Skin Cell and Tissue Responses to Cross-Linked Hyaluronic Acid in Low-Grade Inflammatory Conditions

Hyaluronic acid (HA), used in a variety of medical applications, is associated in rare instances to long-term adverse effects. Although the aetiology of these events is unknown, a number of hypotheses have been proposed, including low molecular weight of HA (LMW-HA) in the filler products. We hypothesized that cross-linked HA and its degradation products, in a low-grade inflammatory microenvironment, could impact immune responses that could affect cell behaviours in the dermis. Using two different cross-linking technologies VYC-15L and HYC-24L+, and their hyaluronidase-induced degradation products, we observed for nondegraded HA, VYC-15L and HYC-24L+, a moderate and transient increase in IL-1β, TNF-α in M1 macrophages under low-grade inflammatory conditions. Endothelial cells and fibroblasts were preconditioned using inflammatory medium produced by M1 macrophages. 24 h after LMW-HA fragments and HA stimulation, no cytokine was released in these preconditioned cells. To further characterize HA responses, we used a novel in vivo murine model exhibiting a systemic low-grade inflammatory phenotype. The intradermal injection of VYC-15L and its degradation products induced an inflammation and cell infiltration into the skin that was more pronounced than those by HYC-24L+. This acute cutaneous inflammation was likely due to mechanical effects due to filler injection and tissue integration rather than its biological effects on inflammation. VYC-15L and its degradation product potentiated microvascular response to acetylcholine in the presence of a low-grade inflammation. The different responses with 2D cell models and mouse model using the two tested cross-linking HA technologies showed the importance to use integrative complex model to better understand the effects of HA products according to inflammatory state.


Introduction
Hyaluronic acid (HA), a major glycosaminoglycan of the extracellular matrix, plays an important role in a number of biological processes such as wound healing, cancer, or embryonic development [1].Cross-linked HA is used in a variety of medical applications, including as a dermal fller in aesthetics applications.Although typically regarded as an inert, biocompatible material due to its similarity with native HA, there have been rare instances of adverse efects which can occur from 4 weeks to more than 1 year posttreatment [2,3].Although the aetiology of these events is unknown, a number of hypotheses have been proposed, including bioflm-related response, a result of mechanical irritation with the product, or the presence of bacteria transferred into the skin during injection [4,5].Tey have also been proposed to be related to the low molecular weight of HA (LMW-HA) in the fller products, either as a component of the product or following its degradation [2,5].
Te size of HA fragments generated during degradation can also play a role in migration, apoptosis or immune regulation by activating the immune cells, such as macrophages and dendritic cell [6].Te high molecular weight polymers of HA (HMW-HA) (>500 kDa) are commonly described to have an anti-infammatory efect, while low molecular weight polymers and fragments of HA (LMW-HA) (10 kDa to 500 kDa) and oligomers of HA (0.1 to 10 kDa) have been described as involved in infammatory responses, possibly functioning as an alarm signal to the immune system, particularly for monocytes/macrophages [1,7].Moreover, the accumulation of LMW-HA seems to be strongly related to the severity of infammatory pathologies and chronic diseases such as rheumatoid arthritis, liver injury, asthma and lung injuries, suggesting dysregulation in HA metabolism or accumulation due to an increase in HA catabolism with turnover of the matrix in these infammatory diseases [8,9].LMW-HA have been widely described as proinfammatory [10][11][12] with some studies showing an increase in proinfammatory cytokine release, such as TNF-α and IL-1β, in the presence of LMW-HA [1,10].However, more recent studies have shown that the infammatory responses observed on in vitro and ex vivo murine immune cells could be due to an endotoxin contamination on samples rather than the size of the fragments [13][14][15].Tis discrepancy may be due to size-specifc efects of the fragments or the culture conditions.
An additional factor that has been proposed in these events is an activation of the systemic immune response, like which occurs with fu, cold, or seasonal allergies [2,5,16].Moreover, fller treatments are not recommended in case of active auto-immune diseases such as systemic lupus erythematosus, rheumatoid arthritis, or mixed connective tissue disease [3] that all share a low-grade infammatory microenvironment [17].Tis low-level infammation may impact cell-cell interactions between immune cells, microvascular endothelial cells, and dermal fbroblasts leading to extracellular matrix alteration [18].
Based on clinical observations, practitioners are recommended to perform proper skin preparation prior to HA injections to avoid infection and to advise patients to delay fller treatment if areas of infammation/infection are present [19].Although the efects of HA on the infammatory response have been examined, the efect of cross-linked hyaluronic acid and its degradation products on the skin, under low-level infammatory conditions, remains unexplored.In the present study, we used two diferent crosslinking technologies VYC-15L (Vycross technology) and HYC-24L+ (Hylacross technology) and their hyaluronidaseinduced degradation products [20,21].We hypothesized that a subclinical infammation present at the time of the injection of degraded (HA fragments) and nondegraded cross-linked HA could impact immune responses that could afect cells found in the dermis.M1 macrophages and mature dendritic cells, both derived from monocytes THP-1 lineages (as immune cells models), infammatory dermal fbroblasts and infammatory human dermal microvascular endothelial cell (HDMEC) were used in vitro to study the efects of HA products and their LMW degradation fragments in the low-grade infammatory environment.To further study in vivo the vascular and infammatory responses, we used HA and their LMW degradation fragments in a systemic low-grade infammatory murine model.

