Excerpts from Blood Lines

Chapter One
Introduction

Following the sequencing of most human and mouse genes, intense effort is now being expended to establish the function of the products of these genes. For genes encoding secreted proteins, the problem becomes the familiar one of mass-producing the protein product then establishing its actions in vitro and the consequences of injecting or overproducing the protein in vivo. For genes encoding cell-restricted proteins -whether these be adhesion molecules, signaling molecules, modulating molecules, transcription factors, or the like-the problems of demonstrating functions become more complex.

Initially, functional genetic approaches used overexpression of the gene in cell lines or the production of transgenic mice with overexpression of the transgene. However, it became clear that this approach can lead to serious misconceptions regarding the likely function of the encoded molecule due either to the resulting excessively high concentrations of the protein or to the use of inappropriate regulatory sequences, with protein production then occurring either in aberrant cells or at inappropriate times.

Although their generation is more tedious and time-consuming, knock-out mice in which a specific deletion has been made of the gene under study are the current gold standard for establishing the function of a gene product, or at least its non-redundant functions. Often, deletion results in embryo lethality and to progress further, chimeric mice need to be made or conditional knock-out mice developed in which deletions are restricted to certain tissue types. Alternatively, the induction of the knock-out can be delayed until adult life. Further variations on this theme involve insertion of normal or mutated genes into the location of the deleted gene and under the control of the usual promoter elements for the gene.

These increasingly elegant studies have resulted in the production of an embarrassingly large variety of abnormal mice, with stunning or quite subtle pathology that is then in need of characterization and explanation. There are now few workers who are experienced in investigating novel disease states in whole animals. Furthermore, there has been no primer or textbook available to guide the novice into the often puzzling field of mouse disease. To make matters worse, most journals seem now to allow or encourage the publication of such ridiculously small photographs of cells or abnormal tissues that they are quite unidentifiable and have no teaching value. Help is badly needed for the newcomer.

This book deals with hematopoietic disease as occurring in the mouse, often as a result of genetic manipulation. The purpose of this book is to take the reader through the most useful steps that will need to be taken to identify the general nature of the hematopoietic disease in the mouse under study. The investigator may well be faced with the task of characterizing bizarre combinations of diseases of hematopoietic cells coupled with diseases of other organs. This is not a common situation with naturally-occurring disease but is always a possibility following gene manipulation.…

Where hematopoietic disease appears to be present in a novel mouse, the intelligent use of semisolid clonal cultures is central to the initial analysis. This subject will therefore form a major part of this book. Semi-solid clonal cultures of hematopoietic cells have now been with us for 40 years. There are still occasions when they appear to remain an art form, as in the bad old days, but in truth the techniques have long since passed the unpredictable and mysterious stage. There are now available purified recombinant growth factors that allow selectivity in analyzing the behavior of various hematopoietic subpopulations. Although the culture procedures may now be more reliable, errors continue to be made and valuable information not gathered. This book will have served one of its major purposes if fewer opportunities are missed to make important observations using semi-solid cultures.

The approaches and procedures to be described are our current in-house ones that we have learned to rely on over the past 40 years. No doubt alternatives exist in other laboratories and better techniques will continue to emerge. However, the present methods will permit an intelligent approach to the novel diseases that will be encountered in the future and lessons learned from these, in turn, will add to our basic fund of experience and technology.

Chapter Two
The Novel Mouse: Opening Remarks and Initial Investigations

The basic investigational procedures

This book is written by an experimental hematologist and the approach of our group to an unknown mouse is first to examine the state of the hematopoietic tissues. If an abnormality is identified, this will be characterized by further studies. If disease states are also encountered in other organs, the possibility will be carefully examined that these might have been induced by abnormal hematopoietic populations or, conversely, have caused the hematopoietic abnormalities. If this proves not to be the case, and particularly if the mouse has no detectable hematopoietic disease, then mice of this type will be promptly transferred to a more appropriate research group. It is important for an active hematological research group not to become preoccupied or bogged down by studies on other organ systems in which the group has no particular experience or expertise. Exit strategies are important, no matter how intriguing these latter diseases may be.…

Four initial groups of investigations will probably be needed for all novel mice and the data from these should allow provisional conclusions to be drawn and, if need be, provide clues to the further specific investigations that may then be necessary. The four basic groups of investigations are:

Case history, basic hematological investigations such as hematocrit, white cell levels, platelet counts, and autopsy examination. Cytological and fluorescence-activated cell sorting (FACS) analyses of organ populations.
Clonal analysis of hematopoietic populations in culture.
Histological examination of the tissues.
Transplantation of affected hematopoietic populations to normal or irradiated recipients.

Each of these four groups of investigations provides an independent set of data that can then be put together to arrive at a conclusion regarding the normality or otherwise of the hematopoietic tissues...

