This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Aims and Objectives: The aim of this research project was to discover the effects that NMO-IgG has on cells in vitro, in order to deduce the pathogenesis of the disease. This was done by observing the effects complement activation has on cell survival and to determine if any diagnostic potential can be found.
Methods: CHO cells were transfected to express AQP1 and AQP4. The CHO cells where then incubated with either control-IgG or NMO-IgG which was obtained from healthy and NMO diagnosed patients respectively. The cells were then stained for cell death or damage using fluorescent antibody markers and a LIVE/DEAD kit. The stained cells where then analysed under a light microscope and pictures taken of the slides. Cell death and damage was recorded and the data analyzed.
Results: The results conclusively show that the samples with AQP4 expressing CHO cells incubated in human complement and NMO-IgG had a much greater number of cell death than when control-IgG was used. Furthermore it was also recorded that cells expressing AQP1 had similar cell death in the presence of NMO-IgG and control-IgG, which was significantly lower than that of AQP4 cells in NMO-IgG. These results support the theory that NMO-IgG targets AQP4 on the surfaces of cells. It does not however directly show if any complement had been activated or shed any light into the role of human complement in the disease.
Conclusion: This study is a good pilot study to provide an experimental procedure in which the current theories of NMO may be research,, for example why does the body create NMO-IgG. Understanding the role of NMO-IgG will not only increase our understanding of NMO but could potentially enable better method for diagnosis and treatment and also open doors to research into other autoimmune diseases, especially demyelinating diseases.
FAB: Fragment antigen binding
FBS: Fetal Bovine Serum
Fc Region Fragment Crystallisable region
NMO: Neuromyelitis optica
NMO- IgG Neuromyelitis Optica Immunoglobulin G
MS: Multiple Sclerosis
MRI: Magnetic resonance imaging
TR: Texas Red
'Horror autotoxicus' was the term first given to describe autoimmune diseases by Paul Ehrlich. An autoimmune disease is an abnormality whereby the immune system of the host attacks itself. (1) This occurs when there is a break down in the self tolerance mechanisms of the immune system. (1) He however believed it was impossible for the human body's immune system to attack itself due to the horror it would create. (2) It is now known that this is not the case. An autoimmune disease is normally caused by an inappropriate reaction of the immune system towards its own cells. Certain subsets of T cells, known as T regulatory cells, exert great control over the immune system. (2) Unlike immunodeficiency diseases autoimmune diseases are generally caused by polygenic mutations as opposed to single gene mutations, although there are exceptions. (2) Currently over 80 autoimmune diseases have been categorised. (3) In the past autoimmune diseases were classified as B or T cell mediated. Now however this classification is no longer used as it appears both are required for autoimmunity. (2)
Genetic and environmental susceptibility
Much research has been done which links both genetic and environmental factors with increased susceptibility to certain autoimmune diseases. Many studies have shown a higher incidence of autoimmune diseases in monozygotic twins than dizygotic twins. (4) However as mentioned above the polygenic nature of autoimmune diseases means that even if someone has one of the required mutations it is unlikely to have a severe manifestation, as several mutations are normally required for a autoimmune response. (5) It is generally accepted that although a person may have the correct genetic mutations for an autoimmune disease the disease tends to only manifest after exposure to a certain environmental trigger. (5) This is shown by studies which have demonstrated that the incidence of both Multiple Sclerosis (MS) and type 1 diabetes has altered in correspondence with population migration to different locations. (5, 6, 7, 8)
Pathogenesis of Neuromyelitis optica
Neuromyelitis optica (NMO) is an idiopathic autoimmune disorder. (9) It is an inflammatory demyelinating disease of the spinal cord and optic nerve called myelitis and optic neuritis respectively. (10) Symptoms can occur together or separately. (11) Optic neuritis is known to cause complete visual impairment with a low incidence of recovery. (12) Myelitis is more severe as it affects the spinal cord and depending on the location of the lesions can affect the nerves innervating a large part of the lower body causing pain in the limbs and bladder paralysis. (10) In severe case if the lesions travel upwards they can damage the brain stem and cause neurogenic respiratory failure. (12)
NMO has also got a very high rate of relapse of 85%, especially when compared to MS. (9) Women comprise more than two thirds of reported cases of NMO with a mean age of onset in the late 30's, (20) ten years later than the average onset of MS. (11) Recent studies in population variance have shown that the disease is far more predominant in Asians than in Caucasians with 30%-40% of MS cases in Asia thought to be attributed to NMO. (14) As with most autoimmune diseases there appears to be an abundance of environmental triggers. Many case studies have been published describing patients who had suffered from either syphilis, HIV type 1, Chlamydia and varicella, and have then gone on to develop symptoms of NMO. (14, 15, 16,) Therefore these instances are described as post- infectious NMO.
