Candida glabrata (CG) is an opportunistic fungal pathogen that initiates infection by binding to host cells via specific lectin-like adhesin proteins. We have previously shown the importance of lectin-oligosaccharide binding in cardiac responses to flow and agonists. Because of the lectinic-oligosaccharide nature of CG binding, we tested the ability of CG to alter the agonist- and flow-induced changes in cardiac function in isolated perfused guinea pig hearts. Both transmission and scanning electron microscopy showed strong attachment of CG to the coronary endothelium, even after extensive washing. CG shifted the coronary flow vs. auricular-ventricular (AV) delay relationship upward, indicating that greater flow was required to achieve the same AV delay. This effect was completely reversed with mannose, partially reversed with galactose and N-acetylgalactosamine, but hyaluronan had no effect. Western blot analysis was used to determine binding of CG to isolated coronary endothelial luminal membrane (CELM) receptors, and the results indicate that flow-sensitive CELM receptors, ANG II type I, α-adrenergic 1A receptor, endothelin-2, and VCAM-1 bind to CG. In addition, CG inhibited agonist-induced effects of bradykinin, angiotensin, and phenylephrine on AV delay, coronary perfusion pressure, and left ventricular pressure. Mannose reversed the inhibitory effects of CG on the agonist responses. These results suggest that CG directly binds to flow-sensitive CELM receptors via lectinic-oligosaccharide interactions with mannose and disrupts the lectin-oligosaccharide binding necessary for flow-induced cardiac responses.
- cell-cell interaction
- GPCR receptors
- flow sensors
candida species are opportunistic fungal pathogens that are among the most common causes of nosocomial infections worldwide with a mortality rate for Candida infections that can reach 40–50% (26, 28, 30). Candida glabrata (CG) has emerged as the second most prevalent infectious Candida species with increasing frequency as a cause of candidemia, especially in immunocompromised patients (10, 18, 28). Disseminated candidiasis leads to fungal replication in many organ systems, including the heart, brain, and kidney, but the heart suffers the most damage following Candida infection (1, 20, 23, 22). For these reasons, continued investigation into the pathophysiological characteristics of Candida infection and identification of novel mechanisms of Candida actions in the heart are very important.
Dissemination of Candida into the bloodstream and tissue invasion requires that the pathogen adhere to the blood vessel endothelium before entering the tissues. Candida albicans yeast forms adhere more efficiently under conditions of flow compared with static conditions (16). Similarly, CG initiates infection by binding to its host cell via specific host-pathogen interactions and can form biofilms (36). The cell wall of CG is composed of an outer layer containing a high concentration of mannoproteins with nearly 44% of the cell wall composed of mannose (9). This outermost layer is the first point of contact between CG and the endothelium.
CG contains a large number of glycosylphosphatidylinositol adhesin-like proteins, covalently bound to the wall via 1,6 and 1,3-β-glucan linkages. These lectin-like adhesin proteins, particularly those in the epithelial adhesin (EPA) family, are critical for adhesion to host cells (5, 9, 12, 43). In CG sir3Δ mutants, the transcription of two normally silent EPA genes leads to greater adherence of this strain to host cells (25). Additional proteins, such as Aed1p (adherence to endothelial cells), have also been proposed in the binding of CG to the endothelium (8).
Our group has previously shown that coronary blood flow stimulates heart glucose uptake, ventricular contraction, ventricular spontaneous rhythm, coronary vascular tone, and auricular-ventricular transmission (2, 27, 29, 32, 33) and indicated these flow-induced changes in cardiac function are the result of lectin-oligosaccharide binding in the luminal endothelial membrane (2, 17, 27, 29, 32, 33). For example, the role of luminal endothelial oligosaccharides can be shown by the fact that when intracoronary infusions of lectins (32, 33) or oligosaccharide-hydrolyzing enzymes are given (32, 33), they alter coronary flow-induced cardiac responses. On the other hand, the role of luminal endothelial lectins is shown when polymers of glucose, mannose, or N-acetylglucosamine bind irreversibly to coronary endothelial membrane lectins and irreversibly inhibit coronary flow-induced cardiac responses (2, 27, 29). Thus, agents that bind to oligosaccharide or to lectinic structures on the coronary endothelium have profound effects on flow-induced changes in cardiac function.
