- Open Access
The role of casein in supporting the operation of surface bound kinesin
© Verma et al; licensee BioMed Central Ltd. 2008
- Received: 23 October 2007
- Accepted: 20 October 2008
- Published: 20 October 2008
Microtubules and associated motor proteins such as kinesin are envisioned for applications such as bioseparation and molecular sorting to powering hybrid synthetic mechanical devices. One of the challenges in realizing such systems is retaining motor functionality on device surfaces. Kinesin motors adsorbed onto glass surfaces lose their functionality or ability to interact with microtubules if not adsorbed with other supporting proteins. Casein, a milk protein, is commonly used in microtubule motility assays to preserve kinesin functionality. However, the mechanism responsible for this preservation of motor function is unknown. To study casein and kinesin interaction, a series of microtubule motility assays were performed where whole milk casein, or its αs1 and αs2, β or κ subunits, were introduced or omitted at various steps of the motility assay. In addition, a series of epifluorescence and total internal reflection microscopy (TIRF) experiments were conducted where fluorescently labeled casein was introduced at various steps of the motility assay to assess casein-casein and casein-glass binding dynamics. From these experiments it is concluded that casein forms a bi-layer which supports the operation of kinesin. The first tightly bound layer of casein mainly performs the function of anchoring the kinesin while the second more loosely bound layer of casein positions the head domain of the kinesin to more optimally interact with microtubules. Studies on individual casein subunits indicate that β casein was most effective in supporting kinesin functionality while κ casein was found to be least effective.
- Glass Surface
- Motility Assay
- Total Internal Reflection Fluorescence
- Microtubule Binding
- Kinesin Motor
Biological molecular motors are a unique class of proteins, which exist in eukaryotic cells and function as nano scale vehicles that drive a range of fundamental biological processes. Kinesin motor proteins transport intracellular cargo, move proteins and mRNA in neurons, and play a vital role in cell division. These motor proteins move unidirectionally on protein tracks known as microtubules which assemble and disassemble creating a unique dynamically reconfigurable transportation system. Kinesins are powered by the hydrolysis of adenosine triphosphate (ATP) and exert a maximal force of ~6 pN and exhibit a maximal efficiency of ~50% [1–3]. Biological motors are the subject of intense research in part because of their potential to be used in-vitro as 'nano-engines' for several future applications ranging from bioseparation to powering hybrid micro and nano scale electromechanical systems (MEMS/NEMS) [4–8]. Surface interactions and adsorption of molecular motors have also been studied [9, 10].
An important hurdle that must be overcome before realizing these applications is developing approaches for preserving motor protein function when integrated into engineered devices and surfaces. Motor protein function is studied using gliding assays or bead assays [11, 12]. In the bead assay, microtubules are immobilized on a surface and biological motors attached to microscale beads move along these immobilized microtubules. In the gliding assay, the motor proteins are immobilized on a surface and microtubules are transported across the surface by the immobilized motors. In either case, these motor proteins are attached to a surface in a manner that is intended to preserve their functionality. This is typically accomplished via the use of blocker proteins that prevent motor denaturation on the untreated surface [11, 13]. We studied microtubule motility through gliding assay.
The most common proteins used for creating kinesin compatible surface interfaces are caseins extracted from bovine milk [12–14]. These proteins have been extensively used in the surface immobilization of motor proteins such as kinesin, and they are also used in Western Blots for blocking nonspecific adsorption of antibodies to nitrocellulose membranes. A typical process for performing a gliding assay consists of: 1) constructing a flow cell using a glass slide (Fisher Finest Premium microscope slides) and glass cover slip (Corning 1 1/2, 18 mm2); 2) blocking the glass surfaces by flowing in a solution of 0.5 mg/ml casein protein; 3) adsorbing kinesin to the glass surface by flowing in a solution of kinesin containing ATP and 0.2 mg/ml casein; 4) flowing in a microtubule solution containing ATP, antifade reagents and casein, and 5) observing microtubule movements by epi-fluorescence microscopy. Microtubule motility is quantified by examining several parameters including the density of microtubules observed, their landing rate on the surface, the microtubule transport velocity, and the distances microtubules move before detaching from the surface and diffusing away.
