Interleukin-1 Receptor Modulation Using β-Substituted α-Amino-γ-Lactam Peptides From Solid-Phase Synthesis and Diversification

As a key cytokine mediator of inflammation, interleukin-1β (IL-1β) binds to the IL-1 receptor (IL-1R) and activates various downstream signaling mediators, including NF-κB, which is required for immune vigilance and cellular protection. Toward the development of IL-1-targeting therapeutics which exhibit functional selectivity, the all-D-amino acid peptide 1 (101.10, H-D-Arg-D-Tyr-D-Thr-D-Val-D-Glu-D-Leu-D-Ala-NH2) was conceived as an allosteric IL-1R modulator that conserves NF-κB signaling while inhibiting other IL-1-activated pathways. Employing β-hydroxy-α-amino-γ-lactam (Hgl) stereoisomers to study the conformation about the Thr3 residue in 1, [(3R,4S)-Hgl3]-1 (2b), among all possible diastereomers, was found to exhibit identical in vitro and in vivo activity as the parent peptide and superior activity to the α-amino-γ-lactam (Agl) counterpart. Noting the relevance of the β-hydroxyl substituent and configuration for the activity of (3R,4S)-2b, fifteen different β-substituted-Agl3 analogs of 1 (e.g., 2c-q) have now been synthesized by a combination of solution- and solid-phase methods employing N-Fmoc-β-substituted-Agl3-Val-OH dipeptide building blocks. Introduction of a β-azido-Agl3 residue into the resin bound peptide and subsequent reduction and CuAAC chemistry gave access to a series of amine and triazole derivatives (e.g., 2h-q). β-Substituted-[Agl3]-1 analogs 2c-q exhibited generally similar circular dichroism (CD) spectra as that of Hgl analog 2b in water, presenting curve shapes indicative of β-turn structures. The relevance of the β-substituent was indicated in rodent models of preterm labor and retinopathy of prematurity (ROP), in which certain analogs inhibited preterm birth and vaso-obliteration, respectively, with activity similar to 1 and 2b. The β-substituted-[Agl3]-1 analogs exhibited functional selectivity on IL-1-induced signaling pathways. The described solid-phase method has provided discerning probes for exploring peptide structure-activity relationships and valuable leads for developing prototypes to treat inflammatory events leading to prematurity and retinopathy of prematurity, which are leading causes of infant morbidity and blindness respectively.