Chapter Three
Hematopoietic Subpopulations, Their Detection and Regulation

An ability to interpret abnormalities in hematopoiesis requires some understanding of the hierarchical organization of the hematopoietic populations and the manner in which they are normally regulated. Although stem and progenitor cells comprise fewer than 1% of total hematopoietic cells, it is in this small subset that abnormalities will have a dramatic influence on subsequent hematopoietic events. It is necessary therefore to understand the relationships between these precursors and the manner in which their proliferation and differentiation are regulated. The hematopoietic population is usually considered to have a pyramidal shape with one or a few initiating stem cells generating large populations of committed progenitor cells with restricted lineage potential. These progenitor cells in turn each generate large numbers of dividing and maturing progeny that finally are the mature cells released to the circulation (Figure 3.1A). This model has three dominant features (Figure 3.1B). First, it is unidirectional: stem cells form progenitor cells that then form maturing cells, with no reversion possible. Second, there is a progressive restriction of the capacity for self-generation. Self-generation is possible for stem cells, probably not possible for progenitor cells, and certainly not possible for maturing cells. Third, no side movement is possible from one lineage to another, such as from the erythroid to the granulocytic or megakaryocytic lineages.

Figure 3.1 (A) The pyramidal shape of hematopoietic subpopulations, each subpopulation being more numerous than the preceding. (B) The three features of hematopoiesis (a) unidirectional, (b) decreasing capacity for self-generation, and (c) no capacity for lineage switching.

The model fits most observed facts about subsets of hematopoietic cells and has been verified fairly stringently by the use of FACS-sorted hematopoietic subpopulations. Whether it is as inflexible as portrayed is only recently being questioned. The possible plasticity of hematopoietic stem cells and their possible ability to generate cells of other organs are startling observations [20, 21]. However it has yet to be clearly resolved whether the observed cell formation in damaged tissues is based on stem cell plasticity, merely on cell fusion [22] or on the occurrence of organ-specific stem cells in unexpected locations. What these discoveries have raised is some questioning of the validity of the hierarchical hematopoietic population model. Is self-generation really only possible for stem cells or can other cells self-generate under certain conditions? Are maturation stages fixed and irreversible? Even if lineage fidelity is impressive, may it nonetheless be possible for cells sometimes to switch wholly or partly their lineage commitment? For the present, the conventional model serves well enough to introduce the populations needing to be monitored when analyzing an unknown mouse but it needs to be kept in mind that, in a genetically-manipulated animal, different events may occur.

As shown in Figure 3.2, eight distinct lineages of hematopoietic cells are generally regarded as making up the hematopoietic tissues but the spleen is the only organ in the mouse that actually contains all eight lineages. The mouse marrow lacks both mast cells and T-lymphocytes. It can be calculated that the marrow contains more than 95% of all stem and progenitor cells, the small remainder being located in the spleen. However, in some disease states the spleen becomes grossly enlarged and greatly increases its content of hematopoietic precursor cells so that a majority of early hematopoietic cells is now located in the spleen. For this reason, the relative level of hematopoiesis in the two tissues is an important parameter to establish in any novel mouse...

Figure 3.2 The eight major hematopoietic lineages generated by self-renewing multipotential stem cells.

Chapter Ten
Diseases Involving Hematopoietic Tissues

Some basic biology

Experiments performed in the 1980s established in principle what was required for leukemic transformation. It was shown that immortalized factor-dependent cells were not leukemogenic on transplantation to syngeneic recipients. However, if the cDNA encoding the relevant growth factor was transfected into such cells, the acquisition by the cells of a capacity to produce their own growth factor led to immediate transformation of the immortalized cells to leukemic cells [114]. For reasons that still remain obscure, for certain hematopoietic cells, self-generation of growth factors is often one of two key changes needed for cells to become leukemic. Why it should matter whether the cells develop autocrine growth factor production is difficult to explain because in vivo such cells are in an environment already containing relevant concentrations of such growth factors.…

In the discussion to follow, various gene manipulations will be noted as having led to leukemia development. Because the full functions of many of these genes are not yet known, it is difficult to assign to them a role of having induced one or another of the key changes in leukemogenesis. As a guide to critically reading such publications, the reader needs to ask three questions:

Did the authors prove their "leukemias" were transplantable with proof that the transplanted leukemias were derived from the injected test cells and were clonally related in multiple recipients?
Did the authors study the pre-leukemic state of hematopoiesis in their animals to establish a likely role for their gene or manipulation in hematopoiesis?
Did the authors give a full description of the pathology of their mice, including the non-leukemic animals?

Regrettably, many publications fail to provide answers to these questions, leaving doubt about the claimed leukemogenesis, and certainly not providing much information about the state of the hematopoietic tissues prior to disease development. As a consequence, there is much still to be learned about even the most longstanding oncogenes like myc, myb, ras, and bcr-abl...