The initial blocking step is implemented to form a layer of the protein casein on the surface to prevent the kinesin denaturation that would occur if the motors proteins were directly adsorbed to the glass surface . Casein is included in the motor and microtubule solutions based on a qualitative observation that it improves microtubule motility. However, although casein is used to optimize motor activity in a number of different in vitro motility assays on surfaces, no studies have been conducted to understand the mechanism by which casein creates compatible interfaces. Understanding the role of casein in supporting biomotor function is important for developing design rules for engineering optimal surfaces for use in advanced hybrid devices that incorporate biological motor proteins for actuation and transport.
Caseins are phosphoproteins that bind calcium and form large aggregates in milk. Casein subunits range from 20–30 kDa and are classified into 4 types: αs1, αs2, β, and κ. The relative concentrations of these caseins vary depends upon the milk producing species. Casein from bovine milk, which is commonly used in motor protein experiments, contains the following subunit compositions in skim milk (mg/ml): αs1 (12%–15%), αs2 (3%–4%), β (9%–11%), and κ (2%–4%) [15, 16]. The different forms of casein have been extensively studied due to their importance in milk, as food additives and as emulsifiers and stabilizers for glue, paint, and other materials . Although there are no crystal structures of any of the subunits, every one of the subunits contains one or more clearly defined hydrophobic and hydrophilic domains, resulting in the formation of micelles with differing geometries [18–21].
In the present work, we have examined the role of both whole casein (consisting of mixture of αs1, αs2, β, and κ casein) and the various casein subunits in modulating kinesin function in the microtubule gliding assay. In particular, microtubule motility assays were performed where casein was either included or excluded in the surface blocking, kinesin adsorption, and microtubule solutions. The inclusion of casein in various incubation steps and the specific casein subunits used had dramatic impacts on the functionality of surface adsorbed kinesin, as measured by counting the number of microtubules on the surface. Most notably, samples including κ casein exhibited significant reductions in the number of microtubules observed. Studies on fluorescently labeled casein were also performed where the labeled casein was introduced into the flow-cell surface either as a first or after unlabeled casein had been adsorbed. By measuring the binding of fluorescent casein to the surface, we hypothesize that casein forms a bi-layer on the glass surface and that optimum kinesin-microtubule binding is achieved when kinesin proteins co-assemble with the casein bi-layer.
Casein was prepared by dissolving the protein powder (Sigma C-7078, St Louis, MO) in BRB80 buffer (80 mM PIPES, 1 mM MgCl2, 1 mM EGTA, pH 6.9), centrifuged to remove un-dissolved casein. The solution was then filtered using a 220 nm filter (Fisher brand, Ireland). Full length Drosophila conventional kinesin heavy chain and lacking light chains was bacterially expressed and purified according to established procedures . Tubulin was prepared from bovine brain and rhodamine-labeled according to standard procedures [25, 26]. Microtubules were polymerized by incubating a mixture of 32 μM tubulin, 4 mM MgCl2, 1 mM GTP, and 5% DMSO in BRB80 buffer at 37°C for 20 minutes, and were stabilized by diluting to a final concentration of 0.32 μM in a solution containing 10 μM paclitaxel. Kinesin was labeled with Rhodamine-NHS using the standard protocol for tubulin labeling  to examine relative density of kinesin absorbing to the surface in the presence of different casein.
