3.5 ANÁLISIS TÉCNICO
3.6.2 DETALLE DE EQUIPO LINSERVER UCNC
RBCs taken from infected Balb/c mice were investigated for the presence of surface Ig and C3. Uninfected Balb/cs were used as the negative control and uninfected NZB mice (which suffer from an antibody-mediated spontaneous autoimmune haemolytic anaemia) as the positive control.
Blood samples from the three groups of mice were analysed using the Direct
Antiglobulin Test (DAT) or Coombs test, which involves incubating the RBCs with antibody raised against mouse antibody or C3 — agglutination of the cells indicating the presence of mouse immunoglobulin or C3 on the erythrocyte surface. The infected Balb/cs were tested on day 7 and 14 post infection (approximately 5% and 40% parasitaemias respectively). Three samples from each mouse in the three groups were tested; one for the presence of IgG (using rabbit anti-mouse IgG), one for IgM (using rabbit anti-mouse IgM) and one for surface C3 (using sheep anti-mouse C3). The level of agglutination was then graded from 0 for a negative sample to 3 for a strongly positive sample, the mean for each group expressed to 1 decimal place (Figure 5.3.a). Initially tanned RBCs coated with mouse serum were tested as the C3 positive control, however this did not give any signs of agglutination when incubated with the
anti-C3 antibody. Fortunately NZB RBCs when tested gave a positive result for C3 and were subsequently used as the positive control. None o f the uninfected mice showed any signs of RBC agglutination when treated with anti-IgG, anti-Ig-M or anti-C3. Only one mouse in the infected group at 7 days post infection gave a positive result (1 +’ve (visible to the naked eye)), which when tested at 14 days post infection gave a negative result. At 14 days post infection there was another positive mouse (0.5 +’ve (only visible under the microscope)) which had not been positive at the previous time point. The means of the infected group showed no significant difference from the uninfected control at either the 1% or 5% level (Figure 5.3.a.).
A NTI-IgG ANTI-IgM ANTI-C3
B alb/c Infected ( 5% para. ) 0 0.2 ( ± 0 . 2 ) 0 B alb/c U ninfected ( N egative Control ) 0 0 0 N ZB ( Positive C ontrol ) 2.2 ( ± 0 . 3 ) 1.3 ( ±0. 2) 1.1 ( ± 0 . 2 ) B alb/c Infected ( 40% para. ) 0 0.1 ( ± 0 . 1 ) 0 B alb/c U ninfected ( N egative Control ) 0 0 0 NZB ( Positive C ontrol ) 1.8 ( ± 0 . 3 ) 1.1 ( ± 0 . 3 ) 0.3 ( ± 0 . 1 )
Figure 5.3.a. Agglutination of RBCs treated with anti-IgG, anti-IgM or anti-C3. Level of agglutination given a value of 0 to 3; 0 = no agglutination, 0.5 = agglutination only visible under microscope, 1 -3 = varying degrees of agglutination visible to the naked eye. Values are means ± SEM (n=6).
At the first sampling (d=7) all three mean values for the NZBs (positive control) were significantly different (1% level) from the uninfected Balb/cs (negative control). At the
significantly different (1% level) from their respective negative controls, however the value for the anti-C3 treated mice was only significantly different from the uninfected Balb/cs at the 5% level. Therefore it was clear that when looking at individual samples IgM could be detected occasionally at very low levels on the RJBCs of certain infected Balb/c mice, but not consistently throughout the course of the infection. This is not surprising as even the NZB mice show some variation in the degree of positivity of their Coombs tests at different times of sampling. Unlike the infected Balb/cs however the levels of IgG and IgM for the groups as a whole are significantly different from the negative control. The levels of C3 on the surface of NZB erythrocytes were lower than levels of Ig and showed variation between the two points of sampling. Some NZB mice at the second sampling, that had previously given positive results, showed no signs of agglutination. However all of the NZB samples did show a significant difference at either the 5% or 1% level when compared to the negative control. None of the infected Balb/c mice gave any degree of agglutination when treated with anti-C3 at either of the two time points. To ensure that the positive agglutination gained with anti-C3 treated NZB blood samples was not just a result of non-specific binding to the autoimmune Igs on the erythrocyte surface, Ig coated RBCs from normal Balb/cs were tested with the anti-C3.
