CD34+ hematopoietic stem cells are a small population of cells, found mainly in bone marrow, that can regenerate the entire blood and immune system. The “CD34+” label refers to a protein sitting on their surface that acts as a molecular flag, letting scientists and clinicians identify, count, and isolate them from among billions of ordinary blood cells. First identified in 1984, CD34 has become the single most important marker in stem cell transplantation, gene therapy, and the diagnosis of blood cancers. Yet the biology behind that surface protein is more nuanced than a simple on-off switch, and understanding it opens up questions about where these cells come from, how they are collected, and what modern medicine does with them.
The CD34 Protein Itself
CD34 is a transmembrane phosphoglycoprotein with a molecular weight of roughly 115 kDa. Its extracellular portion is heavily coated with sugar chains, which give it a bulky, negatively charged surface. That sugar coating is thought to play a role in cell-to-cell signaling and adhesion, though researchers have used CD34 primarily as a practical tag rather than fully understanding every function it performs. For nearly three decades, CD34 has been the go-to marker for identifying and purifying blood-forming stem cells ahead of bone marrow transplantation, and it has also been adopted as a marker for other tissue-specific stem cells, including muscle satellite cells and certain skin precursors.
Where CD34+ Stem Cells Live
Most CD34+ hematopoietic stem cells reside in the bone marrow, tucked into specialized microenvironments called niches. These niches are not random locations. Research in mice has shown that the majority of blood-forming stem cells sit in direct contact with cells that produce high levels of a signaling molecule called CXCL12. These CXCL12-abundant reticular cells, or CAR cells, cluster around the inner lining of blood vessels in the marrow (sinusoidal endothelium) or near the bone surface (endosteum). The interaction between CXCL12 and its receptor CXCR4 on the stem cell surface is what keeps these cells anchored in place and in a quiet, non-dividing state. When that signaling axis was experimentally disrupted in adult mice by deleting CXCR4, the number of stem cells in the marrow dropped sharply, and the remaining cells became more vulnerable to toxic injury.
This anchoring mechanism has direct clinical relevance. If you need to collect CD34+ cells from a patient’s bloodstream rather than drilling into bone, you first have to coax them out of the marrow. That process is called mobilization, and it works by loosening the CXCL12-CXCR4 grip.
How CD34+ Cells Are Mobilized and Collected
Under normal conditions, only a tiny fraction of CD34+ cells circulate in the blood. To harvest enough for a transplant, clinicians use drugs that push them out of the marrow and into the peripheral bloodstream. The standard approach uses granulocyte colony-stimulating factor (G-CSF), a growth factor that, over several days of injections, floods the marrow with maturing white cells and disrupts the niche environment enough that stem cells spill into the blood. A newer drug, plerixafor, works more directly: it blocks the CXCR4 receptor, severing the molecular handshake that holds stem cells in place. Clinical trials showed that combining plerixafor with G-CSF produces significantly more CD34+ cells in the blood than G-CSF alone, with few side effects.
Once mobilized, the cells are collected through apheresis, a process where blood is drawn, run through a machine that separates out the CD34-rich fraction, and the rest is returned to the patient. The harvested product is then counted, processed, and either transplanted fresh or frozen for later use.
Counting CD34+ Cells
Accurate counting matters because transplant outcomes depend heavily on how many CD34+ cells are infused. The standard method uses flow cytometry, a technology that passes individual cells through a laser beam and detects fluorescent antibodies stuck to CD34 on the cell surface. In the mid-1990s, the International Society of Hematotherapy and Graft Engineering (ISHAGE) established a standardized gating strategy, a set of rules for how to distinguish real CD34+ cells from debris and false signals, which became the global standard for clinical labs.
Since then, refinements have improved precision. A single-platform protocol developed for cord blood, using specialized counting tubes alongside the ISHAGE gating strategy, showed very high agreement with the original method and eliminated the need for a separate cell counter. That kind of standardization matters when transplant centers around the world need to compare results and predict outcomes reliably.
Three Sources, Three Personalities
CD34+ cells can be harvested from bone marrow, mobilized peripheral blood, or umbilical cord blood. Each source has a distinct cellular profile. Bone marrow contains the highest raw percentage of CD34+ cells among mononuclear cells, averaging around 5.4% in one large analysis, compared with roughly 1.8% in peripheral blood and 1.1% in cord blood. But those numbers are misleading if taken at face value, because the composition of the CD34+ population differs by source.
Bone marrow CD34+ cells include a large fraction of B-cell-committed progenitors (about 30% express the B-cell marker CD19), which are nearly absent in peripheral blood and cord blood products. Meanwhile, cord blood and mobilized peripheral blood contain a higher proportion of earlier myeloid progenitors. One study found that the proportion of these early progenitors was roughly 70% in cord blood, 64% in apheresis products, and only 27% in bone marrow. The practical implication: despite bone marrow’s higher overall CD34 count, you may need roughly 1.4 to 2.2 times as many bone marrow CD34+ cells as peripheral blood cells to achieve the same speed of myeloid recovery after transplant.
