Abstract
Parvulins are a distinct family within the peptidyl-prolyl cis-trans isomerase group of proteins that catalyse the cis-trans isomerization of proline-containing peptides. The intracellular distribution of a novel human parvulin homologue (hEPVH) has been investigated in a human kidney cell line (HEK 293) by immunogold labeling transmission electron microscopy (TEM). This showed hEPVH to be distributed throughout HEK 293 cells but in highest concentration within mitochondria. Unexpectedly, preabsorption of anti-hEPVH antiserum with recombinant hEPVH exaggerated the observed immunolabel density in a pattern that mirrored that of the endogenous hEPVH. The hEPVH protein itself was then used to label sections, and the specificity of its binding was confirmed with the use of polyclonal and monoclonal antibodies in conjunction with homologous and irrelevant protein controls. The pattern of hEPVH binding also mirrored that of endogenous hEPVH. A known chaperone protein, Pin1, was also found to bind to cells in a pattern mirroring that of the endogenous protein. This lends considerable weight to our hypothesis that hEPVH is binding to its target protein(s) within the cell, reflecting its postulated chaperone function. Finally, we suggest that chaperone proteins in general might be used as TEM probes for their target (or substrate) proteins.
T
The first member of the parvulin family of proteins was identified in E. coli (Rahfeld et al. 1994b). Eukaryotic homologues, such as S. cerevisiae Ess1 (Hanes et al. 1989) and D. melanogaster Dodo (Maleszka et al. 1996) have since been characterized. The essential mitotic regulator, Pin1, was the first member of this family to be discovered in humans (Lu et al. 1996; Crenshaw et al. 1998).
A new subfamily of parvulin homologues has been identified (Rulten et al. 1999) and termed EPVHs (eukaryotic parvulin homologues), owing to their close sequence homology with bacterial parvulin (Rahfeld et al. 1994a). We have recently characterized the 14-kD human homologue (hEPVH) of this group of proteins.
In this study we investigated the subcellular localization of the hEPVH protein by immunogold labeling transmission electron microscopy (TEM). During the course of this work, a standard homologous preabsorption control for antiserum specificity was found to exaggerate, rather than abolish, the observed immunolabel density. We postulated that this putative chaperone protein might be binding to its target protein on the section surface, and a series of experiments was carried out to test this hypothesis.
Materials and Methods
Cell Line
HEK 293 cells were grown to confluency in DMEM supplemented with Glutamax (Life Technologies; Rockville, MD), 10% fetal calf serum (Life Technologies), 100 U/ml penicillin, and 0.1 mg/ml streptomycin (Sigma; Poole, UK).
Recombinant hEPVH and Pin1 Protein Production
PCR primers were designed to amplify the full-length hEPVH coding sequence (Genbank Accession No. AF143096) from human lung cDNA (Clontech; Basingstoke, UK). The product was cloned into pET21a (Novagen; Madison, WI) with the coding sequence in-frame with the vector's N-terminal T7-tag. A pET20b (Novagen) construct for un-tagged full-length Pin1 expression was kindly provided by Pfizer Central Research (Sandwich, UK). Transformed E. coli strain BL21 (DE3) cells (Novagen), were grown in 2YT broth to an OD600 of 0.8, and expression was induced with 1 mM IPTG. Cells were harvested 2 hr after induction and lysed in 50 mM Tris-HCl, 5 mM MgCl2, 2 mM CaCl2, 5% glycerol, 5 mM DTT (Sigma), pH 7.5 (for Pin1) or pH 8.9 (for hEPVH). Recombinant protein was purified to homogeneity (Rulten et al. 1999) using a 0–1 M NaCl gradient with a HiTrap Heparin FPLC column, followed by gel filtration on a Superdex-75 24-ml sizing column (Amersham Pharmacia Biotech; Little Chalfont, UK).
Mass Spectrometry
Rec-hEPVH (0.5 mg/ml; c.30 μM) was analyzed in a Micro-mass TofSpec-E MALDI-TOF mass spectrometer. The sample spectrum was acquired in linear positive-ion mode using a matrix of 12 mg/ml sinapinic acid and was externally calibrated against cytochrome c and trypsinogen.
