Abstract
An antigen present in plant vascular tissue crossreacts with antibodies towards keyhole limpet hemocyanin (KLH). The antigen is present in xylem and vascular cambium, as evidenced by immunocytochemical staining of plant sections. This cell type assignment was confirmed by staining of mesophyll cell cultures from Zinnia elegans L. undergoing tracheary cell differentiation. The strongest staining both in sections and cell cultures occured in cells and tissues during early stages of differentiation. Although the anti-KLH antibodies can easily be removed by affinity purification, our findings suggest that a certain caution is needed when KLH is used as an immunological carrier for studies in plants.
Keywords
V
KLH is a large protein complex present in the hemolymph of the mollusk Megatura crenulata (Swerdlow et al. 1996). The complex, which is copper-containing, serves as an oxygen carrier in the extracellular fluid. Related proteins fulfill the same function in other mollusks. Hemocyanins of arthropods, on the other hand, are of a different evolutionary origin and have a distinct structure, even though their overall function is the same. Very limited primary protein sequence data from the two subunits A and B of KLH have been established but much more extensive data are available in a few other species, e.g., the giant octopus (Octopus dofleini) (Miller et al. 1998). KLH has been used extensively as a carrier in raising peptide-specific antibodies (Dixon et al. 1966). Several things have contributed to its popularity for this purpose. It elicits a strong immune response in vertebrates. It is possible to couple considerable quantities of peptide to the complex, and there are few reputed problems with crossreacting antigens in most species. However, studies in plants using peptide antisera generated by the use of KLH have not been that common thus far (Bergmann et al. 1994).
We noted, while using different peptide antisera generated with the above approach, that although the peptides used for immunization were very different, the immunocytochemical staining in plant tissue sections was very similar unless the KLH-specific antibodies were removed. Additional experiments directly addressed the specificity of the crossreacting antigen and the cell type identity of the staining cells. We were able to clearly show that the staining is specific for cells undergoing differentiation to form the xylem of the plant. The KLH-crossreactive antigen appears to be an early marker for this developmental pathway.
Materials and Methods
Plant Material
Plant tissues used for immunocytochemical staining were stamens, leaves, petals, and sepals from Adonis aestivalis, roots, leaves, stamens, and siliques from Arabidopsis thaliana, leaves from Nicotiana tabaccum, and stems and seeds from Brassica napus.
Zinnia elegans Tracheary Cell Cultures
Cultures were prepared and grown as described in Groover et al. (1997). Fixation in preparation for microscopy was by addition of 2% or 4% final concentration of paraformaldehyde. No discernible differences were seen between these two fixation conditions.
Generation of Antisera
Coupling of cysteine containing peptides to the KLH carrier by carbodiimide coupling and immunization of rabbits with the protein complexes were carried out on a commercial basis by Agri Sera AB (Vännäs, Sweden) according to standard protocols (Harlow and Lane 1988).
The initial immunization of the rabbits was by subcutaneous injection of 200 μg of the respective peptide–KLH complex or KLH alone in Freunds' complete adjuvant. Booster injections of 100–200 μg were administered every sixth week after the initial injection, and the rabbits were bled every second week. The resulting sera were assayed by double immunodiffusion, immunocytochemical staining, or ELISA.
Affinity-purification of Antisera
An affinity resin for depletion of anti-KLH antibodies from whole sera was prepared by covalently coupling KLH to CNBr-activated Sepharose 4B (Pharmacia Biotech; Uppsala, Sweden). The procedure suggested by the manufacturer was used. Small portions of sera were allowed to pass by gravity flow over minicolumns set up in punctured Eppendorf centrifuge tubes. The flow-through was collected and tested for removal of anti-KLH antibodies by double immunodiffusion before they were used for immunocytochemical staining of tissue sections.
Affinity columns for positive selection of specific antibodies were obtained by coupling of the antigen to SulfoLink columns (Pierce; Rockford, IL). The coupling of antigen, binding of antibodies, subsequent washing, and elution of specific antibodies were as described in the manual provided by the manufacturer.
Total IgG fractions from preimmune sera were obtained by use of HiTrap protein-G columns (Pharmacia Biotech) as described by the manufacturer.
Immunological Procedures
Double immunodiffusion assays were performed according to Ouchterlony (1967). Preparation of microscopy specimens and immunocytochemical staining were performed as described (Höglund et al. 1991). To make the handling of Zinnia cells easier, the cell suspension was molded into 1% agarose before being subjected to the remainder of the preparation procedure.