Endotoxin Quantifcation.
Te vehicle control, VYC-15L fragments and HYC-24L + fragments were quantifed by Endosafe PTS (Charles River).Te Endosafe ® - PTS (Portable Test System) is a rapid, point-of-use test system, which is comprised of a test cartridge along with a hand-held spectrophotometer.For this test, we used cartridge with 0.01 EU/ml sensitivity (PTS2001F).All experiments were carried out according to the manufacturer's instructions.

Macrophages M1 Polarization.
THP-1 monocytes were incubated for 24 h at 5% CO 2 37 °C with PMA (Phorbol 12myristate 13-acetate from Sigma) to induce the diferentiation into macrophages M0.Ten macrophages M0 were polarized into proinfammatory M1 for 66 hours in the presence of 100 ng/ml of LPS, 100 ng/ml of PMA and 20 ng/ ml of IFN-c (Peprotech, France).Polarization of macrophages M1 were confrmed via cytokine profle (Supplementary Table S1 and Supplementary Figure S1) and were then put at rest for 24 h, as previously reported [22].

2
International Journal of Infammation CSF, 200 ng/ml of ionomycin (Sigma) and 20 ng/ml of TNFα (Peprotech) and incubated for 4 days at 5% CO 2 37 °C.Medium exchange was performed every 2 days with fresh cytokine-supplemented medium.After cell diferentiation confrmed via cytokine profle (Supplementary Figure S2), cells were cultured with complete medium [23].Service (agreement #18529).Special efort was made to minimize the number as well as the stress and sufering of mice used in this study.Before experimentation, animals were acclimatized for 1 week.To induce a systemic lowgrade infammatory phenotype, 100 µl of very low dose of LPS (O26 : B6) at 5 ng/kg or PBS (control) was injected by intraperitoneal injection (IP) 3 times per week for 6 weeks as previously described [24,25].In order to characterize the systemic low-grade infammatory phenotype, blood was collected after sacrifce by intracardiac puncture and centrifuged at 3000 rpm during 10 min at 4 °C to study CCL2 and IL-6 in the plasma using Luminex assay (R&D system).Cutaneous biopsy (punch 8 mm Ø at the back) was also performed for histological analysis and to study in the skin TNF-α and IL-1β gene expression using RT-qPCR and IL-6, IL-1β, COX-2, and VCAM-1 protein quantifcation using ELISA assay.Tese parameters collected from low-grade infammatory mice were compared with those from healthy control mice.Acute intradermal injection (20 µl) of LPS at 5 µg/ml was used as a positive control to induce strong infammation in healthy mice.
International Journal of Infammation

