Life Sciences Discovery and Technology Highlights

Wound Healing

Aligned Carbon Nanotubes Reduce Hypertrophic Scar via Regulating Cell Behavior

Hypertrophic scars arise due to excessive skin tissue fibrosis, impacting >100 million patients globally every year. Current treatment options, including surgery, dressings, and drugs, are considered unsatisfactory. Even though they possess significant potential to direct growth and regeneration, aligned structure materials have not been intentionally employed for abnormal scar therapy. Promisingly, they appear to promote quiescent keratinocytes, which may make them suitable for cardiac regeneration and scar prevention.

Aligned structure materials, such as nanotubes, have also proven difficult to mass-produce and are often employed as coatings in limited quantities. The authors of this article propose a bulk method to circumvent production limitations. Aligned carbon nanotube (ACNT) films are dry-drawn from spinnable CNT arrays prepared via chemical vapor deposition (CVD) and characterized to have 8 nm diameter with an average distance of 300 nm between individual nanotube strands.

Various dermal (fibroblasts), endothelial, smooth muscle, and preosteoblastic cells are seeded on the ACNTs, which retard their proliferation rates. ACNTs suppress cell growth for reasons apart from cell toxicity. ACNTs also contribute to aligning fibroblasts as well as elongation into a spindle-like shape. The films also result in the directional growth of cells.

ACNTs are then transferred onto an in vivo model—by application to fresh rabbit ear wounds. ACNTs increase wound closure time by ~2 days (not significant), but more importantly, they reduce the extent of scar elevation. Further morphological observations suggest that ACNT-treated wounds result in thinner and more organized fibrous tissue with reduced collagen I. Further whole-genome studies of ACNT-cultured fibroblasts also show that collagen binding, extracellular matrix components, cell–substrate adhesion, and epithelial cell proliferation experience major downregulation.

Gene ontology analysis suggests that the main culprit mechanisms include TGF-β, cell proliferation, and focal adhesions. This affirms ACNT’s suitability as a candidate for the regulation of abnormal scarring through topographical means without the use of drugs. (Weng, W.; et al. ACS Nano 2018, 12, 7601–7612).

Near-Infrared Fluorescence Probes to Detect Reactive Oxygen Species for Keloid Diagnosis

Reactive oxygen and nitrogen species (RONS) are essential to regulate normal physiological processes, while abnormally high levels are associated with pathogenesis. While high RONS expression is associated with various cancers, inflammation, cardiovascular disease, and fibrosis, it is less associated with wound healing and keloid scars. Keloids are a result of aberrant wound healing that results in excessive, invasive fibrous tissue that causes itch, pain, and psychological and physical limb impairment. Currently, surgery, radiation, and injections are used in combination to manage the lesions.

Abnormally high RONS could affect various processes, including extracellular matrix (ECM) deposition, cell migration, and adhesion. Thus, understanding the process of RONS production in keloid pathogenesis and detecting its presence could allow clinicians to intervene at earlier time points.

The authors of this report exploit the advantages of rapid detection, high throughput, sensitivity, and instrumentation compatibility to synthesize near-infrared (NIR) probes to diagnose RONS in keloid cells. The authors synthesize two different RONS probes from NIR dyes. The first component, CyTF, undergoes oxidation in the presence of ONOO–, releasing caged oxygen. The second component, CyBA, is immolative in the presence of strong Lewis bases like ONOO– and H₂O₂. Both are initially nonfluorescent before RONS generates fluorescent CyOH.

The probes display up to a 15-fold increase at 717 nm; however, CyTF is found to have greater selectivity in response to ONOO– and H₂O₂. Cell imaging shows that a 2.6- to 3.76-fold signal enhancement is observed when transforming growth factor (TGF)-β1 is added. Conversely, the signal returns to basal levels upon the addition of N-acetyl-l-cysteine (NAC). This suggests that the addition of TGF-β1 increases intracellular ONOO–, thereby working as an inducer of oxidative stress.

