Graviola Mclaughlin

7/29/2019 Graviola Mclaughlin

1/7

Cancer Letters 115 (1997) 73-79

CWERLETTERSThe Annonaceous acetogenin bullatacin is cytotoxic againstmultidrug-resistant human mammary adenocarcinoma cellsNicholas H. Oberlies, Vicki L. Croy, Marietta L. Harrison, Jerry L. McLaughlin*

Department of Medicinal Chemistry and Molecular Pharmacology, School of Phurmacy and Pharmacal Scierwe,~,Purdue University, West Lafayetfe, IN 47907-133.3, USA

Received 19 December 1996; revision received 14 January 1997; accepted 14 January 1997

AbstractCytotoxic effects of the Annonaceousacetogenin,bullatacin, were studied in multidrug-resistant (MDR) human mammaryadenocarcinoma MCF-7/Adr) cells vs. the parental non-resistant wild type (MCF-7&t) cells. Buktacin was effectivelycytotoxic to the MCF-7/Adr cells while it was more cytostatic to the MCF-7/wt cells. ATP depletion is the mode of action ofthe Annonaceousacetogenins,and these agentsoffer a special advantage n the chemotherapeutic reatment of MDR tumorsthat have ATP-dependent mechanisms.0 1997 Elsevier Science Ireland Ltd.

Keywords: Acetogenins; Bullatacin; Multidrug resistance;P-gp; Mammary adenocarcinoma - ---... -_--_----1. Introduction

In many cancer patients, the eventual cause of deathis not from the original, chemotherapeutically respon-sive, tumor cells but from the surviving tumor cellsthat develop resistance to both the original antineo-plastic agent and to new, mechanistically and structu-rally unrelated compounds [ 1,2]. This phenomena hasbeen accurately coined multidrug resistance (MDR)and is often characterized by the increased expressionof a 170 kDa phospholipid glycoprotein (P-gp). TheP-gp forms a channel in the cell membrane whichserves to extrude anticancer agents before antitumorefficacy can be realized. The active export of antineo-plastic compounds requires energy supplied by ATP*Correspondingauthor.Tel.: +l 317 4941455; fax: +l 317

4941414; e-mail: [email protected]

cleavage, and two ATP binding sites for this ATPaseactivity have been identified on the cytosolic side ofthe P-gp [3-71.The Annonaceous acetogenins are a relatively newclass of biologically active natural compounds [S- 111which act to decrease ATP production by inhibitingcomplex I (NADH:ubiquinone oxidoreductasej af themitochondrial electron transport system (ETS) [12-141. A second, related, mode of action is the inhibitionof an ubiquinone-linked NADH oxidase, involved insubstrate level phosphorylation, which is constitu-tively expressed in the plasma membranes of cancercells and only transiently expressed in the membranesof normal non-cancerous cells [ 15J. A recent studyreported that a series of related acetogenin compoundssignificantly inhibited the growth of several differenttransformed cell types, including an adriamycin-resis-tant murine mammary cell line (M 17/Adr), while only

0304-3835/97/$17.00 0 1997 Elsevier Science Ireland Ltd. All rights reservedPII SO304-3835(97)04716-2


2/7

74 N.H. Oberlies et al. /Cancer Letters 115 (1997) 73-79minimally affecting the growth of non-cancerous ratG.I. epithelial (118) cells [16]. Mean bar graphs,showing cytotoxicities in the NC1 panel of tumorcells [17], revealed that MDR cell lines are oftenfive times more susceptible to the acetogenins thanthe parent, non-MDR cell lines.Numerous studies have explored the use of adju-vant compounds that serve to impede the action of theP-gp by competitively blocking the efflux channel inan effort to increase the exposure time of antineoplas-tic compounds within MDR cell types [6,18-231.However, in vivo trials, using verapamil as an adju-vant, failed due to verapamil-induced hypotension[24]. Alternatively, we have approached the MDRproblem by investigating the hypothesis that the bio-chemical difference between MDR and parental can-cer cells, i.e. the ATP-dependent P-gp, results in ahigher demand for ATP in the MDR cancer cells.Therefore, the Annonaceous acetogenins, because oftheir ability to decrease ATP levels [14], were inves-tigated for their effect on the growth of MDR cells.We herein report that the adriamycin (MDR)-resistanthuman mammary adenocarcinoma (MCF-7/Adr) cellline is more susceptible than its parental human mam-mary adenocarcinoma (MCF-7/wt) cell line to treat-ment with the acetogenin bullatacin. Bullatacin wasfound to be cytotoxic to the MDR MCF-7/Adr cells,but it was only cytostatic to the MCF-7/wt cells.

