Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Dec 10.
Published in final edited form as: J Biotechnol. 2021 Oct 12;342:28–35. doi: 10.1016/j.jbiotec.2021.10.001

Target-directed evolution of novel modulators of the dopamine transporter in Lobelia cardinalis hairy root cultures

Dennis T Rogers 1,4,*, Francois Pomerleau 2,4,7,8, Zachary Kelley 3,4, Dustin Brown 2,4, Bert Lynn 3,4, Greg A Gerhardt 2,4,5,6,7,8, John Littleton 1,4,9
PMCID: PMC8657075  NIHMSID: NIHMS1749907  PMID: 34648893

Abstract

The dopamine transporter (DAT) is targeted in substance use disorders (SUDs), and “non-classical”“ DAT inhibitors with low abuse potential are therapeutic candidates. Lobinaline, from Lobelia cardinalis, is an atypical DAT inhibitor lead. Chemical synthesis of lobinaline is challenging; thus, “target-directed evolution” was used for lead optimization. A target protein is expressed in plant cells, and a mutant cell population is selected under conditions where target protein functional inhibition confers a survival advantage. Surviving mutants are “mined” for the targeted activity. Applied to a mutant L. cardinalis cell population expressing the human DAT, we identified 20 mutants overproducing DAT inhibitors. Microanalysis prioritized novel lobinaline derivatives, and we first investigated the more water-soluble lobinaline N-oxide. It inhibited rat synaptosomal [3H]DA uptake with an IC50 similar to lobinaline. Against repeated DA microinjections into the rat striatum, lobinaline produced transient DA clearance reductions. In contrast, lobinaline N-oxide prolongingly increased DA peak amplitudes, particularly in the ventral striatum. Lobinaline N-oxide also produced complex changes in post-peak DA clearance inconsistent with simple DAT inhibition. This unusual DAT interaction may prove therapeutically useful for treating SUDs. This study demonstrates the value of target-directed evolution of plant cells for optimizing lead compounds difficult to synthesize chemically.

Keywords: activation tagging mutagenesis, chrono-amperometry, directed evolution, dopamine transport inhibitors, lobinaline, substance use disorders

1. Introduction

Substance use disorders (SUDs) are arguably the most important medical problem in the 21st century [1], and the significance of these SUDs is exacerbated by the lack of effective treatments. Dopamine release from the mesolimbic dopaminergic pathway is involved in the positively reinforcing effects of all abused drugs [2]. Since the plasmalemmal dopamine transporter (DAT) regulates synaptic dopamine concentrations [3], the human DAT protein is a molecular target for all substance use disorders [3,4]. However, it is a challenging target, and a massive synthetic drug discovery effort has yielded very few DAT inhibitors without inherent abuse liability. One exception is bupropion, which is FDA-approved for nicotine use disorder and is a synthetic derivative of the plant-derived natural product, cathinone [5]. Plants are a good source of DAT inhibitory metabolites, probably because they have evolved as defenses against herbivorous insects [6] targeted on the DAT in the insect central nervous system [7]. Therefore, we searched for novel plant metabolites with inhibitory activity on the DAT by screening an extract library of over 1,000 native plant species [8]. This search identified Lobelia cardinalis as containing metabolites with inhibitory effects on the DAT. This plant has no history of abuse. Thus, it is likely that the DAT inhibitory metabolites it contains are “atypical” in not having abuse liability.

The principal alkaloid in Lobelia cardinalis is lobinaline [9], a decahydroquinoline isolated and identified as the first-known binitrogenous plant alkaloid more than 80 years ago [9]. Our studies [see 10] showed that lobinaline is a relatively potent inhibitor of the DAT in vitro (rat brain synaptosomal [3H]DA uptake) and in vivo (DA clearance after microinjection into the rat brain). In addition, lobinaline is a weak non-subtype selective partial agonist at nicotinic acetylcholine receptors (nicAChRs) and a good free radical scavenger [10]. Lobinaline did not violate any of the “druggable” criteria in Lipinski’s “Rule of Five” [11], and the original studies on lobinaline suggest that the compound has the pharmacokinetics and low toxicity suitable for use as a therapeutic agent [12]. This unique pharmacology supports the potential value of lobinaline as a lead compound for the treatment of SUDs. However, it has very poor aqueous solubility and a high degree of structural complexity (including five chiral centers), making lobinaline problematic for conventional lead optimization by chemical synthesis of a compound library.

For this reason, we attempted to generate lobinaline-like inhibitors of the DAT using a novel approach to target-directed evolution in plant cells [13,14]. Target-directed evolution commonly refers to any process in which a population of mutants is selected toward increasing engagement with a specific target. In this plant biotechnology version, a target protein is expressed in plant cells so that the desired interaction of plant metabolites with this protein is linked to cell survival in a specific selection procedure. Mutant plant cells that overproduce metabolites with the desired activity at the target have a survival advantage when this selection procedure is imposed on a population. Therefore, this technology “evolves” biosynthesis in plant cells toward metabolites with the desired activity on the target protein. The surviving mutant sub-population should contain many metabolites with activity on the specific molecular target, providing an alternative complement to combinatorial chemistry for novel drug leads.

