Novel Locally Active Estrogens Accelerate Cutaneous Wound Healing-Part 2

Estrogen deprivation is associated with delayed healing, while estrogen replacement therapy (ERT) accelerates acute wound healing and protects against development of chronic wounds. However, current estrogenic molecules have undesired systemic effects, thus the aim of our studies is to generate new molecules for topic administration that are devoid of systemic effects. Following a preliminary study, the new 17β-estradiol derivatives 1 were synthesized. The estrogenic activity of these novel compounds was evaluated in vitro using the cell line ERE-Luc B17 stably transfected with an ERE-Luc reporter. Among the 17β-estradiol derivatives synthesized, compounds 1e and 1f showed the highest transactivation potency and were therefore selected for the study of their systemic estrogenic activity. The study of these compounds in the ERE-Luc mouse model demonstrated that both compounds lack systemic effects when administered in the wound area. Furthermore, wound-healing experiments showed that 1e displays a significant regenerative and anti-inflammatory activity. It is therefore confirmed that this class of compounds are suitable for topical administration and have a clear beneficial effect on wound healing.


Results and Discussion
Synthesis of Compounds 1. The starting material for the synthesis of compounds 1 were the known 17β-estradiol derivatives 3, whose preparation we have previously described 9 . The first step in the synthesis was conversion of 3 to the corresponding carboxylic acids 4 by means of a 2,2,6,6-tetramethylpiperidinyloxy (TEMPO) catalyzed oxidation, using NaClO 2 as stoichiometric oxidant (Fig. 2) 14 .
Compounds 4 were then converted into compounds 5 by esterification with a number of alcohols in the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) and 4-dimethylaminopyridine (DMAP) 15 . Deprotection of the phenol function with p-toluensulphonic in a MeOH/H 2 O mixture gave the target compounds 1. Compounds 3 were eventually unprotected with p-toluensulphonic in a MeOH/H 2 O mixture to  give 17β-estradiol derivatives 6. Finally, 1b was submitted to basic hydrolysis to give 7(H-NMR and C-NMR spectra of new 17β-estradiol derivatives 1, 6 and 7 are shown in Supplementary Information).

Estrogenic Activity of the New Chemical Entities (NCE).
In vitro studies -cell transfection assay. The ability of the novel compounds to transcriptionally activate ERs was tested in ERE-Luc B 17 cells, a clone of the breast cancer cell line MCF-7 stably transfected with a reporter constituted by the luciferase gene driven by an estrogen-regulated synthetic promoter previously generated and tested in our laboratories 16 . Several compounds displayed a transactivation potency (EC 50 ) of an order of magnitude compatible for therapeutic use as most synthesized compounds had an EC 50 between 2.3 × 10 −6 and 9.8 × 10 −7 M. The highest transactivation activity was shown by compound 1e and 1f with an EC 50 of 1.3 × 10 −9 and 7.9 × 10 −8 M, respectively. Compounds 1b, 6a and 1h did not show any activity on ERs (Fig. 3).
The compounds 1e and 1f were therefore selected for a further study to evaluate their estrogenic activity in vivo.
In vivo imaging. To investigate the activity of 1e and 1f at systemic level, we utilized the ERE-Luc reporter mice. The ERE-luc is a transgenic model obtained by random integration of a luciferase reporter driven by an estrogen-responsive promoter. The ubiquitous, regulated expression of luciferase was obtained by flanking the reporter construct with insulators 15 . A large number of experimental evidence demonstrates that in this mouse luciferase activity is strictly correlated with the state of ER transcriptional activity [16][17][18] and that the sensitivity of the response to estrogen is sufficient to measure by total body in vivo bioluminescence the subtle physiological changes in ER activity typical of the different phases of the reproductive cycle.
Here, the first study we carried out was a large-range dose-response analysis to measure whole body luciferase assay after subcutaneous administration of: (1) the vehicle; (2) corn oil as control; (3) the two compounds 1e and 1f at the dosage of 20 μg/kg, 200 μg/kg, 2 mg/kg and 20 mg/kg. The mice (4 for each experimental group) were subjected to in vivo whole body imaging 6 h after the pharmacological treatment. This time was selected on the basis of previous studies showing that the peak of ER transcriptional activation generally occurs at 6-8 h after administration of the estrogenic compounds 16,17 . The administration of 20 and 200 μg/kg of 1e and 1f did not induce any measurable effect at systemic level, at the higher dosages (2 and 20 mg/kg) both compound 1e and 1f induced the expression of the luciferase reporter around the area of subcutaneous injection and in some other body areas (Fig. 4).
This observation suggested that 1e and 1f up to the dose of 200 μg/kg acted locally by showing some activity at the administration site, but not a systemic level and led us to further evaluate their effects in a wound healing model.
