doi.org/10.26434/chemrxiv.13216196.v1

Redesigned Silk: A New Macroporous Biomaterial Platform forAntimicrobial Dermal Patches with Unique Exudate Wicking AbilityKankan Qin, Rui F.P. Pereira, Thibaud Coradin, Veronica de Zea Bermudez, Francisco Fernandes

Submitted date: 10/11/2020 • Posted date: 12/11/2020Licence: CC BY-NC-ND 4.0Citation information: Qin, Kankan; Pereira, Rui F.P.; Coradin, Thibaud; de Zea Bermudez, Veronica;Fernandes, Francisco (2020): Redesigned Silk: A New Macroporous Biomaterial Platform for AntimicrobialDermal Patches with Unique Exudate Wicking Ability. ChemRxiv. Preprint.https://doi.org/10.26434/chemrxiv.13216196.v1

Silk is one of the most important materials in the history of medical practice. Owing to its excellent strength,biocompatibility and degradability, silk from Bombyx mori – which is structured as a concentric assembly ofsilk fibroin (SF) coated by a sheath of sericin (SS) – has long been used for wound treatment. Here, werecapitulate for the first time the topology of native silk fibers using a radically new materials design-orientedapproach to achieve unprecedented porous dermal patches suitable for controlled drug delivery. The methodimplies four steps: (1) removing SS; (2) creating anisotropic macroporosity in SF via ice templating; (3)stabilizing the SF foam with a methanolic solution of Rifamycin (Rif) antibiotic; and (4) coating Rif-loadedredesigned SF foams with a SS sheath. The core-shell SS@SF foams exhibit water wicking propertiesaccommodate up to ~20% lateral deformation. Moreover, monitoring of antibacterial activity againstStaphylococcus aureus revealed that the SS@SF foams’ Rif release extended up to 9 days. We anticipatethat reverse-engineering of silk foams opens exciting new avenues towards the fabrication of advanced drugeluting silk-based biomaterial platforms with improved performance. The present approach can begeneralizable to re-build multicomponent biological materials with tunable porosity.

File list (2)

download fileview on ChemRxivQin et al 2020 preprint.pdf (1.50 MiB)

download fileview on ChemRxivSupporting information preprint.pdf (629.31 KiB)

Redesigned silk: a new macroporous biomaterial platform for antimicrobial

dermal patches with unique exudate wicking ability

Kankan Qina, Rui F.P. Pereirab, Thibaud Coradina, Verónica de Zea Bermudezc* and Francisco M. Fernandesa*

a Sorbonne Université, UMR 7574, Laboratoire de Chimie de la Matière Condensée de Paris, F-75005, Paris, France

b Chemistry Center and Chemistry Department, University of Minho, Campus de Gualtar, 4710-057, Braga, Portugal

c Chemistry Department and CQ-VR, University of Trás-os-Montes e Alto Douro, Apartado 1013, 5001-801, Vila Real, Portugal

Corresponding authors: francisco.fernandes@upmc.fr, vbermude@utad.pt

Abstract

Silk is one of the most important materials in the history of medical practice. Owing to its excellent

strength, biocompatibility and degradability, silk from Bombyx mori – which is structured as a

concentric assembly of silk fibroin (SF) coated by a sheath of sericin (SS) – has long been used

for wound treatment. Here, we recapitulate for the first time the topology of native silk fibers using

a radically new materials design-oriented approach to achieve unprecedented porous dermal

patches suitable for controlled drug delivery. The method implies four steps: (1) removing SS; (2)

creating anisotropic macroporosity in SF via ice templating; (3) stabilizing the SF foam with a

methanolic solution of Rifamycin (Rif) antibiotic; and (4) coating Rif-loaded redesigned SF foams

with a SS sheath. The core-shell SS@SF foams exhibit water wicking properties accommodate up

to ~20% lateral deformation. Moreover, monitoring of antibacterial activity against

Staphylococcus aureus revealed that the SS@SF foams’ Rif release extended up to 9 days. We

anticipate that reverse-engineering of silk foams opens exciting new avenues towards the

fabrication of advanced drug eluting silk-based biomaterial platforms with improved performance.

The present approach can be generalizable to re-build multicomponent biological materials with

tunable porosity.

Keywords

Silk, Ice templating, Antibacterial activity, Drug delivery, Wound healing

2

1. Introduction

In Nature, judicious texture patterning (pore size distribution, pore geometry, and connectivity) determines

materials’ function. Porosity, and in particular macroporosity,modulates the materials’ interactions with cells,

determines the transport mechanisms, and regulates the overall mechanics. Among the numerous examples of

macroporous materials found in Nature, trabecular bone – the porous fraction of bone composed of an

interconnected network of trabeculae – stands as a paradigm of the tremendous benefits arising from a finely

controlled macroporous structure. Owing to the inherent macroporosity, trabecular bone favors nutrient and gas

diffusion as well as exposure to growth factors, resulting in bone tissue that displays higher metabolic activity and

undergoes remodeling more frequently than the dense cortical bone[1]. Devising new ways to shape biomaterials

into controlled macroporous constructs, whose design principles rely often in reproducing the structural building

block and composition found in natural materials, holds promise in dictating their functionalities in applications

as diverse as drug delivery carriers, biomimetic membranes, and implants[2]. In this context, ice-templating (also

known as freeze-casting) has evolved from a ceramics-oriented process devoted to the fabrication of lightweight

materials to one of the most promising routes to fabricate aligned macroporous biomaterials from biopolymers[3–

6]. The technique is based on a uniaxial thermal gradient that promotes the segregation of solutes and/or suspended

particles induced by a growing ice front. Since the solubility of most molecular species in ice is extremely limited,

the freezing step forces solutes to progressively accumulate in the interstitial space formed by the ice crystals.

Upon sublimation these materials reveal the macroporous structure defined by ice. Ice-templating has been

previously widely explored for the fabrication of biocompatible three-dimensional (3D) aligned scaffolds for tissue

engineering applications[3,4,7–9] in particular because of the scaffolds’ ability to promote cell contact guiding.

However, beyond cell guiding, the aligned macroporous channels generated by ice-templating also configure an

effective capillary transport system[10] that could be beneficial for the clearance of exudates from wounds, such as

proteases generated during the inflammation stage.

