Comparative evaluation of performance measures for shading correction in time-lapse fluorescence microscopy.

Time-lapse fluorescence microscopy is a valuable technology in cell biology, but it suffers from the inherent problem of intensity inhomogeneity due to uneven illumination or camera nonlinearity, known as shading artefacts. This will lead to inaccurate estimates of single-cell features such as average and total intensity. Numerous shading correction methods have been proposed to remove this effect. In order to compare the performance of different methods, many quantitative performance measures have been developed. However, there is little discussion about which performance measure should be generally applied for evaluation on real data, where the ground truth is absent. In this paper, the state-of-the-art shading correction methods and performance evaluation methods are reviewed. We implement 10 popular shading correction methods on two artificial datasets and four real ones. In order to make an objective comparison between those methods, we employ a number of quantitative performance measures. Extensive validation demonstrates that the coefficient of joint variation (CJV) is the most applicable measure in time-lapse fluorescence images. Based on this measure, we have proposed a novel shading correction method that performs better compared to well-established methods for a range of real data tested.

A single-cell pedigree analysis of alternative stochastic lymphocyte fates.

In contrast to most stimulated lymphocytes, B cells exposed to Toll-like receptor 9 ligands are nonself-adherent, allowing individual cells and families to be followed in vitro for up to 5 days. These B cells undergo phases typical of an adaptive response, dividing up to 6 times before losing the impetus for further growth and division and eventually dying by apoptosis. Using long-term microscopic imaging, accurate histories of individual lymphocyte fates were collected. Quantitative analysis of family relationships revealed that times to divide of siblings were strongly related but these correlations were progressively lost through consecutive divisions. A weaker, but significant, correlation was also found for death times among siblings. Division cessation is characterized by a loss of cell growth and the division in which this occurs is strongly inherited from the original founder cell and is related to the size this cell reaches before its first division. Thus, simple division-based dilution of factors synthesized during the first division may control the maximum division reached by stimulated cells. The stochastic distributions of times to divide, times to die, and divisions reached are also measured. Together, these results highlight the internal cellular mechanisms that control immune responses and provide a foundation for the development of new mathematical models that are correct at both single-cell and population levels.

B cell differentiation and isotype switching is related to division cycle number.

The mature, resting immunoglobulin (Ig) M, IgD+ B lymphocyte can be induced by T cells to proliferate, switch isotype, and differentiate into Ig-secreting or memory cells. Furthermore, B cell activation results in the de novo expression or loss of a number of cell surface molecules that function in cell recirculation or further interaction with T cells. Here, a novel fluorescent technique reveals that T-dependent B cell activation induces cell surface changes that correlate with division cycle number. Furthermore, striking stepwise changes are often centered on a single round of cell division. Particularly marked was the consistent increase in IgG1+ B cells after the second division cycle, from an initial level of < 3% IgG1+ to a plateau of approximately 40% after six cell divisions. The relationship between the percentage of IgG1+ B cells and division number was independent of time after stimulation, indicating a requirement for cell division in isotype switching. IgD expression became negative after four divisions, and a number of changes centered on the sixth division, including the loss of IgM, CD23, and B220. The techniques used here should prove useful for tracking other differentiation pathways and for future analysis of the molecular events associated with stepwise differentiation at the single cell level.

Generation of splenic follicular structure and B cell movement in tumor necrosis factor-deficient mice.

Secondary lymphoid tissue organogenesis requires tumor necrosis factor (TNF) and lymphotoxin alpha (LTalpha). The role of TNF in B cell positioning and formation of follicular structure was studied by comparing the location of newly produced naive recirculating and antigen-stimulated B cells in TNF-/- and TNF/LTalpha-/- mice. By creating radiation bone marrow chimeras from wild-type and TNF-/- mice, formation of normal splenic B cell follicles was shown to depend on TNF production by radiation-sensitive cells of hemopoietic origin. Reciprocal adoptive transfers of mature B cells between wild-type and knockout mice indicated that normal follicular tropism of recirculating naive B cells occurs independently of TNF derived from the recipient spleen. Moreover, soluble TNF receptor-IgG fusion protein administered in vivo failed to prevent B cell localization to the follicle or the germinal center reaction. Normal T zone tropism was observed when antigen-stimulated B cells were transferred into TNF-/- recipients, but not into TNF/LTalpha-/- recipients. This result appeared to account for the defect in isotype switching observed in intact TNF/LTalpha-/- mice because TNF/LTalpha-/- B cells, when stimulated in vitro, switched isotypes normally. Thus, TNF is necessary for creating the permissive environment for B cell movement and function, but is not itself responsible for these processes.

The effect of correlations on the population dynamics of lymphocytes.

