Colloids and Surfaces B: Biointerfaces 34 (2004) 259–263
Adhesion of Cryptosporidium parvum and Giardia lamblia to solid surfaces: the role of surface charge and hydrophobicity X. Dai a , J. Boll b,∗,1 , M.E. Hayes c , D.E. Aston d b
a State Office of Technical Services, Idaho Department of Environmental Quality, Boise, ID 83706, USA Department of Biological and Agricultural Engineering, University of Idaho, Moscow, ID 83844-2060, USA c Holladay Engineering Company, Payette, ID 83661, USA d Department of Chemical Engineering, University of Idaho, Moscow, ID 83844-1021, USA
Accepted 19 December 2003
Abstract Adhesion of Cryptosporidium parvum and Giardia lamblia to four materials of different surface charge and hydrophobicity was investigated. Glass beads were used with and without three polymer coatings: aminosilines (A0750), fluorosilines (T2494), an amino cationic polymer. Surface charge density and hydrophobicity of the beads were characterized by measuring the zeta potential (ZP) and the contact angle, respectively. Adhesion was derived from batch experiments where negatively charged (oo)cysts were mixed with the beads and recovery was determined by counting (oo)cysts remaining in suspension using a flow cytometer. Experimental results clearly show that adhesion to solid surfaces of C. parvum is different from G. lamblia. Adhesion of C. parvum to positively charged, hydrophilic beads (82% recovery relative to control) indicated that surface charge was the more important factor for C. parvum, dominating any hydrophobic effects. Adhesion of G. lamblia cysts to negatively charged, hydrophobic beads (0% recovery relative to control) indicated that although hydrophobicity and surface charge both played a role in the adhesion of G. lamblia to solid surfaces, hydrophobicity was more important than surface charge. © 2004 Elsevier B.V. All rights reserved. Keywords: Contact angle; Hydrophobicity; Zeta potential; Surface charge; Adhesion
1. Introduction Cryptosporidium parvum and Giardia lamblia have been leading causes for waterborne disease outbreaks in North America for the last two decades . Both pathogens host in gastro-intestinal tracks of humans and livestock and are excreted into the environment in encysted forms as oocysts and cysts, respectively (henceforth referred to as (oo)cysts). Removal of (oo)cysts from waste water in treatment plants or from natural water systems primarily relies on the physical processes of sedimentation and filtration. In these systems, the effectiveness of filtration of (oo)cysts depends on
∗ Corresponding author. Tel.: +1-208-885-7324; fax: +1-208-885-8923. E-mail address: [email protected]
(J. Boll). 1 Dr. Boll is on sabbatical leave until 1 July 2004 at the following address: Department of Water Resources, Wageningen University and Research Center, Nieuwe Kanaal 11 6709, PA, Wageningen, The Netherlands. Tel.: +31-317-482778; fax: +31-317-484885.
0927-7765/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2003.12.016
adhesion properties of C. parvum and G. lamblia to solid surfaces of the filter materials when there is no appreciable size exclusion sustained by the filter matrix. Surface charge and hydrophobicity are important characteristics of (oo)cysts for the understanding of adhesion [2–4]. These characteristics, in turn, depend on the biochemistry of the (oo)cyst’s wall, i.e., the distribution and function of lipids, proteins, and carbohydrates [5,6]. Surface charge and hydrophobicity can depend on pH, ionic strength [2,7], purification, or treatment method, and age [4,8]. Under different environmental conditions, the governing forces of interaction between (oo)cysts and adhesion media change between electrostatic and hydrophobic . Microbial adhesion to inert substratum surfaces (and to other microbial cell surfaces) originates from three fundamental forces, the Lifshitz–van der Waals, electrostatic, and acid–base interactions [9,10]. In a review of physico-chemical approaches used in modeling microbial adhesive interactions, Bos et al.  describe how the addition of the Lewis acid–base interaction in the extended
X. Dai et al. / Colloids and Surfaces B: Biointerfaces 34 (2004) 259–263
Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory to the Lifshitz–van der Waals and electrostatic double-layer interactions in the classical DLVO theory implies that hydrophobic attraction and hydrophillic forces may be accounted for in colloid and surface science more formally. The Lifshitz–van der Waals attractive force is short-ranged and usually too weak to consider compared with the electrostatic double-layer force in actual interacting space. Particles of like charges tend to repel each other while oppositely charged ones adhere. Colloids that cannot form hydrogen bonds tend to be attracted to hydrophobic surfaces [10,11]. Non-DLVO factors such as steric interactions  also influence (oo)cyst interaction with solid surfaces. Several studies reported that purified (oo)cysts are negatively charged under ambient environmental conditions [4,13,14]. Oocysts can be electrically neutral when washed in a phosphate buffer solution  or following continuous-flow flotation during purification . However, using the same continuous-flow purification method as referred to in Brush et al. , we found negatively charged oocysts in our laboratory . The isoelectric points of cysts and oocysts have been estimated at pH 2.2 and 3.3, respectively , although slight variations have been reported by others. Above the iso-electric point, (oo)cysts are negatively charged in water. Relatively little quantitative information is available on the hydrophobicity of (oo)cysts. Drozd and Schwartzbrod  showed weak hydrophobicity of C. parvum whereas others [2,4] found significant hydrophobicity of oocysts as measured by adhesion to polystyrene surfaces (60–80% adsorption) and microbial adherence to hydrocarbons (MATH). Hsu and Huang  also demonstrated strong hydrophobicity of G. lamblia. Capizzi-Banas et al.  reported that G. lamblia had a hydrophobic character. In our previous study on (oo)cyst interaction with soil particles , recovery rates in batch experiments were consistently lower for G. lamblia ( aminosiloxane > cationic polymer > control. We expected a lower contact angle for aminosiloxane polymer, because it has a polar and somewhat reactive surface group. The higher measured contact angle is possibly due to binding of the –NH2 group with other hydrophilic surface groups such as –OH of the glass, exposing a more hydrophobic surface than that of the amine segments. Solid surfaces that are not easily wetted by water form liquid beads when water drops contact the surface and therefore exhibit large contact angles (θ). These surfaces are classified as hydrophobic without a standardized reference. Researchers have different criteria to classify hydrophobicity or hydrophilicity of a material . A commonly used cutoff value is θ = 90◦ . When θ > 90◦ , the solid surface is named hydrophobic or “non-wetting” by water. For solids resulting in 0◦ < θ < 90◦ , the solid surface is said to be “wetted” or “partially wetted.” When θ = 0◦ , the solid surface is said to be “wet out” by the water. Table 1 shows that the fluorosiloxane polymer is definitely hydrophobic, and the measured θ agrees with referenced θ = 108◦ by Walford et al.  for Teflon (PFTE), the equivalent chemical composition to cross-linked T2494. Glass is hydrophilic and so are the cationic polymer and the aminosiloxane polymer, being partially wettable with increasing contact angle. 3.2. Effect of surface properties on adhesion of C. parvum Surface charge was the more important factor in the adhesion process of C. parvum to solid surfaces. Table 2 lists recovery percentages of the (oo)cysts relative to the control tube in mixing experiments with four solids. C. parvum
Table 2 Recovered %a of C. parvum oocysts and G. lamblia cysts from adhesion experiments Surface
θ ± S.E. (◦ ) 21 ◦ C, 24 h curing 28 70 95 47
± ± ± ±
1 4 2 2
115 ◦ C, 15 min curing 29 70 90 39
±1 ±2 +0 ±1
√ S.E. is standard error, S.E. = S.D./ n, where S.D. is standard deviation and n is the number of replicates for each sample.
Control (styrene) Glass beads Amino-coating Fluoro-coating Cationic-coating
C. parvum oocysts
G. lamblia cysts
Recovered ± S.E. (%)
Recovered ± S.E. (%)
2.6 7.5 5.1 6.7 8.0
100 111 116 0 66
100 100 95 100 82
± ± ± ± ±
3 8 5 7 7
±6 ±9 ± 14 ±6
CV (%) 5.7 7.7 12.0 0 9.1
S.E. is standard error as in Table 1, CV is coefficient of variation. a Percentages were calculated relative to the recovered amount of (oo)cysts in control.
