FEM,q MicrobiologyLetters 68 (1990) 249-254 Published by Elsevier

249

FEMSLE 03955

Adherence of Candida albicans to components of the subendothelial extracellular matrix S t e p h e n A. K l o t z Departments of Medicine and Ophthalmology,. VeteransAdmimstratltm and Loutstana SIt,Iv Unwersttv.~edJcalCenterL Shreveport. LA. US.A.

Received23 October 1989 Revisionreceived27 November 19i49 Accepted 8 December 1989 Key words: Candida albicans; Adhesion: Extracellular matrix

1. S U M M A R Y Candida albicans yeasts adhered avidly to extracellular matrix (ECM) proteins, type IV collagen, iamiw:n, and fibroneetin immobilized on plastic. Type IV collagen showed an increase of adherence of 400~ above control values; laminin. 300%; and fibronectin, 150%. In addition, all three (in quantities of 0.02-200 t t g / w e l l of a culture tray) bound yeasts in a dose-response fashion. Adherence was inhibited when the proteins were preincubated with specific antibody, except with type IV collagen. Soluble laminin or fibroneetin inhibited yeast adherence to the same proteins by 36 and 94~, regpeetively. Soluble fibronectin bound to the yeagt surface and in so doing inhibited subsequeW yeast adherence to fibronectin by 6670. By comparison, Candida albicans yeasts adhered in smaller numbers to glycosaminoglycans (GAGs). Keratan sulfate, hyaluronic acid. chondroitin sulfate, Type B, and heparin actually

Cor~.wtmdence to: Stephen A. Klotz, Division of lnfectw,us Diseases, Department of Medicine.VA Medtcal Center, 510 E. Stoner Avenue, Shreveport, LA 71101-4295. U.S.A.

decreased yeast adherence compared to control from 10~ to 25~.

2. I N T R O D U C T I O N Hematogenous dissemination of Candicla albicans throughout the vascular system is an initial event in the development o[ metastatic infectious sites, involving the interaction of the yeast with endothelial cells or subendothelial extracelhitdr matrix (ECM). The preference of yeasts to bind to ECM more so than t o the endothelial cell surface has been shown in previous work [I.2] and may be analogous to some tumor cell preference to adhere to ECM [3]. Type IV collagen, laminin and fibronectin occur in vascular basemenl membrane and stromal ECM but not on the surface of normat ¢ndothelium whereas glycosaminoglycans (G.~Gs) such as heparin, hcparan sulfate and chondroitin sulfate occur on the endothelial cell surface as well as in ECM [4,5|. This work demonstrates which isolated components of subendothehal ECM and the surface of endothelial cells favor yeast adherence, and conversely, which components inhibit yeast adherence.

0378-1097/90/$03.50 © 1990 Federation of European MicrobiologicalSocieties

250 3. MATERIALS AND METHODS

man fibronectin for 2 hours at 37°C, washed by centrifugation in EBSS and resuspended to desired concentrations in EBSS.

3. I. Yeasts Two clinical isolates of C. albwans were maintained by monthly passage on Sabouraud dextrose agar. These strains adhere to endothelium and subendothelial ECM as reported [1,2,6]. For adherence tests a Ioopful of yeasts was cultured in 50 ml of Sabouraud dextrose broth incubated at 26 o C for 20 h while shaking. This yields stationary phase yeast forms. Yeasts were then washed by centrifugation in phosphate buffered saline, pH 7.4 (PBS) for testing to provide data in Fig. 1 or Earie's balanced salt solution with calcium and magnesium (EBSS) and diluted to desired concentrations by hemacytometer count. In competition experiments, yeasts were added to bu.t%, containing 100 pg/ml of laminin or fib.~;,ectin and then placed in the wells. In otheT..-.^periments yeasts were preincubated with lr~.• pg/ml of hu-

