3.1.1) Anisomycin modulation of DC maturation
The central focus of this study was to investigate the effect of anisomycin on modulation of DC maturation. In summary the central postulate from chapter one was the idea that DC could be directly matured by stress inducers, i.e. osmotic shock, viral infection and toxins. As DCs provide a pivitol sentinel cell role in immunity, linking innate immunitiy to adaptive immunity, it was speculated that DC could use the stress response conserved through evolution to transduce ‘stress’ into a meaningfiil immune response. Key kinases involved in the stress response across the eukaryotic kingdom are JNK/SAPK and p38/RK and their homologues. As outlined in chapter one, pharmacological manipulation o f p38 and INK by anisomycin was chosen to stimulate a stress response in DC via 28 S rRNA and observe the effects on DC function and phenotype.
The experimental aims of this first investigation were primarily limited to repeating and confirming earher observations made by Dr. K. Rutault, in which in vitro DC had been successfully matured after anisomycin treatment for 24 hours. Maturation had been demonstrated by showing that the DC increased expression of HLA-DR, HLA-DQ and MHC class I/II molecules. Additionally the functional ability of the DC to stimulate allogeneic and autologous T-cell proliferation was also increased after anisomycin treatment. Dr. Rutault had also shown that INK was activated in DC after anisomycin treatment, however very high concentration of anisomycin had been used (l-5pM). Furthermore huffy coat derived DC had been prepared for the kinase investigations, whilst peripheral blood DC were used for the functional and phenotypic experiments. Therefore in addition to confirming the original data, this investigation set out to correlate JNK activation with functional and phenotypic changes. In order to do this huffy coat DC were prepared and tested for functional, phenotypic and kinase activation
C hapter 3 - introduction
The decision was also made to investigate p38 activation in the DC in addition to JNK, as well as monitoring the DC for production of the important immuno-regulatory cytokine IL-12.
3.1.2) Summary of the primary objectives:
1. Compare the phenotypes of buffy coat derived DC to peripheral blood DC. 2. Assess the purity of the DC after immuno-depletion.
3. Establish that p38 and JNK could be activated in DC by anisomycin and other stimulants.
4. Establish whether the DC produced IL-12 in response to CD40 ligand, LPS and anisomycin.
5. Attempt to repeat earlier work showing that anisomycin could be used to induce DC maturation.
3.2.1) Production and purification of buffy coat derived DC
The initial phase of the research was to estabhsh if culturing of the adherent monocytes for 7 days in GM-CSF and IL-4 (see materials and methods section 2.1) produced DC. In order to test the cultures for DC, flow cytometry analysis was carried out. The phenotype of the cells, and the results from this analysis are shown in figure 3.1.1. As can be observed, the cells present in the culture after seven days are divided by size and granularity into two groups.
One representative experiment shown in part(i) shows dotplots of hght scattering as cells pass through the flow cell of the cytometer: the x-axis records light scattered forward and gives a measure of size; while the Y-axis records light scattered sideways and gives a measure of granularity; the dotplot profiles were prepared using winMDI™. In part(ii) of the figure the two distinct populations of cells are outlined for clarity with elipses. The elipses define regions that are assigned as region 1 (Rl) and region 2 (R2). The dotplot profile clearly shows that the cells in the Rl population are considerably larger and more granular than those in the R2 population.
Further analysis of the R2 population is shown in figure 3.1.2. The seven day GM- CSF/IL-4 cultured cells were divided into aliquots, and stained with mouse primary antibodies, specific for known human DC and lymphocyte cell surface proteins. After the primary antibodies had bound and excess antibody had been removed, flurophore (FITC) conjugated secondary antibodies designed to bind to the primary antibodies were added. After staining and thorough washing the cell aliquots were then analysed by one colour flow cytometry using the Facs machine.
C hapter 3 - results
The absence of strong staining for CD 19 and CD 14 also negated the possibility that these cells were monocytes or B-cells. The presence of CD3, class I and weak staining of HLA-DR and DQ strongly suggested that a proportion of these cells were activated T-cells.
