CD34 is a commonly used and clinically significant marker  for human hematopoietic progenitors and stem cells. Flow cytometric separation of distinct subsets within the CD34+ population is used to create a comprehensive map of human hematopoietic progenitor and stem cells. The possibility of separating different progenitor/ stem cell subsets could be used in clinic for differential mobilization, outcome prediction and prognosis, graft manipulation and engineering. Nevertheless, It was shown more than a decade ago that there are hematopoietic stem cells (HSC) in CD34-negative population in human. I was wondering what has changed since then and how can we apply our knowledge about CD34-negative HSC?
Confirmation of HSC existence in CD34-negative population
In 1998 Zanjani for the first time identified  stem cell activity in CD34-negative fraction of Lineage-negative (Lin-) bone marrow cells, based on the human/sheep xenogeneic model. In the same time Mickie Bhatia  from John Dick lab showed that CD34-negative cells (which are also Lin-) isolated from bone marrow and cord blood, are capable of long-term multilineage repopulation in SCID mice (so-called SCID-repopulating cells – SRC). One year earlier, Margaret Goodell , isolated so-called side population (SP) HSC, which were negative for CD34.
Studying stem cell activity
Bhatia studied  two subsets of Lin-/CD34- HSC, based on expression of CD38. Both subsets repopulated SCID/NOD mice in long-term, but CD38+ much less. Magically, Sonoda group was not able to reproduce  SCID/NOD mice repopulation by intavenous injection (IV) of Lin-/CD34- cells, derived from cord blood (CB). In their hands, CB-derived CD34- cells repopulated mice only via intra-bone transplantation.
Limiting dilution transplantation assay for CD34-negative HSC activity was performed by several groups. Recently, Sonoda’s group has developed very sensitive transplantation assay  – intra-bone injection in NOG mice. Sensitivity of this assay and improved purification technique is based on the 18 Lineage antibody mixture, allowing the calculation of the frequency of CD34-negative HSC to be as high as 1 in 1000. Overall, frequency of HSC in CD34-negative is much inferior comparing with CD34+ counterpart. I summarized some experimental data below:
1 SRC in 1010 Lin-/CD34+ (Wang, 2003 )
1 SRC in 24100 Lin-/CD34- (13 antibodies used for Lineage detection, Wang, 2003 )
1 SRC in 1000 Lin-/CD34- (18 antibodies used for Lineage detection, Ishii, 2010 )
1 SCR in 40-44 Lin-/CD34+/CD38- (Ishii, 2010 ; Yahata et al, 2003 )
Functional similarities between CD34+ and CD34- cells
1. CD34-negative HSC can be mobilized by G-CSF in the same fashion as CD34+ cells.
2. Both populations spontaneously leave bone marrow niches and are able to migrate.
3. Comparable self-renewal potential.
4. Comparable long-term multilineage engraftment.
Functional distinction between CD34+ and CD34- populations is based on the following:
1. CD34-negative populations (CD38+ and CD38-) have very low hematopoietic activity in vitro (colony assay). Because CD34-negative cells didn’t give any colonies in standard colony assay, this population was ignored as potential source of HSC.
2. In sharp contrast to CD34+ cells, CD34-negative HSC easily undergo ex vivo expansion resulting in increase SCID-repopulating potential. As soon as Lin-/CD34- cells acquired CD34 expression in culture, CD34+ were not able to repopulate mice, unlike fresh CD34+ cells.
3. Trend toward lymphoid-biased engraftment (Bhatia M. et al, 1998) after intravenous administration, but not biased multilineage engraftemnt after intra-bone injection (Sonoda group).
4. Lin-/CD34- cells have an impaired ability to home to bone marrow after intravenous injection due low level of CXCR4 expression (Sonoda reviewed 2008 ). These cells also have very poor SDF-1-mediated migration ability in vitro.
Markers and hierarchy
One of the main problem for isolation of HSC from CD34-negative population is the absence of known positive markers. It was shown that the most primitive Lin-/CD34- HSC are also negative for: CD38, HLA-DR, Flt3 and c-Kit (Bhatia, 1998, Sonoda 2003-2008).
As candidates for further isolation of HSC within CD34-negative population, two markers expressed on Lin-/CD34- cells were studied: ALDH  and CD133 . Unfortunately xenogeneic transplantation assays were not performed in both studies.
It could be very interesting to look at lineage repopulation by CD34-/ALDH+ and CD34-/CD133+ in sensitive xenotransplantation assays.
All groups showed that CD34-negative HSC are precursors for CD34+. Therefore CD34-negative cells are the most primitive human HSC  in adult bone marrow and cord blood. Interestingly, there is a reversibility of CD34 expression on primitive HSC, at least for bone marrow .
Clinical relevance and perspectives
Very low clonogenic capacity in vitro and undiscovered markers which could allow positive selection are the main obstacles for clinical application and assessment of CD34-negative HSC. Some thoughts about CD34-negative HSC that could be significant for clinical use, I summarized as the following:
1. In order to assess clinical relevance of CD34-negative HSC, some positive markers should be discovered and new sensitive assays should be developed. For example, standard colony assay can estimate potency only CD34+ cells, but not CD34-negative.
2. In clinical HSC transplantation CD34-negative population can not be ignored, because it possesses high HSC activity and therapeutic potential.
3. Clinical significance of CD34-negative HSC supported by the fact that clinical outcome sometimes is not correlated with transplanted CD34+ cells quantity and quality.
4. In order to investigate therapeutic potential CD34-negative HSC, one should be brave enough to transplant one CD34-depleted cord blood unit with one fresh or ex vivo expanded and keep track of long-term engraftment. Another scenario – Lineage-negative cells  versus CD34+ positively selected, whole bone marrow versus CD34+ selected. The former approach is currently under clinical trials for different conditions.
5. CD34-negative population could be more interesting target for clinical-grade HSC expansion compared to CD34+ cells.
6. CD34-negative population could be a candidate for leukemic stem cells  and therefore could have prognostic and therapeutic value in leukemia clinic.
Connotea tag: CD34-negative