Figure 1: An MRI of the cervical spine and upper thoracic cord in T2-weighted (left) and T1-weighted (right) modalities in a patient with NMO. The images show showing longitudinal extensive transverse myelitis with swelling, necrosis and linear gadolinium enhancement.(10)
Discovery of NMO and advancement
NMO was first discovered in the late 19th century by Eugene Devic. (17) The first patients he saw with the condition all presented with loss of vision, paralysis and loss of bladder control. The most striking characteristics observed by Devic, were that these same patients has also developed non-organ specific autoantibody conditions. (18) These conditions were found to be a result of inflammation in the spinal cord and optic nerve. This discovery was made during post-mortems which showed immunoglobulin deposition and complement activation. (18) For many years after Devic recorded his findings, NMO was still mistakenly classed as a subtype of MS (10) but it was not until Clifford Allbutt in the late 19th century studied Devic's work that he associated the NMO symptoms with an entirely separate disorder. (19)
The biggest achievement in NMO so far has been the discovery of the NMO-IgG autoantibody. It was discovered in 2004 by Lennon VA et al. (20) They reported a detection rate of the autoantibody in almost 75% of patients already diagnosed with NMO and in almost 50% of patients who didn't have the condition but were deemed to be high risk of developing it. In other scientific research it has been shown to have a sensitivity of 73%, and can distinguish NMO from classical MS with an accuracy of 91%. (21) As NMO-IgG is quantifiable it allows for the measurement of the effectiveness of treatment as well as disease progression. (20) NMO-IgG is an autoantibody specific to the Aquaporins 4 (AQP4) protein found on cell surfaces in the brain.
Differences between NMO and MS
As mentioned above for many years NMO was mistaken as MS, (10) and even today many people in Asia still refer to it as optic-spinal MS. (14) The main distinction NMO has over MS is the detection of an autoantibody to the water channel protein AQP4. (20) Although the NMO-IgG was only discovered in 2004, there has been much clinical and neurological evidence supporting the theory that MS and NMO were two different disease entities even before the discovery of the autoantibody. When viewed with MRI, typical MS often appears as widespread regions with large amounts of inflammation in the white matter, which also appear more commonly in cerebellar and periventricular regions. (14, 22) This is contrary to NMO where the lesions are typically seen in the diencephalon. (16) Moreover it is very rare to find lesions in the white matter in patients suffering from NMO. (16)
Clinical diagnosis criteria
In the past the diagnosis was based on the presence of clinical characteristics such as optic neuritis, bilateral motor deficits, pain transverse myelitis and multiple other symptoms. (19) Recently however an auto-antibody for AQP4 has been discovered and called NMO-IgG. (23) It is significantly accurate and is now considered an accurate marker for NMO. (19) This has lead to the belief that the onset of disease may be caused or exacerbated by the immune system. (10) Current criteria, requires the presence of both optic neuritis and myelitis (9) (see Figure 2) as well as other factors. Some clinicians believe the factors are too strict, such as Saida (24) who believes that this criterion puts patients at harm as it makes early diagnosis more difficult and hence reduces the possible benefit of treatment.
Figure 2: Criteria for diagnosis of NMO. Adapted from Wingerchuck et al. (9)
Treatment of NMO and prognosis
Currently treatment for NMO is still limited and similar to that used in MS. (25) There are two long term treatments widely used, these are the use of interferon beta and immunosuppressive therapy. (15) Current research suggests that interferon treatment is significantly less effective than immunosuppressive treatment. (26) Both treatments however still fail to halt disease development. (25) A new drug being trialled, called Rituximab, is a monoclonal antibody which works against CD20+ cells (important in the inflammatory response against the spinal cord) and has shown promising results. (27) Prognosis for NMO is very poor with 50% of patients becoming wheelchair bound and 62% being functionally blind within 5 years. (25) Therefore the sooner the diagnosis the better the outlook for the patient. (28) This is why characterizing the role of NMO-IgG is so vital, as it could have significant therapeutic benefits.
