Langmuir 2007, 23, 7475-7483
7475
Articles Investigation of the Interaction Force between Cryptosporidium parWum Oocysts and Solid Surfaces T. L. Byrd† and J. Y. Walz*,‡ Yale UniVersity, Department of Chemical Engineering, New HaVen, Connecticut 06520, and Virginia Tech, Department of Chemical Engineering (0211), Blacksburg, Virginia 24061 ReceiVed January 18, 2007. In Final Form: April 9, 2007 Interaction force profiles between single Cryptosporidium parVum oocysts and positively charged, silane-coated silica particles were measured in aqueous solutions using an atomic force microscope. The oocysts were immobilized for the measurements by entrapment in Millipore polycarbonate membranes with 3 µm pore size. Experiments were performed in both NaCl and CaCl2 solutions at ionic strengths ranging from 1 to 100 mM. For both electrolytes, the decay length of the repulsive force profile was found to be nearly independent of the ionic strength and always much larger than the theoretical Debye length of the system. In addition, the magnitude of the force was found to be essentially the same for both electrolytes, suggesting that the long-range repulsive forces are primarily steric in nature. These results support the theory that the interaction force between oocysts and surfaces is controlled by an outer, weakly charged or uncharged carbohydrate layer. Measurements were also performed with oocysts that had been deactivated using either chemical (formalin) or heat treatment. The force profiles obtained with formalin-treated oocysts appear to be essentially the same as for the untreated oocysts, whereas the profiles measured with the heattreated oocysts show a much stronger dependence on solution ionic strength. With either the heat-treated or formalintreated oocysts, adhesion was observed much more frequently than with untreated oocysts, which is consistent with the increased deposition rate observed with treated oocysts by Kuznar and Elimelech (Kuznar, Z. A.; Elimelech, M. Langmuir 2005, 21, 710-716). These results also suggest that treated oocysts, especially ones that have been inactivated by heating, may not be good surrogates for viable oocysts in laboratory studies.
Introduction Cryptosporidium parVum is a protozoan parasite that has been found to cause the diarrheal disease cryptosporidiosis.1 Transport of C. parVum from the environment to a human host mainly occurs through contact with contaminated open surface waters. In fact, studies of surface water sources in the United States have shown the presence of C. parVum in 55% of surface water samples. Cryptosporidium parVum has even been found in 17% of drinking water samples.2 Other studies by LeChevallier et al. found C. parVum in 27% of drinking water samples and contamination in 87% of raw water samples.3,4 With continued outbreaks of C. parVum occurring globally, it is clear that further investigation of the transport of Cryptosporidium from its source to these open surface waters is warranted. C. parVum is typically found in the environment in the form of oocysts. Ranging in diameter from 3.9 to 5.9 µm,5 these oocysts are primarily spherical and are composed of four inner sporozoites surrounded by a thick shell wall. This protective covering of the C. parVum wall provides strong protection against harsh * To whom correspondence should be addressed. E-mail:
[email protected]. † Yale University. ‡ Virginia Tech. (1) Meisel, J. L.; Perera, D. R.; Meligro, C.; Rubin, C. E. Gastroenterology 1976, 70, 1156-1160. (2) Rose, J. B.; Gerba, C. P.; Jakubowski, W. EnViron. Sci. Technol. 1991, 25, 1393-1400. (3) LeChevallier, M. W.; Norton, W. D.; Lee, R. G. Appl. EnViron. Microbiol. 1991, 57, 2617-2621. (4) LeChevallier, M. W.; Norton, W. D.; Lee, R. G. Appl. EnViron. Microbiol. 1991, 57, 2610-2616. (5) Medema, G. J.; Schets, F. M.; Teunis, P. F. M.; Havelaar, A. H. Appl. EnViron. Microbiol. 1998, 64, 4460-4466.
environmental conditions, including some protection from chemical treatments such as chlorination.6 The oocyst wall is between 40 and 50 nm thick and is composed of an inner layer containing a suture up to one-half the length of the oocyst circumference and an outer layer that is between 5 and 10 nm thick and is thought to be composed primarily of acidic glycoproteins.7-9 These acidic glycoproteins are thought to form a “hairy” layer extending from the surface of the oocyst and include proteins that are composed primarily of cysteine, proline, and histidine.8 The oocyst wall also contains both glycolipids and phospholipids, which play an important role in determining oocyst wall permeability, surface charge, and hydrophobicity.10 The primary oocyst treatment strategies used in drinking water purification are physical-chemical removal using sand-bed filters and UV irradiation.11,12 Other techniques that have been found to be effective are ozone disinfection13-15 and mixed oxidant (6) Smith, M.; Thompson, K. C. Cryptosporidium - The Analytical Challenge; Royal Society of Chemistry: Cambridge, U.K., 2001. (7) Reduker, D. W.; Speer, C. A.; Blixt, J. A. J. Protozool. 1985, 32, 708-711. (8) Harris, J. R.; Petry, F. J. Parasitol. 1999, 85, 839-849. (9) Nanduri, J.; Williams, S.; Aji, T.; Flanigan, T. P. Infect. Immun. 1999, 67, 2022-2024. (10) Robertson, L. J.; Campbell, A. T.; Smith, H. V. Appl. EnViron. Microbiol. 1993, 59, 2638-2641. (11) Morita, S.; Namikoshi, A.; Hirata, T.; Oguma, K.; Katayama, H.; Ohgaki, S.; Motoyama, N.; Fujiwara, M. Appl. EnViron. Microbiol. 2002, 68, 5387-5393. (12) Belosevic, M.; Craik, S. A.; Stafford, J. L.; Neumann, N. F.; Kruithof, J.; Smith, D. W. FEMS Microbiol. Lett. 2001, 204, 197-203. (13) Li, L. J.; Haas, C. N. J. Water Supply: Res. Technol.-AQUA 2004, 53, 287-297. (14) Corona-Vasquez, B.; Samuelson, A.; Rennecker, J. L.; Marinas, B. J. Water Res. 2002, 36, 4053-4063. (15) Parker, J. F. W.; Greaves, G. F.; Smith, H. V. Water Sci. Technol. 1993, 27, 93-96.