2. Methods2.1. Materials and reagents

Bullatacin was isolated and characterized in ourlaboratory as previously described and reviewed [8-111. Adriamycin, vincristine, vinblastine, penicillin,streptomycin, poly-D-lysine, and Nonidet NP-40were purchased from Sigma, St. Louis.2.2. Culturing and plating of MCF-7 cell lines

The MCF-7/wt (wild type human mammary adeno-carcinoma) and the MCF-7/Adr (adriamycin-resistanthuman mammary adenocarcinoma) cell lines werekindly provided by Craig Fairchild of NIIVNCI.Both were maintained in RPM1 (Gibco; Grand Island,NY) with 10% heat-inactivated fetal calf serum (Inter-

gen, Purchase, NY) and 1% penicillin/streptomycin(PS) and were transferred twice weekly at a ratio of1:6 for the former and 1:3 for the latter. The MCF-7/Adr cells were originally isolated from the MCF-7/wtcells by growth in the presence of 10 PM adriamycin.They retain their adriamycin resistance for at least 6months in its absence.The 96-well microtitre plates (Falcon Labware,Oxnard, CA) were coated with poly-D-lysine inorder to facilitate cell adherence [25]. A 50 ~1 aliquotof poly-D-lysine (100 pg/ml in distilled water filteredthrough a 0.2 pm filter) was added to each well andincubated at room temperature for at least 30 min. Theplates were rinsed twice with sterile water andallowed to dry overnight in a sterile environment.The plates can be stored for several weeks at 4C.For all of the experiments, MCF-7/wt and MCF-7/Adr cells were plated at 5 x lo3 and 1.5 x IO4 cells/ml, respectively, on the poly-D-lysine-coated 96-wellmicrotitre plates, in a total volume of 200 ~1 of med-ium per well and incubated overnight in a humidifiedCO2 incubator at 37C. We previously determinedthat the cells could tolerate 1% by volume of 95%ethanol without significantly affecting their growth(data not shown); this facilitated the dilution and dis-persion of bullatacin. Thus, 24 h after plating the cells,100 ~1 of fresh medium was added to the test wellsfollowed by the indicated concentration of bullatacinin a total volume of 3 ~1 of 95% ethanol; the plateswere then incubated at 37C in the humid atmosphere.Adriamycin was used as a positive control, while sixwells per cell line were used as a standard vehiclecontrol.2.3. Determining the amount of cell growth using thebicinchoninic acid (BCA) protein assay

The amount of cell growth inhibition was deter-mined using a bicinchoninic acid (BCA) proteinassay reagent kit (Pierce, Rockford, IL) [26]. Thecells were washed with an eight-channel plate washer(Flow Laboratories) with phosphate buffered saline(PBS). Ten ~1 of non-ionic detergent solution (1%by volume of Nonidet NP-40 in sterile water) wasadded to each well in order to solubilize the cells,followed by 200 ,~l of BCA working reagent (a 5O:lmixture of base reagent/4% copper sulfate solution).The plates were incubated at 37C for 30 min, and the


3/7

N.H. Oberlies et al. /Cancer Letters I IS (1997) 73 -74 7s

0.7 :-0.6 -0.5:0.4;

---- -__ii A.-_-. --L-L , _-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2

log[caneeatntion (wg/mL)I

1.2 ----- 7 ---- I TV - --v,--IT,,1.1. MCP-7/wt Cells .I[0.91 \0.81 \ \0.7 i G0.6. \O.S/- \ \0.4 / \0.3 I- \0.2 4 \ i 1

.,O -9 -8 -7 -6 .5 -4 -3 -2 -1 0 1 2IgIconcentrstion (pg/mL)]

Fig. I. The effect of 6 days of exposure to the standard antineo-plastic drugs, adriamycin, vincristine, and vinblastine, againstMCF-7/Adr (A) and MCF-7/wt (B) cells. Values are expressed asa percentage of the vehicle-treated controls with each point repre-senting the normalized average of four values and the error barsrepresenting the standard deviation about that average. The con-centration values are log[dose] with un its of pg/ml.absorbance was determined at 570 nm on a DynatekMR 600 microplate reader.2.4. Cell growth assay