Similarity Searches and Multiple Sequence Alignments
Gapped-BLAST and PSI-BLAST search algorithms were used for searching databases with distinct protein sequences (Altschul et al. 1997). Multiple sequence alignments were performed with Clustal W (Thompson et al. 1994), and the resulting similarities highlighted by subjecting the alignments to Boxshade (Hofmann and Baron, unpublished).
Results
Rabbit peptide antisera that have been raised with KLH as a carrier for the peptides contain high titers of antibodies directed towards KLH itself (Figure 1, before). This is true irrespective of whether or not there is a response to the coupled peptide. The results presented here were all obtained with an antiserum that had undetectable levels of antibodies towards the peptide. We show that the anti-KLH antibodies give rise to specific staining in plant tissue sections. Structures and cells that appear to be xylem or xylem-like, with the highly characteristic thickened cell walls (Figures 2A–2D), stain specifically and strongly. Removal of anti-KLH antibodies (Figure 1, after) by passage over an affinity resin also abolishes the specific staining (Figure 2C). Therefore, the reaction in Figure 2B is due to interactions with immunoglobulins directed towards KLH. To further characterize the cell type identity of the staining cells, cross-sections of stems were subjected to immunocytochemical staining. Figure 2H shows a typical staining pattern from a stem section. It clearly shows that some xylem vessels, and the inner part of the cambium layer, are stained. In cross-sections, only vessels that still contain cells or remnants of protoplasts are distinctly stained. In lengthwise sections of tissues from various origins, particularly strong staining is always seen where there are living protoplasts confined by thickened cell walls with tightly spaced annular or spiral shapes. This is characteristic of meta-xylem. To further strengthen the cell type identification, we turned to a well-characterized experimental system for tracheary cell differentiation. Thus, Zinnia elegans mesophyll cells that have been induced to undergo tracheary cell differentiation also exhibit specific staining (Figure 2F). This is particularly so when the cells are still alive and often even before the time when a distinctly visible change of cell wall morphology has occurred. The same is true in tissue sections. As seen in Figure 2D, which is a slanted section across a whole bundle of cells, living protoplasts or remnants of them stain the strongest. The crossreactive antigen therefore appears to be a marker for this particular developmental fate and appears to be most abundant in an early phase of differentiation. Our data are therefore consistent with expression of the antigen in cells that are active in forming the xylem. The antigen may then reside in the xylem for some time even after the xylem-forming cells themselves have died, as evidenced by the diffuse staining seen in Figure 2B.
Double immunodiffusion of sera against KLH. The wells (approximately 40 μl) were punched in 1% agarose plates containing PBS. They were filled with 1:5 dilutions in PBS of untreated serum (Before) or serum that had been passed over an affinity resin that had KLH covalently coupled to it (After). The well-labeled KLH contained KLH at 1 mg/ml in PBS. After precipitin lines had formed, excessive protein was removed by extensive washing of the plates and subsequent staining with Coomassie brilliant blue according to Ouchterlony (1967).
We wanted to ensure that the treatment of antisera by passage over a KLH affinity resin did not lead to generalized depletion or inactivation of antibodies with other specificities than anti-KLH. For this purpose, we allowed a well-characterized anti-myrosinase antiserum (Höglund et al. 1991) to pass over a KLH column before its use. This and the untreated control anti-myrosinase serum were then used in staining rapeseed tissue sections for the presence of myrosinase. Both treated and untreated sera yielded the same staining pattern characteristic for myrosine cells (data not shown), thus ensuring that the depletion seen is specific for the anti-KLH antibodies and not a general effect.
As a final and definitive test for the specificity of the staining patterns observed, we also used antibodies that were raised by immunizing rabbits with KLH alone, i.e., with no peptide coupled to the complex. As an additional safety measure, these anti-KLH antibodies were in turn affinity-purified by binding and subsequent washing and elution on SulfoLink columns that had KLH coupled to positively ensure that only anti-KLH antibodies were responsible for the staining. These antibodies gave the same specific staining pattern (data not shown). This constitutes compelling evidence that we are indeed visualizing an antigen that is crossreactive with anti-KLH antibodies and that resides in cells that are undergoing differentiation to form xylem.
