Identifying new AMP-activated protein kinase inhibitors that protect

Subscriber access provided by Washington University | Libraries

Article

Identifying new AMP-activated protein kinase inhibitors that protect against ischemic brain injury Jae-Won Eom, Tae-Youn Kim, Bo-Ra Seo, Hwangseo Park, Jae-Young Koh, and Yang-Hee Kim ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.8b00654 • Publication Date (Web): 14 Feb 2019 Downloaded from http://pubs.acs.org on February 15, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

Identifying new AMP-activated protein kinase inhibitors that protect against ischemic brain injury Jae-Won Eom1, Tae-Youn Kim2, Bo-Ra Seo2, Hwangseo Park3, Jae-Young Koh2,4, Yang-Hee Kim1,3,* 1Department 2Neural

of Molecular Biology, Sejong University, Seoul 05006, Republic of Korea

Injury Research Laboratory, University of Ulsan College of Medicine, Seoul 138-

736, Republic of Korea 3Department

of Integrative Bioscience and Biotechnology, Sejong University, Seoul 05006,

Republic of Korea 4Department

of Neurology, University of Ulsan College of Medicine, Seoul 138-736,

Republic of Korea

* Correspondence

to:

Dr. Yang-Hee Kim 209 Neungdong-Ro Gwangjin-Gu Seoul, 05006, Republic of Korea Tel: +82-2-3408-3648 Fax: +82-2-3408-4336 E-mail: [email protected]

-1ACS Paragon Plus Environment

ACS Chemical Neuroscience 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract

We recently reported that AMP-activated protein kinase (AMPK) contributes to zinc-induced neuronal death by inducing Bim, a pro-apoptotic Bcl-2 homology domain 3-only protein, in a liver kinase B1 (LKB1)-dependent manner. Current data suggest AMPK plays key roles in excitotoxicity and ischemic brain injury, with zinc neurotoxicity representing at least one mechanism of ischemic neuronal death. Inhibition of AMPK could be a viable therapeutic strategy to prevent ischemic brain injury following stroke. This prompted our search for novel inhibitors of AMPK activity and zinc-induced neuronal death using cultured mouse cortex and a rat model of brain injury after middle cerebral artery occlusion (MCAO). In structure-based virtual screening, 118 compounds were predicted to bind the active site of AMPK 2, and 40 showed in vitro AMPK 2 inhibitory activity comparable to compound C (a well-known, potent AMPK inhibitor). In mouse cortical neuronal cultures, 7 of 40 compound reduced zinc-induced neuronal death at levels comparable to compound C. Ultimately, only agents 2G11 and 1H10 significantly attenuated various types of neuronal death, including oxidative stress, excitotoxicity, and apoptosis. When administered as intracerebroventricular injections prior to permanent MCAO in rats, 2G11 and 1H10 reduced brain infarct volumes, whereas compound C did not. Therefore, these novel AMPK inhibitors could be drug development candidates to treat stroke.


Page 2 of 37

Page 3 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Graphic Abstract

Keywords AMPK, brain ischemia, zinc neurotoxicity, excitotoxicity, oxidative stress, apoptosis



Page 4 of 37

Introduction Brain ischemia is a cerebral injury due to an acute drop in cerebral blood pressure and subsequent interruption of oxygen and glucose supplies. Declining oxygen/glucose levels impose obstacles to ATP production and eventually reduce intracellular energy reserves. Ischemic injury is associated with various types of cellular stress, including zinc overload1-5, glutamate excitotoxicity6, oxidative/nitrosative stress7, apoptosis8, and necrosis9. Among these diverse cellular stress, accumulation of intracellular zinc after brain ischemic injury has particularly broad ramifications, given the potential for reactive oxygen species (ROS) generation10, elevation of intracellular calcium5, NAD+/ATP depletion11, and p75NTRinduced apoptosis12,13. Therefore, we firstly used zinc-mediated neurotoxicity model to select the potential compounds for protection against brain ischemia. AMP-activated protein kinase (AMPK) is a heterotrimeric enzyme consisting of an alpha catalytic (1 or 2), a beta regulatory ( or  and a gamma regulatory (  or  subunit AMPK is activated when cells are confronted with energy stress conditions necessitating catabolic/anabolic pathway adjustments to increase cellular ATP levels15,

16.