Similarly, in keloid-derived fibroblasts, the NIR signal is similarly elevated compared with normal fibroblasts. Further treatment with inhibitor molecules causes signal regression, further validating the role of oxidative stress in keloid pathogenesis and ONOO– as a biomarker.

Thereafter, the ability to distinguish keloid cells is evaluated on a mouse model by subcutaneous implantation. The injected CyTF probe gradually increases from t = 1 h and reaches maximum signal output at 3.5 h. Its intensity is threefold higher in keloid compared with control cells. This signal also returns to basal levels at 6 h, suggesting that the probe is cleared from the diseased spot over time. Thus, ONOO– appears to play a valuable role in keloid pathogenesis and feasibility is demonstrated in using optical probes to diagnose keloids. (Cheng, P.; et al. Chem. Sci. 2018, 9, 6340–6347).

Reduction of Fibrosis and Scar Formation by Partial Reprogramming In Vivo

Cell reprogramming is renowned for transforming terminally differentiated cells into precursor/stem cell states that run counter to normal development. This has led to significant use in many areas of biomedicine. A development from this discovery is partial cell reprogramming through short-term activation of OSKM (Oct4, Sox2, Klf4, C-myc) transcripts. This has been shown to reduce senescence (biological aging) in progeria mice, enhance muscle regeneration, regenerate pancreatic tissue, and heal traumatic brain injury. Wound repair in adult mammals is typically characterized by three phases: inflammation, new tissue formation (cell proliferation and migration), and tissue remodeling.

Often, wound regeneration results in scar formation characterized by myofibroblasts, that is, activated fibroblasts that release inflammatory/profibrotic cytokines and cause wound contraction and support collagen-rich extracellular matrix (ECM). Myofibroblasts are the main culprits of abnormal scar formation (e.g., hypertrophic scars), which are characterized by the overexpression of α-smooth muscle actin (α-SMA) consisting of fibroblasts, disorganized ECM, and tightly packed parallel-oriented collagen bundles. They are also responsible for increased angiogenic response characterized by vascular endothelial growth factor (VEGF).

This state of mammalian adult scarring is responsible for significant clinical burden, including impaired growth and restricted joint mobility. In other organs (e.g., liver, lung, and kidney), these mechanisms center on transforming growth factor (TGF)-β1 stimulation and myofibroblasts that similarly give rise to organ fibrosis. Interestingly, TGF-β1-driven wound healing only occurs in late-gestating and adult mice, whereas earlier stages of development correspond with scarless wound healing. Unlike mammalian wound healing in adults, scarless healing is characterized by increased TGF-β3 with increased type 3 collagen and interleukin 10. The authors hypothesize that partial OSKM-driven cell reprogramming may give rise to improved wound closure dynamics that lead to reduced scar formation.

To prove their hypothesis, OSKM mice activatible through doxycycline (Dox) treatment are obtained for in vitro and in vivo studies. While cell reprogramming is highly desirable, the authors wanted to ensure that only partial reprogramming is achieved to avoid the risk to teratoma formation. Following a 10-day Dox treatment (to activate OSKM factors), the wound size is reduced by 33% compared with a 94% reduction in control wounds. Dox application in wild-type mice has no discernible effect.

By day 25, wounds from different treatment groups had all closed (control mice closed by day 15), whereas full stem cell reprogramming had not yet occurred (due to the absence of Nanog gene expression). This is further supported by the absence of tumor (or teratoma) formation.

In Dox-activated wounds, the genes TGF-β1, type 1 collagen, α-SMA, and VEGF—markers for fibrosis—are significantly reduced compared with control wounds. This is confirmed by further immunocytochemistry suggesting that OSKM induction reduces fibroblast-to-myofibroblast transdifferentiation at wound sites.