10.1021/la0701576 CCC: $37.00 © 2007 American Chemical Society Published on Web 06/08/2007
7476 Langmuir, Vol. 23, No. 14, 2007
disinfection.16,17 Oocyst removal by riverbank filtration is commonly practiced in many parts of the world.18 Recently, a number of research groups have become interested in understanding the nature of the interaction force profiles between oocysts and various surfaces. Although these experiments are typically conducted using model systems, the hope is that these force profiles will provide fundamental information about the properties of the oocyst surface, which could in turn lead to improved removal methods or, perhaps, a better understanding of the degree to which oocysts can be transported through the soil. Considine et al. obtained a direct measurement of the interaction force between a single oocyst that had been fixed to a silicon wafer and a silica particle that had been glued to the AFM cantilever tip.19,20 The authors discovered that the force was entirely repulsive upon approach with a decay length (typically measured 20 to 50 nm from contact) of 15 to 16 nm. The decay lengths decreased with increasing ionic strength, specifically in the presence of calcium ions where decay lengths of 6 to 10 nm were observed. These authors also discovered that most of the oocysts adhered to the opposing silica particle upon contact. This adhesion produced a series of spikes in the force curve upon retraction that could extend out to 100 nm. Considine et al. attributed these pull-off spikes to the specific adsorption of oocyst surface proteins to the silica surface; however, they found no correlation between the adhesion force observed and the ionic strength of the solution. On the basis of their results, Considine et al. proposed that the oocyst/silica force profile consisted of a long-range electrostatic repulsive force followed by a shorter-range steric repulsion, consistent with a polyelectrolyte brush layer thickness of 50 nm.20 The authors theorized that the polyelectrolyte brush layer of the oocyst was stretched out into solution by the electrostatic charge repulsion of charges along the brush chain and that the compressibility of this layer varied from oocyst to oocyst.20 However, upon the addition of dissolved calcium ions, compression of the hairy layer occurred, along with an effective decrease in charge density.19 More recently, Kuznar and Elimelech published results on the deposition of C. parVum oocysts onto an ultrapure quartz surface by means of radial stagnation point flow experiments.21-23 Although deposition was predicted at higher ionic strengths (based on DLVO theory), these authors found no deposition of oocysts when the concentration of a monovalent salt, KCl, was varied over a range of 1 to 316 mM. Deposition did occur in the presence of a divalent salt, CaCl2, though the degree was substantially lower than that predicted using DLVO theory. Kuznar and Elimelech proposed that an electrosteric repulsion existed between the oocysts and the silica surface as a result of the surface characteristics of the oocyst. By modifying the oocyst surface through protein cross-linking, heat treatment, and protein removal, they were able to obtain higher oocyst attachment efficiencies. (16) Biswas, K.; Craik, S.; Smith, D. W.; Belosevic, M. Water Res. 2005, 39, 3167-3176. (17) Son, H.; Cho, M.; Chung, H.; Choi, S.; Yoon, J. J. Ind. Eng. Chem. 2004, 10, 705-709. (18) Tufenkji, N.; Ryan, J. N.; Elimelech, M. EnViron. Sci. Technol. 2002, 36, 422a-428a. (19) Considine, R. F.; Dixon, D. R.; Drummond, C. J. Water Res. 2002, 36, 3421-3428. (20) Considine, R. F.; Drummond, C. J.; Dixon, D. R. Langmuir 2001, 17, 6325-6335. (21) Kuznar, Z. A.; Elimelech, M. EnViron. Sci. Technol. 2004, 38, 68396845. (22) Kuznar, Z. A.; Elimelech, M. Langmuir 2005, 21, 710-716. (23) Kuznar, Z. A. Elimelech., M. EnViron. Sci. Technol. 2006, 40, 18371842.
