Prenylated phloroglucinols from Hypericum scruglii, an endemic species of Sardinia (Italy), as new dual HIV-1 inhibitors effective on HIV-1 replication

In a search for new potential multitarget anti-HIV compounds from natural products, we have identified in Hypericum scruglii, an endemic and exclusive species of Sardinia (Italy), a potent plant lead. The phytochemical study of the hydroalcoholic extract obtained from its leaves led to the isolation of its most abundant secondary metabolites, belonging to different chemical classes. In particular, three phloroglucinols derivatives were identified, confirming their significance as chemotaxonomic markers of the Hypericum genus. Among them, the 3-(13-hydroxygeranyl)-1-(2'-methylbutanoyl)phloroglucinol was reported here for the first time. All six isolated compounds have been evaluated firstly for the inhibition of both Human Immunodeficiency Virus type 1 (HIV-1) Reverse Transcriptase (RT)-associated DNA Polymerase (RDDP) and Ribonuclease H (RNase H) activities, for the inhibition of HIV-1 integrase (IN) in biochemical assays, and also for their effect on viral replication. Among the isolated metabolites, three phloroglucinol derivatives and quercitrin were effective on both RT-associated RDDP and RNase H activities in biochemical assays. The same active compounds affected also HIV-1 IN strand transfer function, suggesting the involvement of the RNase H active site. Furthermore, phloroglucinols compounds, included the newly identified compound, were able to inhibit the HIV-1 replication in cell based assays.


Introduction
Natural products have played, and will continue to play, a key role in drug discovery. In particular, the diversity of plant-based systems has provided an enormous number of lead compounds a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 in healthcare [1]. Indeed, plant products represent, according to an assessment of FDA on the source of natural products, over one-quarter of all approved new molecular entities [2,3]. However, despite the intensive investigation of plant kingdom, it is estimated that only 6% of the approximately 300,000 species of higher plants have been pharmacologically investigated, and only 15% phytochemically [4]. Therefore, plants should be further investigated because new compounds with original structures and novel modes of action are continuously required. Naturally occurring compounds frequently inspire synthetic medicinal compounds, and they could be chemically modified, based upon their structural and biological properties [5][6][7][8]. Their structural modification allows increasing their efficacy and selectivity, improving physicochemical, biochemical and pharmacokinetic properties, removing or reducing side effects.
The therapeutic area of infectious diseases has benefited from abundant scaffold diversity in natural products, able to interact with many specific targets [7]. Significant research and development over the last 25 years into antiviral drug discovery has resulted in the identification of important antiviral drugs [7]. In particular, a number of attempts have been made in the fight against HIV-1 infection and several natural compounds able to inhibit the viral enzymes have been reported [9][10][11][12][13][14][15][16][17]. However, so far all anti HIV-1 approved drugs were obtained only by chemical synthesis.
HIV-1, the etiological agent of AIDS, still remains a global scourge despite the availability of more than 30 approved anti-AIDS drugs [18]. Although the global scale-up of antiretroviral therapy has contributed to reduce the number of new infections and AIDS-related deaths, about 37 million people were estimated to be infected with HIV in 2016, with 1.8 million of new infections and 1 million of deaths [19]. To date there is no vaccine or cure for HIV infection, and the efficacy of antiretroviral therapy, which combines two or three antiviral agents, targeting different steps of the virus replication cycle, can be compromised by the selection of strains resistant to one or multiple drug classes [20,21] and treatment-associated toxicity [22], requiring the discovery of new antiviral agents with innovative modes of action or targets. In this respect, the identification of one molecule able to inhibit more than one viral function would provide significant advantages, raising the genetic barrier to resistance and increasing the compliance to therapy.
Five different classes of anti-AIDS approved molecules are available for therapy [18] and the majority of them is represented by inhibitors of reverse transcriptase (RT), the enzyme responsible for the conversion of the single-stranded RNA genome into a double-stranded cDNA [23,24]. RT is a multifunctional enzyme with two associated functions [25], DNA polymerase and RNase H activities [26,27], that have been proven to be both essential for viral replication. While the first one is currently the main target for AIDS treatment, the latter is the only HIV enzymatic function not targeted by approved antiviral drugs [26,28,29], although it is a very promising target [30]. Indeed it has been shown that RNase H inactivation lead to non-infectious virions [ [17,26,[45][46][47][48][49]. This innovative strategy could offer the possibility to reduce the toxicity associated to the coadministration of several classes of drugs [18,40].
In particular, in this study we focus on Hypericum scruglii Bacchetta, Brullo et Salmeri [61,62], a species exclusive to Sardinia island (Italy). Despite the large number of Hypericum species, only H. perforatum L. has been intensively investigated, both chemically and pharmacologically. Commonly known as St. John's wort, it is widely used in Europe as a drug for the treatment of mild to moderate depression [63,64]. When compared to H. perforatum, few studies have been undertaken on the other members of this genus although their recognized pharmacological properties range from wound healing and antiseptics to antiviral, antiinflammatory, anticancer, antioxidants, antifungal, antimicrobial, cardioprotective and cytotoxic activities [14,[51][52][53][65][66][67][68][69]. Some Hypericum species also exhibited anti HIV-1 properties [14,70,71]. This genus is known for the production of a broad spectrum of secondary metabolites, mainly naphthodianthrones (hypericin and pseudohypericin), phloroglucinols (hyperforin and adhyperforin), phenolic acids, flavonoids (hyperoside, rutin or quercitrin), xanthones and essential oils [14,[72][73][74][75]. Although H. scruglii was not already characterized in terms of phytochemical composition and biological/pharmacological properties, recently Mandrone et al. [53] have identified from this species shikimic and chlorogenic acids, two known phlorogucinols derivatives, quercitrin, hyperoside and hypericin, even though in a very low content, confirming their chemotaxonomic significance [76]. They have also described the antioxidant and α-glucosidase inhibitory activities.
In the present study, we investigated the ability of the main compounds isolated from leaves of Hypericum scruglii to inhibit both HIV-1 RDDP and RNase H activities in biochemical assays. Active compounds were then assayed for their effects on HIV-1 IN activities and to interfere with the HIV-1 life cycle.

