Calcium-Mediated Biophysical Binding of Cryptosporidium parvum Oocysts to Surfaces Is Sensitive to Oocyst Age

The mechanisms by which pathogens bind to surfaces are of interest to a wide variety of scientific communities, as these mechanisms drive infectivity, fate, and transport of the pathogenic organisms. This study begins to reveal the mechanism of direct binding of Cryptosporidium parvum to surfaces containing both carboxylic acid and amine moieties, in an attempt to understand how much of the binding ability is due to long-range electrostatic forces versus other mechanisms (specific or nonspecific) of bonding. In addition to improving the scientific understanding of fate and transport of oocysts, an expanded understanding of the binding mechanisms may aid in the development of new tools and sensors designed to detect and track oocysts in waterways. Furthermore, the methods used to examine binding in this study could be translated to other waterborne pathogens of interest.

ABSTRACT

Cryptosporidium parvum causes potentially life-threatening gastrointestinal disease in humans and may not be effectively removed from drinking water via conventional methods. Prior research has shown that environmental biofilms immobilize oocysts from the water column, but the biophysical mechanisms driving this attraction are still under investigation. This study investigates the affinity of C. parvum oocysts to silanized surfaces. Surfaces were prepared with hydroxyl, amine, and carboxyl moieties. Binding forces between the oocysts and these engineered substrates were analyzed, with and without divalent ions, using atomic force microscopy. Binding forces were measured over several weeks to investigate the influence of age on adhesion. C. parvum oocysts bind most strongly to carboxylic acid functional groups, with rupture forces greater than that required to break noncovalent molecular bonds, regardless of oocyst age. This adhesion is shown to be due to divalent cation bridging mechanisms. In addition, the binding strength increases over a 5-week period as the oocysts age, followed by a decrease in the binding strength, which may be related to structural or biochemical changes in the outer wall-bound glycosylated proteins. This study sheds new light on the biochemical parameters that influence C. parvum oocyst binding to surfaces. Increased understanding of how age and water chemistry influence the binding strength of oocysts may inform future developments in environmental detection and drinking water treatment, such as with the development of oocyst-specific sensors that allow for more frequent tracking of oocysts in the environment.

IMPORTANCE The mechanisms by which pathogens bind to surfaces are of interest to a wide variety of scientific communities, as these mechanisms drive infectivity, fate, and transport of the pathogenic organisms. This study begins to reveal the mechanism of direct binding of Cryptosporidium parvum to surfaces containing both carboxylic acid and amine moieties, in an attempt to understand how much of the binding ability is due to long-range electrostatic forces versus other mechanisms (specific or nonspecific) of bonding. In addition to improving the scientific understanding of fate and transport of oocysts, an expanded understanding of the binding mechanisms may aid in the development of new tools and sensors designed to detect and track oocysts in waterways. Furthermore, the methods used to examine binding in this study could be translated to other waterborne pathogens of interest.

INTRODUCTION

Cryptosporidium parvum is a zoonotic waterborne pathogen that can cause severe gastrointestinal disease (1–4). While the disease is generally self-limiting in healthy individuals, there are few treatment options for immunocompromised individuals (2, 5). The environmental state of the pathogen, a sporulated oocyst, is quite stable, and it is not generally removed by common water treatment methods (3, 6–8), thus necessitating research and development of novel C. parvum tracking devices as well as removal strategies.

One of the ongoing areas of inquiry involves the nature of C. parvum adhesion to surfaces (7, 9–22). For C. parvum to cause gastrointestinal disease, and thus complete the life cycle, the oocysts must be able to bind to the lumen of the small intestine (23–27). Prior research would indicate that, in vivo, lectins are a mediating molecule for intestinal binding (27). In addition, C. parvum has been found adhered to environmental biofilms (28–32), as well as to other surfaces associated with water treatment plants. In the environment, a few studies have identified a possible mechanism of calcium bridging between carboxyl or other negatively charged groups that are present on surfaces (10, 15, 18). Still other studies have identified electrostatic attraction as a mediating event in C. parvum binding (9–12, 18). However, no study to date has directly probed the surface interaction forces that are involved in C. parvum binding.