Discussion
We have established the specificity and the cell type in which it resides for the KLH-crossreactive antigen in plant vascular tissue. For the time being we have, unfortunately, not been able to further identify the antigen. However, we have noted suggestive and quite extensive sequence similarities between hemocyanins and other proteins. Database searches were performed using the gapped-BLAST or PSI-BLAST algorithms (Altschul et al. 1997). Searches were done using the long g-type hemocyanin sequence from O. dofleini (Miller et al. 1998). Many strikingly significant hits from many different taxa turned up. The most significant of these are other hemocyanins, but a whole range of tyrosinases/polyphenol oxidases (PPOs)/catechol oxidases also show significant identity. From the tyrosinase/PPO/catechol oxidase category of proteins, sequences from plants, bacteria, fungi, and animals were found. The hits are highly significant statistically and the similarities extend over long stretches of the protein sequences (data not shown). The similarities were particularly impressive when analyzed by multiple sequence alignments of several sequences from the different taxa. Many completely conserved amino acids indicate extensive conservation. Both hemocyanins and tyrosinases are copper-containing proteins that use the bound copper ions to immobilize molecular oxygen, in one case for oxygen transport purposes and in the other for enzymatic oxidative purposes. The proteins turn out to contain highly conserved motifs with critical His residues that bind the copper ions. The copper-binding regions indeed show the highest degrees of similarity.
Figure 3 shows a limited part of the amino acid sequence similarities that were found. The sequence shown is where the highest similarity is seen, displayed in a multiple sequence alignment format of taxonomically diverse representative sequences. The region encompasses a highly characteristic double His among many other fully conserved amino acids. The latter of the two histidines in the sequences shown constitutes the third copper-coordinating His in the copper B binding domain. In molluskan hemocyanins the oxygen-binding domain is repeated eight times in each subunit, but in Figure 3 only one of these repeats is shown for each hemocyanin sequence. The similarities between hemocyanins and tyrosinases have been observed and discussed by others (Van Gelder et al. 1997) and are in fact so extensive that a structural model for better understanding of the tyrosinase enzyme mechanism is being based on the known 3D structure of the hemocyanin oxygen-binding domain.
Although the overall functions of tyrosinases and related proteins are reasonably clear, most of their representatives in plants have not been extensively studied. Among other things, in plants they are known to be involved in secondary cell wall biosynthesis and modification, both during normal development and in modifications in response to various forms of stress. Most of the actual enzymes and genes have not been extensively studied with regard to their tissue distribution, subcellular localization, or induction properties.
Immunocytochemical staining of plant tissue sections and cells undergoing tracheary differentiation. Sera, according to Figure 1, as well as a preimmune serum from the same rabbit were used for staining of sections. (
Whether or not the striking and strong sequence similarities we have found are indeed an explanation for the crossreactivity remains to be determined by purification and identification of the crossreactive antigen.
Multiple sequence alignment of different hemocyanins and tyrosinases/PPOs/catechol oxidases. The aligned sequences were retrieved by gapped-BLAST, followed by Clustal W multiple sequence alignments of the regions that showed the best similarities. Boxshade was used to highlight the similarities. A star under an aligned position indicates complete conservation among the sequences shown. The sequences and the E-values for the gapped-BLAST search are, from top to bottom: Octopus dofleini, hemocyanin AAC39018, amino acids 1439–1467, E=0.0; Helix pomatia, hemocyanin P12031, amino acids 189–217, E=10−124; Rapana thomasiana, hemocyanin P80960, amino acids 185–213, E=10−99; Podospora anserine, tyrosinase Q92396, amino acids 287–315, E=2 × 10−13; Agaricus bisporus, polyphenol oxidase S53017, amino acids 269–298, E=10−11; Prunus armeniaca, polyphenol oxidase gi3282505, amino acids 358–386, E=8 × 10−10; Lycopersicon esculentum, catechol oxidase Q08305, amino acids 349–377, E=10−7; Spinacia oleracea, polyphenol oxidase S72509, amino acids 371–399, E=2 × 10−8; Streptomyces antibioticus, tyrosinase P07524, amino acids 200–227, E=7 × 10−13; Mus musculus, tyrosinase P11344, amino acids 373–401, E=7 × 10−7.
Footnotes
Acknowledgements
Supported by a grant to LGJ from Carl Tryggers Foundation for Scientific Research and to AMJ from the National Science Foundation Developmental Mechanisms Program. SH was supported by grants from NUTEK.
The affinity-purified anti-KLH antibodies were a generous gift from Agri Sera AB (Vännäs, Sweden).