Importantly, AMPK shows allosteric activation. In energy-rich states,  subunits are bound to ATP and are thus inactive. However, during incremental energy stress, ATP is replaced by ADP or AMP, and upstream kinases of AMPK boost phosphorylation of AMPK on threonine 172 (Thr 172) at the  catalytic subunit for enzymatic activation. This phosphorylation changes the entire conformation of AMPK, inhibiting dephosphorylation of Thr 172 and sustaining AMPK activity15-18. In mammals, there are two major upstream targets of AMPK, namely liver kinase B1 (LKB1) and Ca2+/calmodulin-dependent protein kinase  CaMKK

 .

Activated AMPK

broadly regulates the metabolism of glucose, proteins, and lipids; mitochondrial biogenesis; -4ACS Paragon Plus Environment

Page 5 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


autophagy; cell growth; and apoptosis by directly or indirectly phosphorylating a host of requisite substrates or transcribing metabolism-related genes20-22. Once energy levels are restored, protein phosphatase 2A (PP2A) and 2C (PP2C) dephosphorylate Thr 172 to inactivate AMPK18. Although one study showed that activation of AMPK by metformin protects against global cerebral ischemia23, many reports suggest that AMPK has detrimental effects on neurons impacted by brain ischemia. AMPK is clearly activated during physiologic states of energy depletion to overcome this stress, but there is also evidence that ischemia-related energy loss in the brain may trigger its activation14,

24-27,

which induce the severe brain damage.

Furthermore, we have demonstrated that AMPK mediates caspase-3 dependent neuronal apoptosis, one of the chief mechanisms of neuronal death during ischemic injury21, implicating AMPK inhibition as a therapeutic approach to brain ischemia. Consistent with this, some investigators have showed in animal models of stroke that injecting compound C (Cpd C; a selective pharmacologic AMPK inhibitor) or using mice genetically deficient in AMPK2 achieves neuroprotection, whereas the use of an AMPK activator (metformin) seems to exacerbate stroke injury24, 26, 28, 29. Since we already confirmed that AMPK contributes to zinc-induced neuronal death in mouse cortical cultures via LKB1-dependent induction of Bim21, and the findings of others suggest that excitotoxicity is mediated via AMPK20, this study was performed to find novel chemical inhibitors of AMPK for use as brain ischemia therapeutics. We used a structure-based virtual screening method to identify novel agents capable of binding the AMPK2 catalytic subunit30 as potential protective agents against stroke injury. We screened the effects of these candidate drugs step-by-step through in vitro assays of AMPK enzymatic activity, neuronal cell-based analysis, and an in vivo rat model of stroke. Ultimately, we selected two lead -5ACS Paragon Plus Environment


compounds, each of which showed profound and multifaceted mitigation of ischemic stroke effects. These new-found AMPK inhibitors could be promising protective treatments for ischemic brain injury.