Properties of 1e and 1f when administered in the area surrounding a wound. To study the estrogenic and healing properties of the two compounds, we generated a wound in the ERE-Luc mouse model and tested the activity of the two compounds administered s.c. in the area surrounding the wound as described in the methodology section.
The study was done in the following experimental groups: (1) vehicle (corn oil), (2) 1e (200 µg/kg s.c.); (3) 1f (200 µg/kg s.c.) and 4) 17β-estradiol (20 µg/kg s.c.) as positive control. Each group was composed of 4 ERE-Luc females' mice ovariectomized 2 weeks' prior the experiment to avoid any interference due to endogenous estrogens. Photon emission was measured in the living animals three times: at 0, 3 and 6 h after s.c. injection. Figure 5 shows the images of the mice in pseudocolors as obtained by the CCD camera in panel a and the measurement of the amount of photon emission by whole body, thymus, hepatic area, abdomen, limbs, vagina (ventral view) and in wound area (dorsal view) in panel b.
Most relevant was the fact that we did not observe increased photon emission in any of the body areas taken into consideration at 3 or 6 h after the topic administration of 1e and 1f; in contrast, 17β-estradiol treatment induced a significant increase of luciferase activity in the whole body and in several of the body areas analyzed. In the wound area photon emission measured at 6 hours time showed a trend to increase in animals treated with both 1e and 1f (30% and 20% increase with respect to time 0, respectively). In this area 17β-estradiol induced a 40% increase with respect to vehicle-treated mice; no increase was registered in the animals treated with the vehicle.
We next examined photon emission in a series of organs excised from the mice euthanized at 6 hours after treatment by measuring photon emission with the CCD camera (Fig. 6a) and by the more sensitive enzymatic assay in tissue extracts of each organ (Fig. 6b). This latter experiment clearly demonstrated that, contrary to 17β-estradiol, the two NCE did not induce bioluminescence in any of the organs investigated and further indicated that 1e and 1f were devoid of any systemic activity.
The final study aiming at investigating the systemic effect of the NCE was done by measuring the uterus weight of all treated animal as the golden standard for the quantitative measure of estrogenic compounds. Figure 7 shows that, 6 h after treatment, with 1e and 1f the weight of the uterus was indistinguishable from controls; conversely, the average uterus weight of the mice treated with 17β-estradiol was 22.8 mg, i.e 35% higher than the average weight of controls (n = 4).
Proliferative potential and healing efficiency of 1e and 1f when administered. Wound healing is a dynamic process consisting of four continuous, overlapping, and precisely programmed phases: hemostasis, inflammation, proliferation and tissue growth (proliferation) and remodeling (maturation). The healing effect of estrogens may be exerted at several levels, certainly the ability of the hormone to accelerate macrophage polarization towards the M2-reparative polarization stage plays a relevant role, however estrogens may also facilitate local cell proliferation and the formation of the granulation tissue, epithelialization, and wound contraction. We therefore planned to carry out experiments with the MITO-Luc 19 a transgenic mouse carrying a reporter of mitosis, aimed at demonstrating the local and systemic proliferative potential of the synthetic estrogens we generated.
The experiments were done with the 1e compound as completely deprived of any activity on the uterus (Fig. 6). Ovariectomized MITO-Luc mice (n = 4 each group) were injected with vehicle (corn oil), 200 µg/kg 1e and 20 µg/kg 17β-estradiol subcutaneously in the area surrounding the wound. Photon emission was measured Compounds: chemical structure of tested compounds. Dose-response: ER transcriptional activity measured as luciferase activity in the presence of increasing concentrations of indicated compounds in the ERE-Luc B17 cells, fold induction of luciferase activity measured at 10-5 M of indicated compound respect to vehicle. EC50: plotting of the transactivation data for EC50 calculation by means of sigmoidal dose-response (variable slope) using PRISM5 software.
in the wound area at the following time: 0, 24, 48 and 72 hours post-treatment. Figure 8 shows that, 48 hours after administration of 1e and 17β-estradiol, photon emission from the wound area was significantly increased; the maximal activity of both treatments with 1e and 17β-estradiol was observed at h 72 (87% increase of photon emission from wound of animals treated with 200 µg/kg 1e and 78% increase in 20 µg/kg 17β-estradiol treated animals). In animals treated with the vehicle photon emission was also increased, but at much lower extent (+13% versus time 0), indicating that both 1e and 17β-estradiol were able to regulate the extent of proliferation in the area surrounding the wound (Fig. 8).
To investigate to which extent the increased proliferation observed after the pharmacological treatments was functional to the healing process, we measured the size of the wound (Fig. 9a). 72 hours after the treatments we observed a 73% wound area reduction in 1e treated animals, 60% in 17β-estradiol treated animals and 50% in vehicle animals. These data suggested that 1e accelerates wound healing.
