|Molecular Vision 1999;
Received 20 May 1999 | Accepted 15 October 1999 | Published 29 October 1999
A heterologous expression system for bovine lens transmembrane Main Intrinsic Protein (MIP) in Nicotiana tabacum plants
Oliver S. de Peyer,1
Andrew C. Wetten,2
James M. Dunwell,2
M. James C. Crabbe1
1Division of Cell and Molecular Biology, School of Animal and Microbial Sciences and 2School of Plant Sciences, The University of Reading, Whiteknights, Reading, Berkshire RG6 6AJ, UK
Correspondence to: M. James C. Crabbe, PhD, Division of Cell and Molecular Biology, School of Animal and Microbial Sciences, The University of Reading, P. O. Box 228, Whiteknights, Reading, Berkshire RG6 6AJ, UK; Phone: +44 (0)1189 318894; FAX: +44 (0)1189 318894; email: firstname.lastname@example.org
We have developed a heterologous expression system for transmembrane lens main intrinsic protein (MIP) in Nicotiana tabacum plant tissue. A native bovine MIP26 amplicon was subcloned into an expression cassette under the control of a constitutive Cauliflower Mosaic Virus promoter, also containing a neomycin phosphotransferase operon. This cassette was transformed into Agrobacterium tumefaciens by triparental mating and used to infect plant tissue grown in culture. Recombinant plants were selected by their ability to grow and root on kanamycin-containing media. The presence of MIP in the plant tissues was confirmed by PCR, RT-PCR and immunohistochemistry. A number of benefits of this system for the study of MIP will be discussed, and also its application as a tool for the study of heterologously expressed proteins in general.
Alterations to the structure and function of the vertebrate lens transmembrane 26 Kd main intrinsic protein (MIP) have been implicated in the development of cataract, the leading cause of blindness in the world human population. MIP accounts for over 60% of lens fiber cell membrane intrinsic protein. Hypotheses for its function have included gap junction, adhesin-like and channel roles, and as a nexus for interaction with other lens proteins such as crystallins. Moreover, MIP is a representative of an important family of membrane proteins, which includes the aquaporins, and which has been largely conserved throughout evolution .
However, a firm elucidation of MIP function remains recalcitrant. MIP is only found in the final stages of lens fiber cell differentiation. Mature fiber cells, which make up the great bulk of the lens no longer contain organelles and have no aerobic respirative or genetic activity. Therefore, they cannot be cultured in vitro and so MIP cannot be studied in its main cellular environment. Therefore, a variety of heterologous expression systems have been developed for the study of MIP. To date, MIP has been expressed in E. coli , Xenopus oocytes , Spodoptera frugiperda (SF) cells [4,5], and recently mammalian erythroid leukemia (melC) cell lines (Drake and Crabbe, unpublished).
However, a number of characteristics of the heterologously expressed MIP differ between the different systems: MIP is non-glycosylated in E. coli and Xenopus oocytes, but significantly glycosylated in SF cells. MIP may not be phosphorylated in SF21 cells (de Peyer and Crabbe, unpublished), even though phosphorylation of MIP is known to be an important regulatory mechanism in vivo. MIP appears to be a weak glycerol channel in Xenopus oocytes , but not in E. coli (de Peyer and Crabbe, unpublished). E. coli studies indicate a 7 transmembrane region topology, whilst SF21 cells and MelC cells indicate 6 transmembrane topologies. Clearly, additional heterologous expression systems need to be developed so that a consensus can be reached for the structure and function of MIP.
Heterologous expression of mammalian proteins in plant tissue is a promising candidate since expression is stable and permanent (not transient as in E. coli or insect cells) and shows post-translational modifications similar to mammalian tissues. Transgenic plants have already been successfully engineered to heterologously express members of the MIP family from other plant species . We have now extended this to mammalian lens MIP itself.
A bovine MIP amplicon bound by a 5' Kpn I restriction endonuclease site and a 3' Sma I site was produced by PCR using appropriate primers. This was ligated into the relevant sites of an expression cassette consisting of the bovine MIP cDNA under the control of a constitutive Cauliflower Mosaic Virus promoter  and a kanamycin resistance operon , flanked by Left and Right Border sequences from the Agrobacterium tumefaciens Ti plasmid . This cassette can be seen in Figure 1.
Sequence of MIP N-terminal primer:
The dinucleotide marked with green is a nonsense sequence. The sequence marked in red is the engineered Kpn I sequence. The slash indicates the restriction enzyme cut site. The rest of the sequence codes for a generic MIP N-terminal sequence.
