br Conflict of interest statement The following are the

Conflict of interest statement The following are the supplementary data related to this article.
Acknowledgments This research was supported by funds to R.C. and F.C. from the Italian Ministry of Education, University and Research (MIUR).
Introduction Besides hematopoietic stem tamoxifen citrate (HSCs), multipotent stromal cells are claimed to be a promising source for clinical applications, but their typical differentiation capacities depending on their origin (cord blood, CB; bone marrow, BM; adipose tissue, AT) might have an impact on their usefulness in specific clinical applications. Therefore, a detailed characterization of their developmental origin and their lineage-specific differentiation potential is mandatory to elaborate, if a specific cell source might be favourable for bone and cartilage formation. Questionable as well is, if mixed bulk cultures should be applied in clinical applications, as no detailed knowledge is available on the impact of mixed subpopulations of stromal cells. In cord blood, two distinct clonal neonatal subpopulations are described: USSCs (unrestricted somatic stromal cells) and CDSCs (cord blood derived stromal cells) (Kluth et al., 2010; Kogler et al., 2004; Liedtke et al., 2010). Regarding their immunophenotype both clonal neonatal cell types share the same pattern of surface molecules (CD45−, CD13+, CD29+, CD73+, CD105+) similar to bone marrow-derived stromal cells (Erices et al., 2002) and are commonly referred to as “MSCs” phenotype (Dominici et al., 2006). However, the markers used to describe stromal cells are not specific and are expressed by many connective tissue cells that are not stem cells. To date, a marker set clearly distinguishing connective tissue stem cells from more mature cells is not available. As the term “MSCs”, which can stand for mesenchymal stromal cells as well as multipotent stromal cells, is controversially discussed (Bianco, 2011a), we will refer to cells derived out of cord blood (CDSCs and USSCs) and out of bone marrow (BMSCs) as stromal cells in this paper. A great advantage of these CDSCs and USSCs is their simple isolation and expansion in vitro. Likewise, USSCs produce functionally significant amounts of hematopoiesis-supporting cytokines and are superior to BMSCs in expansion of CD34+ cells from cord blood (Kogler et al., 2005). USSCs are therefore a suitable candidate for stroma-driven ex vivo expansion of hematopoietic cord blood cells for short-term reconstitution or co-transplantation (Jeltsch et al., 2011). In the near future, these cells may be applied to patients to reduce the graft-versus-host disease, the most occurring side effect after transplantation of hematopoietic stem cells or to support hematopoiesis (Abdallah and Kassem, 2009). USSCs and CDSCs share the osteogenic and chondrogenic differentiation potential. In a recent study our group analysed in detail the expression of osteogenic and chondrogenic marker genes during differentiation defining the osteogenic signature of USSCs, CDSCs, BMSCs and (umbilical cord) UCSCs (Bosch et al., 2012). Based on the work of Kluth et al. (2010), it was demonstrated that USSCs in contrast to CDSCs do not differentiate naturally towards the adipogenic lineage, while expressing the adipogenic inhibitor Delta-like 1 homolog (DLK1) on a transcript but not on a secreted protein level. In addition, expression of HOX genes is absent in USSCs, whereas CDSCs revealed a typical positive HOX code similar to BMSCs (Liedtke et al., 2010). HOX genes are essential for normal development of vertebrates by determining the positional identity along the anterior-posterior body axis (Krumlauf, 1994). In humans, the 39 known HOX genes are distributed among four paralogous clusters HOXA to HOXD, located in chromosomes 7, 17, 12 and 2, respectively. HOX genes are expressed sequentially 3´ to 5´ along the body axis during embryogenesis, termed “temporal and spatial colinearity” (Kmita and Duboule, 2003). The typical HOX code of a cell describes the specific expression of functional active HOX genes in distinct tissues (Gruss and Kessel, 1991). More importantly HOX genes may also have a therapeutic application in near future. It was found that HOXD3 protein is upregulated during normal wound repair (Hansen et al., 2003). The protein promotes angiogenesis and collagen synthesis, but is absent in poorly healing wounds of genetically diabetic mice. After adding HOXD3, the treatment resulted in faster diabetic wound closure and tissue remodeling.