Aquaporins general properties
Due to the discovery of NMO-IgG there has been increased interest in the role of Aquaporins (AQPs) within the brain and the difference between the different isotopes. It has become increasingly important to fully understand the functions and roles that water channels have within the brain as strong evidence is emerging which suggests that the AQP4 molecule is targeted in NMO. (23) AQPs are a family of transmembrane molecules with six membrane domains forming a ring, with a pore in the middle within which water molecules can be transported in both directions (29), this is illustrated in Figure 3. The majority of work carried out on AQPs has been done on AQP knockout mice. Studies by Song, et al. (30) showed that deletion of the AQP5, an AQP known for its role in sweat gland function, led to decreased volume of sweat and a more hypertonic fluid. In general AQPs increase water transport down osmotic gradients, with AQP knockout mice unable to concentrate urine, (29, 31) consequently showing that APQs help to regulate water movement and homeostasis.
Figure 3: a) Shows the monomeric structure of AQPs with the six different domains labelled H1-H6. There is a small gap between them to allow water molecules in and out. b) Shows the tetrameric arrangement of AQPs, made up of 4 monomers so it can allow the passage of water molecules in and out. (29)
Discovery of Aquaporins
AQPs were initially discovered by Petre Agre in 1988 (32) who showed that frog oocytes which expressed AQPs, were far more susceptible to water lysis than those not expressing AQPs, hence showing improved water transport. (33) Furthermore it has been shown that where brain oedema has occurred, the astrocytes associated with that region will show an increased expression of AQPs on the cell surface. (34) Thus far over 10 different subtypes of AQP have been isolated in mammals. (29) The most abundant AQP in the human brain is AQP4, which is also the AQP targeted by NMO-IgG; hence it is this AQP which has most significance in NMO. (34) Brain AQP4 is mainly found in astrocyte foot processes, brain parenchyma and major fluid compartments. (35)
M1 vs. M23
AQP4 is further subdivided into an isoform of M1 and M23. (36) The two isoforms, M1 and M23, are due to translation initiation differences in the N termini of the first and second methionine amino acid respectively. (37) M23 is also 3 times more abundant in the human brain than M1. (38) Functionally no differences have been noted between the 2 isoforms, however they do appear to alter the organization of AQP4 on the intramembrane. (36) This is thought to be due to the M23 isoform forming large square arrays with abundant cross bridges, whereas the M1 isoform appears to restrict square array assembly. (36) Recent experiments have shown that addition of NMO-IgG to AQP4 expressing cells have resulted in inflammation, (23) and this is now the current point of interest.
The immune system
The innate and adaptive immune system
The immune system is described as a collection of mechanisms inside an organism which protect against disease by identifying and eliminating a diverse number of pathogens. (39) The human body's immune system is divided into two main categories, the innate and the adaptive immune system. Whilst they are split into two categories they work together in synergy, and rely on one another to work to provide maximum protection for the human body. (2) The innate immune system is antigen non-specific and has a rapid response. (2) It is the body's first line of defence against infection. Included in the innate immune system are all the body's natural barriers, phagocytes and complement amongst others (2). The adaptive immune system on the other hand is far more antigen specific but has a much slower response, taking days to be implemented. The adaptive immune system mainly utilises B and T cells as its main immune components along with antibodies.
Antibody Structure and function
One mechanism of the adaptive immune system is the use of antibodies. These belong to the immunoglobulin (Ig) super family. (2) A standard antibody consists of 2 upper light chains and 2 heavy chains; this is shown in Figure 4. (39) Depending on their isoform they can be found on the surface of cells or circulating freely within the blood vessels. Immunoglobulins are produced from mature B cells and come in the following isoforms, IgA, IgD, IgE,IgG and IgM and each isotope is further divided into subclasses. They are distinguished by differences in their constant heavy chain regions. This is thought to give each isoform its different biological function, with variations in the lighter upper chains resulting in different epitope specificity. When a B cell matures it will begin to produce different antibodies depending on the stimulus which caused it to mature. (2)
The main Ig isotype that this study is concerned with is IgG. It is secreted out of the cell and equally present in intra and extra vascular surfaces. (2) It has the highest half life of all the Ig molecules with an average life span of 23 days. Furthermore IgG is the only antibody which can pass through the placenta to the foetus giving immunity for the first weeks of life. (2) The main mechanisms by which IgG works is via opsonization and activation of the complement system. When an antibody detects a pathogen of correct specificity it binds to it with the FAB regions of the light chains. This leaves the lower heavy Fc region sticking out. Macrophages have receptors which can detect the Fc region and will engulf anything attached to that antibody. This is called opsonization and comes from the Greek opsonin which means to prepare for eating. (2) Furthermore once an antibody is bound both to an antigen and a phagocyte it activates the phagocytes and causes release of Interleukins and cytokines, which mediate an inflammatory response. This is thought to be the cause of inflammation seen around the site of lesions in NMO patients.
