In vivo generation of beta-cell-like cells from CD34(+) cells differentiated from human embryonic stem cells

Abstract

Objective: CD34(+) cells, present within the bone marrow, have previously been shown to possess pancreatic endocrine potential. Based on this observation, we explored the capacity of CD34(+) cells derived in culture from the differentiation of human embryonic stem cells (hESC), for their in vivo pancreatic endocrine capacity.

Materials and methods: Sheep were transplanted with hESC-derived CD34(+) cells, as well as nonsorted differentiated cultures. Transplantations were carried out with in utero intraperitoneal injections prior to development of the immune system in the fetus so that tolerance toward foreign antigens was acquired during gestation and persisted in the adult.

Results: All cell populations that were tested demonstrated human cellular activity and long-term presence up to 5 years. However, the in vivo beta-cell-like activity achieved from the transplantation of the sorted CD34(+) cell population was not augmented by transplanting the entire cell population from which the CD34(+) cells were isolated. Human DNA and insulin messenger RNA were detected in sheep pancreases. An average of 1.51 ng/mL human C-peptide was detected in serum from eight animals transplanted with differentiated cell populations and assayed up to 55 months posttransplantation. Transplantation of as few as 23,500 cells resulted in long-term sustainable beta-cell-like activity. Teratomas were absent in the transplanted animals.

Conclusion: Our data suggest that hESC-derived CD34(+) cells have a potential for long-term in vivo endocrine cellular activity that could prove useful in regenerative medicine. Because the same cell population has previously been shown to contain hematopoietic potential, it could be used for the induction of immunological tolerance and bone marrow chimerism prior to cellular therapy for diabetes.

Figure 1

Characterization of donor cells by RT-PCR: Presence of genes pertinent to germ layers and differentiation, and commitment to the pancreatic lineage. hESC was differentiated by co-culture on S17 stromal layer for 17 days (S17/Day 17) or by EB formation for 8 days (EB/day 8). CD34⁺ cells were isolated from these: (S17/CD34⁺) and (EB/CD34⁺). RT-PCR was performed for genes indicated. Lane 1: water; 2: hESC; 3: S17/day 17, 4: S17/CD34⁺, 5: EB/day 8, 6: EB/CD34⁺; 7: human fetal pancreas, 8: water, 9: CD34⁺ from adult BM (BM/CD34⁺), and 10: human fetal pancreas (22–24 wk).

Figure 2. Detection of human DNA in sheep pancreas: Engraftment and presence of human cells

(A) Sheep were transplanted in-utero with hESC-derived cells as noted in Tables 1–3. Pancreatic tissue was harvested from each animal (identified by animal no. on top right) at various periods after transplant (months, indicated in parentheses, bottom right). Four samples from each animal were amplified. Primers for human-specific GAPDH gene (A, top row) were for the detection of human DNA. Primers for beta-actin (B, bottom row) indicated the presence of amplifiable DNA from both human and sheep. (B) Human GAPDH primers do not amplify sheep DNA. Seven control sheep were amplified with GAPDH (top row) and beta-actin (bottom row), lanes 2–8. Lane 1: water. Lane 10: Human positive control. (C) Detection limit for human DNA present in sheep. Human DNA was spiked into sheep DNA at levels indicated to determine the detection limit of this assay. Ten ng of total DNA was used per reaction. Top row: GAPDH gene. Bottom row: Beta-actin.

Figure 2. Detection of human DNA in sheep pancreas: Engraftment and presence of human cells

(A) Sheep were transplanted in-utero with hESC-derived cells as noted in Tables 1–3. Pancreatic tissue was harvested from each animal (identified by animal no. on top right) at various periods after transplant (months, indicated in parentheses, bottom right). Four samples from each animal were amplified. Primers for human-specific GAPDH gene (A, top row) were for the detection of human DNA. Primers for beta-actin (B, bottom row) indicated the presence of amplifiable DNA from both human and sheep. (B) Human GAPDH primers do not amplify sheep DNA. Seven control sheep were amplified with GAPDH (top row) and beta-actin (bottom row), lanes 2–8. Lane 1: water. Lane 10: Human positive control. (C) Detection limit for human DNA present in sheep. Human DNA was spiked into sheep DNA at levels indicated to determine the detection limit of this assay. Ten ng of total DNA was used per reaction. Top row: GAPDH gene. Bottom row: Beta-actin.

Figure 2. Detection of human DNA in sheep pancreas: Engraftment and presence of human cells

(A) Sheep were transplanted in-utero with hESC-derived cells as noted in Tables 1–3. Pancreatic tissue was harvested from each animal (identified by animal no. on top right) at various periods after transplant (months, indicated in parentheses, bottom right). Four samples from each animal were amplified. Primers for human-specific GAPDH gene (A, top row) were for the detection of human DNA. Primers for beta-actin (B, bottom row) indicated the presence of amplifiable DNA from both human and sheep. (B) Human GAPDH primers do not amplify sheep DNA. Seven control sheep were amplified with GAPDH (top row) and beta-actin (bottom row), lanes 2–8. Lane 1: water. Lane 10: Human positive control. (C) Detection limit for human DNA present in sheep. Human DNA was spiked into sheep DNA at levels indicated to determine the detection limit of this assay. Ten ng of total DNA was used per reaction. Top row: GAPDH gene. Bottom row: Beta-actin.

Figure 3

IHC of pancreatic tissue from chimeric sheep. Sheep were transplanted with hESC-derived cells as indicated in Tables 1–2. Chimeric animals were euthanized at periods after transplantation indicated in Fig. 2A. Pancreatic tissue was processed for IHC as described in methods. (A)Two representative animals were analyzed with antibodies that specifically recognized human proteins. The proteins were: chromogranin A, proinsulin, and C-peptide. A control sheep (not transplanted) stained negative for all these antibodies. (Original magnification × 400). (B) Image enlarged for detail.

Figure 3

IHC of pancreatic tissue from chimeric sheep. Sheep were transplanted with hESC-derived cells as indicated in Tables 1–2. Chimeric animals were euthanized at periods after transplantation indicated in Fig. 2A. Pancreatic tissue was processed for IHC as described in methods. (A)Two representative animals were analyzed with antibodies that specifically recognized human proteins. The proteins were: chromogranin A, proinsulin, and C-peptide. A control sheep (not transplanted) stained negative for all these antibodies. (Original magnification × 400). (B) Image enlarged for detail.

Figure 4

H&E staining of human and sheep pancreas for comparison of morphology. Human (top) 200× magnification (A) and 400× (B). Sheep (bottom) 200× magnification (C) and 400× (D). Islets stained lighter and revealed a denser cluster of nuclei as also evident in the IHC stains.