Assessment of Acetylcholine-(ACh-) Mediated Vasodilation.
Two days prior to the microvascular experiments, the back was depilated with a depilatory lotion in mice under isofurane to provide hairless areas for skin blood fow measurements and iontophoretic delivery of acetylcholine (ACh 2% Sigma) performed in normal and low-grade infammatory mice.One day prior to the microvascular experiments, normal and low-grade infammatory mice received an acute intradermal injection (20 µl at a concentration of 100 µg/ml) of the Vehicle Control, VYC-15L fragments, VYC-15L, HYC-24L + fragments, or HYC-24L + undiluted.
For the microvascular experiments, all mice were anaesthetized with thiopental (65 mg•kg −1 , i.p.) and placed in an incubator (MMS, Chelles, France) to maintain a stable skin temperature (35.0 ± 0.5 °C).Te cutaneous blood fow was measured using a laser Doppler probe (PF 4001 Master, Perifux, Perimed, Sweden) in response to iontophoretic delivery of ACh with an anodal current application (100 μA for 10 s) at the site of intradermal injection.Te laser Doppler signal expressed in arbitrary units (a.u.) was digitized with a 200 Hz sampling frequency using a computerized acquisition system (MP 100, Biopac System, CA).Local iontophoretic drug delivery was chosen to assess the in vivo cutaneous microvascular function while avoiding any systemic efects.Data collection started with a 1 minute control period before the ACh delivery and was continued for 20 minutes.

Mouse Skin Analysis.
For histological analysis, mouse skin was taken from the lower back 24 h after an acute intradermal injection (20 µl) of the Vehicle Control, VYC-15L fragments, VYC-15L, HYC-24L + fragments, or HYC-24L+.Skin biopsy was fxed in AFA solution (alcohol, formalin, and acetic acid).Samples were embedded in parafn and sectioned at a thickness of 5 µm.After dewaxing, the cuts were successively stained with Harris hematoxylin, eosin G, and safron.Te matrix elements were coloured in pinkish-yellow, the cytoplasm in pink, the nuclei in purple, and HA in dark violet.
Skin tissue lysate was analysed by ELISA assay on IL-6 from Abcam (Cambridge, UK).Skin tissue lysate was prepared by homogenization in T-PER lysis bufer and supplemented with protease and phosphatase inhibitor (Termo Fisher Scientifc) in TissueLyzer using magnetic beads.Skin tissue lysates were centrifuged 10 min at 10000g and 4 °C.Supernatant were then collected and quantifed.

Mouse RNA Isolation and Real-Time Quantitative PCR.
Total RNA of skin samples was isolated using RNeasy Fibrous Tissue Mini Kit for skin biopsies (Qiagen, Courtaboeuf, France) according to the manufacturer's instructions.500 ng of total RNA was reverse-transcribed into cDNA using PrimeScript TM RT reagent kit (Takara, Shiga, Japan) and analysed by Real-Time QPCR using SYBR
2.12.Statistical Analyses.Te results are expressed as mean ± SD.ACh-mediated vasodilation was expressed as the percentage changes of the cutaneous blood fow over the basal values, calculated as the average over the 1 min control period before the iontophoretic delivery of ACh.Statistical analyses were performed in all obtained data using Prism (version 8.0, GraphPad Software Inc.).An unpaired t-test or one-way analysis of variance (ANOVA) followed by Dunnett's multiple comparison test was used to compare two or more independent groups, respectively.A p value less than 0.05 was regarded as statistically signifcant.

HA Degradation Products:
Characterization and Endotoxin Quantifcation.After a 6-hour incubation with hyaluronidase, VYC-15L and HYC-24L+ were degraded into fragments with a mean molecular weight of 30 ± 4.6 kDa and 80 ± 1.2 kDa and a polydispersity of 2.4 and 3.3, respectively.More than 75% of the total hyaluronidase-induced degradation products from VYC-15L were smaller than 40 kDa after 6 hours of degradation, while 55% of total HYC-24L + fragments were smaller than 40 kDa (Figure 1).Furthermore, degradation products from HYC-24L + contained more HMW fragments with a molecular weight >500 kDa than those from VYC-15L.Tese results indicate that HYC-24L+ is less degraded than VYC-15L for the same duration of degradation.
VYC-15L fragments, HYC-24L + fragments, and the Vehicle Control contained less than 0.02 EU/ml of LPS (Figure 1), showing that the degradation process did not introduce endotoxin and can be considered endotoxin-free in comparison to the positive LPS control (∼500 EU/ml).
In the condition medium of mDC at the 24 h time point (Figure 2(d)), no signifcant diference of TNF-α protein level was observed between conditions.Level of IL-1β was too low to be detectable by the Luminex assay except for LPS condition as positive control (Supplementary Figure S3B).IL-8 protein was only signifcantly overexpressed in the presence of VYC-15L fragments (4842 ± 325 pg/ml) compared to the vehicle control (2943 ± 242 pg/ml)(Figure 2(e)), although this overexpression was much lower compared to the positive LPS control (29697 ± 500 pg/ml) (Supplementary Figure S3B).
No diferences were observed for any cytokines for either M1 macrophages or mDC, regardless of HA treatment group, at the 48 h and 72 h time points (Supplementary Figure S4), suggesting a transient efect of HA products in contrast to the positive LPS stimulation that lasted up to 72 h (data not shown).