The ability of pathogens such as Cryptosporidium to survive harsh conditions is attributed to two main factors, the biochemical and physical structure of the oocyst and the ability of these oocysts to attach to other surfaces in the environment (6–14). The oocyst wall, which is a complex multilayer structure (6), is both a highly resistant barrier to physiochemical stressors and a mediator of surface attachment (6–8). The wall structure of the oocyst is multilayered, consisting of an inner wall, central wall, and outer wall (6). The inner layer contains a number of disulfide-bridged wall proteins that are believed to provide mechanical rigidity to the structure, while the central layer is composed of a lipid-protein complex (6). The outer wall exhibits a glucose-rich glycocalyx with an overall net negative charge that becomes more negative with increasing pH and less negative with ionic strength. This outer wall is largely suspected to be the mediator of electrostatically facilitated surface attachment in the environment, including the calcium-mediated bridging effect reported in prior works (13).

To directly study the mechanism of binding between C. parvum and specific functional groups known to mediate calcium bridging and electrostatic interactions, binding forces were measured using atomic force microscopy (AFM). AFM can be used to probe the interfacial region between the sample and tip, providing either a map of surface topography or a measure of interaction forces (force spectroscopy), data that represent a balance between the attractive and repulsive forces between tip and sample (33–37). Compared to methods used to assay surface composition or morphology (such as protein staining or electron microscopy), AFM force spectroscopy is a nondestructive means to directly probe binding forces. Using this method, comparative binding forces between a material attached to the AFM probe tip (an oocyst, in this case) and the surface can be measured. Binding forces were examined as a function of both surface chemistry and oocyst age, directly probing the calcium-mediated bridging phenomenon recently reported (15). In the future, these results can be used to inform both C. parvum environmental tracking devices (such as deployable sensors) that are sufficiently robust to detect oocysts at all infective ages and C. parvum removal strategies for both water and wastewater treatment that take advantage of specific binding characteristics of the pathogen.

RESULTS AND DISCUSSION

Silanization of glass surfaces.

Surfaces were silanized and analyzed for surface chemistry and topographical features using attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) and AFM. Silanization times were maintained to allow for formation of a thin layer of the surface moieties on the surface, building a monolayer to a few layers of the surface groups. ATR-FTIR (Fig. 1) revealed weak peaks consistent with the surface groups of interest, namely, –OH stretch for critically cleaned glass (3,400 to 3,300 cm⁻¹), –NH₂ stretch (3,500 to 3,100 cm⁻¹) and –NH₂ bend (1,640 to 1,550 cm⁻¹) for amine-functionalized surfaces (via use of 3-aminopropyltrimethoxysilane), and –C=O stretch (1,725 to 1,700 cm⁻¹) for the –COOH-functionalized surfaces (via carboxyethylsilanetriol). In addition, for both silane functionalities, appropriate alkane peaks were identified, namely, –CH₂ stretch (3,000 to 2,850 cm⁻¹) and –CH₂ bend (1,565 to 1,550 cm⁻¹).

FIG 1.

Characterization of glass functionalization. ATR-FTIR spectra taken for surfaces terminated with hydroxyl, amine, and carboxyl functional groups. At the experimental pH of 6.8, the –NH₂ takes on the form of –NH₃⁺, and the –COOH becomes –COO⁻, based on the pKa of the molecules.

AFM scans of the silanized surfaces were collected (Fig. 2). The surface morphologies of the two silanized surfaces exhibit nanoscale roughness with some degree of variability. The amine surfaces exhibited the highest degree of roughness (Table 1), likely due to the chemical nature of the (3-aminopropyl)trimethoxysilane. On both the amine and carboxyl surfaces, small islands are noted, consistent with traditional silanization (38–42) processes that yield some multilayered areas. The roughness of the surfaces is well below that of many natural and synthetic surfaces on which Cryptosporidium can attach (19, 43), which exhibit nanoscale to microscale roughness. While previous work has shown that oocyst attachment to biofilms is positively correlated with microscale biofilm roughness (43), given the nature of the experiments described here, with measurement of direct interactions between the surfaces and oocysts, the small differences in roughness are not a likely factor in the overall observations.