Byrd and Walz
In an earlier paper, we described a series of AFM experiments that directly measured the force profile between C. parVum oocysts and negatively charged silica particles.24 A unique aspect of this work was the use of a novel immobilization technique in which individual oocysts were physically trapped in the pores of polycarbonate membranes. The test silica particles were glued to the tip of AFM cantilevers. These measurements showed that the force profile was dominated by a long-range steric force that was essentially independent of either ionic strength (up to 100 mM ionic strengths) or counterion valence (NaCl and CaCl2 were used). Fitting the measured force profiles to the brush-layer model of Alexander and de Gennes25 suggested a layer thickness of roughly 115 nm. It should be emphasized that unlike Considine et al. we observed the force profile to be dominated by a steric repulsion over all separation distances sampled (0 to 100 nm). Specifically, no significant changes in the force profile were observed over this range of separation when the ionic strength varied by two orders of magnitude (from 1 to 100 mM) using both monovalent (NaCl) and divalent (CaCl2) electrolytes. In this article, we first describe additional measurements performed with positively charged, silane-coated silica particles. The goal is to further investigate the relative importance of steric versus electrostatic forces on the net interaction force profile. We also describe a series of experiments performed using oocysts that had been inactivated using either heat treatment or chemical treatment (formalin). Because such deactivated oocysts are sometimes used in experimental testing to avoid the potential risks associated with using live oocyts,26 we sought to determine whether such treatment alters the interaction force profile between the oocysts and solid surfaces. Experimental Section Apparatus. Force profiles for our experiments were measured by using a Digital Instruments NanoScope IIIa Multimode atomic force microscope (Digital Instruments, CA). Each colloidal probe used in these experiments was created by gluing (Norland Optical Adhesive 68, Lot 173) a 5 µm silica or silane-coated silica particle (Bangs Laboratories, Fishers, IN) to the end of a silicon nitride cantilever tip (Veeco, model number MLCT-AUNM). The glue was then cured by exposing the colloidal probe to ultraviolet light via a UV/O3 chamber for 10 min. The cantilever probe used in our experiments was triangular and 220 µm in length with a manufacturerreported spring constant of 0.03 N/m. Sample Preparation. Viable oocysts used in all experiments were supplied by Sterling Parasitology Laboratory (SPL) from the University of Arizona. According to the supplier, all oocysts were shed from the same calf that was originally infected with an Iowa isolate and obtained from the National Animal Disease Center in Ames, IA. All oocysts used were purified using discontinuous sucrose and cesium chloride centrifugation gradients. The oocysts were then stored in an antibiotic solution containing 0.01% Tween 20 (to prevent aggregation), 100 U of penicillin, and 100 µg of gentamicin per mL (to prevent bacterial growth from occurring within the sample). All experiments were performed using oocysts that were within 3 months of their shed date. The oocysts used in the formalin-treated experiments were treated with SPL by contacting the oocysts with a 5% formalin solution.27 These oocysts were treated with formalin immediately after being shed, and the experiments were conducted within 6 weeks after the shed date. Oocysts used in the heat-treatment experiments were heated upon arrival to 80 °C for 1 h in a Boekel dry bath incubator (24) Byrd, T. L.; Walz, J. Y. EnViron. Sci. Technol. 2005, 39, 9574-9582. (25) de Gennes, P. G. C. R. Acad. Sci., Ser. II 1985, 300, 839-843. (26) Tufenkji, N.; Dixon, D. R.; Considine, R.; Drummond, C. J. Water Res. 2006, 40, 3315-3331. (27) Robertson, L. J.; Campbell, A. T.; Smith, H. V. Water Res. 1993, 27, 723-725.
Interaction Force between Oocysts and Solid Surfaces (Boekel Scientific, Feasterville, PA) in accordance with the procedure described by Kuznar and Elimelech.22 Prior to performing experiments, each of the 1 mL samples of 1 × 106 oocysts that were purchased from SPL were centrifuged, according to the supplier’s specifications, at 12 000 rpm (13 278 rcf) for 1 min using an Eppendorf 5415D microfuge with a fixed angle rotor. After centrifugation, the antibiotic solution supernatant was removed and replaced with deionized water (Barnstead). This procedure was repeated two more times to ensure the removal of all antibiotic solution initially present. After all solution was removed, the oocysts were resuspended by means of a vortex mixer and stored at 4 °C. No further treatments were performed. Physical immobilization of the oocysts in polycarbonate membranes during the actual measurement of the force profiles was performed using the membrane entrapment procedure described by Byrd and Walz.24 Solution Preparation. Aqueous sodium chloride solutions with concentrations of 1.0, 3.16, 10.0, 31.6, and 100.0 mM were prepared using ACS-grade NaCl (Mallinckrodt AR, lot number 7581 A13634) and deionized water. The solutions were not buffered, and the pH of the solution ranged from 5.5 to 5.7. Similarly, aqueous calcium chloride solutions with concentrations of 0.333, 1.05, 3.33, 10.5, and 33.3 mM were made using ACS-grade CaCl2 (Aldrich, lot number 07516JU) and deionized water. These concentrations of CaCl2 produced ionic strengths equivalent to those of the NaCl solutions mentioned above. The pH of these solutions