个陌生蛋白介绍给免疫系统

耐受这条线最有说服力的临床证据，反而不来自器官移植，而来自酶替代治疗。在重型 I 型黏多糖贮积症（Hurler 综合征）中，输进去的 α-L-艾杜糖苷酶本身就是一个免疫原。2012 年，Saif 等在《Haematologica》上系统量化了这件事：八名在移植前后接受酶替代治疗的患者中，有 87.5%（七人）在接触该酶后产生了高滴度免疫反应，其中一部分出现酶催化活性抑制、一部分出现对有催化活性之酶的细胞摄取抑制，而高滴度抗体的出现通常早于既有疾病进展生物标志物的变化。关键的对照在后半段：二十名此前接受过异体造血干细胞移植的患者，在移植后一年以上受检时全部已对该酶耐受——尽管酶水平正常，尽管其中 22% 只是混合嵌合；从移植到耐受的中位时间是 101 天（范围 26 至 137 天）。

也就是说，只要让造血系统里长期存在一群会生产这个蛋白的细胞，免疫系统就会把它归入「自己」。这正是把造血干细胞基因治疗直接用作耐受诱导手段的理由，而在小鼠层面，这一点已被拆得很细。2022 年，Liang 等在《Molecular Therapy - Methods & Clinical Development》上于 Pompe 病小鼠中显示，慢病毒介导的表达即使在亚治疗剂量下也足以诱导对转基因产物、以及对随后给予的酶替代治疗的耐受；耐受在四到六周内建立，此后小鼠能耐受高剂量酶替代并维持肌功能改善与糖原蓄积下降，胸腺中出现造血干/祖细胞来源的细胞，指向中枢耐受。在血友病 A 小鼠中，Chen 等 2017 年在《Journal of Thrombosis and Haemostasis》上把血小板定向的凝血因子 VIII 基因治疗所诱导的耐受，定位为 CD4 T 细胞介导的过程。

这些都还是小鼠。但它们和 Hurler 那份人体数据指向同一个机制轮廓：抗原若由造血系统自己表达、并出现在胸腺，它就不再是外来物。

同一层性质的两面

造血干细胞的独特性因此可以收成一句话：它是唯一能同时改写「免疫系统有什么武器」和「免疫系统认谁为己」的细胞。前者让它成为持久性免疫治疗的上游入口，后者让它成为耐受诱导的天然载体。两种用法看起来相反，却共用同一条发育通路——都要经过胸腺，都依赖它的后代构成整个免疫系统这一事实。

而这层性质也正是它在异体现货疗法上最难解的约束：一个细胞若要长期供应整个免疫系统，它就必须被受体的免疫系统长期容忍——不是几周，而是几十年。这道题的另一半，留给讲免疫隐身工程的那一篇（#97）。