Choosing silk as the base component to get the proof-of-concept of such properties was obvious. Silk,

produced by a variety of animals including spiders and domestic Bombyx mori silkworm, is one of the most

3

sophisticated and unique soft materials found in Nature. Silk is characterized by a precise coaxial structuration of

two proteins, a silk fibroin (SF) core and a silk sericin (SS) sheath that envelopes the former. Similarly to bone,

skin or wood, silk’s extraordinary macroscopic mechanical performance correlates positively with the mesoscopic

hierarchical structure of its fibers[11]. Five levels of structural hierarchy have been identified in silk [14]: (level 1)

amino acid sequence within the protein molecules; (level 2) secondary structure (regular hydrogen bonding pattern

between the amine and carboxyl groups yielding -helices, -sheets and -turns); (level 3) intermolecular β-

crystallites; (level 4) molecular -crystal network (nanofibrils[12]), in which amorphous chains bond the -

crystallites together along a nano-fishnet topology; and (level 5) nanofibrils network (multi-domain system). It has

been claimed that the outstanding mechanical properties of silk materials, originate from the nano-fishnet topology

structure of the β-crystallites in the molecular-scale crystal networks, and from the strong friction among

nanofibrils in the mesoscopic nanofibrils network[11]. Silk also holds attractive environmental credentials, such as

biodegradability, and sustainable and ecological production.[13,14] Historically, the use of silk in biomedical

applications dates back to the Greek and Roman civilizations who bundled up spider silk fibers for wound

treatment[15]. Nowadays, the use of silk as a high-tech fiber is widespread in clinical practice, and plays a pivotal

role in suturing[16], as well as in plastic and internal surgery[17]. From its early use to the current applications, a

myriad of approaches have been proposed to extract the highest potential out of this natural material. Reverse

engineering of silk, encompassing the separation of SF from the SS fractions, has been claimed as an attractive

strategy to fabricate SF films, foams, fibers and meshes. Using porous SF materials is particularly attractive in the

framework of wound healing[18] due to suitable cell attachment and migration both in vitro[19] and in vivo[20]. Silk-

based wound dressing materials have, to the best of our knowledge, been solely built from degummed (SS-free)

silk. SS’s largely accepted immunogenic role has limited the development of SS-containing silk materials for

biomedical applications.[21] This shortcoming has been, however, challenged over the past decade with a growing

body of work both in academic and clinical contexts[22–26] where the potential of SS has been emphasized.

Combining recent knowledge on SS with the widely explored solution processing routes to fabricate SF

4

materials[27] opens an interesting pathway to redesign silk biomaterials based on their native composition and

topology.

Here we propose an unprecedented strategy to fabricate a silk dermal patch with redesigned ice-templated

anisotropic macroporosity in which the native topological assembly of the silk is reproduced. The radically new

concept introduced here relies on the controlled disassembly and re-assembly of silk proteins. It involves the

removal of the outer component (SS), followed by shaping the core component (SF) via ice-templating, and finally

covering back the latter with SS to yield a reconstructed macroporous silk (SS@SF) (Figure 1). Our strategy

couples top-down structural disassembly (SS removal) and bottom-up assembly (SF ice templating followed by

SS coating) to engineer a functional material with improved features targeting biomedical applications

Figure 1. Simplified scheme for the fabrication of SS@SF/Rif foams. Following degumming, a SF solution is

structured into a macroporous foam via ice templating. Upon MeOH stabilization and Rif loading, the resulting

foams are covered with a SS sheath, recapitulating the initial concentric organization of native silk.

The constructs produced here are obtained via a four-step process (Figure 1): (1) Bombyx mori silk

degumming; (2) ice-templating of an aqueous SF solution, ensuring a controlled anisotropic macroporous

structure; (3) methanol (MeOH)-induced enhancement of the β-sheet domain content of SF; and (4) coating of SF

with a SS sheath to yield a reconstructed 3D macroporous silk material. The SS@SF core-shell system obtained

5

represents a suitable 3D cage to encapsulate and deliver Rifamycin (Rif), a bactericidal hydrophilic molecule that

works by binding to – and inhibiting – the DNA-dependent RNA polymerase[28].

The present SS@SF platform was tested as an antibiotic delivery system under the form of disk patches, an

attractive vector for the delivery of topical drugs[29,30]. The fabrication of patches for the topic delivery of

antibiotics is of the utmost relevance for the prevention and treatment of many primary cutaneous bacterial

infections, but finds particular relevance in the treatment of chronic wounds, such as neuropathic[31], venous[32],

pressure[33], and diabetic[34] ulcers, which affect millions of patients each year throughout the world. Here, we took

advantage of the MeOH stabilization step to load SF foams with Rif before achieving a subsequent coating with

SS sheath. As-produced SS@SF foams show liquid wicking properties, an important asset to keep wounds clear

from exudates during healing. The SS coating helps to retain more Rif than SS-free SF. The freezing rate for foam

fabrication tunes its release kinetics in aqueous solution. Long-term antibacterial activity on agar plate nominates

these novel silk matrices as good candidates for developing antibacterial dermal patches.

2. Results

Porosity plays a critical role in the performance of antimicrobial dermal patches. It dictates the material’s

mechanical and liquid transport properties, two key issues in ensuring the patch cohesiveness and its ability to

extract exudates from the underneath wound. To address the need to fabricate porous patches with a controlled

macroporous texture, we performed uniaxial ice-templating experiments involving different SF solution

concentrations and cooling rates. We prepared SF foams from 2.5, 5.0 and 7.5 wt. % SF aqueous solutions at three

different cooling rates (1, 10 and 100 °C.min−1). Samples obtained from 2.5 wt. % SF solutions were too fragile to

withstand handling and were subsequently discarded from further studies. Both 5.0 and 7.5 wt. % SF solutions

originated foams with lamellar pore section morphology, as evidenced in the Scanning Electron Microscopy

(SEM) images obtained from slices cut perpendicular to the ice growth direction (Figure 2A). Both SF

concentrations yielded comparable pore sizes, as deduced from the wall-to-wall distance (measured from each

pore minimum Feret diameter). As discussed elsewhere, the elastic properties of ice-templated materials usually

6

correlates positively with the concentration of the biopolymer used prior to freezing.[10] Producing foams whose

pore sections are invariant with the polymer initial concentration allows thus to independently modulate the

mechanics of the foams without altering the pore dimensions. In contrast, the impact of the cooling rate on the

characteristic pore size was marked. The average pore wall-to-wall distances (smallest axis of each pore section)

decreased from 90 to 70 µm ([SF] = 5.0 wt. %) and from 100 to 60 µm ([SF] = 7.5 wt. %) with increasing cooling

rate (Figure 2B). The full wall-to-wall distance data, obtained by SEM image analysis (N > 180), are plotted next

to the boxplots in Figure 2B for a detailed account of the dispersion of the measured average distances. These

results suggest that it is possible to selectively control the average pore size of biopolymer foams by changing SF

concentration and cooling rate. To maximize the patches’ flexibility and ability to accommodate skin strain during

use, we have selected the materials obtained from 5.0 wt.% SF aqueous solution for the elaboration of antimicrobial

patches. These materials they display a similar range of pore-wall distances for each freezing rate (as compared

with 7.5 wt.%) yet a less rigid behavior upon handling.

Given that the application of antimicrobial patches must reside and act in a humid environment, one of the

central requirements that needs to be fulfilled by a prospective material is water insolubility. Unlike most

biomaterials processing routes, ice templating is particularly useful to preserve the structural and molecular

integrity of biopolymers[3]. This feature, commonly considered an advantage to shape temperature-sensitive

systems, also means that the intrinsic properties of SF, such as its water solubility, are not modified during the

process.