For this project, expression of the human (h)DAT protein in L cardinalis cultures rendered the transgenic (hDAT) plant cells highly susceptible to the dopaminergic neurotoxin 1-methyl-4-phenylpyridinium (MPP+), which is accumulated intracellularly by the hDAT [14]. This established a selection procedure (exposure to MPP+) in which metabolites with the desired inhibitory activity on the hDAT confer a survival advantage. Activation tagging mutagenesis (ATM), which randomly activates one or two genes in the plant genome, was performed in an MPP+ concentration in which only 1/300 transgenic (hDAT) mutants developed. The selection was maintained for 4 months, yielding 108 MPP+-resistant transgenic (hDAT) mutants maintained on toxin-free medium for 2 months before analysis. Extracts from each MPP+-resistant clone were compared with control extracts to inhibit rat synaptosomal uptake of [3H]dopamine. Of the 108 MPP+ resistant mutants, 58 showed DAT inhibitory activity greater than two standard deviations above the control mean inhibitory activity, suggesting that MPP+-resistance in these 58 mutants is at least partly a result of overproduction of DAT inhibitory metabolites. Extracts from this sub-population were then analyzed by GC/MS for lobinaline content. In 16 of these mutants, the increased DAT inhibition was attributable to increased lobinaline production, and in another 22 mutants, other increased metabolites are structurally similar to known inhibitors of the DAT [14]. Extracts from the remaining 20 MPP+-resistant mutant clones, therefore, contain previously unknown inhibitors of the DAT. When analyzed using UHPLC-MS and a novel CZE-MS method designed for plant cell cultures, a representative subset of these mutant clones showed an increase in several previously uninvestigated metabolites [15]. These included several lobinaline-like structures not detected by conventional analysis of wild-type plants or non-mutant cultures [15]. This study describes the preliminary pharmacological characterization of one of these, an N-oxide of lobinaline, on DAT function.

2. Methods

2.1. Preparation of individual hairy root cultures for analysis

All reagents were purchased from Thermo Fisher (Waltham MA, USA) unless otherwise noted. As described previously [14], dried methanolic extracts of hairy roots are reconstituted in MeOH to 1mg/mL and 250μL aliquots diluted in 200μL of Optima™ H2O (Thermo Fisher) and acidified with 50mM HCl to pH 2 before extraction with 500μL CHCl3 and the organic layer removed. The aqueous layer is basified to pH 10 with 100mM NH4OH and extracted again with 500μL of CHCl3 before being reduced to dryness under N2.

2.2. UHPLC-MS and CZE-MS analysis

The method previously described by Kelley et al., 2019 [15] was used. UHPLC-MS separations were performed using a Shimadzu Nexera X2 UHPLC system (Torrence CA, USA; SIL-30AC, LC-30AD, CTO-20A, CBM-20A, and DGU-20A) with a Restek Pinnacle DB biphenyl column (Bellefonte PA, USA; 1.9μm, 50x2.1mm). Chromatographic separation was achieved using solvent A (0.1% formic acid) and B (0.1% formic acid in acetonitrile) in a gradient flow method consisting of 10% solvent B for 4 minutes, increasing linearly to 95% over 11 minutes, held at 95% for 5 minutes, and then returned to starting conditions for equilibration. CZE-MS separations were performed using the 908 Devices ZipChip® equipped with an HS chip. Electrophoretic separation was achieved over 3 minutes using a 1,000 V/cm field strength and pressure assist applied at t=0 min. Both UHPLC-MS and CZE-MS mass spectral analyses were performed on a Thermo Scientific QExactive high-resolution orbitrap mass spectrometer. The MS parameters for UHPLC separations consisted of a scan range of 100 to 800 m/z, a resolution of 140,000, microscans set to 3 (1.7 scans per second), the sheath gas flow set to 10 psi, and an AGC target of 3e6. Full scan data provided empirical formulae, and tandem mass spectrometry provided the putative structures. For CZE-MS, the resolution was reduced to 17,500, and microscans were set to 1 (~8 scans per second).

2.3. M/Zmine processing of data from hairy root cultures

RAW data files were converted to mzXML files using the open-source MSConvert software. CZE-MS electropherograms were de-convoluted and de-isotoped following parameters as previously described [15]. The resulting peak lists were aligned using the RANSAC alignment tool with tolerances consistent with previous processes. The aligned list was gap-filled for peaks with a 10% intensity tolerance and again with the m/z and RT range gap filler. UHPLC data was processed using the same workflow, but time ranges were adjusted to accommodate wider peak widths and longer retention times. MS data were normalized based on relative abundance to the putative monoamine (m/z 200.2363) that showed the least variation in abundance across analyses of culture extracts. We used empirical formulas and abundance criteria to prioritize the many metabolites that were increased in the mutant subpopulation containing unknown inhibitors of the DAT. First, those peaks with empirical formulae and putative structures (as determined by tandem mass spectrometry) that were similar to lobinaline or other known Lobelia alkaloids and were increased in cultures containing unknown metabolites with DAT inhibitory activity were prioritized for further study. Second, those increased peaks to the greatest extent and increased in several clones from this mutant subpopulation were also prioritized. In this case, we used a custom algorithm in which high abundance (qualifying level of metabolite >100% mean increase in the sub-population) and frequency (qualifying level of metabolite >100% increase in >30% of selected mutant clones) were both considered. The average increase of each of these metabolites in this sub-population was then multiplied by the frequency of the increase to yield a single “priority score.” This method identifies those metabolites with the highest abundance and the greatest frequency of high abundance in the selected population. Both strategies for prioritization were used to compare wild-type cultures, cultures expressing the hDAT, and ten of those MPP+-resistant mutant clones containing high inhibitory activity on the DAT that could not be ascribed to previously known active compounds. Both prioritization methods identified a lobinaline N-oxide as a likely active compound, which was chosen for initial investigation. Lobinaline N-oxide was separated from culture extracts and whole plants using acid/base extraction and column chromatography with purity (>95%), and its identity was confirmed using LC/MS.