Cytokines Released by Preconditioned Fibroblasts (Low-Grade Infammatory Phenotype) in
Response to HA and HA Degradation Products at 72 h.Compared to noninfammatory dermal fbroblasts, IL-6, IL-1β, and IL-8 were overexpressed in fbroblasts preconditioned with M1 macrophages but to a lesser extent than the positive LPS stimulation (Supplementary Figure S5), showing a lowgrade infammation in 72 h post-preconditioned fbroblasts.

Cytokines Released by Preconditioned HDMEC (Low-Grade Infammatory Phenotype) in
Response to HA and HA Degradation Products at 72 h.Compared to noninfammatory HDMEC, IL-8, IL-1β, and IL-6 were overexpressed in preconditioned HDMEC, but to a lesser extent than LPS stimulation (Supplementary Figure S6), showing a low-grade infammation in 72 h post-preconditioned HDMEC.
In the skin, the upregulations of TNF-α (2.8-fold) and IL-1β (1.3-fold) gene expression in the skin of systemic low-grade infammatory mice were not signifcant compared to control mice (Supplementary Figure S7B).Similarly, the increase of +130% in COX-2 protein concentration in the skin of systemic low-grade infammatory mice (2410 ± 385 pg/ml) was not signifcant compared to control mice (1868 ± 312 pg/ml), such as LPS positive control (1764 ± 256 pg/ml) (Supplementary Figure S7C).No increase of IL-6 protein concentration in the skin of systemic low-grade infammatory mice (103.9 ± 22 pg/ ml) was observed compared to control mice (102 ± 16 pg/ml) (Supplementary Figure S7C).Again, the increase of +123% in  International Journal of Infammation VCAM-1 protein concentration in the skin of systemic lowgrade infammatory mice (8493 ± 1418 pg/ml) was not significant compared to control mice (6903 ± 1695 pg/ml) (Supplementary Figure S7C).Systemic low-grade infammation increased intragroup heterogeneity as shown by the increased coefcient of variation (CV) in low-grade infammatory mice compared to control mice: 21% vs 7% for IL-6 plasma level, 29% vs 20% for CCL2 plasma level, 47% vs 39% for TNF-α gene International Journal of Infammation expression, and 103% vs 71% for IL-1β gene expression and correlated with standard deviation observed.Tus, the cytokine expression was increased in the plasma (systemic infammation), and there was no overexpression, only trends with a high variability, in the skin.Tese moderate increases in cytokine expression were expected to obtain a mouse model of systemic low-grade infammation in the skin, and not strong infammation obtained in the positive control mice (Supplementary Figure S7).In addition, no diferences were observed between control mice and systemic low-grade infammatory mice on skin structure by histological analysis.No higher immune cell infltration and diferences in thickness were observed in comparison to the infltration observed in the positive control mice (Supplementary Figure S7D).On the vascular aspect, basal skin blood fow was not diferent between control mice (45 ± 13 arbitrary units (a.u.)), systemic low-grade infammatory mice (49 ± 11 a.u.), and positive control mice (49 ± 9 a.u.).Similarly, the AChmediated vasodilation was not diferent between systemic low-grade infammatory mice (32 ± 15%) and control mice (43 ± in contrast to the augmented ACh-mediated microvascular reactivity in the positive control mice (69 ± 32%) (Supplementary Figure S7E).