FIG 2.

AFM images of functionalized surfaces. (a) AFM height profile of hydroxyl surface (critically cleaned glass), (b) AFM height profile of amine-modified surface, (c) AFM height profile of carboxyl-modified surface, (d) phase image of hydroxyl surface, (e) phase image of amine-modified surface, and (f) phase image of carboxyl-modified surface.

TABLE 1.

Roughness of surfaces, as measured by AFM

Functional group	RMS roughness at indicated scan sizea
	1 μm	5 μm	10 μm
–OH	0.354	0.443	0.977
–NH2	1.82	4.25	5.92
–COOH	0.509	0.879	1.48

^aRMS (root mean square) roughness = Rq (nm).

Force spectroscopy.

Mouse-shed oocysts were selected for this study, primarily based on the consistent availability of the specimens, as well as the presumed genotypic stability of the strain after passaging through mammalian systems (44). An individual Cryptosporidium oocyst was attached to the AFM tip for each experiment performed, as confirmed by the integrated light microscope on the AFM instrument. AFM-based force spectroscopy was used in this study to probe the interaction forces between C. parvum oocysts and model functionalized surfaces (Fig. 3). Using AFM to measure the magnitude of binding allows for direct, quantitative comparative analysis between conditions, both surface treatments as well as surrounding medium parameters. The technique is widely used for biophysical characterization; here, we demonstrate the utility of the technique for measuring binding forces between waterborne pathogens and model surfaces. The data shown here demonstrates that AFM is sufficiently robust to compare binding strengths across surfaces and in different water conditions. Thus, the technique could be translated across surfaces of interest, such as biofilms, cell culture systems, and/or sensor surfaces to expand the scientific understanding of binding characteristics under different water/medium chemistries. The characteristic shape of the force curves was similar between the different surface moieties and indicates that there may be multiple nonspecific bonding mechanisms occurring between the oocysts and the functionalized surfaces (Fig. 4). The forces observed are consistent with both long-range forces (i.e., van der Waals) and longer-range electrostatic double-layer (EDL) forces (11, 14, 45–48); the Derjaguin-Landau-Verwey-Overbeek (DLVO) hypothesis is frequently cited when discussing these two types of forces in dispersion stabilization (45, 46, 48). Van der Waals forces are weak attractive forces that arise between surfaces exhibiting dipole character and are not generally considered to be sensitive to ionic strength (45, 48). EDL interactions can be either attractive or repulsive, depending upon surface charge, and are highly sensitive to the ionic composition of the surrounding media (45, 48). Both types of forces can be deduced from the AFM binding profiles.

FIG 3.

(a) A schematic of the experimental procedure showing one force curve process in the AFM adhesion assay with a C. parvum oocyst. (b) Typical force-distance curve for oocyst binding to a model surface. (c) Light microscope image of a single oocyst adhered to the AFM probe over the functionalized surface (CCD camera with ×20 magnification). Single oocyst attachment to the tip was confirmed for each experiment performed.

FIG 4.

(a) Three sample AFM curves demonstrating the interaction between C. parvum oocysts and the functional groups tested (OH, NH₂, COOH). In these graphs, all oocysts are aged 1 week after purification. The OH trace involves very little interaction. The NH₂ trace indicates an intermediately scaled detachment force. The strongest adhesion occurred between the oocyst and the COOH surface. (b) Force spectrum traces for C. parvum oocysts bound to the COOH surface at different ages (week 1, week 4, and week 7 after purification). Arrows in the picture indicate rupture events.

