Injectable antibacterial conductive nanocomposite cryogels with rapid shape recovery for noncompressible hemorrhage and wound healing

Developing injectable antibacterial and conductive shape memory hemostatic with high blood absorption and fast recovery for irregularly shaped and noncompressible hemorrhage remains a challenge. Here we report injectable antibacterial conductive cryogels based on carbon nanotube (CNT) and glycidyl methacrylate functionalized quaternized chitosan for lethal noncompressible hemorrhage hemostasis and wound healing. These cryogels present robust mechanical strength, rapid blood-triggered shape recovery and absorption speed, and high blood uptake capacity. Moreover, cryogels show better blood-clotting ability, higher blood cell and platelet adhesion and activation than gelatin sponge and gauze. Cryogel with 4 mg/mL CNT (QCSG/CNT4) shows better hemostatic capability than gauze and gelatin hemostatic sponge in mouse-liver injury model and mouse-tail amputation model, and better wound healing performance than Tegaderm™ film. Importantly, QCSG/CNT4 presents excellent hemostatic performance in rabbit liver defect lethal noncompressible hemorrhage model and even better hemostatic ability than Combat Gauze in standardized circular liver bleeding model.

. However, all of the cryogels still kept good shape and elasticity, suggesting their great robustness.
Supplementary Figure 6. Rheological properties of QCSG/CNT0 and QCSG/CNT4. 6 Both QCSG/CNT0 and QCSG/CNT4 showed stable or slightly increased G' and G'' when gradually increasing the frequency from 0.1 rad/s to 100 rad/s and fixing the strain at 1%, suggesting their stable cryogel networks. During the strain amplitude sweep test, both the cryogels showed gradually decreased G' and increased G'' when changing the strain from 0.01% to 100%, probably attributed to the rapid reversible shape deformation of the cryogel networks. QCSG/CNT4 presented higher storage moduli than those of QCSG/CNT0 due to the reinforcement of CNT. Supplementary Figure 13. The X-ray detectability was introduced to the cryogel by 11 gluing a X-ray detectable line to the cryogel, and the X-ray detectable line contained

Supplementary Tables
Supplementary When delivering the cryogel to bleeding site, the cryogel rapidly absorbed blood and accelerate blood coagulation during hemostatic application, and the coagulated blood further strengthened the cryogel network and crosslinked with CNT to prevent the CNT release from the cryogel.

Supplementary Note 2. Cryogels remaining safe compression moduli after application
During the hemostatic application, the shape-fixed cryogels can absorb blood and then recover their shapes. The cryogel will remain a certain compression strain due to the 16 limited space from surrounding tissues. However, with the increase of CNT from 0 to 6 mg/mL, the cryogels just presented about 0.49 kPa to 1.19 kPa compression moduli when remaining 20% strain. These compression moduli are lower than those of human soft tissues (between several kPa to several MPa) 11,12,13,14 . Therefore, the cryogel hemostatic agents will not cause severe pressure and additional injury to soft tissue during the application.

Supplementary Note 3. Mechanism of the high resilience and rapid recovery behavior of the cryogels
In the first stage of compression, the macroporous sponge-like cryogel matrix could allow water freely flow in and out, and just experienced collapsed macropores and deformed matrix under the compressive stress. The water would be squeezed out from the cryogel and then effectively relieve the internal compressive stress of cryogel, which just led to little strain deformation. Once the stress was removed, the cryogel would instantly recover and absorb the water back simultaneously, which was derived by the bended cryogel network. When further increasing the compression stress to a high level, the cryogel matrix would relieve most of the compression stress by squeezing out of the water, and then the network of CNT hard microdomain physical crosslinker would serve as a highly resilient substrate to prevent the mechanical failure of the cryogel 2, 15 . 17

Supplementary Note 4. Photothermal antibacterial activity of the cryogels
As shown in Figure 3e- Figure 3b), and the obviously enhanced S. aureus killing ratio after NIR irradiation might be due to higher sensitivity of S. aureus to NIR irradiation than those of E. coli and P. aeruginosa.