Similar profiles o f cell size plotted against FITC were drawn for the Rl population. One representative experiment showing dotplots for each antibody staining is shown in figure 3.1.3. The results show that the cells gated in Rl were all positive for HLA-DQ, HLA-DR, MHC class I, CD la and CD86. A small proportion of the Rl cells were also weakly positive for CD 14, there was no expression of CD3 and CD 19. The strong staining for MHC molecules, CD la and the co-stimulator CD86 showed that the Rl population predominantly contained DC.
The presence of significant numbers o f T cells in the DC cultures (i.e. the R2 population) was not considered appropriate for experiments that were to be carried out: the contaminating T cells may have effected the antigen presentation abilities o f the DC, thus giving misleading results. Therefore immuno-depletion was employed to remove as many non-DC as possible. The cultures were first incubated with anti-CD3 and CD2 antibodies to coat the T-cells and anti-CD 19 to coat any residual B-cells. Iron particles (Dynal) coated with sheep anti mouse IgG were then added, to bind the antibody coated cells. The T and B cells were then seperated from the DC using a magnet. The results shown in figure 3.1.4 show two typical dotplot profiles of the DC after purification by immuno-depletion. Both part(i) and part(ii) show that the majority of the smaller cells in the R2 population have been removed: resulting in a population of DC that was over 90% pure. In addition Part(ii) shows a third population (marked as DC debris) that was observed in several DC preparations. This was provisionally identified as a population of apoptotic DC, and a more detailed discussion of this population can be found in figure 3.2.2.
As a final check on the phenotype of the DC in the Rl population, a phase contrast image was taken using a video-camera connected to an inverted light microscope. The DC were first treated with LPS lOOng/ml, for 24 hours, to observe if typical features of mature DC could be observed. The image shown in figure 3.1.5 shows large granular cells with long 'spikey' processes radiating away from the central body of the cell: these are indeed features that are exhibited by mature DC.
3.2.2) Comparison between phenotypes of huffy coat and peripheral blood DC
DC derived from buffy coats were compared ‘phenotypically’ to DC derived from fresh peripheral blood. In addition slightly different methods were used in the culture of peripheral blood DC as compared to buffy coat derived DC. In the ‘peripheral blood’ culture method, purification was prior to stimulation. This was at this point in time the most common practice with in the laboratory. The stimulation o f DC prior to purification was the method used in some of the initial experiments. The comparison was therefore also designed to observe if similar levels o f maturation were obtained using the two methods.
After 6 days culture in GM-CSF and IL-4, buffy coat derived DC were stimulated for 24 hours with LPS lOOng/ml, or left untreated. After stimulation the DC were then purified by immuno-depletion. Peripheral blood derived DC were purified on day 6 by immuno-depletion, before being stimulated for 24 hours with LPS lOOng/ml or left untreated. The post purification DC phenotypes of each population where then analysed by flow cytometry and conq)ared: see figure 3.1.6.
For each DC population (peripheral blood and buffy coat derived) the results show the average M.F.Is from three independent experiments. On the whole only small
Chapter 3 - results
The increases in HLA-DR and CD86 expression after LPS stimulation were also less in the bufiy coat derived DC than the peripheral blood DC. No increase in MHC I expression was observed with the bufiy coat derived DC.
The increases in expression of HLA-DQ, HLA-DR, MHC I and CD86 were all statistically significant in the peripheral blood derived DC, p<0.05; while in the buffy coat derived DC only the increases in HLA-DR and CD86 were significant, p<0.05. In summary therefore it can be stated that both DC types gave very similar phenotypes with the major difference being that the bufiy coat DC expressed more CD 14 than the peripheral blood DC (the difference was statistically significant at p<0.01) No statistical differences were observed between the DC populations for any o f the other markers.
Further comparisons between the percentage of DC gated positive for each marker were also carried out between the buffy coat DC and peripheral blood DC. The results shown in figure 3.1.6 show that the huffy coat DC population is slightly less positive for CD la and slightly more positive for CD 14 than the peripheral blood DC. The percentage of DC positive for CD86 after LPS stimulation is also less in the bufiy coat DC; none of these différencies were statistically significant.