Cord blood CD34+ cells also stand out for their proliferative vigor. In culture, cord blood CD34+ cells expanded about 85-fold over ten days, compared with 56-fold for mobilized peripheral blood and 49-fold for bone marrow. Cord blood cells also contained a higher percentage of the most primitive CD38-negative fraction. These properties make cord blood an attractive source for transplanting children and smaller adults, though the total cell number in a single cord blood unit is limited, which can be a problem for larger recipients.
How Many Cells Do You Need for a Transplant?
The dose of CD34+ cells infused during a transplant is one of the strongest predictors of how quickly the patient’s blood counts recover and, ultimately, how well they do. For autologous transplants, where a patient receives their own stored cells, it is generally agreed that a minimum of 2.5 million CD34+ cells per kilogram of body weight is necessary for successful engraftment. At that threshold, blood counts do recover, but the process is slow and vulnerable to complications. Infusing 5 million cells per kilogram results in prompt recovery of all three major blood cell lines and is the preferred target.
In allogeneic transplants from sibling donors using peripheral blood stem cells, the picture gets more complicated. One study found that a dose in the range of 6 to 7 million CD34+ cells per kilogram was associated with the best overall survival and the lowest transplant-related mortality. Doses below 5 million per kilogram were linked to higher relapse rates but, interestingly, also to lower rates of chronic graft-versus-host disease. That trade-off, more cells potentially meaning fewer relapses but more immune complications, illustrates why transplant physicians carefully tailor the dose rather than simply aiming for the highest number possible.
CD34+ Cells in Gene Therapy
One of the most exciting frontiers for CD34+ cells is their use as vehicles for gene therapy. The basic idea is straightforward: remove a patient’s own CD34+ stem cells, fix the genetic defect in the lab, and transplant the corrected cells back. Because these are the cells that regenerate the entire blood system, a successful correction can, in theory, produce a lifelong cure.
Sickle cell disease has become a proving ground for this approach. In gene-addition therapy, a working copy of the beta-globin gene is delivered into patient CD34+ cells using a lentiviral vector, essentially a repurposed virus that inserts the gene into the cell’s DNA. After the modified cells are transplanted back, they produce functional hemoglobin alongside the sickle form.
Gene editing takes a different route. Instead of adding a new gene, researchers use tools like CRISPR/Cas9 to correct the existing mutation or to reactivate fetal hemoglobin, a form of hemoglobin that naturally compensates for the sickle defect. One group demonstrated that delivering CRISPR reagents to CD34+ cells from sickle cell patients led to gene correction rates averaging about 21%, and the edited cells went on to produce normal adult hemoglobin A, representing up to roughly 13% of total hemoglobin. A separate approach targeted a regulatory element of the BCL11A gene, which normally silences fetal hemoglobin production. Disrupting that element in CD34+ cells led to nearly complete editing at the target site, robust fetal hemoglobin induction, and resistance to sickling in the resulting red blood cells, all without detectable harmful effects on stem cell function. This strategy formed the basis for one of the first approved CRISPR-based therapies.
CD34 Beyond Blood
Although CD34 is most famous as a blood stem cell marker, the protein shows up in other tissues too. A subset of non-circulating endothelial cells lining smaller blood vessels express CD34. These CD34+ endothelial cells look different from their neighbors: they are more elongated, lack tight junctions, and tend to be quiescent. CD34 expression on these cells is concentrated on the inner (luminal) surface, though it can also appear on the outer membrane at the tips of growing blood vessel sprouts. Researchers believe these cells are involved in migration and adhesion during blood vessel remodeling.
CD34 is also found on vascular endothelial progenitor cells more broadly, and it has been detected on progenitor populations in muscle, skin, and other tissues. This wider expression has occasionally caused confusion. A biopsy staining positive for CD34 does not automatically mean blood stem cells are present; the context matters. In pathology, CD34 staining is used to identify certain soft tissue tumors and vascular abnormalities, not just hematopoietic cells.
The CD34-Negative Paradox
Here is where the story gets genuinely surprising: not all true hematopoietic stem cells are CD34-positive. Experiments transplanting human bone marrow cells that completely lacked CD34 into immunodeficient mice showed that these CD34-negative cells could produce long-term, multilineage engraftment, meaning they regenerated all blood cell types over an extended period. Even more striking, when CD34-negative cells engrafted, some of their progeny turned CD34-positive, and vice versa. In serial transplant experiments, both CD34-positive and CD34-negative cells from primary recipients could engraft secondary and tertiary hosts.