Results and discussion Selection of 40 potential AMPK inhibitors that significantly reduce 2 subunit activity We previously launched our search for new chemical inhibitors of AMPK, using structuralbased virtual screening to select 118 compounds predicted to bind AMPK 2 active sites30. At 1 μM concentrations, seven agents (1A03, 1A05, 1A06, 1A07, 1A08, 1D05, and 1F05; Figure 1A, grey bars) showed the most substantial in vitro AMPK inhibition, which was corroborated by previous reports30. To broaden the array of candidates, 10 μM concentrations were assayed, which added another 33 compounds (1A09, 1A10, 1B04, 1B05, 1C06, 1C09, 1C12, 1D06, 1D09, 1D10, 1E08, 1F06, 1G04, 1G06, 1G09, 1G10, 1H02, 1H05, 1H06, 1H09, 1H10, 2C08, 2C12, 2E09, 2F07, 2G02, 2G06, 2G11, 2H05, 3A02, 3A07, 3B03, and 3B07; Figure 1B) that reduced recombinant AMPK 2 activity to less than 10%. Compound C served as a positive control (Figure 1 A & B). We considered these 40 compounds as novel AMPK inhibitors with potential to attenuate neuronal death following ischemic injury. Seven agents attenuate zinc neurotoxicity in primary mouse neuronal cultures Various reports have shown that accumulation of chelatable intracellular zinc is at least one cause of neuronal cell death during ischemic stroke2-4. However, the mechanism of zinc neurotoxicity is quite complex, involving a number of derangements related to ischemia, such as calcium elevation5, oxidative stress, and apoptosis13. Our first objective was to assess the activity of these 40 compounds in terms of protection against zinc-mediated cell death -6ACS Paragon Plus Environment

Page 6 of 37

Page 7 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


showing the various types of cellular stress using primary cultures of mouse cortical neurons. Cultures were pretreated with one of the 40 novel agents or Compound C (Cpd C) 30 min before zinc treatment, briefly exposed for 10 min to 400 μM zinc (ZnCl2), and then replaced by fresh minimal essential medium (MEM) with one of the 40 agents or Cpd C. Zinc-induced neuronal cell death was assessed after 12-18 hours by measuring lactate dehydrogenase (LDH) release into the medium by dead cells. Surprisingly, the seven most potent inhibitors failed to diminish zinc-induced cell death at all, whereas seven different agents (1A09, 1C09, 1D06, 1E08, 1H10, 2G11, and 3A02; Figure 2A) yielded substantial declines in neuronal death that were comparable to Cpd C (Figure 2A). We also assessed nuclear morphology by staining with Hoechst dye (Figure 2B) and looking for features of apoptotic cell death. Fragmented and shrunken nuclei were significantly decreased due to treatment with these seven agents, indicating their potential for preventing zinc neurotoxicity. Two agents reduce diverse neuronal toxicities The pathogenesis of ischemic stroke brain is highly complex, involving not only zincmediated toxicity but also oxidative stress7, 31-34, glutamate excitotoxicity6, 35, apoptosis8, and more. Consequently, we examined the seven agents conferring protection against zincinduced neuronal death for resistance to several types of cytotoxicity seen with brain ischemia. We first determined whether H2O2- or Fe2+/3+-mediated oxidative stress was alleviated by these seven agents. As shown in Figure 3A, five compounds (1C09, 1D06, 1H10, 2G11, and 3A02) and Cpd C were markedly successful in this regard (Figure 3A). We then tested the protective effects of all seven agents against N-methyl-D-aspartate (NMDA)mediated neuronal death in mouse cortical cultures. Only agents 1H10 and 2G11 successfully attenuated NMDA-mediated excitotoxicity. Although others found that NMDA-induced excitotoxicity is mediated by AMPK and blocked by Cpd C20, we saw no protection from -7ACS Paragon Plus Environment