On the carboxylic acid surfaces, multiple rupture events can be observed, even as the oocysts undergo aging (Fig. 5), which could be consistent with either multiple EDL-type forces, cation bridging, or other mechanisms that are yet unstudied (10, 11, 21, 47, 49); the magnitude of these events indicates that many noncovalent interactions are involved in the binding process (50). Overall, the magnitude of the forces associated with binding to carboxylic acid-decorated surfaces is at least twice that associated with amine-decorated surfaces, indicating a large contribution of multiple cation bridging or similar phenomena associated with nonspecific binding of oocysts to surfaces. On the amine surfaces, there are apparent rupture events, which we suspect to be primarily electrostatic (EDL) in nature, based on prior knowledge (9, 10, 12, 18); the initial event is generally the largest in magnitude, with a rupture strength that would indicate that many electrostatic events are involved (50). On hydroxyl surfaces, binding is very weak, with limited features observed in the binding force curves, indicating primarily van der Waals type forces. On all surfaces, the interaction strength between the oocysts and surfaces is highly variable, as indicated by the scale bars. This variability is not inconsistent with previous studies of Cryptosporidium structure, as the nature of the outer wall is known to be sensitive to a variety of perturbations (46, 48). In addition, as we used mouse-shed oocysts, the timing of different preparations may impact the surface characteristics in ways that are not yet understood.

FIG 5.

Detachment force between oocyst and model surfaces in an artificial stream water solution with divalent cations. *, P < 0.01 compared to week 1 binding force (compared on same surface chemistry); #, P < 0.01 compared to binding force on –OH surface of similarly aged oocysts; ^, P < 0.01 binding force comparison between –NH₂ surface and –COOH surface. By week 6, the electrostatic attraction between the –NH₂ surfaces and oocysts has decreased to that of the –OH surface chemistry, while the carboxylic acid surfaces still exhibit significant attractive forces.

We suspect that the magnitude of the binding forces between oocysts and different surfaces may be highly variable between oocysts derived from different animal models (calf versus mouse), as well as different species (C. parvum versus C. hominis). This is based on both the observations in this study (high variability even within an oocyst population) and the potential biochemical variability of the surfaces of different oocyst species. While there is still insufficient understanding of the outer-wall composition of oocysts, it is known that the Cryptosporidium oocyst wall proteins (COWPs) do vary among different species (51), and it is highly likely that the degree of glycosylation of the COWPs is also variable, although the patterns of glycosylation on these proteins have been shown to be rather simple (52). However, given the structural and molecular similarities (glycosylated states) of the different known outer wall components, we suspect that different species and preparations of oocysts would ubiquitously interact most strongly with the carboxylic acid-decorated surfaces, although perhaps at a greater or smaller magnitude, depending upon the number of noncovalent bonds that form between the oocyst and the surface.

Aging of oocysts influences surface binding.

There have been many reports regarding oocyst viability and infectivity during the aging process. Different oocyst vendors, different preparation methods, and even different times of the year can influence these properties in commercially derived oocysts (53, 54). Our study tried to minimize these effects by using an animal source (mice) that is kept in a controlled environment, so that the impacts of weather, feedstock, and other variables did not factor into the aging process. Oocysts were analyzed every week for 8 weeks after shedding and purification. At each time point, the binding force between oocysts and the carboxyl surface was significantly greater (P < 0.01) than the binding force between oocysts and the amine- or hydroxyl-based surfaces, although the magnitude of the binding force decreased after multiple weeks of aging (Fig. 4b; Fig. 5). Prior research has indicated that there is likely a calcium bridging mechanism at play during surface adhesion of C. parvum oocysts to biofilms (15). This hypothesis was directly tested in this study. The sequential detachment features (rupture events) highlighted in Fig. 4b are consistent with an ionic bridging binding process that arises due to EDL events (10, 11, 14, 45, 46); however, the magnitude of the force indicates that various surface characteristics of the oocysts may contribute to these interactions (11, 12). The C. parvum wall structure has been reported to be multilayered, with the outer layer of the wall covered in acidic glycoproteins (6, 54, 55); the outer-wall structure exhibits an isoelectric point of ∼2.5, thus exhibiting a negative charge at the experimental pH of 6.8. In addition, this outer wall possesses a glucose-rich glycocalyx, which likely plays a role in nonspecific physiochemical binding. It is hypothesized that this outer layer changes over time (7, 9, 13), which may also explain related increases in permeability of the oocyst, as reported by others. This ephemeral nature is also related to variable reports of zeta potential and other properties of oocysts, as well as to the various reports of infectivity potential of differently aged oocysts. Alternatively, the central layer is a glycolipid/lipoprotein complex that also possesses a weak negative charge, which may be responsible for the binding events that are observed after multiple weeks of aging (Fig. 5), if the outer wall has been subject to degradation. While degradation could occur, it is more likely that there is are macromolecular conformational changes on the surface that allows for decreased electrostatic binding, while retaining the ability of oocysts to respond to excystation stimuli (56); this hypothesis will be tested in the future.