Candida has been shown to quickly bind to the endothelium under conditions of flow and can cause serious conduction disturbances in the heart (11, 13, 15, 40). Fairchild et al. (11) demonstrated that Candida administration induced sinus bradycardia and auricular-ventricular (AV) block in all mice injected intraperitoneally with a range of 107-108 CFU/mouse. The bradycardia began within 1–3 min of intraperitoneal injection and resolved spontaneously within 30–90 min. The authors concluded that the mechanism associated with the bradyarrhythmias following pathogen injection was rapid activation of vagal signaling. On the basis of our results presented here, we propose that the AV delay is a direct result of Candida binding to the coronary endothelial membrane.
Because of the lectinic nature, mannose-rich outer cell wall, and cardiotropic effects of CG (8, 9 ,11, 12, 31, 43), we investigated the effects of intracoronary administration of CGM35 sir3Δ (CG3), a mutant form of CG that expresses more adhesin-like proteins (5) in isolated perfused hearts. We found that CG3 binds to the coronary luminal membrane using electron microscopy and, as a result of a mannose-dependent lectinic binding, inhibits the flow-dependent shortening of AV conduction. In addition, to define potential molecular mechanisms for this effect of CG3, we identified flow-sensitive coronary luminal membrane receptors that, upon binding to CG3, alter their function.
The Committee for the Use of Animal Experimentation at the Universidad Autonoma de San Luis Potosi approved the animal experimental procedures.
Candida glabrata Strain and Culture
Candida glabrata strain CGM35 sir3Δ (CG3) was chosen because it expresses two EPA genes, leading to increased expression of the adhesin-like proteins, which have lectin-like properties and high adherence to cells (5). The strain was cultured on Saburaud dextrose agar slants at 25°C. Cells were subcultured in Sabouraud dextrose broth for 18 h at 37°C in an orbital shaker incubator. After 48 h, the culture was centrifuged two times (3,500 rpm) for 5 min, washed with water, and suspended in Krebs-Henseleit (K-H) buffer to a final concentration of 3.7 × 103 cells/ml. The experiments in isolated perfused hearts were performed no more than 6 h following isolation and suspension of the cultures (36).
Isolated Perfused Guinea Pig Hearts
Dunkin Hartley guinea pigs (300–400 g body wt) were anesthetized by intraperitoneal administration of pentobarbital sodium (50 mg/kg) followed by sodium heparin (500 IU). Under artificial respiration, the thorax was opened; the heart was excised and retrogradely perfused at a constant flow of 8 ml/min, according to the Langendorff method with oxygenated K-H (in mM): 127 NaCl, 6 KCl, 1.8 CaCl2, 1.2 NaH2PO4, 1.2 MgSO4, 25 NaHCO3, 5 dextrose, 2 pyruvate, pH 7.4, 37°C, 95% O2, and 5% CO2. Heart rate was kept constant at 4.5 Hz by application of electric square pulses using a pair of electrodes placed in the right atria. There was a 30-min stabilization period before any experimental treatment (2, 29, 32, 33).
A-V Delay Measurements
Dromotropic effects (changes in A-V delay) were determined. A-V delay (in ms) was recorded via two electrodes connected to an oscilloscope. One electrode was placed in the left atrium, and a second electrode was placed on the apex of the left ventricle to record their corresponding action potentials. We have established that the time interval between these two electrical potential recordings (A-V delay) measures the electrical conduction time (ms) through the A-V node (33) and 1/A-V delay reflects the AV node conduction velocity. Heart rate was kept constant at 4.5 Hz by application of electric square pulses using a pair of electrodes placed in the right atria. After 30 min of stabilization, A-V-delay was measured at different coronary flows (4, 5, 6, 8 and 10 ml/min), allowing 5 min of stabilization at each step before measurement. For each heart, A-V delay was measured directly in milliseconds, and changes were expressed as means ± SE (n = 6 per group) (2, 29, 32, 33).
