Research, mostly by other people (but some by us!)

    Introduction, including the origin of co-bodies

    Cancer is a colossal research subject with thouands of pages published every day: no-one can read it all never mind summarise it for a website, and we shall not attempt to. For a brief, general account see www.cancerinfo.be. Here we are confined to recent cancer research that relates to co-bodies and research on co-bodies that has some bearing on cancer. Even so it must take many words to express what we have to say and what we say will appear difficult even though the idea is to provide a source rather than a complete description and we avoid both detail and pedantry. This page lets you in to serious, current scientific interests. Stuff will be added as time goes by.

    One start-point was the realisation that anticancer antibodies as well as drugs are rather bad at discriminating cancer cells from normal so that treatments based on antibodies simply do not work in the clinic. This is hard for doctors and scientists to accept because of the deep-seated belief that antibodies are both tight-binding and highly specific - that is, reacting with only one kind of target molecule or cell   -  consider for example how effectively our immune system deals with an infection when it has once been exposed to that particular micro-organism. To counter this prejudice calls for a paragraph about affinity and cooperative behaviour of antibodies.
    Few natural antibody binding sites have an affinity exceeding 10 to power 9 l/mol. This is chickenfeed compared for example with the affinity of avidin for biotin; therefore antibodies could easily have evolved with far higher binding-site affinities than they actually possess; simply there is no need of that to deal with infections, because of cooperation in binding, which is of two main kinds. First, all antibodies have two or more identical binding sites which is just what is needed to deal with something like a bacterium: the sites bind simultaneously to neighbouring but identical epitopes and that gives a much higher total affinity. Second, the immune response produces not just one but a range of antibodies against different epitopes on the microorganism. Binding occurs simultaneously by several different antibodies (to different epitopes) and those antibodies are then themselves bound by other molecules also present in the body fluids so that the whole lot is formed into a polyvalent, polyspecific reaction-product mass. The binding power cannot be described by chemical theory and is properly referred to as avidity - a word which should not be used for the simple reactions of single antibodies no matter how many binding sites are involved.
    A second start-point was the suspicion that each cancer must be unique, meaning not only that cancer of the breast differs from cancer of the colon but that each cancer of the breast differs from all other examples of 'the same' disease. This thinking began with looking at the famous cancer-related mutations such as those in the DNA-maintenance protein p53. Alterations in this gene and protein are present in a majority of all human cancers, but not all and not even in all cases of any recognised kind of cancer*. So the typing of cancers that occupies pathologists for so much of their time does not reflect what is really going on inside the cancer cell: it does not tell us what damage has been done; cancers that differ fundamentally between themselves have the same appearance and are given the same name; different kinds of damage produce similar effects as far as the patient and doctors are concerned. The old thinking is wrong and should be replaced.
    * But do consider cancerinfo.be, nature of cancer, Section 4 for other research published in December 2009 on a new candidate for what might be more nearly a universal cancer gene; HIF2alpha.

     Individuality of cancers is finally an established fact, as of December 2009, and will be discussed again below, but it was not so when co-bodies were invented.

    Drawing together these lines of thinking ( and not only with regard to cancers nor only to antibodies) led to the idea that we might combine DIFFERENT binding sites in a single molecule and thus obtain the advantages of both those kinds of cooperative behaviour, that is both polyvalency and heterospecificity, in a way more amenable to our needs than natural antibodies (which have of course evolved for very particular purposes). Observe that the new 'co-body' has a new specificity, for the presence of both epitopes together on the same target, but has much lower affinity for each alone, properties that can in principle be manipulated to suit our needs. Such a reagent could be made to select for any desired combination of epitopes, and therefore might serve to bind to cancer cells from an individual clone and no other cells whatever.

    Time for another brief digression; on 'specificity' and nomenclature related to co-bodies:-
    It is common and acceptable to describe a binding site as being 'specific' for one particular object, which in the case of an antibody is the corresponding epitope. Then a 'bispecific' construct has binding sites for two different kinds of epitope, but this says nothing about whether those epitopes are present on the same target and IN NEARLY EVERY CASE IN THE RELEVANT RESEARCH LITERATURE THEY ARE NOT. That is to say, these bispecific constructs were made and are used to bring together two distinct targets such as two kinds of cells, or a fluorescent molecule and a marker protein. They are not intended to give increased affinity or a new combined specificity and they are utterly distinct from co-bodies. There is ample room for confusion that can be avoided only by a new or re-defined nomenclature. We use, 'specificity' only to identify the epitope or target while 'selectivity' is a quantitative measure of discriminatory power. 'Conselective' is a new word to mean the kind of cooperation shown in the picture and in fact we may call that 'hetero-conselective' because there are two different kinds of binding site, selecting together. See www.hybridantibodies.com and www.trcboyde.net    And another thing: Bispecific antibodies have a bad name in clinical research on cancer treatment because of poor success and nasty side-effects. Those are not co-bodies, in fact they are more like the original crude antibody agents against cancer of forty years ago.

    Recent Research Developments

    How many mutations, the first malignant clone, and singularity

    Scientists naturally wish to know about the changes that make a cancer behave as it does, what are now called the 'driver' mutations, and have spent most time on them. For co-bodies to be useful requires only that new, detectable epitopes be present in the cancer cell  -  we are not interested in what altered proteins do. So it was very interesting to see mention of 'irrelevant' mutations and then further research showing an enormous number of mutations in all mature cancers tested. It turns out that there are tens of thousands including many of an extreme type such as rearrangements of bigger or smaller segments of chromosomes (only the very largest of which are detectable under the microscope).
    'Science' (2006) 314 p268; (2008) 321 pp. 1801 and 1807.
    ‘Nature’ (2007) 446, p153; (2008) 455, p148; (2009) 458, p719; 462, p1005; and two further papers by the same research group published first 'on-line' (16th December 2009) providing the following round numbers for mutations. The first number given is total detected, followed by those affecting protein structure:-
  Typical 'smoker's cancer' of the lung; 22900; 100.
  Malignant melanoma caused by UV light; 33350; 200.