Figure 4: Schematic representation showing the structure of different immunoglobulin molecules. (39)
Another way in which antibodies protect the human body from disease is by activating complement. The complement system is a set of about twenty proteins circulating in the blood. It can be activated via three pathways, the alternate pathway, the lectin pathway and the classical pathway. Eventually all the pathways converge but the initiation is different. Immunoglobulins are excellent triggers of the classical pathway. Thus this pathway was named classically as it was first of the three pathways to be discovered. (2) The main proteins are called C1 through to C9, however the numbers do not denote order of activation rather the order in which they were discovered. When two IgG molecules are bound to an antigen they cause C1 to cleave C3 and thus begin the cascade. The next complex will result in the cleavage of the next complex with most binding to the pathogens surface. Once C9 is reached a membrane attack complex (MAC) is formed, this is shown in Figure 5. (40) This molecule literally punctures the surface of the pathogen and causes lysis. (2) A study conducted by Shen et al. (41), showed that complement activation caused cytotoxicity and led to neuronal death. Again it is the activation of this lytic function which is thought to be the cause of lesions and demyelination in NMO, and it is this increase in cell death which we shall be measuring.
Figure 5: Diagram shows the outlines of the three different pathways taken in activating complement starting from recognition and ending in lysis. (40)
Project Aims and hypothesis
This project aims to establish whether NMO-IgG has the ability to activate complement in vitro. This will be seen through increased cell death as the activation of complement has been well established as being cytotoxic. (41) The hypothesis of this study is that there will be a statistically significant increase in cell death when the AQP4 expressing CHO cells are incubated with NMO-IgG and human complement indicating the activation of complement.
2.1 Isolating the IgG from the serum of patients.
The IgG used in this experiment was obtained from the serum of five different patients. The serum was collected by a nurse who had the relevant clinical training and ethical approval. The serum samples taken were then labelled P1 through to P5. The patients NMO-IgG was being collected from had an established diagnosis of NMO and had also been previously diagnosed with a strong AQP4 antibody serum positivity. For the Con-IgG three different pooled healthy patients were used. Their sera was collected and labelled C1 to C3 accordingly. In this research projected IgGNMO is termed as the IgG isolated from the serum of NMO patients, which contained the AQP4 antibody and IgGCON as the IgG isolated from the serum of non-NMO participants.
2.2 Human complement
Non-haemolysed blood was collected from volunteers in a plain glass tube. This was also carried out by a nurse with the correct clinical training and ethical approval. Once collected the blood was then allowed to clot at room temperature, approximately 25oc, for 30 min. The samples were then centrifuged at 1,000 r.p.m. The serum supernatant was then collected from the sample and aliquoted. Finally the complement was stored at -800C until it was used. When serum is collected in this manner it preserves complement activity. In this research project the human complement collected from the blood is referred to as hC.
2.3 Cell culture
The Chinese hamster ovary (CHO) cells used in this experiment where of the K1 variety. The cells where grown in Ham mixture (see appendix) and were stably transfected with plasmids encoding AQP1 or M23 AQP4. In this research project these cells are termed CHO-AQP1 and CHO-AP4 respectively. They were then grown on coverslips in F12 medium (Ham mixture) with 10 % Fetal Bovine Serum (FBS) from, Invitrogen, Paisley, UK. Tests where then run on the cells to detect the success level of the transfection. 95% of cells were found to express the respective proteins in their plasma membranes. The transfections were kindly carried out by Dr Samria Saadoun, as the stable transfections take 6 months to complete.
