Diabetes mouse models used as models for wound healing

To many people, diabetes connotes problems with the pancreas, insulin and sugar balance. Few associate the disease with difficulty healing wounds. Yet, chronic wounds and tissue death are life-threatening problems faced by many people with diabetes. Because the wound-healing phenotypes of diabetes mouse models have been poorly characterized, choosing appropriate models for wound-healing research has been difficult. To address this problem, a research group led by Thomas Mustoe, M.D., from the Laboratory for Wound Repair and Regenerative Medicine, Northwestern University, analyzed and compared the wound-healing phenotypes of four diabetes mouse models (Fang et al. 2010).

The pathophysiology of diabetes-related wound-healing problems is complex and poorly understood. Most of the research has been conducted using type 1 or monogenic type 2 diabetes models, like the db/db mouse strain BKS.Cg-Dock7^m +/+ Lepr^db/J . However, type 1 diabetes accounts for only 5-10% of all diabetes cases in the U.S., and the extreme obesity, insulin resistance and hyperglycemia, along with the defective leptin/leptin receptor axis of the db/db model, are not representative of the most common diabetes phenotypes in the general population. Emerging polygenic models, such as NONcNZO10/LtJ, exhibit more moderate diabetes phenotypes and have an intact leptin/leptin receptor axis. Thus, the Mustoe team assessed and compared the wound healing phenotypes of the NONcNZO mouse and three other commonly used diabetes mouse models:

• 

  The "Akita" mouse, strain C57BL/6-Ins2^Akita/J – a model of type 1 diabetes
• The streptozotocin (STZ)-induced C57BL/6J mouse – a model of type 1 diabetes
• The db/db mouse – a monogenic model of type 2 diabetes

To account for the different responses to various wound types among mouse models of diabetes, Mustoe and his team assessed how these models respond to three major wound types.

1. Cuts. An "incision model" was used to assess collagen deposition in, and the breaking strength of, a healed wound.
2. Open wounds. A "splinted excisional model" was used to assess re-epithelialization and granulation tissue deposition in an open wound.
3. Ischemia/reperfusion (I/R) injury. A "compression model" was used to assess the repetitive conditions that lead to I/R skin injury.

To avoid the confounding influence of aging, Mustoe and his team conducted all experiments using young adult mice. To ensure that the wound-healing phenotypes of each model were assessed at comparable stages of diabetes progression, each model was kept in a chronic diabetic state for at least eight weeks before experiments began. The research team's results are shown in the table below (adapted from Fang et al. 2010, Table 2).

Wound Type	Impaired Wound Healing in Four Diabetes Models (compared to controls*)
	Akita	STZ-induced	db/db	NONcNZO/LtJ
Cuts	No	Yes	Yes	Yes
Open Wounds	No	No	Yes	Yes
I/R Injury	No	No	No	Yes

*Control for the Akita mouse: B6J
Control for the STZ-induced mouse: B6J
Control for the db/db mouse: C57BLKS/J
Control for the NONcNZO10/LtJ mouse: NON/ShiLtJ

All three wound types heal normally in the Akita strain. Whereas only cuts heal poorly in the STZ-induced diabetes model, and only cuts and open wounds heal poorly in the db/db model, all three types of wounds heal poorly in the NONcNZO10/LtJ model. Conclusions about these results must be drawn cautiously. Just as a single model does not sufficiently replicate diabetes, a single model does not replicate diabetes-associated wound-healing impairment. However, the monogenic nature and extreme phenotypes of the db/db strain may reduce its effectiveness as a wound-healing model. In contrast, the polygenic nature and multiple wound healing deficits of the NONcNZO10/LtJ strain might make it a more clinically relevant model.

Reference

Fang RC, Kryger ZB, Buck II DW, De La Garza M, Galiano RD, Mustoe TA. 2010. Limitations of the db/db mouse in translational wound healing research: Is the NONcNZO10 polygenic mouse model
superior? Wound Repair Regen Oct 18. doi: 10.1111/j.1524-475X.2010.00634.