To better understand whether the healing effect of 1e could be attributed to its anti-inflammatory action, at the end of in vivo imaging (72 hours after treatments) the tissue surrounding the wound was dissected and subjected immunohistochemical analysis using the pan macrophage marker Iba1. As shown in Fig. 9b, in the animals treated with 1e and 17β-estradiol the number of macrophages in the wound area significantly lower than in controls (25% and of 44% respectively). This observation confirmed that the estrogenic compounds were able to regulate the immune response at local level.
Next, we investigated the effects of 1e with respect to 17β-estradiol using ovx MITO-Luc mice at longer time. The first day of the experiment, one round full thickness wound was made on each mouse (n = 2) and then all mice were treated with vehicle, 200 µg/kg 1e or 20 µg/kg 17β-estradiol by subcutaneous injection around the round wound. The measure of photon emission from the wound area was done for several days after the wounding. 3 days after treatment the bioluminescent signal is highest ( Fig. 10 Panel a). The intensity of photon emission was significantly decreased 2 days later and kept decreasing to reach, at day 10, values very comparable to those measured before the wounding (time = −1 day). The histological analyses done on the explanted wound tissues at day 10 showed that in the animals treated with 1e there was a significantly higher granulation tissue than (histoscore mean = 1.5) in those injected with vehicle (histoscore mean = 0.3) (Fig. 10 panel b).
Comparative analysis of 1e and 17β-estradiol systemic proliferative activity. The study was done in the MITO-Luc mice. Photon emission was measured in the following body areas: head, sternum, vagina, femur and breast     (ventral view, as indicated in Fig. 11a) at 0, 24, 48 and 72 hours after the local, s. c., injection of vehicle, 200 µg/kg 1e and 20 µg/kg 17β-estradiol (n = 4). No increase of bioluminescence was measured in 1e treated animals respect to vehicle at any time point (Fig. 11b).
This was not the case with the animals treated with 17β-estradiol where 48 hours after the local s.c. injection of 17β-estradiol we observed that the bioluminescent signal showed a trend to increase, likely due to the uterotrophic effect of the hormone. This observation further underlines the lack of systemic effects of the 1e compound.

Conclusions
In the present study we have prepared a family of compounds active through the ER characterized by a structure that should ensure their rapid metabolism and limited systemic action when administered in vivo. Most of compounds tested, and in particular 1e and 1f, showed a significant affinity for ERs and were able to efficiently induce ER transcriptional activity.
The biological activity of 1e tested in a well-characterized wound healing murine model 8 showed a significant effect on wound healing and inflammatory processes. 1e proved to be very effective in accelerating healing. Several studies suggest that the beneficial effects of estrogens on the healing process is in the dampening of the inflammatory response, in particular it was reported that accelerated healing was correlated with a reduction in infiltrating macrophages numbers 1,3,4 . The current study supports this view by showing a decreased number of Iba1 positive macrophages in the granulation tissue (Fig. 9b); the increase of proliferation in the 1e and 17β-estradiol treated animals may be in line with the view of an accelerated transition from inflammatory phase to proliferative phase. In this regard, 1e showed an activity comparable to that one of the natural estrogen. Most importantly, 1e administered s.c. at the dose of 200 μg/kg in the wound area proved to be devoid of systemic effects as indicated by the lack of ER transcriptional activity in the ERE-Luc model, of bioluminescence signaling in the MITO-Luc mouse and the lack of uterotrophic effect. All parameters were clearly affected by 17β-estradiol (Figs 5, 6 and 11).
Our study, by identifying a new class of molecules locally active through the ERs in the wound healing process confirms the results of our previous work and add new avenues in the search of novel drugs for a still unmet medical need. To this mixture, NaClO (5.25%, 0.24 mL, 0.20 mmol) and NaClO 2 (80%, 0.34 g, 3.0 mmol) were added portionwise simultaneously.

Materials and Methods.
The solution became slowly brownish and was left under stirring at 40 °C till disappearance of the starting material (generally 2-4 hours, TLC: EtOAc). To the reaction mixture H 2 O (40 mL) was added and pH adjusted to 8 with 4N NaOH. The reaction flask was then cooled to 0 °C and quenched with a Na 2 SO 3 aqueous solution (2 g in 30 mL). The whole was then extracted with methyl-t-butylether and the organic phase discarded. After adjusting Hematoxylin-eosin representative staining of wound section at 10 days after the indicated treatments. The histoscore value for granulation tissue (fibrosis) is shown as mean, each histoscore consists of 4 random fields/groups, the criteria for the assignation of histoscore value are: 1 = wound bed partially covered with granulation tissue; 2 = thin granulation over the whole wound bed; 3 = thick granulation over the whole wound bed. Bar = 2 mm. the pH to 4 the aqueous phase was thoroughly extracted with EtOAc. Combined organic phases were washed with water, brine, dried over anhydrous Na 2 SO 4 and evaporated to dryness to give carboxylic acids 4. The latter was used as crude in the next reaction. Thus to an ice-bath cooled solution of 4 and DMAP (30 mg, 0.25 mmol) in the proper alcohol (20 mL), EDC (0.23 g, 1.2 mmol) was slowly added in 40 minutes. The ice bath was removed and the whole allowed to stir at rt till disappearance of the starting material (generally 20 hours, TLC: EtOAc/hexane = 6/4). The mixture was then concentrated under reduced pressure, H 2 O was added and the whole extracted with EtOAc. The combined organic phases were washed with NaHCO 3 s.s., brine, dried over anhydrous Na 2 SO 4 and evaporated to dryness to give crude 5.