The CHO cells expressing AQP1 and AQP4 were then washed in Phosphate buffered saline (PBS) and then fixed in 4% of neutral buffered formaldehyde (NBF) from (Sigma, Poole, UK) for 5 minutes. The next step was to add rabbit anti-AQP1 (1:200, Chemicon-Millipore, Livingstone, UK) or rabbit anti-AQP4 (1:200, Chemicon-Millipore, Livingstone, UK) primary antibody, depending on whether it was the AQP1 or AQP4 cell line. The cells where then incubated in this solution for 1 h at 25 0C. The Cells were then washed again with PBS and incubated with AlexaFluor-linked goat anti-rabbit secondary antibody (1:200, Invitrogen, Paisley, UK). For AQP4 immunostaining with IgGNMO, live cells were washed with PBS, exposed to IgGNMO diluted 1:200. They where incubated for 15 minutes at 4oC in PBS, with the addition of 5mM of dextrose. After that they were then washed again with PBS, and post-fixed in 4% NBF. After this they were washed with PBS and incubated with Texas-red-linked anti-human IgG secondary antibody bought from Vector Laboratories, Peterborough, UK. Finally after secondary antibody incubation, the coverslips were washed with PBS and mounted in Aquamount medium with 4',6-diamidino-2-phenylindole (DAPI) also bought from Vector Laboratories, Peterborough, UK. This concluded the preparation of the cells. The Coverslips were examined using a BX-51 Olympus epifluorescence microscope.
2.5 Complement activation and cell viability assays
For complement activation the CHO cells were placed on cover slips which had been exposed to F12 medium (Ham mixture) at 37 0C. Serum containing (by vol.) 5 % IgGNMO or IgGCON and 5 % hC was added to the cover slips and left for 2 hours. After the two hours had elapsed the cells were stained with a LIVE/DEAD® cell viability kit bought from Molecular Probes - Invitrogen, Paisley, UK. According to the manufacturer's instructions, the live cells would stain fluorescent green and dead cells with damaged plasma membranes would be stained fluorescent red. Once again these were also placed on coverslips and examined using a BX-51 Olympus epifluorescence microscope and the number of red and green cells were counted and analysed.
2.6 Statistical Analysis
2.6.1 DEAD/LIVE staining
All the staining using the DEAD/LIVE kit was done 8 times for each of the 3 conditions. This meant the mean, standard deviation, standard error of the mean had to be calculated. In addition analysis of variance (ANOVA) was also used to compare the 3 data sets.
As mention the staining with the LIVE/DEAD kit was repeated 8 times. Therefore the mean of the live and the dead cells where calculated. This allows a mean of cell death and survival.
2.6.2 Standard deviation.
In this experiment the standard deviation was also carried out of the cell death and cell survival. The standard deviation measures the spread of data around the mean The smaller the standard deviation the closer the majority of the points are to the mean, meaning are less outliers or anomalous results.
188.8.131.52 Standard error of the mean.
The standard error of the mean was calculated from the standard deviation. Due to the amount of repeats done it was more appropriate to use the standard error of the mean as it gets smaller with more data, whereas the standard deviation is not affected by population size. The standard error of the mean (S.E.M.) was used as error bars for the graph.
This was used to compare the significance of variance between the 3 data sets used. It was useful in showing if the variance between the 3 conditions was due to random probability or statistically significant.
3.1 Immunocytochemistry staining
Before any experiments could be conducted, it was imperative to confirm the presence AQP1 and AQP4 on the CHO cells. This was done by first incubation with the primary antibody of rabbit-anti AQP1, followed by incubation with secondary anti- rabbit conjugated to Texas red fluorescent dye. This would allow us to see if the AQPs would be tagged as it would appear bright red.
Figure 6: Staining for AQP1
Figure 6 shows the AQP1 CHO cells incubated with Primary rabbit anti AQP1 followed by the secondary anti- rabbit conjugated to Texas Red dye. The cells were also stained with DAPI which stains the cell nucleus blue. The image shows an abundant amount of red staining on the membrane of the cells.
Figure 7: Staining for AQP4.
Figure 7 shows the AQP4 CHO cells incubated in primary rabbit-anti AQP4 followed by Texas red anti-rabbit. The cells were also stained with DAPI which stains the nucleus blue. The image shows clear red staining on the cell membrane.
