Development of Diaryl Hydrazones for Alleviation of Mitochondrial Oxidative Stress in Preeclampsia

Abstract

Preeclampsia is a pregnancy-specific syndrome, linked to oxidative stress, affecting 5–8% of pregnancies, with no effective treatment available. Here, diaryl-hydrazones have been designed, synthesized, and investigated as mitochondria-targeting antioxidants to reduce placental oxidative stress and mitigate preeclampsia symptoms. The design, based on density functional theory studies, revealed that conjugated electron structure with the NH-motif appeared to explain their effect. Thirty compounds were synthesized and tested in three assays, where they exhibited excellent radical scavenging activity, significantly greater than that of the standard, Trolox. Based on the data, eight compounds were selected for cell-based assays. Oxidative stress was induced in human trophoblast cells and assessed whether the compounds reduced downstream antiangiogenic responses using ascorbic acid and MitoTEMPO as standards. The pretreatment with the hydrazones reduced mitochondrial superoxide and sFLT-1 production in H₂O₂-exposed trophoblast cells, indicating that mitochondrial oxidative stress and the anti-angiogenic response can be alleviated by these compounds.

Graphical Abstract

INTRODUCTION

Preeclampsia (PE) is a common, life-threatening complication in pregnancy, characterized by high blood pressure putting the mother at risk of eclampsia and kidney dysfunction.¹ PE affects about 5–8% of pregnancies and is highly prevalent among African American women. It is estimated to cause 70,000–80,000 maternal and 500,000 perinatal deaths worldwide annually.² There is no clear pathogenesis or cure, thus an effective treatment for PE is an unmet medical need. It has been shown that placental oxidative stress is a major contributor to PE.^3–5 In previous studies, several natural antioxidants and synthetic derivatives have been tested with limited success mostly due to delivery issues.⁶ The presence of oxidative species, such as reactive oxygen and nitrogen species (ROS, RNS) are both crucial for cell survival and reactive enough to cause cell death, thus the body needs to maintain balanced redox homeostasis.⁷ To control the levels of free radicals, the body utilizes endogenous and exogenous antioxidants to scavenge the radicals when necessary to avoid oxidative stress and damage onto the cells.⁸ Oxidative stress causes damage to a wide variety of biomolecules (i.e., lipids and DNA), potentially causing cellular death.⁹ Overproduction of ROS and RNS contributes to the development and progression of many ailments, such as aging, cardiovascular diseases, Alzheimer’s disease, Parkinson’s disease, diabetes, cancer, or multiple sclerosis.^10–14 Among these disorders, preeclampsia is one of the most serious complications of pregnancy, characterized by hypertension and proteinuria occurring after 20 weeks.¹ It is associated with significant maternal, fetal, and neonatal morbidity and mortality.² No disease-altering treatments are available, current care is limited to treating symptoms: mitigating hypertension and alleviating eclampsia. While the etiology and pathogenesis of PE is elusive, it is currently believed that placental ischemia, due to impaired spiral artery remodeling, is the primary culprit.^5a It has been proposed that the ischemic state will cause oxidative stress and impair endothelial and trophoblast function that may contribute to the pathogenesis of PE. In PE, pregnancy significant ROS production in the placenta has also been reported.¹⁵ In addition, the mitochondrial electron transport chain enzyme, cytochrome c oxidase, activity is diminished in the syncytiotrophoblast cells of the placenta, implicating mitochondrial damage/dysfunction as a potential contributor to pathogenesis of PE.¹⁴ The release of ROS/RNS can stabilize hypoxia inducible factor (HIF-1α) which will induce transcription of antiangiogenic factors such as soluble fms-like tyrosine kinase 1 (sFLT-1) and soluble endoglin (sEng).^16,17 The antiangiogenic factors are released into the maternal circulation; and their actions are thought to disrupt the maternal endothelium and result in hypertension, proteinuria, and other systemic manifestation of PE (Figure 1).¹⁸

Figure 1.

Schematic representation of human preeclampsia development and the potential intervention used in this work.

Given that placental oxidative stress may be an early trigger in the pathogenesis of preeclampsia, therapies that enhance antioxidant pathways have been proposed as treatments. The first attempts in clinical trials have shown that maternal vitamin C and E therapy do not prevent PE,¹⁹ likely due to the inability of the antioxidants to reach the site of oxidative stress (i.e., mitochondrial matrix).²⁰ Therefore, there is an interest in developing alternate strategies to reverse oxidative stress in the placenta. Mitochondria-targeted antioxidants²¹ have been used in^22,23 In models of ischemia-reperfusion injury^22,23 and are an attractive alternative treatment as preeclamptic tissue is characterized by profound mitochondrial oxidative stress.^24,25 Data has also shown promise that delivery of H₂S to the mitochondria of primary trophoblast cells could alleviate oxidant stress and sFLT-1 upregulation in a hypoxia model in vitro.^3c This data implicate that the use of antioxidants at an early stage have potential to prevent the trophoblast and endothelial injury that is observed in PE.

In our earlier work, we reported the design, synthesis and application of hydrazone-based antioxidants that acted as multifunctional anti-AD agents and exhibited excellent radical scavenging effects. The starting point was resveratrol (Figure 2), a biologically active natural product^26,27 and effective antioxidant.²⁸

Figure 2.

General structures and electrostatic potential maps of hydrazones and resveratrol.

However, resveratrol is too polar to reach most target sites. In contrast, the diaryl hydrazones offer a large degree of structural similarity with more hydrophobic character, showing excellent radical scavenging, often three times higher than that of the controls.²⁹ Density functional theory (DFT) studies also provided insights into the origin of their high antioxidant potential.³⁰ In addition to being potent radical scavengers,^29–31 hydrazones are nontoxic and possess a variety of biological activities, yet their potential as a drug candidate is underexplored.^32,33

Continuing our efforts³⁴ on the development of synthetic hydrazones for mitochondrial antioxidant therapy to alleviate PE, herein we describe the design, synthesis, and evaluation of organofluorine hydrazones as potential PE therapeutics.

RESULTS AND DISCUSSION

Chemistry: Design and Synthesis.

As mentioned above, resveratrol (Figure 2), a natural polyphenolic antioxidant with a stilbene core, served as a starting point in our inhibitor design. Hydrazones are highly similar molecular entities, exhibiting comparable size and structural rigidity with an overall higher lipophilicity. Another important factor is that the potential tautomeric equilibrium (Forms I and II, Figure 2) would provide the same continuous conjugation and nearly uniform π-electron delocalization similar to resveratrol, via a 4 atom-4-electron system that includes the two nitrogen atoms. The conjugation appears to be important due to the likely hydrogen atom transfer (HAT) radical scavenging mechanism,³⁵ the hydrazones presumably act through. Our extensive DFT calculations on the electronic and structural features of hydrazones provided more insight into the role of the substituents in affecting activity. It has been found that the N–H bond dissociation energy (BDE, the main driver of the HAT mechanism), ionization potential (IP), and the partition coefficient (logP) of the compounds showed strong correlation with the experimental activity.^30a In addition to these characteristics, the HOMO energy, the HOMO–LUMO gap, dipole moment (DP), proton affinity (PA), and the Hammett constant of the substituents showed good to excellent correlation with the experimental activities and related to the delocalized electron structure. Thus, we included hydrazones with extended network of conjugation adding additional aromatic rings to the design as shown in Figure 3. This involved two ways; (i) using terephthalaldehyde-linked with two hydrazine molecules, or (ii) changing the starting benzaldehyde unit to naphthyl and anthracenyl or coumarin-based aldehydes.

Figure 3.

Structures of synthesized hydrazones used in this study.

The other crucial element in our design was the introduction of fluorine into the hydrazone core that is expected to further enhance the biological properties via electronic effects as well as improving the membrane permeability of the compounds. The beneficial effects of fluorine incorporation into small molecule drugs are known to improve their biological properties.³⁶ As a result, fluorine substitution became a mainstream approach in drug development to enhance the membrane permeability of therapeutic candidates.^37,38 Fluorine containing groups used in this study will include CF₃ and aryl-F which are highly stable and would not yield toxic products as they are largely resistant to metabolic degradation. The presence of polar groups ensures proper water solubility, and the fluorine addition improves lipophilicity, thus resulting in favorable pharmacokinetic properties.³⁹

The above detailed considerations led to the design of several compounds. The structures of the hydrazones synthesized for this study are illustrated in Figure 3.