All the motility experiments were performed in a closed volume flow cell as previously described [13, 27]. Flow cells consisted of a microscope slide (Fisher Finest Premium microscope slides) and a glass cover slip (Corning 1 1/2, 18 mm2) separated by double side tape, and had volumes of ~20 μl. The protein solutions were exchanged in these flow cells and motility of microtubules was assessed. In a flow cell, solutions were exchanged by injecting fresh buffer from one end and wicking out the previous solution from the other. In control experiments the first step is surface blocking where 0.5 mg/ml casein is incubated in the flow cell for 10 minutes followed by a solution of kinesin motors (0.05–5 μg/ml) including 0.2 mg/ml casein and 100 μM Mg-ATP in BRB80 buffer. Finally, microtubule motility solution containing 0.032 μM microtubules, 10 μM paclitaxel, 0.2 mg/ml casein, 1 mM Mg-ATP, 20 mM D-glucose, 20 μg/ml glucose oxidase, 8 μg/ml catalase and 0.5% β-merceptoethanol in BRB80 buffer is introduced. Microtubules were observed by epifluorescence microscopy (Nikon E600, 100×, 1.3 N. A. oil immersion Plan Fluor objective) and captured to videotape using a Genwac GW-902H camera. Microtubule surface binding was quantified by counting the number of microtubules present in a given video screen (65 μm × 48 μm) after 15 minute incubation time. Microtubules with length longer than 1 μm and moving for more than 2 μm were counted. Immobilized microtubules or those smaller than 1 μm were not counted. In all cases microtubule binding refers to microtubule binding to and moved by kinesin.
Epifluorescence and total internal reflection microscopy (TIRF) using a Nikon TE2000 inverted microscope (60×, 1.45 NA, CFI Plan Apo TIRF oil objective) was used to measure fluorescent intensities to study bi-layer formation of casein, and co-assembly between kinesin and casein proteins. Meta-Vue (Universal Imaging, PA) software was used to acquire images for bi-layer and co-assembly studies using rhodamine casein and rhodamine kinesin. Fluorescent intensities from the surface from the recorded images were measured in arbitrary units.
To assess the role that casein is playing in supporting the activity of surface-adsorbed kinesin, we first examined kinesin function under a range of casein conditions and then quantified the binding of both casein and kinesin motors to the surfaces using fluorescence assays. While surface-adsorbed motors are difficult to visualize microscopically, it is relatively straightforward to visualize and quantify the interaction of microtubules with these motors. It has been shown previously that because individual kinesin motors are sufficient to bind microtubules and transport them across the surface, below maximal motor densities, the rate that microtubules land and move over the surface is proportional to the density of active kinesins on the surface [11, 24]. Hence, the surface density of functional motors can be estimated by allowing microtubules to land for a defined time and then counting the number of microtubules on the surface. We systematically removed casein from either the blocking solution, the motor solution, or the microtubule solution and counted the average number of microtubules propagating over a kinesin functionalized glass surface after 15 minutes. Whole bovine milk casein, containing all four casein subunits, was used, and both high (8 μg/ml) and low (0.8 μg/ml) concentrations of kinesin protein were examined. It has been shown previously that at low kinesin concentrations, very few or no motors are functionally adsorbed to the surface if the surfaces are not pretreated with a blocking protein, but at high motor concentrations some fraction of the motors presumably bind and denature on the surface and replace the role of blocking proteins . However, to date, the role of casein in the motor and microtubule solutions has not been systematically investigated.
Measured average number of microtubules observed under different casein and kinesin concentration conditions.
Casein in Kinesin
Casein in MT
Average MT per screen (N = 5)
Low (0.8 μg/ml)
0 ± 0
0 ± 0
1.4 ± 1.3
4.4 ± 1.5
High (8 μg/ml)
0 ± 0
0 ± 0
9.4 ± 1.1
60 ± 4.4
Low (0.8 μg/ml)
0 ± 0
0 ± 0
3.6 ± 1.5
9.2 ± 2.4
High (8 μg/ml)
13.6 ± 2.4
48 ± 5.7
20 ± 3
80 ± 9.6
The observed variation in the number of microtubules present on the surface under differing experimental conditions must then result from the impact of casein on either the availability of surface adsorbed kinesin to participate in microtubule binding, or the density of kinesin adsorbed. From table 1 it can be concluded that surface adsorbed kinesin retain their functionality except in the case of low kinesin concentration and where no casein has been included in any incubation solution. The loss of kinesin functionality can either be due to denaturing of the kinesin or binding of the kinesin head domain to the surface, which may result in no observed microtubule binding. Hence, the inclusion of casein is not impacting the ability of the kinesin to operate normally on an available microtubule if contact with a microtubule was possible. This hypothesis is based on the observation of increased observed microtubule density when casein is added to the final microtubule solution. Moreover, since no additional kinesin had been added in the microtubule solution with casein, the density of kinesin on the surface could not have been increased, but rather more of the existing kinesin on the surface were available to participate in microtubule binding when casein was included in the microtubule solution. It is also apparent that once kinesin are adsorbed in the absence of casein, they cannot be revived by subsequent inclusion of casein in the microtubule solution. This may result from kinesin having a stronger affinity for the surface than casein or that, in the case of kinesin denaturing on the glass surface, that denatured kinesin cannot regain their functionality if casein is subsequently introduced and is able to displace the kinesin.