The crude product of the previous reaction was dissolved in MeOH (20 mL) and a 0.1 M p-toluensulphonic acid solution was added to adjust the pH to 2. The resulting solution was left under stirring at rt until TLC showed the complete disappearance of the starting material (generally 1 h). The solution was then diluted with EtOAc (80 mL) and washed to neutrality with a saturated NaHCO 3 aqueous solution, and the aqueous phase was extracted with EtOAc. The combined organic phases were collected, washed with brine and dried. After filtration, evaporation of the solvent and purification of the crude product, gave 1 as a pale yellow viscous oil. Hydroxy-17α-(4′carbomethoxy-1′-butyn-1′-yl)estra-1,3,5-(10)-triene (1a). Yield     3,17β-Hydroxy-17α-(4′carboxy-1′-butyn-1′-yl)estra-1,3,5-(10)-triene (7). To a rt stirred solution of 1b (0.10 mmol) in EtOH (12 mL) a 0.1 M KOH solution (10 mL) was added. The whole was allowed to stir for full-thickness skin (around 9 mm in diameter) using fine scissors on the dorsal skin of the isoflurane-anesthetized animal. For in vivo treatment the compounds were administered onto the subcutaneous area surrounding the wound by multiple subcutaneous injection, immediately after the generation of the wound.

3,17β-
In vivo imaging. For the semi-quantitative analysis of in vivo photon emission, animals were injected i.p.
with 80 mg/kg of luciferin (Beetle Luciferin Potassium Salt; Promega, Madison, WI, USA) 15 min prior the imaging session. For the imaging, mice were anaesthetized using isofluorane (Isofluorane-Vet; Merial, Lyon, France) and kept under anesthesia during the 5 minutes of the session carried out with a CCD camera: both ventral and dorsal acquisition (Xenogen IVIS Lumina System; Caliper, PerkinElmer company). Photon emission in selected body areas was measured using the Living Image Software (Caliper, PerkinElmer company). The analysis was done in ventral view: whole body, thymus, hepatic area, abdomen, limbs, vagina and in dorsal view in wound area and expressed as photon/second/cm2/steradiant (p/s/cm2/sr).
Ex vivo imaging. The selected organs were dissected from the animals who had received luciferine 20 min before euthanasia. The tissues were immediately subjected to ex vivo imaging session of 5 minutes. Photon emission was quantified with the Living Image Software (Caliper, PerkinElmer company) for each tissue and expressed as p/s/cm2/sr.
Luciferase enzymatic assay. Luciferase enzymatic activity was measured in cells or selected tissue homogenates by a luminometer (Glomax ™ 96 microplate luminometer; Promega) using Promega standard luciferin (Beetle Luciferin Potassium Salt) and expressed as Relative Light Units over 10 sec/μg protein (RLU/μg protein). Proteins concentrations were measured by Bradford.
Histology, immunohistochemistry and image analysis. At the time of euthanasia, the wounds with surrounding healthy tissue were excised, distended on a strip of hard paper and fixed in 10% buffered formalin (pH = 6). Fixed samples were cut through the center of the wound and each half was embedded in paraffin "sideways on". All fixed samples underwent routine processing in an automatic tissue processor and then paraffin embedding was carried out according to standard procedures. For histopathological examination 4 µm sections were routinely stained with Hematoxylin-Eosin (HE) and evaluated under a light microscope. The degree of re-epithelialization, the type and amount of granulation tissue and inflammation, which are the main features of wound healing, were scored separately. The scoring system is slightly modified compared to that applied by Tkalcević 20 .
In particular, the criteria for the assignation of granulations value are: 1 = wound bed partially covered with granulation tissue; 2 = thin granulation over the whole wound bed; 3 = thick granulation over the whole wound bed. Evaluation of macrophages was done through a specific antibody and digital image analysis: 4 µm sections from each wound healing sample were immunostained with primary rabbit monoclonal antibody against Iba1 antigen (Wako; #019-19741) to assess the number of macrophages.
The number of positive cells were counted using the ImageJ analysis program (http://rsb.info.nih.gov/ij/) in 400x microscopic fields selected at the periphery of the wound area.