X. Dai et al. / Colloids and Surfaces B: Biointerfaces 34 (2004) 259–263
oocysts were recovered 100% after mixing with glass beads (negative, hydrophilic) and beads coated with fluorosiloxane (negative, hydrophobic). Recovered oocyst amount was slightly (5%) less than that in the control tube in experiments using aminosiloxane (positive, partially wetted). For mixing experiments using cationic polymer coated glass beads (highly positive, partially wetted), 82% of oocysts were recovered, the largest decrease compared to other materials. In a separate experiment using the polystyrene plate method by Brush et al.  and counting oocysts with flow cytometer, we recovered 35% of oocysts (relative to total amount applied). Polystyrene has a θ = 91◦  and is suspected of participating in hydrophobic interactions. In other words, despite the strong hydrophobic character of polystyrene, we were still able to recover oocysts. 3.3. Effect of surface properties on adhesion of G. lamblia Recovery rates in Table 2 shows a mixed adhesion pattern for G. lamblia. Compared to the control tube, more than 100% of the cysts were recovered from mixing experiments with glass beads and beads coated with aminosiloxane, indicating no cysts were lost due to interaction with non-hydrophobic, neutral or moderately positively charged materials. Over-detection of cysts probably was due to measurement variability as indicated by standard error and the coefficient of variation, which were the highest among the experiments. Considerable loss of cysts occurred in the mixing experiments using fluoro-coated and cationic polymer coated glass beads. Table 1 shows the fluorosiloxane to be the most hydrophobic and moderately negatively charged and the cationic polymer to be hydrophilic and the most strongly positively charged. Cysts recovery after mixing with the hydrophobic beads (fluoro-coated) was 0%, and after mixing with cationic beads 66% was recovered. In a separate experiment using the polystyrene plate method by Brush et al.  and counting oocysts with flow cytometer, we recovered 21% of cysts (relative to total amount applied). These data suggest that hydrophobicity is more important than surface charge for adhesion of G. lamblia to solid surfaces. 3.4. Evaluation of results Interestingly, only 76% of C. parvum and 36% of G. lamblia were recovered from our control tube experiments, similar to the low recovery results in Dai and Boll . The non-complete recovery indicates some (oo)cyst losses in the experimental process. In our studies, the centrifuge tubes used for mixing were made of polystyrene, which is hydrophobic. The higher recovery of C. parvum compared to G. lamblia again indicates that hydrophobicity is more influential on adhesion of G. lamblia than C. parvum to solid surfaces.
The main results of this paper are that adhesion of C. parvum is governed mostly by surface charge, and that G. lamblia showed a much stronger hydrophobic behavior than C. parvum. The difference in the role of surface charge and hydrophobicity in (oo)cyst–solid adhesion cannot be explained in the absence of better knowledge of the physico-chemical composition of (oo)cyst walls and characterization of the interacting energies involved in adhesion . Progress in this area has been made primarily for C. parvum oocysts [12,19]. Tilley and Upton  reported that the proteineous material on the outside wall is rich in cysteine, proline and histindine based on gel electrophoresis analysis. Four major proteins have been identified in the outer cyst wall and the sugar component of the outer portion of the cyst is predominantly galactosamine . C. parvum oocysts in our study were approximately 25% more negatively charged than G. lamblia cysts. Hsu and Huang  found that C. parvum oocysts were twice as negatively charged as G. lamblia cysts. This difference in surface charge partly explains the differences between (oo)cyst–solid interactions given that electrostatic forces between two charged particles is proportional to the product of charge densities of the interacting surfaces. For G. lamblia cysts, non-charge based forces, such as a hydrophobic effect, may overrule the influence of a weaker electrostatic force and become dominant in cyst–solid adhesion. Another difference between C. parvum oocysts and G. lamblia is their shape. C. parvum oocysts are nearly spherical, while the average eccentricity (long radius/short radius) of G. lamblia cysts is 1.5 . This difference may affect (oo)cyst-adhesion. For example, G. lamblia cysts attached to a hydrophobic surface on longer sides form stronger association than on the shorter sides. However, quantitative investigation of this is beyond the scope of this study. The adhesion process can be different in environmental water samples where pH and ionic strength are different , or depending on other constituents of the aqueous solution . Since transport and removal of (oo)cysts occur mostly in the water environment, however, our results are representative of most environmental conditions. If our findings hold in the natural setting, filtration efficiencies of oocysts or cysts will change depending on surface properties of the filtering material [3,7]. In the design of filter materials for (oo)cyst removal, surface charge and hydrophobicity should be considered differently for oocysts and cysts, respectively. Furthermore, hydrophobically-driven adhesion is reversible through dilution or washing, whereas surface charge-moderated adhesion is irreversible with respect to the same. Practically then, a hydrophobic filter bed may be recycled with little economic or time loss, whereas chemical addition or electrical processes are necessary to desorb electrostatic adsorbates. Our findings on adhesion of (oo)cysts to solid surfaces are also important for a better understanding of transport and fate of the organisms in natural water
X. Dai et al. / Colloids and Surfaces B: Biointerfaces 34 (2004) 259–263
and for optimization of removal processes in natural systems (e.g., filter strips, wetlands).