2.4. Evaluation of [3H] dopamine uptake into rat striatal synaptosomes in vitro

2.4.1. Synaptosomes preparation

The animal experimentation protocol was approved by the University of Kentucky Animal Care and Use Committee (protocol # 01059M2006). A previously described method [10] was used. Striatal synaptosomes were prepared from rat brains harvested by BioIVT (Westbury, NY, USA) from CO2-anesthetized adult male Sprague-Dawley rats (200–250 g), flash-frozen in 0.32 sucrose solution, and stored at −80 °C. After thawing on ice, the striatum was dissected and finely minced, then homogenized in approximately 20 volumes of ice-cold buffer containing 0.32 M sucrose, 120 mM NaCl, and 50 mM TRIS HCl (pH=7.4) using a hand-held glass homogenizer with a Teflon pestle. The homogenate was centrifuged (10 min; 1,000xg) to yield a membrane-enriched supernatant. After resuspension, the pellet fraction was centrifuged again (10 min; 1,000xg), and the supernatants were combined. The combined supernatant was centrifuged (20 min; 40,000xg) 3x with re-suspension of the intermediate pellets in ice-cold incubation buffer containing 25 mM HEPES, 120 mM NaCl, 5mM KCl, 2.5 mM CaCl2, 1.2 MgSO4, 1 μM pargyline, 2mg/ml sucrose, and 0.2 mg/ml ascorbic acid, pH=7.4. The resulting pellet was homogenized as described above. The final pellet was resuspended in a 10 ml incubation buffer volume, and the protein concentration was estimated using a bicinchoninic acid protein assay kit (Sigma-Aldrich, St Louis MO, USA). Synaptosomes were kept on ice until their use immediately after preparation.

2.4.2. Synaptosomal uptake of [3H]-dopamine

For the uptake assay, all solutions were prepared and maintained at 37°C. Freshly prepared striatal synaptosomes (100 μl) were loaded into 96-well plates in a 37°C water bath to yield a final protein concentration of 1 mg/ml/well. Each well was then loaded with serial concentrations of lobinaline-N-oxide (100 pM-1mM) or GBR-12909 (10 μM final concentration) as a positive control inhibitor to measure nonspecific uptake or incubation buffer to measure total [3H]-uptake. After 15 min, each well was then loaded with [3H]-dopamine (specific activity: 33.5 Ci/mmol; 10 nM final concentration; Perkin Elmer, Boston MA USA) and incubated for 30 min at 37°C. Incubation was terminated by harvesting the well contents onto a 96-well plate GF/C filter array using a TopCount 96-well harvester (Perkin Elmer) and washing 3x with approximately 300 uL ice-cold TRIS-acetate buffer (pH=7.25). After overnight drying in a chemical hood, the plates were sealed, and each filter was loaded with 35 μL of MicroScint 20® scintillation fluid (Perkin Elmer). After incubation in the dark for 2 hr, plates were counted on a TopCount scintillation counter (Perkin Elmer). CPM data were analyzed using GraphPad v6 software (GraphPad, San Diego, CA USA).

2.5. [3H]GBR 12935 binding to rat striatal homogenate

This assay was performed as previously described [10]. Binding assays were performed in 96-well plates with a final per well volume of 300 μl. All solutions were prepared in a binding buffer (50 mM Tris-HCl, pH 7.4, containing 120 mM NaCl and 0.01% bovine serum albumin), and each well contained a final concentration of 1 mg of membrane protein. Membranes were prepared as described for synaptosomal preparations with some modifications. Thawed striatal tissue from BioIVT was homogenized in ice-cold homogenization buffer (120 mM NaCl, 50 mM Tris-HCl, pH = 7.4), centrifuged at 49,000 × g for 20 minutes at 4°C, then washed three times with assay buffer (120 mM NaCl, 50 mM Tris-HCl, 0.01% FBS, pH 7.4). Each wash was performed by resuspending the pellet in approximately ten volumes of assay buffer followed by centrifugation at 49,000 × g for 20 minutes at 4°C. Total protein content was measured using a bicinchoninic acid assay kit (Sigma Aldrich). Striatal membranes were stored at −80°C prior to further experimentation. The protein concentration was adjusted to 3 mg/ml with assay buffer (to yield a final per well concentration of 1 mg/ml).

Solutions containing lobinaline-n-oxide (10 pM-3 mM), 100 μM GBR-12909, or buffer solution were added to wells containing the striatal membrane preparation 15 min prior to the addition of 1 nM of [3H]GBR-12935 and incubated for 1 hr at 20-22°C. The binding was terminated by rapid vacuum filtration using a TopCount Harvester system through a 96 well GF/B filter array (Perkin Elmer). Filters were pre-treated with a solution of 0.1% polyethyleneimine 1 hr prior to harvesting to reduce the radioligand’s nonspecific binding. Nonspecific binding was determined by measuring binding in the presence of 100 μM GBR-12909. After overnight drying, 35 μL of scintillation fluid (Microscint 20, Perkin Elmer) was added to each well, and plates were placed in the dark for two hours. Afterward, radioactivity was measured by scintillation counting (2 minutes per well) using a Perkin Elmer TopCount® NXT™ microplate scintillation counter. In each plate, nonspecific binding was measured in the presence of excess GBR12909 (final concentration, 10 μM), and total binding was measured with radioligand alone. Total specific binding and specific binding in the presence of lobinaline were calculated by subtracting nonspecific binding. Specific binding in the presence of lobinaline was expressed as a percentage of total specific binding.

