Bioorganic & Medicinal Chemistry Letters 21 (2011) 6184–6187

Contents lists available at SciVerse ScienceDirect

Bioorganic & Medicinal Chemistry Letters

journal homepage: www.elsevier .com/ locate/bmcl

Inhibition of xc� transporter-mediated cystine uptake by sulfasalazine analogs

Krupa Shukla a, Ajit G. Thomas b, Dana V. Ferraris b, Niyada Hin b, Rita Sattler a,b, Jesse Alt b, Camilo Rojas b,Barbara S. Slusher a,b, Takashi Tsukamoto a,b,⇑a Department of Neurology, Johns Hopkins University, Baltimore, MD 21205, USAb Brain Science Institute, Johns Hopkins University, Baltimore, MD 21205, USA

a r t i c l e i n f o

Article history:Received 22 June 2011Revised 20 July 2011Accepted 22 July 2011Available online 12 August 2011

Keywords:System xc

� cystineGlutamateAntiporterSulfasalazine

0960-894X/$ - see front matter � 2011 Elsevier Ltd.doi:10.1016/j.bmcl.2011.07.081

⇑ Corresponding author. Tel.: +1 410 614 0982; faxE-mail address: ttsukamoto@jhmi.edu (T. Tsukamo

a b s t r a c t

A series of sulfasalazine analogs were synthesized and tested for their ability to block cystine-glutamateantiporter system xc

� using L-[14C]cystine as a substrate. Replacement of sulfasalazine’s diazo group withan alkyne group led to an equally potent inhibitor, 2-hydroxy-5-((4-(N-pyridin-2-ylsulfa-moyl)phenyl)ethynyl)benzoic acid 6. Our SAR studies also revealed that the carboxylate group of sulfa-salazine is essential for its inhibitory activity while the phenolic hydroxyl group is dispensable.Truncated analogs lacking an N-pyridin-2-ylsulfamoyl moiety were less potent than sulfasalazine, butmay serve as more tractable templates because of their low molecular weight by applying a variety offragment growing approaches. Given that sulfasalazine is rapidly metabolized through cleavage of thediazo bond, these analogs may possess a more desirable pharmacological profile as system xc

� blockers,in particular, for in vivo studies.

� 2011 Elsevier Ltd. All rights reserved.

The xc� cystine-glutamate antiporter is a sodium independent

amino acid transporter that transports extracellular cystine in ex-change for intracellular glutamate. Because cystine is a precursorfor glutathione (GSH), system xc

� is believed to play a critical rolein intracellular synthesis of GSH and hence cellular redox regula-tion. Accumulating evidence suggests that the xc

� antiporter is vi-tal to the growth of cancer cells dependent on extracellularcystine.1–3 For example, the xc

� cystine-glutamate antiporter ap-pears to be universally up-regulated in glioma, while expressionof the GLT1/EAAT2 glutamate transporter is compromised. As a re-sult, the glutamate released from gliomas accumulates in theextracellular space, causing excitotoxic death to surrounding neu-rons. This provides a significant growth advantage to the cystinerich glioma cells by offering space to expand.4

The system xc� also plays a critical role in maintaining intracel-

lular levels of glutathione for the growth and survival of a variety ofcancer cells such as lymphoma,5 prostate,6 lung cancer,7 pancreaticcancer8 and others.1 These findings prompted a search for inhibitorsof system xc

� to explore the therapeutic benefit of xc� system block-

ade in glioma and other forms of cancer. Two structurally distinctsmall molecules, (S)-4-carboxyphenylglycine 1 (S-4CPG)9 and sul-fasalazine 2,5 were reported to inhibit system xc

�. Although bothcompounds have served as useful pharmacological tools to studythe function of xc

� system in vivo,10,11 each has a unique issue thathinders its utility as a drug besides its low potency. S-4CPG not onlyinhibits xc

� system but also acts as a group I metabotropic receptor

All rights reserved.