Coronary Vascular Resistance Measurements
Vascular resistance was determined from coronary perfusion pressure (CPP) measurements. CPP was recorded via a pressure transducer connected to a side branch of the perfusion cannula. The CPP at the end of the stabilizing period, at a coronary flow of 8 ml/min, was taken as control and was 37 ± 5 mmHg.
Ventricular Contraction Measurements
Inotropic effects were determined from changes in left ventricular pressure (LVP) measurements. For that purpose, a small hand-made latex balloon attached to the end of a fluid-filled catheter was introduced into the left ventricle via the left atrium. The other end of the catheter was connected to a pressure gauge, and 0.2 ml of water was injected with a syringe into the balloon to create a diastolic pressure and, upon ventricular contraction, pressure developed, and its amplitude was continuously recorded. All changes in amplitude of developed left ventricular pressure were taken as a qualitative indicator of inotropic response.
A-V delay Effects of Coronary Flow After Candida Glabrata Infusion
To study the effects of CG3, its maximal effective concentration was first determined from dose-response curves performed in six hearts and was found to be 3.7 × 103 cells/ml [administration of a higher concentration of CG3 completely blocks A-V delay (not shown); administration of wild-type Candida glabrata did not affect A-V delay (not shown)]. Control A-V delay vs. coronary flow and ventricular contraction vs. coronary flow (4, 5, 6, 8, and 10 ml/min) relationships were determined for each heart. Following the control measurements, CG3 (3.7 × 103 cells/ml) was infused for 10 min followed by a 20-min wash period with K-H to remove unbound CG3 from the coronary endothelial luminal membrane (CELM). At this point a second A-V delay-coronary flow curve was determined. Thereafter, either mannose, galactose, N-acetylglucosamine (30 mM), or hyaluronic acid (10 μg/ml) was infused for 5 min, followed by a 5-min washout with K-H, and a third A-V delay-coronary flow curve was determined (n = 6).
Dromotropic, Inotropic, and Vascular Responses Induced by Intracoronary Administration of Bradykinin, Phenylephrine, or ANG II Are Depressed by Candida glabrata
In isolated guinea pig hearts perfused at a constant coronary flow of 8 ml/min, an intracoronary bolus infusion (0.5 s duration) of bradykinin, ANG II, or phenylephrine (10−6 M) was given, and control agonist responses were measured. Five minutes after these control responses, CG3 (3.7 × 103 cells/ml) or K-H (control) was perfused for 5 min followed by a 20-min washout to remove unbound CG3. A second intracoronary bolus infusion of (10−6 M) bradykinin, ANG II, or phenylephrine was repeated. Thereafter, a sustained 5-min perfusion of 30 mM mannose was given, followed by a 10-min washout period. A third bolus infusion of bradykinin, ANG II, or phenylephrine was administered (10−6 M). In all cases, the agonist bolus infusion caused a transient peak change, whose amplitude defined the response: ΔA-V, ΔCPP, and ΔLVP. To compare the responses following CG3 and mannose treatments (experiments; e) with the corresponding control response (c), the ratios between the corresponding responses were taken and multiplied by 100. The control response was 100%. These ratios were as follows: ΔA-Vc/ΔA-Ve X100, ΔCPPe/ΔCPPc X100, and ΔLVPe/ΔLVPc X100. In the case of A-V, the ratios are inverse to those of CPP and LVP, because the measured parameter reciprocally reflects, not directly, the variable measured (velocity of propagation in the AV node). In short, the amplitude of responses is expressed as % of the corresponding control agonist response.
Isolation of Coronary Endothelial Luminal Membrane Proteins
Cationic silica perfusion.