From Figure 6 and 7 it can be seen that the transfection of the CHO cells to express AQP1 and AQP4 was successful. This can be seen with the Texas red staining on the cell periphery, meaning they have been tagged by the anti AQP1 or AQP4 antibody respectively.
After confirming the success of the transfections, the live CHO cells were then incubated with either Con-IgG or NMO-IgG conjugated to the TR dye, hC and a nucleus stain (DAPI).
Figure 8 AQP1_Con-IgG_hC
Figure 8 shows the incubation of live AQP1 expressing CHO cells in the presence of Con-IgG and human complement. TR linked anti human IgG was then added along with DAPI which stains the nucleus blue. There was no staining of AQP1 showing that Con-IgG does not target AQP1.
Figure 9 AQP4_ConIgG_hC
Figure 9 shows the incubation of live AQP4 expressing CHO cells in the presence of Con-IgG and human complement. TR linked anti human IgG was then added along with DAPI which stains the nucleus blue. There was no staining of AQP4.
From Figures 8 and 9 it can be seen that there was no staining of either AQP1 or AQP4 using control IgG. This shows that neither is recognised by the Con-IgG and therefore no staining occurred. Again these cells were stained with DAPI so we can see the cell nucleuses and differentiate between the cells. The results were expected as in a healthy individual IgG should not be targeting self cells.
Figure 10 AQP1_NMO-IgG_hC
Figure 10 shows the incubation of live AQP4 expressing CHO cells in the presence of NMO-IgG and human complement. TR linked anti human IgG was then added along with DAPI which stains the nucleus blue. There is no staining of AQP1.
Figure 8b AQP4_NMO-IgG_hC: This image show the incubation of live AQP4 expressing CHO cells in the presence of NMO-IgG and human complement. TR linked anti human IgG was then added along with DAPI which stains the nucleus blue. From the image we can see red staining around the cell membrane. This shows that NMO-IgG targets AQP4. This is in line with the hypothesis and the current research which shows that NMO-IgG targets AQP4 in the brain.
Figure 11 AQP4_NMO-IgG_hC
Figure 11 shows the incubation of live AQP4 expressing CHO cells in the presence of NMO-IgG and human complement. TR linked anti human IgG was then added along with DAPI which stains the nucleus blue. There is visible staining of AQP4.
Figure 10 shows no visible staining when incubated in NMO-IgG. This is because NMO-IgG does not target the AQP1 protein. In Figure 11 however, AQP4 cells have been incubated with NMO-IgG and there is visible staining around the cell membrane. This proves that NMO-IgG targets AQP4 and not AQP1 as previously found in other research.
4.2 LIVE/DEAD staining.
The cells where then stained using the LIVE/DEAD kit. This would allow the accurate quantification of cell survival or death.
Figure 12 AQP1_NMO-IgG_hC.
Figure 12 shows the staining of AQP4 expressing CHO cells in the presence of Con-IgG and human complement. The cells where stained with the LIVE/DEAD kit. Living cells fluoresce green, whilst cells which are dead or have damaged cell walls are stained red. There are a few visible dead cells in this image but the majority are alive.
Figure 13 AQP1_NMO-IgG_hC
Figure 13 shows the staining of AQP1 expressing CHO cells in the presence of NMO-IgG and human complement. The cells where stained with the LIVE/DEAD kit. Living cells fluoresce green, whilst cells which are dead or have damaged cell walls are stained red. There are a few dead cells in this image but the majority are alive.
Figure 12 uses Con-IgG and therefore normal cell death is expected. As seen in figure 12 the majority of the cells are alive with a small minority of visible dead cells. This can be attributed to normal cell death, as the cell cycle was not arrested. In Figure 13 there is a very similar amount of cell death. Again this was expected as Figure 10 shows that NMO-IgG does not target AQP1.
Figure 14 AQP1_NMO-IgG_hC.
Figure 11 AQP4_NMO-IgG_hC: This image shows the staining of AQP4 expressing CHO cells in the presence of NMO-IgG and human complement. The cells where stained with the LIVE/DEAD kit. Living cells fluoresce green, whilst cells which are dead or have damaged cell walls are stained red. There is a huge increase in the numbers of dead cells compared to the control. This is due to the fact that NMO-IgG targets the AQP4 as shown in previous research and in Figure 11.
Images of all slides stained using the LIVE/DEAD kit were taken using a light microscope. The cells where then manually counted for living cells and dead cells. Table 1 below shows the data used.