Efects of HA and HA Degradation Products on Skin
Structure in Mice.To study the efects of HA and HA degradation products in a systemic low-grade infammation environment, the histological analyses were performed 24 h after the intradermal injections of vehicle, VYC-15L fragments, VYC-15L, HYC-24L + fragments, and HYC-24L+ in systemic low-grade infammatory mice (Figure 5).

Skin IL-6 and IL-1β in Response to HA and HA Degradation Products in Control and Systemic Low-Grade Infammatory Mice.
In systemic low-grade infammatory mice, IL-6 protein in the skin was only signifcantly overexpressed at 24 h in the presence of VYC-15L (238 ± 52 pg/ ml) compared to the vehicle control mice (102 ± 20 pg/ml) (Figure 6), associated with an enhanced intragroup heterogeneity.In contrast, no diferences were observed for the IL-1β protein concentration in the skin between each condition at 24 h.

Discussion
Few fundamental studies have been done on delayed complications after intradermal HA injection.It has been hypothesized that HA degradation products may be related to the development of delayed complications.In this work, we hypothesized that an infammatory state could be a key mechanism of skin complications in the presence of crosslinked HA.Te objectives of this study were to assess the efect of degraded (HA fragments) and nondegraded crosslinked HA under systemic low-grade infammatory conditions on immune and dermal cells as well as vascular responses.
For many years, concerns regarding pro-and antiinfammatory efects of HA fragments have been raised [13].LMW-HA fragments were described as a danger proinfammatory signal, which stimulates stromal fbroblasts to perform an immune response and to provide a favourable environment for infltrating immune cells [26][27][28].Indeed, LMW-HA are described as an active proinfammatory molecule stimulating particularly innate immune cells, monocytes and macrophages, to produce a palette of proinfammatory cytokines and chemokines [29].Release of these cytokines recruits neutrophils and drives T lymphocyte diferentiation [7,10,30].However, it was later shown that the proinfammatory response related to LMW-HA fragments and HA could be due to endotoxin contamination from the sources of HA, enzyme, and methods for degradation [13].In the current study, we showed that all the HA fragments and HA used were endotoxin-free (at least below detection threshold), excluding any confounding efects.
Hyaluronic acid has several functions depending on its molecular weight and several studies reported HA infammatory modulation in a size-specifc manner.In the present study, two sizes of HA fragments were obtained for diferent fller formulations, 30 kDa and 80 kDa, under the same experimental conditions, highlighting potential differences between the two HA fllers, VYC-15L, and HYC-24L+, respectively.Te diference of size between VYC-15L and HYC-24L + fragments may be related to the composition or cross-linking technology of each fller or to their initial HA concentration, 15 mg/ml or 24 mg/ml for VYC-15L and HYC-24L+, respectively [20,31,32].Both are cross-linked with 1,4-butanediol diglycidyl ether (BDDE), but the composition of the material (Hylacross is HMW and Vycross is HMW and LMW) could allow for minor differences in the fnal rheological properties [32,33].Te LMW-HA in VYC-15L allows for the chains to be closer together at cross-linking, which results in both a tighter and more efcient cross-linking.Interestingly, we degraded the International Journal of Infammation  International Journal of Infammation two HA fllers at diferent time point inducing diferent range sizes of LMW-HA fragments and no infammatory efect was observed overtime in nonpreconditioned cells (data not shown), correlated with previous studies using endotoxin-free enzyme [13,14].
We observed for nondegraded HA, VYC-15L, and HYC-24L+, a moderate and transient increase in IL-1β, TNF-α in M1, but no change in cytokine release in infammatory fbroblasts and infammatory endothelial cells; contrary to LPS stimulation inducing a strong infammatory response on immune cells and dermal skin cells.
About immune cells, activated M1 responded to VYC-15L and HYC-24L + by further elevating the production of TNF-α and IL-1β, whereas only TNF-α was further increased with fragments of VYC-15L and HYC-24L + at 24 h.Te majority of immune cells do not bind HA until they are activated by an antigenic or infammatory agent [13,34,35].In the present study, CD44 gene expression was highly increased in M1 and dendritic cells compared to M0 and THP-1, suggesting a HA binding on M1 with the production of TNF-α and IL-1β [34,36].TNF-α level seems to play a major role in the immune HA modulation.In the current study, the lower production or TNF-α observed in response to fragments from both HA fllers compared to their respective nondegraded HA could reveal an alteration of HA fragments binding afnity or that there was no infammatory efect due to HA fragments, supported by no further production of IL-1β.Indeed, CD44-mediated HA efects are more complex in terms of receptor density, afnity to HA and kinetic of response [35,36].
Te results obtained underline the main role of activated macrophages at 24 h that was not persistent at 48 h and 72 h, indicating a transient transmission of infammatory cytokines in the presence of degraded and nondegraded HA, without compensative efects of anti-infammatory factors.In contrast, direct LPS stimulation on M1 and mDC induced a strong and sustained infammation up to 72 h balanced by the release of anti-infammatory cytokines, contrasting with the mild response obtained with nondegraded and degraded HA products (Supplementary Figure S3).
Tese transient and moderate infammatory responses on immune cells highlighted that HA fragments did not induce infammation over time, but suggest a possible cell communication with other dermal skin cells such as dermal fbroblasts and microvascular endothelial cells.Indeed, an important outcome of studies on molecular mechanisms that underlie infammation was that an infammatory response is not a specialized function of cells of hematopoietic origin (macrophages, neutrophils, lymphocytes etc.) but rather a fundamental attribute of several cellular types.In this context, fbroblasts play a critical role in modulating immune cells such as leukocytes and could be responsible for the establishment of chronic infammation [37].In addition, endothelial cells synthesize and secrete chemokine and cytokines, which are implicated in recruiting immune cells in response to infammatory disorders [38].Tis inside-out infammatory response [18] has been supplemented with an outside-in signalling involving activated proinfammatory fbroblasts interacting with immune cells to direct endothelial activation [39].In the present study, although endothelial cells and fbroblasts under proinfammatory M1 stimulation exhibited proinfammatory phenotype by releasing increased levels of IL-1β, IL-8, and IL-6 no major change was observed for both cell types in the presence of degraded or nondegraded HA.