NMDA-receptor-mediated neurotoxicity in Cpd C-treated cultures (Figure 3B). This difference could be explained by variable expression levels of NMDA receptors and different culture conditions. Concannon et al. performed the experiments at DIV6-9, but we observed NMDA toxicity at DIV10-12. According to one report, the NR1 NMDA receptor subunit is already expressed by DIV5, peaks by DIV10-15 (>5-fold) and then remains unchanged through DIV6036. Therefore, it seems likely that Cpd C was ineffective because we used a more severe NMDA toxicity model. In addition, we only used serum-containing MEM for cortical cultures, but Concannon and colleagues used feeding medium containing serum-free Neurobasal medium with 2% B27 supplement37. Taken together, these differences may explain the discrepant effects of Cpd C on NMDA toxicity. Finally, we investigated whether these selected agents reduced the apoptosis when exposed to the following: 1) N,N,N′,N′-Tetrakis(2-pyridylmethyl)ethylenediamine (TPEN), a cellpermeant zinc chelator38-41; 2) etoposide (ETPS), a topoisomerase II inhibitor42; or 3) staurosporine (STSP), a potent cell-permeable inhibitor of various protein kinases43 (Figure 3C). Although STSP-induced apoptosis was attenuated by five compounds (1A09, 1D08, 1H10, 2G11, and 3A02) and by Cpd C (Figure 3C, right graph), agents 1H10 and 2G11 treatment protected from apoptosis in all three experimental conditions (Figure 3C). These results imply that agents 1H10 and 2G11 target both AMPK and other factors related to NMDA-receptor-mediated excitotoxicity and TPEN- or ETPS-induced apoptosis, whereas Cpd C did not prevent these types of neuronal death. Thus, 1H10 and 2G11 could protect against a broad range of neuronal death, surpassing Cpd C in their capacities to diminish ischemic brain injury. To visualize the inhibitory effects of various agents and Cpd C, primary neuronal cultures treated with H2O2, NMDA, or TPEN were Hoechst stained (Figure 3D). As expected, the -8ACS Paragon Plus Environment

Page 8 of 37

Page 9 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


numbers of shrunken nuclei (Figure 3D) declined substantially with 1H10 or 2G11 treatment, whereas Cpd C merely reduced oxidative injury. Agent 1A09 was used as a negative control for H2O2- or NMDA receptor-mediated neurotoxicity, and 1E08 was used as a positive control for TPEN-induced apoptosis. Agent 1C09 was used as both a positive and negative control for H2O2 neurotoxicity and NMDA receptor-mediated toxicity, respectively (Figure 3D). Based on previous reports30 and the result shown in Figure 1A, we identified seven chemical agents (1A03, 1A05, 1A06, 1A07, 1A08, 1D05, and 1F05) with the strongest AMPK 2 inhibition. However, none of these compounds reduced zinc-mediated neurotoxicity in cortical cultures (Figure 2A). Therefore, we examined the possibility that these compounds acted as robust aggregators rather than inhibitors of AMPK enzyme activity. We first checked whether these compounds are structurally similar to typical aggregators using a website (http://advisor.bkslab.org/). Four compounds (1A05, 1A06, 1D05, and 1F05) were similar to previously reported aggregators, and the other compounds (1A03, 1A07, and 1A08) also had high LogP values, suggesting the strong aggregator potential. Therefore, we determined whether 0.02% Triton X-100 could reverse the possible aggregate-based inhibition of AMPK. Four compounds (1A05, 1A06, 1A08, and 1D05) still significantly reduced AMPK activity in assays with 0.02% Triton X-100, but the other three compounds (1A03, 1A07, and 1F05) did not, indicating that these three compounds may have been false positives (Figure 4A). Although 1A05 did not reverse AMPK inhibition in the presence of Triton X-100, it could be a robust aggregator because it has a very high LogP value (4.8) and we observed clear aggregates in culture medium. In the 0.02% Triton X-100 condition, 1A05 aggregates were not sufficiently dissolved, and there was no reversal effect on AMPK activity. Moreover, agents 1A06 and 1D05 induced self-toxicity (Figure 4B), and 1A08 had no impact on H2O2--9ACS Paragon Plus Environment


mediated neurotoxicity (Figure 4C). Because zinc toxicity encompasses a variety of toxicity types5, 10-12, 44, it seems likely that 1A08 did not protect against zinc-induced neuronal death (Figure 2A) despite its potency as an AMPK inhibitor (Figure 1A). Supporting these, Cpd C significantly attenuated H2O2- or Fe2+/3+-mediated oxidative stress as well as zinc neurotoxicity (Figure 2 & 3A). Thus, even with remarkable in vitro inhibition of AMPK enzymatic activity, at least antioxidant action of agents may be needed for protection against zinc-mediated neuronal death. We selected agents 1H10 and 2G11 based on serial screening of potent candidates that demonstrated in vitro neuronal protection from diverse ischemic brain insults for in vivo study. Comparison with the chemical structure of 2G11 and 1H10 (Figure 5), there was no common structure in 2G11 and 1H10.