In the isolated perfused guinea pig heart preparation described above, colloidal silica was intracoronarily infused, as previously described (2, 4, 7, 27, 29). In brief, hearts were retrogradely perfused at a constant pressure of 70 cm of water with 50 ml MBS [MES; 2-(N-morpholino) ethanesulfonic sodium salt, 19.6 mM, NaCl 150 mM 10 M EGTA, pH 6], followed by successive perfusion with 10 ml MBS pH 5. Thereafter, 24 ml of 1.5% cationic colloidal silica in MBS pH 5 was perfused. Unbound silica was washed using 20 ml of MBS pH 5. CELM-bound silica was polymerized by perfusion of 40 ml of anionic polyacrylic acid (3 mg/ml in MBS at pH 6), and the excess was washed with 20 ml of MBS, pH 5. Subsequently, 20 ml of lysis buffer (LB, 25 mM HEPES; N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid, sucrose 250 mM, pH 7.4) was perfused. From these hearts CELM proteins were isolated (n = 6). Each heart was processed separately, and the CELM proteins of all six hearts were pooled.
Isolation of silica-bounded CELM.
Each ventricle was dissected, minced, and suspended in 10 ml of ice-cold LB containing 1 mM of PMSF and homogenized with a Glass-Teflon homogenizer, at 6,000 rpm, 5 min, at 4°C. The homogenate was mixed with an equal volume of Nycodenz (Histodenz; Sigma, D2158) solution I (228 mM in LB) layered on top of 0.5 ml of Nycodenz solution II (290 mM in LB), and centrifuged for 20 min at 20,000 g. The supernatant was discarded, and the enriched CELM fraction was obtained from the pellet, as described below. The pellets from the six hearts were pooled.
ISOLATION OF ENRICHED CELM.
The pooled pellet was washed three times with MBS, pH 6. The resulting CELM-silica pellicle fraction was suspended in 2% SDS, sonicated, boiled for 5 min, and centrifuged for 10 min at 14,000 g. The supernatant contains CELM proteins that were quantified using the bicinchoninic acid kit (Sigma B9643) (2, 27, 29).
CELM samples were briefly sonicated and heated for 5 min before loading the gel (n = 6). For all CELM samples, 20 μg of protein was loaded per well on 12% bis-acrylamide mini gels (1.5 mm). CELM proteins were visualized using BioSafe Coomassie stain (Bio-Rad 161-0787). Preparative gels were run, and proteins were transferred to polyvinylidene difluoride nylon membranes (GE Health Care, PVDF 0.45 μm, RPN303F) for Western blot analysis.
Western blot analysis.
Membranes were washed in TBS (0.05 M, pH 7.6) and blocked for 1 h with 10% nonfat powdered milk in TBS (0.05 M, pH 7.6). Membranes were incubated overnight at 4°C with a 1:5,000 dilution of primary IgG polyclonal rabbit or monoclonal mouse specific antibody to G protein-coupled receptors (GPCRs). ANG II type 1 receptor (AT1R), endothelin receptor type 2 (ETR), VCAM-1, and α adrenergic 1A receptor (AR) were used as loading controls (SC-1173, SC-33535, SC-8304, SC-28982; Santa Cruz Biotechnology, Santa Cruz, CA, respectively). After 24 h, membranes were washed twice at room temperature with TBS (0.05 M, pH 7.6) and TBS with 0.1% Tween 20 (0.05 M, pH 7.6), for 5 min. Thereafter, membranes were blocked for 1 h, incubated for 1 h with secondary antibody (1:5,000 dilution of anti-rabbit or anti-mouse; horseradish peroxidase-conjugated), and washed. Western blots were developed by ECL (enhance chemiluminescence reagent, SuperSignal West Pico, Pierce 34080) and were visualized on Kodak photographic paper. Band densities were quantified using the National Institutes of Health program ImageJ. A mean value was calculated from the four-receptor density measurements. Mean density for each sample was expressed as a percentage of the control value (control group).
Identification of GPCR of CELM that bind to Candida glabrata.