Altogether, these in vitro experiments showed that HA fragments in an infammatory environment did not induce strong infammatory responses over time, while nondegraded HA induced transient infammatory modulation.
To better characterize the cellular response observed after cross-linked HA injection [1,4], we also used an in vivo murine model.It appeared that indirect coculture model present limitation and cannot refect complexity of the skin whose homeostasis relies on a fnely tuned equilibrium of well-regulated interactions between the diferent layers of the skin with their cellular and subcellular structures [40].Very low dose endotoxin in the blood circulation is one of the emerging risk factors that can cause chronic low-grade infammation in humans and animals [41,42].In this way, to mimic a systemic low-grade infammatory phenotype on mice, repeated intraperitoneal injections of a very low dose of LPS for 6 weeks in healthy mice were carried out, based on previous studies [24,25].Tis chronic LPS treatment increased proinfammatory cytokines (IL-6 and CCL2) in the 10 International Journal of Infammation plasma, with no signifcant skin infammation although infammatory markers (TNF-α, IL-1β, and COX-2) revealed a trend to increase within the skin.Tese characteristics are concordant with skin weakened by systemic low-grade infammation described in chronic pathologies such as obesity and aging [43,44] and validating our mouse model of lowgrade infammation.Tis in vivo mouse model of systemic low-grade infammation was then used to study the efects induced by intradermal injections of HA and its degradation products and evaluate diferences between the two crosslinking technologies on skin integrity and microvascular reactivity.Histological analysis revealed intact skin tissue architecture around the injection site in all mice that received vehicle in the presence of systemic low-grade infammation, revealing the nontraumatic nature of the intradermal injection in healthy adult mice.Regarding the injection of exogenous HA and its degradation products, we showed an acute infammation and cell infltration into the skin more pronounced with VYC-15L compared to HYC-24L+, in systemic low-grade infammatory mice.Likewise, nondegraded VYC-15L was the only product to signifcantly increase the IL-6 concentration in the skin in the presence of systemic lowgrade infammation (+231%), confrming the results highlighted by histological analyses.In addition, IL-1β was not oversecreted in the presence of VYC-15L, suggesting that the common infammatory pathway described in literature using CD44 was not involved in the infammatory response to the presence of VYC-15L [9].We suggest that this cutaneous infammation induced by VYC-15L injection was likely due to mechanical efects due to fller injection and tissue integration rather than its biological efects, since degradation products did not increase IL-6 protein concentration.In contrast, HYC-24L + did not modulate IL-6 concentration in the skin.Tese diferent observations between VYC-15L and HYC-24L + may be related to their diferent rheological properties, degree of cross-linking and concentration of HA inside the gel that may have diferent biological and physical efects [20,32,45,46].Furthermore, contrary to the literature we observed no anti-infammatory efect in the presence of nondegraded HA, suggesting that these biological efects are dependent to the structure of HA and to tissue localization, if its native or cross-linked HA [47,48].
Te use of acetylcholine (ACh) is the gold standard method to assess the vascular function that relies on vasomotor factors present in the tissue microenvironment.Normal vascular response to ACh relies mostly on endothelial L-arginine/nitric oxide (NO) pathway, which can be altered in pathological condition with infammation features leading to a decreased involvement of the NO pathway during ageing, diabetes [49,50].However, in a low-grade infammatory pathology such as obesity [43], an increase in Ach-induced vasodilation response has been reported to be promoted by COX pathway.We highlighted that VYC-15L was the only product capable of signifcantly modulating the microvascular reactivity of the skin in the presence of a systemic low-grade infammation.Indeed, basal vascular tone of the skin was not changed by the injection of HA and their degradation products in all mice, except a slight increase following the VYC-15L injection.In addition, we showed that endothelium-dependent vasodilations in response to ACh were increased by VYC-15L and its degradation product compared to the vehicle control.Tis potentiated acetylcholine response, which was not observed with HYC-24L+, was similar to the positive LPS control mice exhibiting a strong infammation.Surrogate signaling pathways including infammatory factors such as COX-2 can be involved in vascular ACh sensitization in low-grade infammatory pathology [43,51].We thus quantifed the amount of VCAM-1 which is a marker for vessels integrity and COX-2 as vasodilator mediators [50,52,53].Since these markers were not overexpressed either in the presence of a systemic low-grade infammation or in the presence of VYC-15L, we suggest that this vascular modulation involved another pathways or mechanisms, which require further study.Altogether, these short-term in vivo experiments using an infammatory integrative murine model induced by a very low dose of LPS, have shown that intradermal injection of nondegraded VYC-15L products increase IL-6 in the skin, cells infltration, skin perfusion, and microvascular sensitivity in response to ACh, showing a moderate and transient cutaneous proinfammatory response in contrast to HYL-24L+.