Preparation of Cells for Electron Microscopy
All the following procedures were carried out at 4C. After fixation for 3 hr in 4% formaldehyde plus 0.1% (vacuum-distilled) glutaraldehyde in PBS, 0.14 M, pH 7, the cells were rinsed thoroughly (five changes and left overnight) in PBS and dehydrated in an ethanol series (30%, 50%, 75%, 90% and 3 × 100%, 20 min each). The samples were then embedded in Unicryl resin (British BioCell International; Cardiff, UK) by a procedure slightly modified from Bogers et al. (1996) as follows. After 30 min each in 2:1 and then 1:2 100% ethanol:Unicryl, the samples were allowed to infiltrate in complete resin overnight. The resin was changed the following day and the samples dispensed into (size 00) Beem capsules. The resin was light-polymerized for 48 hr by illumination (from the underside of the Beem capsules) from a 12-V, 100-W type 6834 Philips projection lamp at a distance of 35 cm. Thin sections were cut on an LKB Ultrotome III ultramicrotome and collected on formvar-coated grids.
Antibodies and Gold Probes
Rabbit polyclonal antisera were raised to the purified recombinant hEPVH ([p]anti-hEPVH) by Cambridge Research Biochemicals (Northwich, Cheshire, UK). Western blot analysis of whole HEK 293 cell lysate has confirmed the specificity of this antiserum. Mouse anti-T7 Tag monoclonal antibody (hereon termed [m]anti-hEPVH) was obtained from Novagen. Goat anti-Pin1 polyclonal antisera to the C-terminus ([p]anti-(C)Pin1) and N-terminus ([p]anti-(N)Pin1) domains were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal antibody to bovine growth hormone ([m]anti-bGH) was kindly donated by our colleague Prof. Mike Wallis (for specificity see Thorpe et al. 1990). Five-nm goat anti-rabbit IgG-gold (GaR5), 10-nm goat anti-mouse IgG-gold (GaM10), and 10-nm protein A and G-gold (pAG10) were obtained from British BioCell International.
Immunogold Labeling TEM
For details of the precise immunolabeling methodology used throughout this work, please see Thorpe (1999). All immunolabeling experiments were carried out twice or three times and, within each experiment, each variant of the immunolabeling procedure was duplicated. A modified PBS, pH 8.2, containing 1% BSA, 500 μl/liter Tween-20, 10 mM NaEDTA, and 0.2g/liter NaN3 (henceforward termed PBS+) was used throughout all the following procedures for all dilutions of antisera and gold probes.
hEPVH Immunolabeling. Sections were blocked for 30 min in 1:10 normal goat serum at room temperature (RT) and then incubated overnight at 4C in 1:200 [p]anti-hEPVH. As a control, sections were concurrently incubated in preimmune rabbit serum (PIS) at similar dilution. After three 2-min rinses in PBS+, the sections were immunolabeled for 1 hr at RT in GaR5 and subsequently rinsed in PBS+ (three times for 10 min) and then distilled water (four times for 3 min). Sections were poststained in 0.5% aqueous uranyl acetate for 15 min.
Homologous and Heterologous Preabsorptions
Liquid Phase. The [p]anti-hEPVH was preabsorbed (at the working dilution of 1:200) in 20 or 200 μg/ml recombinant hEPVH (rec-hEPVH) or 20 μg/ml Pin1 (rec-Pin1) for 48 hr at 4C. PIS and [p]anti-hEPVH (at similar dilution) minus this protein addition were set up concurrently for control immunolabelings.
Solid-phase Coupling of hEPVH to NHS-activated Sepharose. HiTrap NHS-activated affinity columns were obtained from Amersham Pharmacia Biotech. Rec-hEPVH was concentrated to 5 mg/ml using a centricon-3 column (Millipore; Watford, UK) and diluted in coupling buffer (0.2 M NaHCO3, 0.5 M NaCl, pH 8.3) to give a final concentration of 1 mg/ml. One ml of this solution was injected onto the activated column, which was then incubated at RT for 1 hr. Excess active groups were deactivated by serially washing the column with 0.5 M ethanolamine, 0.5 M NaCl, pH 8.3, and 0.1 M acetate, 0.5 M NaCl, pH 4. A blank Sepharose column was made by loading coupling buffer only onto another column, incubating for 1 hr, and deactivating the excess groups in the same way as for the hEPVH column. Both columns were equilibrated in PBS+ before loading 1 ml of a 1:50 dilution of [p]anti-hEPVH and incubating at 4C for 1 hr. The preabsorbed antiserum was eluted from the column in 4 ml of PBS+ to obtain a 1:200 diluted antiserum which was used directly for immunolabeling.