After the design, the structures were subjected to a SwissADME screening⁴⁰ to evaluate their drug-like properties and synthetic accessibility. The full set of data is tabulated in Table S1 (see Supporting Information). The analysis of the data indicates that the design appears appropriate, all compounds show proper drug-like properties. Most molecules have no or just 1 Lipinski violation,⁴¹ other violations such as Ghose-,⁴² Veber-,⁴³ Egan-,⁴⁴ and Muegge-violations⁴⁵ are also minimal (mostly zero or 1) for the above set of compounds. The number of PAINS alerts⁴⁶ are also quite favorable, most compounds have zero PAINS alerts, and no compound received more than 1, indicating that the design resulted in reasonable structures. The compounds also exhibit relatively easy synthetic accessibility scores between about 2.5–3 indicating that their synthesis is likely achievable. In addition, the hydrazones appear to be highly stable in a powder/crystalline form for an extended period of time (>1 year). When dissolved for the assays they are stable for several hours, up to a day. After 7 days of storage in solution, independently of the conditions, they show some signs of deterioration.⁴⁷

Based on the above design and analysis, a variety of hydrazones were synthesized from commercially available arylaldehydes (including the bifunctional terephthalaldehyde) and arylhydrazines. The basic synthetic procedure for preparation of these compounds is a simple rapid, catalyst-free reaction that yields the crystalline products upon cooling the reactions to −20 °C, as has been described before²⁹ (Figure 4).

Figure 4.

Synthesis of the hydrazones used in this study.

The starting aldehydes and hydrazines were all commercially readily available chemicals and were selected to ensure a broad substituent scope for the hydrazones including electron donating or electron withdrawing substituents, including fluorine-containing moiety (Ar–F, or CF₃) to enhance the lipophilicity of the compounds. Overall, 30 compounds were prepared and characterized for this study.

Biochemical Evaluation.

After the syntheses were completed, the antioxidant activity of the compounds was evaluated by three common antioxidant assays such as the 2,2-diphenyl-1-picrylhydrazyl (DPPH),¹⁴ 2,2’-azino-bis(3-ethyl-benzothiazoline-6-sulfonic acid) (ABTS)⁴⁸, and oxygen radical absorbance capacity (ORAC) assays.⁴⁹ In order to put the observed effects into context, the activity of hydrazones is presented as Trolox equivalents (in DPPH and ABTS assays) or Trolox equivalent concentration (in ORAC assay) as described in the Experimental section below. Trolox (Figure 5), a vitamin E analog, is a well-known reference antioxidant commonly used for this purpose.⁵⁰ Other reference antioxidants that were applied in the cell biology assays were ascorbic acid, and MitoTEMPO (Figure 5). The data is shown in Figures 5–8. It must be noted that the concentrations used in these assays were different and selected based on earlier works.³⁰ These concentrations allowed for the most variety in radical scavenging percentages so that all the antioxidants show radical scavenging that is neither near 100% nor 0%, thus reflecting the differences between the hydrazones as well as the standards in scavenging the target radicals. In addition, it allows for better standardization to the reference antioxidant, Trolox. Standardizing the radical scavenging percentages to that of Trolox takes into account any small deviations that could result from the assays. In the DPPH and ABTS assays, their respective antioxidant concentrations give a Trolox radical scavenging percentage that is ideal for identifying potent antioxidants. In the ORAC assay, Trolox is an extremely good radical scavenger thus a calibration curve of it is needed but the hydrazones still show a variety of radical scavenging performance. Higher antioxidant concentrations in this assay would lead to nearly 100% radical scavenging for many hydrazones, eliminating the potential for comparison.

Figure 5.

Structure of Trolox, ascorbic acid and MitoTEMPO reference compounds used for comparison in the assays.

Figure 8.

ORAC radical scavenging potential of the hydrazone derivatives (1–30) used in this study, measured at 80 μM antioxidant concentration, expressed in Trolox equivalent concentration. The values are shown as the means of the Trolox equivalent concentration ± standard deviation, where the number of independent repeats is three.

The results in Figures 6–8 are expressed in comparison to the activity of Trolox used as the reference antioxidant, either in Trolox equivalents (DPPH, ABTS, Figures 6, 7) or Trolox equivalent concentration (ORAC, Figure 8) due to the kinetic nature of the ORAC assay. The data indicate that the hydrazones are highly potent radical scavengers as the majority of the tested compounds show higher activity than that of Trolox. It is also clear from Figures 6–8 that the activity and behavior of the hydrazones toward the different radicals used in the assays changes significantly. While the hydrazones exhibit activity that is several times higher than that of Trolox in the DPPH and ABTS assays (Figures 6, 7), they show similar or somewhat lower activity than the reference in the ORAC assay (Figure 8). That is the main reason to test the antioxidants in different assays, to obtain a more holistic evaluation. The assays have their own strengths and limitations, as well as the radical scavenging mechanism of the various antioxidants affects their activity in the assays, for instance Trolox (or phenols in general) shows an outsized effect in the ORAC assay. The error bars plotted in Figures 6–8 also (within 5–10%) confirm that the assays provided sufficiently reproducible data.

Figure 6.

DPPH radical scavenging potential of the hydrazone derivatives (1–30) used in this study, measured at 200 μM antioxidant concentration, expressed in Trolox equivalents. The values are shown as the means of the Trolox equivalents ± standard deviation, where the number of independent repeats is four.

Figure 7.

ABTS radical scavenging potential of the hydrazone derivatives (1–30) used in this study, measured at 500 μM antioxidant concentration, expressed in Trolox equivalents. The values are shown as the means of the Trolox equivalents ± standard deviation, where the number of independent repeats is three.

The combination of the data from the three assays also revealed that the most potent compounds are at the beginning of the list, compounds 1-17, and the compounds with extended aromatic structures do not appear to stand out in the assays except 20 in the ORAC assay (Figure 8). Considering the structure–activity relationship (vide infra) and potential physical parameters of the compounds that significantly contribute to their ADME properties, particularly potential membrane permeability, eight compounds have been selected for cell-based assays.

Biological Evaluation.

As discussed in the introduction, oxidative stress appears to be an important factor in PE pathogenesis; therefore, novel antioxidant therapy is expected to improve trophoblast and/or endothelial cell function and may lead to improved angiogenic balance in vitro. The hydrazones showed significant antioxidant potential in the above radical scavenging assays and their in silico predicted pharmacokinetic features (Table S1) appeared to be superior to those of natural antioxidants. Thus, the cell biology assays were designed to reveal whether the compounds can also be effective in cellular environment. Based on the radical scavenging assay results described above, eight compounds (and 2 reference antioxidants: ascorbic acid and Trolox) were subjected to cell-based assays. The selected structures are summarized in Figure 9.

Figure 9.

Structures of hydrazones selected for cell biology assays.

The compounds have been tested to observe whether the antioxidant pretreatment was able to reduce the sFLT-1 protein expression in H₂O₂-exposed human trophoblast cells. Pregnancy complications are associated with placental oxidative stress, which induces the stabilization of the transcription factor HIF1A and the expression of downstream antiangiogenic factors such as sFLT-1 expression in PE. In fact, the use of these biomarkers has recently been approved by the FDA to predict PE development risk early in pregnancy.⁵¹ Therefore, we employed the human trophoblast HTR8/SVneo cell line to determine the effect of various hydrazone-based antioxidants on sFLT-1 expression.

First, we evaluated the oxidative stress response in HTR-8/SVneo cells. Cells were incubated with increasing concentrations of H₂O₂ (40, 60, 80, 100, and 120 μM) for 24 h. CCK-8 assay was used to assess cell viability, as shown in Figure S106. We detected a dose-dependent increase in cytotoxicity of HTR-8/SVneo cells in response to H₂O₂. The IC₅₀ value of H₂O₂ concentration in this system was 100 μM (which depicts 50% killing of the HTR-8/SVneo cells). Based on this finding, in subsequent experiments 100 μM H₂O₂ treatment was used to induce oxidative stress. The corresponding sFLT1 levels were also measured and showed H₂O₂ concentration-dependent increase when normalized for viable cell count (Figure S107).

Cells were pretreated with hydrazones 0.1–50 μM 30 min before the addition of 100 μM H₂O₂ to the cells. The cells were then incubated for a further 20 h in the same medium (containing hydrazones and hydrogen peroxide). We found that the various hydrazone pretreatments all reduced sFLT-1 protein expression but with different efficiencies in H₂O₂-exposed human trophoblast cells as depicted in Figure 10.

Figure 10.

Antioxidant pretreatment reduced sFLT-1 protein expression in H₂O₂-exposed human trophoblast cells. sFLT-1 ELISA in HTR8 cells pretreated with compounds 1, 2, 3, 4, 5, 6, 7, 16 or reference antioxidant MitoTEMPO (MT) for 30 min, then exposed to 100 μM H₂O₂ or sham (summary of 5 repeat experiments). *P < 0.05 compared to H₂O₂; **P < 0.01 compared to H₂O₂, ***: p < 0.001 compared to H₂O₂ by Kruskal–Wallis test followed by Dunn’s multiple comparison post hoc.