IC50 estimates and self-toxicity at high concentrations of 2G11 and 1H10 Prior to in vivo study, we measured the half maximal inhibitory concentration (IC50) of various neurotoxicities for each agent in primary mouse neuronal cultures. Before toxin exposure, we pretreated cultures with various concentrations of 2G11 or 1H10 (i.e., 0, 0.5, 1, 2.5, 5, 10, 20, 40, and 60 M). After toxin exposure, we quantified LDH release and calculated IC50 values (Prism 5; GraphPad Software, La Jolla, CA, USA). Initial estimates of IC50 values for 2G11 were 9.71 M in zinc neurotoxicity, 10.92 M in H2O2-induced oxidative stress, 4.90 M in NMDA receptor-mediated excitotoxicity, and 10.18 M in TPEN-induced apoptosis (Figure 6A). Respective IC50 values for 1H10 were 2.24 M, 1.44 M, 4.63 M, and 10.09 M (Figure 6B). In addition, we observed whether 2G11 and 1H10 displayed self-toxicity in mouse cortical cultures. Using graded concentrations of 2G11 or 1H10 in the absence of toxins, LDH release and counts of propidium iodide (PI)-positive dead cells were determined after 24 hours - 10 ACS Paragon Plus Environment

Page 10 of 37

Page 11 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(Figure 6C and 6D). Because 1H10 itself showed the autofluorescence, we observed cell viability after 1H10 treatment using CCK-8 kit instead of PI-positive dead cells counting. Self-toxicity for 2G11 was negligible to 60 M (Figure 6C), whereas 1H10 showed a gradual increase in cytotoxicity starting at 20 M and eventually proving fatal at a 60 M (Figure 6D). Thus, we faced a risk of drug overdose using 1H10 for in vivo animal testing, but this risk was not present for 2G11. In addition, a starting 1H10 dose of 20 M produced significant neurotoxicity; however, this same dose inhibited cytotoxicity under stress conditions, implying that the functional dose range of 1H10 for neuroprotection could be very narrow. Based on self-toxicity, 2G11 could therefore be a better candidate for drug development.

Use of permanent MCAO model to assess neuroprotective effects of 2G11 or 1H10 following focal cerebral ischemia We used a permanent MCAO rat model for brain ischemic injury. This approach produces the large and consistent infarcts with cores that always include the striatum, neocortex, amygdala, and hypothalamus45. All animals with a laser-Doppler flow (LDF) signal decrease to ≤40% were included in this study (Table 1). We injected vehicle or 2G11 (100 ng in 4 l) into the lateral ventricle 15 min prior to MCAO. Cerebral blood flow (CBF) did not differ significantly in 2G11- and vehicle-treated groups (Table 1). To measure infarct size, animal brains were harvested 24 hours later and stained using 2,3,5-triphenyl tetrazolium chloride (TTC) (Figure 7A, first set). Infarct volume (mm3) was calculated by integrating infarcted areas of serial coronal sections in the anteroposterior (AP) axis (Figure 7B). Then, the percent infarct volume was calculated to correct for the possible contribution of edema (Figure 7C). The infarct volume in the 2G11-treated group was markedly lower than that in the vehicle-