4. Conclusions In this study, we found surface charge and hydrophobicity influence the adhesion of (oo)cysts to solid surfaces. The role of surface charge and hydrophobicity, however, appear very different for the two organisms. Surface charge was the most important factor for C. parvum, dominating any hydrophobic effects. Hydrophobicity and surface charge both played a role in the adhesion of G. lamblia to solid surfaces, but hydrophobicity was more important than surface charge.
Acknowledgements This work was supported partially by NSF/EPSCoR program (EPS-9720634) at the University of Idaho. We thank Dr. Jeff Boyle at Washington State University for the use of the Zetasizer. We thank Elizabeth Scherling for editorial comments to this paper.
Appendix A. Chemical component and formula of the polymers Aminosiloxane (A0750): 3-aminopropyltriethoxysilane H2 N(CH2 )3 Si(OCH2 CH3 )3 Fluorosiloxane (T2494 or PTFE): tridecafluoro-1,1,2,2tetrahydrooctyl-1-triethoxysilane C6 F13 CH2 CH2 Si(OCH2 CH3 )
References  R.L. Guerrant, Emerg. Infect. Dis. 3 (1997) 51.  B.M. Hsu, C.P. Huang, Colloids Surf. A: Physicochem. Eng. Aspects 201 (2002) 201.  R.M. Hozalski, X.J. Dai, Water Res. 36 (2002) 3523.  C.F. Brush, M.F. Walter, L.J. Anguish, W.C. Ghiorse, Appl. Environ. Microbiol. 64 (1998) 4439.  M. Tilley, S.J. Upton, in: R. Fayer (Ed.), Cryptosporidium and Cryptosporidiosis, CRC Press, Boca Raton, FL, 1997, pp. 163– 180.  M.E. Karaman, R.M. Pashley, H. Bustamante, S.R. Shanker, Colloids Surf. A: Physicochem. Eng. Aspects 146 (1999) 217–225.  B.M. Hsu, C.P. Huang, J.R. Pan, Water Res. 35 (2001) 3777.  M.A. Butkus, J.T. Bayes, M.P. Labare, Appl. Environ. Microbiol. 69 (2003) 3819.  R. Bos, H.C. van der Mei, H.J. Busscher, FEMS Microbiol. Rev. 23 (1999) 179.  C.J. van Oss, Colloids Surf. B: Biointerfaces 5 (1995) 91.  D. Grasso, K. Subramaniam, M.A. Butkus, J. Bergendahl, K. Strevett, Rev. Environ. Sci. Bio/Technol. 1 (2002) 17.  R.F. Considine, D.R. Dixon, C.J. Drummond, Langmuir 16 (2000) 1323.  C. Drozd, J. Schwartzbrod, Appl. Environ. Microbiol. 62 (1996) 1227.  J.E. Ongerth, J.P. Pecoraro, J. Environ. Eng. 122 (1996) 228.  X. Dai, J. Boll, J. Environ. Qual. 32 (2003) 296.  S. Capizzi-Banas, M. Maux, J. Schwartzbrod, Helminthologia 39 (2002) 197.  A.W. Adamson, A.P. Gast, in: Physical Chemistry of Surfaces, John Wiley & Sons Inc., 1997.  A.J. Walford, M. Mullay, R. Schlicke, in: Guide to Reference Material, vol. 1, Library Association Publishing Ltd., London, 1993.  J. Nanduri, S. Williams, T. Aji, T.P. Flanigan, Infect. Immun. 67 (1999) 2022.  R.D. Adam, Clin. Microbiol. Rev. 14 (2001) 447.  X. Dai, in: Transport Kinetics of Cryptosporidium parvum and Giardia lamblia in overland flow, Unpublished Ph.D. Dissertation, University of Idaho, Moscow, ID, Chapter 3.