Different oocyst subtypes, preparation protocols, and storage conditions have all been shown to impact oocyst infectivity (53, 54). While excystation and infectivity were not part of this study, we suspect that commensurate changes in these properties would occur as a result of our aging process. Specifically of interest are the changes in the macromolecular composition of the outer wall, which is often cited as ephemeral but which is known to be composed of negatively charged residues and a variety of glycan moieties (predominantly glucose) (57). Figure 5 shows a dramatic decrease in the binding forces after 6 weeks of aging, indicating that the outer wall is undergoing a change in the overall surface charge, which may reflect conformational changes in the outer proteins or a change in the glycosylation state (52). By week 7, the binding forces have been reduced to one-half of the maximum binding strength measured. As the outer glycocalyx of the oocyst is responsible for both the robust nature of the oocyst and its immunogenicity (56), the infectivity profile of the oocysts may be related only to a subset of wall proteins that do not undergo storage-related changes. Many of the glycoconjugates are presumed to have limited lateral mobility in the wall (56); however, the surface conformation of the glycosylated groups and other surface moieties responsible for nonspecific surface binding may be subject to storage-related transformations.

Storage and environmental conditions are known to impact oocyst viability and infectivity. For example, in previous studies, oocyst infectivity and viability are both reduced as the ambient temperature increases above 37°C, in part due to lipid structure decomposition at particularly high temperatures (57). In addition, at elevated temperatures, oocysts exhibit an increase in metabolic activity, thus depleting the amylopectin reserves that are used as the oocyst is in the free (nonhost) state (58). At lower temperatures, oocysts undergo less robust excystation but are shown to be more stable (56). The stability at lower temperatures may be related to the wall proteins and limited lateral mobility; it is also likely related to the structural wall rigidity at low temperatures. For our study, the long-term storage of the oocysts demonstrates that wall changes are occurring in a nonnegligible fashion, given the dramatic decrease in binding forces on all surface chemistries tested after week 5 of storage. All binding strengths, regardless of whether the binding is mediated by van der Waals, electrostatic, or other forces, decrease after 5 weeks of storage, indicating that surface charges and the surface presentation of glycosylated proteins are both impacted during cold storage. However, the nature of these wall changes and how they impact both surface binding in the environment (and thus transport) and infectivity necessitate significant further study.

Influence of divalent cations on surface binding.

To directly examine the theory of calcium bridging as the predominant mechanism of binding, calcium and magnesium ions were removed from the artificial stream water. The pH and ionic strength of the water was held constant; the water hardness, however, was influenced by this elimination. AFM was again used, under these different medium conditions, to directly compare binding forces between oocysts and model surfaces. Oocyst binding was again assessed over an 8-week period postshedding (Fig. 6). The magnitude of the binding forces between the oocysts and the –COOH surfaces decreased by an order of magnitude (P < 0.001), simply via removal of divalent cations, thus indicating that the cation bridging model is a highly likely mechanism of attraction between the model surfaces and the oocysts, as the data indicate disruption of this mechanism via simple removal of divalent cations. In the absence of divalent cations, the nature of the bonding between the surfaces and the oocysts appears to be primarily driven by electrostatic forces. The attractive forces exhibited by the oocysts to the –COOH surface were not statistically different from those between oocysts and the amine- or hydroxyl-based surfaces, except in highly aged oocysts (week 7 is an outlying data point regarding significant differences between modified surfaces and –OH surfaces), which indicates that the surface of the oocyst does undergo changes that influence the ability of the organism to bind to a surface, potentially related to the changes that are possibly occurring at the outer layer of the oocyst wall. While the glycoconjugates in the outer wall are presumed to have limited lateral mobility (56), the outer surface conformation is possibly subject to conformational disruption, which would also render the oocysts more permeable over time. In future work, we aim to study the nature of these changes, as well as to correlate them to infectivity of a cell culture model. Given that Cryptosporidium, similarly to other apicomplexans, likely utilizes multiple ligand-receptor events during the initial infection (59, 60), changes to the surface conformation of temperature/time sensitive wall biomolecules may not impact infectivity associated with storage.