The results indicate that H₂O₂ increased the sFLT-1 expression in the trophoblast cells, as expected (Figure 10 red column). The pretreatment with all hydrazones reduced sFLT-1 expression in HTR8/SVneo cells to a varied extent. The most significant mitigating effect was achieved with compounds 4 and 5. The effect of these compounds is about 10–40% better than that of the reference antioxidant, MitoTEMPO (a mitochondria targeting antioxidant^52,53). This is a significant result since sFLT-1 is known to induce the cardinal features of preeclampsia in vivo;¹⁸ therefore, the reduction observed here could have significant therapeutic benefits in the in vivo models as well.

Based on the above in vitro antioxidant assays compound 5 was proven to be the one with the highest activity therefore we selected this compound for further cell-based assays. Immunofluorescence analysis of MitoSOX Red (A-D, superoxide production) in HTR8 cells was carried out, with 50 μM 5 treatment for 30 min, which was then exposed to 20 h H₂O₂.

The data in Figure 11 clearly indicates that a pretreatment by 5 reduced mitochondrial-derived ROS production in H₂O₂-exposed trophoblasts cells (HTR8). Comparing the H₂O₂-treated cells with those where compound 5 was also applied, about 50% reduction can be observed in the production of ROS in the mitochondria. While the positive effects of 5 have been established above, this data supports the original hypothesis that the organofluorine hydrazone acts via a mitochondrial antioxidant process.

Figure 11.

Immunofluorescence analysis of MitoSOX Red (a) A-D: Immunofluorescence analysis of MitoSOX Red: mitochondrially derived superoxide production in HTR8 cells treated with vehicle (A) or 50 μM compound 5 (B) for 30 min, then exposed to 20 h with 100 μM H₂O₂ with compound 5 (D) or without (C) (Bars: 50 μm). E-H: MitoSOX Red images overlaid on bright field images of the trophoblast cell culture (E-H) demonstrating maintained cell viability at all concentrations; vehicle (E), 50 μM compound 5 (F), 100 μM H₂O₂ (G), 100 μM H₂O₂ with compound 5 (H). (b) Quantitation of MitoSOX Red (I) immunofluorescence in trophoblasts: OD per area (pixels²) of cell surface area was calculated in 3 high-power fields per sample (n = 4 per group). Mann–Whitney-U-test, median [IQR]. Control vs 100 μM H₂O₂: ***: p < 0.001, 100 μM H₂O₂ vs H₂O₂ + 5: **: p < 0.01.

Structure–Activity Relationship.

In the above experimental data, we described the efficacy of the compounds in a variety of chemical and biological assays. The radical scavenging mechanism of antioxidants can be broken down into two main processes: single electron transfer (SET) or hydrogen atom transfer (HAT). Depending on how these events are sequenced, we can classify radical scavenging pathways as sequential electron transfer-proton transfer (SET-PT), sequential proton loss-electron transfer (SPLET), sequential proton loss–hydrogen atom transfer (SPLHAT) or a hydrogen atom transfer (HAT) and radical adduct formation (RAF).⁵⁴ Different antioxidant assays associated with different mechanisms, for instance, while ORAC is based on hydrogen atom transfer, ABTS and DPPH are considered mixed assays that include both electron and hydrogen transfer.⁵⁵ Therefore, it is critical to investigate the antioxidant activity of compounds across different assay environments.

Upon closer examination of Figures 6, 7, and 8, it is evident that compounds 2, 3, 4, 5, 6, 7, 10, 13, and 17 demonstrated superior scavenging activity across all three assays, even outperforming the reference antioxidant, Trolox. Generally, most compounds showed higher activity in the ABTS assay compared to the DPPH assay. Within compounds 2, 3, 4, 5, and 6, a push–pull-like system is observed, where one phenyl ring is substituted with electron-donating groups such as −N(Me)₂, hydroxyl, or methoxy groups, while the other phenyl ring is substituted with an electron-withdrawing (− CF₃) group.⁵⁶ On the other hand, the pattern is not determining, compounds 1 and 7, which have electron-withdrawing substituents on both aryl groups of the molecule, also exhibited notable antioxidant activity. The introduction of a methyl group to the hydrazone moiety showed mixed activity results in compounds 9, 10, 11, 12, 13, 14 and 15, depending on the substitution pattern. Compounds 11, 12, and 14 exhibited the weakest activity in the DPPH and ABTS assays, with a common feature being the NO₂ group substituent. Extending the π system linearly in compounds 15, 16 and 17, particularly with electron-donating substituents like −OMe, resulted in a superior overall activity profile. Interestingly, asymmetric expansion of the π system did not consistently enhance the activity. Notable discrepancies were observed among compounds 20, 21, 22, 28, and 29. Compounds 21 and 22 exhibited strong activity in the ABTS and DPPH assays but performed poorly in the ORAC assay. Conversely, compounds 20, 28, and 29 demonstrated excellent activity in the ORAC assay but showed limited effectiveness in the ABTS and DPPH assays. In comparing the naphthyl and anthryl derivatives, the former group’s compounds 21, 23, 24, 29, and 30 matched in performance the reference antioxidant Trolox in the DPPH and ABTS assays but exhibited a mixed activity profile in the ORAC assay, depending on the substitution pattern. Specifically, compounds 23, 24 (2,5-diMe), and 29 (4-OMe), either unsubstituted or with electron-donating groups, showed good performance in ORAC. In contrast, compounds 21 (3-NO₂) and 30 (4-Cl), which contain electron-withdrawing groups, demonstrated minimal activity in the ORAC assay. Overall, while the anthryl derivatives generally underperformed compared to the naphthyl derivatives, they exhibited moderate activity in both the DPPH and ABTS assays, except for compound 27.

To elucidate the reasons behind the compounds’ activity profiles, a theoretical analysis was performed using density functional theory (DFT) calculations at the B3LYP/6–311++G(d,p) (Table 1)⁵⁷ as well as M06–2X/6–311++G(d,p) (Table 2) level of theories in gas phase, utilizing the Gaussian 09 program suite.⁵⁸ The ORCA 6.0,⁵⁹ global optimizer and conformer generator GOAT program was utilized for the initial conformation analysis. As previously mentioned, eight promising compounds were selected for further biological testing and detailed theoretical analysis.

Table 1.

Calculated Molecular Descriptors for the Selected Hydrazone Derivatives^a

Compound	N–H distance (Å)	Dipole moment (Debye)	LogPo/wb	HOMO (hartree)	LUMO (hartree)	HOMO–LUMO gap (eV)	BDE (kcal/mol)	PA (kcal/mol)	IP (eV)/(kcal/mol)		Spin density N radicalc
1	1.0170	3.9999	2.16	−0.2331	−0.0926	3.8251	79	323.68	7.71	177.80	0.5005
2	1.0155	8.5192	2.77	−0.1918	−0.0548	3.7282	75	337.89	6.43	148.31	0.4707
3	1.0165	8.2403	2.46	−0.2022	−0.0553	3.9951	76	338.44	6.77	156.13	0.5216
4	1.0153	5.4930	2.92	−0.2002	−0.0572	3.8912	75	337.24	6.69	154.22	0.4913
5	1.0153	6.4249	2.78	−0.1889	−0.0509	3.7565	74	339.32	6.37	146.83	0.4767
6	1.0167	4.1790	3.09	−0.1877	−0.0476	3.8120	74	340.27	6.30	145.17	0.4272
7	1.0163	8.6191	2.05	−0.2317	−0.1138	3.2098	77	320.23	7.65	176.44	0.4877
16	1.0158	5.4401	3.31	−0.2063	−0.0867	3.2531	75	327.49	6.68	176.44	0.4700

^a(Bond distance (in Å), Dipole moment in Debye, logP_o/w, HOMO and LUMO energies in hartree, band gap in eV, bond dissociation energy (BDE, in kcal/mol), proton affinity (PA, in kcal/mol), ionization potential (IP, in eV and kcal/mol). Level of theory and basis set B3LYP/6-311++G(d,p).

^bSwissADME, consensus.

^cMulliken atomic spin densities.

Table 2.

Calculated Molecular Descriptors for the Selected Hydrazone Derivatives^a

Compound	N–H distance (Å)	Dipole moment (Debye)	HOMO (hartree)	LUMO (hartree)	HOMO–LUMO gap (eV)	BDE (kcal/mol)	PA (kcal/mol)	IP (kcal/mol)	Spin density N radicalb
1	1.0123	3.9522	−0.2729	−0.0531	5.9811	83	325.19	182.09	0.5262
2	1.0111	7.8066	−0.2340	−0.0204	5.8128	79	336.57	153.69	0.4988
3	1.0110	4.9488	−0.2529	−0.0284	6.1100	78	327.99	165.91	0.5271
4	1.0116	5.2512	−0.2446	−0.0217	6.0649	79	335.85	160.65	0.5180
5	1.0117	6.9349	−0.2334	−0.0191	5.8341	78	337.29	152.66	0.5018
6	1.0103	4.8829	−0.2325	−0.0168	5.8681	80.0	391.56	152.21	0.5125
7	1.0118	2.9485	−0.2724	−0.0746	5.3821	82	393.05	181.39	0.5207
16	1.0117	3.2754	−0.2489	−0.0524	5.3470	80.0	327.53	162.29	0.5117

^a(Bond distance (in Å), Dipole moment in Debye, logP_o/w, HOMO and LUMO energies in hartree, band gap in eV, bond dissociation energy (BDE, in kcal/mol), proton affinity (PA, in kcal/mol), ionization potential (IP, in eV and kcal/mol). Level of theory and basis set: M06-2X/6-311++G(d,p).