Since OSKM activation does not significantly affect skin reepthelialization, the decelerated wound healing is due to reduced wound contraction during the healing process. Further gene analysis finds that even though fibrotic activity is suppressed, reepithelialization is not impaired, nor is the expression of senescence genes. Further examination of OSKM activation in vitro shows decreased closure of cell gap widths after scratch assays. This is accompanied by reduction in TGF-β1 and α-SMA expression. Finally, OSKM activation is performed on incisional wounds in mice. These wounds are then sutured, which results in longitudinal scars. Hematoxylin and eosin (H&E) staining (tissue and nuclei morphology) and picosirus red (collagen) staining show the appearance of reduced scar tissue (dermis) formation (morphology). Scar width is found to be reduced by 57% compared with control wounds. Crucially, the implication that OSKM transgene activation reduces fibroblast-to-myofibroblast transdifferentiation suggests that cell reprogramming processes can be utilized in a therapeutic manner. This could involve treating diseases with fibrosis mechanisms such as hypertrophic scars, liver cirrhosis, and pulmonary and renal fibrosis. (Doeser, M. C.; et al. Stem Cells 2018, 36, 1216–1225).

Nonlinear Optical Microscopy and Histological Analysis of Collagen, Elastin, and Lysyl Oxidase Expression in Breast Capsular Contracture

Silicone breast implants used in aesthetic and reconstructive breast surgery induce a cascade of foreign body responses involving an inflammatory reaction followed by fibroblast recruitment, which secretes collagen and myofibroblasts that generate α-smooth muscle actin (α-SMA)-positive stress fibers. This leads to the formation of a collagenous capsule around the implant. Prolonged foreign body response leads to excessive fibrous tissue accumulation, which leads to discomfort, pain, deformity, and breast distortion. It is also the leading cause of patient dissatisfaction following implant surgery. These contractures are typically classified by the Baker classification system, ranging from a normal, soft, and nonpalpable implant to a painful and hard breast with distortion. The fibrous capsules comprise collagen and elastin fibers cross-linked by lysyl oxidase (LOX), which initiates the process.

This study seeks to elucidate the role of LOX in capsular contractures. Baker grade 3 capsules have abundant type 1 and 2 collagen and elastin. Type 1 collagen is highly expressed at the implant–tissue interface. In grade 4 capsules, high heterogeneity is observed in the capsules. Type 1 and 2 collagen is expressed to different extents. Elastin can be found at the outer and inner layers of the sample and is predominantly present at the capsule inner layer. LOX is found to be expressed at a higher intensity at the inner capsule layer. All Baker 3 and 4 classification capsules consist largely of collagen cross-linked by LOX.

Using nonlinear optical imaging, second harmonic generation (SHG) microscopy is able to reveal type 1 and 2 collagen structures. Two-photon excitation fluorescence (TPEF) microscopy can also detect elastin label-free. Surprisingly, TPEF and immunostaining reveal that Baker grade 3 capsules have a higher content of type 1 and 2 collagen and elastin than Baker grade 4 capsules. While grade 3 capsules show type 1 collagen at the tissue–implant interface, it is almost absent from those of grade 4. LOX is present throughout the tissues, with the highest expression at the tissue–implant interface.

This suggests that collagen composition changes dynamically as the extracellular matrix (ECM) tissue (collagen and/or elastin) undergoes significant cross-linking to form cross-linked mature collagen fibrils that are less susceptible to degradation.

Interestingly, myofibroblasts that are often responsible for matrix contraction are absent in Baker grade 3 and 4 capsules. This may be a result of myofibroblasts undergoing apoptosis due to overexpressed LOX that has been known to be responsible for increased tissue stiffness in other contexts.

In conclusion, LOX very likely plays a role in the pathology of fibrous capsule development around the silicone implant. Nonlinear microscopy is a useful tool to perform label-free assessment of the ECM contents in such capsules. However, further studies may reveal crucial information about the mode of collagen cross-linking that may lead to changes in the number of enzymatic cross-links. This is a critical criterion for irreversible fibrosis events. (Poh, P.; et al. Eur. J. Med. Res. 2018, 23, 30).