Verification of Binding of Rec-hEPVH on the Preabsorbed Antibody to Sections. After homologously preabsorbed (with rec-hEPVH) [p]anti-hEPVH incubation and immunolabeling with GaR5, sections were incubated in a monoclonal antibody against the T7 tag of the rec-hEPVH ([m]anti-hEPVH; 50 μg/ml) and immunolabeled with GaM10. As a control for this latter immunolabeling, sections were incubated in an irrelevant monoclonal antibody ([m]anti-bGH; 87.5 μg/ml) before immunolabeling with GaM10.
Using Recombinant hEPVH and Pin1 to Label Cells. For rec-hEPVH labeling, this procedure was identical to the routine immunogold labeling except that an incubation in 20 μg/ml rec-hEPVH (in PBS+ for 24 hr at 4C, followed by either three or six 2-min PBS+ rinses) was inserted between the normal goat serum blocking and specific antibody ([p]anti-hEPVH or [m]anti-hEPVH) incubation steps. As controls, sections were concurrently incubated in PBS+ alone (in place of rec-hEPVH) and also (for the monoclonal antibody labeling) in an irrelevant protein [(bovine) growth hormone (bGH; 20 μg/ml)] which was immunolabeled with its specific monoclonal antibody ([m]anti-bGH; 87.5 μg/ml) and GaM10.
For rec-Pin1 labeling, sections were blocked in PBS+ for 30 min at RT, incubated in 20 μg/ml rec-Pin1 (as for rec-hEPVH), then immunolabeled with a mixture of 1:50 (4 μg/ml) [p]anti-(C)Pin1 and [p]anti-(N)Pin1 and subsequently pAG10. Other sections were concurrently incubated in PBS+ alone (in place of rec-Pin1) and either immunolabeled (as for the rec-Pin1 labeling) to reveal endogenous Pin1 protein or with normal goat serum (4 μg/ml).
Verification of the Specificity of Recombinant hEPVH Binding. To verify the specificity of rec-hEPVH binding, sections were preincubated in either the closely related protein rec-Pin1 or in the irrelevant protein bGH (both at 20 μg/ml for 24 hr at 4C) before rec-hEPVH incubation and immunolabeling as described above. As further controls, concurrent sections were preincubated in PBS+ alone (before rec-hEPVH incubation and immunolabeling) or in PBS+ followed by bGH incubation and immunolabeling with [m]anti-bGH (50 μg/ml). Finally, as a control for nonimmunological, hydro-phobic protein interactions, sections were incubated in Pin1 and bGH (as above) and then immunolabeled with [p]anti-hEPVH.
Image Acquisition and Analysis
The immunolabeled thin sections were examined in a Hitachi-7100 TEM at 100 kV. Random images at X50,000 magnification were acquired digitally with an 800 × 1200 pixel charge-coupled device camera (Digital Pixel; Brighton, UK). The areas of the different cell compartments were computed (Kinetic Imaging; Liverpool, UK), gold particles counted, and the immunolabeling results expressed as numbers of gold particles per μm2. Means and standard errors (SEM) were calculated and t-tests were used for statistical analyses of significance.
Results
Subcellular Distribution of Endogenous hEPVH Protein
Figure 1 shows the pooled results of three separate immunolabelings for endogenous hEPVH. These revealed the presence of hEPVH throughout the cells, but in higher concentration within the mitochondria. The immunolabeling of the latter organelles was confined to the matrix.
Preabsorption Results
Preabsorption of the [p]anti-hEPVH with its homologous antigen (rec-hEPVH at 20 μg/ml) greatly exaggerated, rather than abolished, the observed immunolabeling density (Figure 2). This increased immunolabeling density was some 11-, 14-, and 10-fold for the nuclear, cytoplasmic, and mitochondrial compartments, respectively. The pattern of distribution of this exaggerated immunolabeling was similar to that of the endogenous hEPVH protein. Preabsorption with an increased concentration of rec-hEPVH (200 μg/ml) also resulted in elevated immunolabel densities (not shown).
A (concurrent) heterologous preabsorption (with rec-Pin1 at 20 μg/ml) had only minor effects on immunolabeling density (Figure 2).
Subcellular distribution of endogenous hEPVH. Immunogold labeling densities for sections incubated in 1:200 preimmune serum (PIS) or [p]anti-hEPVH (black columns). Data = GaR5 gold particles/μm2, n = 30 ± SEM, except for PIS, where n = 60 ± SEM (combined data for all cell compartments). Significant differences of PIS vs hEPVH: nucleus and mitochondria p<0.0001; cytoplasm p<0.05.