Plant material
Aerial parts of H. scruglii were collected at the flowering stage (June 2012) in the site of Sant'Antonio (Jerzu, Sardinia, Italy, 39˚45'57.4"N 9˚30'41.8"E). The leaves were randomly harvested from 30 individuals of the same population. No flowers, fruits, seeds and roots were collected to avoid damage to the population. The species was botanically identified by C.S. and a voucher specimen (CAG 239/c) was deposited at the General Herbarium of the Department of Life and Environmental Sciences, University of Cagliari. Although H. scruglii is endemic, it is not protected by local or international regulations. Furthermore, the location where the plant material was harvested is not included in national or local parks or any other natural protected areas. Therefore, no specific permission was required for its collection.

Chemicals and instruments
Reagents and solvents were purchased from Sigma-Aldrich Chemical Company (St. Louis, MO, USA). The reagents used for expression, purification and biochemical assays were purchased from Microbiol (Sardinia, Italy), Sigma-Aldrich (Milano, Italy) and PerkinElmer (Milano, Italy). The reference compound raltegravir was purchased from ChemScene (Monmouth Junction, United States) while the reference compound RDS1759 was provided from a chemist collaborator Prof. Roberto Di Santo (University of Rome La Sapienza). The isolation of the metabolites was conducted with the aid of distinct chromatographic techniques. The thin-layer chromatographies (TLC) were carried out on plates impregnated silica Merck Kieselgel 60 F 254 thickness 0.20 mm for analytical purposes and on plates impregnated silica Merck Kieselgel 60 F 254 thickness 0.5 o 1.0 mm for preparative purposes. The bands were visualized by spraying with the solution H 2 SO 4 -AcOH-H 2 O (1:20:4) and subsequent heating in the stove for 5 min at 120˚C. The column chromatographies (CC) were performed using either silica gel Merck Kieselgel 60 (70-240 mesh), Amberlite XAD-4 (20-50 mesh; Fluka) and XAD-7 (20-50 mesh; Fluka), Sephadex LH-201 (Pharmacia) or Bakerbond C8 and C18 as stationary phase.
NMR spectra were recorded at 25˚C and 300 MHz for 1 H and 75 MHz for 13 C on a Varian NMR spectrometer FT-300. Methanol-d 4 was used as internal lock.
Correlation spectroscopy (COSY) and double quantum filtered COSY (DQF-COSY) spectra were recorded with gradient enhanced sequence at spectral widths of 3000 Hz in both f2 and f1 domains; the relaxation delays were of 1.0 s. The total correlation spectroscopy (TOCSY) experiments were performed in the phase-sensitive mode with a mixing time of 90 ms. The spectral width was 3000 Hz. For all the homonuclear experiments, the initial matrix of 512 × 512 data points was zero-filled to give a final matrix of 1 k × 1 k points.