Four milliliters of CELM proteins (1 mg/ml) were mixed with 5 ml of buffered medium (K-H), and 5 ml of CG3 (3.7 × 103 cells/ml) were added. The mixture was incubated at room temperature under continuous shaking for 10 min. CG3 was used like an affinity chromatography resin to bind CELM molecules, with GPCRs present in the CELM sample. The solution was centrifuged (5,000 rpm, 4°C, 10 min) to sediment CG3 with bound proteins. The supernatant was collected and identified as the nonbound fraction, or NBF. The CG3 pellet was suspended in 5 ml of K-H containing mannose, galactose, and N-acetylglucosamine (each at 30 mM) and gently agitated (5 min, 4°C). The suspension was centrifuged, and the supernatant was identified as the bound fraction (BF). The NBF and BF were concentrated, and Western blots were performed as described above.
Transmission Electron Microscopy and Scanning Electron Microscopy of isolated hearts perfused with Candida glabrata
Isolated guinea pig hearts (n = 3) were infused with CG3 (3.7 × 103 cells/ml) for a 5-min period followed by a 20-min washout to remove unbound CG3. Thereafter, the tissue was fixed by a 5-min intracoronary perfusion of a solution of 2.5% glutaraldehyde in phosphate buffer 100 mM, pH 7.4. Tissue was cut in 5- to 6-mm-thick blocks, immersed in glutaraldehyde solution, and stored at 4°C until ready to process for electron microscopy.
For transmission electron microscopy (TEM) heart tissue blocks of ∼2 mm2 were fixed in 3% glutaraldehyde in phosphate buffer pH 7.4 for 1 h at room temperature. Fixed tissues were washed with buffer and postfixed for 1 h in 1% OsO4. Tissues were washed in buffer and dehydrated in a graded ethanol series at 4°C and embedded in ethanol:resin LR-White 1:1 for 1 h, and overnight with resin at 100%. The tissues were included inside LR-White gelatin capsules and polymerized with UV light for 24 h. Ultra-thin sections (60–90 nm) were cut with a Sorvall MT2 ultramicrotome, mounted on 200-mesh copper grids, stained with 2% lead citrate and 2% uranyl acetate, and photographed with a Jeol JEM-200CX TEM.
For scanning electron microscopy (SEM), the heart tissues were fixed 2 h in 3% glutaraldehyde in phosphate buffer. Samples were rinsed three times with cold buffer for 15 min each. Afterward, samples were postfixed with 1% OsO4 in phosphate buffer for 1 h, followed by three rinses with buffer, 15 min each, and then the tissues were dehydrated in graded ethanol. Critical point drying was done in a Tousimis Samdri-PVT-3D, samples were mounted and gold sputter-coated using a Cressington model 108 auto and examined in a FEI model Quanta 200 SEM.
The free statistical program R 2.13.1 was used; an ANOVA and Dunnett test were utilized to compare the experimental results of flow-induced responses. To compare the results of the effects of CG3 on drug administration and its reversal by sugar treatment, we used the Student's t-test. Values are expressed as means ± SE, and each heart served as its own control. A value of P ≤ 0.05 was considered to be statistically significant.
Candida glabrata Binds to the Coronary Endothelium
TEM and SEM of guinea pig hearts that were perfused at a flow rate of 8 ml/min with CG3 showed that even after extensive washing, the yeast remains bound to the coronary endothelial cell membrane (Fig. 1, A and B). In the transmission electron micrograph in Fig. 1A, the membranes of the two cell types actually are separated by a small space of about 0.1- to 0.2-μm thickness, which likely results from the intertwining of the cell wall of CG3 with the glycocalyx of the endothelial cell. This is suggested by the faint fibrous electron opaque structures seen between the yeast surface and the endothelial surface (Fig. 1A, thick arrow). A three-dimensional scanning electron micrograph of a capillary is shown in Fig. 1B. In this picture, several yeast cells are clearly shown bound to the luminal wall of the blood vessel. Inclusively, the yeast cell located in the front plane of the picture, at the right lower corner, is engulfed by a matrix, possibly the luminal coronary endothelial glycocalyx. These results are similar to those found recently using Candida parapsilosis and cultured human endothelial cell monolayers under no-flow conditions (35).