A homologous (rec-hEPVH-bound Sepharose) column preabsorption of the [p]anti-hEPVH resulted in an 88% reduction in immunolabel density over the mitochondria [compared with a blank column (n = 20); data not shown].
Subcellular distribution of immunolabeling after incubation with homologously and heterologously preabsorbed [p]anti-hEPVH. Immunogold labeling densities for sections incubated in either (nonpreabsorbed) 1:200 [p]anti-hEPVH (white columns) or the latter preabsorbed with 20 μg/ml rec-hEPVH (black columns) or 20 μg/ml Pin1 (striped columns). Data = GaR5 gold particles/μm2; n = 30 ± SEM. Significant differences of non-preabsorbed [p]anti-hEPVH vs rec-hEPVH preabsorption: p<0.0001 for all cell compartments.
Homologous preabsorption: verification of recombinant hEPVH binding using a monoclonal anti-hEPVH antibody. [m]anti-hEPVH (50 μg/ml) immunogold labeling densities for sections previously incubated in homologously preabsorbed (20 μg/ml rec-hEPVH) 1:200 [p]anti-hEPVH as in Figure 2 (black columns). Irrelevant, irrelevant monoclonal antibody ([m]anti-bGH; 87.5 μg/ml). Data = GaM10 gold particles/μm2; n = 20 ± SEM, except for “Irrelevant,” where n = 30 ± SEM (pooled data for all cell compartments). Significant differences of [m]anti-hEPVH vs irrelevant: cytoplasm p<0.001; nucleus and mitochondria p < 0.0001.
Recombinant hEPVH labeling: immunolabeling with polyclonal antiserum. Immunogold labeling densities for sections incubated in either PBS+ (white columns) or 20 μg/ml rec-hEPVH (black columns), both followed by 1:200 [p]anti-hEPVH and GaR5. Data = GaR5 gold particles/μm2; n = 20 ± SEM. Significant differences of PBS+ vs hEPVH: p<0.0001 for all cell compartments.
Verification of Binding of Recombinant hEPVH on the Preabsorbed Antibody to Sections
The results of immunolabeling sections initially with rec-hEPVH preabsorbed antiserum (and GaR5) and subsequently with [m]anti-hEPVH (and GaM10) confirmed the binding of the recombinant protein to the section. The cellular distribution of monoclonal immunolabeling (Figure 3) was similar to that of the preabsorbed polyclonal antiserum. A control incubation in an irrelevant monoclonal antibody ([m]anti-bGH) at a concentration higher (87.5 rather than 50 μg/ml) than the specific monoclonal exhibited extremely low levels of labeling.
Using Recombinant hEPVH to Label Cells
Incubation of sections in 20 μg/ml rec-hEPVH alone before immunolabeling with the [p]anti-hEPVH and GaR5 also resulted in increased immunolabeling. There was no difference in the immunolabeling density between grids rinsed three or six times before transferring to the antibody (not shown). The immunolabeling pattern was similar to that of the rec-hEPVH-preabsorbed antibody (Figure 4). Two further immunolabelings carried out using a fresh batch of rec-hEPVH protein exhibited very high immunolabel densities (Figure 5A and 5B). A separate incubation in 20 μg/ml rec-hEPVH followed by immunolabeling with the [m]anti-hEPVH and GaM10 confirmed the presence of the rec-hEPVH bound to the sections (Figures 5D, 5E, and 6).
As with the endogenous hEPVH, the immunolabeling of the mitochondria was confined to the matrix (Figure 5A and 5E). It was also noted that there was some association of labeling with the cristae membranes (Figure 5B).
Using recombinant hEPVH and Pin1 as TEM probes. Digital TEM images showing: (A,B) immunolabeling in the mitochondria after incubation in rec-hEPVH before immunolabeling with [p]anti-hEPVH and GaR5. (B) Enlarged view of the boxed area in A, showing a lack of gold particles over the (electron-lucent) cristae and the relative high frequency of labeling along the cristae membranes (arrows). (D,E) Immunolabeling in the mitochondria after incubation in rec-hEPVH before immunolabeling with [m]anti-hEPVH and GaM10. (D) A portion of an HEK 293 cell (n, nucleus). (E) Enlarged view of the mitochondrion arrowed in D. Note that with both the monoclonal and polyclonal anti-sera the labeling is confined to the matrix of the mitochondria. (C,F) Immunolabeling in the nucleus after incubation in rec-Pin1 before immunolabeling with [p]anti-(C and N terminus) Pin1 and pAG10. nu, nucleolar profile; arrows denote nuclear membrane. (F) Enlarged view of the boxed area in C. Bars: A,B,E,F = 100 nm; C = 0.5 μm; D = 1 μm.
