Highly concentrated trehalose induces transient senescence-associated secretory phenotype in fibroblasts via CDKN1A/p21

Trehalose is the nonreducing disaccharide of glucose, evolutionarily conserved in invertebrates, but does not exist in vertebrates. The living skin equivalent (LSE) is an organotypic coculture containing keratinocytes cultivated on fibroblast-populated dermal substitutes. We demonstrated that human primary fibroblasts treated with highly concentrated trehalose promote significantly extensive spread of the epidermal layer of LSE without any deleterious effects. The RNA-seq analysis data and Ingenuity pathway analysis of the differentially expressed genes of trehalose-treated 2D and 3D fibroblasts at early time points revealed the involvement of the CDKN1A pathway, which is necessary for the marked upregulation of growth factors including DPT. By contrast, the mRNA-seq data of LSEs 2-weeks after air exposure indicated that gene expression profiles are similar for untreated and trehalose-treated cells in both keratinocytes and fibroblasts. The trehalose-treated fibroblasts were positive for senescence-associated β-galactosidase with the significantly downregulated expressions of LMNB1. Finally, we demonstrated that transplantation of the dermal substitute with trehalose-treated fibroblasts accelerated wound closure and increased capillary formation significantly in the experimental mouse wounds in vivo. These data indicate that high-concentration trehalose can induce the beneficial senescence-associated secretory phenotype in fibroblasts via CDKN1A/p21, which may be therapeutically useful for optimal wound repair.


Rapid spread of LSEs containing trehalose in the fibroblast-populated collagen gel
LSEs have been used to treat skin defects. However, production of LSEs takes approximately 4 127 weeks, which makes LSE production impractical for applications in regenerative medicine. To 128 investigate the beneficial effects of trehalose on fibroblasts, we constructed fibroblast-populated 129 type I collagen gel with trehalose, upon which normal human keratinocytes were seeded to form 130 LSEs. The sizes of the LSEs were observed after 2 weeks of airlifting at 37°C ( Figure 1A). We week after air exposure (Figure 1-figure supplement 1, A to C). 136 Two weeks after airlifting, hematoxylin and eosin staining was used to compare the LSEs containing 137 trehalose (10 or 100 mg/ml) with the control LSEs. Interestingly, they were morphologically 138 indistinguishable besides the size of the final products ( Figure 1D). Next, paraffin-embedded 139 sections of LSEs were subjected to immunohistochemistry with the Ki67 antibody to assess the 140 proliferation of fibroblasts. Conversely, fibroblasts in the collagen gel with trehalose showed 141 increased Ki67 positivity and proliferative capacity ( Figure 1D). Additionally, we examined elastic 142 and collagen fibers in the three-dimensional culture system with or without trehalose by Elastica  To investigate the novel effect of trehalose further, we prepared larger LSEs using a larger culture 157 insert (75-mm diameter), with proportionally more fibroblasts and keratinocytes. A rubber ring (8-158 mm interior diameter) was covered over the fibroblast-containing gel to stabilize it, and 159 keratinocytes were seeded in the ring hole. The epidermal layer of LSEs containing trehalose (100 160 mg/ml) in the gel harvested after 2-week airlifting at 37°C spread markedly, and thus, the 161 experimental LSEs were substantially larger than the control LSEs under the same conditions 162 (Figure 1-figure supplement 3). 163 To examine the signaling pathways modulated by trehalose treatment in the 3D skin model, we  Figure 1F). In addition, the factor loadings of genes in PC1 173 and PC2 showed no effect on trehalose treatment. Our observations indicated that gene expression 174 profiles in a long culture of 3D gels are similar for untreated and trehalose-treated cells in both 175 keratinocytes and fibroblasts. Next, we examined whether fibroblast pretreated with trehalose before 176 seeding in the collagen gel achieved increased proliferation of the epidermal layer of the 3D culture 177 model ( Figure 1G). Interestingly, LSEs with fibroblast pretreated with trehalose (30 or 100 mg/ml) 178 accelerated the spread of the epidermal layer compared with the control LSEs ( Figure 1H). These 179 observations indicated that trehalose pretreatment on the fibroblast monolayer before seeding in the 180 gel can induce significantly accelerated proliferation of the keratinocyte layer of LSE.  Figure 2C). With respect to upstream factors, asparaginase, a drug for 208 acute lymphoblastic leukemia, was detected in the upstream analysis and is able to arrest the cell 209 cycle. ZBTB17 is a transcriptional negative regulator in the cell cycle ( Figure 2D). CDKN1A was 210 also detected as an upstream factor based on the significant decrease of PLK1, CDK1, CCNA2, and 211 CDC25A after trehalose treatment ( Figure 2E). The inhibition of CDK1 and CCNA2 was suggested  Figure 3A). 220 The graphical summary connected the network analysis to the cellular functions and showed that 221 9 the senescence cells were activated by p53 and CDKN1A related to the cell cycle regulation and 222 mitosis arrest ( Figure 3B). In the network analysis, activation of DPT and VEGF were suggested to 223 be induced by the Notch and Caspase (Figure 3-figure supplement 1, A and B). The mRNA 224 expressions of genes involving cellular senescence, CDKN1A, and LMNB1 were confirmed by 225 quantitative PCR (qPCR) in 2D fibroblasts ( Figure 4A). Western blot analysis revealed that 226 trehalose treatment increased p21 expression and decreased lamin B in a dose-dependent manner 227 ( Figure 4B). The dose-dependently increased expression of p21 in the nuclei after trehalose 228 treatment was confirmed using fluorescence microscopy ( Figure 4C and   236 To further explore the effects of highly concentrated trehalose, we studied morphological alterations 237 of fibroblasts after trehalose treatment. Phase contrast microscopy revealed dose-dependent 238 morphological differences between fibroblasts cultured with or without trehalose. We observed that 239 the shape of cells cultured with trehalose remained polygonal/expanded, although the control cells 240 became fusiform/elongated ( Figure 5A and Video 1-3). In Figure 5A and the video 1-3, we 241 demonstrate that trehalose inhibited the population growth of monolayer fibroblast cells. Further, to 242 clarify trehalose-induced cell proliferation inhibition, we examined cell viability via CCK8 assay.  performed; this revealed a significant increase in the 3D fibroblasts that were added in the collagen 291 gel with trehalose (100 mg/ml) for 72 h compared with that of the control fibroblasts ( Figure 6D).