: +1 410 614 0659.to).

antagonist.12 This may raise concerns about off-target effectscaused by the blockade of the metabotropic receptors. On the otherhand, sulfasalazine 2 is known to be extensively metabolized in theintestine by azoreductases and only a fraction of the parent drugenters systemic circulation.13 Since its two key metabolites, sulfa-pyridine 3 and 5-aminosalicylic acid 4, do not interact with systemxc�,5 sulfasalazine may not elicit in vivo efficacy against glioma by

oral administration, the only approved administration route for thisdrug.

Sulfasalazine 2S-4CPG (S)-1

H2N

OH

CO2H

5-Aminosalicylic acid 4Sulfapyridine 3

HO2CCO2H

NH2

S

NN CO2H

OH

NH

N OO

S

NH2

NH

N OO

For these reasons, efforts to identify new system xc� inhibitors

are of significance. For example, Natale and his co-workers evalu-ated analogs of AMPA (amino methylisoxazole propionic acid) asinhibitors of system xc

�.14 Thompson and his co-workers reported

c d9 10

OH

CO2CH3

OH

CO2Ha

CO2CH3

OH

bCO2H

OH

5 6

7 8

5

S

CO2CH3

OH

NH

N OOS

CO2H

OH

NH

N OO

SNH

N OOS

NH

N OO

SNH

N OOS

NH

N OO

Scheme 1. (a) and (b) See Ref. 16; (c) Pd/C, H2, MeOH/EtOAc, rt, overnight, 54%; (b) 1 M KOH, 90 �C, overnight, 61%.

I11

b

c

14 15

CO2CH3 a

CO2CH3

12

13

S

CO2HNH

N OO

SNH

N OO

CO2H

SNH

N OO

SNH

N OO

Scheme 2. (a) Pd(PPh3)2Cl2, CuI, 12, degassed NEt3/THF, 60 �C, 4 h, 44%; (b) 1 M NaOH, 90 �C, 18 h, 74%; (c) Pd/C, H2, MeOH, 15 h, 33%.

K. Shukla et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6184–6187 6185

conformationally-restricted amino acid analogs bearing a distalsulfonic acid as selective inhibitors of system xc

� over the vesicularglutamate transporter.15

In the present study, we have evaluated various analogs of sul-fasalazine for their ability to inhibit system xc

�. Our objective istwofold. The first is to determine whether we can identify meta-bolically stable analogs with xc

� inhibitory potency similar to thatof sulfasalazine. The second is to identify a minimum fragment ofsulfasalazine (or its analogs) required for system xc

� inhibition.We first synthesized various sulfasalazine analogs in which the

diazo group is replaced by two carbons connected by a single, dou-ble, or triple bond. Compounds 5–8 were synthesized according tothe method described previously (Scheme 1).16 Catalytic hydroge-nation of compound 5 afforded fully saturated derivative 9, whichwas subsequently converted into compound 10 via base-mediatedhydrolysis.

In order to assess the role of the hydroxyl group on the benzoatemoiety, deoxy analogs 14 and 15 were synthesized using Sono-gashira coupling reaction of 11 and 12 as a key step (Scheme 2).The coupling product 13 was hydrolyzed to compound 14, whichwas subsequently reduced to a fully saturated analog 15 by cata-lytic hydrogenation.

Several different assays have been developed for the measure-ment of system xc

� antiporter activity. Although L-[3H]glutamate

uptake assays have been commonly used for screening pur-poses,14,15,17 we chose to use L-[14C]cystine as an extracellular sub-strate in order to accurately simulate physiological function ofsystem xc

�. The assay was adapted from the previously reportedmethod using 35S-L-cystine as a substrate.4 Human U87 glioma cellline, which is known to overexpress system xc

�, was used as asource for the antiporter.10 IC50 values were determined by mea-suring L-[14C]cystine uptake in the presence of the test compoundsat various concentrations.18