For the standard 7-day assay (Figs. 1 and 2), theamount of cell growth inhibition was determined atthe end of 6 days of exposure to the test compoundsusing the BCA assay described above. The averageabsorbance in the vehicle control wells was calculatedinitially so that all subsequent determinations could benormalized to this average; the abscissas on Figs. 1and 2 represent percent of control. Each test com-

pound was examined in duplicate on two differentplates so that II = 4. The average normalized absor-bance is plotted at each respective concentration(log[concentration]) with the error bars representingthe standard deviation of the four determinations.Alternatively, the cell growth inhibition was deter-mined every 24 h after the original plating of thecells and loading of test compound using the BCAassay described above (Fig. 3). In order to calculatean average absorbance and a standard deviation, sixtest wells were used for the vehicle-treated controls(n = 6), while four test wells were used in the bulla-tacin-treated wells (n = 4). The results are displayedsuch that the abscissa represents an absolute absor-bance while the ordinate represents time in hours.

For the re-feeding experiments (Fig. 4). cell growthinhibition was determined every 24 h after the originalplanting of the cells and loading of test compoundusing the BCA assay described above. On day three,half of the remaining plates had their mediumremoved using a vacuum aspirated sterile syringe,and fresh medium was added followed by incubation.Thus, on days 4 through 7, two plates were examinedeach day (one standard and one re-feed). For theseexperiments, the average absorbance of six wells(n = 6) for both the control wells and the test com-pound wells is plotted such that the abscissa repre-sents the absolute absorbance; the error bars depict,.2..,------- .--T../. .,. , -,1.1.

1 I0.9 :0.8 :

Bh 0.7.I 0.6

Fig. 2. Comparison of the effect of 6 days of exposure to buktacinagainst the MCF-7/wt vs. MCF-7/Adr cells. Values are expressedas a percentage of the vehicle-treated controls with each pointrepresenting the normalized average of four values and the errorbars representing the standard deviation about that average. Theconcentration values are log[dose] with un its of &ml.


4/7

N.H. Oberlies et al. /Cancer Letters 115 (1997) 73-79

0.6 1

I 7

1

0.0

0.6 -

0.4 -

0.2 -!

t.,..I..,.,,.,.I,.,,i0 50 100 150 200

uw (hours)

Fig. 3. Periodic analysis (every 24 h) of serial dilutions of bullata-tin against MCF-7/Adr (A) vs. MCF-7/wt (B) cells. Values areexpressed as an average absorbance (n = 6 for the vehicle treatedcontrol, n = 4 for the bullatacin-treated wells) and the error barsrepresent the standard deviation about that average. The concentra-tion values are log[dose] with units of pg/ml.the standard deviation, and the R shows when re-feeding was initiated.

3. Results and discussionFig. 1 shows the effect of three standard antineo-plastic compounds on the growth of adriamycin-resis-tant and parental (wild type) MCF-7 cells. The MCF-7/Adr cells were resistant to treatment with eitheradriamycin, vincristine or vinblastine (Fig. 1A) asanticipated for these multidrug-resistant (MDR)cells [3-7,271. The concentration of adriamycin thatinhibited cell growth by 50% (ICss value) in the MDRMCF-7/Adr cells was greater than 1 &nl, whereas,