3.2.3) MAP kinases p38 and JNK activation in DC
The next step in the investigation was to compare anisomycin activation of p38 and JNK in DC to other known activators of p38 and JNK. Seven day GM-CSF/IL-4 cultured DC were used, and immuno-depleted prior to stimulation. Post immuno- depletion, DC were re-plated in a 6 well plate; 3ml samples were prepared at 5x10^ cells/ml. The samples were then stimulated for 30 minutes with the reagents at the concentrations shown in figure 3.2.1. After the incubation period the DC were then analysed for p38 and JNK activation by western blot, and in vitro c-jun phosphorylation assays respectively. The results in part (i) of figure 3.2.1 are fi-om one representative experiment and show that anisomycin (lOOng/ml) is the strongest activator of p38 compared to the other reagents used; this was shown by approximate band intensity comparisons: which correspond to the levels of p38 phosphorylation observed.
In comparison to anisomycin, LPS lOOng/ml and anti-CD40 (mimics CD40 ligand) did not appear to activate p38. UV light and hydrogen peroxide in comparison to anisomycin were moderate activators of p38.
The Western blot for pan-p38 shows that the lanes on the electrophoresis gel were equally loaded with protein. The results in part(ii) from one representative experiment show the effect of the reagents used on JNK activation in the DC. The band intensities correspond to the levels of c-jun phosphorylation observed with these reagents: Anisomycin and UV light were the strongest activators of JNK. In comparison to anisomycin and UV hght, LPS and anti-CD40 did not activate JNK. Hydrogen peroxide also gave strong activation of JNK, but the activation observed was only about half as strong as that observed with anisomycin or UV light. With both experiments (p38 and JNK) quantitative band intensity measurements were not made, because the experiment clearly showed that anisomycin was a strong activator of both p38 and JNK.
3.2.4) Phenotype analysis of DC treated with LPS and anisomycin
Having established that anisomycin was a strong activator of p38 and JNK in DC, experiments were carried out to observe the effect of anisomycin on DC phenotype and function. Maturation of the DC after treatment with anisomycin would strongly indicate that DC could use the conserved stress response involving p38 and JNK as a mechanism o f converting stress into a useful immune response. Comparisons were made to the well known DC maturation factor LPS.
For the first stage of the phenotype analysis, size and granularity were compared between anisomycin treated, LPS treated, and untreated DC. GM-CSF/lL-4 cultured
Chapter 3 - results
The cells were then incubated for a further 24 hours, before purification by immuno- depletion and analysis by flow cytometry. The results in part(i) of figure 3.2.2 show DC dotplot profiles of size versus granularity from one representative experiment.
The profiles show that two distinct populations o f cells could be observed; the DC population designated as Rl and a smaller less granular population of cells designated as R3. In the untreated DC sample the R3 population was small, comprising 3.5% of the total. In the DC sample treated with LPS lOOng/ml (for 24 hours) the R3 population increased to 11% of the total.
With in 24 hours of anisomycin 50ng/ml and lOOng/ml treatment the R3 population increased to 33% and 39% respectively. The results also clearly show that the increase in the R3 population was accompanied by a corresponding decrease in the percentage of total cells present in the Rl population. In the untreated population, Rl comprised 62% of the total recorded events. The percentage decreased to 54% with LPS treatment, and decreased to 28% and 6% with anisomycin 50ng/ml and lOOng/ml respectively. The results in part(ii) show the mean percentages of cells recorded in Rl and R3 from four experiments ± the range. The data shows that a similar pattern of change between Rl and R3 with the reagents was observed in all the experiments.
The DC populations after treatment with LPS or anisomycin were stained for DC phenotypic markers and analysed by one colour flow cytometry: the aim being to observe any differences between the cells gated in regions Rl and R3. Both part(i) and part(ii) of figure 3.2.3 show comparisons between the M.F.I values recorded for the Rl population (red) and the R3 population (green). Part(i) shows comparisons of M.F.I values in Rl and R3 with and without 24 hours stimulation with LPS lOOng/ml.