2.6. in vivo measures of dopamine uptake in the striatum

The effects of lobinaline or lobinaline N-oxide on dopamine uptake in the striatum of isoflurane-anesthetized SD rats were monitored by in vivo electrochemistry using high-speed chrono-amperometry [see 17,18]. After obtaining reproducible signals for locally applied DA (50-75 nL, 200μM) to produce a control DA peak height below 1 μM, vehicle or drug was locally applied slowly over 10-30 sec (~500 nL 250 μM), then 60 sec later the next application of DA was performed, with subsequent DA applications at 5-min intervals for 25 min. The DA kinetic data from each animal were analyzed using FAST software (Quanteon, LLC, Nicholasville KY USA). This analysis defines the maximum peak height and three clearance parameters: 1) T80, the 80% decay time from the peak response; 2) the first-order rate of decay of the DA signal (k−1); and 3) the clearance rate, the maximum peak height multiplied by k−1. These parameters were compared pre-and post-drug application by one-way ANOVA with Tukey’s post hoc analysis, with significance assigned at p<0.05.

3. Results

3.1. Stability of transgenic mutant plant cell cultures

Five years after transformation, the transgenic (hDAT) clones continue to show increased [3H]DA uptake and remain sensitive to MPP+-toxicity. Thus, the hDAT transgene continues to be expressed and remains functional. Similarly, MPP+-resistant mutants selected 5 years previously retained resistance upon re-exposure to MPP+, indicating that the protective mutations are stable. Next-generation RNA sequencing of MPP+-resistant mutants showed over-expression of genes that regulate alkaloid biosynthesis, microtubule assembly, and responses to oxidative stress, which are all relevant to continued MPP+-resistance. Mutants whose extracts produced DAT inhibitory activity or overproduced lobinaline continued to do so. The genotype and phenotype stability suggests that these cultures remain valuable for lead optimization, as well as for plant-based drug discovery and production [14].

3.2. Prioritization of metabolites from MPP+-resistant cultures

Using MZmine, 139 unique, resolved peaks that appeared in at least two replicates of the same sample were detected. The WT, DAT+, and the MPP+-resistant mutant 479 (a representative mutant selected a priori) typically contained 136-139 peaks. This master peak list was then used to select putatively identified molecules of interest because they were increased in MPP+-resistant mutants containing increased DAT inhibitory activity. The average area counts for lobinaline and its derivatives generally showed an increasing trend from the WT to the mutant sample sets. This finding is consistent with the hypothesis that the overproduction of these metabolites protected against the MPP+ toxicity used to select L. cardinalis mutant hairy roots. Lobinaline-like or lobeline-like molecules were initially set as one priority because they were highly likely to be active. This list included eight variants on the lobinaline structure, including different N-oxides of lobinaline. The subsequent analysis of several putative N-methyl pyridine metabolites, together with their putative N-oxides, was included in this list. These have putative structures similar to lobeline (see Kelley et al., 2019 [15]). but lobeline itself was never observed in analyzed cultures or whole Lobelia cardinalis plants during our analyses.

Further analysis of these peaks found in the highest abundance within the samples from this mutant subpopulation compared with control groups was then prioritized based on the algorithm described in the methods. Thus, we prioritized 13 metabolites, including one of the lobinaline N-oxides (m/z 403.2740), dubbed priority #9, and a bi-nitrogenous analog of lobinaline (nominated as priority #3) (m/z 447.3006) that were in the other prioritized group (Figure 1). Neither of these compounds has previously been reported in any plant species.

Figure 1. Putative structures of selected lobinaline N-oxides and lobeline-like alkaloids increased in MPP+-resistant cultures of Lobelia cardinalis containing unknown inhibitors of the DAT.

Figure 1.

Open in a new tab

Chemical structures of lobinaline (A), its naturally occurring N-oxide (B), and putative structures of some other metabolites detected in hairy root mutants (C-H) are shown.

The other prioritized metabolites’ putative structures are previously unreported complex alkaloids, with no apparent structural relation to lobinaline. They may represent novel leads for modulators of the DAT. We chose to investigate the pharmacology of the lobinaline N-oxide first because it is relatively easy to extract and purify from plant cell cultures. This metabolite and lobinaline were separated by acid/base extraction before separation by column chromatography with identity and purity (>95%) confirmed by LC/MS. N-oxidation commonly increases the aqueous solubility of alkaloids [32], and the lobinaline N-oxide from mutant cultures was much more soluble than lobinaline (K octanol/water 1.29+19 vs. lobinaline Ko/w 4.8). This improved aqueous solubility immediately makes the investigation of biological activity easier and confers a “druggability” advantage over the parent compound.

3.3. Radioligand studies of DAT inhibition and binding

We investigated DAT functional inhibition by measuring rat striatal synaptosomal uptake of [3H]-DA. Figure 2 shows the effects of lobinaline and lobinaline-N-oxide extracted from mutant plant cells on {3H]-DA uptake in rat striatal synaptosomes. Both lobinaline and lobinaline-N-oxide inhibited [3H]-DA uptake into rat striatal synaptosomes, with a lobinaline IC50 of 359 nM and a lobinaline N-oxide IC50 of 517 nM.