% of dead cells
% of living cells
Table 1: Table one shows the mean of different conditions used, as well as the SEM. used to draw the error bars.
Error bars derived from SEM
Figure 15: Percentage of cell death and cell survival for the different conditions.
Figure 15 shows the percentage of death and survival for each of the 3 conditions. Cell survival is blue and cell death is red. The error bars are the standard error about the mean (SEM) of the respective data. SEM was used as opposed to Standard deviation (SD), because of the large number of repeats which would make the SEM more accurate as it is affected by population size.
The results obtained (appendix xxx) where used to analyse the variance in the percentage of cell death between the 3 different conditions; AQP4CHO_Con-IgG_hC, AQP1CHO_NMO-IgG_hC and AQP4CHO_NMO-IgG_hC. A standard T test could not be used as It does not compare more than 2 sets of data. Therefore a single factor ANOVA was carried out using Microsoft Excel. The results (appendix xxx) gave a p value of 1.03x10-6. This is considered extremely significant and gives great confidence in the variance of our data.
Prior to this study it was discovered that NMO-IgG is present in the majority of patients presenting with NMO. (20) It was then predicted that the NMO-IgG targets the AQP4 proteins in the cells of the brain causing demyelination and the formation of lesions, yet it was unclear of the exact mechanism. Recent studies have shown that complement activation is the main cause of cell death in NMO. Due to this it is expected that there will be a significant increase in cell death when AQP4 cells are incubated in NMO-IgG and complement. The results obtained from this study appear to support the theory that NMO-IgG can cause complement activation in-vitro, but only in the presence of AQP4 expressing cells.
4.1 Interpretation and analysis of results
4.1.1 AQP1 and AQP4 expressing CHO cells in the presence of Con-IgG and human complement.
AQP1 expressing CHO cells incubated in the presence of Con-IgG and hC, was the first control done. From Figure 8 it can be seen that there was no labelling of the AQP1 protein which was expected as healthy IgG is not known to cause autoimmunity. This result was further confirmed in the next control used AQP4 expressing CHO cells incubated in the presence of Con-IgG and hC. These experiments mimic the conditions found in a healthy person. This allowed us to gauge the percentage of cell survival and death in a 'healthy' person. This was particularly important as cell death by apoptosis had to be factored into the experiment. Using the LIVE/DEAD kit on the AQP4 expressing cells incubated in Con-IgG and hC it can be see that a very small proportion of cell death has occurred Figure 12. The highest recorded cell death in any sample with these conditions was 16.3% (appendix xxx) with the mean for all eight readings at 11.8% (Table 1). The LIVE/DEAD kit was not used for the AQP1 expressing cells. The reason for this is that AQP1 was not labelled with NMO-IgG (Figure 10) and there would therefore be no benefit in staining it with the LIVE/DEAD kit as any results would be too similar to the control.
Whilst the results found were expected, some cell death was recorded.
4.1.2 AQP1 expressing CHO cells in the presence of NMO-IgG and human complement.
This experiment was conducted to see if NMO-IgG had any affects on AQP1 proteins. Current research all supports an AQP4 antibody and no experiments have implicated AQP1. Therefore as was expected there was no increased cell death over the control using AQP4 (Figure 13). As this research project has already shown only AQP4 was labelled by NMO-IgG (Figure 8,9,10 and 11), using AQP1 in this study was in essence its own control as it showed cell death could not be attributed to the actual AQP4 protein.
4.1.3 AQP4 expressing CHO cells in the presence of NMO-IgG and human complement.
Current research has indicated that NMO-IgG targets AQP4 proteins and in turn causes demyelination (20), therefore AQP4 labelling by NMO-IgG and increased cell death were expected. Initially from Figure 11 it can be seen that the NMO-IgG labels the AQP4 protein on the surface of the live CHO cells. This supports the theory that NMO-IgG is selective for AQP4 and not AQP1.Using the LIVE/DEAD kit the mean percentage of cell death was 41.85% (Table 1). This is significantly higher than both the previous experiments. This shows that there is a marked increase in cell death in the presence of NMO-IgG for cells expressing AQP4. On analyses with ANOVA the significance of the cell death between the 3 groups had a p value of 1x10-6 (appendix xxx).