292
CDKN1A is involved in the upregulation of SASP factor genes 293 Next, we investigated the involvement of CDKN1A in trehalose-induced SASP in 2D fibroblasts. The SASP is closely associated with positive and negative outcomes depending on cell types and 301 contexts. Senescent cells also generate proinflammatory molecules and matrix metalloproteinases.

302
Next, we investigated whether trehalose-induced cellular senescence in monolayer and organotypic 303 cultures of human fibroblasts would lead to similarly dramatic changes in senescence factors 304 induced by other stressors. Surprisingly, analysis of RNA-seq data revealed that SASP factor genes 305 related to inflammation, such as IL-6, IL-8, and IL-1B, were not elevated, whereas several genes 306 associated with senescence, such as GDF15, MMP3, and TNFRSF10C, were upregulated (Table 1). 307 Overall, our data support the hypothesis that trehalose elicits the non-inflammatory SASP.  (Figure 8, F and G). Therefore, the dermal substitute with trehalose-treated fibroblasts     Previous studies demonstrated that epidermal growth factor (EGF) family members-transforming 415 growth factor (TGF)-α, heparin-binding (HB)-EGF, and EREG-act as autocrine growth factors for 416 normal human keratinocytes (Shirakata, 2010). We also observed dramatic upregulation of EREG 417 after trehalose treatment. EREG is upregulated in the psoriatic epidermis and was initially purified 418 from the mouse NIH-3T3 (Shirakata, 2010). We also demonstrated that trehalose treatment induced  The method used for LSE preparation was described previously (Yang et al., 2005). Briefly, a 498 collagen gel was prepared by mixing six volumes of ice-cold porcine collagen type I solution (Nitta 499 Gelatin, Osaka, Japan) with one volume of 8 × DMEM, 10 volumes of 1 × DMEM supplemented 500 with 20% FBS, and one volume of 0.1 N NaOH, of which 1 ml was added to each culture insert 501 (Transwell, 3-µm membrane pore, Corning, Corning, NY) in a six-well culture plate (Corning).