^bMulliken atomic spin densities.

Then, the obtained values describing physical properties have been analyzed in connection to the experimental antioxidant potential. The experimental radical scavenging data were plotted as a function of the above calculated properties, respectively, and analyzed to reveal a potential fit between the data that may help to describe the relationship between structure and activity. All plots are depicted in SI, here we only show and analyze those that exhibited a meaningful correlation (Figure 12). Several physicochemical and quantum parameters exhibited promising correlations, particularly with intrinsic reactivity indices such as BDE (referenced in Figure 12, a, b, c), IP (Figure 12, j, k, l), and PA (Figure 12, m, n, o). Generally, lower values in these variables are indicative of better antioxidant activity. The calculated BDE values showed the strongest correlation with scavenging activity in the ABTS assay (R² = 0.81 (B3LYP), R² = 0.88 (M06–2X)), and moderate correlations in the ORAC (R² = 0.49 (B3LYP), R² = 0.28 (M06–2X)) and DPPH (R² = 0.47 (B3LYP), R² = −0.22 (M06–2X)) assays. IP displayed high and nearly identical correlations in the ABTS (R² = 0.69 (B3LYP), R² = 0.67 (M06–2X)) and DPPH (R² = 0.68 (B3LYP), R² = 0.62 (M06–2X)) assays, and a slightly lower correlation in the ORAC assay (R² = 0.54 (B3LYP), R² = 0.52 (M06–2X)). PA demonstrated excellent correlation in both the ORAC (R² = 0.94 (B3LYP), R² = 0.57 (M06–2X)) and DPPH (R² = 0.72 (B3LYP), R² = 0.13 (M06–2X)) assays. The greatest difference was observed between the two calculation methods for determining PA values.

Figure 12.

Correlation between the experimental radical scavenging activity and the indicated physical properties of the hydrazones (B3LYP – blue, M06–2X -red). (a) Activity in ABTS assay as a function of bond dissociation energy (BDE), (b) Activity in ORAC assay as a function of BDE, (c) Activity in DPPH assay as a function of BDE; (d) Activity in ABTS assay as a function of HOMO energy, (e) Activity in ORAC assay as a function of HOMO energy, (f) Activity in DPPH assay as a function of HOMO energy, (g) Activity in ABTS assay as a function of LUMO energy, (h) Activity in ORAC assay as a function of LUMO energy, (i) Activity in DPPH assay as a function of LUMO energy, (j) Activity in ABTS assay as a function of ionization potential (IP), (k) Activity in ORAC assay as a function of IP, (l) Activity in DPPH assay as a function of IP, (m) Activity in ABTS assay as a function of proton affinity (PA), (n) Activity in ORAC assay as a function of PA, (o) Activity in DPPH assay as a function of PA.

The energies of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) also significantly correlated with antioxidant activities. HOMO exhibited the strongest correlation in the DPPH assay (R² = 0.82 (B3LYP), R² = 0.65 (M06–2X)), with slightly lower correlations in the ABTS (R² = 0.67 (B3LYP), R² = 0.67 (M06–2X)) and ORAC (R² = 0.68 (B3LYP), R² = 0.53 (M06–2X)) assays. Meanwhile, LUMO showed the highest correlation in the ORAC assay (R² = 0.9 (B3LYP), R² = 0.80 (M06–2X)) and maintained good correlation in the DPPH assay (R² = 0.72 (B3LYP), R² = 0.58 (M06–2X)). Although these analyses were conducted on a small sample size, they suggest directions for further investigation.

The analysis of the data in Tables 1 and 2 clearly showed that compound 5, a compound that emerged with the best overall performance from the study, exhibited parameters that position this compound in the high activity region in all of the meaningful activity-physicochemical variable relationships (Figure 11). For example, its BDE (73.58 kcal/mol) was in line with the highest activities in Figure 11 (b) and (c), same was true for the HOMO energy (0.1889 hartree, Figure 12 (d), (e) and (f)), LUMO energy (−0.0509 hartree, Figure 12 (h) and (i)), proton affinity (PA, 339.32 kcal/mol, Figure 12 (n) and (o)) and ionization potential (146.83 kcal/mol, Figure 11 (k) and (l)). These positive correlations indicated that the theory was very much in line with the experimental observations that described the antioxidant activity. Based on similar studies we have conducted on hydrazones,³⁰ as well as on phenols, thiophenols and anilines³⁰ it appeared that a selected group of physicochemical descriptors provided a meaningful tool for predicting antioxidant activity of these compounds. It must be mentioned, however, that compound 5 was the best in the studied series only, in addition to showing strong drug like properties, such as high GI absorption, zero Lipinksi-, Ghose-, Veber-, Egan-, and Muegge-violations and zero PAINS alerts, and a relatively modest 2.65 synthetic accessibility score, indicating straightforward synthesis. Thus, these observations could be highly supportive in designing a broad SAR study using compound 5 as a lead compound and using the above physicochemical properties in the iterative design.

To validate the above used DFT methods we have assessed our results obtained with B3LYP and M06–2X approaches by comparing four out of the eight lead compounds to their ab initio BDE, PA, IP, and HOMO, LUMO values. For this validation, we selected the Domain-Based Local Pair Natural Orbital CCSD(T) for single point calculations on the geometries obtained using the M06–2X/6–311++G(d,p), based on prior benchmark studies on antioxidant DFT methods.^60–62 Since H has only one electron, Hartree–Fock (HF) was considered as exact solution (−0.49980 hartree), and for H⁺, since has no electron, the total electronic energy was considered as zero. The chosen four compounds were 1, 3, 4, and 5. Table 1 below includes data compared in our analysis. The mean absolute deviation (MAD) was calculated for all analyzed descriptors according to the following equation.

$$MAD=\frac{1}{N}\sum _{i=1}^N|X_i-X_{ref}|$$ (1)

The comparisons and summary of the MAD results for B3LYP and M06–2X are shown in Figure 13.

Figure 13.

Comparison of the two DFT functionals (B3LYP (green) and M06–2X (magenta)) and the reference data collected with DLPNO–CCSD(T) DEF2-TZVPP DEF2-TZVPP/C//M062X//6–311++(d,p).

When comparing our results to the benchmark study by de Souza et al.,⁶⁰ we found strong agreement (Table S3). The M06–2X method exhibited better overall alignment with ab initio results compared to B3LYP. In terms of trend analysis within our calculations, both B3LYP and M06–2X methods show good alignment with the CCSD(T) trends in most cases. The detailed trends are presented in Figure S105.

CONCLUSIONS

A broad variety of structurally diverse hydrazone derivatives were designed and synthesized to alleviate oxidative stress in trophoblast cells exposed to hydrogen peroxide—an in vitro model of preeclampsia–via potential mitochondrial antioxidant therapy. The compounds represented hydrazones with a variety of electron withdrawing and donating groups in the diaryl hydrazone skeleton, and larger derivatives with more extended delocalized electron structures. The compounds were evaluated in a diverse selection of assays, including antioxidant assays to determine in vitro activity, as well as in cell biology assays to assess the effect of the compounds on human trophoblast cells exposed to oxidative stress. Based on the experimental results, compound 5 showed the best overall results in the biochemical as well as cell-based assays and emerged as a lead compound. The computational analysis was also in agreement with the experimental data. In addition, the in silico prediction of the pharmacokinetic properties of the compounds revealed that compound 5 possesses highly desirable molecular descriptors that suggest that the compound would be able to reach the site of action, in contrast to natural antioxidants, vitamins C and E, that have been tested for antioxidant therapy. It has a nearly ideal lipophilicity (e.g., logP = 4); significantly higher than that of Vitamin C (logP = −2) that is completely water-soluble, that would enable compound 5 to pass through cell membranes. At the same time, it is partially soluble in water, unlike vitamin E (logP = 12) that is completely insoluble in aqueous medium. Pretreatment with this compound reduced mitochondrial-derived ROS production in H₂O₂-exposed trophoblasts cells (HTR8) confirming its mode of action and more importantly reduced sFLT-1 expression which is a key element among biomarkers of PE development. These key data suggest the potential of hydrazones in antioxidant therapy which may have future implications in preeclampsia and hypertensive disorders, which remains to be confirmed in in vivo studies.