Recent studies of the population dynamics of a system of lymphocytes in an in vitro immune response have reported strong correlations in cell division times, both between parents and their progeny, and between those of sibling cells. The data also show a high level of correlation in the ultimate number of divisions achieved by cells within the same clone. Such correlations are often ignored in mathematical models of cell dynamics as they violate a standard assumption in the theory of branching processes, that of the statistical independence of cells. In this article we present a model in which these correlations can be incorporated, and have used this model to study the effect of these correlations on the population dynamics of a system of cells. We found that correlation in the division times between parents and their progeny can alter the mean population size of clones within the system, while all of the correlations can affect the variance in the sizes of different clones. The model was then applied to experimental data obtained from time-lapse video microscopy of a system of CpG stimulated B lymphocytes and it was found that inclusion of the correct correlation structure is necessary to accurately reproduce the observed population dynamics. We conclude that correlations in the dynamics of cells within an ensemble will affect the population dynamics of the system, and the effects will become more pronounced as the number of divisions increases.

Converse Smith-Martin cell cycle kinetics by transformed B lymphocytes.

Recent studies using direct live cell imaging have reported that individual B lymphocytes have correlated transit times between their G1 and S/G2/M phases. This finding is in contradiction with the influential model of Smith and Martin that assumed the bulk of the total cell cycle time variation arises in the G1 phase of the cell cycle with little contributed by the S/G2/M phase. Here we extend these studies to examine the relation between cell cycle phase lengths in two B lymphoma cell lines. We report that transformed B lymphoma cells undergo a short G1 period that displays little correlation with the time taken for the subsequent S/G2/M phase. Consequently, the bulk of the variation noted for total division times within a population is found in the S/G2/M phases and not the G1 phase. Models that reverse the expected source of variation and assume a single deterministic time in G1 followed by a lag + exponential distribution for S/G2/M fit the data well. These models can be improved further by adopting two sequential distributions or by using the stretched lognormal model developed for primary lymphocytes. We propose that shortening of G1 transit times and uncoupling from other cell cycle phases may be a hallmark of lymphocyte transformation that could serve as an observable phenotypic marker of cancer evolution.

B cell activation, tolerance and antigen-presenting function.

Recent advances in the understanding of the collaboration between T cells and B cells have provided a novel framework within which to analyse the mechanisms of B-cell tolerance and its breakdown. Of particular interest has been the finding that B cell anergy is due to defective antigen receptor mediated functions, while the antigen-processing machinery and CD40-dependent activation pathways are unaffected. Thus, the anergic B cell, which otherwise has a short lifespan, can be rescued by a number of regimes to participate in autoimmune responses.

T-cell stimuli independently sum to regulate an inherited clonal division fate.

In the presence of antigen and costimulation, T cells undergo a characteristic response of expansion, cessation and contraction. Previous studies have revealed that population-level reproducibility is a consequence of multiple clones exhibiting considerable disparity in burst size, highlighting the requirement for single-cell information in understanding T-cell fate regulation. Here we show that individual T-cell clones resulting from controlled stimulation in vitro are strongly lineage imprinted with highly correlated expansion fates. Progeny from clonal families cease dividing in the same or adjacent generations, with inter-clonal variation producing burst-size diversity. The effects of costimulatory signals on individual clones sum together with stochastic independence; therefore, the net effect across multiple clones produces consistent, but heterogeneous population responses. These data demonstrate that substantial clonal heterogeneity arises through differences in experience of clonal progenitors, either through stochastic antigen interaction or by differences in initial receptor sensitivities.

Regulation of platelet lifespan in the presence and absence of thrombopoietin signaling.

UNLABELLED: Essentials We examined platelet survival in models of absent or enhanced thrombopoietin (TPO) signaling. Platelet lifespan is normal in transgenic mice with chronically enhanced TPO signaling. Mpl deficiency does not negatively affect platelet lifespan in the absence of thrombocytopenia. We conclude that TPO and its receptor Mpl are dispensable for platelet survival in adult mice. SUMMARY: Background It is well established that thrombopoietin (TPO), acting via its receptor Mpl, is the major cytokine regulator of platelet biogenesis. The primary mechanism by which TPO signaling stimulates thrombopoiesis is via stimulation of Mpl-expressing hematopoietic progenitors; Mpl on megakaryocytes and platelets acts to control the amount of TPO available. TPO could potentially reduce platelet and/or megakaryocyte apoptosis, and therefore increase the platelet count. However, the effect of TPO receptor signaling on platelet survival is unresolved. Methods and results Here, we investigated platelet survival in mouse models of absent or enhanced TPO signaling. In the absence of thrombocytopenia, Mpl deficiency did not negatively influence platelet lifespan, and nor was platelet survival affected in transgenic mice with chronically increased TPO signaling. Conclusions We conclude that TPO and its receptor Mpl are dispensable for platelet survival in adult mice.

Switch recombination and germ-line transcription are division-regulated events in B lymphocytes.