Conclusion
Any infammation that was observed in the present study was transient and independent of the degradation (i.e., HA fragments were not shown to impact the infammatory response) and may be more related to physical properties than biologic stimulation.We could suggest that diferences in rheological properties or cross-linking technologies may interact with skin components and actors, leading to different responses to VYC-15L in systemic low-grade infammatory conditions.
In addition, an in vivo model of low-level infammation was developed to evaluate a change in the basal state and its impact on infammatory responses to HA and its degraded products.Some changes were observed with VYC-15L, independent of degradation state, as it relates to short-term infammatory response and vascular sensitivity.Tis vascular sensitivity could be linked to the macrophagemediated infammation related to an increased recruitment of immune cells at the injection site and to the transient infammatory modulation observed in 2D models.Tis suggests a potential acute response, although it is unclear whether this would be related to long-term events or just short-term tissue response.Further studies would be needed to clarify the potential link with the reported rate of adverse reactions to the two types of cross-linked hyaluronan and does this correlate with our results.From the perspective of a practicing dermatologist or other aesthetic physician, a proinfammatory profle prior to HA injection may be necessary to inform about possible side efects, as recently mentioned by Rivers (19), in particular for VYC-15L based on our results.
Te diferent responses between 2D cell models and mouse model showed the importance to use integrative complex model to better understand the efects of HA products according to infammatory state.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.Te data are not publicly available due to ethical restrictions.