Part(ii) shows comparisons of M.F.I values in Rl and R3 with and without 24 hours of incubation with anisomycin lOOng/ml. For both stimulations, LPS and anisomycin, it can be observed that the Rl and R3 populations were both positive for CD 14, HLA-DQ, HLA-DR, MHC 1, CD86 and CD la. However the MF.ls, and thus the expression levels o f these markers were far higher in the Rl population than the R3 population. Relative to the control, the expression levels of DC markers HLA-DR, HLA-DQ and CD86 also increased in both Rl and R3 populations after stimulation with LPS or anisomycin.
The increase in expression of DC maturation markers in the Rl population after anisomycin treatment was interesting. To confirm this observation, the phenotype of the Rl population with and with out anisomycin treatment was analysed in three independent experiments. CD la and CD40 were not measured in all the experiments. The average M.F.I values ± S.E.M of the three experiments are shown in figure 3,2.4.
Part(i) shows the comparison of the untreated DC phenotype in Rl to the phenotype of DC treatment with anisomycin 50ng/ml for 24 hours.
Part(ii) shows the comparison of the untreated DC phenotype to the phenotype of DC treated with anisomycin lOOng/ml for 24 hours. The results in part(i) show that treatment with anisomycin 50ng/ml did not substantially alter the expression levels o f any of the phenotypic markers used. There was a slight increase in the expression of HLA-DQ and a slight decrease in class 1 expression, but the increases were not statistically significant, p>0.05. There was no significant change in the expression levels o f CD 14, HLA-DR or CD86.. The results in part(ii) with lOOng/ml anisomycin show similar results to part(i) with the exception that the expression of CD86 significantly increased, p<0.05. There was also a small increase in CD40 expression,
Chapter 3 - results
3.2.5) Functional analysis of DC treated with anisomycin and LPS
The observation o f constistant increases in CD86 expression on DC after anisomycin (lOOng/ml) treatment raised the possibility that the anisomycin treated DC might show enhanced T cell stimulatory capacity, as CD86 is a well documented co-stimulator. In order to further examine this posibility, antigen independent T cell prohferation assays were carried out. After 6 days incubation in GM-CSF and IL-4 the cells were aspirated fi*om the culture plates, thoroughly washed and re-plated with fi’esh GM-CSF and IL-4 at 8-12 million cells per ml. Six well plates were used and 3 ml of cells were added per well. DC samples were then incubated with anisomycin at 50ng/ml and lOOng/ml for 24 hours. As a control one sample was left untreated. The DC samples were then aspirated fi*om the well, purified by immuno-depletion, thoroughly washed and diluted to various concentrations on a 96 well plate.
Purified T cells were then added and the plate incubated for a further 68 hours. The results shown in figure 3.2.5, fi*om three independent experiments and methods, show that anisomycin lOOng/ml treatment and not anisomycin 50ng/ml treatment, enhanced the abihty of DC to stimulate T-cell prohferation above the levels observed for untreated DC. Part(i) shows the result with anti-CD3 treated autologous T-cells, part(ii) shows the result with sodium periodate treated T ceUs and part(iii) shows the result with allogeneic T cells (see section 2.2 for more details).
3.2.6) Increased DC/T cell cluster sizes with anisomycin treatment of DC
Consistant with the observations o f enhanced T cell proliferation and CD86 expression, enhanced DC/Tcell cluster sizes were also observed after the DC had been treated with anisomycin. As enhanced cluster sizes are symptomatic of expanded T cell proliferation and perhaps enhanced DC/T ceU adhesion, an attempt was made to quantify the increases observed with anisomycin. After 6 days incubation in GM-CSF and IL-4, the cells in the DC culture were divided into 3 ml samples at 8-12 million cells per ml, and either left untreated or stimulated with anisomycin lOOng/ml. After 24 hours stimulation the DC samples were purified by immuno-depletion and analysed for cluster size formation with autologous periodate treated T cells.
The result in part(i) of figure 3.2.6 shows an image of DCs clustering with T cells. As can be observed DCs treated with anisomycin lOOng/ml formed larger cluster sizes with autologous T cells than untreated DC.
The results in part(ii) show the mean diameter sizes from three independent experiments ± S.E.M of T cell clustering with anisomycin treated DC and T cells clustering with