Candida glabrata Inhibits the Positive Dromotropic Effect of Coronary Flow
As previously established (2, 29, 32, 33), increasing coronary flow reduces A-V delay (Fig. 2, A–D). After infusion of CG3, the relationship between flow and AV delay was shifted up and to the right, indicating an inhibition of the positive dromotropic effect of flow (Fig. 2, A–D). The addition of mannose (MAN; Fig. 2A), galactose (GAL; Fig. 2B), N-acetylglucosamine (NAG; Fig. 2C), and hyaluronan (HYAL; 2D) reversed the effects of CG3 in the following decreasing sequence MAN > GAL > NAG >>> HYAL. MAN completely eliminated the effects of CG3, and HYAL had no effect.
These results are similar to our previous work (33) in which intracoronary infusion of various plant lectins with different affinities to NAG, MAN, and GAL, upon binding to the endothelial lumen as does CG3, depressed the positive dromotropic effects of coronary flow. These lectins were Lycopersicron esculentum (affinity to β1,4NAG), Lens culinaris (affinity to αMAN), concanavalin A (affinity to αMAN), Arachis hypogea (affinity to GALβ1GalNac), Griffonia simplificolia (affinity to αGalNac), and Ricinus communis (affinity to βGal). The orders of inhibitory potency of these lectins on flow-induced positive dromotropism were Lycopersicron esculentum > Lens culinaris ≈ concanavalin A ≥ Arachis hypogea > Griffonia simplificolia >>> Ricinus communis. The order of sugar affinities of these lectins is NAG > MAN > GAL and indicates that these three sugars participate in flow sensing. The discrepancy in the results using plant lectins vs. the present results using CG3 will be explained in the discussion.
Flow-Sensitive CELM Membrane Receptors Bind to Candida glabrata
CG3 was used as an affinity resin in suspension to which CELM proteins were added and incubated. Following centrifugation, a supernatant fraction (not bound proteins) and CG3 pellet fractions (bound proteins) were obtained. The CG3 pellet was eluted with a solution of three monosaccharides (mannose, galactose, and N-acetylglucosamine) to release bound proteins. In the control CELM, bound and unbound fractions form the proteins; ANG II receptor (AT1R), α-adrenergic receptor (α1R), endothelin receptor (ET2R), and VCAM-1 were visualized and quantified by Western blot analysis and ImageJ software (Fig. 3). With the total amount of each of these proteins in the CELM fraction (control) defined as 100%, more than 65% binds to CG3 (bound fraction); i.e., there may be two fractions; one recognized by CG3 and another that is not. AT1R, α1R, and VCAM-1 are activated by flow (6, 21, 27, 33, 38, 39, 41, 42). It is unknown whether ET2R is also flow stimulated (27). These results support previous studies that show GPCRs in the coronary endothelial membrane are glycosylated and lectinic (27, 34); for these reasons, GPCRs are flow sensitive and are recognized by CG3.
Mannose Reverses the Action of Candida glabrata on GPCR-Mediated Cardiac Effects
A certain and simple way to functionally identify a GPCR is to measure the response induced by its activation with a specific agonist. To detect the effect of CG3 on agonist-activation of GPCRs, we chose to measure responses to bradykinin, ANG II, and phenylephrine, which are also flow-sensitive (21, 27, 38, 42). Each of these three GPCRs was activated with a submaximal dose (95% max) of the corresponding specific agonist, and the effects were determined on three different functions: A-V delay, coronary vascular responses, and inotropic effects (Fig. 4). Bradykinin, ANG II, and phenylephrine caused in the control group a positive dromotropic effect (1/-ΔA-Vc) response. ANG II and phenylephrine caused a control +ΔCPP and +ΔLVP responses, while for bradykinin, these control responses were negative. After CG3 infusion, the magnitude of these responses diminished compared with their corresponding control responses (Fig. 4, A–C), but, in the case of bradykinin, the diminutions of −ΔCPP and a −ΔLVP to bradykinin show only a trend. For all agonists, the blockade by CG3 on the dromotropic, vascular, and inotropic effects was restored following infusion of mannose (Fig. 4, A–C).