Sensing of Vimentin mRNA in 2D and 3D Models of Wounded Skin Using DNA-Coated Gold Nanoparticles

Wound healing occurs upon the onset of injury with the goal to regenerate tissue promptly to prevent infections. It is a complex process regulated by specific factors that control cell proliferation, alignment, and organogenesis. While staining fixed tissue to assess cell biomarkers is commonly performed, the detection of protein expression dynamics is important yet challenging to achieve. Monitoring expression dynamics in tissue is also achievable via mRNA. The available range of mRNA labeling techniques (e.g., fluorescence in situ hybridization, molecular beacons, aptamers, and gene transfection) requires tissue fixation, which limits the amount of real-time information that can be obtained. Recent work shows how oligonucleotide nanoparticles can penetrate skin tissue. These have been loaded with gene-regulating or -detecting sequences for detection specificity.

Vimentin is a critical factor involved in the epithelial mesenchymal transition process during wound healing. To enable live imaging, NanoFlares—oligonucleotide nanoparticles with reporter and recognition elements—are applied to monitor vimentin mRNA in an open wound. Using cell lines (human, murine origin), the authors show that NanoFlares successfully detect vimentin expression, particularly in a scratch assay. At the edges of the scratch, confluent epithelial cells increase NanoFlare signals by 10-fold. NanoFlares are then used to detect vimentin in damaged mouse skin tissue. By injecting NanoFlares through the skin, proximal tissue uptakes these at various layers of skin (i.e., epidermis, dermis, and subcutaneous tissues). Similarly, fixed tissue sections show that the highest NanoFlare signal is generated at the wound edge in mouse skin. In contrast, noncoding NanoFlares express minimal signal.

Unfortunately, this method does not give rise to true data in real time. Thus, the authors seek to achieve this using light-sheet fluorescence microscopy (LSFM), which produces an optical sectioning image stack. In addition to NanoFlares for vimentin, fluorescence probes are used for signal normalization. Eventually, these allow the authors to track the steady increase of vimentin over 6 h in live wounded tissue. Thus, this study suggests that NanoFlares present a very promising technique to monitor gene expression changes in live tissue. DNA versatility means that this method could be readily applied to other critical genes by modifying the recognition and reporter sequences. (Vilela, P.; et al. Small 2018, 22, 1703489).

Nanotechnology and Stem Cells

PLGA Spherical Nucleic Acids

Spherical nucleic acids (SNAs) are a class of nanomaterials invented by the Mirkin group (Northwestern University, Evanston, IL) to circumvent the limitations of nucleic acids in biology. SNAs exploit cellular scavenger receptors to deliver nucleic acids since linear nucleic acids cannot enter cells without assistance. These qualities have made SNAs emerge as a powerful platform for developing diagnostic probes, compounds for gene regulation, and immunomodulation therapy. Recent iterations have given rise to wholly biocompatible SNAs by replacing the gold nanoparticle core with liposomes, proteins, polymeric micelles, organic nanoparticles, and so forth.

Poly(lactic-co-glycolc acid) (PLGA) is an attractive material in biomedical applications because it has tunable release kinetics of encapsulated cargoes and is biocompatible and biodegradable. Following synthesis, characterization shows the particle size increases from 50 nm (unmodified PLGA core) to 65 nm (nucleic acid-modified particle). Electrophoretic mobility shift assays and atomic force microscope imaging confirm nucleic acid–polymer binding. The particle size and absolute number of DNA strands is quantified by fluorescence spectroscopy. Each particle contains ~200 strands, which is a surface density in between gold or liposomal nanoparticles (previous iterations of nanomaterials).

While typical drug carriers are unable to carry both hydrophobic and hydrophilic drugs, PLGA-SNAs both carry nucleic acids (outer shell) and compartmentalize hydrophobic drugs in their core. A fluorescence turn-on experiment is performed to evaluate nucleic acid release from PLGA-SNAs when immersed in serum. The stability outperforms liposomal-SNAs already used clinically. PLGA-SNAs also exhibit improved stability against DNase I degradation compared with linear DNA. Having established various properties of the PLGA-SNA constructs, the authors then switch their focus to applications.