    Against this background, the concept of 'individuality' or 'singularity' of cancers has been developed further, see 'Medical Hypotheses' (2009) 73, p503. This paper suggests that a key stage of cancer development should be recognised in principle -  the stage at which one clone of a developing tumour becomes unmistakably malignant -  and proposes that the whole list of mutations then present be termed the malignant-clone-defining mutation set (McDMS). This word is intended to include all mutations, drivers and passengers, functional or not, and including even silent mutations in which a DNA change does not alter the protein produced in any way. It was predicted that the McDMS would number hundreds, of which tens might be detectable by co-bodies or the like.

    'Nature' (2009) 461, p809 is another very important paper. The subject matter is an ovarian cancer which came back 9 years after surgical removal and where DNA studies were possible on both the new and the old specimens. The authors show, with perfect precision, 30 mutations in coding regions of the DNA in the recurrent cancer; and 11 of them were present also in the primary. Therefore, all those 11 mutations were present in the clone of cells that lay dormant for so long. This is the first listing of mutations from a McDMS, though in this case the only mutations listed are those in coding regions of the DNA. The real total McDMS must have been at least one hundred times greater and both are minimal estimates. (However, the authors misinterpreted their own results, suggesting that 6 of the group of 11 mutations were unimportant because present in only few copies in the primary. Nothing that hides away for 9 years can be unimportant.)

    The relevance of all this for co-bodies? If it is possible to detect all cells that are descended from the first malignant clone, then all can be destroyed. It is necessary only to have co-bodies capable of detecting simultaneously a sufficient number of the McDMS, three or four might do. The principle of clonality leads us to expect that all mutations present in the McDMS would be present in all descendant clones, with only unusual exceptions.

    The Matrix, and Stem Cells

    The characteristic cells of a tissue interact both with other kinds of cell round about and the non-living framework laid down by those cells, forming a cooperative whole, exchanging materials and messages. This is a pattern set down during the process of specialization. We can identify a tissue by looking at it under the microscope, taking into account the appearance of both the characteristic cells and the supporting cells; the overall pattern. Much of this behaviour is continued in a developing cancer; so that until an advanced stage its tissue of origin may be readily recognizable, and there are many cases where the actual cancer cells form only a small proportion of the whole tumour. The normal tissue organization remains to some extent; that is, even cancer cells are partially responsive to their neighbours, and may exploit them to assist growth (‘Nature’ 2007; 449 , p557).
    Three surprising conclusions follow: a] The appearance, classification and even clinical behaviour of a cancer depend as much upon the tissue where it began as upon the exact nature of the mutations and the consequent protein alterations that make it cancerous. b] We may have success by attacking the supporting tissue not only the cancer cells themselves. c] The malignant cells we see down the microscope and think of as representing the cancer may not be the most important in treatment, being partially differentiated and partially cooperative; maybe we should be seeking out cells which have not adapted in that way; 'stem cells'.
     We mean cancer stem cells, which have already acquired the really dangerous mutations, but grow slowly and divide infrequently so that they are resistant to chemo- and radiotherapy, yet provide a source of more sensitive successors that form the bulk of the tumour. Even if we kill all these successor cells, the tumour will be renewed by the persistent stem cells; which might explain a lot about cancer medicine. See 'Nature Biotechnology' 2009, 27, p44.
    But presumably these stem cells all carry the McDMS. Again, this sounds like a task for enhanced selectivity, and therefore co-bodies.

    Immune Response to Cancer

    Part of the body’s mechanism for preventing development of cancers is the immune system. Probably the vast majority of tumours that begin to escape from cellular control mechanisms, are soon detected by ‘killer cells‘, with or without the cooperation of antibodies, and invited to die. Recent research (‘Nature’ 2007, 450, p903) reveals another mechanism, wherein the immune system holds miniature tumours in check without actually killing them. We can be sure that these things are important because we see that people with impaired immunity die of cancers that hardly matter to the rest of us. The immune system is under difficulties in this area because it is naturally orientated to detecting things that are definitely abnormal including especially small parts of proteins that were either totally foreign to the cell or present in unusual quantities or in unusual circumstances. Cancers, however, originate within ordinary cells, so most of their proteins are wholly normal, and they also evolve ways of suppressing or adapting the immune response (Nature 2009, 457, p.102; Nature Biotechnology 2008, 26, p.1348). Immune activity against cancers surely exists, yet not one effective anticancer antibody has ever been made artificially, isolated or studied in detail so as to understand exactly how it might act. We really cannot see how and why the immune system works even as well as it does.
    Therapeutic cancer vaccines have limited success in a few tumours especially by way of stimulating cellular immunity (The Lancet 2009, 373, p.673) and there are recent suggestions about how this might work better (ibid, p.1033). Antibodies loaded with radioisotopes or toxins and directed towards tumour cells have had virtually no success, as mentioned elsewhere. Vaccines that prevent infection by tumorigenic viruses are altogether different since they act long before any tumour or any kind of precancerous lesion has begun, they are not therapeutic but preventative.