ranged from 5.6 to 6.9. All solution concentrations were verified by means of pH and conductivity measurements. Test Particles. Particles for all experiments were provided by Bangs Laboratory, Inc. (Fishers, IN). The uncoated silica spheres had a manufacturer-reported mean diameter of 5 µm. The particles were cleaned using dialysis to ensure that all remaining surfactant was removed, and they were stored in DI water prior to use. The silane-coated microspheres had a mean diameter of 4.74 µm and R-NH2 surface functional groups. The microspheres were guaranteed by the manufacturer to be free of surfactant and were stored in DI water prior to use. Zeta potential measurements on the silane-coated particles were performed in both monovalent and divalent salt solutions with ionic strengths of 1, 10, and 100 mM using a ZetaPALS instrument (Brookhaven Instruments Corporation, Holtsville, NY). The measurements were performed at a fixed temperature of 25 ( 1 °C and at ambient pH (5.5-5.7). Positioning the Oocysts. To avoid an additional lateral component being present in the force interactions between the oocyst and a colloidal particle, it was important that the particle probe be positioned directly over the top of the oocyst. The centering of the oocysts beneath the particle was done using the technique described by Byrd and Walz.24 First, a relatively large area was scanned until an oocyst was located. The scanned image was then analyzed in 3-D imaging mode using the NanoScope software to verify that an actual oocyst was present. This was done by comparing the relative heights and shape of the oocyst to those of the surrounding membrane. The oocyst was then placed in the center of the AFM scan, and the scan size was incrementally reduced, with the oocyst being repeatedly centered within the image after each scan. This process was repeated until a scan size of 5 µm × 5 µm was obtained. AFM Force Measurements. Once an oocyst was centered under the colloidal particle of the colloidal probe, the AFM was switched from imaging mode to force mode. A minimum of 40 force measurements were then obtained in force mode using a scan rate of 0.996 Hz and a ramp size of 1 µm. Each force curve generated by the AFM Nanoscope software was an average of five individual force measurements. Once force curves had been obtained for the lowest strength solution (1 mM NaCl), the colloidal probe was disengaged from the oocyst surface. The fluid cell was then drained and flushed with the next highest ionic strength solution. Positioning and force measurements were repeated for each ionic strength. In the case where both a monovalent and a divalent salt solution were used with the same oocyst surface, the fluid cell was first drained and flushed with the lowest ionic strength monovalent salt (1 mM
Langmuir, Vol. 23, No. 14, 2007 7477
Figure 1. Zeta potential of the siliane-coated silica particles in aqueous NaCl and CaCl2 solutions. Each data point is the arithmetic average of approximately 50 measurements, and the error bars represent the standard deviation. Note that the lines through the data points are simply meant to guide the eye. NaCl), followed by draining and flushing with the lowest ionic strength solution of the divalent salt (0.333 mM CaCl2).
Results Zeta Potentials of Silane-Coated Silica Particles. Zeta potentials for the silane-coated silica particles in NaCl with ionic strengths of 1, 10, and 100 mM ranged from 26 to 59 mV in NaCl and from 17 to 50 mV in CaCl2 (Figure 1). The zeta potential of C. parVum oocysts has been measured by numerous researchers19-22,28-33 and is negative (-19 to -36 mV) for the conditions used in these experiments. Silane-Coated Silica Particle/Polycarbonate Membrane Force Profiles. As in our previous experiments with negatively charged, uncoated silica particles, we first verified our system by measuring the interaction force profile between a silanecoated silica particle and the polycarbonate membrane itself. The membrane is known to be negatively charged, and the interaction should be primarily electrostatic in nature. A relatively smooth section of the membrane was chosen to avoid problems with the membrane pores. Experiments at varying ionic strengths of an aqueous NaCl solution were performed using a single particle at the same membrane location. The results of these measurements are presented in Figure 2, which shows plots of the measured force versus particlemembrane separation distance. At the highest ionic strengths, little net interaction is seen until contact, at which point a strong, essentially hard-wall repulsion is observed. At an ionic strength of 3.16 mM, a slight long-range repulsive force is observed just prior to contact. Given the opposite charge between the particle and membrane, this repulsion is unexpected. It is possible that the force is due to irregularities on either the particle or substrate, (28) Thomas, F.; Bard, E.; Rouillier, M. C.; Prelot, B.; Mathieu, L. Colloids Surf., A 2001, 195, 135-142. (29) Considine, R. F.; Dixon, D. R.; Drummond, C. J. Langmuir 2000, 16, 1323-1330. (30) Dai, X.; Boll, J. J. EnViron. Qual. 2003, 32, 296-304. (31) Drozd, C.; Schwartzbrod, J. Appl. EnViron. Microbiol. 1996, 62, 12271232. (32) Karaman, M. E.; Pashley, R. M.; Bustamante, H.; Shanker, S. R. Colloids Surf., A 1999, 146, 217-225. (33) Searcy, K. E.; Packman, A. I.; Atwill, E. R.; Harter, T. Appl. EnViron. Microbiol. 2005, 71, 1072-1078.
7478 Langmuir, Vol. 23, No. 14, 2007
Figure 2. Force profiles measured upon approach between a 5 µm silane-coated silica particle and a polycarbonate membrane surface at various ionic strengths of an aqueous NaCl solution.