Proton-detected heteronuclear correlations were measured. Heteronuclear single-quantum coherence (HSQC) experiments (optimized for 1 J (H,C) = 140 Hz) were performed in the phase sensitive mode with field gradient; the spectral width was 12000 Hz in f1 ( 13 C) and 3000 Hz in f2 ( 1 H) and 1.0 s of relaxation delay; the matrix of 1 k × 1 k data points was zero-filled to give a final matrix of 2 k × 2 k points. Heteronuclear 2 bond correlation (H2BC) spectra were obtained with T = 30.0 ms, and a relaxation delay of 1.0 s; the third-order low-pass filter was set for 130 < 1 J (C,H) < 165 Hz. Heteronuclear multiple bond coherence (HMBC) experiment (optimized for n J (H,C) = 8 Hz) was performed in the absolute value mode with field gradient; typically, 1 H-13 C gHMBC were acquired with spectral width of 18000 Hz in f1 ( 13 C) and 3000 Hz in f2 ( 1 H) and 1.0 s of relaxation delay; the matrix of 1 k × 1 k data points was zero-filled to give a final matrix of 4 k × 4 k points. Constant time inverse-detection gradient accordion rescaled heteronuclear multiple bond correlation spectroscopy (CIGAR-HMBC) spectra (8> n J (H,C) >5) were acquired with the same spectral width used for HMBC. Heteronuclear singlequantum coherence-total correlation spectroscopy (HSQC-TOCSY) experiments were optimized for n J (H,C) = 8 Hz, with a mixing time of 90 ms.
LC-MS analysis was carried out on an Alliance 2695 separation module equipped with a column heater and a sample chiller. The liquid chromatography system was coupled to a Waters 2487 dual wavelength UV detector and to a Quattro Micro™ triple quadrupole mass spectrometer (Waters/Micromass, Manchester, UK).
Recombinant proteins were purified using the chromatography system Biological LP (Biorad). Biochemical assays were measured using the multiplate reader Victor 3 (Perkin Elmer).

Extraction and isolation of active compounds
Plant material (70.0 g) was extracted by sonication with a solution of H 2 O:MeOH (1:1), immersed in an ultrasonic bath (Elma1Transonicdigitals) for 40 min. Subsequently, samples were filtered and the obtained crude extract was solubilized in H 2 O and then subjected to liquid-liquid extraction using ethyl acetate (EtOAc) as extracting solvent. The aqueous fraction was chromatographed on Amberlite XAD-4 and XAD-7, eluting first with water and then with methanol. The organic fraction (10.0 g) was chromatographed on silica gel (SiO 2 CC) using CHCl 3 /MeOH solutions as eluent to afford 13 fractions. Fraction 1, eluted in chloroform, was chromatographed through Sephadex LH-20, using n-hexane/CHCl 3 /MeOH (2:1:1) as eluent solution; four fractions (A-D) were obtained. Fraction A was chromatographed by semi preparative TLC (0.5 mm) using CHCl 3 /MeOH (1:19) as eluent. The compounds 3 was obtained. From fraction B compounds 1 and 2 were isolated. Fraction 2, eluted with CHCl 3 /MeOH (95:5), chromatographed by RP-18 CC using decreasing polarity H 2 O/MeOH solutions, led to compounds 4 and 5.

RNase H polymerase-independent cleavage assay
The HIV-1 RT-associated RNase H activity was measured in 100 μL reaction volume containing 50 mM Tris-HCl, pH 7.