EXPERIMENTAL SECTION

Materials and General Procedures.

The aldehydes and hydrazine derivatives, the NMR reference compounds, were all acquired from Aldrich. NMR solvents, including DMSO-d₆ and CDCl₃ (99.8%), were sourced from Sigma-Aldrich and Cambridge Isotope Laboratories. Additional materials and supplies were acquired from Fisher Scientific and Oakwood Chemical.

GC-MS measurements were conducted using an Agilent 6850 gas chromatograph coupled to an Agilent 5973 mass spectrometer system, operating with 70 eV electron impact ionization and equipped with a 30 m DB-5 column (J&W Scientific). The HRMS data were collected using an Agilent 7250 GC-QTOF mass spectrometer, also in electron impact ionization (EI, 70 eV) mode.

The ¹H, ¹³C, and ¹⁹F NMR spectra were recorded on a 400 MHz Agilent MM2 NMR spectrometer using DMSO-d₆ and CDCl₃ as solvents. Tetramethylsilane (TMS) was used as internal standards, or the residual solvent signal served as a reference. Chemical shifts (δ) are reported in ppm. The following abbreviations are used to denote signal multiplicity in ¹H NMR spectra: s (singlet), d (doublet), t (triplet), q (quartet), and m (multiplet). All measurements were conducted at 25 °C with a temperature accuracy of ± 1 °C.

According to the ACS Ethical Guidelines, we confirm that this work does not include animal or human studies.

Synthesis and Characterization.

Synthesis of the Hydrazones – General Procedure.

During the synthesis our earlier procedure was followed.³³ In a 15 mL round-bottom flask, 1 mmol of substituted benzaldehyde and 1 mmol of phenylhydrazine were dissolved in 2 mL of dichloromethane (DCM). The reaction mixture was allowed to stay at room temperature for 10 min then was cooled down to – 20 °C for 30 min. By the end of the cooling period, the hydrazone began to precipitate from the mixture. The crystalline product was filtered and dried at ambient temperature. The products were purified by recrystallization or preparative TLC to obtain sufficient purity, which was verified using GC–MS, and NMR. Depending on the structure of the products (mono vs dihydrazones) either substituted benzaldehydes (Method 1) or terephthaldehyde (Method 2) was used for the synthesis.

Synthesis of 3-(1-(2-Phenylhydrazineylidene)ethyl)aniline (9), 1-(1-(4-Chlorophenyl) ethylidene)-2-phenylhydrazine (10), 1-Phenyl-2-(1-(p-tolyl)ethylidene)hydrazine (15) (Method 1).

To synthesize the acetophenone based hydrazone derivatives, weigh the aldehyde (1 mmol) and sodium acetate (NaCOOCH₃, 1.5 mmol) into a microwave reaction tube with a stir bar. Add 2 mL of ethanol, and then introduce phenylhydrazine (1.5 mmol) dropwise. Seal the tube and run the reaction at 200 W and 80 °C for 15 min while stirring. After cooling to room temperature, evaporate the ethanol using a rotary evaporator. Add 10 mL of diethyl ether to the residue, stir briefly, and filter to remove sodium acetate. Concentrate the filtrate by evaporating the diethyl ether to obtain the crude product, which can be characterized by GC-MS and, if necessary, recrystallized from an ethanol/water mixture.

Synthesis of the 1,4-Bis((E)-(2-(4-(trifluoromethyl)phenyl)-hydrazinylidene)-methyl)benzene (16) (Method 2).

Terepthaldehyde (0.67 g, 0.5 mmol; 0.13 mL) and 4-(trifluoromethyl)phenyl)-hydrazine (0.176 g, 1 mmol) were mixed in 2 mL of DCM. After 30 min at RT and 30 min at – 20 °C the product hydrazone was isolated by filtration and purified by preparative TLC.

Computational Methods.

Density functional theory has been employed to understand the electronic structure of hydrazones. All calculations were carried out at B3LYP⁵⁸ as well as M06–2X level of theories and 6–311++G(d,p) basis set using Gaussian 09 program suite.⁵⁶ These methods were selected based on recent literature.⁵⁴ B3LYP, a widely used method for these systems, allows for comparison with previous studies. While it tends to underestimate energies, it reliably preserves reactivity trends when using the same basis set. Meanwhile, Minnesota functionals are gaining prominence due to their strong performance in thermochemistry and calculating reaction energies involving free radicals.

To find the conformational ensemble, ORCA 6.0^59,62 Global Optimizer Algorithm (GOAT) was applied. Frequency calculations were performed at the same level of theory to confirm the minima on potential energy surfaces. The radical or cationic electronic structures were calculated using the Kohn–Sham formalism with unrestricted spin DFT/UB3LYP. Key thermochemical indices such as bond dissociation energy (BDE), ionization potential (IP), and proton affinity (PA) values were determined (eqs 2–4).

$$BDE=H(Ar-O·)+H(H·)-H(Ar-OH)$$ (2)

$$IP=(Ar-O{H}^{+·})+H(e^{-})-H(Ar-OH)$$ (3)

$$PA=H(Ar-O^{-})+H(H^{+})-H(Ar-OH)$$ (4)

Additionally, further physicochemical and quantum parameters were predicted, including the N–H distance, dipole moment, logP value, and the energies of the HOMO and LUMO orbitals, as well as their gap energy, to explore possible correlations with experimental scavenging activity.

Statistical Analysis.

The activity values shown in Figure 6, 7, 8 were measured four times for the DPPH assay, three times for the ABTS assay and three times for ORAC assay. All data points were included in the statistical analysis. The data in Figure 6, 7, 8 are presented as the mean ± standard deviation (SD).

The Python programming language was used to develop the code for both regression analysis and visualization. The code is publicly available on GitHub at https://github.com/curve-fit/curvefit.git under the repository named curvefit, maintained by BernadetteVlo.

Regression analysis was performed using the method of least squares, and the quality of the fit was assessed using the coefficient of determination (R²), calculated as shown in eq 5:

$$R^2=1-\frac{RSS}{TSS}=1-\frac{{\sum }_{i=1}^n{(y_i-{\widehat{y}}_i)}^2}{{\sum }_{i=1}^n{(y_i-\overline{y})}^2}$$ (5)

Where:

1. RSS: Residual Sum of Squares

2. TSS: Total Sum of Squares

This approach ensures accurate fitting and provides a quantitative measure of how well the model explains the observed data.

Compound Characterization.

The methods described above were applied to synthesize of all other hydrazones. The structure of the molecules was confirmed by using mass spectrometry and ¹H, ¹⁹F, and ¹³C NMR spectroscopy. All compounds were >95% pure by GC and/or HPLC analysis. The spectra are displayed in the Supporting Information. Several organofluorine hydrazones can be found in the literature, and their NMR data including ¹⁹F NMR were in agreement with the earlier spectra.^33,47

(E)-1-((perf luorophenyl)methylene)-2-(4-(trifluoromethyl)-phenyl)hydrazine (1).

¹H NMR (399.822 MHz, CDCl₃) δ (ppm) = 8.08 (s, 1H), 7.74 (s, 1H), 7. 54–7.52 (d, 2H, J= 8 Hz), 7.17–7.14 (d, 2H, J = 12 Hz); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm) = 146.13, 146.01–145.90 (m), 143.58–143.35 (m), 141.97, 139.42–139.20 (m), 136.89–136.61 (m), 128.66–120.57 (q, J = 268.5 Hz), 126.98–126.86 (q, J = 4 Hz), 125.63–125.60 (q, J = 3 Hz), 123.63–122.66 (q, J = 32 Hz), 112.76, 110.30–110.26 (m); ¹⁹F NMR (376 MHz, CDCl₃) δ (ppm): −61.56, −142.82–142.90, −154.27–154.38, −162.23–162.36; HRMS (EI): m/z: [M+] calcd. for C₁₄H₆F₈N₂: 354.0403; found: 354.0382.

(E)-N,N-dimethyl-4-((2-(4-(trifluoromethyl)phenyl)hydrazono)-methyl)aniline (2).

¹H NMR (399.822 MHz, CDCl₃) δ (ppm) = 10.45 (s, 1H), 7.84 (s, 1H), 7.50–7.49 (m, 4H), 7.13–7.11 (d, 2H), 6.73–6.71 (d, 2H), 2.93 (s, 6H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm) = 151.05, 149.08, 140.61, 129.63–121.56 (q, J = 269 Hz), 127.70, 126.83–126.79 (q, J = 4 Hz), 123.31, 118.38–117.43 (q, J = 32 Hz), 112.39, 111.58, 40.23; ¹⁹F NMR (376 MHz, CDCl₃) δ (ppm): −59.16; HRMS (EI): m/z: [M+] calcd. for C₁₆H₁₆F₃N₃ 307.1296; found: 307.1287.