参考文献

  1. Owen RD. Immunogenetic consequences of vascular anastomoses between bovine twins. Science. 1945;102(2651):400–401. DOI
  2. Billingham RE, Brent L, Medawar PB. “Actively acquired tolerance” of foreign cells. Nature. 1953;172(4379):603–606. DOI
  3. Sharabi Y, Sachs DH. Mixed chimerism and permanent specific transplantation tolerance induced by a nonlethal preparative regimen. J Exp Med. 1989;169(2):493–502. DOI
  4. Kawai T, Cosimi AB, Spitzer TR, et al. HLA-mismatched renal transplantation without maintenance immunosuppression. N Engl J Med. 2008;358(4):353–361. DOI
  5. Scandling JD, Busque S, Dejbakhsh-Jones S, et al. Tolerance and chimerism after renal and hematopoietic-cell transplantation. N Engl J Med. 2008;358(4):362–368. DOI
  6. Leventhal J, Abecassis M, Miller J, et al. Chimerism and tolerance without GVHD or engraftment syndrome in HLA-mismatched combined kidney and hematopoietic stem cell transplantation. Sci Transl Med. 2012;4(124):124ra28. DOI
  7. Kawai T, Sachs DH, Sprangers B, et al. Long-term results in recipients of combined HLA-mismatched kidney and bone marrow transplantation without maintenance immunosuppression. Am J Transplant. 2014;14(7):1599–1611. DOI
  8. Morris H, DeWolf S, Robins H, et al. Tracking donor-reactive T cells: evidence for clonal deletion in tolerant kidney transplant patients. Sci Transl Med. 2015;7(272):272ra10. DOI
  9. Gorochov G, Larsen M, Parizot C, et al. Comment on “Tracking donor-reactive T cells: evidence for clonal deletion in tolerant kidney transplant patients”. Sci Transl Med. 2015;7(297). DOI
  10. Bertaina A, Grimm PC, Weinberg K, et al. Sequential stem cell–kidney transplantation in Schimke immuno-osseous dysplasia. N Engl J Med. 2022;386(24):2295–2302. DOI
  11. Saif MA, Bigger BW, Brookes KE, et al. Hematopoietic stem cell transplantation improves the high incidence of neutralizing allo-antibodies observed in Hurler’s syndrome after pharmacological enzyme replacement therapy. Haematologica. 2012;97(9):1320–1328. DOI
  12. Liang Q, Vlaar EC, Catalano F, et al. Lentiviral gene therapy prevents anti-human acid α-glucosidase antibody formation in murine Pompe disease. Mol Ther Methods Clin Dev. 2022;25:520–532. DOI
  13. Chen Y, Luo X, Schroeder JA, et al. Immune tolerance induced by platelet-targeted factor VIII gene therapy in hemophilia A mice is CD4 T cell mediated. J Thromb Haemost. 2017;15(10):1994–2004. DOI
  14. De Oliveira SN, Ryan C, Giannoni F, et al. Modification of hematopoietic stem/progenitor cells with CD19-specific chimeric antigen receptors as a novel approach for cancer immunotherapy. Hum Gene Ther. 2013;24(10):824–839. DOI
  15. Vatakis DN, Koya RC, Nixon CC, et al. Antitumor activity from antigen-specific CD8 T cells generated in vivo from genetically engineered human hematopoietic stem cells. Proc Natl Acad Sci USA. 2011;108(51):E1408–E1416. DOI
  16. Seet CS, He C, Bethune MT, et al. Generation of mature T cells from human hematopoietic stem and progenitor cells in artificial thymic organoids. Nat Methods. 2017;14(5):521–530. DOI
  17. Zhu Y, Smith DJ, Zhou Y, et al. Development of hematopoietic stem cell-engineered invariant natural killer T cell therapy for cancer. Cell Stem Cell. 2019;25(4):542–557.e9. DOI
  18. Zhen A, Peterson CW, Carrillo MA, et al. Long-term persistence and function of hematopoietic stem cell-derived chimeric antigen receptor T cells in a nonhuman primate model of HIV/AIDS. PLoS Pathog. 2017;13(12):e1006753. DOI
  19. Barber-Axthelm IM, Barber-Axthelm V, Sze KY, et al. Stem cell–derived CAR T cells traffic to HIV reservoirs in macaques. JCI Insight. 2021;6(1):e141502. DOI

A cell whose descendants are a definition of self

Most cell therapies mean one thing by their product: a batch of cells given to a patient. A hematopoietic stem cell is different. Its progeny is not a batch but an entire immune system — T cells, B cells, NK cells, monocytes, dendritic cells all descend from it. And in the course of building that system, it settles a second question along the way: what counts as self. Negative selection in the thymus and the peripheral regulatory T cell compartment are both downstream products of the same supply chain.

Lay those two meanings side by side and an odd symmetry appears. You can go upstream and fit the stem cell with an engineered receptor, so that the marrow ships armed effector cells for the rest of a patient’s life. Or you can go upstream for the opposite reason, and rewrite the boundary of self so that the immune system falls silent toward an antigen it would otherwise attack. One makes immunity fiercer, the other makes it quieter, and both run through the same cell and the same developmental pathway.

Putting the receptor upstream: where persistence comes from

Today’s clinical CAR-T products are made from mature T cells drawn from a patient’s blood. The approach works, but two limitations recur: the cells are finite in number and prone to exhaustion, and if a tumor reappears months or years later, the infused population is often long gone. Installing the receptor in hematopoietic stem and progenitor cells should, in principle, address both.

The earliest feasibility work stopped in the dish. In 2013, De Oliveira and colleagues reported in Human Gene Therapy that first- and second-generation CD19-directed CARs delivered to human HSPCs by lentiviral vector did not impair hematopoietic differentiation or proliferation at one to two copies per cell, and that the myeloid and NK cells subsequently differentiated in vitro specifically lysed CD19-positive targets — with the second-generation CAR carrying a CD28 domain more effective in NK cells. An elegant result, but it left the central question untouched: what would a T cell look like if it developed through a thymus?