against the parental MCF-7/wt cells (Fig. lB), the ICsovalue was 5 x 10m2pg/ml. The differences in KS0values between the two cell lines were even greaterwith vincristine and vinblastine and served to illus-trate the MDR phenomenon.In contrast to adriamycin, vincristine, and vinblas-tine, bullatacin was effective at inhibiting the growthof the MDR MCF-7/Adr cells and exhibited a lineardose-response curve over a concentration range of 1 Opg/ml to 1.0 x lOA pg/ml (Fig. 2). However, over thesame concentration range, there was a plateau near theICZO alue against the parental MCF-7/wt cells. At themost concentrated dose of 1 Opg/ml, bullatacin inhib-ited nearly all of the growth of the MDR MCF-7/Adrcells but only 50% of the growth of the MCF-7/wtcells (Fig. 2). This observation was examined furtherby analyzing the growth of both cell lines periodicallyover the 7 days of the assay (Fig. 3). In the MDRMCF-7/Adr cells, bullatacin inhibited cell growth byvarying amounts in a dose-dependent fashion, i.e.nearly zero cell growth at the most concentrateddose of 1 O ,r&ml vs. nearly 100% cell growth at theleast concentrated dose of 1.0 x lOA pg/ml (Fig. 3A).Alternatively, the cell growth of the parental MCF-7/wt cells was only inhibited by 50%, relative to thegrowth of the vehicle-treated controls, regardless ofthe dose of bullatacin (Fig. 3B). The data in Figs. 2and 3, therefore, illustrate that a linear dose-responsecurve in the MCF-7/Adr cells after a 6 day treatment(Fig. 2) reflects dose-dependent cell growth when ana-lyzed on a daily basis (Fig. 3A). Likewise, dose-inde-pendent cell growth was confirmed in the MCF-7/wtcells both at the end of 6 days of bullatacin exposure(Fig. 2) as well as over the 7 days of the assay (Fig.3B).Both cell lines were then analyzed to determine ifthey were still viable after bullatacin treatment (Fig.4A,B). A 24 h exposure to bullatacin (1 .O pg/ml) wascytotoxic to the MDR MCF-7/Adr cells since thesecells were not able to grow after being fed withfresh medium (Fig. 4A). However, the MCF-7/wtcells were able to grow to a level near that of thevehicle treated control upon being fed fresh medium;thus, bullatacin was more cytostatic than cytotoxic tothese wild type cells. A similar re-feeding experiment,using 1.0 &ml of adriamycin, showed oppositeresults; the MCF-7/Adr cells were completely unaf-fected by the antineoplastic agent (Fig. 4C), whereas


5/7

h!H. Oberlies et al. /Cancer Letters Il.5 (1997) 73-79 77

0.6i.

0.4 j/

0.2 c

0 .-AA . . i_- . - 2. --_ .l--L ..-. -.. i

0 so 100 150 200lime l3oun)

0.6

x/1

Fig. 4. Periodic analysis (every 24 II) of 1.0 &III of bullatacin (A,B) vs. 1.0 &ml of adriamycin (C,D) against MCF-7/Adr (A,C) vs. MCI;-71wt (B,D) cells with re-feeding of fresh media in half of the plates. Values are expressed as an average absorbance (n = 6 for both control anddrug-treated wells) and the error bars represent the standard deviation about that average. The R depicts when refeeding was initiated.the MCF-7/wt cells were no longer viable after adria-mycin treatment (Fig. 4D).The acetogeninsare potent inhibitors of ATP pro-duction via their interaction with complex I in thernitochondria [12-141 and an NADH oxidase at theplasma membrane [15]. In cells in which ATPrequirements are elevated, such as in MDR MCF-71A& cells that require ATP to drive the P-gp transpor-ter 13-71, bullatacin was cytotoxic. These cells werenot viable after only 2 days of bullatacin exposure(Fig. 4A). Alternatively, in the parental MCF-7/wtcells, it appears that ATP production is limited bybullatacin treatment to induce only a static responsein cellular growth and replication. These cells grew tonearly the same level as the vehicle-treated controlafter bullatacin was removed (Fig. 4B). Thus, suchMDR cell types seem o be more susceptible o ATPdepletion than the parental cells from which they were

derived, and this biochemical difference may beexploitable by the chemotherapeuticuse of the Anno-naceousacetogenins.In conclusion, the Annonaceous acetogenins,suchas bullatacin, may have a unique potential as che-motherapeutic agents. Earlier in vivo studies haveshown bullatacin to be effective at only 50 pg/kgper day against L1210 mutine leukemia in normalmice and against A2780 human ovarian xenograftsin athymic mice [14]. Furthermore, bullatacin waseffective against multidrug resistance n both the pre-viously reported adriamycin-resistant murine mam-mary model [16] and, now, in the adriamycin-resistant MCF-7 human mammary adenocarcinomamodel. The latter cell line was no longer viable after2 days of exposure to 1.0 ,&ml of b&at&n. BuHa-tacin, therefore, may be useful as an adjuvant withstandard chemotherapeutic egimes. In this regard, it