- 11 ACS Paragon Plus Environment


treated group. In contrast, weight loss and motor deficit scores 24 hours after ischemia did not differ between groups (Table 1). Next, we injected vehicle or 1H10 (75 ng in 4l) into the lateral ventricle 15 min prior to MCAO and found no difference in the CBF of the 1H10- and vehicle-treated groups (Table 1). The animal brains were harvested 24 hours later and stained using TTC (Figure 7A). Agent 1H10 reduced both direct and indirect infarct volume by 50% compared with the control group (Figure 7B and 7C). Both groups had similar neurologic scores and weight loss (Table 1). When we injected vehicle or Cpd C (100 ng/ 4l) in the same manner as the previous experiments, there was no between-group difference in infarct volume (Figure 7). Several other reports noted that AMPK activation following depletion of ATP/AMP is deadly during ischemic brain injury, so administering Cpd C has proved beneficial in reducing infarct volume in stroke animal models24, 26, 29. Our protocol was similar in design, but we could not reproduce the protective effect of Cpd C in our rat permanent MCAO model. There are two possible reasons for this observation. First, the fatal mechanisms of ischemia are complicated and not solely based on activation of AMPK. Agents 2G11 and 1H10 showed protection against all types of neurotoxicities (Figure 3), indicating they are also neuroprotective via mechanisms other than AMPK inhibition. Second, there could be procedural differences affecting the present results. Other groups performed transient ischemic surgery, restoring blood flow and tissue oxygenation after a period of time, whereas our MCAO model was permanent. Because excess ROS is produced during reperfusion46, a transient MCAO model is more vulnerable to oxidative injury. In support of this hypothesis, Cpd C showed considerable inhibition of oxidative-induced neuronal death (Figure 3A). Future studies


Page 12 of 37

Page 13 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


confirming the beneficial effects of 2G11 and 1H10 treatment in a transient MCAO model are needed. In summary, we identified new chemical agents that protect against ischemic brain injury, possibly due to their ability to inhibit AMPK. Ultimately, agents 2G11 and 1H10 showed greater efficacy than the traditional AMPK inhibitor, Cpd C, because they showed protective effects on the diverse types of neuronal death but not Cpd C. 1H10 displayed dose-dependent self-toxic cellular effects, but there was no toxicity at higher concentrations in 2G11-treated cortical cultures (Figure 6C & D), so 2G11 appears to be a preferable candidate for drug development. Furthermore, 1H10 contains one of the PAINS (pan assay interference compounds) substructures, which may interfere in biological assays. Baell and Holloway identified rhodanines, phenolic Mannich bases, hydroxy-phenylhydrazones, alkylidene barbiturates, alkylidene

hetero-cycles,

1,2,3-aralkylpyrroles,

activated

carbonylthiophenes, catechols, and quinones as PAINS47,

48.

benzofurazans,

2-amino-3-

Among these, 1H10 has an

alkylidene bearing five-membered heterocycles. Therefore, it is still possible that 1H10 is a false hit, even though we saw multiple positive signals from AMPK enzyme and LDH assays, PI- or Hoechst-stained morphology examinations, and TTC staining. However, 2G11 is absolutely not a PAINS, so its potential for drug development is superior to that of 1H10. In the present study, we injected these compounds using the intracerebroventricular route without crossing the blood brain barrier (BBB). To advance either drug for the treatment of stroke, we need precise data on mode of action and BBB permeability.

Conclusion



In the present study, we sought new AMPK-inhibiting agents targeted at identifying a new therapeutic strategy for treating strokes. Based on structure-based virtual screening, in vitro AMPK enzymatic assay, and quantifiable neuroprotection (LDH release) against zinc-, oxidative stress-, excitoxicity-, and apoptosis-induced primary neuronal death, we eventually selected two prospective agents, 2G11 and 1H10 (Figure 8). Cpd C failed to inhibit excitotoxicity and apoptosis-induced cell death. We also assessed drug efficacy using an in vivo animal model of stroke. Agents 2G11 and 1H10 significantly reduced brain infarct volumes, whereas Cpd C did not. These collective findings suggest that these novel chemical inhibitors of AMPK could be therapeutic candidates for stroke treatments given their excellent inhibitory effects on diverse types of ischemic neuronal death.