Normal Balb/c RBCs were washed and resuspended in PBS and then incubated with an IgG anti-mouse RBC monoclonal antibody (1° antibody), produced from an NZB mouse. The Ig coated normal erythrocytes, which did not have any surface C3, were then treated with either goat anti-mouse IgG or the sheep anti-mouse C3 (2° antibody). Agglutination was only shown by the samples treated with the anti-mouse IgG
1 0^ 1 0^ 1 0^
Fluorescence Intensity
10
Figure 5.3.bi F luorescence intensity o f a blood sam ple tak en from an infected B alb/c m ouse, treated w ith a fluorescent an ti-m o u se Ig G , A and M. M ark er (at
1
o')
set u sing a blood sam ple from an uninfected m ouse, sh o w in g upperlevel o f b ackground fluorescence.
1 0^ 1 0^ 1 0^
Fluorescence Intensity
10
Figure 5.3.bii F lu o rescen ce inten sity o f a b lood sam p le tak en from an N Z B
m ouse, treated w ith a flu o rescen t a n ti-m o u se Ig G , A and M. M arker (at
1
o')
set using a b lood sam ple from a n orm al B alb/c m ouse, show ing(2° antibody), all samples tested with the anti-mouse C3 (2° antibody) were negative. This proved that the positive results gained with the NZB RBCs incubated with anti mouse C3 were indeed due to the presence of surface bound C3 and not as a result of non-specific binding of the antibody to the surface bound anti-RBC antibodies.
Blood samples from these same mice were also analysed for erythrocyte surface Ig by FACSscan analysis. The blood was incubated with FITC conjugated polyclonal rabbit anti-mouse IgG, A and M and the mean fluorescence of the samples compared to those treated with the irrelevant control. Figure 5.3.bi and 5.3.bii are the results of individual mice from the infected Balb/c (17 days post infection) and the uninfected NZB groups respectively, to give an indication of the two extremes. The mean fluorescence intensity (MFI) of the infected Balb/c sample treated with anti-IgO,A,M did not show any shift when compared to the sample treated with the irrelevant antibody (MFI of 2.4). The NZB sample showed a shift from an MFI of 2.01, for the irrelevant antibody, to an MFI of 59.25, for the anti-IgO,A,M antibody. Again there was some variation when
comparing individual mice, however none of the infected Balb/c group showed as great a shift as the NZBs.
Balb/c Infected ( 40% para.)
1.15 (SE ±0.45)
17 Days Post Infection
Balb/c Uninfected
2.18 (SE ±2.82)
NZB
9.57 (SE ±2.20)
Figure 5.3.C MFI shift of samples treated with anti-mouse IgO,M,A (FITC conjugate)
compared to irrelevant antibody (FITC conjugate) treatment. Values are means ± SEM (n=6).
Samples were tested at both 9 and 17 days post infection, however when the samples at the d=9 time point were tested the FACSscan was not functioning properly which resulted in the RBC population effectively oscillating in and out of the gate, which had been set on the forward and side scatter (FSC and SSC) of the erythrocytes. This resulted in inaccurate readings on the fluorescence measurements. The machine was given a full service which rectified the problem. However, to ensure that accurate data could be obtained for the d=17 samples, even in the event of a recurrence o f the problem, the thresholds (FSC, SSC) for the machine were set for the RBC population but a gate was not applied. Fortunately in this situation a gate is not necessary and this allowed clear and accurate readings to be obtained.
The values for the MFIs of the infected Balb/cs showed no significant difference (5% and 1% level) from those for the uninfected mice. The values for the NZBs, however, showed a significant difference at the 5% and 1% level when compared to the
uninfected Balb/cs (negative control). This again confirms the absence of antibody on the surface of erythrocytes in infected Balb/cs.