INTRODUCTION
Inflammatory factor expression is induced primarily through signaling pathways triggered by interleukin-1β (IL-1β) (Gabay et al., 2010). This major pro-inflammatory cytokine stimulates various physiological effects leading ultimately to hyperthermia, hypotension, tissue destruction, and inflammation (Dinarello, 1997). The activity of IL-1β is critical for inflammatory responses to treat damaged tissue and to ward off invading pathogens. Uncontrolled IL-1β activity is, however, a pathogenic characteristic of many chronic conditions.
In reproductive tissue during pregnancy, toll-like receptors (TLRs) can recognize and discriminate bacterial pathogenassociated molecular patterns (PAMPs) (Elovitz et al., 2003;Ilievski et al., 2007). The recognition of PAMPs during intraamniotic infection (e.g., chorioamnionitis) leads to up-regulation of TLRs (Kim et al., 2004) and release of pro-inflammatory IL-1β from immune cells. The latter plays a key role in the induction of both term and preterm labor (Romero et al., 1989). Engagement of TLRs also leads to the activation of nuclear factor kappalight-chain-enhancer of activated B cells (NF-κB), a transcription factor that is involved in the expression of cytokines, such as IL-1β, chemokines, and antimicrobial defensin peptides (Choi et al., 2004). Activation of NF-κB is essential for maintaining immune vigilance against invading pathogens.
Regulation of TLR expression plays a key role in ischemic diseases of the retina, such as retinopathy of prematurity (ROP) (Xu and Wang, 2016). Multiple cell types in the retina express TLRs, including glia, retinal pigment epithelium (RPE), photoreceptor, and endothelial cells. In the pathogenesis of retinal ischemic diseases, activation of TLRs initiates signal transduction, leading to production of pro-inflammatory cytokines, such as IL-1β (Rivera et al., 2010(Rivera et al., , 2017Xu and Wang, 2016;Beaudry-Richard et al., 2018). In response to hypoxia, retinal microglia cells produce IL-1β and trigger an inflammatory cascade involving IL-6 and IL-8 (Tosato and Jones, 1990). Endothelial cytotoxicity results from IL-1β-dependent retinal and sub-retinal injury in models of oxygen-induced retinopathy (OIR) (Tremblay et al., 2013;Zhou et al., 2016). Moreover, IL-1β has been linked to oxidative stress and is associated with retinal microvascular degeneration mediated by the proapoptotic guidance cues of semaphorin 3A (Sema3A) (Rivera et al., 2013).
Two natural inhibitors of IL-1β signaling are the intrinsic IL-1 receptor antagonist (IL-1Ra) and the decoy receptor, IL-1 receptor type II (IL-1RII) (Krumm et al., 2014). IL-1Ra competes with IL-1 for the IL-1R binding site and prevents recruitment of IL-1RAcP. IL-1RII sequesters IL-1 but cannot form a signaling complex (Gabay et al., 2010). Current therapeutic strategies to counter pathological IL-1 signaling have been based on the mechanisms of these natural proteins. The three FDA-approved peptide-based therapies comprise: (1) the recombinant IL-1 receptor antagonist, Kineret, (2) the IL-1 Trap composed of a dimeric fusion of the ligand-binding domains of the extracellular portions of IL-1R1 and IL-1RAcP proteins, Rilonacept, and (3), a human monoclonal antibody targeting IL-1β, Canakinumab (Kaneko et al., 2019). These relatively large proteins have presented undesirable secondary effects in clinical settings including immunosuppression, which increases the risk for opportunistic infections, and pain at the site of injection (Opal et al., 1997;Roerink et al., 2017). Their failures in clinical trials may be due in part to drawbacks related to acting directly on the native orthosteric ligand and indiscriminately interfering with all signals triggered by IL-1β (Opal et al., 1997;Roerink et al., 2017). Selective agents are desired to differentially target IL-1β signaling pathways leading to immune vigilance and inflammation. For example, allosteric ligands which bind remotely from the orthosteric ligand binding site of IL-1RI have been pursued to modulate IL-1β activity by inducing biased signaling. Such allosteric modulators may be smaller molecules exhibiting improved bioavailability, potential for oral administration, protease resistance, and lower risks of toxicity.
Fifteen new (3R,4S)-β-substituted-Agl 3 analogs of 1 (e.g., 2cq) were synthesized using the combination of the solutionand solid-phase synthesis protocols. The hydroxyl group of 2b has been replaced by a variety of functional groups with potential to contribute to hydrogen bonds as donors and acceptors, to salt bridges as protonated ammonium ions, and to π-cation interactions as aromatic donors. Furthermore, the novel β-substituents exhibited limited effect on conformation in comparative analyses of the curve shapes of (3R,4S)-Hgl analog 2b with those of the other (3R,4S)-β-substituted-Agl 3 analogs (2c-q) using circular dichroism spectroscopy.
The β-substituent on lactams 2c-q exhibited notable effects on inflammatory gene transcription (IL-1β and COX-2) and kinase phosphorylation in vitro in RAW 264.7 macrophages. In a reporter gene assay, the β-substituted-Agl analogs 2c-q behaved like peptide 1 and Hgl analog 2b and did not exhibit effects on IL-1β-induced NF-κB signaling. Based on preliminary in vitro analyses, certain analogs were selected for examination in in vivo rodent models of preterm labor (PTL) and oxygen-induced retinopathy (OIR). These investigations have identified a novel set of (3R,4S)-β-substituted-Agl 3 analogs that exhibited in vivo effects comparable to the (3R,4S)-Hgl counterpart 2b. In summary, this study provides valuable methods for studying structure-activity relationships of biologically active peptides and a deeper understanding of the relevance of lactam β-substituents on IL-1R modulator activity. SCHEME 1 | Synthesis of dipeptide 3f by ring opening of cyclic sulfamidate 4 and ester cleavage.