This reversibility means CD34 is not a permanent identity badge. A stem cell can toggle the protein on and off without losing its fundamental “stemness.” In practice, most clinical protocols still select for CD34+ cells because they are far easier to identify and collect in useful numbers. But the existence of CD34-negative stem cells is a reminder that the marker is a practical convenience, not a perfect definition of what a blood stem cell is. Some of the most primitive, deeply quiescent stem cells may sit in the CD34-negative compartment, only switching on CD34 when they begin to activate.
CD34+ Cells in Leukemia
In blood cancers, CD34 takes on a different significance. Many types of acute myeloid leukemia (AML) arise from or are sustained by a small population of leukemic stem cells that, like normal stem cells, express CD34. Distinguishing leukemic CD34+ cells from normal ones is critical for monitoring treatment response and predicting relapse. Research has identified the CD34+/CD38- fraction as the population most closely linked to AML progression, treatment resistance, and relapse.
The challenge is that leukemic stem cells often look very similar to normal stem cells under the microscope and even by standard flow cytometry. To tell them apart, researchers screen for additional “aberrant” surface markers that normal CD34+/CD38- cells do not carry. In AML, these markers show up inconsistently: CD25 appears on about 48% of cases, CD96 on 40%, CLL-1 on 68%, and IL-1RAP on 65%. In chronic myeloid leukemia (CML), the picture is more uniform, with leukemic stem cells almost invariably carrying a specific combination of markers including CD25, CD26, CD56, CD93, and IL-1RAP. These aberrant profiles are being explored as targets for immunotherapy and as tools for detecting minimal residual disease after treatment.
How Aging Reshapes the CD34+ Landscape
As people age, the bone marrow itself changes. Fat (adipose tissue) accumulates, and the space devoted to active blood cell production shrinks. A study comparing middle-aged and older adult bone marrow found a significant decrease in the hematopoietic area and an increase in fat content with age. Interestingly, the percentage of CD34+ cells within the remaining blood-forming tissue did not differ between the two groups. But when the researchers looked at the relationship more closely, they found a positive correlation between the density of CD34+ cells and age, accompanied by an increase in maturing myeloid cells.
What this likely reflects is a well-known phenomenon in aging hematopoiesis: the stem cell compartment becomes biased toward myeloid output (white blood cells, particularly of the innate immune system) at the expense of lymphoid output (the cells that drive adaptive immunity, like T and B cells). The CD34+ cells are still there, and may even be packed more densely into the shrinking available space, but they are doing a different job than they did in a younger person’s marrow. This myeloid skew is thought to contribute to the weaker immune responses and increased susceptibility to certain blood cancers seen in older adults.
Growing CD34+ Cells Outside the Body
One persistent limitation of cord blood transplantation is that a single cord blood unit may not contain enough CD34+ cells for an adult recipient. That has driven intense interest in expanding these cells in the lab before transplant. Two small molecules, UM171 and SR1, have moved into clinical protocols for this purpose. In laboratory studies, neither molecule did anything when added to cord blood cells on its own. But when combined with the growth factors that normally drive stem cell proliferation, both UM171 and SR1 significantly boosted the expansion of stem and progenitor cells from cord blood. The effect was most pronounced in purified CD34+ populations and in an intermediate fraction called lineage-negative cells.
Separately, researchers have found that manipulating the metabolic environment of CD34+ cells can steer their behavior. Blocking glutamine metabolism pushed stem and progenitor cells into a quiescent state, which actually improved their ability to engraft after transplantation. Adding ascorbic acid (vitamin C) to cultures had the opposite effect, promoting proliferation and enhancing the production of platelet-forming megakaryocytes both in the lab and in animal models. These findings suggest that the metabolic state of CD34+ cells at the time of transplant may matter as much as their sheer number.
Why Mouse and Human CD34 Biology Diverge
Much of what we know about stem cell biology comes from mouse studies, but CD34 behaves differently across species in ways that have caused real confusion. In steady-state mouse bone marrow, the stem cells with the strongest long-term regenerative capacity are actually CD34-negative or CD34-low, essentially the opposite of the human pattern. This species difference made it difficult to translate mouse findings directly to human clinical practice.
To investigate whether this was a true biological difference or a matter of gene regulation, researchers created transgenic mice carrying the entire human CD34 gene on an artificial chromosome. In these animals, the human CD34 transgene was expressed on the vast majority of functional stem cell populations, including those that were negative for the mouse version of CD34. In other words, the human gene’s regulatory machinery drives expression in stem cells even when the mouse gene’s machinery does not. This result confirmed that positive selection based on CD34 expression, the standard approach for purifying human stem cells, is valid for human biology despite what mouse studies might suggest. It also serves as a useful cautionary tale about assuming animal models map neatly onto human physiology.