3.2. Adherence components Murine laminin and type IV collagen were from Collaborative Research Inc., Bedford, MA. Human fibronectin was from Sigma (St. Louis, MO), ICN (Cleveland, OH) and Collaborative Research, Inc. Type I collagen was from Sigma and Collagen Corp., Palo Alto, CA. tteparan sulfate, heparin, chondroitin sulfates, Types A and B, keratan sulfate and hyaluronic acid were from Sigma. These ere prepared as targets for yeast adh.~,mce by immersing: (i) cellulose filters (0.45 p. mesh, Metricel, Fisher Scientific, Piano, TX) or (ii) polytetrafluoroethylene (PTFE) [Fisher] filters in solutions of ECM components and GAGs for 24 h. The filters were washed, placed in 24-well tissue culture trays (GIBCO, Grand Island, NY)

300

iio° #,c

~

~

T

-

T

1

. I '

R& l. A d ~

tI

'

I i

I

of Co.d~c.~: ,'/bwm~ ~ to vasc-,.;~ar cxtracellulax maLrix proteins arid to glycosaminoglycans. The graph represents the percent of added )crests which adhered as compared to control values. The space bar~ are standard error of the mean

and numbersin parenthesesrepresentthe numberof wellste~ted.

IO0-

a n d the adherence assay performed. This was a modification of a method used with monocytes {7]. In addition, 0.1% ( w t / v o l ) solutions o f the c o m p o nents were allowed to (a) d r y into the tray wells o, (b) remain in wells for 4 - 2 4 h, aspirated a n d washed. These were mndifications of methods used with Treponema pallidum [8 I. Controls were performed with the a p p r o p r i a t e buffers. Antibodies were diluted 1 : 5 in PBS, p H 7.4, a n d included: rabbit polyclonal anti-mouse laminin, anti-mouse collagea IV (Collaborative Research, Inc.), a n d a n t i - h a m a n fibronectin ( O r g a n o n Teknika, West Chester, PA).

A

80-

,.y

8

8O

~o, 3. 3. Adherence assay Yeasts were diluted in EBSS to 2 - 3 x 102 colo n y forming units (efu) per ml. T o each well, 0.5 ml o f yeasts were adde.d, incubated without agitation at 3 7 ° C for 30 rain, a n d then each well washed with EBSS 3X. Molten Mycosel a g a r (BBL, Cockeysville, M D ) at 5 6 ° C was overlaid in the wells, the tray incubated for 24 h a n d cfu of adherent yeasts determined. Aliquots were spread o n t o Mycosei plates to determine the n u m b e r of cfu a d d e d pet- well. Results were reported as percent of yeasts adhering o f the total n u m b e r of yeasts a d d e d per well m i n u s % of yeasts a d h e r i n g in control wells for Fig. 1. In other experiments percent adherence was recorded, determined by dividing the n u m b e r of cfu adherent in each well b y the n u m b e r a d d e d per well x 100. The method of drying a k n o w n weight of art E C M protein or G A G o n t o the plastic was chosen for the d a t a recorded in Fig. 1 (other methods were similar). Table 1

Adheccnc¢ of C. alb~cans yeasts to type I collagen m the presence or absence of divalent catmns and ~ of yeasts to gelatin and type I collage¢

Ca 2., Mg2+ abs~t Ca 2÷, Mg2+ present Gelatin Type I collagen

a

Percent adherence ( ± SEM)

23

35 : 1 6z±2 44 ± I 68±2

24

24 34

T~:x~ I collagen and gelatin were prepared at 2~0 ~g/nd and 500 pl w~replaced in each well for 24 h and the wells w ~ ~ t h EBSS prior to assa~,.

Y

zo.

c

80-1

~t

40

,/

3'

MICROGRAMS PER WELL

Fig, 2. Dosc-res;Imm¢of Camdu/a bl;, Jmg to lain=ran(AL I~e IV collagen (B) and fibrone~in (C~ ~ u m h ~ s m O~"e'nU~S,L~ represent the number o( ~¢lls tcsl~,

Thereafter. c o m p o n e n t s were added to wells, inc u b a t e d for 24 h at 4 ° C and then washed ~ t h buffer prior to the addition of yeasts. In some experiments. 200/AI of dilute a n t i b o d y were a d d e d to each well a n d incubated for 30 n u n at 26 ° C. the wells washed 3 × w i t h EBSS a n d then the adherence assay performed. E x i t s "~¢re performed a m i n i m u m of three times with ¢ag'h variable in triplicate or quadruplicate.