General Chemistry Methods
Unless otherwise specified, all non-aqueous reactions were performed under an inert argon atmosphere. All glassware was dried with a flame and a flushing stream of argon gas or stored in the oven and let cool under an inert atmosphere prior to use. Anhydrous solvents (THF, DCM, MeCN, MeOH, and DMF) were obtained by passage through solvent filtration systems (Glass Contour, Irvine, CA). Anhydrous solvents were transferred by syringe. Reaction mixture solutions (after aqueous workup) were dried over anhydrous MgSO 4 or Na 2 SO 4 , filtered, and rotary-evaporated under reduced pressure. Column chromatography was performed on 230-400 mesh silica gel, and thin-layer chromatography was performed on alumina plates coated with silica gel (Merck 60 F 254 plates). Visualization of the developed chromatogram was performed by UV absorbance or staining with iodine or potassium permanganate solutions. Specific rotations, [α] D values, were calculated from optical rotations measured at 25 • C in CHCl 3 at the specified concentrations (c in g/100 mL) in a 0.5 dm cell length (l) on a Anton Paar Polarimeter, MCP 200 at 589 nm, using the general formula: [α] D 25 = (100 × α)/(l × c). Nuclear magnetic resonance spectra ( 1 H NMR, 13 C NMR) were recorded on a Bruker 300 MHz spectrometer. 1 H NMR and decoupled 13 C[ 1 H] NMR spectra were measured in and referenced to CDCl 3 (7.26 ppm, 77.16 ppm) as specified below. Coupling constant J values were measured in Hertz (Hz) and chemical shift values in parts per million (ppm). High resolution mass spectrometry (HRMS) data were obtained in electrospray ionization (ESI-TOF) mode by the Centre Régional de Spectrométrie de Masse de l'Université de Montréal. Protonated molecular ions [M + H] + , and sodium adducts [M + Na] + were used for empirical formula confirmation. Analytical LCMS and HPLC analyses were performed on either a CSH-C18, 4.6 X100 mm, 5 µm column with a flow rate of 0.8 mL/min or CE-C18 3 X 50 mm, 2.7 µm column with a flow rate of 0.4 mL/min using appropriate gradients from X-Y% of MeOH [0.1% formic acid (FA)] or MeCN (0.1% FA) in H 2 O (0.1% FA) over 10 min: a) 10-90%, b) 50-90%, c) 30-60%, d) 5-60%, e) 30-90%, f) 20-40%.
All final peptides were purified using the respective conditions below on a Waters TM preparative HPLC instrument with UV detection at 214, 254, and 280 nm and one of the following systems: a reverse-phase Gemini TM C18 column (21.2 × 250 mm, 5 µm) using a flow rate of 10 mL/min over 40 min; a C18 Atlantis column (19 × 100 mm, 5 um) using a flow rate of 24 mL/min over 15 min; a RP-Polar column (19 × 100 mm, 4 µm) using a flow rate of 24 mL/min over 15 min.

NF-kB QUANTI-Blue Assay
HEK-Blue IL-33/IL-1β cells (InvivoGen) were pretreated with peptides 1 or 2 (10 −6 M), or Kineret (1.0 mg/mL) for 30 min, followed by treatment with a constant concentration of IL-1β (100 ng/mL), and incubation at 37 • C for 4 h. Levels of secreted alkaline phosphatase in cell culture supernatant were determined using the QUANTI-Blue assay, according to the manufacturer's instructions (InvivoGen). Alkaline phosphatase activity was assessed by measuring optical density (OD) at 620-655 nm with an EnVision Multilabel micro plate reader (PerkinElmer, Waltham, MA). Data are representative of 3 experiments (each with n = 4 per treatment group).

LPS-Induced Preterm Model in Mice
Timed-pregnant CD-1 mice at 16.5 days of gestation (G16.5) were anesthetized with 2% isoflurane and received an intraperitoneal injection of lipopolysaccharide (LPS, n = 4 TABLE 1 | List of primers for the human genes assessed by qPCR.