Pathogenic dissemination of Candida into the bloodstream and invasion of tissues requires adhesion to the blood vessel endothelial luminal wall. Here, we show that Candida glabrata sir3Δ (CG3) mutant adheres to the coronary endothelium and alters the relationship between flow and cardiac function, including A-V delay, coronary perfusion pressure, and ventricular contraction (Fig. 5). CG3 binds coronary luminal membrane flow-dependent protein receptors, including endothelin, ANG II, α-adrenergic receptor, and VCAM. Furthermore, agonist-mediated changes in cardiac function can be altered by infusion with CG3. Both flow and agonist-mediated effects can be completely reversed by MAN, partially by GAL, poorly or none by NAG and HYAL, suggesting that CG3 binding is mediated possibly by reciprocal lectinic-oligosaccharide interactions involving mannose and galactose (5, 25, 43).
Candida glabrata strain CGM35 sir3Δ (CG3) was chosen for this study because adaptation of wild-type Candida glabrata to in vivo conditions takes time, and this mutant strain, which overexpresses EPA genes, more closely represents mutated C. glabrata found in urinary tract and bloodstream infections. Overexpression of EPA genes leads to increased expression of the adhesin-like proteins, which are highly glycosylated mannoproteins, and results in high adherence to cells. Three adhesins, Epa1, Epa6, and Epa7, all require a terminal galactose on the carbohydrate ligand for host recognition (5, 43). Similarly, the endothelial luminal membrane also contains a large number of lectins (2, 27, 29), and most proteins are glycosylated, containing mannose and galactose (32, 33, 34).
Role of Endothelial Lectins on Parenchymal Function-Coronary Flow Responses
Our laboratory has previously shown that two components of the coronary endothelial glycocalyx—oligosaccharides and lectins—are involved in the endothelial signal transduction mechanisms linking flow-mediated and agonist-induced cardiac responses (2, 17, 27, 29, 32, 33). Specifically, endothelial lectinic proteins interact reversibly with oligosaccharides, and these interactions, which are flow regulated, lead to alterations in cardiac excitability, vasoconstriction, A-V node velocity of propagation, contractility, and metabolism. The endothelial glycocalyx contains transmembrane lectins with affinity to mannose, galactose, and N-acetylglucosamine, and exogenous polymers of these sugars (NAG-Pol, MAN-Pol, and Gal-Pol) differentially inhibited flow-induced A-V node effects and removal of endogenous hyaluronidate was without effect. These results allowed us to define the following inhibitory potency sequence NAG-Pol > Man-Pol >> Gal-Pol >>>> HYAL (2, 29). The relevance of these observations for the present study is discussed below.
Role of Candida glabrata Lectins on Parenchymal Function-Coronary Flow Responses
On the one hand, the CG3 mutant behaves as a living mannose and galactose polymer, binds to flow-dependent lectinic endothelial membrane proteins, and inhibits flow and agonist-dependent cardiac effects. As far as flow-mediated effects, we show that addition of MAN (Fig. 2A), GAL (Fig. 2B), NAG (Fig. 2C), and HYAL (Fig. 2D) reversed the effects of CG3 in the following sequence MAN > GAL >> NAG >>>> HYAL. This sequence is close to that using exogenous polymers NAG-Pol > Man-Pol >> Gal-Pol >>>> HYAL (2, 29), except that NAG-Pol in this sequence is in the first and not in the third place, as in the sequence for CG3 (MAN > GAL >> NAG >>>> HYAL). This difference could imply that NAG and hyaluronidate do not reverse the effects of CG3 because the yeast surface does not contain NAG and hyaluronidate groups that are functionally involved in the endothelial binding, it contains only functional MAN and GAL structures.