GM-CSF

Extensive studies have been made using transgenic GM-CSF mice [9, 10] and mice repopulated by hematopoietic cells that have been engineered to produce excess levels of GM-CSF [156]. Both situations result in premature death but the resulting disease states differ according to the model in use.

In the transgenic mice where excess GM-CSF production seemed to be restricted to peritoneal macrophages, the transgenic GM-CSF mice exhibited sustained elevated circulating GM-CSF levels (40-fold) and excessive numbers (up to 100-200 _ 106 cells) of peritoneal and pleural macrophages. Oddly, such mice exhibited normal cell populations in the blood, spleen, bone marrow, and normal numbers of granulocytemacrophage progenitor cells in these tissues [10]. The colonies formed by the cells were of normal size, cellular content and responsiveness to regulators. The peritoneal and pleural macrophages were typically basophilic, enlarged with active cytoplasmic phagocytosis and were often bi- or multi-nucleate-GM-CSF being one of the agents able to induce cell fusion between peritoneal macrophages [9] (Figure 10.4).

Figure 10.4 Peritoneal cells from a GM-CSF transgenic mouse showing a dominant population of macrophages with basophilic cytoplasm and two multinucleate macrophages.

These transgenic mice with excess GM-CSF levels died prematurely with a miscellany of lesions including inflammatory nodules in the peritoneal and pleural cavity composed of macrophages and macrophages transformed to fibroblast-like cells, focal granulomas in skeletal muscle, a blackened necrotic gut (an effect of IL-1 plus endotoxin) and, in all mice, destruction of the tissues of the eye by activated macrophages [157]. Two transgenic lines with similar GM-CSF levels developed distinctly different disease patterns, indicating that the chromosomal location of insertion of the transgene was strongly influencing the pattern of disease that developed...

G-CSF

In animals and man, G-CSF stimulates the elevated production of granulocytes, resulting in excess levels of granulocytic cells in the blood, marrow, and spleen. Even in high doses, G-CSF is relatively non-toxic. Mice with high circulating G-CSF levels have been generated by injecting irradiated syngeneic mice with marrow cells infected with a retrovirus containing the G-CSF cDNA [160]. Such mice developed massive elevations of G-CSF in the serum but did not develop clinical illness over a 6 month observation period. They did develop elevated numbers of granulocytes in the blood (X8) and spleen (X3) but bone marrow cell counts showed no significant changes. Total granulocytemacrophage progenitor cells were not altered in the marrow but were elevated 30- to 80-fold in the spleen. Overall, chronic overexpression of G-CSF led to a 2-fold increase in total body progenitor cells not only in the granulocyte-macrophage lineage but also in the erythroid and megakaryocytic lineages. This was likely to be based on the ability of G-CSF to stimulate stem cells to form progenitor cells in multiple lineages [42]. Up to 10% of the G-CSF-transfected progenitors appeared able to proliferate in vitro in the absence of added stimulating factors...

Comments

The reader will appreciate from these examples that dysplasia is an unfortunate term lying uneasily between hyperplasia and hypoplasia, where the cells in some lineages may have a miscellany of intrinsic faults. For few of the examples is there a clear documentation of the long-term consequences and what agents might cooperate with the abnormalities to result in an overtly recognizable disease state. In few cases have the normal regulatory controls of hematopoietic cells been explored for their possible aberrant actions on the altered cells. In each of the models much vital information remains to be uncovered. Possibly as the information becomes more complete, the need for the term dysplasia may decrease...

Chapter Twelve
Applying Closure

...[A]re studies on mice or mouse disease a waste of time?

It has to be admitted that spontaneous hematological disease in the mouse often does not closely resemble comparable diseases in man and usually has a quite different molecular basis. This is not so true for more recent artificially-induced mouse diseases where a particular human translocation has been inserted into the murine cells and results in leukemia development closely resembling the original human disease. Because of the differences between mouse and man and the increasing ability to work almost as readily in vitro with human as with mouse cells, it can be questioned why it makes much sense to persist with studies on murine diseases. Why not work entirely with human cells…?

The rationale for working with mice is that experimental animals can be produced in large numbers and the various organs can be examined in minute detail in a manner impossible with patients. It is also possible to monitor disease development in the whole animal in a manner that is not usually possible in patients.

In the past, work with murine cells pioneered essentially all the modern techniques applicable to human cells-whether these were the culture of hematopoietic subpopulations, FACS fractionation of hematopoietic populations or the discovery and characterization of regulatory molecules. There is no reason to suppose that this situation will change and this is particularly so for the genetically-manipulated mouse where known lesions can be linked to spectacular disease development.

To date, conclusions drawn from the use of murine systems have been found to be true also for human systems. Murine data are not misleading for man...