PTGHS2
ATATTGGTGACCCGTGGAGC GTTCTCCGTACCTTCACCCC per group, a single dose of 10 µg) (Kakinuma et al., 1997;Nadeau-Vallée et al., 2015). A dosage of 2 mg/kg/day of peptides 1 or 2 or vehicle was respectively injected subcutaneously in the neck, every 12 h until delivery. On G16.5, a dose of 1 mg/kg was injected 30 min before stimulation with LPS (to allow distribution of drugs to target tissues) and 1 mg/kg was injected 12 h after stimulation (n = 4 each treatment). Mice delivery was assessed every hour until term (G19-G19.5). A mouse was considered as delivering prematurely if the first pup was delivered earlier than G18.5. Data was analyzed using Prism 7 (GraphPad Software, San Diego, CA, USA) with one-way ANOVA and Dunnett's test for multiple comparisons. Outliers were detected using Grubb's test. Results were treated as significant when p was < 0.05 and expressed as mean ± SEM.

Oxygen-Induced Retinopathy in Sprague Dawley Rats
Oxygen-induced retinopathy rodent model experiments were performed identically to those described in (Geranurimi et al., 2019), and described briefly below.

Animals
Two-day-old (P2) Sprague Dawley rat pups and their mothers were ordered from Charles River (Raleigh, SC, USA) and acclimatized for 3 days in standard conditions. All procedures and protocols involving the use of the rats were approved by the Animal Care Committee of the research center of Hôpital Maisonneuve-Rosemont and are in accordance with the Statement for the Use of Animals in Ophthalmic and Vision Research approved by the Association for Research in Vision and Ophthalmology, and guidelines established by the Canadian Council on Animal Care. The 80% oxygen model of retinopathy was conducted as previously described (Geranurimi et al., 2019). Briefly, litters of P5 pups and their mothers were kept in a controlled 80% oxygen environment until P10. The pups were, respectively injected intraperitoneally twice daily with PBS vehicle (20 µL per injection), peptide 1, or derivatives 2 (titrated to a daily dose of 2 mg/kg/day). Control litters were kept under normal air atmosphere and standard conditions. On P10, pups were euthanized by decapitation under 2% isoflurane anesthesia. Eyes were enucleated and fixed in 4% paraformaldehyde, then stored at 4 • C in phosphate-buffered saline (PBS) until further processing.

Retinal Flatmount and Immunohistochemistry
The fixed eyes were dissected, and the obtained retinas were incubated with antibodies and mounted onto slides as previously described in Geranurimi et al. (2019). Briefly, the cornea and lens were removed from the eyes, and the retina gently removed from the underlying sclera-choroid-retinal pigmented epithelium (RPE) complex. Retinas were treated for 1 h with blocking solution (1% bovine serum albumin [BSA], 1% normal goat serum, 0.1% Triton X-100 and 0.05% Tween-20 in PBS), and then incubated overnight with lectin and Iba-1 primary antibody, followed by Alexa-594-conjugated secondary antibody for 2 h. Retinas were then mounted onto microscope slides under coverslips with anti-fade mounting medium.

Microscopy
Retinal flatmounts were imaged using the Zeiss AxioImager Z2 and the MosaiX feature of the AxioVision software as previously described (Geranurimi et al., 2019). Representative images after Iba-1 staining were taken using a laser scanning confocal microscope (Olympus IX81 with Fluoview FV1000 Scanhead) using the Fluoview Software at 30X magnification.

Quantification and Data Analysis
The FIJI software was used to quantify the area of vasoobliteration in each retina, expressed as a percentage of the area of the whole retina. The number of Iba-1-positive cells was counted using the cell counter plug-in in the FIJI software, and the average of cell counts in 4 fields per retina was calculated. Data was analyzed using GraphPad Prism 7 with one-way ANOVA and the Dunnett's test for multiple comparisons. Results were treated as significant when p was less than 0.05 and expressed as mean ± SEM.

Circular Dichroism Spectra
The impact of the β-substituent on the conformation of (3R,4S)β-substituted-Agl 3 peptides 2c-q was examined in water by CD spectroscopy and the curve shapes of the spectra were compared with that of [(3R,4S)-Hgl 3 ]-1 (2b). Previously, 2b exhibited negative and positive maximum, that were respectively at 198-207 and 221-227 nm indicative of a β-turn conformation in water, trifluoroethanol (TFE), MeOH and hexafluoroisopropanol (HFIP), with the greatest ellipticity seen in 5% TFE in water. In general, peptides 2 exhibited curve shapes indicative of βturn conformers with slightly different ellipticities (Figure 2 and Supporting Information). Notably, thiocyanate 2d exhibited a similar curve shifted to higher wavelengths at 215 and 230 nm (Figure 2). On the other hand, no curve shape was obtained from measuring the CD spectrum of 4-hydroxymethyltriazolyl peptide 2q (Supporting Information). The similar curve shapes illustrated in the spectra of peptides 2b-p indicated that changes of the β-substituent had little influence on the peptide conformation, which was previously shown to be significantly affected by the presence and configuration of the γ-lactam ring (Geranurimi et al., 2019).

