Background. A series of recent studies implies that circulating bone marrow (BM)-derived cells may contribute to both endo- and exogenous regeneration of non-hematopoietic tissue, at least partly through transdifferentiation into differentiated non-hematopoietic cells. We and others have observed donor-derived pancreatic islet cells after BM-transplantation (BMT) to otherwise healthy mice. The above results suggest that BM cells may be useful for the therapeutic regeneration of insulin-producing pancreatic beta cells in diabetes mellitus (DM). The regenerative properties of BM-derived cells appear to be enhanced in models that include pancreatic islet injury. Therefore, we are exploring the capability of the BM-mobilizing cytokine granulocyte colony stimulating factor (G-CSF) to influence regeneration in streptocotocin (STZ)-induced DM.

Methods. Female C57BL/6 mice, 8–10 weeks old, underwent lethal irradiation (11 Gy) followed one day later by BMT (tail vein-injection of 5 million male unfractionated, nucleated BM cells, transgenic for the enhanced green fluorescent protein [EGFP+]). Two months after BMT, the mice were allocated to four experimental protocols (n=10 per group). All experimental animals were treated with STZ (200 mg/kg i.p.) on day 5, to induce DM. Group A received human G-CSF (Amgen, CA) pre- and post-STZ administration (200 μg/kg, i.p., in four daily doses; days 1-4 for pre- and days 22-25 for posttreatment). Group B received human G-CSF only before STZ administration, while Group C received G-CSF only after STZ administration. Group D served as diabetic controls. Group E (n=6) underwent BMT but received no G-CSF or STZ treatment. Blood glucose (Glc) was measured using test strips and a blood glucose meter (Accu-ChekR, Roche, IN). Individual mice from each group were sacrificed on day 31, and organs removed for histological studies.

Results. In all experimental groups, 60% of the mice became manifestly diabetic (Glc > 350 mg/dL). The majority of the other mice remained clearly non-diabetic (Glc < 200 mg/dL), i.e. non-responsive to STZ. One mouse in Group B, however, became subdiabetic later in the experiment. The Table shows the development of blood glucose in the STZ-responsive mice (mean blood glucose; n=6 per group (days −2–29), n=5 per group (days 35 and 43)).

Failure of G-CSF mobilization to influence STZ-induced hyperglycemia

Conclusion. G-CSF administration failed to either prevent or treat the STZ-induced hyperglycemia in this experimental model of DM, implying that the mobilization of BM cells did not promote physiologically significant regeneration of pancreatic islets within the time-frame studied. This failure may imply that BM-based regeneration therapy for DM has serious limitations, or, alternatively, that more favorable timing and dosing of G-CSF may be needed to improve the results. Ongoing work, including the histological examination of these mice, will reveal whether partial, but clinically insufficient, regeneration was indeed induced, which will guide future studies.

Day−251218293543
*mean glucose level (mg/dL); ND = Not done 
Group A 143* 126 380 498 527 393 396 
Group B 156 122 331 468 514 454 373 
Group C 147 152 330 389 480 447 427 
Group D 139 140 327 392 464 406 459 
Group E 154 144 123 138 125 131 ND 
Day−251218293543
*mean glucose level (mg/dL); ND = Not done 
Group A 143* 126 380 498 527 393 396 
Group B 156 122 331 468 514 454 373 
Group C 147 152 330 389 480 447 427 
Group D 139 140 327 392 464 406 459 
Group E 154 144 123 138 125 131 ND 

Author notes

Corresponding author

Sign in via your Institution