Methods In vitro AMPK kinase activity assay To measure AMPK enzymatic activity, we used the CycLex AMPK Kinase Assay Kit (MBL, Nagoya, Japan) and purified recombinant AMPK (A2/B1/G1, active; MBL) as described previously30. For inhibitor screening, we used dorsomorphin dihydrochloride (Cpd C; Tocris Bioscience, Bristol, UK) as a selective inhibitor of AMPK and the 118 potential AMPK inhibition drugs identified using structure-based virtual screening. We purchased the 118 compounds with >90% purity by NMR from InterBioScreen (Chernogolovka, Russia) and tested them as provided by the company. In brief, samples (3 ng of AMPK active enzyme with 1 or 10 µM AMPK inhibitor) were diluted in kinase buffer supplemented with 50 µM ATP. The assay was conducted according to the manufacturer’s protocol. The change in the chromogenic substrate was subsequently measured using absorbance at 450 nm measured on a microplate reader (Molecular Devices, Sunnyvale, CA, USA). - 14 ACS Paragon Plus Environment

Page 14 of 37

Page 15 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Primary mouse cerebrocortical neuronal cultures Mouse neuronal cultures were prepared from unborn mice at days 13-14 of gestation as described previously44. Briefly, isolated cerebral cortical tissues were minced in glutaminefree Dulbecco’s modified Eagle medium (Gibco-BRL, Grand Island, NY, USA) supplemented with 25 mM glucose, 44 mM sodium bicarbonate, 2 mM glutamine, 5% fetal bovine serum, and 5% horse serum for seeding onto poly-D-lysine (Sigma, St Louis, MO, USA) coated plates (SPL Life Sciences, Gyeonggi-do, South Korea). Seven hemispheres were seeded per 24-well plate or eleven hemispheres per 6-well plate. Cultures were incubated at 37°C in a humid 5% CO2 incubator and used at days in vitro (DIV) 10-12. Pregnant ICR mice were commercially available (ORIENT BIO, Seongnam, South Korea). This research was conducted in accordance with the guidelines for the care and use of mice in research under protocols approved by the Animal Care and Use Committee of Sejong University. Treatments with AMPK inhibitors and zinc solution Bathing medium was replaced with MEM (Gibco-BRL), and cells were pretreated for 30 min in 20 µM of the above-mentioned AMPK inhibitors. Then, cells were exposed for 10 min to 400 µM zinc solution (Hank’s balanced salt solution [Welgene, Gyeongsanbuk-do, South Korea] supplemented with 1.8 mM CaCl2, 1.22 µM MgSO4, 3.15 µM MgCl2, and 1.94 mM glucose). After, the zinc solution was washed out with MEM, and AMPK inhibitors were added post-treatment. In experiments involving zinc toxicity, cells were re-incubated in CO2. Estimates of cell death and viability Primary mouse neuronal cultures were pretreated with inhibitor for 30 min, and diverse toxicities were induced as follows: zinc solution, H2O2, Fe2+ or Fe3+ (all Sigma); NMDA - 15 ACS Paragon Plus Environment


(Abcam, Cambridge, UK); TPEN (Merck, Burlington, MA, USA); ETPS (Sigma); and STSP (Santa Cruz Biotechnology, Dallas, TX, USA). At the desired time points, cell death or viability were determined. To detect cell death, we checked LDH release into the bathing medium49. Each measured LDH value was scaled to maximum LDH release (equal to 100%) from cells near death. Cells were also stained (Hoechst 33258; Invitrogen, Carlsbad, CA, USA). Cultures were fixed in 4% paraformaldehyde for 15 min, permeabilized in 0.1 % Triton X-100 for 10 min, and then stained in 2 µg/ml Hoechst for 15 min. Dead cells were stained in PI (2.5 µg/ml bathing medium, Sigma) for 10 min and then washed with MEM. All processes were performed at room temperature. Stained cells were imaged under an inverted fluorescence microscope (EVOS Cell Imaging System; Thermo Fisher, Waltham, MA, USA) to obtain counts of PI-positive cells. To determine cell viability, the reagent from the Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Rockville, MD, USA) was added to the medium (1:10 volume). After a 2-hour incubation of plates in CO2, the color change was read at 450 nm using a microplate reader (Molecular Devices). We used 100 µM of NMDA to determine full primary neuronal death. Calculating IC50 Log-phase graphs showing the dose-dependent effects of drugs and calculated IC50 values were generated using a commercial scientific 2D graphing and statistics program (Prism 5; GraphPad Software). Permanent middle cerebral artery (MCA) ischemia Male Sprague-Dawley rats (mean weight, 317.6±18.8 g; age range, 8-9 weeks) were commercially obtained (Charles River Laboratories, Yokohama, Japan). All animals were allowed free access to food and water. Experiments were performed in accordance with the - 16 ACS Paragon Plus Environment