Sequence of MIP C-terminal primer:
The dinucleotide marked with green is a nonsense sequence. The sequence marked in red is the engineered Sma I sequence. The slash indicates the restriction enzyme cut site. The rest of the sequence codes for a generic MIP C-terminal sequence.
PCR cycle parameters
Melting step: 94 °C for 1.5 min, annealing step: 60 °C for 2 min, elongation step: 72 °C for 2 min. At termination of 30 cycles, there was a further, final elongation step of 72 °C for 10 min.
Transformation of A. tumefaciens and thence plant tissue with this expression cassette was essentially as per reference . The expression cassette was cloned into the disabled A. tumefaciens LBA4404 strain by triparental mating (Helper plasmid pRK2013; Helper strain JM101). The resultant recombinant strain was used to infect Nicotiana tabacum (cultivar Wisconsin) leaf disks. Transgenic disks produced calli on kanamycin (250 µg/ml) selective medium. Shoots from calli were excised and transferred to fresh selective medium.
RNA and DNA extraction and amplification
Plant genomic DNA was extracted using the Qiagen (Crawley, West Sussex, UK) "DNAeasy" kit and accompanying proprietary protocol (Qiagen, catalogue number 69103). Plant mRNA was extracted using the Dynal (Bromborough, Wirral, Meseyside, UK) Dynabeads mRNA DIRECT kit and accompanying proprietary protocol (Dynal, catalogue number 610.11). Plant genomic DNA, and cDNA generated from plant mRNA after reverse transcription, were both used as templates in PCR with the primers described above, followed by agarose gel electrophoresis analysis . Reverse transcription was carried out using the Promega (Southampton, UK) Access RT-PCR kit and accompanying proprietary protocol (Promega, catalogue number A1250).
Immunohistochemical analysis of tissue samples from two different plant organs (young leaves, and roots) was carried out according to a proprietary protocol developed at Oxford Brookes University, UK (This protocol proved unsuitable for other tissues).
Examination of proteins in plant cells is complicated, since even minor disruption of plant tissues releases vacuolar proteases which rapidly degrade cellular proteins. Therefore, no sample could be obtained for simple analyses such as SDS-PAGE or Western blotting. Instead, samples were explored immunohistochemically, after fixing and sectioning in Butyl-Methyl Methacrylate resin, according to the following protocol:
Fixation: Roots and young leaves were dissected, and were fixed in 4% paraformaldehyde in 50 mM Pipes (Piperazine-N,N'bis[2-ethanesulfonic acid]) buffer plus 1 mM CaCl2, pH 6.9 for 2 h at room temperature, followed by 3 washes of 10 min each in 50 mM Pipes plus 1 mM DTT (pH 6.9). Root samples were then dyed with Fast Green to make them easier to see in subsequent manipulations.
Dehydration: Samples were dehydrated by transfer using tweezers to the following solutions. Each step had a duration of 30 min at the temperature indicated:
10% Ethanol + 1 mM DTT, 4 °C, rotated on "windmill"
25% Ethanol + 1 mM DTT, 4 °C, rotated on "windmill"
50% Ethanol + 1 mM DTT, 4 °C, rotated on "windmill"
70% Ethanol + 1 mM DTT in cryostat at -20 °C
90% Ethanol + 1 mM DTT in cryostat at -20 °C
100% Ethanol + 1 mM DTT in cryostat at -20 °C
Infiltration: Butyl-Methyl Methacrylate resin was then infiltrated into the samples by transfer using tweezers into the following solutions, all at -20 °C:
2:1 Ethanol:Resin + 1 mM DTT overnight
1:1 Ethanol:Resin + 1 mM DTT 2 h
1:2 Ethanol:Resin + 1 mM DTT 2 h
100% Resin + 10 mM DTT, 3 h or overnight
Polymerization: The infiltrated samples were then degassed with dry Nitrogen gas for 30 min. Samples were then transferred to aluminium dishes of fresh resin and polymerized at 0 °C under UV light (360 nm) overnight. A Nitrogen atmosphere (gas flow of 450 ml/min) was maintained throughout the polymerization. The resulting blocks of polymerized resin were then stored with dessication until sectioning.
Sectioning and mounting: The methacrylate resin-encased sample blocks were sectioned on a glass knife microtome at 2-4 µm. Individual sections were then transferred to drops of water on wells of multiwell Polylysine-coated slides, and allowed to dry at 37 °C overnight.