502
Following polymerization of the gel in the inserts at 37°C, two volumes of fibroblast suspension 503 solution 5 × 10 5 cells/ml in 1 × DMEM supplemented with 10% FBS were added to eight volumes 504 of the collagen solution (thus, the final collagen concentration was 0.8 mg/ml). Then, 3.5 ml of the 505 fibroblast-containing collagen solution was applied to each insert. When the fibroblast-containing 506 gel polymerized, DMEM supplemented with 10% FBS and ascorbic acid (final concentration 50 507 ng/ml) was added with or without trehalose (in three concentrations: 10, 30, and 100 mg/ml). The 508 gel was submerged in culture for 5 days until the fibroblasts contracted the gel.

18
A larger LSE was constructed following the same method as previously described except using a 510 larger culture insert (Transwell, 75-mm diameter, 3-µm membrane pore, Corning), thus utilizing 511 proportionally more fibroblasts. A rubber ring (8-mm interior diameter) was covered over the 512 fibroblast-containing gel to stabilize it within the large-scale LSE. In the hole of the ring, 6 × 10 5 513 keratinocytes in 30 µl MCDB 153 type II were seeded. The keratinocytes were submerged in culture 514 for 2 days. When the keratinocytes reached confluence, the LSE was lifted to the air-liquid interface 515 and a cornification medium (a 1:1 mixture of Ham's F-12 and DMEM supplemented with 2% FBS 516 and other supplements, as described previously (Yang et al., 2005)) was added. The medium was 517 changed every other day.

518
To construct the conventional LSE, keratinocytes were seeded onto the contracted gel and then 519 submerged and airlifted as described above, except without the ring. The seeding cell density was 520 adjusted using rubber rings. Both LSE types were harvested 7 or 14 days after airlifting. For Medicine. Ten-week-old female BALB/cAJc1-nu nude mice (CLEA Japan, Tokyo, Japan) were 528 anesthetized by isoflurane inhalation. Full-thickness wounds were created on the skin of the backs 529 of each mouse using a 6-mm skin biopsy punch. The fibroblast-containing collagen gels were 530 prepared with vehicle or trehalose (100 mg/ml) and submerged in culture for 5 days, and the dermal 531 substitutes (1 day after airlift) were grafted onto the wounds, which were covered with transparent 532 films. Seven days after transplantation, the grafts were harvested. One part of each graft was 533 paraffin-embedded and sectioned at 6 µm, from which hematoxylin and eosin staining was prepared.   Tokyo, Japan). Images were obtained using Nikon ECLIPSE E600 microscope coupled with Nikon 566 DS-Ri1camera (Nikon, Tokyo, Japan).

567
Evaluating the epidermal spreading potential of trehalose-treated LSEs 568 Rubber rings with an inner diameter of 8 mm were put on gels with or without trehalose in the gel 569 (10, 30, and 100 mg/ml) and 6 × 10 5 keratinocytes in 30 µl MCDB 153 II medium were seeded into 570 each ring hole. When the keratinocytes reached confluence, the LSEs were lifted to air-liquid 571 surface, and the rubber rings were removed. At 14 days after airlifting, the epidermal size was 572 20 measured using computer-assisted morphometric analysis. The epidermal sizes of the conventional 573 LSEs and trehalose-treated LSEs were compared statistically using Student's t-test. 574 SA-βgal assay, p21 and dihydroethidium (DHE) immunocytochemistry 575 SA-βAgal assay was performed by seeding fibroblasts onto eight-well chamber slides and treating 576 with trehalose or vehicle. Cells were also treated with adenovirus vectors that encode lacZ (Ax 577 LacZ), following a previously described method . Cells were fixed and

597
Single-cell preparations from the monolayer fibroblasts or the dermal side of LSE were carried out.

598
The dermis was removed from the epidermis of the LSE cells. The dermis was further digested with 599 collagenase XI and hyaluronidase (both from Sigma-Aldrich) for 120 min followed by fluorescence-

603
Total RNA was isolated by using the RNeasy Mini Kit (Qiagen), and real-time PCR was used to 604 determine the mRNA abundance, as described previously (Dai et al., 2011). TaqMan™ Gene 605 Expression Assays (Thermo Fisher Scientific) were used to analyze the expressions (Table S2).

606
GAPDH mRNA was used as an internal control. Target gene mRNA expression was calculated 607 relative to GAPDH mRNA, and all data are presented as normalized data compared to each control 608 (mean of control cells or tissues).   Competing interests: Authors declare that they have no competing interests.

642
Data and materials availability: All data are available in the paper or the supplementary materials.

643
RNA sequence data are submitted to GEO under accession number GSE184892.

644
The following data sets were generated.