Biology
Anti-inflammatory agents that modulate the IL-1R but preserve NF-κB signaling are desired to avoid compromising immune vigilance against invading pathogens. Peptides 2c-q were examined for their effects on the NF-κB pathway using a reported assay that had previously been shown to contrast the immunosuppressive activity of Kineret with that of peptide 1 (Nadeau-Vallée et al., 2015). The effects of peptides 1 and 2c-q on the activation of NF-κB signaling by IL-1β was assessed using the QUANTI-blue assay, which quantifies the secretion of alkaline phosphatase, a reporter gene product for NF-κB. Peptides 2cq all behaved like 1 in the assay and exhibited no effect on SCHEME 3 | Solid-phase synthesis of peptides 2h-k. SCHEME 4 | CuAAC chemistry on resin 8c provides access to triazole derivatives 2l-q.
From the results of the in vitro screens, a subset of six (3R,4S)β-substituted-Agl 3 peptides (2c, 2d, 2f, 2l, 2n, 2q) were selected for examination in vivo in a CD-1 mouse model of preterm birth (PTB), and a Sprague Dawley rat model of oxygen-induced retinopathy (OIR). In the PTB model, timed-pregnant CD-1 dams were pre-treated with peptides 1 or 2, or PBS vehicle, and then injected with LPS on day 16.5 of gestation (G16.5). LPS, a bacterial cell wall component that contains PAMPs, is known to reliably induce labor via pro-inflammatory pathways implicating IL-1 (Hirsch and Wang, 2005). The use of LPS may better mimic more systemic endotoxemia instead of the more commonly observed intrauterine bacterial infection of spontaneous PTB in human cases. The same inflammatory responses evoked by viable FIGURE 2 | The molar ellipticity circular dichroism spectra of 2b-d, 2j, and 2l.
Gram-negative bacteria are however elicited by LPS (Huang et al., 1999), including the upregulated expression of COX-2 observed in human placental tissues (Gross et al., 2000). The CD-1 mice have a mean gestation of 19.2 days (Goupil et al., 2010). Dams that delivered at least one pup before G18.5 were considered to have given birth prematurely. In the absence of peptide, LPS alone induced premature labor in ∼80% of treated mice (Figure 6). Peptide 1 and triazoles 2l and 2q, all reduced the PTB rate to ∼20%. Azide 2c and thiocyanate 2d both exhibited modest effects by reducing the PTB rate to 40-50%. Neither Noxyphthalimide 2f nor 4-m-aminophenyltriazole 2n exhibited effects on PTB.
In the context of OIR, microglia have been previously shown to be mediators of vaso-obliteration (Rivera et al., 2013). The ramified and branched morphology of inactive microglia has also been observed to change to an amoeboid state with retracted limbs upon microglial activation (Donat et al., 2017). Microglial activation and density were thus ascertained by histochemical staining for the Iba-1 marker. Microglia in the active amoeboid state were observed in the retina of animals kept under hyperoxia and treated with vehicle or azide 2c (Figures 8A,B). Conversely, pups raised in normoxia exhibited ramified and branched microglia in their retina. Furthermore, pups presented retina with similarly ramified and branched microglia after treatment with peptides 1 and 2 (e.g., 2d, 2f, 2l, 2n, and 2q) under hyperoxia, which exhibited diminishing effects on vaso-obliteration. In summary, five of the six tested (3R,4S)β-substituted-Agl 3 peptides acted like peptide 1 and exhibited protection against vaso-obliteration in the hyperoxic phase of OIR, due in part to attenuation of microglial activation.