As shown in Table 1, while S-4CPG (S)-1 inhibited system xc�-

mediated cystine uptake with an IC50 value of 15 lM, its enantio-mer R-4CPG (R)-1 had no activity up to 1 mM. This is consistentwith the earlier finding of Sontheimer’s group.19 Sulfasalazine 2also showed an inhibitory potency similar to that of S-4CPG (S)-1with an IC50 value of 30 lM in our assay. Consistent with the pre-vious findings,5 the two key metabolites of sulfasalazine, sulfapyr-idine 3 and 5-aminosalicylic acid 4, were inactive in our assay.Although various system xc

� inhibitory potencies have been re-ported for S-4CPG (S)-1 and sulfasalazine 2 by differentgroups,5,10,17,19–21 the variability is presumably due to the differ-ence in the assay conditions such as cell type, substrate concentra-tions, and transport direction. Given the potent inhibition by thetwo known inhibitors S-4CPG (S)-1 and sulfasalazine 2 coupledwith the complete loss of potency by the inactive enantiomer

Table 1Inhibition of system xc

�-mediated cystine uptake by 4CPG enantiomers, sulfasalazine,and its metabolites

Compd22 Structure IC50a (lM)

(S)-1 HO2CCO2H

NH215

(R)-1 HO2CCO2H

NH2>1,000

2

S

NN CO2H

OH

NH

N OO

30

3S

NH2

NH

N OO

>1,000

4

CO2H

OH

H2N>1,000

a Concentration at which the compound reduces the cellular cystine uptake by50% compared to the solvent control in U87 cell lines.

Table 2Inhibition of system xc

�-mediated cystine uptake by sulfasalazine analogs

Compd22 Structure IC50a (lM)

5 SNH

N OOCO2CH3

OH >1,000

6 SNH

N OOCO2H

OH 30

7

S

CO2CH3

OH

NH

N OO

900

8

S

CO2H

OH

NH

N OO

70

9

S

CO2CH3

OH

NH

N OO

>1,000

10

S

CO2H

OH

NH

N OO

700

14 SNH

N OOCO2H

75

15

S

CO2HNH

N OO

920

16 SNH

N OOCO2H

OH

CH3

70

a Concentration at which the compound reduces the cellular cystine uptake by50% compared to the solvent control in U87 cell lines.

Table 3Inhibition of system xc

�-mediated cystine uptake by truncated analogs ofsulfasalazine

Compd22 Structure IC50a(lM)

17 NN CO2H

OH

200

18

CO2H

OH150

19

CO2H270

a Concentration at which the compound reduces the cellular cystine uptake by50% compared to the solvent control in U87 cell lines.

6186 K. Shukla et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6184–6187

(R)-1, and metabolites 3 and 4, our assay system appears to be wellsuited for screening new compounds.

Several interesting SAR trends emerged from the carbon analogsof sulfasalazine 5–16 (Table 2). It is evident from the results thatthe free carboxylate group plays an important role in the bindingof sulfasalazine to system xc

�. All three methyl ester derivatives5, 7, and 9 exhibited negligible inhibitory activity. Among thefree-carboxylate containing analogs, compounds 6, 8, 14, and 16showed system xc

� inhibitory potency comparable to that of sulfa-salazine. Fully saturated analogs 10 and 15 were significantly lesspotent even though they contain the free carboxylate group. A cer-tain degree of conformational rigidity seems important for bindingto system xc

�. In contrast, the phenolic hydroxyl group on the ben-zene ring appears dispensable as deoxygenation of compound 6 re-sulted in only twofold increase in IC50 value (compound 14). Thisobservation is important because the phenolic hydroxyl groupcould constitute another metabolic liability in addition to the diazobond. Incorporation of a methyl group into the pyridine ring of 6had little effect on system xc

� inhibition as compound 16 is simi-larly potent in inhibiting cystine uptake. It is worth noting thatcompound 16, susalimod, was originally designed as a metaboli-cally stable analog of sulfasalazine against azoreductase. Susali-mod was reported to be metabolically stable in humans with aplasma half-life of 16 h,23 clearly demonstrating the advantage ofreplacing the diazo group of sulfasalazine.24

In an attempt to identify the minimum required structure forthe system xc

� inhibition, we also explored some truncated ana-logs. Knowing that sulfapyridine 3 and 5-aminosalicylic acid 4are inactive and that the benzoate group is essential for bindingto system xc