Research has shown that when complement is activated it causes cell death (41). Therefore the increased cell death observed in this condition is very likely to be due to the activation of complement. Furthermore other research (23) has shown that when NMO-IgG was incubated with AQP4 expressing CHO cells, there was positive staining for C5B9 deposition. C5B9 is a product of complement activation and can only be present if complement has been activated. In addition to this, a study conducted by Leinhase et al (42), showed that mice lacking in the ability to activate complement, had a lower recorded neuronal cell death after being induced with an experimental brain injury. This again supports the theory that complement activation is vital for the lysis of neuronal cells.
Whilst there is evidence of increased cell death when NMO-IgG is incubated with AQP4 expressing CHO cells there is no visible evidence that directly proves that this is due to compliment activation or even that complement has been activated.
Due to time and resource constraints, there were many controls which could not be used in this study. This does not discredit the results found but only means that to progress the results further these controls would need to be taken into account. By identifying controls it not only allows for better understanding on how to develop the results but allows potential influences to be considered and evaluated.
The first such control was the use of only M23 AQP4 expressing cells. The reason for this is that currently there is no available fluorescent antibody marker for the M1 isoform. In the human brain the two isoforms found are M1 and M23, with M23 being significantly more abundant (36) however the results are not expected to differ greatly as they appear to be functionally identical. (36) However these tests where only conducted in-vitro with no functionality tests being performed in-vivo.
Another control considered was the use of a C5B9 stain. Time constraints meant it was not possible to do this control. A C5B9 stain stains for the C5B9 molecule which is produced during the activation of complement. This would have shown whether the increased cell death may be attributed to complement with a greater certainty.
A further control would be the use of aquaporin 9. This is also an abundant water molecule in the brain. Whilst there is no evidence that NMO-IgG targets the AQP9 protein, it would be a good control to use as it would further strengthen the hypothesis that NMO-IgG is selective for AQP4 only.
4.4 Further research or investigations.
Whilst the results gathered are encouraging in supporting the hypothesis that NMO-IgG can activate complement in vitro, it does not necessarily mean the same occurs in vivo. Astrocytes are known to be very resistant to complement lysis as they contain many complement regulatory proteins such as CD46, CD55 and CD59 (41). This could potentially alter the results as the CHO cells used in this experiment have no such protection from complement lysis.
Another direction for future investigations would be to attempt to discover the mechanism for the production of NMO-IgG
The aim of this research project was to determine the ability of NMO-IgG to activate complement in vitro. The results obtained from this project appear significant enough to come to the conclusion that NMO-IgG can activate complement in vitro, but only in the presence of AQP4 expressing cells. This is supportive of the previous experiment conducted by Saadoun et al. (23)
The results from the incubation of NMO-IgG with AQP4 expressing CHO cells in the presence of human complement showed a marked increase in cell death over either the control or when AQP1 expressing CHO cells where used. Complement activation has been proven to be lytic to neuronal cells (42) and it would appear that the increase in cell death is due to activation of complement; however this conclusion cannot be conclusively drawn.
NMO is a rare disease, and this project one of very few which have tried to characterise the role NMO-IgG has on complement activation. The findings of this study cannot conclusively prove or disprove the hypothesis that NMO-IgG will activate complement in the presence of AQP4 expressing CHO cells and human complement. While there was a significant difference between increased cell death when NMO-IgG is incubated with AQP4 expressing CHO cells, than with the respective control, we cannot prove that the cell death was due to complement activation and could have been due to other factors. Moreover even if complement was activated, it cannot be said that it was due solely to NMO-IgG activation.
While we were not able to draw unequivocal conclusions about the role of NMO-IgG and its ability to activate complement in NMO, this project could serve as a pilot study providing a good basis and experimental procedure that future studies may be able to use to prove that this cell death is due specifically to complement by simple utilisation of the xxx stain, or perhaps other cytotoxic factors.
With more research, a characterisation of the exact role of NMO-IgG as well as any cytotoxic factors on NMO neuronal cells will enable more specific treatments targeted against these specific factors. With NMO a disease with no cure and few management options, target of these factors may inhibit the progression of the disease providing a more favourable prognosis for sufferers. Furthermore understanding the role of NMO-IgG may not increase our understanding of NMO but could potentially enable better methods for diagnosis and treatment in other autoimmune diseases, especially demyelinating diseases which may have a similar pathophysiology to NMO.