4. R E S U L T S 4.1. Adherence to E C M proteins The adherence properues of the E C M comlx~ nents were determined simultaneously on numer-

252 OUS occasions (Fig. l). The yeasts were suspended in PBS for these experiments in order to avoid the avid binding of yeasts to plastic which occurs in the presence of divalent cations [9]. Type IV collagen, laminin and fibroncctin gave the greatest yeast adherence, therefore the effect of different concentrations of each ECM protein was investigated. The ECM components bound yeasts in increasing numbers as the concentration of each ECM protein was increased (Fig. 2). At the maximum concentration teswed, the ECM proteins possessed about equal abilities to bind yeasts, since between 51 and 73% of added yeasts were bound to fibronectin and type IV collagen, respectively. Yeasts were also tested against type I collagen from bovine Achilles tendon. There was about a 140% increase to this sabstrate, significantly less than to type IV collagen. In order to gain a better understanding of the adherence process, Vitrogen 100, a type 1 collagen from bovine skin was immobilized in wells and yeasts suspended in buffer containing Ca 2. and Mg 2+ or in buffer without the divalent cations (Table 1). The presence of Ca 2+ and Mg 2+ ions enhanced yeast adherence to this ECM protein. Furthermore, the protein in its natiee triple-helical form (Vitrogen 100) is a pre-

Table 2 Adherence of Cand/da a/bfcan.s yeasts to ECM components pr,4ncubated with antibodies to each ~ t or ~n the presenc~ of exogenous ECM component

ferred target over denatured type 1 collagen or gelatin (Table 1). The specificity of the yeast interaction with ECM proteins was then demonstrated by inhibition of yeast adherence to these proteins when preincubated with antibodies (Table 2). The inhibition of adherence was 60% or greater with antibodies to laminin and fibronectin but not to type IV collagen. A second anti-type IV collagen antibody (ICN), likewise, did not reduce yeast adherence. The epitope recognize, by these antibodies is unknown but it clearly is not the site where Candida is adhering since there was no inhibition of adherence. Adding exogenous laminin or fibronectin reduced yeast adherence to these two proteins by 36 and 94%, respectively. Because fibronectin has been of such interest in mit,robial-host interactions, the ability of fibronectin to inhibit yeast adherence to laminin was tested. When laminin was placed in the wells and yeasts added with 100 l a g / m l fibronectin, the percent adherence was 25 5= 4%, an inhibition of 68~. Neither of the collagens were tested because they are not soluble in EBSS at pH 7.3. When yeasts were preincubated with fibronectin ( 1 0 0 / t g / r n l ) for 2 h, washed with buffer by centrifugation and resuspended in buffer and compared with yeasts incubated in buffer alor, e for their ability to adhere to type I collagen, yeasts preincubated in buffer had 65 5= 7 per cent adherence and yeasts preincubated in buffer plus fibronectin had 22 5= 1 per cent adherence, a reduction of 66~. This demonstrates the ability of plasma fibronectin to bind to the yeast surface.

Per cent adherence ( 4-SEM) of yeasts to:

Yeasts in EBSS Wellspmncubated with antibody Yeasts m EBSS+ laminin •

Type IV

Laminin Fibronectin

79:z3

79±5

82±3

83+3

27±2

33+4

53 ±2

-

Yeasts in EBSS +

fibfonecfn *

54-1

• Lammin and fibronectin were prepax~ at 100/~g/ml conomtratioa. ECM ~ we/e p~a~.~,-.: ~%~_.--I and 200 itl placed in each well for 24 h and then t l ~ wells ,a,ashed with FJSS prior to assay.

4.2. Adherence to GAGs The G A G s did not bind C albicans yeasts (Fig. 1) since heparan sulfate, keratan sulfate and chondroitin sulfate type A allowed yeasts to adhere in numbers equal to control wel!s. Fewer yeasts adhered to keratan sulfate, hy~uronic acid, chondroith; ~t,lfate type B and hcpa.,'L-, th~n to control values. When larger inocula of yeasts were placed into control wells and compared to beparin or hyaluronic acid-coated wells the latter two G A G s consistently had about 25% of the yeast adherence of the plastic/:6fiir"6"i iv"efis.