(E)-5-methoxy-2-((2-(4-(trifluoromethyl)phenyl)hydrazono)-methyl)phenol (3).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.66 (s, 1H), 10.50 (s, 1H), 8.17 (d, 1H), 7.53–7.50 (m, 3H), 7.06–7.04 (m, 2H), 6.50–6.46 (m, 2H), 3.74 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 161.38, 157.83, 148.35, 139.89, 128.93, 129.49–121.42 (q, J = 269 Hz), 127.01–126.94 (q, J = 4 Hz), 119.02–118.07 (q, J = 32 Hz), 113.81, 111.55, 106.63, 101.44, 55.53; ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −59.32; HRMS (EI): m/z: [M+] calcd. for C₁₅H₁₃F₃N₂O₂ 310.0929; found: 310.0927.

(E)-1-(3,4-dimethoxyphenyl)-2-(4-(trifluoromethyl)benzylidene)-hydrazine (4).

¹H NMR (399.822 MHz, CDCl₃) δ (ppm) = 10.77 (s, 1H), 8.19 (s, 1H), 7.50–7.28 (m, 4H), 7.08–6.97 (m, 3H), 3.80 (s, 3H), 3.77 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm) = 152.71, 146.70, 145.87, 133.76, 130.22, 130.58–129.64 (q, J = 32 Hz), 128.72, 124.29, 128.48–120.35 (q, J = 271 Hz), 116.52, 115.62, 114.70–114.66 (q, J = 4 Hz), 112.54, 107.66–107.62 (q, J = 4 Hz), 60.90, 55.63; ¹⁹F NMR (376 MHz, CDCl₃) δ (ppm): −61.46; HRMS (EI): m/z: [M+] calcd. for C₁₆H₁₅F₃N₂O₂ 324.1085; found: 324.1091.

(E)-N,N-dimethyl-4-((2-(3-(trifluoromethyl)phenyl)-hydrazinylidene)methyl)aniline (5).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.42 (s, 1H), 7.80 (s, 1H), 7.48–7.45 (m, 4H), 7.09- 7.07 (d, 2H), 6.71–6.69 (d, 2H), 2.91 (s, 6H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 150.66, 149.46, 139.68, 130.54–129.61 (q, J = 32 Hz), 130.15, 128.65–120.52 (q, J = 271 Hz), 127.28, 123.10, 115.30, 113.86–113.82 (q, J = 4 Hz), 112.10, 107.41–107.37 (q, J = 4 Hz), 39.93; ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −61.43; HRMS (EI): m/z: [M+] calcd. for C₁₆H₁₆F₃N₃ 307.1296; found: 307.1292.

(E)-N,N-dimethyl-4-((2-(2-(trifluoromethyl)phenyl)-hydrazinylidene)methyl)aniline (6).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 9.37 (s, 1H), 8.25 (s, 1H), 7.75–7.73 (d, 1H), 7.52–7.50 (m, 4H), 6.87–6.83 (t, 1H), 6.74–6.72 (d, 2H), 2.93 (s, 6H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 150.77, 142.99, 142.63, 133.46, 128.75–120.63 (q, J = 271 Hz), 127.46, 126.12–125.96 (q, J = 5 Hz), 122.93, 117.61, 114.39, 112.01, 110.97–110.08 (q, J = 120 Hz), 39.85; ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −59.98; HRMS (EI): m/z: [M+] calcd. for C₁₆H₁₆F₃N₃ 307.1296; found: 307.1274.

(E)-1-(4-nitrobenzylidene)-2-(3-(trifluoromethyl)phenyl)-hydrazine (7).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm) = 11.16 (s, 1H), 8.25–8.23 (d, 2H), 8.00 (s, 1H), 7.94–7.91 (d, 2H), 7.50–7.46 (t, 1H), 7.39–7.37 (m, 2H), 7.15–7.13 (d, 1H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm) = 146.90, 145.58, 142.39, 136.08, 130.86, 130.68–130.37 (q, J = 32 Hz), 126.97, 124.53, 126.08–123.37 (q, J = 271 Hz), 116.66, 116.28–116.24 (q, J = 4 Hz), 108.71–108.67 (q, J = 4 Hz); ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −61.39; HRMS (EI): m/z: [M+] calcd. for C₁₄H₁₀F₃N₃O₂ 309.0725; found: 309.0656.

(E)-1-(4-methoxyphenyl)-2-(4-nitrobenzylidene)hydrazine (8).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.76 (s, 1H), 8.21–8.18 (m, 2H), 7.86–7.81 (m, 3H), 7.10–7.07 (d, 2H), 6.89–6.6.88 (d, 2H), 3.70 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 153.88, 146.07, 143.37, 138.73, 132.67, 126.09, 124.52, 115.15, 114.06, 55.70; HRMS (EI): m/z: [M+] calcd. for C₁₄H₁₃N₃O₃ 271.0956; found: 271.0948.

(E)-3-(1-(2-phenylhydrazinylidene)ethyl)aniline (9).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 9.14 (s, 1H), 7.24–7.19 (m, 4H), 7.08 (s, 1H), 7.04–7.01 (t, 1H), 6.93–6.91 (d, 1H), 6.76–6.73 (t, 1H), 6.53–6.51 (d, 1H), 5.08 (s, 2H), 2.18 (s, 3H); ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 148.48, 146.29, 141.27, 139.89, 128.87, 128.67, 118.65, 113.66, 113.46, 112.75, 110.70, 13.00; HRMS (EI): m/z: [M+] calcd. for C₁₄H₁₅N₃ 225.1266; found: 225.1267.

(E)-1-(4-chlorophenyl)ethylidene)-2-phenylhydrazine (10).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 9.35 (s, 1H), 7.80–7.78 (d, 2H), 7.42–7.40 (d, 2H), 7.24 (m, 4H), 6.77 (s, 1H), 2.23 (s, 3H); ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 145.89, 139.17, 138.08, 131.99, 128.93, 128.23, 126.78, 119.05, 112.87, 12.66; ¹⁹F NMR (282 MHz, DMSO-d₆) δ (ppm): −59.81; HRMS (EI): m/z: [M+] calcd. for C₁₄H₁₃ClN₂ 244.0767; found: 244.0770.

(E)-(4-methoxyphenyl)-2-(1-(4-nitrophenyl)ethylidene)hydrazine (11).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 9.54 (s, 1H), 8.20–8.18 (d, 2H), 7.99–7.97 (d, 2H), 7.24–7.22 (d, 2H), 6.87–6.85 (d, 2H), 3.69 (s, 3H), 2.25 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 153.80, 146.13, 146.11, 139.63, 137.05, 125.87, 124.07, 114.91, 114.73, 55.66, 12.82; HRMS (EI): m/z: [M+] calcd. for C₁₅H₁₅N₃O₃ 285.1113; found: 285.1102.

(E)-(2,5-dimethylphenyl)-2-(1-(3-nitrophenyl)ethylidene)-hydrazine (12).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 8.53 (s, 1H), 8.26–8.24 (d, 1H), 8.18 (s, 1H), 8.16–8.14 (d, 1H), 7.71–7.67 (t, 1H), 7.30 (s, 1H), 6.99–6.97 (d, 1H), 6.62–6.60 (d, 1H), 2.36 (s, 3H), 2.28 (s, 3H). 2.25 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 148.25, 142.93, 140.93, 140.39, 135.73, 131.53, 130.46, 130.07, 122.13, 120.90, 119.78, 119.42, 114.20, 21.31, 16.88, 12.51.; HRMS (EI): m/z: [M+] calcd. for C₁₆H₁₇N₃O₂ 283.1320; found: 283.1307.

(E)-5-methoxy-2-(1-(2-phenylhydrazinylidene)ethyl)phenol (13).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.92 (s, 1H), 10.21 (s, 1H), 8.06 (s, 1H), 7.40–7.38 (d, 1H), 7.23–7.19 (m, 2H), 6.91-6.89 (m, 2H), 6.74–6.71 (t, 1H), 6.47–6.45 (m, 2H), 3.73 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 160.91, 157.87, 145.36, 138.79, 129.70, 129.20, 119.05, 113.89, 111.93, 106.44, 101.49, 55.59; HRMS (EI): m/z: [M+] calcd. for C₁₄H₁₄N₂O₂ 242.1055; found: 242.1049.

(E)-(2,5-dimethylphenyl)-2-(1-(4-nitrophenyl)ethylidene)-hydrazine (14).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 8.32 (s, 1H), 8.25–8.23 (d, 2H), 8.06–8.04 (d, 2H), 7.33 (s, 1H), 7.00–6.99 (d, 1H), 6.64–6.62 (d, 1H), 2.35 (s, 3H), 2.28 (s, 3H), 2.25 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 146.63, 145.81, 142.98, 140.47, 136.16, 130.82, 126.38, 124.13, 121.60, 120.31, 114.78, 21.61, 17.18, 12.63.; HRMS (EI): m/z: [M+] calcd. for C₁₆H₁₇N₃O₂ 283.1320; found: 283.1316.