Figure 2. Effects of lobinaline and lobinaline N-oxide on uptake of [3H]dopamine into rat synaptosomes.

Figure 2.

Open in a new tab

Following 15 min pre-treatment with test compounds (100 pM-1mM final concentrations), synaptosomes were loaded with 10 nM [3H]-dopamine (specific activity 33.5 Ci/mmol) and incubated at 37°C for 30 min. The wells were harvested, washed three times with 300 μL ice-cold 50 mM TRIS acetate buffer (pH 7.25), and dried overnight. Radioactivity was quantified using scintillation counting, and data were analyzed using GraphPad software. Each point represents the mean + SE of 6-8 replicate wells of 96-well plates (300 μL final volume).

Subsequent studies investigated whether this inhibition was related to competitive antagonism of the DAT protein. Lobinaline completely displaced the ligand [3H]GBR 12935 with a Ki of 2.54 μM, whereas the N-oxide partially displaced the radioligand with a modest Ki (131.9 μM Figure 3). GBR 12909 completely displaced the radioligand with a calculated Ki of 20 nM.

Figure 3. Displacement of [3H]GBR 12935 from rat striatal membranes by lobinaline and lobinaline N-oxide.

Figure 3.

Open in a new tab

Following 15 min pre-treatment with test compounds (100 pM-1mM final concentrations), Sprague Dawley rat striatal membranes were treated with 10 nM [3H]GBR12935 (specific activity 40.43 Ci/mmol) and incubated at 37°C for 30 min. Wells were harvested, washed three times with 300 μL ice-cold 50 mM TRIS acetate buffer (pH 7.25), and dried overnight. Radioactivity was quantified using scintillation counting, and data were analyzed using GraphPad software. Each point represents the mean+SE of 6-8 replicate wells of 96-well plates (300 μL final volume).

3.4. The effects of lobinaline and lobinaline N-oxide on dopamine signals in vivo

Representative data from local dopamine, lobinaline, and lobinaline N-oxide applications are shown in Fig. 4a and 4b. These figures show the effects of a single local application of the test agents on the phasic changes in extracellular dopamine concentration produced by repeated local dopamine application into the rat striatum as measured by chrono-amperometry. As reported previously [10], the effects of lobinaline are a transient non-significant decrease in the DA signal’s amplitude. In contrast, lobinaline N-oxide produces an increase in amplitude maintained for at least 20 minutes over four local applications of dopamine (Fig. 4b). Neither lobinaline nor its N-oxide ever caused the release of endogenous DA in any brain region in this system (n=8, 36 respectively).

Figure 4: Representative data showing the effects of a single Lobinaline (a) or lobinaline N-oxide (b) microinjection on the peak amplitude of DA repeatedly microinjected into the striatum.

Figure 4:

Open in a new tab

These are electrochemical signals following local pressure ejection of dopamine (DA 200 μM; 25-50 nl) in the striatum repeated every 5 mins. Lobinaline (a) produced effects for approximately 5 min after microinjection. Lobinaline N-oxide (250 uM, 500 nL) produced a marked increase in DA peaks that persisted for the approximately 20-minute duration of the recording. Other changes in the DA signals induced by the lobinaline N-oxide are described below.

In addition to this major difference in DA peak amplitude effects, there were differences between lobinaline and lobinaline N-oxide in their effects on DA clearance after the peak. As reported previously [10], lobinaline caused a small but significant reduction in all DA clearance parameters (relative to % control) consistent with inhibitory effects on the DAT. However, in contrast, lobinaline-N-oxide caused complex changes in which DA clearance parameters were reduced immediately after microinjection but increased later, even when DA peak amplitude was still increased over control (see Figure 5a and b).

Figure 5. (a) The effects of lobinaline N-oxide on DA signals in the dorsal striatum (DV-3.5 to DV-5.5mm) (b) The effects of lobinaline N-oxide on DA signals in the ventral striatum (DV-6 to DV-7.5mm).

Figure 5.

Open in a new tab

Changes are expressed as the percentage of the control DA signal parameters (with control amplitude selected at < 1uM) over the 26-minute duration of each experiment. Data are expressed as mean + SEM of n=4-6 experimental animals. Statistical analysis (one-way ANOVA and Tukey’s post hoc) compares the control data (100%) with the change produced by lobinaline N-oxide as % control. Asterisks indicate significance with *p <0.05.

Figure 5a illustrates the effects of lobinaline N-oxide on DA signals in the dorsal (DV-3.5 to DV-5.5mm) striatum and are shown with changes expressed as the percentage of control DA signal parameters. There was a marked and prolonged (but variable) increase in the peak amplitude throughout the experiment (i.e., up to 26 minutes after micro-injection of lobinaline N-oxide). Regarding the clearance parameters, the early changes are an increase in T80 (the 80% decay time from the peak response) and clearance rate (the maximum peak height multiplied by the first-order rate of decay of the DA signal (k−1)), reflecting an increase in peak amplitude, with no change in k-1. However, later at 11-26 min, there is an increase in both uptake rate and k-1 with little change in T80. These effects suggest an increase in the rate of DA clearance after the peak and are not consistent with a simple inhibition of the DAT throughout the experiment’s 26 min duration.