In the model illustrated in figure 2, the casein is assumed to have one distinct hydrophilic domain and one distinct hydrophobic domain, as is true for β-casein . However, the casein used for the experiments summarized in table 1 included casein from bovine milk which consists of αs1(12%–15%), αs2(3%–4%), β (9%–11%), and κ (2%–4%) caseins. The structure of these caseins have been determined  and are not identical. One of the most pronounced differences is in the structure of the two α caseins, which have two distinct hydrophobic domains.
The differences observed in the function of the different casein subunits presumably derive from the different structures of the various casein subunits. Because the α caseins have two hydrophobic domains, the packing and order of its bi-layer should quantitatively differ from the structure of the bi-layer from β casein, which contains one distinct hydrophobic domain in addition to its hydrophilic domain. Moreover, κ casein contains only one hydrophilic and hydrophobic domain like β casein, but the average number of microtubules observed for assays containing κ casein was nearly an order of magnitude lower than that observed for assays containing β casein (figure 3). However, κ casein is thought to play a role in casein micelle assembly [19–21] that is distinct from the role of α and β subunits. Interestingly, in virtually all microtubule motility studies no binding of microtubule whose length less than 2–3 μm was observed in case of κ casein.
The impact of the presence of both typical bovine milk casein and purified αs1 and αs2 (combined),β and κ caseins on the binding of microtubules on casein-kinesin coated glass surfaces has been examined. Microtubule motility assays were performed where casein was either included or excluded in the surface blocking, kinesin adsorption and microtubule solutions. It is found that the inclusion of casein at the various incubation steps has a dramatic impact on the number of microtubules observed binding to kinesin on the surface. No microtubules are observed on the surface when casein is not included in the kinesin adsorption solution and the inclusion of casein in the microtubule solution always increases the observed number of microtubules. Moreover, there were quantitative differences in the number of microtubules observed when purified forms of the different caseins were used. Most notably, samples including κ casein exhibited significant reductions in the number of microtubules observed. Studies on fluorescently labeled casein were also performed where the labeled casein was introduced to the surface either first or after an unlabeled casein had been adsorbed. Fluorescence intensities were notably higher in the case where the labeled casein was introduced to the surface first, except in the case of α and κ casein.
From this data a bi-layer casein adsorption model is proposed where the first tightly bound layer of casein mainly performs the function of anchoring the kinesin while the second more loosely bound layer of casein positions the head domain of the kinesin to more optimally interact with microtubules. Studies on adsorption of fluorescently labeled caseins in the first and second layers support the bi-layer model of casein assembly. Results using whole casein from bovine milk are also compared to that obtained when using purified αs1 and αs2 (combined),β and κ caseins and significant variations in observed number of microtubules are found. β casein is found to be the best at promoting kinesin-microtubule interaction while κ casein is found to be the worst, and may behave completely different than the other caseins including the lack of a bi-layer assembly.
Kinesin and microtubule system is studied extensively to design hybrid synthetic devices for detection and transport. Motility assays are also used in studying the role of kinesin, in vitro, which may help get an insight on how to control cell division. Some of the assays, however, require more specific engineered surfaces than others and it is important to understand the mechanism which supports operation of surface bound kinesin. In this paper, we propose a bi-layer model based on our studies. However, more studies are needed to understand the interaction of kinesin with the various caseins. A more detailed understanding may provide new insights enabling the design of engineered surfaces for optimally supporting kinesin activity, a key parameter in the development of hybrid biological-synthetic devices incorporating biological molecular motors.