(E)-1-phenyl-2-(1-(p-tolyl)ethylidene)hydrazine (15).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 9.22 (s, 1H), 7.70–7.68 (m, 2H), 7.25–7.17 (m, 6H), 6.75 (m, 1H), 2.30 (s, 3H), 2.23 (s, 3H); ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 146.23, 140.59, 136.79, 136.54, 128.88, 125.10, 118.70, 112.76, 20.76, 12.81; HRMS (EI): m/z: [M+] calcd. for C₁₅H₁₆N₂ 224.1313; found: 224.1315.

1,4-Bis((E)-(2-(3-(trifluoromethyl)phenyl)hydrazinylidene)-methyl)benzene (16).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.86 (s, 2NH), 7.95 (s, 2H), 7.69 (m, 4H), 7.45–7.41 (t, 2H), 7.37 (s, 2H), 7.34–7.32 (d, 2H), 7.05–7.03 (d, 2H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 146.21, 138.33, 135.71, 130.61, 130.95–130.01 (q, J = 32 Hz), 126.65, 128.89–120.75 (q, J= 271 Hz), 116.10, 115.17–115.13 (q, J = 4 Hz), 108.22–108.19 (q, J = 4 Hz); ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −61.44; HRMS (EI): m/z: [M+] calcd. for C₂₂H₁₆F₆N₄ 450.1279; found: 450.1254.

1,4-Bis((E)-(2-(4-methoxyphenyl)hydrazinylidene)methyl)benzene (17).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.16 (s, 2H), 7.77 (d, 2H), 7.58 (m, 4H), 7.01- 6.99 (d, 4H), 6.84–6.82 (d, 4H), 3.68 (s, 6H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 153.07, 139.69, 135.67, 135.48, 126.05, 115.09, 113.43, 55.69; HRMS (EI): m/z: [M+] calcd. for C₂₂H₂₂N₄O₂ 374.1742; founds: 374.1740.

1,4-Bis((E)-(2-(3-nitrophenyl)hydrazinylidene)methyl)benzene (18).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.94 (s, NH), 7.97 (s, 2H), 7.87 (s, 2H), 7.75 (s, 4H), 7.61–7.59 (d, 2H), 7.54–7.46 (m, 4H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 148.85, 146.28, 138.65, 135.27, 130.51, 126.42, 118.27, 113.01, 105.70; HRMS (EI): m/z: [M+] calcd. for C₂₀H₁₆N₆O₄ 404.1233; found: 404.1732.

(E)-3-(1-(2-(2,4-difluorophenyl)hydrazinylidene)ethyl)-7-methoxy-2H-chromen-2-one (19).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 8.72 (s, 1H), 8.18 (s, 1H), 7.76–7.74 (d, 1H), 7.57–7.51 (q, 1H), 7.25–7.20 (t, 1H), 7.02–6.96 (m, 3H), 3.86 (s, 3H), 2.24 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 162.53, 159.66, 156.67–156.56 (d, J = 236 Hz), 155.13, 150.72–150.60 (d, J = 244 Hz), 143.65, 140.45, 130.91–130.84 (J = 7 Hz), 130.06, 123.48, 116.24–116.10 (d, J = 9 Hz), 112.77, 112.67, 111.31–111.06 (d, J = 21.5 Hz), 104.11–103.61 (d, J = 27.5 Hz), 100.26, 55.99, 15.21; ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −122.61–122.67, −127.59–127.65; HRMS (EI): m/z: [M+] calcd. for C₁₈H₁₄F₂N₂O₃ 344.0972; found: 344.0996.

(E)-7-methoxy-3-(1-(2-(4-methoxyphenyl)hydrazinylidene)ethyl)-2H-chromen-2-one (20).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 9.21 (s, 1H), 8.10 (s, 1H), 7.75–7.73 (d, 1H), 7.18–7.16 (d, 2H), 7.00 (s, 1H), 6.97–6.94 (d, 1H), 6.83–6.80 (d, 2H), 3.85 (s, 3H), 3.68 (s, 3H), 2.17 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 162.60, 160.33, 155.28, 140.19, 139.68, 138.25, 130.24, 124.35, 114.70, 114.50, 113.31, 113.07, 100.63, 56.38, 55.63, 15.56; HRMS (EI): m/z: [M+] calcd. for C₁₉H₁₈N₂O₄ 338.1266; found: 338.1262.

(E)-1-(naphthalen-2-ylmethylene)-2-(3-nitrobenzyl)hydrazine (21).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.95 (s, 1H), 8.09 (s, 1H), 8.03–7.99 (m, 2H), 7.93- 7.87 (m, 4H), 7.58–7.49 (m, 5H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 149.29, 146.79, 139.59, 133.53, 133.52, 133.32, 130.94, 128.81, 128.44, 128.19, 127.28, 127.08, 126.89, 122.84, 118.70, 113.39, 106.10; HRMS (EI): m/z: [M+] calcd. for C₁₇H₁₃N₃O₂ 291.1007; found: 291.0994.

(E)-1-(anthracen-9-ylmethylene)-2-(2-(trifluoromethyl)phenyl)hydrazine (22).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.13 (s, 1H), 9.64 (s, 1H), 8.82–8.79 (d, 2H), 8.63 (s, 1H), 8.13–8.11 (d, 2H), 7.77–7.75 (d, 1H), 7.65–7.54 (m, 6H), 6.98–6.94 (t, 1H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 142.50, 140.66, 133.88, 131.09, 129.34, 128.96, 128.43, 126.80, 128.70–120.57 (q, J = 271 Hz), 126.37–126.26 (q, J = 6 Hz), 126.24, 125.47, 125.08, 118.75, 114.32, 111.51–110.61 (q, J = 30 Hz); ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −59.81; HRMS (EI): m/z: [M+] calcd. for C₂₂H₁₅F₃N₂ 364.1187; found: 364.1178.

(E)-1-(naphthalen-2-ylmethylene)-2-phenylhydrazine (23).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.46 (s, 1H), 8.01–7.88 (m, 6H), 7.53–7.46 (q, 2H), 7.25–7.21 (t, 2H), 7.12–7.10 (d, 2H), 6.77–6.74 (t, 1H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 145.23, 136.45, 133.64, 133.22, 132.80 129.17, 128.22, 127.86, 127.72, 126.54, 126.08, 125.80, 122.50, 118.85, 112.05; HRMS (EI): m/z: [M+] calcd. for C₁₇H₁₄N₂ 246.1157; found: 246.1142.

(E)-1-(2,5-dimethylphenyl)-2-(naphthalen-2-ylmethylene)hydrazine (24).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 9.62 (s, 1H), 8.29 (s, 1H), 8.02–7.88 (m, 5H), 7.53–7.46 (q, 2H), 7.32 (s, 1H), 6.94–6.92 (d, 1H), 6.54–6.52 (d, 1H), 2.28 (s, 3H), 2.19 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 143.29, 138.32, 136.13, 134.17, 133.66, 133.28, 130.69, 128.70, 128.32, 128.15, 126.95, 126.52, 126.30, 122.96, 120.09, 118.21, 113.14, 21.67, 17.59; HRMS (EI): m/z: [M+] calcd. for C₁₉H₁₈N₂ 274.1470; found: 274.1455.

(E)-1-(naphthalen-2-ylmethylene)-2-(3-(trifluoromethyl)phenyl)hydrazine (25).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 11.02 (s, 1H), 9.10 (s, 1H), 8.68–8.66 (d, 2H), 8.63 (s, 1H), 8.14–8.11 (d, 2H), 7.62–7.53 (m, 4H), 7.50–7.46 (t, 1H), 7.35–7.33 (m, 2H), 7.10–7.08 (d, 1H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 145.85, 138.43, 133.13–133.11 (q, J = 2 Hz), 133.04, 130.53–129.60 (q, J = 31 Hz), 130.28, 128.34, 128.49–120.36 (q, J = 271 Hz), 127.97, 127.75, 126.60, 126.34, 122.46, 115.72, 114.80–114.76 (q, J = 4 Hz), 107.82–107.74 (q, J = 4 Hz); ¹⁹F NMR (376 MHz, DMSO-d₆) δ (ppm): −61.39; HRMS (EI): m/z: [M+] calcd. for C₂₂H₁₅F₃N₂ 364.1187; found: 364.1190.

(E)-1-(anthracen-9-ylmethylene)-2-(4-nitrophenyl)hydrazine (26).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 11.61 (s, 1H), 9.25 (s, 1H), 8.69–8.67 (m, 3H), 8.19–8.13 (two d, 4H), 7.66–7.55 (two t, 4H), 7.22–7.20 (d, 2H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 150.51, 140.66, 138.57, 131.02, 129.38, 129.04, 128.94, 127.09, 126.36, 125.51, 124.75, 111.27; HRMS (EI): m/z: [M+] calcd. for C₂₁H₁₅N₃O₂ 341.1164; found: 341.1166.