DISCUSSION
The relevance of the β-substituent for activity are revealed in a comparison of the results of the in vitro and in vivo experiments on (3R,4S)-β-substituted-Agl 3 peptides 2c-q and parent peptide Graphical representations of band density analysis of Western Blots are shown as fold activation compared to control. RAW Blue cells were pretreated with peptides 1 and 2 (10 −6 M), Kineret (1 mg/mL), or vehicle for 30 min and then stimulated with IL-1β for 15 min. Images of representative Western Blots can be found in the Supplementary Figure 1. Results shown are the average of 3 independent experiments: *p < 0.05, **p < 0.01, ***p < 0.001 compared to group treated only with IL-1β. Treatment groups that are not labeled with asterisks are statistically non-significant compared to group treated only with IL-1β.
FIGURE 6 | The effects of peptides 1 and 2 on prevention of PTB. In brief, pregnant dams on day 16.5 of gestation (G16.5) were subcutaneously pretreated with peptides 1 and 2 (2 mg/kg/day subcutaneous injections) or vehicle, followed by LPS (10 µg intraperitoneal injection), and observed for delivery of pups. A dam was considered as delivering preterm if at least one pup was delivered before G18.5. n = 4-5 dams per treatment group. Table 3). Comparison of the curve shapes of the CD spectra of the (3R,4S)-β-substituted-Agl 3 peptides 2c-q indicated that the β-turn conformation exhibited in [(3R,4S)-Hgl 3 ]-1 (2b) was maintained in most of the analogs. Moreover, peptides 1 and 2, all were typically pathway selective and showed no effects on NF-κB signaling, signifying retained immune vigilance. Inhibition of at least one of the three kinase pathways has been usually found to be necessary for exhibiting in vivo effects in the PTB and ROP models (Geranurimi et al., 2019). However, N-oxyphthalimide 2f did not inhibit IL-1 induced expression of COX2 and IL-1 nor kinase phosphorylation, but prevented vaso-obliteration and microglial activation in the OIR model, which indicates the likelihood of another pathway among the pleiotropic effects of IL-1 being implicated in the pathogenesis of OIR and ROP (Mantovani et al., 2019).