�, we chose truncated analogs lacking the N-pyridin-2-ylsulfamoyl moiety 17–19 (Table 3). While none of them wereas potent as their full-length counterparts, they still inhibited sys-tem xc

� at submillimolar concentrations.In summary, we tested a series of sulfasalazine analogs for their

ability to block cystine-glutamate antiporter system xc�. Replace-

ment of sulfasalazine’s diazo group with an alkyne or alkene groupled to similarly potent inhibitors 6 and 8, respectively. Given thatsulfasalazine is rapidly metabolized through cleavage of the diazobond, these carbon analogs may serve as more valuable toolcompounds to further understand the therapeutic utility of system

K. Shukla et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6184–6187 6187

xc� inhibition. This is particularly important in any in vivo studies

where sulfasalazine often exhibits its pharmacological effectsunrelated to system xc

� inhibition through the action of one ofits metabolites, sulfapyridine 3. Our SAR studies also revealed thatthe carboxylate group of sulfasalazine is essential for its inhibitoryactivity while the phenolic hydroxyl group is dispensable. Thus,compound 14 provides additional benefit by eliminating the possi-bility of phase II metabolism such as glucuronidation. Truncatedanalogs lacking an N-pyridin-2-ylsulfamoyl moiety were some-what less potent than sulfasalazine, but their low molecular weight(MW <250) enables them to serve as more tractable templateswith respect to ligand efficiency. Future studies will aim atstructural optimization of compound 14 as well as a variety offragment growing approaches based on compound 19. Obviously,one of the major challenges remains to be the weak binding po-tency of the existing system xc

� inhibitors reported to date. Morepotent xc

� inhibitors would be necessary not only to translatethem into practical therapeutic interventions but also to examinethe toxicological consequences of cystine uptake blockade innon-malignant cells. Furthermore, in the case of glioma treatment,the currently available inhibitors might pose problems of poor CNSpenetration due to the existence of at least one carboxylate group.To this end, a screening of a large chemical library of structurallydiverse compounds should provide a viable alternative to the con-ventional SAR studies based on S-4CPG and sulfasalazine. Develop-ment of a high-throughput screening for system xc

� is currentlyunderway to identify novel and potent inhibitors structurally dis-tinct from S-4CPG and sulfasalazine.

Acknowledgments

This work was in part supported by the Johns Hopkins Brain Sci-ence Institute through the NeuroTranslational program. We thankMs. Alice Nunes and Ms. Yoko Ayukawa for their technical assis-tance with cell cultures.

Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.bmcl.2011.07.081.

References and notes

1. Lo, M.; Wang, Y. Z.; Gout, P. W. J. Cell Physiol 2008, 215, 593.

2. Sontheimer, H. J Neurochem 2008, 105, 287.3. Savaskan, N. E.; Hahnen, E.; Eyupoglu, I. Y. J. Cell Physiol 2009, 220, 531.4. Ye, Z. C.; Rothstein, J. D.; Sontheimer, H. J. Neurosci 1999, 19, 10767.5. Gout, P. W.; Buckley, A. R.; Simms, C. R.; Bruchovsky, N. Leukemia 2001, 15,

1633.6. Doxsee, D. W.; Gout, P. W.; Kurita, T.; Lo, M.; Buckley, A. R.; Wang, Y.; Xue, H.;

Karp, C. M.; Cutz, J. C.; Cunha, G. R.; Wang, Y. Z. Prostate 2007, 67, 162.7. Guan, J.; Lo, M.; Dockery, P.; Mahon, S.; Karp, C. M.; Buckley, A. R.; Lam, S.; Gout,

P. W.; Wang, Y. Z. Cancer Chemother. Pharmacol. 2009, 64, 463.8. Lo, M.; Ling, V.; Wang, Y. Z.; Gout, P. W. Br. J. Cancer. 2008, 99, 464.9. Ye, Z. C.; Sontheimer, H. Glia 1999, 25, 270.