This work was supported by The Pennsylvania State University Center for Nanoscale Science, a NSF Materials Research Science and Engineering Center (DMR0213623). It was also supported by the National Nanotechnology Infrastructure Network (NSF Cooperative Agreement No. 0335765 with Cornell University) and The Pennsylvania State University Materials Research Institute. Vivek Verma wishes to thank the Haythornthwaite Foundation for their Founder's Prize and grant for year 2005–06.
We thank Maruti Uppalapati for purifying kinesin and Gayatri Muthukrishnan for purifying and labeling tubulin.
- Coy DL, Wagenbach M, Howard J: J Biol Chem. 1999, 274: 3667.View ArticleGoogle Scholar
- Meyhofer E, Howard J: Proc Natl Acad Sci USA. 1995, 92: 574.View ArticleGoogle Scholar
- Svoboda K, et al.: Nature. 1993, 365: 721.View ArticleGoogle Scholar
- Mavroidis C, Dubey A, Yarmush ML: Annual Review of Biomedical Engineering. 2004, 6: 363.View ArticleGoogle Scholar
- Hancock WO: Nanodevices for the Life Sciences. Edited by: Kumar CSSR. Wiley-VCH, Weinheim, Germany; 2006:241.Google Scholar
- Hess H, Bachand GD, Vogel H: Chemistry – A European Journal. 2004, 10: 2110.View ArticleGoogle Scholar
- Hess H, Vogel V: Reviews in Molecular Biotechnology. 2001, 82: 67.View ArticleGoogle Scholar
- Schmidt JJ, Montemagno CD: Annual Review of Materials Research. 2004, 34: 315.View ArticleGoogle Scholar
- Katira P, et al.: Advanced Materials. 2007, 19: 3171.View ArticleGoogle Scholar
- Mansson A, et al.: Frontiers in bioscience. 2008, 13: 5732.View ArticleGoogle Scholar
- Howard J, Hudspeth AJ, Vale RD: Nature. 1989, 342: 154.View ArticleGoogle Scholar
- Block SM, Goldstein LS, Schnapp BJ: Nature. 1990, 348: 348.View ArticleGoogle Scholar
- Howard J, Hunt AJ, Baek S: Methods Cell Biol. 1993, 39: 137.View ArticleGoogle Scholar
- Romberg L, Vale RD: Nature. 1993, 361: 168.View ArticleGoogle Scholar
- Eigel WN, et al.: J Dairy Sci. 1984, 67: 1599.View ArticleGoogle Scholar
- Modler HW: J Dairy Sci. 1985, 68: 2195.View ArticleGoogle Scholar
- Plank J: Applied Microbiology and Biotechnology. 2004, 66: 1.View ArticleMathSciNetGoogle Scholar
- Ginger MR, Grigor MR: Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology. 1999, 124: 133.View ArticleGoogle Scholar
- McMahon DJ, McManus WR: J Dairy Sci. 1998, 81: 2985.View ArticleGoogle Scholar
- Ruettimann KW, Ladisch MR: Enzyme and Microbial Technology. 1987, 9: 578.View ArticleGoogle Scholar
- Phadungath C, Songklanakarin : J Sci Technol. 2005, 27: 201.Google Scholar
- Nylander T, Wahlgren NM: Langmuir. 1997, 13: 6219.View ArticleGoogle Scholar
- Kullm T, et al.: Langmuir. 1997, 13: 5141.View ArticleGoogle Scholar
- Hancock WO, Howard J: J Cell Biol. 1998, 140: 1395.View ArticleGoogle Scholar
- Hyman AA: J Cell Sci Suppl. 1991, 14: 125.View ArticleMathSciNetGoogle Scholar
- Williams RC Jr, Lee JC: Methods Enzymol. 1982,85(Pt B):376.View ArticleGoogle Scholar
- Hunt AJ, Howard J: Proc Natl Acad Sci USA. 1993, 90: 11653.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.