Verification of the Specificity of Recombinant hEPVH Binding
Preincubation with a closely related (Pin1) or an irrelevant (bGH) protein had no significant effect on rec-hEPVH binding (Figure 7). Only extremely low levels of nonspecific bGH binding were revealed by incubation in, and subsequent specific immunolabeling for, this protein. Incubation in the closely related (Pin1) or an irrelevant (bGH) protein before immunolabeling with [p]anti-hEPVH revealed label densities at endogenous levels. Therefore, nonimmunological, hydrophobic protein binding interactions were absent (not shown).
Mass Spectrometry of Rec-hEPVH
Mass spectrometry showed no evidence for rec-hEPVH dimers, there being a single peak in the sample spectrum equivalent to the molecular weight of the protein (not shown).
Subcellular Distribution of Endogenous Pin1 and Recombinant Pin1 Labeling of Cells
Immunolabeling for endogenous Pin1 revealed the presence of this protein throughout the cells but, in contrast to the distribution of the hEPVH protein, the highest label density was within the nuclear compartment (Figure 8, white columns). A control immunolabeling in nonimmune goat serum at an identical protein concentration exhibited extremely low levels of labeling.
Incubation of sections in 20 μg/ml rec-Pin1 alone before immunolabeling with the [p]anti-(C- and N-terminus) Pin1 and pAG10 resulted in increased immunolabeling densities over endogenous levels (Figure 8, black columns, and Figure 5C and 5F). Similarly to the rec-hEPVH incubations, there was no difference in the immunolabeling density between grids rinsed three or six times before transferring to the antibody (not shown). As with the rec-hEPVH incubation results, the pattern of this increased labeling was similar to that of the endogenous protein.
Recombinant hEPVH labeling: immunolabeling with monoclonal antibody. Immunogold labeling densities for sections incubated in 20 μg/ml rec-hEPVH followed by (50 μg/ml) [m]anti-hEPVH and GaM10 (black columns). Cell is a control incubation in bGH (at 20 μg/ml) followed by (87.5 μg/ml) [m]anti-bGH and GaM10 (white column). Data = GaM10 gold particles/μm2; n = 20 ± SEM, except for GH, where n = 60 ± SEM (pooled data for all cell compartments). Significant differences of [m]anti-hEPVH vs GH: p<0.001 for all cell compartments.
Preincubation controls for recombinant hEPVH labeling. Immunogold labeling densities for sections preincubated in either buffer (PBS+), 20 μg/ml Pin1 (Pin1), or 20 μg/ml bovine growth hormone (GH) before incubation in 20 μg/ml rec-hEPVH followed by [m]anti-hEPVH (50 μg/ml) and GaM10. PBS+/GH is a further control preincubation in buffer, followed by incubation in 20 μg/ml bGH and immunolabeling with [m]anti-bGH (50 μg/ml) and GaM10. Data = GaM10 gold particles/μm2 mitochondrial profile; n = 30 ± SEM. Significant differences of PBS+ vs Pin1, GH and PBS+/GH are p = 0.779, p = 0.147, and p <0.0001, respectively.
Endogenous Pin1 distribution and recombinant Pin1 labeling. Immunogold labeling densities for sections incubated in either PBS+ (white columns) or 20 μg/ml rec-Pin1 (black columns), both followed by 4 μg/ml [p]anti-(C)Pin1 and [p]anti-(N)Pin1 and pAG10. NGS is a control incubation in (4 μg/ml) normal goat serum. Data = pAG10 gold particles/μm2; n = 20 ± SEM (for PBS+ and Pin1) n = 30 ± SEM (for NGS; pooled data for all cell compartments). Significant differences of PBS+ vs Pin1: p<0.0001 for all cell compartments.
Discussion
The initial TEM immunogold labeling results for the distribution of endogenous hEPVH showed that it was present throughout the HEK 293 cells but in higher concentration within the mitochondria. A subsequent routine homologous (rec-hEPVH) preabsorption test for the polyclonal anti-hEPVH antiserum specificity yielded an unexpected result: the immunolabel density was greatly exaggerated rather than abolished. We therefore postulated that the rec-hEPVH (bound to the preabsorbed antibody) was binding to the section. In addition, because this increase in immunolabeling density appeared to mirror the distribution of the endogenous hEPVH, we suggest that the rec-hEPVH might be binding to its (unbound) target protein within the cell (the possibility of hEPVH-hEPVH binding events was dismissed because mass spectrometry showed no evidence for rec-hEPVH dimerization).