Washing, application of antibodies, and examination: The resin was dissolved by standing the slides in acetone for 5-10 min, without agitation. The samples on the slides were then rehydrated by standing in PBST (phosphate buffered saline [PBS] plus 0.05% Tween 20) for 10 min and blocked for 15 min in PBSBT (PBS plus 0.1% Tween 20, 1% BSA, and 0.002% Sodium Azide). Slides were then dried between the wells, and primary anti-MIP antibody was added to each well at the appropriate dilution in PBS. These slides were placed in a moist sealed container at 4 °C overnight, followed by incubation at room temperature for 2 h the following day. The slides were then washed in PBST three times for 10 min each. The PBST was individually pipetted and removed from each well. Secondary antibody (anti-Rabbit FITC [Fluoroisothiocyanate] conjugate) was then added at 1:40 dilution for 2 h at room temperature. The wells were then washed in PBS three times for 10 min each. Propidium iodide nuclear stain (3 µg/ml) was then briefly applied to each well followed by two washes in PBS and one wash with water. Coverslips were then mounted on the slides with Citifluor antifadeant and sealed with nail varnish. Slides could be examined immediately using fluorescence microscopy or stored at 4 °C.
Putative transgenic plants showed consistent root growth on kanamycin selective medium. Rooting is an excellent indicator of the presence of the expression cassette in plant tissue (Figure 2A,B). Beyond the ability to root in kanamycin selective medium, there was no difference in phenotype between transgenic plants and control plants grown on non-selective medium.
Rooted plants screened by PCR and RT-PCR exhibited a single PCR product of ~800 nucleotides (Figure 3A) which sequencing confirmed as bovine MIP. The presence of this band from mRNA (Figure 3B) confirms transcription of the cassette (mRNA yields from the plants were low, ~10 ng/µl, precluding quantitation). Control plants exhibited neither root growth nor PCR product.
Cell membranes from young transgenic leaves exhibited significantly enhanced immunofluorescence (Figure 4A,B) versus control plants (Figure 4C). Since immunofluorescence is localized to cell membranes, this suggests that MIP has been correctly targetted to young transgenic leaf cells within the plants.
We have succeeded in generating transgenic plants expressing bovine lens MIP. The phenotype of the plants did not differ from control plants, in contrast to a previous example of heterologous expression of an MIP family member  in N. tabacum where morphology was significantly affected.
We observed one sample of increased immunofluorescence in a young root. It is possible that MIP is only expressed in juvenile transgenic roots, although this would be hard to explore further since juvenile specimens would be very close to the main mass of roots and so hard to excise (only the outermost root hairs can be conveniently dissected away for analysis). Likewise, the distribution of MIP protein in other transgenic plant tissues is not known at present, due to their unsuitability for immunohistochemistry.
Currently, it is not possible to usefully quantify levels of transcription and translation of the MIP transgene. Recombinant genes in transgenic plants are not necessarily stable from one generation to the next. For instance, the introduced gene might lie within a methylation island in the plant genome, and become itself methylated and inactive with continued sub-culturing of the plant line. For this reason, in ongoing work the transgenic plant lines expressing MIP26 are being grown in a glasshouse environment to sexual maturity. Backcrossing of plant lines should result in a stable plant line; also the level of MIP expression can be maximized by crossbreeding different high-expressing lines.
This expression system has potential in the elucidation of MIP structure and function. Protoplasts can be isolated from plant tissues  and subjected to electrophysiological analysis such as patch clamping [11,12]. Transmembrane topology can be immunolocalized using antibodies raised to discrete peptide sequences from the MIP26 polypeptide. However, immunolocalization of cytoplasmically exposed domains of MIP has been carried out in E. coli and SF21 cells through inverting spheroplasts or disruption of cell membranes with agents such as acetone to allow antibodies access to the cytoplasmic side of the membrane [2,4]. These techniques are not applicable to plant cells due to the presence of a cell wall. Therefore, techniques such as gold-labelled antibodies used in conjunction with transmission electron microscopy may be required.
Although low resolution structures have been proposed for some MIP family members through cryoelectron microscopy studies [13-16], this has not been sufficient to distinguish between the different topological models proposed. Moreover, heterologous expression systems developed so far have not provided enough correctly folded recombinant protein to permit crystallisation trials. By contrast, a N. tabacum system can effectively be scaled up agriculturally and is not limited by available fermenter technology. Purification of proteins from plant cells is problematic due to vacuolar proteases; to counter this, we have constructed an N terminal oligohistidine-tagged MIP clone for the purposes of Nickel-chelating affinity chromatography, which should aid in the optimisation of MIP26 recovery from a variety of tissues based on existing N. tabacum membrane preparation protocols .