That question had in fact been answered two years earlier. In 2011, Vatakis and colleagues reported in PNAS that an HLA-A*0201–restricted, melanoma-specific T cell receptor introduced into human hematopoietic stem cells and transplanted into BLT humanized mice — animals carrying a human thymic implant — produced a sizeable population of naive melanoma-specific CD8⁺ T cells that, on tumor challenge and without further manipulation, limited and cleared human melanoma in vivo. The authors emphasized one point in particular: the cells had undergone proper thymic selection. There was no response against non-HLA-matched tumors, and in animals lacking a human thymus, no killing of any kind occurred.

The thymus here is not an optional way station but the quality-control step of the whole route. It also buys something unavailable elsewhere. The artificial thymic organoid system published by Seet and colleagues in Nature Methods in 2017 showed that when cultures are initiated with TCR-engineered HSPCs, the T cells that emerge display a near-complete lack of endogenous TCR Vβ expression, consistent with allelic exclusion at the Vβ loci. Developing through a thymus suppresses mispairing between the introduced receptor and the endogenous one more or less for free.

The logic extends beyond conventional T cells. In 2019, Zhu, Yang and colleagues reported in Cell Stem Cell that delivering the invariant natural killer T (iNKT) cell TCR genes into hematopoietic stem cells generated HSC-iNKT cells efficiently and durably in a human HSC–engrafted mouse model; the cells closely resembled endogenous human iNKT cells, attacked tumors through several mechanisms, suppressed growth in multiple human tumor xenograft models, and showed no toxicity or tumorigenicity in preclinical safety studies. All of it in mice.

Two years, in monkeys

What moved the evidence up a level was work in nonhuman primates. In 2017, Zhen and colleagues reported in PLOS Pathogens that pigtail macaques transplanted with HSPCs carrying a protective CD4 chimeric antigen receptor, directed against SHIV infection, retained CAR-containing cells for more than two years with no measurable toxicity, and that those cells achieved multilineage engraftment — CAR was detectable in T cells, NK cells, B cells, and monocyte–macrophages. After antiretroviral therapy was withdrawn, viral rebound was lower in CAR animals than in controls, with what the authors described as a memory-like response; tissue-associated SHIV RNA fell and gut CD4/CD8 ratios were substantially higher than in controls.

In 2021, Barber-Axthelm and colleagues localized this in JCI Insight using immunohistochemistry: HSC-derived CAR cells trafficked to tissue-associated viral reservoirs and persisted, multilineage, for nearly two years in lymphoid germinal centers, the brain, and the gastrointestinal tract.

The honest summary of this line reads as follows. Feasible in vitro; effective and thymically educated in mice; long-lived and tissue-homing in nonhuman primates; and in humans, no efficacy data yet for HSC-derived CAR cells. Between the monkey and the patient lies not only the usual safety bar but a hard biological constraint: the adult thymus has involuted. Regenerating a full cohort of thymus-educated engineered T cells from the marrow of a fifty-year-old is not the same exercise as doing it in a juvenile mouse or a juvenile macaque.

Running the same property backwards

If the thymus can decide what is worth attacking, it can equally decide what is not. The reverse route is older than CAR by more than half a century. In 1945, Owen reported in Science that bovine twins sharing placental vascular anastomoses carried each other’s erythrocytes long term; in 1953, Billingham, Brent and Medawar turned the observation into an operational principle in Nature — actively acquired tolerance. Its modern form is mixed chimerism: establish a stable population of donor-derived hematopoietic cells in the recipient, and donor antigen joins thymic education as self, so that newly arising donor-reactive T cells are deleted before they mature. In 1989, Sharabi and Sachs showed in the Journal of Experimental Medicine that a nonlethal preparative regimen was enough to induce mixed chimerism and permanent, donor-specific transplantation tolerance in mice.