Owing to the strong hydrogen bonding interactions and hydrophobic nature, β-sheet crystallites favor water

insolubility, whereas random coils and -helices promote water solubility[36–38]. The control of the hydrophilic–

lipophilic balance (HLB) in the SF environment, mediated by small molecules, allows fine-tuning of the structure

of the SF protein, thus enabling to engineer its surface properties[39]. In SF the transformation of random coils into

β-sheet structures is facilitated in the presence of MeOH (HLB = 7.95[40,41]).

7

Figure 2 The macroporous structure of ice-templated SF foams is defined by the freezing rate. A) SEM images of

SF foams obtained via ice-templating of 5.0 and 7.5 wt. % SF aqueous solutions at 1, 10 and 100 °C.min−1. B)

Pore wall-to-wall distances determined by analysis of SEM images after thresholding in FIJI software[35]. Boxplots

display the 10th, 25th, 50th, 75th and 90th percentiles, and the average values are marked with square symbols. The

individual data points measured from image analysis are displayed in grey to the left of each boxplot.

To induce the formation of β-sheets, SF foams were exposed to MeOH after freeze-drying. The as-prepared

materials preserved the macroporous texture, as demonstrated by SEM (Figure 3A), where both the pore section

aspect ratio and the SF walls’ thickness remained unchanged. The formation of additional β-sheet domains was

tracked by Attenuated Total Reflection/Fourier Transform Infrared (ATR/FTIR) spectroscopy (Figure 3B). The

intensity maximum of the broad and non-resolved characteristic amide I band at 1643 cm−1 indicated a

predominance of random coil conformations in the SF foams. Upon MeOH treatment, the maximum intensity

8

shifted to 1625 cm−1 and a shoulder emerged at 1699 cm−1. Both spectral features evidence the occurrence of β-

sheet structures[42,43]. The stability of SF-MeOH in aqueous media was confirmed by the absence of significant

redispersion of the foams soaked in water for at least 4 days (data not shown).

To promote the formation of a reconstructed 3D silk material, the SF-MeOH foams were coated with a SS

envelope to reproduce the silk’s native composition and topological arrangement (SS@SF). The as-obtained

SS@SF foams retained the macroporous structure defined through the previous ice templating and MeOH-

exposure processing steps (Figure 3A). The higher magnification SEM image (right panel, third row of Figure 3A)

depicts the foams after SS coating. The walls of the SS@SF foams do not display any noticeable change in

thickness when compared to SF-MeOH, suggesting that the SS layer thickness is negligible compared to the SF

walls. The characteristic amide I spectroscopic signature of the β-sheet secondary structure was not significantly

modified in SS@SF (1623 and 1699 cm−1 events) with respect to SF-MeOH. The co-occurrence of the SS and SF

fractions could be evidenced by confocal microscopy (Figure 3C). SF could be directly imaged by Second

Harmonic Generation (SHG), whose signal is a characteristic feature of the co-alignment of non-centrosymmetric

protein secondary structures (-helices and β-sheets). To ascertain the presence of SS in SS@SF foams,

Rhodamine-B-isothiocyanate (RITC) was grafted to SS, and the latter was subsequently coated onto the surface

of SF-MeOH by immersion for 2 h. The left panel in Figure 3C was obtained by exciting SF-MeOH samples under

a confocal microscope at ex = 880 nm and collecting the emission signal at em = 440 nm, corresponding to the

frequency doubling characteristic of SHG. In the right panel of Figure 3C, the SS@SF foam displayed SHG signal

and RITC fluorescence endowed by the SF crystallites and the RITC-grafted SS, respectively, suggesting a

successful coating. The pixel intensity of the two channels (SHG signal from the β-sheet domains of SF and

fluorescence from the RITC-SS moiety) at each plane correlated positively as determined from the co-localization

analysis using FIJI software[35]. Data correlation was analyzed after an automated thresholding of both channels

using the Costes algorithm (SHG/SF max. threshold = 34 and RITC-SS max. threshold = 31 for 256 gray levels)[44].

Figure 3D depicts the two-dimensional (2D) correlation heatmap obtained from co-localization analysis. Pearson’s

correlation coefficient –which ranges from -1 for anti-correlation to 1 for perfect correlation – was equal to 0.55

9

which may relate to signal intensity difference between the two channels [45]. Manders’ tM1 parameter[46] was

equal to 0.61, which indicates the high fraction of SF voxels (above threshold) that are co-localized with SS voxels

(above threshold). These results confirm that after automatic thresholding (that eliminates the large number of

background voxels) an extensive coverage of SF structures by SS was evidenced.

To understand the ability of SS@SF patches to favor exudate wicking from the wound surface, we designed

a setup that mimicked the wound exudate formation at the wound/patch interface (Supporting Information Figure

S1). Hydrated SS@SF foams were placed over a poly(dimethylsiloxane) (PDMS) surface representing the

wound/skin surface. A 3 µL.s-1 inflow of water – used as a simplified exudate liquid – was delivered under the

patch by a needle crossing the PDMS substrate. After the water filled the pores of the macroporous silk patch, a

water cap accumulated above the surface of the patch. Figure 3E depicts the water cap profile measured above the

patch surface. The low contact angles and the fact that the water cap occupies the full upper surface of the foam

suggest that the patch is hydrophilic. These results are in good agreement with the hydrophilic nature of SS

covering the hydrophobic β-sheets[47,48] present in the core fraction (SF-MeOH). The volume of the water cap was

calculated from the cap height and radius (see Supporting Information eq. S1) detected by video analysis using

FIJI software. Two regimes are observed in the plot of the water cap volume as a function of time (Figure 3F).

The first, between the onset of water injection (0 s) and the initial appearance of the water cap above the foam (25

s), corresponds to the filling of the sample pores. Multiplying the inflow rate (3 µL.s-1) by the time (25 s) yields

75 µL. This value is coherent with the foam total volume (ca. 81 µL) and the low mass fraction of SF foams (ca.

5 wt.%). The SS@SF macroporous silk foams are thus capable of retaining ca. 20 times more water than the

foams’ dry weight. From 25 s onwards, the water cap volume grew linearly until it detached from the substrate.

The measured water cap growth rate, 1.9 µL.s-1, indicates that most of the exudate production (over 60%) was

effectively extracted from the wound model to the outer surface of the patch.

