Fig 2: Impact of hypoxia and SOD3 on perivascular hyaluronan and arterial muscularization. Relative changes in HA in the vascular adventitial vs. parenchymal compartments of the lung between normoxia and 14 days of hypoxia. Quantitation of HA in HABP-stained lung sections was achieved by DAB/methyl green color vector deconvolution in ImageJ followed by signal-to-perimeter measurement of vessel edges (5–10 vessels per HPF; 5 HPF per lung) normalized to alveolar background generated by cropping away all bronchovascular structures. (A) Representative images of vascular HA staining after image processing. (B) Mean alveolar background HA staining intensity. (C) HA intensity expressed as vascular HA (<50 or 50–100 micron diameter) relative to the alveolar HA content in normoxic and hypoxic WT and SOD3KO (KO) mice. (D) Representative histo-immunologic co-staining of HABP and aSMA in lung slices. Vessels are demarcated by black arrows. (E) Correlation between HA-coated and de novo muscularized small vessels (=50 microns – left panel; 50–100 microns – right panel) during hypoxic PH. Color vectors were deconvolved in ImageJ (DAB/methyl green) and independently counted in a blinded fashion. Legend: WT normoxic (open black circles), 14d hypoxic (light gray closed circles), and 35d hypoxic (darker grey closed circles); SOD3KO (open red circles), 14d hypoxic (light pink closed circles) and 35d (darker pink closed circles). 5–10 randomly selected HPF were analyzed per animal. Data represented as mean ± SEM *p < 0.05 **p < 0.01 by 2-way ANOVA.

Fig 3: TG mice exhibited enhanced recovery of blood flow and function after limb ischaemia. Left hind limb femoral artery was transected to induce limb ischaemia in C57BL/6 and TG mice. Mouse hind limb blood flow was measured with LDPI before, immediately after ligation and at day 3, 7, 14 and 21 (A). The blood flow recovery was significantly increased in TG mice compared to the WT mice (B). After surgically induced hind limb ischaemia, C57BL/6 and TG mice were placed on treadmill to test the limb function at day 7, 14 and 21. The TG mice were able to run for a significantly longer time than WT mice (C). WT: C57BL/6 mice; TG: triple-transgenic mouse with overexpression of antioxidant network composed of superoxide dismutase (SOD)1, SOD3 and glutathione peroxidase (Gpx-1). Ischemic limb/normal limb: ratio of mouse ischemic left hind limb blood flow to normal right hind limb blood flow. BL: before ligation. PL: post ligation. # WT vs TG, P < 0.001, n = 8; *Day 14 or 21 vs day 7, P < 0.01, n = 8; $ Day 21 vs 14, P < 0.01, n = 8; #WT vs TG, P < 0.01, n = 8

Fig 4: Concomitant overexpression of triple antioxidant enzymes effectively preserved the muscle fibres of the ischemic limb. The muscle fibres reduced in size and became largely disconnected in the WT control mice with limb ischaemia. CD31 staining showed that capillary density was low in the limb muscle after ischaemia for 3, 7 and 21 days. On the other hand, the muscle fibres appeared to remain long and large, as well as well-connected in the limb muscles at day 21 after ischaemia in the TG mice with concomitant global overexpression of SOD1, SOD3 and Gpx-1. CD31 staining demonstrated that the capillary density remained largely intact (similar to the non-ischemic limb). Normal: normal right limb; ischaemia; ischemic left limb. Red arrow: capillary vessels. BL: before ligation. Scale bar: 50µm

Fig 5: Impact of hypoxia and SOD3 on ex vivo enzymatic cleavage of hyaluronan. (A) Partial degradation of a standardized quantity of monodisperse HMWHA (1000 kD) by whole lung lysates from normoxic and hypoxic (14 day) WT and SOD3KO mice. Densitometry data are expressed as fold-change in center of mass relative to undigested PBS control. Representative lanes were cropped from the same gel (see Supporting Information 6B). (B) Effect of enzyme inactivation on HA degradation. Lysates were pretreated with the following conditions: U – untreated lysated; HK – heat killed; PK – proteinase K digested; Hy – addition of Streptomyces hyaluroniticus hyaluronidase. Densitometry data are expressed as percentage digestion (i.e. 1 - %HA remaining). (C) Total ROS/RNS normalized to total protein and (D) the specific superoxide percentage measured in frozen-and-thawed whole lung protein lysates. Left inset: representative EPR spectrum showing the total CMH-nitroxide signature and SOD1-inhibited signal respectively. Data represented as mean ± SEM *p < 0.05 **p < 0.01 by paired t-test (for inactivation experiments), or 2-way ANOVA.