Homogeneous time resolved fluorescence (HTRF) LEDGF dependent assay
The IN LEDGF/p75 dependent assay allowed to measure the inhibition of the 3'processing and strand transfer IN reactions in the presence of recombinant LEDGF/p75 protein, as previously described [80]. Briefly, 50 nM IN was preincubated with increasing concentration of compounds for 1 hour at room temperature in reaction buffer containing 20 mM HEPES pH 7.5, 1 mM DTT, 1% Glycerol, 20 mM MgCl 2 , 0.05% Brij-35 and 0.1 mg/ml BSA. To this mixture, 9 nM DNA donor substrate (5'-ACAGGCCTAGCACGCGTCG-Biotin-3' annealed with 5'-CGACGCGTGGTAGGCCTGT-Biotin3') and 50 nM DNA acceptor substrate (5'-Cy5-AT GTGGAAAATCTCTAGCAGT-3' annealed with 5'-Cy5-TGAGCTCGAGATTTTCCACAT-3') and 50 nM LEDGF/p75 protein were added and incubated at 37˚C for 90 minutes. After the incubation, 4 nM of Europium-Streptavidine were added at the reaction mixture and the HTRF signal was recorded using a Perkin Elmer Victor 3 plate reader using a 314 nm for excitation wavelength and 668 and 620 nm for the wavelength of the acceptor and the donor substrates emission, respectively.

Antiviral activity and cell toxicity. Phenotypic analyses with fully replicating recombinant HIV-1 strain
The human TZM-bl indicator cell line was obtained from the American Type Culture Collection (Manassas, VA) and maintained at 37˚C and 5% CO 2 in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum, 50 μg/mLpenicillin, and 50 μg/mL streptomycin. The HIV-1 virus NL(AD8) was titrated as follows: serial 5-fold dilutions of the virus were made in quadruplicate wells in 96-well culture plates, in a total volume of 100 μL of growth medium, for a total of 8 dilution steps. Freshly trypsinized cells (20,000 cells in 100 μL of growth medium containing 75 μg/mL DEAE-dextran) were added to each well, and the plates were incubated at 37˚C in a humidified 5% CO 2 -95% air environment. After 48 h of incubation, the medium was removed and viral infection was quantified using a β-galactosidase (CPRG) assay (Roche). Twenty thousand TZM-bl cells/well were seeded in 96-well plates in complete DMEM supplemented with 30 μg/mL DEAE-dextran (Sigma-Aldrich). Three hundred times the 50% tissue culture infective dose (TCID50)/mL of HIV AD8 strain and seven serial dilutions (range, 40.000 nM to 625 nM) of each compound were added to the cells, as previously described [81,82]. Vehicle (0.1% dimethyl sulfoxide [DMSO])-treated cells served as a negative control. A CCR5 inhibitor (maraviroc) and an integrase inhibitor (dolutegravir) were used as positive-control drugs. The TZM-bl indicator cell line has an integrated copy of the β-galactosidase gene under control of the HIV-1 promoter. Β-Galactosidase expression, measured by use of a chlorophenol red/β-D-galactopyranoside (CPRG) assay (Roche) in cell lysates 48 h post-infection was used as a marker of HIV infection. The inhibitory curves were fitted by nonlinear regression, allowing for the calculation of the 50% effective concentration (EC 50 ) using the Prism software. To evaluate the cell toxicity of the compounds, the metabolic XTT [2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide] test (Sigma-Aldrich) was performed according to the manufacturer's instructions.

Time of addition
In TOA experiment it is of fundamental importance that only one replication cycle has been completed to avoid confounding effects derived from unwashed viruses after 1 hour of infection. For this reason, an env-pseudotyped virus (REJO4541 clone 67) was used. Time of addition (TOA) experiment was performed as previously described with minor modifications [83]. 40000 TZM-bl cells/well in a 96-multiwell plate were infected with 1500 X the TCID50 per mL of the env-pseudotyped HIV-1 virus in complete medium supplemented with DEAE-dextran (Sigma-Aldrich, 30 mg/mL). Virus was incubated with cells for 1 h at 4˚C, and unbound virus was subsequently removed by extensive and repeated washing with phosphate-buffered saline (PBS) to synchronize the replication. For the next 7 h, antiretroviral compounds inhibiting distinct viral replication steps (maraviroc, lamivudine, dolutegravir) and compound 3 were added at the following time points: at time 0 and after 60, 120, 180, 240, 300, 360 and 420 min. To ensure complete inhibition of viral replication we used a 40-fold EC 50 concentration as previously evaluated for each compound on TZM-bl cells [maraviroc (0.7 mM), lamivudine (5 mM), dolutegravir (1 mM), and compound 3 (10 mM)]. After 2 days, viral infection was quantified using a CPRG assay (Roche) and was then normalized to untreated control cells.