Figure 3. Force vs separation distance for a 5 µm silane-coated silica particle interacting with a trapped oocyst at various ionic strengths of an aqueous NaCl solution.
such as extended polymer chains that cause a steric force; however, this is purely speculation at this point. At the lowest ionic strength, the repulsion is shifted to much larger separations (onset at approximately 90 nm), after which a strong attraction is observed that is presumably due to the opposite charges on the particle and membrane. In addition, rather than a steep-wall repulsion, a much more gradual repulsion, extending over 40 nm, is found. Although some of these characteristics are quite unexpected and are not understood, the essential components of the force profile are (1) the profiles are clearly sensitive to the ionic strength of the solution and (2) the interaction essentially disappears once the ionic strength exceeds roughly 10 mM, indicating that the origin of these forces is electrostatic in nature. Force Profiles between C. parWum Oocysts and SilaneCoated Silica Particles in NaCl and CaCl2. Figure 3 shows typical force profiles measured between a silane-coated silica particle and a single C. parVum oocyst in various ionic strengths of an aqueous NaCl solution. It is clear that whereas some small
Byrd and Walz
Figure 4. Force vs separation distance for a 5 µm silane-coated silica particle interacting with a trapped oocyst at various ionic strengths of an aqueous CaCl2 solution.
variations are seen at the different ionic strengths the curves are all very similar, especially considering that the ionic strength varies by two orders of magnitude. Similar results were reported by Byrd and Walz for the force profile between an uncoated, negatively charged silica particle and an oocyst.24 The decay lengths of the force profiles were obtained from the slope of the linear portion of each profile (e.g., between 20 and 60 nm). For a purely electrostatic interaction, the decay length (or Debye length) should scale with ionic strength to the -1/2 power. The decay lengths obtained from the profiles in Figure 3 varied between 22 and 24 nm and scaled with ionic strength to the -0.04 power. This very weak dependence on ionic strength clearly indicates that the origin of the repulsion is not electrostatic. Similar behavior was also observed for experiments conducted with various ionic strengths of an aqueous CaCl2 solution. These results are presented in Figures 4 and 5. As in the case of NaCl, the force profiles obtained for CaCl2 show the presence of a repulsive force between the two surfaces that has little dependence on the solution ionic strength, even out to separations as large as 80 nm. Once again, we can characterize the force between silane-silica and the oocyst in terms of a decay length, shown in Figure 5, which further demonstrates that the dependence on ionic strength is much weaker than what would be expected for a purely electrostatic interaction. We should again emphasize that these curves are very similar to those obtained previously with negatively charged, uncoated silica particles. Because the force profiles obtained in the NaCl and CaCl2 solutions involved different oocysts and cantilevers, the force profiles cannot be directly compared. Nonetheless, a comparison between force profiles obtained with different electrolytes is important in determining whether there is any dependence on the nature of the electrolyte. Thus, a series of experiments were performed on a single oocyst with the sample cell first flushed with NaCl solutions, followed by a sample cell rinse, and then the CaCl2 solutions. Figure 6 shows a direct comparison of the force for various separation distances in solution ionic strengths of 1.0 and 31.6 mM. Although some small differences are observed, the force profiles are very similar for both electrolytes. Force Profiles Using Formalin-Treated C. parWum Oocysts. Because inactivated oocysts pose significantly reduced health
Interaction Force between Oocysts and Solid Surfaces
Langmuir, Vol. 23, No. 14, 2007 7479
Figure 5. Log-log plot showing the variation of the decay length of the measured force profiles with ionic strength for CaCl2. Each data point represents an average of 20 individual force profiles.
Figure 6. Comparison of the forces measured on a single oocyst in solutions of NaCl and CaCl2. (The measurements in both the NaCl and CaCl2 solutions were all performed using the same oocyst.) The error bars on the data points represent the standard deviation.
risks, a safer alternative is to use oocysts that have been chemically or physically inactivated for research studies as oocyst surrogates.26 One common method of chemically inactivating oocysts is formalin treatment.27 We sought to address the question of whether such treatment alters the nature of the interaction force profile between oocysts and oxide surfaces. We first show in Figure 7 AFM scan images of both an untreated (Figure 7a) and formalin-treated (Figure 7b) oocyst. It should be mentioned that each of these images was obtained with a 5 µm silica particle attached to the AFM cantilever, which is the reason for the lack of detail. Nonetheless, it is clear that whereas the untreated oocyst appears to be relatively round in shape the formalin-treated oocyst is greatly elongated or smeared. This elongation was also observed with other formalin-treated oocysts
Figure 7. Two AFM images of an oocyst surface made with a cantilever that had a 5 µm silica particle glued to its end. Image (a) utilized an untreated oocyst surface, and image (b) utilized an oocyst that had been treated with a 5% formalin solution.
and suggests that the outer structure of the oocyst has somehow been altered by the treatment. Figure 8 shows typical force profiles measured between a silica particle and a single formalin-treated C. parVum oocyst in various ionic strengths of aqueous NaCl and CaCl2 solutions, respectively. The behavior seen here is similar to that observed between untreated particles and oocysts presented in our previous paper. Again, little change with ionic strength is observed. A direct comparison between NaCl and CaCl2 using formalintreated oocysts is given in Figure 9 to show the effect of electrolyte valence on the force profile for a formalin-treated oocyst. By comparing the force for various separation distances in solution ionic strengths of 1.0 and 31.6 mM, it can be seen that the electrolyte valence has little effect on the force profiles because the force profiles at each ionic strength are essentially equal.
7480 Langmuir, Vol. 23, No. 14, 2007
Byrd and Walz
Figure 9. Comparison of the force measured with different electrolytes and different ionic strengths using a single formalintreated oocyst and an untreated silica particle. (The measurements in both the NaCl and CaCl2 solutions were all performed using the same oocyst.) The error bars on the data points represent the standard deviation.
Figure 8. Force vs separation distance for a 5 µm silica particle interacting with a formalin-treated oocyst at various ionic strengths of aqueous (a) NaCl and (b) CaCl2 solutions. Note that the oocyst used for the NaCl measurements was not the same one as used for the CaCl2 measurements.