Carrying that into people has been a long effort with uneven results. In 2008, Kawai and colleagues and Scandling and colleagues each reported cases of immunosuppression withdrawal after combined kidney and hematopoietic cell transplantation, in the same issue of the New England Journal of Medicine. The long-term outcome of the Massachusetts General series, published in the American Journal of Transplantation in 2014, reads almost austerely: of ten recipients of HLA-mismatched combined kidney and bone marrow transplantation, all developed only transient chimerism; in seven, immunosuppression was successfully discontinued for four years or more; at the time of the report four remained immunosuppression-free for 4.5 to 11.4 years, while three required reinstitution after five to eight years because of recurrence of the original disease or chronic antibody-mediated rejection; three of the ten allografts failed from thrombotic microangiopathy or rejection. Compared with immunologically similar living-donor kidney recipients on conventional immunosuppression, the long-term drug-free survivors had significantly fewer post-transplant complications. A parallel route from the same era — the HLA-mismatched combined kidney and hematopoietic stem cell transplantation reported by Leventhal and colleagues in Science Translational Medicine in 2012 — focused on avoiding graft-versus-host disease and engraftment syndrome.

How is the tolerance maintained? In 2015, Morris, Sykes and colleagues tracked the fate of donor-reactive T cell clones in these tolerant patients in Science Translational Medicine and reported evidence for clonal deletion. The paper drew a published Comment in the same journal disputing the analysis, so treating clonal deletion as a settled mechanism would be premature.

Where the underlying disease already carries an immune defect, the route can be run more cleanly. In 2022, Bertaina and colleagues reported in the New England Journal of Medicine three children with Schimke immuno-osseous dysplasia who first received an αβ T cell– and CD19 B cell–depleted haploidentical parental graft, achieved full donor chimerism, then received a kidney from the same donor with immunosuppression withdrawn by day 30. All three had normal renal function without immunosuppression at 22 to 34 months after the kidney transplant. One developed grade II cutaneous acute GVHD that responded to glucocorticoids; there was no chronic GVHD and no clinical rejection.

Introducing a foreign protein to the immune system

The most persuasive clinical evidence for this second route comes not from organ transplantation but from enzyme replacement therapy. In severe mucopolysaccharidosis type I — Hurler syndrome — the infused α-L-iduronidase is itself an immunogen. In 2012, Saif and colleagues quantified the problem in Haematologica: among eight patients receiving enzyme therapy before and during transplantation, 87.5% (seven) mounted high-titer responses after exposure to the enzyme, some showing catalytic inhibition, some showing inhibition of cellular uptake of catalytically active enzyme, and high antibody titers generally preceded changes in established biomarkers of disease progression. The control lies in the second half of the study: twenty patients previously given an allogeneic hematopoietic stem cell transplant were all tolerized when tested a year or more afterward — despite normal enzyme levels, and despite 22% of them being only mixed chimeras. The median time from transplant to tolerance was 101 days, with a range of 26 to 137.

Sustain a population of cells in the hematopoietic system that manufactures the protein, in other words, and the immune system files it under self. That is the rationale for using hematopoietic stem cell gene therapy directly as a tolerance-induction strategy, and in mice the mechanism has been taken apart in some detail. In 2022, Liang and colleagues showed in Molecular Therapy — Methods & Clinical Development that in murine Pompe disease, lentiviral expression at a subtherapeutic dose was sufficient to induce tolerance both to the transgene product and to subsequently administered enzyme replacement therapy; tolerance was established within four to six weeks, after which the mice tolerated high enzyme doses while retaining improved muscle function and reduced glycogen accumulation, and HSPC-derived cells appeared in the thymus, pointing to central tolerance. In hemophilia A mice, Chen and colleagues reported in the Journal of Thrombosis and Haemostasis in 2017 that the tolerance induced by platelet-targeted factor VIII gene therapy is CD4 T cell–mediated.

These remain mouse results. But they and the human Hurler data outline the same mechanism: an antigen expressed by the hematopoietic system itself, and presented in the thymus, stops being foreign.

Two faces of one property

The singularity of the hematopoietic stem cell can be compressed into a sentence: it is the only cell that can rewrite both what weapons the immune system carries and whom it counts as its own. The first makes it the upstream entrance to durable immunotherapy; the second makes it a native vehicle for tolerance induction. The two uses look opposed, yet they share one developmental pathway — both pass through the thymus, and both rest on the fact that this cell’s descendants constitute the entire immune system.

That same property is also its hardest constraint in off-the-shelf allogeneic therapy. A cell that is to supply an immune system for decades must itself be tolerated by the recipient’s immune system for decades — not weeks. The other half of that problem belongs to the installment on immune cloaking (#97).