10. Chung, W. J.; Lyons, S. A.; Nelson, G. M.; Hamza, H.; Gladson, C. L.; Gillespie, G.Y.; Sontheimer, H. J. Neurosci. 2005, 25, 7101.

11. Savaskan, N. E.; Heckel, A.; Hahnen, E.; Engelhorn, T.; Doerfler, A.; Ganslandt,O.; Nimsky, C.; Buchfelder, M.; Eyupoglu, I. Y. Nat. Med. 2008, 14, 629.

12. Hayashi, Y.; Sekiyama, N.; Nakanishi, S.; Jane, D. E.; Sunter, D. C.; Birse, E. F.;Udvarhelyi, P. M.; Watkins, J. C. J. Neurosci. 1994, 14, 3370.

13. Azadkhan, A. K.; Truelove, S. C.; Aronson, J. K. Br. J. Clin. Pharmacol 1982, 13,523.

14. Patel, S. A.; Rajale, T.; O’Brien, E.; Burkhart, D. J.; Nelson, J. K.; Twamley, B.;Blumenfeld, A.; Szabon-Watola, M. I.; Gerdes, J. M.; Bridges, R. J.; Natale, N. R.Bioorg. Med. Chem. 2010, 18, 202.

15. Etoga, J. L.; Ahmed, S. K.; Patel, S.; Bridges, R. J.; Thompson, C. M. Bioorg. Med.Chem. Lett. 2010, 20, 2680.

16. Agback, H.; Ahrgren, L.; Berglindh, T.; Haraldsson, M.; Smedegard, G.; Olsson,L.-I. U.S. Patent 5,302,718, 1994.

17. Patel, S. A.; Warren, B. A.; Rhoderick, J. F.; Bridges, R. J. Neuropharmacology2004, 46, 273.

18. U87 human glioblastoma-astrocytoma cells (ATCC number HTB14) wereseeded at 0.067 � 106 cells per well and grown to confluency in 12-wellplates for 6 days. On the day of the experiment, the cells were washed withpre-warmed (37 �C) sodium free uptake buffer (contents in mM: cholinechloride 137.5, KCl 5.36, KH2PO4 0.77, MgSO4 0.71, CaCl2 1.1, Glucose 10 andHEPES 10, pH 7.4) and incubated at 37 �C, for 15 min using 80 lM cystine andat a specific activity of 1.35 lCi/lmol in the presence and absence of inhibitors.L-[14C(U)]-cystine uptake was terminated by three washes with ice-colduptake buffer. Subsequently, cells were lysed with 0.1 N NaOH and theradioactivity in the cells measured using a liquid scintillation counter and thecounts normalized to the protein contents. The data were then normalized tothe ‘totals’ (uptake at 37 �C in the absence of inhibitor) and ‘blanks’ (uptake at0 �C in the absence of inhibitor) and presented as percent inhibition ((1-(Unknown-blanks)/(totals-blanks)) � 100). Subsequently, the inhibitoryconstant at 50% inhibition (IC50) was calculated using GraphPad Prism.

19. Sontheimer, H. J.; Ye, Z.-c. U.S. Patent 6,197,820, 2001.20. Gochenauer, G. E.; Robinson, M. B. J. Neurochem 2001, 78, 276.21. Baker, D. A.; Xi, Z. X.; Shen, H.; Swanson, C. J.; Kalivas, P. W. J. Neurosci 2002, 22,

9134.22. S-4CPG (S)-1 and R-4CPG (R)-1 were purchased from Tocris. Sulfasalazine 2,

sulfapyridine 3, and 5-aminosalicylic acid 4 were purchased from Sigma.Compound 16 was prepared according to the previously reported method (Ref.17). Compounds 17, 18, and 19 were purchased from Sigma, Ryan Scientific, andOakwood Products, respectively.

23. Påhlman, I.; Wilen, M.; Bowald, S. J. Pharm Pharmacol 1999, 51, 61.24. The use of susalimod as an inhibitor of system xc

� was first suggested inChung, J. W.; Sontheimer, H. J. WO 2006049957, 2006.