Force Profiles Using Heat-Treated C. parWum Oocysts. Heat treatment is another effective method of inactivating viable C. parVum oocysts that is both simple and cost-effective.34,35 Little is known, however, about the effects of such treatments on the oocyst surface itself. Because the glycocalyx layer on the surface of the oocyst has been reported to consist of a polypeptide backbone,9 it can be reasoned that heat treatments may actually alter the surface of the oocyst. To explore this notion further, we performed force measurements between an oocyst that had been heated at 80 °C for 1 h and a 5 µm silica particle in both NaCl and CaCl2 solutions. Shown in Figure 10 are several force profiles obtained with heat-treated oocysts at varying NaCl concentrations. To facilitate the comparisons, the forces have been normalized by the repulsive force at contact, termed F0. Unlike the previous profiles, there is now an obvious dependence on ionic strength. A comparison of the decay lengths measured in these curves, as well as decay (34) Fayer, R. Vet. Parasitol. 2004, 126, 37-56. (35) Fujino, T.; Matsui, T.; Kobayashi, F.; Haruki, K.; Yoshino, Y.; Kajima, J.; Tsuji, M. J. Vet. Med. Sci. 2002, 64, 199-200.
Figure 10. Force vs separation distance for a 5 µm silica particle interacting with a heat-treated oocyst in aquous NaCl solutions at varying ionic strength. The forces have been normalized by the force at contact, F0, to facilitate the comparison.
lengths obtained from similar force profiles measured in CaCl2 solutions, with the calculated Debye lengths is shown in Table 1. As seen, there is now far better agreement, indicating that with heat-treated oocysts the electrostatic contribution to the total interaction force is significant. Adhesion Measurements. In addition to the long-range interaction, our AFM experiments allowed us to determine whether an adhesion force is present between the oocyst and test particle. The presence of an adhesion force is determined by analyzing the force profile upon retraction of the particle after contact with the oocyst, as opposed to the force profile obtained upon approach of the test particle to the oocyst. Adhesion causes a sudden jump of the cantilever to a larger separation during the
Interaction Force between Oocysts and Solid Surfaces
Langmuir, Vol. 23, No. 14, 2007 7481
Table 1. Comparison of Measured Decay Lengths with the Calculated Debye Length for the Interaction between a 5 µm Silica Particle and a Heat-Treated C. parWum Oocyst in Aqueous NaCl and CaCl2 Solutionsa ionic strength, mM
predicted Debye length, nm
measured decay length with NaCl, nm
measured decay length with CaCl2, nm
1.0 3.16 10.0 31.6 100.0
9.84 5.53 3.11 1.75 0.984
10.2 6.8 5.0 5.2 2.6
9.1 5.7 3.7 3.3 n/a
a Note that the decay length for the 100 mM CaCl2 system could not be accurately determined.
Figure 11. Example of a force curve that shows adhesion between the oocyst surface and a colloidal particle upon retraction. This example is from an experiment between a 5 µm silica particle and a formalin-treated oocyst surface.
retraction of the test particle. This can be seen in the force profile shown in Figure 11. Unlike the long-range forces, which were very reproducible in a given system, the magnitude of the adhesion force was found to vary greatly between measurements. In almost no system was an adhesion force seen consistently with every force profile and on every oocyst. Likewise, the magnitude of the force varied greatly as well. To quantify the adhesion effect, we plot in Figure 12 the percentage of retraction curves that showed a measurable adhesion force in each of the systems studied. There are several trends in these graphs that should be mentioned. First, with both NaCl and CaCl2, the pure silica system (i.e., oocyst interacting with a negatively charged silica particle) shows the lowest amount of adhesion. The adhesion with CaCl2 (Figure 12b) is also significantly higher than with NaCl (Figure 12a). Using silane-coated silica particles results in a significant increase in adhesion at all ionic strengths except at the highest ionic strength. When formalin-treated oocysts are used (interacting with untreated silica particles), the adhesion increases substantially in both the NaCl and CaCl2 systems. The highest amount of adhesion occurs with the heat-treated oocysts: in the CaCl2 system, adhesion is observed with 100% of the force curves. Finally, although there seems to be an increase in adhesion with increasing ionic strength in some systems, such a trend is clearly not a general one, and in some cases, the opposite trend is seen
Figure 12. Comparison of the average percent adhesion observed for various ionic strengths of aqueous (a) NaCl and (b) CaCl2 solutions. It should be noted that each set of data represents an average over all force curves. The various interactions represented are the following: a silica particle interacting with an untreated oocyst (silica), a silane-coated silica particle interacting with an untreated oocyst (silane-coated), a silica particle interacting with a formalintreated oocyst (formalin-treated), and a silica particle interacting with a heat-treated oocyst (heat-treated).
(e.g., the adhesion between oocysts and silane-coated silica particles in NaCl solutions).