6/7

78 N.H. Oberlies et al. /Cancer Letters 115 (1997) 73-79could not only assist in hindering normal cancerouscell growth as previously shown in in vivo models[14], but, more importantly, when MDR tumortypes, induced by chemotherapy, begin to evolve, bul-latacin may effectively eliminate them before theyhave a chance to become problematic, and, possibly,before they are even detected. The Annonaceous acet-ogenins, thus, present a unique approach to circum-venting MDR in that most previous studies haveexplored the use of adjuvants such as verapamil,chloroquine, progesterone, and tamoxifen [6,18-231which, in theory, only competitively block the effluxaction of the P-gp so that standard antineoplasticagents can remain within the cell long enough toinduce a toxic response. The use of bullatacin wouldattempt, instead, to eliminate MDR cell types directlyby targeting their ATP production and their elevatedrequirement for ATP.

tein action in multidrug resistance: are we there yet?, TrendsPharm. Sci, 15, 260-263.

[5] Gottesman, M.M., Ambudkar, S.V., Ni, B., Amn, J.M.,Sugimotto, Y., Cardarelli, C.O. and Pastan, I. (1994) Exploit-ing multidrug resistance to treat cancer, Cold Spring Harb.Symp. Quant. Biol., 59, 677-683.[6] Gottesman, M.M. and Pastan, I. (1993) Biochemistry of mul-tidrug resistance mediated by the multidrug transporter,Annu. Rev. Biochem., 62, 385-427.

[7] Ling , V . (1992) P-Glycoprotein and resistance to anticancerdrugs, Cancer Res., 69, 2603-2609.[8] Zeng, L., Ye, Q., Oberlies, N.H., Shi, G., Gu, Z.-M., He, K.and McLaughlin, J.L. (1996) Recent advances in Annonac-

eous acetogenins, Nat. Prod. Rep., 13, 275-306.[9] Gu, Z.-M., Zhao, G.-X., Oberlies, N.H., Zeng, L. and

McLaughlin, J.L. (1995) Annonaceous acetogenins: potentmitochondrial inhibitors with diverse applications. In: RecentAdvances in Phytochemistry, Vol. 29, pp. 249-310. Editors:J.T. Amason, R. Mata, and J.T. Romeo. Plenum Press, NewYork.

Current studies are exploring the structure-activityrelationships among additional Annonaceous aceto-genins in order to determine which regions of themolecules contribute maximally to this observedbioactivity; over 220 of these compounds have beenisolated [8]. In vivo models of parental vs. MDRtumors must now be subjected to acetogenin treatmentin order to demonstrate these proposed extrapolationsfrom this promising in vitro data.

[lo] Fang, X.-P., Rieser, M.J., Gu, Z.-M., Zhao, G.-X. andMcLaughlin, J.L. (1993) Annonaceous acetogenins: an up-dated review, Phytochem. Anal., 4, 27-67.

[ll] Rupprecht, J.K., Hui, Y.-H. and McLaughlin, J.L. (1990)Annonaceous acetogenins: a review, J. Nat. Prod., 53, 237-278.[12] Lewis, M.A., Amason, J.T., Philogene, B.J.R., Rupprecht,J.K. and McLaughlin, J.L. (1993) Inhibition of respiration

at site I by asimicin, an insecticidal acetogenin of thepawpaw, Asimina triloba (Annonaceae). Pestic. Biochem.Physiol., 45, 15-23.

[ 131 Ahammadsahib, K.I., Hollingworth, R.M., McGovren, J.P.,Hui, Y.-H. and McLaughlin, J.L. (1993) Mode of action ofbullatacin: a potent antitumor and pesticidal Annonaceousacetogenin, Life Sci., 53, 1113-l 120.Acknowledgements

This work was funded in part by grant no.CA30909 from NIIWNCI. N.H.O. acknowledgesstipend support from both the Indiana Elks CancerResearch Fund and the Purdue Research Foundation.

[14] Londershausen, M., Leicht, W., Lieb, F., Moeschler, H. andWeiss, H. (1991) Molecular mode of action of annonins,Pestic. Sci., 33, 427-438.

1151 Morre, J.D., DeCabo, R., Farley, C., Oberlies, N.H. andMcLaughlin, J.L. (1995) Mode of action of bullatacin, apotent antitumor acetogenin: inhibition of NADH oxidaseactivity of HeLa and HL-60, but not liver, plasmamembranes, Life Sci., 56, 343-348.