DISCUSSION
It was shown in the previous chapter that uninfected ^'Cr labelled erythrocytes when injected into infected SCID mice'^^^eliminated from the circulation at an enhanced rate, in comparison to uninfected animals. This removal occurred in the absence of antibody, although not at the same rate or time point as that in infected nude or Balb/c mice. It is possible that although removal is not dependent on antibody, it may be enhanced by it producing the more rapid rate of elimination seen in nude and Balb/c mice. For this to be the case the nude mice would have to be producing T-independent (TI) antibodies and as has been previously mentioned it is believed that P. yoelii does not posses any TI antigens (Taylor et a l, 1982). However, Fossati et a l studying anti-DNA antibodies during P. yoelii infection state that elevated IgM levels are detected in nude mice although they are 2-3 fold lower than.euthymic animals (Fossati, Merino & Izui, 1990).
A
Therefore to reinforce the existing evidence against antibody involvement, large
quantities of serum were transferred from infected immunocompetent mice (Tuck mice), at the peak of their infections when anaemia was pronounced, to infected SCID mice over the period d=8 to d=l 1 post infection. If antibody was involved in augmenting removal then the doses of ‘peak’ serum should have had some affect in accelerating removal. The SCID mice receiving serum, when tested at the end of the experiment (12 days after the last serum injection) still had elevated Ig levels that were approximately the same as those for normal Balb/c mice (data not shovm). Despite the presence of high levels of Ig the elimination of the ^^Cr RBCs was unaffected, the rate of their removal was not significantly different from that of the infected SCIDs that had not been given serum. Similar serum transfer work involving the transfer of serum from
p. berghei infected donors to normal mice, did appear to have an anti-erythrocyte effect but this was not found to correlate with the presence of anti-erythrocyte antibodies (Schetters et al., 1989). It would appear, therefore, that the premature elimination of uninfected RBCs is not reliant on antibody and that the differences seen between SCIDs, nudes and Balb/cs are probably due to some other mechanism, which is as yet
undefined. As has been previously stated, the differences between the strains could be due to their known cellular abnormalities, which due to alterations in the cytokine network could be having a ‘knock-on’ effect on other processes.
All of this evidence is augmented by the Coombs and FACSscan studies, which
examined the surface of erythrocytes during infection for the presence of Ig or C3. This type of investigation has been carried out by several groups, mainly during human infections, some of which suggest that antibody and/or C3 can be detected on the RBC surface of infected patients and is implicated in the anaemia(Facer, Bray & Brown,
1979; Facer, 1980; Jeje, Kelton & Blajchman, 1983). Other groups do not believe that the incidence of DAT positivity is significantly higher in anaemic patients or that the antibody/C3 plays any role in the haemolysis of uninfected erythrocytes (Abdalla & Weatherall, 1982; Merry et a l, 1986). Animal models have been less extensively studied; there is some evidence that indicates the presence of Ig on the surface of uninfected RBCs (Lustig, Nussenzweig & Nussenzweig, 1977; Hunter et a l, 1980).
The Coombs results indicated that there was not a significant increase in DAT positivity for infected samples when testing for surface IgM or IgG, although one mouse in the group for each time point showed a low level of agglutination for IgM. This is often the
case with human samples, where only some anaemic patients give weak DATs. The FACSscan results, which were looking for total surface Ig and not specific classes, did not give a significant shift for the infected samples, again indicating that surface antibody was not present in significant amounts.
None of the infected samples gave any sign o f agglutination when tested for C3., Hunter
et a l were also unable to demonstrate any C3, despite showing the presence of surface Ig using FACSscan analysis. In conjunction with the results for the CVF treated mice, which showed no difference in the pattern of removal of ^’CrRBCs in comparison to infected mice that were not CVF treated, there would not appear to be any complement involvement in the premature removal of uninfected erythrocytes. As previously mentioned, though, some caution must be taken when interpreting the
decomplementation experiment.