Discussion Force Profiles. The results from both the silane-coated silica/ oocyst system and the silica/formalin-treated oocyst system show that ionic strength has little effect on the force interaction between the two surfaces. This is similar to what has been seen in our previous work on the interaction between silica and viable, untreated oocyst surfaces24 and is consistent with our hypothesis that the oocyst interaction is dominated by a steric repulsion arising from an outer layer of uncharged carbohydrates that is not affected by the solution ionic strength. To better correlate our current experiments involving silanecoated silica particles with our previous results involving untreated silica particles, we analyzed the force curves using a modified Alexander-de Gennes expression for the interaction between two uncharged polymer brush layers. The modifications essentially involve (1) a Derjaguin correction to account for the particle-particle interaction versus plate-plate interaction and (2) a correction for the fact that only one of the surfaces contains the brush layer. The resulting equation is
7482 Langmuir, Vol. 23, No. 14, 2007
F(h) ) 2π
(
)( ){
Byrd and Walz
}
R1R2 kT 4L L 5/4 4L h 7/4 -1 + -1 3 R1 + R2 s 5 h 7 L for h < L (1)
[( )
]
[( ) ]
Here, F is the interaction force between two particles at separation h, R1 and R2 are the radii of the particles, k is Boltzmann’s constant, and T is the absolute temperature. This equation can be fitted to the measured force profiles to obtain the brush layer thickness, L, and the characteristic distance between adsorbed chains, s. The result of this fitting for the interaction between a silane-coated silica particle and the C. parVum oocyst surface in 10.0 mM ionic strength NaCl and CaCl2 is given in Figure 13. As seen, this brush-layer model is able to fit the force profile relatively well over the entire range of separations. Similar agreement was obtained in fitting the force profiles measured between oocysts and an untreated silica particle. Specifically, in a 10 mM NaCl solution, the s and L values obtained were 50.1 and 113 nm, respectively, whereas in a 10 mM CaCl2 solution, the corresponding values were 50.0 and 113 nm. The thickness of the protein layer obtained in the present experiments using the silane-coated silica particles (150 and 168 nm for 10 mM NaCl and CaCl2 solutions, respectively) is somewhat larger; however, this could be due to the fact that a different batch of oocysts was used here. Independent verification of the thickness of this layer can be obtained from the force profiles obtained upon retraction, such as shown in Figure 11. Specifically, for those curves in which an adhesion force was observed, we determined the separation distance at which the last jump to zero force occurred, which should correspond to the last chain-detachment event. An analysis of 140 such retraction force profiles yielded a final detachment distance of 170 ( 46.8 nm (arithmetic mean ( one standard deviation), which is clearly similar to the values of layer thickness obtained by fitting the Alexander-de Gennes equation shown in Figure 11. (The detachment distance would be expected to be slightly larger than the thickness obtained from the approach force because detachment could involve stretching the layer.) The agreement between these two methodssone based on the force during approach and the other, during retractionssupports the accuracy of our measured layer thicknesses. It should be acknowledged here that the fact that we see only a very weak dependence of the force profiles on solution ionic strength is quite surprising. After all, zeta potentials have repeatedly shown the oocysts to be negatively charged, and the results of Considine et al. showed that the decay length of the force profile decreased upon the addition of calcium. One possible explanation is that at the very large separations at which the electrostatic force becomes the dominant interaction (i.e., beyond the region of overlap of the carbohydrate layers) the force is too small to be measured accurately. The manufacturerreported spring constant of the cantilevers used in these experiments was 0.03 N/m. Although weaker cantilevers were available (e.g., as low as 0.006 N/m) and were in fact tested, we found that performing the experiments was difficult because of the adhesion between the test particle and oocyst. Had we been able to use these weaker cantilevers, it is possible that the very long-range region of the force profile could have been more accurately resolved; however, this is purely speculation. Nonetheless, the results presented here make it clear that over a relatively large range of separation distances (i.e., out to as large as 120 nm) the force profile is dominated by a steric repulsion. When the oocysts were treated with formalin, the force profiles were found to be similar to those obtained with the untreated
Figure 13. Fitting of force profiles measured for the interaction between a silane-coated silica particle and an oocyst surface in 10.0 mM ionic strength solutions of (a) NaCl and (b) CaCl2 with the modified Alexander-de Gennes expression of eq 1. In each graph, the symbols are the measured results, and the line is the fit.
oocysts. However, it is clear from the AFM images that the surface of the oocyst was in fact altered in some way. Formalin is one of the most commonly used cross-linking reagents for biological samples.36 Formalin, like many aldehydes, forms covalent bonds between adjacent amine-containing groups and has been known to cross-link the proteins of cell membranes and chromosomal proteins.37-39 However, formalin has a rapid reversal rate, on the order of weeks, and its cross-linking ability is pH-dependent, with greater cross-linking occurring in solutions at higher pH.36 We suspect that formalin in this case is altering the outermost surface properties of the glycocalyx layer by means of cross-linking. The bulk integrity of the layer, however, remains intact, which explains the fact that the steric repulsive force is essentially unchanged by the treatment. This would also explain the slight deformation of the oocyst surface seen during imaging as well as the increased adhesion force observed upon retraction. It is also possible that the ambient pH condition used during formalin treatment and subsequent storage may have affected the extent of cross-linking. Proteins that are exposed to high temperatures undergo denaturation and generally lose their biological properties.40 In this case, the higher order of the protein molecule is destroyed, but the peptide bonds between the amino acids remain as a result of covalent bonding.40 Thus, heat treatment is an effective means by which viable C. parVum oocysts may be inactivated and made noninfectious to the host. It is theorized that heat treating the oocyst surface denatures the actual polypeptide backbone of the outer glycocalyx layer, creating an oocyst with characteristics similar to those of a charged sphere. In this case, the oocyst would exhibit DLVO behavior with decay lengths similar to the calculated Debye lengths for each ionic strength. (36) Eltoum, I. F. J.; Myers, R. B.; Grizzle, W. E. J. Histotechnol. 2001, 24, 173-190. (37) Campbell, A. A.; Robertson, L. J.; Smith, H. V. Appl. EnViron. Microbiol. 1993, 59, 4361-4362. (38) Fraenkel-Conrat, H.; Olcott, H. S. J. Am. Chem. Soc. 1948, 70, 26732684. (39) Kuykendall, J. R.; Bogdanffy, M. S. Mutat. Res. 1992, 283, 131-136. (40) Madigan, M. T.; Martinko, J. M.; Parker, J. Brock Biology of Microorganisms; 9th ed.; Prentice Hall: Upper Saddle River, NJ, 2000.