Page 16 of 37

Page 17 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


guidelines for the care and use of laboratory animals at the University of Ulsan in South Korea. Each rat was given 50 mg/kg zoletil (1:1 weight-based iletamine/zolazepam combination; Virbac, France) via intramuscular injection. Body temperatures were maintained at 36.0±0.5°C using a temperature control unit (Harvard Apparatus, Holliston, MA, USA) and a heat lamp during surgery and for 2 hours thereafter. Permanent ischemia in the MCA territory was achieved as previously described50, 51. Under a surgical microscope, right carotid arteries (common [CCA], external [ECA], and internal [ICA]) were identified and separated from the vagus nerve. The ECA and CCA were ligated using a 4-0 silk suture, and the ICA was temporarily clipped. A small incision in the CCA was made 1-mm proximal to the bifurcation, and a 4-0 suture (Ethylon nylon; Ethicon, Bridgewater, NJ, USA), the tip of which was firepolished to a diameter of 0.34-0.36 mm, was inserted into the ICA. When ICA clip was released, the suture was advanced into the proximal MCA to approximately 20 mm from carotid bifurcation. Each surgical operation was completed within 15 min. Laser-Doppler flowmetry The principles and technical details of laser-Doppler flowmetry (LDF) (BLF21D Monitor; Transonic Systems, Ithaca, NY, USA) have been described in detail elsewhere52. To place the LDF probe, a burr hole (2-mm diameter) was drilled 2-mm posterior and 6-mm lateral to the cranial bregma, taking care not to injure underlying dura. Animals were placed supine, and their skulls were firmly immobilized using a stereotaxic frame. Cerebral blood flow (CBF) recordings were obtained 20 min before and after MCAO and expressed the percent of the baseline value.



Drug administration via intracerebroventricular injection A 10-μl microsyringe (Hamilton Medical, Reno, NV, USA) was stereotactically introduced through a 2-mm burr hole into the right lateral ventricle (0.8-mm posterior and 1.2-mm lateral to the bregma, 3.8-mm deep), through which 2G11 (100 ng in 4 μl), 1H10 (75 ng in 4 μl), Cpd C (100 ng in 4 μl), or vehicle (4 μl 10% DMSO [Sigma] in saline) was injected prior to MCAO. The microsyringe was withdrawn 5 min after injection. Operators were blinded to drug identities. Gauging cerebral infarction At 24 hours after ischemia induction, rats were euthanized, and their brains were collected for analysis. Brain slices, 2-mm thick (RBM-4000C, ASI Instruments, Warren, MI, USA), were incubated in 2% TTC (Sigma) normal saline solution at 37°C for 30 min and then fixed in 4% paraformaldehyde50,

51, 53.

Digital images of brain slices were generated using a flatbed

scanner. Infarction volumes were quantified using image analysis software (Image-Pro, Media Cybernetics, Rockville, MD, USA). We estimated total infarction volumes by adding infarct volumes of successive AP coronal slices51, 53, 54. To correct for edema, infarct volumes were normalized to volumes of intact contralateral hemispheres51, 53, 55. Statistical analysis All values were expressed as mean±SEM. Group differences were analyzed using two-tailed t-test, setting statistical significance at p

Identifying new AMP-activated protein kinase inhibitors that protect

Recommend Documents