The nanoconstruct’s sustained release of Coumarin 6 acts as a model drug. Among various polymer compositions, RG 502 has a significantly higher release rate because of its smaller molecular weight. They also readily enter Raw-Blue macrophage cell lines at a 10-fold rate compared with linear nucleic acids. Yet, increasing concentrations of PLGA-SNAs do not diminish cell viability. Finally, they also demonstrate potential for immunotherapy, successfully delivering CpG motifs that activate TLR9 to a greater extent at lower concentrations compared with linear or control GpC constructs.

In summary, PLGA-SNAs are generated using an extremely facile strategy to prepare SNAs for diagnostics and therapeutics without the laborious methods employed for their liposomal counterparts. Promisingly, their drug-release properties (from the PLGA core) can be independently tuned without altering the stability and release properties of the nucleic acid shell. Therefore, novel PLGA-SNAs hold great promise for biomedicine. (Zhu, S.; et al. Adv. Mater. 2018, 22, 1707113).

Nondestructive/Noninvasive Imaging Evaluation of Cellular Differentiation Progression during In Vitro Mesenchymal Stem Cell-Derived Chondrogenesis

Tissue-engineered cartilage derived from stem cells has yet to fulfill its promise for cartilage repair. In numerous instances, the tissue has inferior functional properties compared with native articular cartilage. There is a need to improve the quality of such tissues. One mitigation strategy is to develop technologies to predict outcomes early on to limit the number of constructs that result in poorly performing implants. To achieve this, molecular factors can be monitored to assess the likelihood of failure and the acquisition of desirable differentiation traits. The authors of this article report an approach that allows spatiotemporal monitoring of chondrogenic cell differentiation by imaging particular genetic events in order to predict differentiation success or failure and therefore define/test release criteria of the stem cell implants.

Critically, mesenchymal precursors (e.g., mesenchymal stem cells [MSCs]) undergo a multistep program involving Sex Determining Region Y-Box 9 (Sox9) genes. This leads to additional extracellular matrix (ECM) fabrication defined by the assembly of various components, including proteoglycans—aggrecan (AGG) and collagen fibers (Col2) with highly specific three-dimensionality. This glycosaminoglycan (GAG) ECM is critical to the tissue’s tensile and compressive properties. These attributes can be difficult to reproduce in vitro, however, and they are further complicated by the tendency of MSCs to undergo endochondral bone formation. This ultimately develops into a cartilaginous ECM that is different from permanent hyaline articular cartilage. Evaluation of tissue-engineered cartilage tissue has typically relied on a set of gene markers evaluated at arbitrarily determined time points. Being averaged from multiple samples masks the high inherent variability of engineered tissues, thus raising issues of implant reproducibility.

To circumvent such issues, the authors use molecular probes of known chondrogenic and osteogenic biomarkers. These are designed to nondestructively and noninvasively monitor differentiation status by acquiring the bioluminescence imaging (BLI) signal of gene promoter activity.

Using fibroblast growth factor 2 (FGF2) growth factors applied to hMSCs, the authors noninvasively acquire increased Sox9 activity through BLI during early stages (day 5) and confirm this with increased chondrogenic tissue pellets later on (day 21).

The authors then resort to viral-mediated gene reporters for steady transgene expression beyond 7 days. Moreover, the stably expressed probes are shown not to affect MSC chondrogenic capacity. Thereafter, the Sox9 and AGG BLI probes are used to track chondrogenesis in control and FGF2-stimulated MSCs. These show a steadily rising trend over a period of 28 days, which corresponds with microarray gene analysis. In contrast, BLI probes detect underexpressed AGG when FGF9 is applied and compared with control cultures. This is similarly reflected in FGF9-treated cell pellets with minimal Safranin O staining at day 28. This suggests the likelihood of identifying underdeveloped pellets at earlier stages. In addition to Sox9 and AGG probes, osteocalcin (OC)—an osteogenic differentiation biomarker—is incorporated to discriminate MSC differentiation trajectory. Interestingly, BLI probes detect basal OC expression in MSCs, which declines before it rises on the fourth week. This rise is interpreted to be a result of terminal hypertrophic differentiation.