All of the samples, tested by Coombs and FACSscan, were compared not only to negative controls but also to positive controls (NZBs, which develop an autoimmune haemolytic anaemia (Howie & Helyer, 1968) involving Ig and C3 (DeHeer, Linder & Edington, 1978)), unlike any previously published work, and hence it could clearly be seen that all of the techniques were working and what constituted a positive result. Furthermore the FACSscan analysis was run with irrelevant antibody controls with all samples, to give a clear indication of any spontaneous binding of the anti-Ig FITC conjugated antibody, eliminating any false positives.
There have been many inconsistent reports about the detection of erythrocyte surface antibody and/or complement components during infections and their involvement in the mechanism of premature removal of uninfected RBCs. The controversy centres on whether the levels of RBC associated antibody/C3 are elevated in anaemic individuals and whether or not the antibody/C3 has a pathogenic affect. The results from this project suggest that infected mice do not have significantly higher levels of antibody/C3 on the RBC surface and that antibody is not the key factor in this mechanism of anaemia.
It is possible that the increased level of erythrocytes bearing antibody/C3, that is believed by many to be the result of infection and one of the causes o f anaemia, is an indirect result of the infection and does not mediate the anaemia. Complement and immunoglobulin are believed to facilitate the removal of effete erythrocytes (Lutz, Nater & Stammler, 1993) and the erythrocyte is also known to be the major carrier of immune complexes in the circulation (Schifferli et a l, 1988) transporting them to splenic and hepatic macrophages (Comacoff et a l, 1983). In light of this it is unremarkable that RBCs in the circulation can be shown, by some, to have Ig and/or C3 on their surface, or even that the Ig has specificity for parasite antigen (Facer, 1980), as there are high levels of circulating immune complexes during malarial infections (Musoke, Cox, & Williams,
1977; June et a l, 1979; DeGraves & Cox, 1983). Therefore the increased numbers of RBCs with surface Ig, giving positive DATs, may be a result of the increase in immune complexes being transported to the liver and spleen for destruction. There could also be a persistence of the complex carrying RBCs due to the changes in the liver and spleen during infection or simply due to ‘overloading’ of the phagocytes with parasite, RBCs and complexes, reducing the rate at which the hepatic and splenic macrophages can clear the complexes.
A similar situation occurred during the studies by Gupta et al. of the disruption in the normal asymmetric distribution of membrane phospholipids of uninfected erythrocytes during malarial infections (Gupta et a l, 1986). The conclusion was that the infection caused an increase in phosphatidylserine (PS) exposure on uninfected RBCs, resulting in their elimination and contribution to the anaemia. Later work, by the same group, indicated that PS exposure occurs naturally but is normally corrected by the spleen, however changes in the spleen during infection interfere with its ability to maintain normal phospholipid asymmetry (Joshi et a l, 1986). As a consequence there is a build up of RBCs in the circulation with increased PS exposure; the increased numbers being a result of the infection but unrelated to the mechanism of premature removal of
uninfected erythrocytes.
Furthermore erythrocyte bound Igs have been detected in rats made anaemic by bleeding or phenylhydrazine treatment (Zuckerman & Spira, 1961) and Ig:C3 complexes have been shown to bind more avidly to ‘young’ erythrocytes (Shapiro, Pilar & Gershon,
1993). As the anaemia caused by the infection leads to increased reticulocyte
production, the abundant ‘young’ erythrocytes could be binding circulating complexes leading to an increase in positive DATs. If this were the case it should be more evident in murine infections where the affect of dyserythropoiesis is less marked.
The rather erratic detection of surface Ig and/or C3 during plasmodial infection could therefore possibly be a result of fluctuations in the elimination of immune complexes or variations in reticulocyte levels. Whether this is the case or not, the results in this and the preceding chapter in conjunction with other reported evidence casts substantial
doubt on the involvement of antibody in the premature removal of uninfected erythrocytes. If the uninfected RBCs are binding Ig/C3, then it is not consistently detectable and occurs at very low levels. It has also been shown that the serum antibody produced is not pathogenic to the uninfected RBCs as passive transfer does not enhance the rate of removal of ^*CrRBCs in infected SCID mice.