One aspect that is often disregarded when considering antibacterial dermal patches is their deformability

under strain. In fact, in order to develop antibacterial dermal patches displaying optimal adhesion to the wound,

the mechanical solicitation induced by the deformation of skin should be taken into account. We performed a

10

compliance test for both SF-MeOH and SS@SF hydrated foams adhered on the surface of PDMS substrates that

were stretched to mimic the stretch ratio of skin (c.a. 20-30% depending on the body area in human skin[49]) (Figure

3G). The stretch ratio of the foams was plotted against the PDMS stretch ratio to extract the compliance of the

hydrated foams under strain induced by the PDMS substrate. Regardless of the freezing rate and SS-coating, all

foams exhibited good compliance when PDMS substrate was stretched up to 10%, a value above which foams

exhibited lower stretch ratio than the substrate. When the PDMS substrate was stretched to 25% (close to the

maximum normal arm skin surface strain[49]), the foams were stretched by approximately 14%. Although the

literature on dermal patches seems to lack the quantification of their ability to accommodate the skin (or a skin

model) stretching without detachment, we believe such measurement is central to provide an idea of the conditions

under which the patches can be applied without mechanical stabilization conveyed by adhesives. Although the

patches did not deform to the same extent as the skin model beneath, the SF-based patches did not detach when

the substrate was stretched up to 30%, pointing out their suitability to accommodate normal mechanical solicitation

induced by skin without requiring a supplementary protective layer.

To ascertain their applicability as antibacterial patches, SF foams were loaded with Rif dissolved in MeOH

(Figure 4A). The loading was ca. 17.5 mg.g-1 and did not depend on the freezing rate (Figure 4B). After immersion

of the foams in an aqueous solution containing SS, the loading dropped to ca. 0.5 wt. % and showed a decrease

with increasing freezing rate. Foams immersed in water displayed smaller drug retention (< 0.25 wt. %). This

result demonstrated the crucial dual role played by the SS coating: drug leaching minimization and drug loading

tuning according with the freezing rates. It can be hypothesized that the deposition of SS on the surface of SF-

MeOH foams minimized the desorption of Rif during water immersion. Increasing freezing rates – that resulted in

decreased pore wall-to-wall distance (Figure 2B) and consequently in higher specific surface area – may have

required a longer period for full SS coating of the foam’s internal surface, leaving more time for drug desorption.

11

Figure 3 The modification of SF with SS following MeOH treatment step preserves the morphology and

mechanical properties of the foams, while promoting a major change in the protein’s -sheet proportion. A)

Scheme of SF-based foams modification process and corresponding SEM images at 5.0 wt. %, obtained at 10

12

°C.min−1. B) ATR/FTIR spectra of amide I band region of SF foams after ice-templating (grey line), after MeOH

treatment (blue line) and following SS deposition (magenta line). C) Confocal microscopy of SF-MeOH (left

panel) and SS@SF (right panel) foams fabricated at 10 °C.min−1 (Scale bar 100 µm). Blue channel was obtained

from second harmonic generation (SHG) signal from the β-sheet domains in SF. Magenta channel codes for SS

grafted with Rhodamine isothyocyanate (RITC). D) 2D correlation heatmap between blue (SF) and magenta (SS)

channels. E) Profile of water cap above SS@SF patch fabricated at 10 °C.min−1 during wicking experiments. Water

contact angle of SF-MeOH and cooling rate. F) Wicked water cap volume as function of time. G) Compliance

tests of SF-MeOH and SS@SF, fabricated at 1 (circle symbols), 10 (triangle symbols) and 100 (star symbols)

°C.min−1 cooling rate. Blue and magenta symbols represent SF-MeOH and SS@SF, respectively.

The kinetics of Rif release were studied for both SF-MeOH and SS@SF foams prepared at different freezing

rates. All foams kept releasing Rif for 96 h. The release profiles revealed a typical three-stage process: (1) a burst

release at the beginning; (2) a slower stage in the middle; (3) a plateau at the end. Given the similar Rif content

within the SF-MeOH foams (Figure 4B), tuning freezing rate during foam fabrication could regulate the release

kinetics in aqueous media (Figure 4C). Compared to the SF-MeOH foams, the SS@SF foams showed faster Rif

release kinetics. This effect can be explained by the higher loading of Rif in SS@SF foams that presumably

generated a higher gradient between the drug-rich foam core and the solution. The influence of freezing rate on

Rif release for SS@SF foams was thus difficult to assess since their initial loading varied significantly (Figure

4D). To gain further insight into the release behavior, we applied three common drug release models (Higuchi,

Korsmeyer-Peppas and Weibull model). The fitting curves (Figure S2) and the summary of the results (Table S1)

suggests that the Weibull model (Table 1) described the release profile more accurately than the other models

considered.

13

Figure 4 Antibiotic release behavior in SF-MeOH and SS@SF foams is dictated by SS coating. A) Rif loading

process in SF foam after MeOH treatment and SS coating. B) Initial Rif loading in SF-MeOH (white column),

after soaking SF-MeOH 2 h in H2O (yellow column) and after soaking SS@SF 2 h in SS solution (red column),

as function of freezing rate. C, D) Rif release profiles in aqueous media from Rif-loaded SF-MeOH (left panel)

and SS@SF (right panel). Release profiles were fitted to the Weibull function. E) Kirby–Bauer test of Rif-loaded

SF-MeOH and SS@SF foams (5.0 wt. % SF prepared at 1, 10 and 100 °C.min−1). Cumulative and inhibition zones’

radius at each plate. Yellow and red lines represent release from SF-MeOH and SS@SF, respectively. Darker lines

correspond to foams prepared at 1 °C.min-1, intermediate lines to 10 °C.min−1 and lighter lines to 100 °C.min−1.

Lines between experimental data points are exclusively a guide to the eye.

14

Table 1. Weibull fitting parameter of Rif release profiles for the SF-MeOH and SS@SF foams.

Freezing rate

(°C.min−1) Composition

Weibull model, 𝒚 = 𝑴𝟎(𝟏 − 𝒆(−𝒌𝒙)𝒏)

k n r2

1

SF-MeOH

0.145 ± 0.001 0.55 ± 0.02 0.999

10 0.04 ± 0.01 0.48 ± 0.03 0.999

100 0.003 ± 0.005 0.34 ± 0.03 0.999

1

SS@SF

0.16 ± 0.02 0.62 ± 0.05 0.997

10 0.14 ± 0.01 0.61 ± 0.04 0.998

100 0.12 ± 0.02 0.49 ± 0.03 0.998

In this model, 𝑦 = 𝑀0(1 − 𝑒(−𝑘𝑥)𝑛), where y corresponds to the cumulative release of the active principle,

M0 is the total amount of drug released, k stands for the release rate constant and n corresponds to the structure-

related parameter. In all cases, n was less than 0.75, indicating a Fickian diffusion-dominated release from

cylindrical pores.[50] Compared to the release from SF-MeOH foams, SS@SF foams exhibited a relatively higher

k at a given freezing rate. This faster release constant could be partially due to the higher Rif loading. SF-MeOH

foams possessed similar Rif loading, but significantly different release rate with varying fabrication speed.