Finally, the authors demonstrate that BLI imaging can identify bone formation and terminal hypertrophic differentiation using the OC probes. Primarily, these BLI-based gene probes offer the advantages of avoiding endpoint/destructive assessments, making spatiotemporal correlations between interventions and resulting phenotypes, on-the-go process adaptations, and a defined release criterion for implantation. Crucially, these can be observed in cells as well as small animal subjects. (Correa, D.; et al. Tissue Eng. Part A 2018, 7–8, 662–671).

Dual Bioluminescence and Near-Infrared Fluorescence Monitoring to Evaluate Spherical Nucleic Acid Nanoconjugate Activity In Vivo

While significant progress has been made in identifying novel genes relevant to cancer progression, many cancer genes are nonenzymatic targets with unknown mechanisms. RNA silencing technology has been developed to deliver oligonucleotide payloads to tumor sites. The problem is that potent gene silencing activity in vitro is often ineffective when transferred in vivo to animal subjects. Considerations that impede effective translation include serum factors, renal filtration, liver modification, and an inability to cross biological barriers. For improved preclinical evaluation of gene silencing-based therapeutics, animal models need to predict and confirm functionality of the oligonucleotide constructs as well as be rigorously used to optimize their administration and dosing schedules.

The authors have developed spherical nucleic acids (SNAs) that consist of a gold nanoparticle core functionalized with a shell of radially aligned siRNA oligonucleotides to silence oncogenes in glioblastoma (GBM). GBM is an incurable brain cancer with survival rates of 14–16 months, even after attempts to manage it (surgery, radiation, and temozolomide [TMZ] treatment).

SNAs have been shown to have blood–brain barrier crossing ability in in vitro and in vivo models. In animal subjects, this is enabled following systemic IV administration to animal subjects with GBM tumors. SNAs functionalized to silence glioma oncogenes (e.g., Bcl2L12) or with microRNA (e.g., miR-182) can potentially treat GBM. This current study demonstrates how intracranial GBM xenografts with stable luciferase and near-infrared fluorescent protein fuse to O⁶-methylguanine-DNA methyltransferase (MGMT). This model enables longitudinal and noninvasive evaluation of MGMT gene knockdown in response to SNA treatment. Lower MGMT protein has been linked with prolonged survival in GBM patients.

To develop GBM reporting cells, the authors generate an MGMT-iRFP670 construct that is transfected into GBM cells. The cells are sorted by flow cytometry before being evaluated by Western blotting. This reporter lasts for up to 6 months, generating robust NIR expression in vivo (intracranial) and tumors excised ex vivo. MGMT silencing SNAs (siMGMT-SNAs) are then evaluated against noncoding control SNAs in GBM cells, which confirms that they efficiently reduce endogenous and exogenous MGMT expression. To evaluate effective knockdown of MGMT, in vivo models are created with labeled GBM cells injected into mouse brains.

Following tail-vein IV administration, MGMT is found to be diminished by 60% (at 24 h) before restoration by the 96 h time point. This demonstrates the successful tracking of MGMT genetic expression in vivo. Through reporter mice, the authors ascertain that a dosage of 1.4 mg/kg is sufficient to generate MGMT knockdown in the reporter mice. This dosage, administered at 48 h intervals together with TMZ (after 9 hs), leads to the greatest bioluminescence decrease in the GBM mouse models.

