Foot ulcers and poor wound healing are common problems for patients with diabetes. Our studies have shown that increased matrix metalloproteinase-9 (MMP-9) levels can predict subsequent poor wound healing. Whether doxycycline, an agent known to decrease MMP-9 activity can improve wound healing and the mechanism of this effect has not been studied in detail. Therefore, we investigated the effect of doxycycline on wound healing rate in our excisional diabetic wound healing model. Results were compared with the effect of insulin treatment and additional end-points included markers of granulation tissue formation, MMP-9 and Toll-like receptors (TLR) -2 and -4 as indicators of bacterial contamination.

Diabetes was induced using streptozocin (i.p. 65mg/kg, n=24). Some animals were commenced on insulin (n=12, 10IU/day) and age matched non-diabetic animals (n=16) acted as control. After 6 weeks, full thickness wounds (8mm diameter) were created and diabetic and control rats were randomized to receive daily oral doxycycline by gavage (Dox: 100mg/kg). Six days later the animals were sacrificed and wounds were excised for mRNA analysis by RT-qPCR. Wound MMP-9 activity was measured by zymography and wound closure rate (WCR) was monitored by tracing.

Compared with controls, diabetes caused a 35% decrease in WCR and collagen III and TIMP-1 mRNA levels (each > 2 fold, P<0.05). Insulin but not doxycycline treatment partially normalized WCR. In contrast doxycycline prevented the decrease in collagen III and TIMP-1 mRNA but insulin treatment was without effect. Doxycycline had no effect on wound MMP-9 mRNA but decreased wound MMP-9 protein in control and diabetic animals (P<0.05 vs control). There was no difference in TLR- 2 and 4 expression across the groups. This data indicates for the first time that despite lack of effect on wound closure doxycycline treatment can improve granulation tissue quality, an effect which appears to be independent of its antibacterial action.