FIG 6.

Detachment force between oocyst and model surfaces in an artificial stream water solution without divalent cations. *, P < 0.01 compared to week 1 binding force (compared on same-surface chemistry), #, P < 0.01 compared to binding force on –OH surface of similarly aged oocysts. No significant differences are noted in binding forces between the –NH₂ and –COOH surfaces and oocysts. Weeks 3, 4, and 5 exhibit the highest overall binding characteristics, indicating that the oocyst surface undergoes chemical changes during aging.

Interestingly, the binding forces appear to increase in weeks 3, 4, and 5, consistent with those observed in solutions with divalent cations (Fig. 7a to c). These forces are again highly variable between different oocysts tested, as shown by the error bars. It is apparent that the aging process changes the chemistry of the oocyst surface in a way that is significant to the overall binding profile of a population of oocysts, likely related to biomolecular conformational changes that could change as a result of cold storage. These biomolecular changes could influence surface charge density, presentation of electrostatically active groups, or hydrodynamic radius changes in surface proteins, leading to stronger long-range forces. The shapes of the force curves between the surfaces/oocysts with and without divalent cations are very similar (Fig. 7d), with the oocyst binding to the –COOH surface exhibiting multiple rupture events; these events are more frequent than those seen in oocysts binding to the –NH₂ surface. Future work will focus specifically on these changes, as the ages of oocysts in the environment are highly variable. As oocysts can remain infectious for several weeks, it is imperative that any removal or sensing device is able to correct for these surface changes.

FIG 7.

Comparative binding forces with and without divalent cations; divalent cations appear to play a role in binding potential between both the –COOH and –NH₂ modified surfaces. Age of the oocysts impacts this effect. *, P < 0.01 between binding forces with and without divalent cations on same surface after similar aging times. (a) –OH surface. (b) –NH₂ surface. (c) –COOH surface.

Conclusions.

Binding of Cryptosporidium oocysts and the infectious sporozoites to biological surfaces is a multifaceted affair, with many protein residues and interaction forces involved. However, in environmental systems, when oocysts are being transported in surface waters, these particles are known to bind to inorganic and organic materials in a charge-dependent manner, while the infectivity of the organism can persist for weeks or longer while the oocysts are environmentally transported. Atomic force microscopy was used to assess binding force and was found to be a reliable tool to quantify binding forces between particles such as Cryptosporidium oocysts and surfaces. AFM could potentially be used in the future to study adhesion of oocysts or other pathogens to surfaces of interest, as a way to examine both environmental fate and transport and infectious behaviors in cell culture systems. The study presented here demonstrates that environmental binding is related to a calcium-mediated mechanism that undergoes changes during aging of the oocysts. While infectivity may be retained during aging, based on previous reports, the surface composition or structure is likely undergoing changes that may influence the ability of sensing or removal devices to robustly capture all infective oocysts.

MATERIALS AND METHODS

Cryptosporidium oocysts.

Mouse-shed C. parvum oocysts (Iowa isolate) were obtained from Waterborne, Inc. (New Orleans, LA) and used within 8 weeks of shedding and purification. Four separate batches of oocysts were procured for the experiments, allowing for statistical analysis of the data.

Artificial stream water.

Artificial stream water (12 g calcium sulfate [CaSO₄], 2 g calcium chloride [CaCl₂], 2 g potassium chloride [KCl], 12 g magnesium sulfate [MgSO₄], 4 g sodium bicarbonate [NaHCO₃], and 8 g sodium borate [Na₂B₄O₇] per 50 gal deionized water) was used throughout, to minimize the impact of water chemistry on oocyst adhesion. Nutrient-free water was desirable to limit the impact of phosphates and nitrates on binding strength. To limit variability in our study, we chose to work with water that was prepared in-house, using a chemical composition consistent with water (derived from Hokendauqua Creek) used in a prior work (22). Calcium and magnesium hardnesses were equal to 72 ppm and 32 ppm, respectively, in solutions that contained divalent cations. Using a chemical composition prepared in-house also allowed for complete removal of calcium and magnesium from the water to test the divalent cation-mediated effects. Adhesion analysis in the absence of divalent cations required the artificial stream water to be produced without calcium or magnesium salts, thus altering the hardness dramatically. The pH of the water was maintained throughout at pH 6.8.

