(Purpose) We examine the capacity of human cord blood (CB)-derived cells to generate insulin-producing cells and other lineages of cells in pancreatic tissue in vivo. (Method) Ten millions of human CB-derived T-cell-depleted mononuclear cells were intravenously transplanted into conditioned newborn NOD/SCID/b2-microglobulinnull mice with or without chemical injury by an intraperitoneal injection of streptozotocin (STZ) at dose of 100 mg/g body weight. At 1–3 months post-transplantation, pancreatic tissues of the recipient mice were analyzed for the presence of human CB-derived cells by performing immunofluorescence study (insulin, amylase, or CD45) and FISH analysis for human chromosomes on the same specimens. RNA was isolated from pancreatic tissues of recipient mice, and RT-PCR analysis using human insulin specific primer was performed to examine human insulin at RNA level. Finally, double FISH analysis for human- and murine chromosomes was performed to get an insight into the mechanism for the generation of human CB-derived insulin-producing cells in vivo.

(Results) At 1–3 months post-transplantation, human CB-derived T-cell-depleted mononuclear cells gave rise to both myeloid and lymphoid progeny (CD33+, CD19+, and CD3+ cells) in bone marrow and peripheral blood of the recipient mice. In recipient pancreatic tissues, human CB-derived cells were identified inside and outside islets. Outside pancreatic islets, the vast majority of human chromosome+ cells were CD45+ hematopoietic cells, while human chromosome+ amylase+ acinar cells were also identified. Inside islets, human chromosome+ cells accounted for 1.01 +/− 0.73 % (n=6) without STZ treatment. Among them, human CB-derived insulin-producing cells were identified at a frequency of 0.65 +/− 0.64 % (n=6) of total insulin+ cells in xenogeneic hosts. RT-PCR analysis demonstrated the presence of human insulin, whose sequence was fully identical to that of already-known human insulin cDNA. Chemical injury with STZ treatment led to the significant destruction of islet tissue and reduction of cell numbers in islets. In STZ-treated recipient mice, however, human insulin-producing cells were identified at a frequency of 0.23 +/− 0.27 % (n=4) in islets, which was lower than the mice without STZ treatment. Finally, double FISH analyses using species-specific probes demonstrated the presence of human chromosome+ murine chromosome+ insulin-producing cells and human chromosome+ murine chromosome- insulin-producing cells in recipient islets.

(Conclusion) It is concluded that human CB cells contain the progenitor cells to generate the insulin-producing cells in vivo. The mechanism of CB-derived insulin-producing cells includes both fusion-dependent and independent mechanisms. Although the capacity of CB-derived cells needs to be compared with other stem cell sources such as tissue stem cells or embryonic stem cells, the present study suggests the possibility of CB cells as new source for future regenerative medicine for diabetes mellitus.

Author notes

Corresponding author

Sign in via your Institution