Supplementary Note 5. Cytocompatibility evaluation of the cryogels
After contacting with the cryogel surface for 24 h, all the cells in the three CNTcontained cryogel groups showed more than 90% cell viability compared with the TCP control group (P>0.05) (Figure 5d). However, QCSG/CNT0 group showed the lowest cell viability of 75%, which was significantly lower than CNT-contained cryogel groups and TCP group (P<0.05). Consistent with the results in Figure 5d, most of cells in all 18 the four cryogel groups were green and showed spindle-like shape similar to that of TCP control group. Few dead cells in all the test groups were caused by cell metabolism and apoptosis. However, QCSG/CNT0 group showed obviously reduced cell number compared to CNT-contained groups and TCP group, suggesting slight inhibition of cell proliferation of QCSG/CNT0 due to its strong positive-charged nature 1 . Interestingly, the incorporation of appropriate content of CNT into cryogel network increased cytocompatibility of QCSG by the cation-π interaction between QCSG and CNT 16 .

Supplementary Note 6. In vitro blood clotting performance of the gauze and gelatin hemostatic sponge
A few blood cells were observed in gauze group while slightly enhanced number of blood cells was presented in gelatin hemostatic sponge group, and the blood cells in both the two groups kept their distinctive biconcave disks. Besides, gauze group almost showed no platelet adhesion, while gelatin hemostatic sponge presented few platelets.
Although gelatin hemostatic sponge group exhibited enhanced blood cell adhesion and platelet activation, but it showed less efficient blood clotting than gauze group ( Figure   6a), which is because gauze could rapidly absorb blood to concentrate the blood to induce blood clotting, while the gelatin hemostatic sponge couldn't rapidly absorb blood into its network and just provide a hemostatic surface to induce blood clotting.

Supplementary Note 7. In vivo hemostasis for lethal noncompressible hemorrhage
QCSG/CNT4 possessing excellent in vivo hemostatic capability compared to 19 QCSG/CNT0 was chosen to perform the lethal noncompressible hemorrhage hemostasis. QCSG/CNT0, gelatin hemostatic sponge D1 (with a diameter of 4 mm) and gelatin hemostatic sponge D2 (with a diameter of 6 mm) were used as control groups.
As shown in Figure 7g-i, the rabbits in blank group presented the most blood loss of 24.5 g and the longest blood bleeding time of 25.0 min, and all the rabbits were dead within 25 min. These results demonstrated that the liver volume defect model is a lethal noncompressible hemorrhage model, in which the animal loses a large amount blood within short time and the rabbit cannot stop the bleeding by its own hemostatic capacity.

Supplementary Note 8. In vivo hemostasis of gelatin hemostatic sponges for lethal noncompressible hemorrhage
Although the gelatin hemostatic sponge D1 possessed the same diameter with the shape fixed cryogel groups, it had no shape recovery property after contacting with blood.
Therefore, gelatin hemostatic sponge D1 with permanent diameter slightly smaller than wound could just partly block the liver defect hole, causing significantly reduced blood loss than blank group (P<0.001) (Figure 7g). Although, gelatin hemostatic sponge D2 with a diameter bigger than wound presented better hemostatic effect than gelatin hemostatic sponge D1 due to its capacity to completely block wound except for its inherent hemostatic ability, it presented weak mechanical strength (especially in the state after absorbing blood) limiting its hemostatic efficiency on lethal noncompressible hemorrhage application. Furthermore, the use of gelatin sponge was not convenient when the bleeding was noncompressible and the wound was narrow, deep and irregular. 20

Supplementary Note 9. Discussions on the safety of cryogels for hemostasis
For clinic application, to stop the deep and narrow wound bleeding for human body, the cryogels with greater height (such as more than 4 cm) can be prepared to ensure that part of the cryogel is exposed outside after injection, and the surgeon can inject columnar cryogel hemostatic agent with proper size into the bleeding site. In addition, for massive hemorrhage on the body surface (such as leg amputation), the surgeon can use disk shape cryogel with proper surface area according to the size of the specific wound, thus the material will not enter the body for its much bigger diameter than that of the blood vessel (The blood vessel diameters of the legs are no more than 14 mm) 17 .

Supplementary Note 10. In vivo wound healing performance of the cryogels
Although there is some limitation about using mouse skin as wound healing model 18 , mice were chosen for wound-healing study for several reasons: availability, low cost, the ability to test large numbers of animals with reproducible results, and the potential to test a variety of genetic knockout animals 19 . Both the wound contraction and histomorphological evaluation results on 5th, 10th and 15th day were used to evaluate the healing effect of the cryogels with commercial film dressing (Tegaderm™) as a control group.