Phytochemical profile
Based on previous observations [53], a phytochemical study was undertaken in order to isolate the most abundant compounds in the hydroalcoholic extract obtained from leaves of H. scruglii (Fig 1). These compounds belong to different classes and, among them, are three phloroglucinol derivatives (compounds 1-3). Three compounds (3)(4)(5) were reported and isolated for the first time from this species and compound 3 was reported here for the first time.
Compound 3 (Fig 3) has been identified as a new metabolite on the basis of its spectroscopic features (Table 1). Its 13 C-NMR spectrum shows 21 signals identified through 13 C and HSQC experiments as 4 methyls, 5 methylenes, 4 methines and 8 quaternary carbons, one of them, at δ 210.1, attributable to a carbonyl group.
Compound 4 was identified as 1,3,5-benzentriol 2-[(2S,3R)-3-(3,4-dihydroxylphenyl)-2,3-dihydroxylpropyl], named filiferol, a chalconoid analogue, isolated for the first time from leaves of Washingtonia filifera [85]. It is based on a flavonol structure with the reduction of the common flavonoid keto group to give an unprecedented methylene carbon on the three carbon chain. To our knowledge, this is the first report of filiferol from Hypericum genus. 3,4-dihydroxybenzoic acid (5 ), is a phenolic acid widely distributed in nature [86]. This compound is one of the biologically active components of some medicinal plants, including Hypericum perforatum L. [87]. It is been defined by the presence of characteristic signals in the 1 H-NMR spectrum.

Inhibition of HIV-1 in cell culture and characterization of the mechanism of action of bioactive compounds in cell-based assays
Given that compounds 1, 2, 3, 4 and 6 were able to inhibit both the HIV-1 RT and IN functions in biochemical assays, we wanted to evaluate their effect on the HIV-1 replication.
Results showed that compounds 1, 2 and 3 significantly inhibited HIV-1 replication with EC 50 values in the 3.5-8 μM range (Fig 4), in accordance with the range of IC 50 values showed against the three viral enzymatic functions in biochemical assays, showing no cytotoxic effect up to the highest tested concentration in cells (CC 50 > 50 μM) (Table 3). Quercitrin, even if it was able of inhibiting both HIV-1 RT-associated activities and IN functions in biochemical assay, did not exert any effect on HIV-1 replication at the highest tested concentration (Table 3), similarly to what reported for its aglycon quercetin [102]. Since compounds 1, 2 and 3 were active on both HIV-1 RT and IN in biochemical assays, we asked which was the viral process targeted by bioactive compounds and, hence, a time-of-addition experiment on the most promising molecule was carried out. This experiment determines how long the addition of an anti-HIV compound can be postponed within the viral replication cycle before losing its antiviral activity. Reference compounds with a known mode of action such as Maraviroc, Lamivudine and Dolutegravir were included. As shown in Fig 5, the compound 3, similarly to lamivudine, a RT inhibitor, lost its activity if added after 4-5 hour post-infection. This timing is compatible with an anti-RT activity, exerted on both RNase H and RDDP functions, but not with IN inhibition [103]. These data demonstrate that compound 3 exerts its anti-HIV activity targeting the RT functions, while the anti-IN activity, exhibited only in enzymatic assays, is not significantly involved in the inhibition of viral replication. A number of compounds were reported to have dual RNase H/IN inhibitory activity in vitro but were more selective for IN and, indeed, they were shown to inhibit the IN in cell-based experiments [104]. Very few compounds [32,33] were reported to be more selective for RNase H versus IN in vitro and were shown to inhibit the RT in cell-based experiments. Hence, it is possible that a dual inhibitory activity displayed in enzymatic assays, is not supported by cell-based results, but at our knowledge, no much data are available on such discrepancies on compounds that were shown to have equal potency of inhibition in vitro on the two enzymes. Therefore, further studies should be performed to elucidate this specific behavior and to obtain derivatives of compound 3 that may be active on viral replication targeting both viral enzymes.

Conclusions
Searching for new potential multitarget anti-HIV active compounds form Sardinian endemic flora, we successfully identified in Hypericum scruglii some chemical components able to