253 5. D I S C U S S I O N

5.1. Potential role of ECM in disseminated candidiasis Endothelial cells cover the vast majority of the surface area of the vascular tree. Basement memb r a n e or E C M is only ~xposed in a few a n a t o m i c sites such as the glomerulus. In illness, d a m a g e to endothelium may, however, expose E C M [10!. Therefore, both the endothelial cell surface a n d subendothelial E C M m u s t be looked u p o n as potential targets for yeast adherence in the event of hematogenous dissemination of the yeast. Previous studies have demonstrated the preference of yeasts to adhere to E C M rather t h a n to endothelial cells [1,2] a n d the present s t u d i ~ provide some insight as to why this occurs. Yeasts b i n d in significantly greater n u m b e r s to type IV collagen, laminin, a n d fibronectin b u t are not avidly b o u n d b y the G A G s . Moreover, in the case of h e p a r m a n d hyaluronic acid, yeasts are actually inhibited from binding to plastic. T y p e IV collagen, laminin, a n d fibronectin are found only in the E C M a n d not on the endothelial cell surface a n d laminin n~ay actually be exposed to the vascular lumen in the kidney [11], o n e of the target organs of disseminated candidiasis [12]. Studies b y others show binding of C. albicans to immobilized fibrcnectin [13], and to soluble fibronectin [14]. This report is the first d e m o n s t r a t i n g the avid binding of yeasts to type IV collaT_,en a n d laminin, as well. 5. 2. Conclusion These results d e m o n s t r a t e that (1) Candida albicans yeasts b i n d avidly to the E C M proteins, type IV collagen, laminin a n d fibronectin; (2) that binding of yeasts to E C M proteins is proportional to the an,ount of each E C M c o m p o n e n t present; (3) that adherence of yeasts to E C M proteins has

some cross-specificity, with fibronectin, for example, blocking adherence to !aminin; (4) that soluble fibronectin binds to the surface of yeasts; a n d (5) that the glycosaminoglycans, by contrast to the E C M proteins, have a m u c h lower affinity for interaction with yeasts.

ACKNOWLEDGMENTS The a u t h o r thanks A n n Shows for help with the manuscript. This work was supported by the Veterans Affairs Research Service.

REFERENCES [I] Kiotz. S.A. (1988) FEMS ~i-'robio! I.¢It.4~. 201-205. [2] Klotz, S.A. and Mm:a. R.D. (1988) Infect. lmmun. 561. 2495-2498.

13] Kran~r, R.H., Gcmz,dez. R. and Nicolson. G.L (1980) Int. J. Cancer 26 639-045. [4] Hay, E.D. (1981) J. Cell Biol. 91. 205s-223s. [51 Hook, M., Kjellan, L., Johansson. S. and Rc.binson. J. 0984) Ann_ gev. Bi¢~.tem. 53. 847-869, [6] ~ SA., Drutz. DJ.. H ~ , J.L. and H ~ M. (1983) Infect. Immun. 42. 374-384. [7] Tobias, J.W., lkrn, M.M.. Netland. P.A. and Zett~, B.R_ 0987) Blood 69. 1265-1268. [8] Fitzgerald. TJ., Repesh. L.A_. Blanco. D.R. and Miller. J.N (1984l Br. J. Vener. Dis. 60, 357-363. [9] KJotz. S.A., Drut~ D.J. and Zajic. J.l= (1985) Infect. Immtm. 50. 9/-101. [10] Gaynor. E. Bouv~er.C. and Spae~ T. (1970) Scianc¢ 170. 986. [ll~ Madta. J.A.. Roll. F.J. Uurthrn.ayr.H and Fo4dart. J.M (1980) J. Cell Biol. ~.n'.682-687. ]12] Odds. F.C. (1988) Candtda and Candidosis. 2nd Edition. pp. 220-221. Bailliere Tmdall. Londcm. [13l Calderone. R.A. and Scheld. W.M. O987t Rev. Inf. D-,s. e~S), 4~-403. [14] Kalo. A., SegaL E., Sahar, E. ~ Dayan. D. (1988) J. lnt'. Dis. 157, 1253-1256.