Increasing freezing rate from 1 to 100 °C.min−1 reduced the release rate k by approximately 50 times. This finding

could be associated with intrinsic physical characteristics of the matrix, such as porosity and pore size

distribution[50,51], as well as with the possible erosion of the matrix[52]. The reduced pore section observed for high

freezing rate foams is likely to have limited the diffusion of Rif out of foams, therefore delaying the release

kinetics. To explore the role of the protein’s dissolution on drug release, we conducted protein quantification in

the liquid medium for SF-MeOH and SS@SF for 96 h. For the SS@SF foams, the dissolution of silk proteins

during release experiments displayed a similar profile as Rif released (Figure S2), indicating its essential role on

antibiotic release. In parallel, the similar dissolution of SF from SF-MeOH foams in water suggests that the limited

solubility of MeOH-treated SF, rather than the foams’ morphology, determined the total protein in solution.

Altogether, both SS coating and freezing rate played a positive role, improving Rif initial loading within the foams.

15

The combined effect of these two parameters can thus act as effective lever to modulate indirectly the release

profile from SF-based foams.

Finally, we performed Kirby–Bauer tests[53] to validate the antibacterial behavior of Rif against S. aureus

(Figure 4E). Inhibition zone diameter is the main indicator to describe the antibacterial activity of materials.

Compared to previous studies in solution, here Rif had to diffuse through the agar gel and bacteria growth

inhibition required that antibiotic concentration was larger than its Minimum Inhibitory Concentration (MIC).

Thus, the kinetics profiles resulting from the disc method were expected to be delayed compared to

spectrophotometric titration of antibiotics released in solution. Moreover, the SS@SF foams exhibited longer Rif

release (~ 9 days) than SF-MeOH foams (~ 4 days). In addition increased freezing rate led to lower cumulative

inhibition radius and thus less Rif release. Besides, variation of the freezing rate failed to tune Rif release kinetics

on solid agar plate for SF-MeOH foams in contrast the situation observed in solution. The different Rif release

kinetics at varying freezing rates for the SS@SF foams should thus be ascribed to the different initial Rif loading

(Figure 3B).

3. Discussion

Altogether, it was possible to obtain SF foams with oriented pores whose size decreased with increasing freezing

rate. However, after treatment with MeOH – added for SF matrix stabilization purposes – in the presence of Rif,

the freezing rate played a negligible role both on the foam stability in aqueous solution, and on the antibiotic

loading. It did impact, however, on the Rif release kinetics in solution, but not in the solid form, probably because

the low initial loading limited antibiotic accumulation above its MIC. Deposition of a SS coating had no significant

impact on the material’s porosity, but allowed improving the short-term retention of Rif in water. The resulting

loading of the reconstituted silk foams was highest for the lowest freezing rate, i.e., for the largest pores, suggesting

some influence of SS diffusion on the coating extent.

Drug release studies in solution showed that the lowest freezing rate also led to the fastest delivery rates for

both uncoated and coated foams. It was difficult to compare the role of the SS coating on the release kinetics at

16

various freezing rates since drug loading was markedly different. If the drug loading influence is disregarded, then

we are led to conclude that the release was faster for the coated foams, as described by the Weibull model

parameters. This effect was, however, mostly driven by the strong difference in Rif loading, as demonstrated in

Figure 4B. Parallel studies of the protein release from the foams indicated that SS was also liberated, therefore

suggesting that the two processes, i.e. coating degradation and antibiotic release, were correlated.

Nevertheless, when the antibacterial properties of the foams were evaluated in the solid form, much better

performances were observed for the reconstructed silk foams. We attribute this evidence to the higher initial Rif

loading compared to SF-only samples. It can also be pointed out that, compared to the studies carried out in

solution, the leaching of the SS molecules was slower within the agarose gel environment, allowing the coating to

play a protective role over a longer period of time.

From a wound repair perspective, it is well-established that 2 to 10 days are needed for reconstructing

connective tissue and new blood vessel networks that are essential for subsequent tissue remodeling.[54] During

this period, prevention from bacterial infection is critical. Assuming that our testing methodology by the disc

method is representative of the working environment of antibacterial patches on skin, our reconstructed silk-based

antibacterial patches achieved ~ 9 days-continuous release of antibiotics, clearly meeting the above demand.

Moreover, the ability to produce aligned macropores perpendicular to the patches’ surface promoted efficient water

wicking properties which dictate its applications. Foot ulcer wounds from diabetic patients infected with S. aureus

were reported as having the highest exudate rate from a number of other exudate-forming wounds.[55] The case

study evaluated 3 patients with the same pathology and determined that the average exudate formation rate from

the ulcers was 2.8 mL.cm-2.day-1. Each 1mm-high SS@SF patch reported here can accommodate a liquid

absorption rate close to 170 mL.cm-2.day-1 (1.9 µL.s per sample surface area, ca. 0.95 cm2) equivalent to over 60

days of exudate production. This unique liquid wicking capacity comforts the long-term application of the SS@SF

patches as determined by the antibiotic release studies that confirmed their active drug release for 9 days.

4. Conclusion

17

Combining top-down and bottom-up design approaches resulted in macroporous silk materials that recapitulate

the topological arrangement of native silk with an inner phase of SF surrounded by a sheath of SS. Such

combination enabled to tune the foams’ surface and bulk properties, allowing us to engineer a material with

promising performance in controlled drug release. Stabilization of SF foams with MeOH, enabled the modulation

of SF solubility without a crosslinking step to create stable, hydrated foams with controlled macroporosity. SS

coating favored the retention of higher antibiotic amount than non-coated foams, resulting in materials that

exhibited antibacterial properties in a solid hydrated environment beyond 1 week. Moreover, we show that coating

SF macroporous foams with SS led to macroporous silk matrices that acted as liquid wicking materials able to

extract the fluids generated between the patch and a skin model. The combination of controlled morphology in

absence of crosslinking, high loading capacity, suitable liquid wicking properties, and controlled release of

antimicrobial loading configures an outstanding material with clear interest as antibacterial dressing for clinical

applications.

The design-oriented strategy introduced here takes advantage of the unique intrinsic organization of silk. It

encompasses the controlled removal of the outer structure of the raw natural material and the judicious redesign

of its core structure to re-build an engineered biomaterial with new functionalities. This innovative approach may

be a basis for the rational design of biomaterials with other applications in the clinical context that require a

combination of liquid transport, biocompatibility, and controlled release properties.

5. Experimental section

Preparation of SF aqueous solution: Preparation and purification of SF was performed according to the

protocol published previously.[27] Firstly, Bombyx mori cocoons (5.0 g) was cut into small pieces and degummed

by boiling them for 30 min in a 0.02 M sodium carbonate (Na2CO3, SIGMA-ALDRICH, ≥ 99%) solution. After

that, the fibers were rinsed several times with distilled water, squeezed out to remove the excess water, and finally

dried overnight in a fume hood. Secondly, SF was immersed into 9.3 M lithium bromide (LiBr, FLUKA, > 99%)

solution (ratio, 1 g: 4 mL) and then incubated at 60 °C for 4 h. A 3.5 kDa membrane (ThermoFisher Scientific)

18

was used in the dialysis of the SF solution during 2 days. The final SF solution (7.5 wt.%) was recovered and

stored at 4 °C. The 5.0 wt.% SF aqueous solutions were prepared by dilution from the initial solution.