Interaction Force between Oocysts and Solid Surfaces
Impact on Particle Deposition. As mentioned earlier, Kuznar and Elimelech22 conducted a series of stagnation point flow (SPF) experiments to study the effect of solution conditions and oocyst treatment on the deposition behavior of oocysts onto a quartz glass surface. Although NaCl appeared to have little effect on deposition, increasing the concentration of CaCl2 increased the deposition rate substantially. It was also observed that with formalin-treated oocysts, higher ionic strengths produced significant increases in the deposition rate. For oocysts that had been treated at 80 °C for 1 h, the authors found that the oocyst deposition rate onto quartz was approximately double that of untreated oocysts, even though the zeta potential remained essentially unchanged.22 Kuznar and Elimelech theorized that by treating the oocyst surface with either formalin or heat, they were able to reduce the steric repulsion imparted by the outer surface proteins. Whereas our results are consistent with some of the findings of Kuznar and Elimelech (e.g., the null effect of NaCl and the significant effect of heat treating the oocysts), there are nonetheless some discrepancies. Specifically, we find that the force profiles are essentially unchanged by either the addition of calcium or the treatment of the oocysts with formalin. By comparison, Kuznar and Elimelech found that each of these effects increased the rate at which the oocysts deposited onto quartz. As described in our previous paper, one possible explanation here is that the deposition is controlled by very weak attractive forces between the outermost layer of the carbohydrate and quartz surface that are not being detected by our AFM force profiles. Some evidence of this is given by the adhesion measurements presented in Figure 12, where we see a slight increase in adhesion in the presence of calcium and a substantial increase when formalin-treated oocysts are used. Although attractive forces between the outermost carbohydrate layers and silica particle would not be observed in the measured approach curves, such as those shown in Figures 3, 4, and 8, they could be detected upon detachment of the oocyst, such as shown in Figure 12.
Conclusions The force profile measurements provide further insights into the nature of the outermost layer of C. parVum oocysts and support our earlier theories concerning the structure of the oocyst outer polysaccharide layer.24 The steric interactions that we have found between a positively charged silane-coated silica particle and a C. parVum oocyst support the notion that the outer oocyst structure is that of a relatively thick layer of uncharged carbohydrate molecules, possibly mixed with a thinner layer of charged protein
Langmuir, Vol. 23, No. 14, 2007 7483
molecules. The nature of the force profile is essentially independent of ionic strength and ion valence out to separations greater than 100 nm. Fitting the force profiles measured upon approach of the particle and oocyst to a modified form of the Alexander-de Gennes equation for the interaction between polymer brush layers yielded an outer layer thickness of 150 to 160 nm. This value was supported by the average distance at which detachment between the particle and oocyst occurred during retraction. Treatment of the oocysts with formalin caused no change in the repulsive force profile upon approach; however, adhesion between the oocyst and silica particle was observed much more often. Heat treatment of the oocysts yielded force profiles that showed a much stronger dependence on ionic strength, suggesting that the outer carbohydrate layer had been substantially altered. In addition, adhesion was observed even more frequently (i.e., 100% of the force profiles obtained in CaCl2 showed adhesion when heat-treated oocysts were used). One of the questions that we sought to address in this study was whether inactivated oocysts can be used as surrogates for viable oocysts in laboratory studies. (Using inactivated oocysts eliminates the health risks associated with viable oocyts.) Emelko performed column experiments using both formalin-treated and viable oocysts and found that the removal rates of both the treated and viable oocysts were essentially the same.41 Although we clearly observe increased adhesion with formalintreated oocysts compared to viable oocysts, the repulsive forces measured upon approach show very similar behavior, which tends to support the findings of Emelko. However, both the frequency of adhesion and the nature of the long-range repulsive forces measured upon approach were clearly different with heat-treated oocysts as compared with viable oocysts. This finding strongly suggests that the heat-inactivated oocysts may not be suitable surrogates, especially in studies on filtration. Acknowledgment. This work was supported by the United States Department of Agriculture (award no. 2002-35102-12600) and by the National Science Foundation, Division of Chemical and Transport Systems (award CTS-0350630). We also gratefully acknowledge the benefit of conversations with Professor Menachem Elimelech in the Department of Chemical Engineering at Yale University. LA0701576 (41) Emelko, M. B. Water Res. 2003, 37, 2998-3008.