Role of Endothelial or Candida glabrata Oligosaccharides on Parenchymal Function-Coronary Flow Responses
On the other hand, also in this study, CG3 behaves as living mannose- and galactose-selective lectin. The present results are similar to our previous work (33), in which intracoronary infusion of various plant lectins with different affinities to NAG; Lycopersicron esculentum (affinity to β1,4 NAG), MAN; Lens culinaris and concanavalin A (affinities to αMAN), and GAL; Arachis hypogea, Griffonia simplificolia, and Ricinus comunis (respective affinities for GALβ1GalNac, αGalNac, and βGal) upon binding to the endothelial lumen, depressed the positive dromotropic effects of coronary flow, with the sequence: Lycopersicron esculentum > Lens culinaris ≈ concanavalin A ≥ Arachis hypogea > Griffonia simplificolia >>> Ricinus communis. Enzymatic removal of endothelial glycocalyx hyaluronidate had no effect. This order of potency of these lectins is a reflection of their order of sugar affinities, which are NAG > MAN > GAL >>>> Hyal, which indicates that these three sugars participate in flow sensing by the A-V node, but hyaluronidate does not. The fact that NAG and hyaluronidate do not reverse the effects of CG3 implies that the yeast does not bind to functional NAG-containing structures, only to those with MAN and GAL, which are the ones more effective at reversing CG3 effects. Whether CG3 binds to hyaluronidate cannot be asserted because our response-indicator (A-V node transmission) is not hyaluronidate dependent (32, 33).
In summary, the luminal endothelial membrane is a source of oligosaccharide binding sites for GG3 adhesins (5, 32, 33, 34, 43) and CG3, in turn, is a source of oligosaccharide binding for endothelial lectins (5). Thus, the reciprocal binding between endothelial and GG3 extracellular matrices has functional implications that depend on mannose > galactose.
An increasing number of luminal endothelial GPCRs are activated by flow (6, 21, 27, 29, 38, 41, 42), which enhances the agonist stimulation (22). We have provided evidence that this synergism between flow and agonist results from the lectinic properties of GPCRs (27), which are also glycosylated (34). These two properties suggest that GPCRs could bind to CG3. Binding of GPCRs from the luminal endothelial protein fraction was demonstrated using CG3 as an affinity resin and elution with mannose and galactose freed the bound GPCRs. The binding of CG3 to GPCRs (agonists; phenylephrine, ANG II, and bradykinin) in situ is indicated by inhibition of agonist-induced responses following yeast infusion and reversal of this inhibition by infusion of a high concentration of mannose. This implies that after the endothelium is invaded by CG3, it is pharmacologically impaired, and mannose cannot correct this impairment.
Candida glabrata was once considered a relatively nonpathogenic saprophyte of the normal flora of healthy individuals, but in recent studies, it has emerged as an important nosocomial pathogen. The reduced susceptibility of Candida glabrata to azoles and amphotericin B poses a challenge in choosing the most appropriate antifungal therapy for patients with candidemia prior to the availability of species identification or antifungal susceptibility test results (18, 37). It is possible that mannose or mannose polymers could be used to prevent binding between Candida and host membranes in the endothelium, decrease drug resistance associated with Candida glabrata in biofilm, and decrease the cardiac effects associated with Candida glabrata-endothelium binding, leading to a novel therapy for preventing the deleterious effects of Candida glabrata septicemia.
Future directions of this work will be to test the mechanisms of how Candida glabrata interacts with the coronary endothelium to induce a decrease in coronary blood flow, perhaps through eNOS phosphorylation, NO production, and downstream signaling events. The isolated heart model used in this study did not allow us to address these questions, but future studies using endothelial cell cultures and isolated vessels will allow us to explore the mechanism of these effects at the cellular level.
This study was funded by Grants PROMEP DSA/103.5/14/11016, CONACYT-SEP No. 101850, CONACYT-SEP No. 101850, and CONACYT CB 2014 No. 239629.
No conflicts of interest, financial or otherwise, are declared by the authors.
Author contributions: D.T.-T. and A.P.-S. performed experiments; D.T.-T., A.P.-S., and R.R. analyzed data; D.T.-T., A.D.l.P., and R.R. interpreted results of experiments; D.T.-T. and R.R. prepared figures; D.T.-T., M.K., and R.R. drafted manuscript; D.T.-T., M.K., and R.R. edited and revised manuscript; M.K., I.C., A.D.l.P., and R.R. conception and design of research; M.K., I.C., A.P.-S., A.D.l.P., and R.R. approved final version of manuscript.
- Copyright © 2016 the American Physiological Society