(
The (3R,4S)-β-Substituted-Agl 3 analogs 2c-q provide means of studying the role of the β-substituent for mediating antiinflammatory effects within a conformationally constrained variant of peptide 1. Studying the conformation of the Thr 3 residue in 1, β-hydroxy-α-amino-γ-lactam [(3R,4S)-Hgl 3 ]-1 (2b) was shown to exhibit identical in vitro and in vivo activity as the parent peptide. The β-substituent was changed to other groups that could in principle interact in hydrogen bonds, (B) Quantification of area of vaso-obliteration performed using ImageJ, expressed as a percentage of the total retinal area: n = 5-7 of peptide 2, n = 10-12 for vehicle and peptide 1; Veh vehicle; ****p < 0.0001 relative to the vehicle group, n.s. p > 0.05 relative to vehicle group and not statistically significant. Epifluorescence microscopy images at 20X magnification of retinal microglial density quantified using ImageJ: 4 images per retina were taken at a distance halfway between the optic nerve and the edge of the retina; n = 5-7 for peptides 1 and 2; n = 10-12 for normoxia and vehicle; Norm normoxia, Veh vehicle; ****p < 0.0001 relative to the vehicle group, n.s. p > 0.05 relative to vehicle group and not statistically significant.
salt bridges and π-cation interactions. Switching the alcohol to various groups which could serve as alternative hydrogenbond donors and acceptors and engage in salt bridges (e.g., 2d-2k) caused typically significant losses in activity in the three kinase pathways with notable exceptions of thiocyanate 2d (some activity on p38 and ROCK2), amide 2i (some activity on p38), and alkoxyamine (strong activity on ROCK2). Notably, thiocyanate 2d and amine 2h exhibited statistically better activity in reducing pro-inflammatory IL-1β and COX-2 gene expression in mouse macrophages compared to peptide 1.
Previously, inhibitors of JNK and ROCK2 have been respectively, shown to delay labor and reduce neovascularization in models of PTB and OIR (Pirianov et al., 2015;Yamaguchi et al., 2016). Consistent with such findings, peptides 2 exhibiting inhibitory activities on the JNK (2l and 2q) and ROCK2 (2d, 2l, 2n and 2q) pathways demonstrated efficacy in the PTB and OIR models in vivo, respectively. The battery of in vitro examinations used in this study has further validated the importance of such pathways for in vivo activity; however, the capacity of N-oxyphthalimide 2f to reduce vaso-obliteration and microglial activation in the OIR model without effects on kinase phosphorylation and gene expression demonstrates the probability that other IL-1R mediated pathways may be valid targets for the indication of ROP.
The potential influences of pharmacokinetic properties of peptides 2c-q must be considered in the examination of structure-activity relationships of the in vitro and in vivo experiments, because biological distribution and metabolism may influence potency. Nevertheless, with relatively similar or better activity than parent peptide 1, 4-(phenyl)triazole 2l has exhibited notable inhibition of phosphorylation of the three kinases and of gene expression of IL-1β and COX-2, as well as potency in both in vivo models. In contrast, the absence of activity on kinase phosphorylation and the weak inhibitory effects on gene expression of 4-(p-tolyl)triazole 2m indicate a detrimental steric effect of the p-methyl group. Alternative 4-position substituents on triazoles 2n-q were similarly less effective as the phenyl group in 2l, but maintained activity particularly against ROCK2 kinase indicating the importance of the triazole pharmacophore. The efficacy of alcohol 2b, thiocyanate 2d, and triazoles (e.g., 2l and 2n-q) indicates that a hydrogen-bond acceptor may be important for activity.
Peptides 2d-q were studied by CD spectroscopy and typically exhibited curve shapes indicative of β-turn conformation in contrast to peptide 1 which exhibited a random coil spectrum. Subsequently, β-substituted-Agl analogs 2d-q were examined in a battery of in vitro assays for inhibitory activity on NF-κB signaling, on kinase phosphorylation (e.g., p38, JNK and ROCK2), and on expression of cytokines (e.g., IL-1 and COX2). Finally, six analogs (e. g., 2c, 2d, 2f, 2l, 2n, and 2q) were tested in models of PTB and OIR. Although peptides 1 and 2 all lacked inhibitory activity on the NF-κB signaling pathway, a spectrum of in vitro activities contingent on the β-substituent were exhibited by peptides 2. Among peptides 2, thiocyanate 2d, hydroxylamine 2g, and triazole analogs (e.g., 2l and 2n-q) exhibited inhibitory activity on the ROCK2 pathway. Moreover, 4-(phenyl)triazole 2l and 4-hydroxymethyltriazole 2q exhibited inhibitory activity on the JNK pathway. Capacity to reduce vasoobliteration and microglial activation, hallmarks of retinopathy of prematurity, was exhibited in the OIR model by peptides (e.g., 2d, 2l, 2n, and 2q), which had demonstrated ROCK2 inhibitory activity. In the PTB model, peptides (e.g., 2l and 2q) delayed LPS-induced labor. The structure-activity relationships of βsubstituted-Agl peptides 2d-q offer insight into the requirements for pharmacological selectivity of IL-1R modulators. Moreover, with equal and better activity relative to peptide 1 in the in vitro and in vivo assays, 4-(phenyl)triazole 2l epitomizes a valuable lead for developing a selective anti-inflammatory intervention strategy to delay preterm birth and improve neonatal outcomes without impeding immune vigilance.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
Timed pregnant CD-1 mice were used according to a protocol approved by the Animal Care Committee of Hôpital Sainte-Justine in accordance with the principles of the Guide for the Care and Use of Experimental Animals of the Canadian Council on Animal Care. All procedures and protocols involving the use of the rats were approved by the Animal Care Committee of the research center of Hôpital Maisonneuve-Rosemont and are in accordance with the Statement for the Use of Animals in Ophthalmic and Vision Research approved by the Association for Research in Vision and Ophthalmology, and guidelines established by the Canadian Council on Animal Care.

AUTHOR CONTRIBUTIONS
AG wrote the manuscript, synthesized and purified compounds, and conducted circular dichroism analyses. CC wrote the manuscript and conducted in vivo and in vitro experiments. CQ edited the manuscript and conducted in vitro experiments. XH edited the manuscript and conducted in vivo experiments. AB, FC, and IL performed in vitro experiments. SC and WL supervised the progress of the project, edited, and proofread the manuscript. All authors have read the final manuscript and agree to be accountable for the content of this work.