(E)-1-(anthracen-9-ylmethylene)-2-(2-nitrophenyl)hydrazine (27).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 11.64 (s, 1H), 9.74 (s, 1H), 8.83–8.81 (d, 2H), 8.68 (s, 1H), 8.17–8.13 (m, 3H), 7.93–7.90 (d, 1H), 7.68–7.55 (m, 5H), 6.96- 6.94 (t, 1H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 144.70, 141.82, 137.07, 131.45, 131.25, 130.01, 129.63, 129.41, 127.49, 126.29, 126.14, 125.98, 125.54, 118.83, 116.16; HRMS (EI): m/z: [M+] calcd. for C21H15N3O2 341.1164; found: 341.1153.

(E)-3-((2-phenylhydrazineylidene)methyl)-1H-indole (28).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 11.37 (s, 1H), 9.86 (s, 1H), 8.33 (s, 1H), 8.17 (s, 1H), 7.65 (s, 1H), 7.45 (s, 1H), 7.25–7.22 (m, 4H), 7.11–7.09 (m, 2H), 6.71 (m, 1H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 146.24, 137.07, 134.99, 129.15, 127.33, 124.31, 122.33, 121.73, 120.07, 117.57, 112.88, 111.74, 111.44; HRMS (EI): m/z: [M+] calcd. for C₁₅H₁₃N₃ 235.1109; found: 235.1102.

(E)-1-(4-methoxyphenyl)-2-(naphthalen-2-ylmethylene)hydrazine (29).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.27 (s, 1H), 7.98–7.89 (m, 6H), 7.52–7.45 (q, 2H), 7.08–7.05 (d, 2H), 6.87–6.85 (d, 2H), 3.70 (s, 3H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 153.12, 139.71, 135.70, 134.31, 133.69, 133.07, 128.58, 128.19, 128.12, 126.92, 126.32, 125.74, 122.90, 115.10, 113.49, 55.70; HRMS (EI): m/z: [M+] calcd. for C₁₈H₁₆N₂O 276.1262; found: 276.1248.

(E)-1-(4-chlorophenyl)-2-(naphthalen-2-ylmethylene)hydrazine (30).

¹H NMR (399.822 MHz, DMSO-d₆) δ (ppm): 10.58 (s, 1H), 8.00 (s, 1H), 7.96–7.95 (m, 2H), 7.90–7.86 (t, 3H), 7.51–7.44 (q, 2H), 7.25–7.23 (d, 2H), 7.10–7.08 (d, 2H); ¹³C NMR (DMSO-d₆, 100 MHz) δ (ppm): 144.59, 137.78, 133.78, 133.59, 133.32, 129.39, 128.67, 128.33, 128.15, 127.00, 126.63, 126.55, 122.91, 122.47, 113.90; HRMS (EI): m/z: [M+] calcd. for C₁₇H₁₃ClN₂ 280.0767; found: 280.0756.

Biochemistry - Radical Scavenging Assays.

The radical scavenging assays have been carried out as described in our earlier publication [30].

2.6.1. DPPH Radical Scavenging Assay.

The DPPH assay was conducted following earlier procedures.³⁰ A VersaMax UV–vis plate reader with the SoftMax Pro 5 software (Molecular Devices) were used to determine the scavenging of the DPPH radical by the compounds investigated in comparison to the reference compound Trolox. The data was processed using eq 6 below where Abs_c is the absorbance of the control and Abs_t is the absorbance of the test sample.

$$PercentRadicalScavenging\text{=}\frac{(Ab{s}_c-Ab{s}_t)}{Ab{s}_c}\times 100$$ (6)

The 60 min percent radical scavenging values were standardized against the 60 min Trolox percent radical scavenging value to obtain the final Trolox equivalent values for each compound as shown below (eq 7).

$$Trolox\mathit{Equivalents}=\frac{PercentRadicalScavengin{g}_{Sample}}{PercentRadicalScavengin{g}_{Trolox}}$$ (7)

ABTS Radical Scavenging Assay.

Following an earlier procedure, the ABTS assay was carried out to determine the radical scavenging activity of these hydrazones.³⁰ The ABTS radical was generated 12 to 24 h before the assay and a VersaMax UV–vis plate reader with the SoftMax Pro 5 software (Molecular Devices) were used to assess the scavenging of the ABTS radical by the compounds. The data was processed using eq 5.

Similarly to the DPPH assay, Trolox was used as a reference antioxidant and the radical scavenging values were standardized against Trolox percent radical scavenging value to obtain the final Trolox equivalents values for each compound using eq 6.

ORAC Radical Scavenging Assay.

A SpectraMax i3x UV–Vis/fluorescence plate reader was used for these assays with the SoftMax Pro 5 software (Molecular Devices) as described earlier.³⁰ The data was processed using the equations below (eqs 8 and 9), the parameters are described below.

$$NetAUC=0.5\sum _{0-29}\frac{f_i}{f_0}+\left(0.5*\frac{f_{30}}{f_0}\right)$$ (8)

f_i is the fluorescence intensity at reading 0–29. f₀ is the fluorescence at reading zero. f₃₀ is the fluorescence at time 30.

$$PercentRadicalScavenging\text{=}\frac{(NetAU{C}_t-NetAU{C}_c)}{(NetAU{C}_{f\_\max }-NetAU{C}_c)}\times 100$$ (9)

The Net AUC_c is the net area under the curve for the control sets with AAPH and fluorescein with the appropriate amount of DMSO in ethanol. The Net AUC_t is the area under the curve for the test sample while the Net AUC_{f_max} is the area under the curve for the maximum fluorescence sample where no AAPH was added. A Trolox dilution curve was used for standardization to obtain the Trolox equivalent concentration for each compound, similarly to the above assays.

The percent radical scavenging values of different concentrations of Trolox (100 μM, 50 μM, 25 μM, 12.5 μM, 6.25 μM, and 3.125 μM) calculated using eq 4 were plotted versus the natural logarithm of the Trolox concentration values. A calibration curve was constructed using those plotted values.⁶³ The following equation (eq 10) was used to calculate the Trolox Equivalent Concentration of the samples.

$$TroloxEquivalentConcentration=e^{\left({PercentRadicalScavenging}_{Sample}-b\right)/m}$$

(eq 9)b (constant) and m (coefficient) are from the trendline equation, y = m(ln(x))+b (x: Trolox concentration, y: Percent Radical Scavenging_Trolox).

Biological Evaluation.

Hydrazone Treatment to Reduce Oxidative Stress in Trophoblast Cells.

Human trophoblast HTR-8/SVneo cells were cultured in RPMI medium supplemented with 5% fetal bovine serum and 1% penicillin-streptomycin in a humidified incubator containing 5% CO₂ at 37 °C. Cells in the H₂O₂-treated (“Control”) group were treated with 0.1–50 μM H₂O₂ alone. Cells in the hydrazone + H₂O₂-treated (“Treated”) groups were pretreated with 0.1–50 μM HMP for 30 min, respectively, and then they were treated with 100 μM H₂O₂ for 20 h including hydrazones in the cell culture media for the entire time. Cell culture supernatants were collected at the end of the experiment and stored at −20 °C until assayed.

Enzyme-Linked Immunosorbent Assay (ELISA).

Soluble FLT-1 (sFLT-1) in culture medium was measured by ELISA using the human VEGF receptor 1 (VEGFR1) Quantikine kit (R&D Systems) following manufacturer’s instructions. This assay has an intra-assay assay coefficient of variation of 2.6 – 3.8% and an interassay coefficient of variation of 5.5 – 9.8%.⁵²

MitoSOX Red Immunofluorescence Assay.

Immunofluorescence assay was carried out to determine the mitochondrial-derived superoxide production in HTR-8/SVneo cells as described earlier.⁶¹ The images were produced from fluorescence microscopy for MitoSOX Red, with an original magnification of 20× using ImageJ software v. 1.53t (National Institute of Health [NIH], Bethesda, MD, USA; http://imagej.nih.gov/ij).

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.5c00010.

Detailed description of the experimental data, including synthesis, biochemical, biological assays, computational details and data, compound characterization (HRMS and NMR spectra); all experimental antioxidant activity vs physicochemical parameter plots (PDF)

Molecular formula strings and all experimental antioxidant and cell biology data (CSV)

ABBREVIATIONS USED

BDE

Bond dissociation energy

DP

Dipole moment

HF

Hartree–Fock

HOMO

Highest occupied molecular orbital

IP

Ionization potential

LUMO

Lowest unoccupied molecular orbital

PA

Proton affinity

logP

Partition coefficient

PE

Preeclampsia

sFLT-1

Soluble fms-like tyrosine kinase 1