Surface functionalization.

Glass slides were cleaned in an ultrasonic bath with acetone. Water was used to thoroughly rinse the acetone residue, and slides were left to dry under vacuum. The slides were then critically cleaned with freshly prepared Piranha solution (70% H₂SO₄ and 30% H₂O₂) for 1 h, followed by rinsing with water and then drying under vacuum. Glass slides treated in this way were enriched with surface hydroxyl (–OH) groups. Slides were used as cleaned or further functionalized with carboxyl or amine groups as described below.

Carboxyl moieties were added to the glass slides by silanization with carboxylethylsilanetriol (CTES). The clean glass slides were immediately immersed in an aqueous solution of 2% CTES in deionized (DI) water for 2 h. During silanization, containers were placed on an orbital shaker at room temperature with gentle shaking (100 rpm). Excess reagent was rinsed with DI water, and slides were thermally cured at 110°C for 1 h.

Amine moieties were added to the glass slides with 3-aminopropyltrimethoxysilane (APTMS). The cleaned glass slides were first critically dried at 150°C for 4 h. The slides were then subject to further cleaning in methanol, methanol-toluene (1:1, vol/vol), and toluene, each for 10 min in an ultrasonic bath. Slides were immersed in APTMS (2% vol/vol in toluene) for 2 h, followed by baking at 110°C for 1 h.

Surfaces were analyzed by attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR; Perkin Elmer) and atomic force microscopy (AFM, Veeco). Surfaces were prepared fresh (within 1 day prior) for each experiment.

AFM probe preparation.

AFM probes (silicon nitride, MLCT-UC; Bruker Nano) were cleaned with acetone and sterilized by UV. The probes were functionalized with poly-l-lysine (PLL; pH 9.0, 1 μg/ml) overnight at 4°C and used immediately.

AFM instrument.

All force measurements were conducted using a laboratory-built AFM (61–64) that employs a single-axis piezoelectric translator equipped with a strain gauge (P841.10; Physik Instrumente, Karlsruhe, Germany) to control the absolute position of the AFM cantilever (65–66). The piezo can generate movements up to 15 μm, sufficient for the oocyst manipulation. The deflection of the cantilever was monitored optically by using an inverted optical system attached to the AFM. A focused laser spot from a fiber-coupled diode laser (Pegasus Optical Systems, Shanghai, China) was reflected off the back of the cantilever onto a 2-segment photodiode (SPOT-2D; OSI Optoelectronics, Hawthorne, CA) to monitor the cantilever deflection. The photodiode signal was then preamplified, digitized by a 16-bit analog-to-digital converter (ITC-18; Heka Elektronik, Lambrecht, Germany), and processed by a computer. The sample holder of the apparatus was designed to accept a standard 35-mm tissue culture dish. Oocysts were observed using an inverted optical system equipped with a 20× objective (Thorlabs, Newton, NJ) attached to a charge-coupled device (CCD) camera. The setup has been previously used to study a wide variety of cell-cell and cell-surface interactions (67–75).

AFM adhesion assay.

C. parvum oocysts were attached to the PLL-coated cantilever, relying on the net charges of the oocysts (negative) and probe tip (positive) (Fig. 1, bottom panel). The observed force between the PLL-coated tip and oocyst were sufficient to limit detachment of the oocyst from the tip during the experiments. The interaction force between the oocyst surface and different model surfaces was measured with a force of approximately 300 pN and a dwell time of 0.5 s. A total of 20 force scans were completed for each condition/oocyst, with a recovery time of at least 2 min between scans (Fig. 1, middle panel).

Statistical analysis.

Data were examined by an analysis of variance (ANOVA). Post hoc test comparison was performed using Tukey’s honest significant difference (HSD) test to evaluate the potential relationships between the analyzed variables.