In Figure 5b, the effects of lobinaline N-oxide on DA signals are shown with changes expressed as the percentage of control DA signal parameters. This depth in the DV gradient represents the shell of the nucleus accumbens [30]. Regarding the effect of lobinaline N-oxide, there is a marked and consistent increase in DA peak amplitudes throughout the experiment (i.e., up to 26 min after single microinjection of lobinaline N-oxide). The effects of lobinaline N-oxide are variable but show an increase in clearance rate and k-1 at 21 and 26 min. Once again, these data are not consistent with a simple inhibition of the DAT throughout the experiment but suggest complex DA signaling modulation.

4. Discussion

This study aimed to generate and identify previously uninvestigated DAT inhibitors using target-directed evolution of mutant Lobelia cardinalis hairy roots expressing the human DAT. This plant species contains a DAT inhibitor, lobinaline, as the major alkaloid [10], which is very difficult to synthesize chemically [22], It was assumed that several of the active metabolites produced by target-directed evolution would be previously unknown lobinaline derivatives. If any of these proved to be more therapeutically valuable or druggable than lobinaline, this would provide proof-of-application for target-directed evolution as a biosynthetic lead optimization strategy. Several lobinaline- or lobeline-like molecules were increased in the mutant cultures that contained unknown metabolites with increased inhibitory activity on the DAT. In addition to lobinaline N-oxides (see below), several N-methyl pyridines with structural similarity to lobeline were found to be markedly increased in this mutant population. These metabolites have previously been reported in other Lobelia species, including the Chinese medicinal plant L. chinensis [23], but appear to have never been tested for DAT activity. Lobeline itself inhibits the DAT and VMAT2 [24] but was not found in mutant culture extracts or extracts of L. cardinalis [10] (or in L. chinensis [25]).

In addition to these Lobelia alkaloid–like metabolites, many other metabolites were increased in one or more mutants containing increased DAT inhibitory activity. To investigate all of these simultaneously was beyond our capability, so those metabolites showing an increase relative to wild-type were prioritized based on the averaged magnitude of increase in the whole population, together with a frequency of a 100% increase in individuals in this population. Thirteen metabolites were prioritized on this basis, and one of these was identified as an N-oxide of lobinaline. This metabolite was chosen to be the first metabolite to be investigated because of its high likelihood of exhibiting the desired bioactivity and because heterocyclic N-oxides generally possess good druggable characteristics [see 26]. It should be noted that the N-oxide position on the molecule is not yet certain. Energetically, it is likely that N-oxidation in this natural N-oxide occurred on the N-methyl nitrogen of the decahydroquinoline heterocycle rather than the tetrahydropyridinyl heterocycle. However, we have not yet accumulated enough pure lobinaline N-oxide to confirm this by NMR. Additionally, when the N-oxide is on the decahydroquinoline heterocycle, this can exist in two diastereomeric forms, which may differ in their pharmacology. This study used the N-oxide extracted from transgenic (hDAT) plant cells markedly increased in MPP+-resistant mutants. This N-oxide was much more soluble in aqueous media than lobinaline, making it a more druggable optimized lead relative to lobinaline.

In our studies, lobinaline and this lobinaline-N-oxide inhibit [3H]-DA uptake in rat striatal synaptosomes with IC50s in the sub-micromolar range. These inhibitors compare favorably with other clinically used DAT inhibitors such as modafinil, IC50 4 μM [33], and bupropion, IC50 1.3 μM [34]. As regards mechanism, our radioligand binding displacement studies with [3H]-GBR12935, a selective ligand for the DAT [31], showed that lobinaline and lobinaline-N-oxide exhibit Ki values at least an order of magnitude lower than the potent “competitive” DAT inhibitor GBR-12909. The low affinity for this binding site suggests that these lobinaline alkaloids differ from classical DAT inhibitors like cocaine and methylphenidate that potently displace [3H]GBR-12935. In addition, the lobinaline alkaloids are not structurally related to any of these compounds [32]. They likely act on a different site as allosteric modulators of the DAT to reduce net uptake of [3H]-DA in this synaptosomal system in vitro.

Regarding its in vivo pharmacology, the lobinaline N-oxide produced different effects on repeated DA signals from those caused by lobinaline when locally microinjected into the rat brain. As reported previously [14], lobinaline produced a small transient reduction in DA peak height following a single microinjection. However, lobinaline-N-oxide caused a significant and prolonged increase in dopamine peak height following the same procedure. This effect of a single micro-injection of lobinaline N-oxide was maintained through repeated local DA applications, suggesting that lobinaline N-oxide is poorly displaced from the DAT by DA and makes it different from most other known DAT inhibitors, including cocaine. This effect of lobinaline N-oxide on DA peak height is consistent with inhibition of the DAT and its effects on synaptosomal [3H]DA uptake. However, the two metabolites also appeared to produce different effects on the decline in the DA signal after the peak, with lobinaline producing the expected reduction in the clearance rate. In contrast, lobinaline N-oxide caused changes consistent with an increased DA clearance rate, at least in the striatum. These effects of lobinaline N-oxide are not compatible with purely inhibitory effects on the DAT. They suggest either a complex modulatory effect on the DAT or some other action that impacts DA clearance in the striatum, such as agonist activity at nicotinic receptors (14, 29). Further investigation of this lobinaline N-oxide pharmacology is warranted, particularly regarding its interaction with drugs of abuse. Here the mixture of positive and negative modulation of the DAT shown by this lobinaline N-oxide may confer therapeutic value in treating SUDs.