References1161 Oberlies, N.H., Jones, J.L., Corbett, T.H., Fotopoulos, S.S.and McLaughlin, J.L. (1995) Tumor cell growth inhibition

of Annonaceous acetogenins in an in vitro disk diffusionassay, Cancer L&t., 96, 55-62.

[l] Goldstein, L .J. (1995) Clinical reversal of drug resistance, [17] Monk, A., Scudiero, D., Skelhan, P., Shoemaker, R., Paull,Curr. Prob. Cancer, 19, 65-123. K., Vistica, D., Hose, C., Langley, J., Cronice, P., Vagio-[2] van der Heyden, S., Gheuens, E., De Bruijn, E., Van Wolff, A., Gray-Goodrich, M., Cambell, H., Mayo, J. andOosterom, A. and Maes, R. (1995) P-glycoprotein: clinical Boyd, M. (1991) Feasibility of a high&x anticancer drugsignificance and methods of analysis, Crit. Rev. Clin. Lab. screen using a diverse panel of cultured human tumor cellSci., 32, 221-264. lines, Cancer, 83, 757-766.[3] Simon, SM. and Schindler, M. (1994) Cell biologicalmechanisms of multidmg resistance in tumors, Proc. Natl.Acad. Sci. USA, 91, 3497-3504.[4] Ruetz, S. and Gros, P. (1994) A mechanism for P-glycopro-

1181 Desai, P.B., Bhardwaj, R. and Damle, B. (1995) Effect oftamoxifen on mitoxantrone cytotoxicity in drug-sensitiveand multidrug-resistant MCF-7 cells, Cancer Chemother.Pharmacol., 36, 368-372.


7/7

N.H. Oherlies et al. / Cancer Letters I15 i 1997) 73 -79 791191 Yang, C-P.H., DePinho, S .G., Greenberger, L.M., Arceci,

R.J. and Horwitz, S.B. (1989) Progesterone interacts withP-glycoprotein in multidrug-resistant cells and in the endo-metrium of gravid uterus, J. Biol. Chem., 264 , 782-788.

1201 Inaba, M. and Maruyama, E. (1988) Reversal of resistance tovincristine in P388 leukemia by various polycyclic clinicaldrugs, with special emphasis on quinacrine, Cancer Res., 48.20&L2067.

[21] Shiraishi, M., Akiyama, S.-I., Kobayashi, M. and Kuwano,M. t 1986) Lysosomotropic agents reverse multiple drug resis-tance in human cancer cells, Cancer Lett., 30, 251-259.1221 Willingham, M.C., Comwell, MM., Cardarelli, C.O.,Gottesman, MM. and Pastan, 1. (1986) S ingle cell analysisof daunomycin uptake and efflux in multidrug-resistant and -sensitive KB ce lls: effects of verapamil and other drugs,Cancer Res., 46. 5941-5946.

1231 Tsuruo, T., Iida, H.. Tsukagoshi, S. and Sakurai, Y. (1983)

Potentiation of vincristine and adriamycin effects in humanhemopoietic tumor cell lines by calcium antagonisb and cal-modulin inhibitors, Cancer Res., 43, 2267 -2272.

[24] Sikic, B.I. (1993) Modulation of multidrug resistance: at thethreshold, J. Clin. Oncol., 11, 1629-1635.

[25] McKeenhan. W.L. and Ham, R.G. (1976) Stimulation 01 clo-nal growth of normal fibroblasts with substrata coated withbasic polymers. J. Cell Biol., 7 1, 727.-734.

1261 Hall. A.M., Croy, V ., Chan, T., Ruff, D., Kuczek, T. andChang, C.-j. (1996) Bicinchoninic acid protein assay in thedetermination of adriamycin cytotoxicity modulated by theMDR glycoprotein, J. Nat. Prod.. 59, 35.-40.

1271 Gupta, K.P.. Ward, N.E., Graviu, K.R., Bergman, P.J. andOBrian, C.A. 111996)Partial reversal of multidrug resistancein human breast cancer cells by an N-myristoylated proteinkinase C-o pseudosubstrate peptide. J. Biol. Chem.. 271.2102.-211 I.

Documents

Graviola Mclaughlin