Critically, such animal model reporter systems are a suitable proof of concept to demonstrate effective knockdown of target oncogenes in vivo. Through animal imaging, one can readily and noninvasively monitor the efficacy of oncogene silencing SNAs. Furthermore, this allows us to design combination treatments (e.g., SNAs in addition to drugs or otherwise) for the most effective therapy in difficult-to-treat cancer, like GBM. (Sita, T.; et al. Proc. Natl. Acad. Sci. U.S.A. 2017, 114, 4129–4134).

Kynurenic Acid, an IDO Metabolite, Controls TSG-6-Mediated Immunosuppression of Human Mesenchymal Stem Cells

A number of molecules, such as IDO and TSG-6, are responsible for the immunomodulatory effects of mesenchymal stem cells (MSCs). Some scientists argue that any variability arises from their microenvironmental context. Previous studies have demonstrated that indoleamine 2,3-dioxygenase (IDO) is indispensable for these effects. IDO is critical in tryptophan catabolism, influencing downstream metabolites like kynurenine (KYN). These further inhibit effector T-cell proliferation, but also induce regulatory T-cell differentiation. TSG-6 is another crucial glycoprotein involved in MSC-mediated tissue repair. It antagonizes CXCL8 chemotaxis by neutrophils and inhibits leukocyte (neutrophils, macrophages) extravasation at inflammation sites.

Despite some cursory knowledge of the involvement of these factors, their relationship and function in MSC immune regulation is unclear. In this article, the authors uncover a novel link between IDO and TSG-6 in human MSCs that promotes more effective clinical cell therapy. The authors first generate acute lung injury (ALI) with lipopolysaccharide (LPS) in BALB/c mice through intranasal delivery. Upon the collection of bronchoalveolar lavage (BAL) fluid, significant increases in the number of cells and neutrophils are observed. The introduction of MSCs (2.5 × 10⁵) leads to the significant suppression of both. However, MSCs with attenuated IDO (IDO knockdown [KD]) activity are not able to suppress their numbers. This is further confirmed by high quantities of Gr-1 staining (a neutrophil biomarker) in LPS-infected and LPS-infected mice treated with IDO-KD MSCs.

Following further analysis of IDO-KD MSCs, TSG-6 is similarly decreased compared with untreated MSCs. Other factors, such as fibroblast growth factor 7 (FGF7 or KGF), angiopoietin 1 (ANG-1), and interleukin 1 receptor antagonist (IL-1Rα), are not affected by IDO suppression in MSCs. To uncover the role of TSG-6, recombinant TSG-6 is applied to the ALI mouse model co-injected with IDO-KD MSCs. This completely restores MSC immunosuppressive function. Using a peritonitis model induced by zymosan, knockdown of IDO or TSG-6 is similarly sufficient to compromise immunosuppression activity. On the other hand, TSG-6 absence does not affect IDO expression. In fact, TSG-6-KD MSCs still significantly suppress T-cell proliferation. Further exploration demonstrates that reducing the enzymatic activity of IDO is sufficient to reduce TSG-6. Analysis of tryptophan metabolism reveals that kynurenic acid (KYNA) is strongly expressed in MSCs and KYNA treatment leads to increased TSG-6 expression compared with untreated cells. Thereafter, KYNA-treated MSCs (KYNA-MSCs) are reapplied back to the ALI mouse models. As expected, KYNA-MSCs significantly decrease infiltrated cells and neutrophils through increased TSG-6 production. This is further confirmed by H&E and Gr-1 staining.

The authors further uncover how KYNA interacts with the aryl hydrocarbon receptor (AhR) that undergoes nuclear translocation to form a heterodimer with AhR nuclear translocation (ARNT). The ARNT activity then interacts with the TSG-6 promoter, reinforced by the addition of KYNA to regulate TSG-6 expression. Thus, this study successfully uncovers the interactions of IDO and the immunosuppressive TSG-6. Further study shows that it is the presence of KYNA metabolites that promotes MSC-mediated immunosuppression of inflammatory diseases. (Wang, G.; et al. Cell Death Differ. 2018, 25, 1209–1223).