5. Conclusions

This novel biotechnology identified a lobinaline derivative that is both more druggable and probably more pharmacologically valuable than the parent alkaloid, providing proof-of-application for this novel plant biotechnology as a platform for biosynthetic lead optimization. Target-directed evolution should prove invaluable in using the vast synthetic potential of plant biosynthesis for lead optimization and drug discovery.

Highlights.

  • Extract library screening identified novel DAT inhibition in Lobelia cardinalis.

  • Human DAT-expressing, activation tagged L. cardinalis mutants were selected using MPP+

  • Novel microanalysis prioritized survivor metabolites, including lobinaline-N-oxide.

  • Lobinaline-N-oxide inhibited synaptosomal 3H-dopamine uptake similar to lobinaline.

  • Microinjected lobinaline-N-oxide prolonged rat ventral striatum DA peak amplitudes.

  • It produced complex DA clearance changes inconsistent with simple DAT inhibition.

  • This unusual DAT interaction may be therapeutically useful in substance use treatment.

  • Thus, plant cell target-directed evolution is valuable for optimizing drug leads.

Funding

This research was supported by two SBIR awards to Naprogenix Inc with JL as PI. These were 1R44AA018226 “Selection-driven plant metabolites for treatment of CNS diseases” and 1R44AA025804 “Mutant transgenic plant cells as a novel source of drugs” from the National Institute on Alcohol Abuse and Alcoholism. The content is solely the authors’ responsibility and does not necessarily represent the official views of the National Institutes of Health. The funding agency had no role in the study design, the collection, analysis, and interpretation of data in the writing of the report or in the decision to submit the article for publication.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

CRediT authorship contribution statement

Dennis T. Rogers: Conceptualization, Data Curation, Formal Analysis, Investigation, Methodology, Supervision, Writing - Review & Editing, Validation, Visualization., Francois Pomerleau: Data Curation, Investigation, Methodology, Visualization, Writing - Review & Editing., Zachary Kelley: Data Curation, Methodology, Formal Analysis, Investigation, Writing - Review & Editing., Dustin Brown: Conceptualization, Methodology, Resources, Bert Lynn: Conceptualization, Data Curation, Methodology, Writing - Review & Editing, Greg A. Gerhardt: Data Curation, Methodology, Visualization., John Littleton: Conceptualization, Writing - Original Draft, Project administration, Funding acquisition, Visualization