Plant systems have previously been used for the heterologous expression of a variety of substances on an industrial scale [17-19], including the production of recombinant light-harvesting complexes for crystallisation ; here we promote the use of transgenic plants on a laboratory scale, with an inherent and considerable potential for scale-up.
We thank the RNIB for support. We thank Dr Gorin of the University of California for the kind donation of the bovine MIP cDNA. We thank Dr Chris Hawes and colleagues at Oxford Brookes University, UK for generous technical assistance with sample preparation and microscopy.
1. Chepelinsky AB. The MIP transmembrane channel gene family. In: Peracchia C, editor. Handbook of membrane channels: molecular and cellular physiology. San Diego: Academic Press; 1994.
2. Dilsiz N, Crabbe MJ. Heterologous expression in Escherichia coli of native and mutant forms of the major intrinsic protein of rat eye lens (MIP26). Biochem J 1995; 305:753-9.
3. Mulders SM, Preston GM, Deen PM, Guggino WB, van Os CH, Agre P. Water channel properties of major intrinsic protein of lens. J Biol Chem 1995; 270:9010-6.
4. Schuette DG, Chepelinsky AB, Crabbe MJC. Membrane topology of recombinant native rat MIP26 in E. coli and Spodoptera frugiperda cells (Sf9). Invest Ophthalmol Vis Sci 1997; 38:S298.
5. Swamy-Mruthinti S. Expression and characterization of lens membrane intrinsic protein, MIP, in a baculovirus expression system. Curr Eye Res 1998; 17:88-94.
6. Kushmerick C, Rice SJ, Baldo GJ, Haspel HC, Mathias RT. Ion, water and neutral solute transport in Xenopus oocytes expressing frog lens MIP. Exp Eye Res 1995; 61:351-62.
7. Zhang Y, Roberts DM. Expression of soybean nodulin 26 in transgenic tobacco. Targeting to the vacuolar membrane and effects on floral and seed development. Mol Biol Cell 1995; 6:109-117.
8. Draper J, Scott R, Armitage P, Walden R, editors. Plant genetic transformation and gene expression: a laboratory manual. Blackwell Scientific; 1988.
9. Doyle K. Protocols and applications guide. 3rd ed. Promega, Southampton, UK; 1996.
10. Spangenberg G, Potrykus I. PEG mediated transfer of tobacco protoplasts. Gene transfer to plants, Springer Lab Manual, New York; 1995; p58.
11. Schuette DG, Drake KD, Jakob T, Crabbe MJC. Channel activity of native rat MIP26 synthesised in SF21 cells. Exp Eye Res 1998; 67:S.139.
12. Maathuis FJ, Sanders D. Patch-clamping plant cells. Methods Cell Biol 1995; 49:293-302.
13. Konig N, Zampighi GA, Butler PJ. Characterisation of the major intrinsic protein (MIP) from bovine lens fibre membranes by electron microscopy and hydrodynamics. J Mol Biol 1997; 265:590-602.
14. Li H, Lee S, Jap BK. Molecular design of aquaporin-1 water channel as revealed by electron crystallography. Nat Struct Biol 1997; 4:263-5.
15. Walz T, Hirai T, Murata K, Heymann JB, Mitsuoka K, Fujiyoshi Y, Smith BL, Agre P, Engel A. The three-dimensional structure of aquaporin-1. Nature 1997; 387:624-7.
16. Cheng A, van Hoek AN, Yeager M, Verkman AS, Mitra AK. Three-dimensional organization of a human water channel. Nature 1997; 387:627-30.
17. The emerging world of plant science [editorial]. Science 1995; 268:653-91.
18. Dunwell JM. Novel food products from genetically modified crop plants: methods and future prospects. International Journal of Food Science and Technology 1998; 33:205-13.
19. Poirier Y, Nawrath C, Somerville C. Production of polyhydroxyalkanoates, a family of biodegradable plastics and elastomers, in bacteria and plants. Biotechnology (N Y) 1995; 13:142-50.
20. Flachmann R. Purification and crystallisation of 6xHis-tagged membrane proteins produced in transgenic tobacco. Biochem Soc Trans 1999; 27:A412.