Fabrication of SF foam: 2.5, 5.0 and 7.5 wt. % of SF solutions were freeze-cast at 1, 10 and 100 °C.min−1

with unidirectional freezing home-made setup[10]. Briefly, 2 mL sample solution was pipetted into a plastic tube

disposed onto a temperature-controlled copper rod and a thermal controller was used to precisely control the

temperature gradient. After stabilizing for 3 min, the temperature was decreased from 20 to −80 °C at a controlled

freezing rate. Once the temperature reached −80 °C, sample was recovered and then freeze-dried (ALPHA 2-4

LD, Bioblock Scientific) for 24 h.

SS-coating of SF foam: Firstly, 100 mL of 10 mg.mL−1 silk sericin (SS, SIGMA-ALDRICH) in carbonate

buffer (pH 9.3) was mixed with 1 mL of 1 mg.mL−1 Rhodamine-B isothiocyanate (RITC, SIGMA-ALDRICH, >

80%) in N,N-dimethylformamide (SIGMA-ALDRICH, 99.8%) overnight. The mixture was then dialyzed to

deionized water with a 3.5 kDa cutoff membrane (ThermoFisher Scientific) for 2 days. Dialysis water was changed

every 4 h. RITC-SS was recovered upon freeze-drying the dialyzed solution for 24 h. Prior to SS coating, SF foams

(5.0 wt. %, 10 °C.min−1) were immersed in MeOH for 2 h and freeze-dried for 2 h, which resulted in SF-MeOH

foams. Subsequently, SS coated SF foams (SS@SF) were obtained by immersing for 2 h the SF-MeOH foams into

2 wt. % SS/RITC-SS aqueous solution and then drying it at 37 °C in an oven for 24 h.

Scanning electronic microscope (SEM). Morphological analysis was performed in a Hitachi S-3400N SEM

(Japan) at 10 kV. Scaffolds were cut along and perpendicularly to the main freezing direction and then sputtered

with a 10 nm gold layer on the surface in order to improve its conductivity. Image analysis was performed by

Image J software.

Confocal microscopy: SF non-centrosymmetric structures (determined by their SHG signal) and fluorescence

signal of RITC-SS coating on SF foams were analyzed on a Leica SP5 upright confocal, multiphoton laser scanning

microscope in order to get information about the SS distributions inside the foam. Image processing was performed

with Image J software. Co-localization analysis was performed using the Coloc2 plugin that performs Costes[44]

19

auto-threshold analysis to define the signal-rich voxels in each channel followed by Manders[46] and Pearson

analysis[45].

Attenuated Total Reflection (ATR) Fourier-transform infrared (ATR/FTIR) spectroscopy: SF, SF-MeOH,

SS@SF foams were analyzed by FTIR using a PerkinElmer spectrometer 400 with ATR accessories. All

ATR/FTIR spectra were recorded in the 4000 to 400 cm−1 range by averaging 4 scans and a resolution of 1 cm−1.

Drug release in aqueous solution: SF foams (5.0 wt. %, at 1, 10 and 100 °C.min−1) were directly immersed

into 1 mg.mL−1 Rif methanolic solution for 2 h and then freeze-dried for 24 h. Half of the samples were

subsequently disposed into water for 2 h and the other half in 2 wt. % SS aqueous solution for 2 h. All foams were

then dried in an oven at 37 °C. The dried samples were then disposed into 3 mL of water and the release of Rif

and proteins were detected at 445 and 280 nm respectively with a UV-vis spectrometer (UVIKONXL

SECOMAM). Three replicates were performed for each condition. The release profile was fitted with three models

(Higuchi, Korsmeyer-Peppas, Weibull) in Qtiplot software.

Kirby–Bauer disc diffusion test: 5 wt. % SF scaffolds fabricated at 1, 10 and 100 °C.min−1 were immersed

into 1 mg.mL−1 Rif methanolic solution for 2 h, freeze-dried, and cut into ~1 mm slices. Half of the samples were

put into water and the other half in 2 wt. % SS aqueous solution for 2 h. All samples were dried in an oven at 37

°C. The dried slices were disposed on the surface of agar culture plate pre-cultured with 107 cells.mL−1

Staphylococcus aureus. The plates were incubated at 30 °C for 24 h and then the inhibition zone was measured.

The sample disks were then successively transferred onto a new agar culture plate with pre-cultured bacteria for

another 24 h. The inhibition radius was measured daily until no inhibition zone was detected. Data analysis was

performed by Graphpad Prism software.

Liquid wicking experiments: SS@SF foams’ (5.0 wt. %, at 10 °C.min−1) ability to transport water away from

the patch/skin interface was determined in a home built setup depicted in the Supplementary Information section

(Fig. S1). Briefly a 10 CC Terumo syringe (∅ = 15.8 𝑚𝑚), controlled by a syringe pump, was connected to a

hypodermic needle piercing a PDMS substrate where the patch was deposited. The needle tip protruded by 0.5

mm the PDMS surface. A hydrated SS@SF foam (diameter of 1.1 cm and thickness of approximately 1 mm) was

20

deposited on the PDSM surface (over the needle tip). A flow speed of 3 µL.s-1 was imposed with the syringe pump

to inject water in between the bottom of the sample and the PDMS surface. Finally, the volume of water cap above

the foam was calculated from the cap height and radius, measured from video analysis on FIJI.

Compliance test: SF and SS@SF foams (5.0 wt. %) fabricated at 1, 10 and 100 °C.min−1 were cut into disk-

shaped pieces with height of approximately 1 mm. The samples were then immersed for 2 h in water and 2% wt.

% SS solution, respectively. Subsequently, wet samples were disposed on the surface of a PDMS substrate that

was stretched to different ratios (5%, 10%, 15%, 20%, 25% and 30%). Relative deformation of the substrate and

the sample were measured by video analysis using predefined marks on both the substrate and the sample to infer

local strain. Three samples were analyzed for each condition.

Acknowledgements

KQ acknowledges funding from the China Scholarship Council, PhD grant n. 201606230232. This

research was funded by Fundação para a Ciência e a Tecnologia (FCT) (UID/QUI/00616/2019,

UID/QUI/00686/2019). FCT and Égide are acknowledged for Pessoa/Hubert Curien 2017/2018 project

“Redesigning silk fibroin-based composites via freeze-casting” (Ref. 3775/38003ZHR). F.P. Pereira

thanks FCT-UM for the contracts in the scope of Decreto-Lei 57/2016 and 57/2017.

References

[1] J. Li, Q. Bao, S. Chen, H. Liu, J. Feng, H. Qin, A. Li, D. Liu, Y. Shen, Y. Zhao, Z. Zong, J. Orthop. Res.

2017, 35, 812.

[2] S. Mann, G. A. Ozin, Nature 1996, 382, 313.