We could not directly confirm the latter in the current absence of knowledge of this protein's identity and thus the lack of a specific antibody. However, immunolabeling of the sections incubated with homologously preabsorbed anti-hEPVH and subsequently with a monoclonal antibody against the (T7 tag of) rec-hEPVH lent convincing evidence: the pattern of this immunolabeling was similar to that of the preabsorbed antibody.
Concerns that the observed effect could be due to use of suboptimal concentrations of rec-hEPVH (despite calculations suggesting the opposite) during the initial (20 μg/ml hEPVH) preabsorption (causing signal amplification via, e.g., endogenous hEPVH-anti-hEPVH-rec-hEPVH-anti-hEPVH binding events) were allayed by the following. First, preabsorption with 200 μg/ml hEPVH also caused an increase in immunolabeling density. Second, incubation of sections in the rec-hEPVH protein alone [followed by thorough rinsing before subsequent immunolabeling with (non-preabsorbed) anti-hEPVH] exhibited a similar cellular binding pattern. The specific nature of this rec-hEPVH binding was confirmed by the results of incubations and preincubations with a related (rec-Pin1) and an irrelevant protein (bGH).
As a final test of our hypothesis, we investigated the subcellular distribution of the parvulin protein Pin1, a known chaperone protein with a known (predominantly nuclear, some cytoplasmic) localization (Lu et al. 1996). Immunolabeling results confirmed a predominantly nuclear location of the Pin1. In addition, using a recombinant Pin1 to label the sections again revealed a binding pattern similar to that of the endogenous protein.
These latter results therefore lend considerable weight to our argument that both hEPVH and Pin1 may be binding to their target proteins within the section, and support our theory that hEPVH, like Pin1, acts as a chaperone. It is noteworthy that the higher levels of rec-hEPVH binding in the mitochondria might reflect a higher level of unbound target protein. Therefore, this organelle might be a target for hEPVH protein activity.
In summary, this work reports on immunolabeling evidence of the differential binding of a putative (hEPVH) and a known (Pin1) chaperone protein to various cell compartments. Although enzyme-gold probes have been shown to bind to their substrates (Bendayan 1981) and some proteins used as probes in light (e.g., vasopressin; Fadool and Aggarwal 1990) and electron microscopy (e.g., galectin-1; Wasano and Hirakawa 1997, 1999), we believe that the use of chaperone proteins as TEM probes is a novel methodology that might ultimately be very informative in the elucidation of their role(s) in the cell. Where a purified or recombinant chaperone protein and its specific antibody are available, the former can be used as a probe for unbound target protein and the latter for endogenous chaperone protein distribution. In this way, insight might be gained not only on the distribution of these proteins but also, perhaps, on ratios of unbound to bound target protein under different conditions.
It should be noted that, as with standard immunogold labeling, the preparation of particular specimens/chaperone proteins must be optimized empirically. Undoubtedly, the preferred method will always involve minimizing the denaturation of proteins throughout the procedure. The use in this work of the acrylic resin Unicryl (after a minimal cold fixation) has apparently facilitated the maintenance of in vivo chaperone/target protein binding interactions. This recently developed resin minimizes protein denaturation, and its largely hydrophilic nature allows good access to aqueous solutions. These latter properties markedly improve the quality of immunolabeling (Bogers et al. 1996). In addition, it may well be that the methodology described here might also be adapted for use in other procedures, such as light or confocal microscopy.
Footnotes
Acknowledgements
SLR is a recipient of a BBSRC CASE Award, sponsored by Pfizer Central Research (Sandwich, UK).
We would like to thank our colleagues Mike Wallis and John Armstrong for useful discussions concerning this research. We are also grateful to the former for the provision of bovine growth hormone and its specific monoclonal antibody. In addition, we would like to thank Colin Robinson, Richard Bazin, and Vikash Malde (Pfizer Central Research, UK) for provision of the Pin1 expression construct and for their assistance with hEPVH and Pin1 expression and purification. We also thank Chris Kowalczyk for carrying out the mass spectrometry. Finally, JRT would like to thank David Randall for looking after the day-to-day running of the lab during work on this project.