Conflict of interest declaration

JL is the CEO of Naprogenix, and DTR is the Research Director of Naprogenix Inc. GAG is the sole proprietor of Quanteon, LLC, the manufacturer of the Fast-16 hardware and software used to measure DA uptake.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  • 1.Klein JW. Pharmacotherapy for Substance Use Disorders. Med Clin North Am. 2016. 100(4):891–910. [DOI] [PubMed] [Google Scholar]
  • 2.Koob GF, Volkow ND. Neurocircuitry of addiction. Neuropsychopharmacology. 2010. 35(1):217–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Zhu J, Reith ME. Role of the dopamine transporter in the action of psychostimulants, nicotine, and other drugs of abuse. CNS Neurol Disord Drug Targets. 2008. 7(5):393–409 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Schmitt KC, Rothman RB, Reith ME. Nonclassical pharmacology of the dopamine transporter: atypical inhibitors, allosteric modulators, and partial substrates. J Pharmacol Exp Ther. 2013. 346(1):2–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Kelly JP. Cathinone derivatives: a review of their chemistry, pharmacology, and toxicology. Drug Test Anal. 2011. 3(7–8):439–53 [DOI] [PubMed] [Google Scholar]
  • 6.Pichersky E & Gang DR: Genetics and biochemistry of secondary metabolites in plants: an evolutionary perspective. Trends Plant Sci 2000. 5: 439–445 [DOI] [PubMed] [Google Scholar]
  • 7.Chen R, Wu X, Wei H, Han DD, Gu HH. Molecular cloning and functional characterization of the dopamine transporter from Eloria noyesi, a caterpillar pest of cocaine-rich coca plants. Gene. 2006. 366(1):152–60. [DOI] [PubMed] [Google Scholar]
  • 8.Littleton J, Rogers T, Falcone D. Novel approaches to plant drug discovery based on high throughput pharmacological screening and genetic manipulation. Life Sci. 2005. 78(5):467–75. [DOI] [PubMed] [Google Scholar]
  • 9.Manske RHF (1938). Lobinaline, an alkaloid from Lobelia cardinalis L. Can. J. Research 16 B: 445–448. [Google Scholar]
  • 10.Brown DP, Rogers DT, Pomerleau F, Siripurapu KB, Kulshrestha M, Gerhardt GA, Littleton JM Novel multifunctional pharmacology of lobinaline, the major alkaloid from Lobelia cardinalis. Fitoterapia. 2016. 111:109–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Lipinski CA. Lead- and drug-like compounds: the rule-of-five revolution. Drug Discovery Today Technol. 2004. 1(4):337–41 [DOI] [PubMed] [Google Scholar]
  • 12.Robison MM, et al. , The skeletal structure of lobinaline. J Org Chem, 1966. 31(10): p. 3206–13. [DOI] [PubMed] [Google Scholar]
  • 13.Littleton JM. The future of plant drug discovery. Expert Opinion on drug discovery. 1: 673–683, 2007 [DOI] [PubMed] [Google Scholar]
  • 14.Brown DP, Rogers DT, Gunjan SK, Gerhardt GA, Littleton JM. Target-directed discovery and production of pharmaceuticals in transgenic mutant plant cells. J Biotechnol. 2016. 238:9–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kelley ZD, Rogers DT, Littleton JM, Lynn BC. Microfluidic capillary zone electrophoresis mass spectrometry analysis of alkaloids in Lobelia cardinalis transgenic and mutant plant cell cultures. Electrophoresis. 2019. 40(22):2921–2928. DOI: 10.1002/elps.201900220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Williams JM, Steketee JD. Characterization of dopamine transport in crude synaptosomes prepared from rat medial prefrontal cortex. J Neurosci Methods. 2004. 137(2): 161–5 [DOI] [PubMed] [Google Scholar]
  • 17.Zahniser NR, Larson GA, and Gerhardt GA, In vivo dopamine clearance rate in rat striatum: regulation by extracellular dopamine concentration and dopamine transporter inhibitors. J Pharmacol Exp Ther, 1999. 289(1): p. 266–77. [PubMed] [Google Scholar]
  • 18.Cass WA, Gerhardt GA, Gillespie K, Curella P, Mayfield RD, Zahniser NR. Reduced clearance of exogenous dopamine in rat nucleus accumbens, but not in dorsal striatum, following cocaine challenge in rats withdrawn from repeated cocaine administration. J Neurochem. 1993. 61(1):273–83. [DOI] [PubMed] [Google Scholar]
  • 19.Gupta RN and Spenser ID, Biosynthesis of Lobinaline. Canadian Journal of Chemistry, 1971. 49(3): p. 384–397 [Google Scholar]
  • 20.Yang S, Shen T, Zhao L, et al. Chemical constituents of Lobelia chinensis. Fitoterapia. 2014. 93:168–174. [DOI] [PubMed] [Google Scholar]
  • 21.Felpin FX, Lebreton J. History, chemistry and biology of alkaloids from Lobelia inflata. Tetrahedron. 2004. 60:10127–53. [Google Scholar]
  • 22.Wang H, Li Y, Huang Y, Zhao C, Cheung HY. Chemical Profiling of Lobelia chinensis with High-Performance Liquid Chromatography/Quadrupole Time-of-Flight Mass Spectrometry (HPLC/Q-TOF MS) Reveals Absence of Lobeline in the Herb. Molecules. 2018. 23(12):3258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Mfuh AM, Larionov OV. Heterocyclic N-oxides - An Emerging Class of Therapeutic Agents. Curr Med Chem. 2015. 22(24):2819–57 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Wilhelm CJ, Johnson RA,Eshleman AJ, Janowsky A. Lobeline effects on tonic and methamphetamine-induced dopamine release. Biochem Pharmacol. 2008. March 15;75(6): 1411–5. doi: 10.1016/j.bcp.2007.11.019. Epub 2007 Dec 4. ; [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Foster JD, Vaughan RA. Phosphorylation mechanisms in dopamine transporter regulation. J Chem Neuroanat. 2017. 83–84:10–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Owesson-White CA, Roitman MF, Sombers LA, Belle AM, Keithley RB, Peele JL, Carelli RM, Wightman RM Sources contributing to the average extracellular concentration of dopamine in the nucleus accumbens. J Neurochem. 2012. 121(2):252–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Yorgason JT, Zeppenfeld DM, Williams JT. Cholinergic interneurons underlie spontaneous dopamine release in nucleus accumbens. J Neurosci. 2017. 37(8):2086–2096 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Hartmann T, Ehmke A, Eilert U, von Borstel K, Theuring C. Sites of synthesis, translocation, and accumulation of pyrrolizidine alkaloid N-oxides in Senecio vulgaris L. Planta. 1989. 177(1):98–107 [DOI] [PubMed] [Google Scholar]
  • 29.Ksir C, Mellor G, Hart C, Gerhardt GA. Nicotine enhances dopamine clearance in rat nucleus accumbens. Prog Neuropsychopharmacol Biol Psychiatry. 1995. 19(1): 151–156. DOI: 10.1016/0278-5846(94)00111-t [DOI] [PubMed] [Google Scholar]
  • 30.Paxinos G, Watson C 5th ed. Academic Press; San Diego: 2005. The rat brain in stereotaxic coordinates. [Google Scholar]
  • 31.Andersen PH. The dopamine inhibitor GBR 12909: selectivity and molecular mechanism of action. Eur J Pharmacol. 1989. 166(3):493–504. DOI: 10.1016/0014-2999(89)90363-4. [DOI] [PubMed] [Google Scholar]
  • 32.Schmitt K Rothman RB, Reith M. Non-classical Pharmacology of the Dopamine Transporter: Atypical Inhibitors, Allosteric Modulators and Partial Substrates. JPET 2013, jpet.111.191056; DOI: 10.1124/jpet.111.191056 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Zolkowska D, Jain R, Rothman RB, Partilla JS, Roth BL, Setola V, Prisinzano TE, Baumann MH. Evidence for the involvement of dopamine transporters in behavioral stimulant effects of modafinil. J Pharmacol Exp Ther. 2009. 329(2):738–46. doi: 10.1124/jpet.108.146142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Sitges M, Reyes A, Chiu LM. Dopamine transporter-mediated release of dopamine: role of chloride. J Neurosci Res. 1994. September 1;39(1):11–22. doi: 10.1002/jnr.490390103. [DOI] [PubMed] [Google Scholar]

RESOURCES