[3] C. Rieu, C. Parisi, G. Mosser, B. Haye, T. Coradin, F. M. Fernandes, L. Trichet, ACS Appl. Mater.

Interfaces 2019, 11, 14672.

[4] N. L. Francis, P. M. Hunger, A. E. Donius, B. W. Riblett, A. Zavaliangos, U. G. K. Wegst, M. a. Wheatley,

21

J. Biomed. Mater. Res. - Part A 2013, 101, 3493.

[5] K. Yin, P. Divakar, U. G. K. Wegst, Acta Biomater. 2019, 84, 231.

[6] U. G. K. Wegst, M. Schecter, A. E. Donius, P. M. Hunger, Philos. Trans. A. Math. Phys. Eng. Sci. 2010,

368, 2099.

[7] B. B. Mandal, E. S. Gil, B. Panilaitis, D. L. Kaplan, Macromol. Biosci. 2013, 13, 48.

[8] A. L. Oliveira, L. Sun, H. J. Kim, X. Hu, W. Rice, J. Kluge, R. L. Reis, D. L. Kaplan, Acta Biomater. 2012,

8, 1530.

[9] K. Yin, P. Divakar, J. Hong, K. L. Moodie, J. M. Rosen, C. A. Sundback, M. K. Matthew, U. G. K. Wegst,

In MRS Advances; Materials Research Society, 2018; Vol. 3, pp. 1677–1683.

[10] S. Christoph, A. Hamraoui, E. Bonnin, C. Garnier, T. Coradin, F. M. Fernandes, Chem. Eng. J. 2018, 350,

20.

[11] W. Qiu, A. Patil, F. Hu, X. Y. Liu, Hierarchical Structure of Silk Materials Versus Mechanical Performance

and Mesoscopic Engineering Principles. Small 2019, 15.

[12] S. Ling, D. L. Kaplan, M. J. Buehler, Nat. Rev. Mater. 2018, 3.

[13] F. Vollrath, D. Porter, Polymer (Guildf). 2009, 50, 5623.

[14] F. G. Omenetto, D. L. Kaplan, Science (80-. ). 2010, 329, 528.

[15] C. Holland, K. Numata, J. Rnjak-Kovacina, F. P. Seib, Adv. Healthc. Mater. 2019, 8, 1800465.

[16] JAMA - J. Am. Med. Assoc. 1939, 113, 1417.

[17] M. W. Clemens, S. Downey, F. Agullo, M. R. Lehfeldt, G. M. Kind, H. Palladino, D. Marshall, M. L.

Jewell, A. B. Mathur, B. P. Bengtson, M. W. Clemens, Plast. Reconstr. Surg. - Glob. Open 2014, 2, 1.

[18] E. S. Gil, B. Panilaitis, E. Bellas, D. L. Kaplan, Adv. Healthc. Mater. 2013, 2, 206.

[19] B. M. Min, L. Jeong, K. Y. Lee, W. H. Park, Macromol. Biosci. 2006, 6, 285.

[20] A. Sugihara, K. Sugiura, H. Morita, T. Ninagawa, K. Tubouchi, R. Tobe, M. Izumiya, T. Horio, N. G.

Abraham, S. Ikehara, Proc. Soc. Exp. Biol. Med. 2000, 225, 58.

[21] C. Pecquet, Allergic reactions to insect secretions. Eur. J. Dermatology 2013, 23, 767–773.

22

[22] R. I. Kunz, R. M. C. Brancalhão, L. D. F. C. Ribeiro, M. R. M. Natali, Silkworm Sericin: Properties and

Biomedical Applications. Biomed Res. Int. 2016, 2016.

[23] S. Gilotra, D. Chouhan, N. Bhardwaj, S. K. Nandi, B. B. Mandal, Mater. Sci. Eng. C 2018, 90, 420.

[24] C. S. Chen, F. Zeng, X. Xiao, Z. Wang, X. L. Li, R. W. Tan, W. Q. Liu, Y. S. Zhang, Z. D. She, S. J. Li,

ACS Appl. Mater. Interfaces 2018, 10, 33879.

[25] B. B. Mandal, A. S. Priya, S. C. Kundu, Acta Biomater. 2009, 5, 3007.

[26] Y. Q. Zhang, Biotechnol. Adv. 2002, 20, 91.

[27] D. N. Rockwood, R. C. Preda, T. Yücel, X. Wang, M. L. Lovett, D. L. Kaplan, Nat. Protoc. 2011, 6, 1612.

[28] M. J. Maslow, C. Portal-Celhay, In Mandell, Douglas, and Bennett’s Principles and Practice of Infectious

Diseases; Elsevier, 2019; pp. 337–349.

[29] O. Sarheed, A. Ahmed, D. Shouqair, J. Boateng, In Wound Healing - New insights into Ancient Challenges;

InTech, 2016.

[30] C. Helary, A. Abed, G. Mosser, L. Louedec, D. Letourneur, T. Coradin, M. M. Giraud-Guille, A. Meddahi-

Pelle, Biomater. Sci. 2015, 3, 373.

[31] K. Panasci, In Acute Care Handbook for Physical Therapists (Fourth Edition); 2014; pp. 283–311.

[32] R. A. Kamel, J. F. Ong, E. Eriksson, J. P. E. Junker, E. J. Caterson, Tissue engineering of skin. J. Am. Coll.

Surg. 2013, 217, 533–555.

[33] D. M. Smith, D. E. Snow, E. Rees, A. M. Zischkau, J. D. Hanson, R. D. Wolcott, Y. Sun, J. White, S.

Kumar, S. E. Dowd, BMC Med. Genomics 2010, 3.

[34] E. Gianino, C. Miller, J. Gilmore, Smart wound dressings for diabetic chronic wounds. Bioengineering

2018, 5.

[35] J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch, C. Rueden,

S. Saalfeld, B. Schmid, J. Y. Tinevez, D. J. White, V. Hartenstein, K. Eliceiri, P. Tomancak, A. Cardona,

Fiji: An open-source platform for biological-image analysis. Nat. Methods 2012, 9, 676–682.

[36] H. J. Jin, J. Park, V. Karageorgiou, U. J. Kim, R. Valluzzi, P. Cebe, D. L. Kaplan, Adv. Funct. Mater. 2005,

23

15, 1241.

[37] Q. Lu, X. Hu, X. Wang, J. A. Kluge, S. Lu, P. Cebe, D. L. Kaplan, Acta Biomater. 2010, 6, 1380.

[38] S. Lu, X. Wang, Q. Lu, X. Zhang, J. A. Kluge, N. Uppal, F. Omenetto, D. L. Kaplan, Biomacromolecules

2010, 11, 143.

[39] S. Lu, J. Li, S. Zhang, Z. Yin, T. Xing, D. L. Kaplan, J. Mater. Chem. B 2015, 3, 2599.

[40] M. Tsukada, G. Freddi, P. Monti, A. Bertoluzza, N. Kasai, J. Polym. Sci. Part B Polym. Phys. 1995, 33,

1995.