References

  1. Owen RD. Immunogenetic consequences of vascular anastomoses between bovine twins. Science. 1945;102(2651):400–401. DOI
  2. Billingham RE, Brent L, Medawar PB. “Actively acquired tolerance” of foreign cells. Nature. 1953;172(4379):603–606. DOI
  3. Sharabi Y, Sachs DH. Mixed chimerism and permanent specific transplantation tolerance induced by a nonlethal preparative regimen. J Exp Med. 1989;169(2):493–502. DOI
  4. Kawai T, Cosimi AB, Spitzer TR, et al. HLA-mismatched renal transplantation without maintenance immunosuppression. N Engl J Med. 2008;358(4):353–361. DOI
  5. Scandling JD, Busque S, Dejbakhsh-Jones S, et al. Tolerance and chimerism after renal and hematopoietic-cell transplantation. N Engl J Med. 2008;358(4):362–368. DOI
  6. Leventhal J, Abecassis M, Miller J, et al. Chimerism and tolerance without GVHD or engraftment syndrome in HLA-mismatched combined kidney and hematopoietic stem cell transplantation. Sci Transl Med. 2012;4(124):124ra28. DOI
  7. Kawai T, Sachs DH, Sprangers B, et al. Long-term results in recipients of combined HLA-mismatched kidney and bone marrow transplantation without maintenance immunosuppression. Am J Transplant. 2014;14(7):1599–1611. DOI
  8. Morris H, DeWolf S, Robins H, et al. Tracking donor-reactive T cells: evidence for clonal deletion in tolerant kidney transplant patients. Sci Transl Med. 2015;7(272):272ra10. DOI
  9. Gorochov G, Larsen M, Parizot C, et al. Comment on “Tracking donor-reactive T cells: evidence for clonal deletion in tolerant kidney transplant patients”. Sci Transl Med. 2015;7(297). DOI
  10. Bertaina A, Grimm PC, Weinberg K, et al. Sequential stem cell–kidney transplantation in Schimke immuno-osseous dysplasia. N Engl J Med. 2022;386(24):2295–2302. DOI
  11. Saif MA, Bigger BW, Brookes KE, et al. Hematopoietic stem cell transplantation improves the high incidence of neutralizing allo-antibodies observed in Hurler’s syndrome after pharmacological enzyme replacement therapy. Haematologica. 2012;97(9):1320–1328. DOI
  12. Liang Q, Vlaar EC, Catalano F, et al. Lentiviral gene therapy prevents anti-human acid α-glucosidase antibody formation in murine Pompe disease. Mol Ther Methods Clin Dev. 2022;25:520–532. DOI
  13. Chen Y, Luo X, Schroeder JA, et al. Immune tolerance induced by platelet-targeted factor VIII gene therapy in hemophilia A mice is CD4 T cell mediated. J Thromb Haemost. 2017;15(10):1994–2004. DOI
  14. De Oliveira SN, Ryan C, Giannoni F, et al. Modification of hematopoietic stem/progenitor cells with CD19-specific chimeric antigen receptors as a novel approach for cancer immunotherapy. Hum Gene Ther. 2013;24(10):824–839. DOI
  15. Vatakis DN, Koya RC, Nixon CC, et al. Antitumor activity from antigen-specific CD8 T cells generated in vivo from genetically engineered human hematopoietic stem cells. Proc Natl Acad Sci USA. 2011;108(51):E1408–E1416. DOI
  16. Seet CS, He C, Bethune MT, et al. Generation of mature T cells from human hematopoietic stem and progenitor cells in artificial thymic organoids. Nat Methods. 2017;14(5):521–530. DOI
  17. Zhu Y, Smith DJ, Zhou Y, et al. Development of hematopoietic stem cell-engineered invariant natural killer T cell therapy for cancer. Cell Stem Cell. 2019;25(4):542–557.e9. DOI
  18. Zhen A, Peterson CW, Carrillo MA, et al. Long-term persistence and function of hematopoietic stem cell-derived chimeric antigen receptor T cells in a nonhuman primate model of HIV/AIDS. PLoS Pathog. 2017;13(12):e1006753. DOI
  19. Barber-Axthelm IM, Barber-Axthelm V, Sze KY, et al. Stem cell–derived CAR T cells traffic to HIV reservoirs in macaques. JCI Insight. 2021;6(1):e141502. DOI