[41] X. Chen, Z. Shao, N. S. Marinkovic, L. M. Miller, P. Zhou, M. R. Chance, Biophys. Chem. 2001, 89, 25.

[42] X. Hu, D. Kaplan, P. Cebe, Macromolecules 2006, 39, 6161.

[43] A. Barth, Infrared spectroscopy of proteins. Biochim. Biophys. Acta - Bioenerg. 2007, 1767, 1073–1101.

[44] S. V. Costes, D. Daelemans, E. H. Cho, Z. Dobbin, G. Pavlakis, S. Lockett, Biophys. J. 2004, 86, 3993.

[45] K. W. Dunn, M. M. Kamocka, J. H. McDonald, Am. J. Physiol. - Cell Physiol. 2011, 300, 723.

[46] E. M. M. Manders, F. J. Verbeek, J. A. Aten, J. Microsc. 1993, 169, 375.

[47] L. Wang, H. Xie, X. Qiao, A. Goffin, T. Hodgkinson, X. Yuan, K. Sun, G. G. Fuller, Langmuir 2012, 28,

459.

[48] T. Lefèvre, M. E. Rousseau, M. Pézolet, Biophys. J. 2007, 92, 2885.

[49] R. Maiti, L. C. Gerhardt, Z. S. Lee, R. A. Byers, D. Woods, J. A. Sanz-Herrera, S. E. Franklin, R. Lewis,

S. J. Matcher, M. J. Carré, J. Mech. Behav. Biomed. Mater. 2016, 62, 556.

[50] U. Gbureck, E. Vorndran, F. A. Müller, J. E. Barralet, J. Control. Release 2007, 122, 173.

[51] K. Xing, R. Fan, F. Wang, H. Nie, X. Du, S. Gai, P. Wang, Y. Yang, ACS Appl. Mater. Interfaces 2018,

10, 22746.

[52] E. Csapó, H. Szokolai, Á. Juhász, N. Varga, L. Janovák, I. Dékány, Carbohydr. Polym. 2018, 195, 99.

[53] M. L. Kung, P. Y. Lin, S. W. Peng, D. C. Wu, W. J. Wu, B. W. Yeh, M. H. Tai, H. S. Hung, S. Hsieh,

Appl. Mater. Today 2016, 4, 31.

[54] L. A. Barnes, C. D. Marshall, T. Leavitt, M. S. Hu, A. L. Moore, J. G. Gonzalez, M. T. Longaker, G. C.

24

Gurtner, Mechanical Forces in Cutaneous Wound Healing: Emerging Therapies to Minimize Scar

Formation. Adv. Wound Care 2018, 7, 47–56.

[55] C. Dealey, J. Cameron, M. Arrowsmith, J. Wound Care 2006, 15, 149.

download fileview on ChemRxivQin et al 2020 preprint.pdf (1.50 MiB)

Supporting information

Redesigned silk: a new macroporous biomaterial platform for antimicrobial

dermal patches with unique exudate wicking ability

Kankan Qina, Rui F.P. Pereirab, Thibaud Coradina, Verónica de Zea Bermudezc* and Francisco M. Fernandesa*

a Sorbonne Université, UMR 7574, Laboratoire de Chimie de la Matière Condensée de Paris, F-75005, Paris, France

b Chemistry Center and Chemistry Department, University of Minho, Campus de Gualtar, 4710-057, Braga, Portugal

c Chemistry Department and CQ-VR, University of Trás-os-Montes e Alto Douro, Apartado 1013, 5001-801, Vila Real, Portugal

Corresponding authors: francisco.fernandes@upmc.fr, vbermude@utad.pt

s2

Figure S1. Water wicking analysis setup. The quantification of the water cap volume as a function of time was

performed by video analysis using FIJI software. The volume was calculated from the measurements of the radius

and the height of the water cap according to equation S1. The water influx was controlled by a syringe pump

connected to a hypodermic needle that crosses the PDMS substrate to reach the center of the lower surface of the

where the SS@SF patches are

Volume of a spherical cap:

𝑉𝑐𝑎𝑝 =𝜋ℎ2

3(3𝑟 − ℎ) (eq.1)

where Vcap is the volume of the spherical cap defined by a planar base, h is the height of the cap at the center of

the cap and r is the radius of the sphere.

Figure S2. Normalized Rifamycin release (A) and silk protein release (B) during drug release test. Dot lines and

solid lines represent SF-MeOH and SS@SF (5 wt.% SF was fabricated at 1/10/100 °C/min), respectively. Light

red line: SF-MeOH (SF was fabricated at -1 °C/min); light green line: SF-MeOH (SF was fabricated at -

10 °C/min); light blue line: SF-MeOH (SF was fabricated at -100 °C/min); red line: SS@SF (SF was fabricated

at 1 °C/min); green line: SS@SF (SF was fabricated at 10 °C/min); blue line: SS@SF (SF was fabricated at

100 °C/min).

s3

Figure S3. Fitting curves of Rifamycin release from SF-MeOH foams (5 wt.% SF was fabricated at 1 °C/min).

Black dot: experimental data; green line: Higuchi fitting curve; blue line: Korsmeyer-Peppas fitting curve; black

line: Weibull fitting curve.

Table S1. Rifamycin release analysis according to Higuchi, Korsmeyer-Peppas and Weibul models.

Sample

Freezing

rate

(°C/min)

Higuchi model

𝑦 = 𝐴𝑥0.5 + 𝐵

Korsmeyer-Peppas model

𝑦 = 𝐴𝑥𝐵 + 𝐶

Weibull model

𝑦 = 𝑀0(1 − 𝑒(−𝑘𝑥)𝑛)

A R2

A B R2

k n R2

SF-MeOH 1 9.4 ± 1.8 0.820 39.7 ± 7.8 0.223 ± 0,036 0.969 0.145 ± 0.001 0.547 ± 0.02 0.999

SS@SF 1 9.5 ± 1.8 0.819 38.2 ± 9.2 0.231 ± 0,044 0.956 0.157 ± 0.017 0.620 ± 0.05 0.997

SF-MeOH 10 9.5 ± 1.1 0.931 26.5 ± 3.8 0.295 ± 0.027 0.990 0.039 ± 0.010 0.482 ± 0.03 0.999

SS@SF 10 9.4 ± 1.7 0.832 36.5 ± 8.5 0.237 ± 0.043 0.960 0.143 ± 0.014 0.606 ± 0.04 0.998

SF-MeOH 100 8.13 ±

0.89

0.933 24.3 ± 1.4 0.282 ± 0.011 0.998 0.003 ± 0.005 0.343 ± 0.03 0.999

SS@SF 100 8.8 ± 1.6 0.836 36.6 ± 5.9 0.224 ± 0.030 0.978 0.122 ± 0.020 0.489 ± 0.03 0.998

download fileview on ChemRxivSupporting information preprint.pdf (629.31 KiB)