Treatment of resting murine B lymphocytes with CD40 ligand (CD40L) and IL-4 induces proliferation and a switch in immunoglobulin (Ig) isotype surface expression from IgM and IgD to IgG1 and IgE. Using a fluorescent dye to enable cell sorting according to cell division cycle number, we have examined molecular events associated with B cell differentiation, namely, germ-line transcription and DNA recombination. Digestion-circularisation polymerase chain reaction experiments showed that DNA recombination leading to isotype switching from IgM to IgG1 surface expression is division-dependent and was first detected after B cells had divided three times. Similarly, DNA rearrangement involving the IgE switch region was detectable only after five division cycles. These division cycle numbers correlate with the numbers of divisions required before surface expression of the switched isotype [P.D. Hodgkin, J.-H. Lee, A.B. Lyons, J. Exp. Med. 184 (1996) 277-281]. RT-PCR analyses also revealed that germ-line transcripts for both IgG1 and IgE increased with division number suggesting a threshold expression level may be required for recombination to occur.

B-cell activation by helper T-cell membranes.

Resting B cells can be stimulated to proliferate and differentiate to antibody-producing cells by the combination of cell contact and soluble signals provided by activated primed helper T (Th) cells. The ability of purified plasma membranes from activated Th cell clones and recombinant lymphokines to reconstitute B cell proliferation and differentiation has allowed an increased understanding of B cell activation and characterization of the molecules involved. B cell-Th cell contact appears sufficient for delivering the proliferative signal to B cells in the absence of lymphokines. A receptor ligand pair that plays a critical role in delivery of the contact signal is CD40 on the B cell surface and the ligand for CD40 on activated Th cells. Lymphokines alone do not drive resting B cell differentiation, however, when these soluble signals are delivered during the time of B cell DNA replication, they effect B cell differentiation and isotype switching. Delivery of the CD40-dependent contact signal to resting B cells appears to require a high degree of CD40 crosslinking on the B cell surface. Providing contact signals to naive B cells with recombinant molecules in membrane fractions may allow the generation of methodology to support the production of novel antibodies in vitro.

T lymphocyte function in vivo. Ambivalence of the class I MHC antigen-reactive subset.

Lymphocytes that recognize class I alloantigens (class I T cells) are able to lyse appropriate target cells and release lymphokines in vitro. However the relative contribution of these activities to biological, in vivo functions of these cells is unclear. It is possible to discriminate between these activities using cyclosporine (CsA). CsA inhibits lymphokine release from class I T cells but has no effect on their cytotoxic activity. The in vivo function of class I T cells is analyzed using 2 models; the local GVHR induced by the transfer of sensitized T cells to the foot-pad and islet allograft rejection induced by the passive transfer of sensitized T cells. Both reactions may be mediated by class I T cells. CsA inhibits the in vivo functions of the class I T cells in both systems--hence, these functions appear to be lymphokine-dependent. This demonstrates the ambivalence of this T cell subset in relation to biological function; the cells express direct cytotoxic activity and producing lymphokines. The alloreactivity of the class I T cells is dependent upon the latter activity.

An analysis of tissue-specific transplantation phenomena in a minor histoincompatibility system.

Tissue-specific differences in immunogenicity were demonstrated following allotransplantation across the same minor histocompatibility barrier (BALB/c----DBA/2). In contrast to the high immunogenicity of fetal pancreas and skin, isolated islets, fetal proislets, and thyroid were weakly immunogenic. These tissue-specific effects were not related to the presence of tissue-specific antigens or the absence of recognizable minor alloantigens from the less immunogenic tissues. There was a strong correlation between tissues that were highly immunogenic and those that contained rich populations of donor leukocytes. The survival of fetal pancreas allografts was significantly improved by pretreating the donor tissue in high-oxygen organ culture and by the preparation of fetal proislets. Using other MHC-compatible strain combinations (B10.D2----BALB/c; BALB/c----B10.D2; C3H.SW----C57/10J) strain-specific effects were observed in the immunogenicity of thyroid allografts. In two of three strain combinations (B10.D2----BALB/c; BALB/c----B10.D2), pretreatment of the donor tissue with cyclophosphamide and organ culture prior to grafting significantly improved graft survival. These findings suggest that donor passenger leukocytes may play an important role in determining the immunogenicity of MHC-compatible allografts.

Blocking of delivery of the antigen-mediated signal to the nucleus of T cells by cyclosporine.

Cyclosporine (CsA) inhibits release of interleukin 2 (IL-2) and hemopoietic growth activities such as interleukin 3 (IL-3) from major histocompatibility complex (MHC)-antigen-activated T cells. Production of both lymphokines appears to be coordinately regulated; the antigen dose response, T cell dose response, and time course of lymphokine appearance are similar. The triggering of lymphokine production by these cells is solely dependent on T cell-target cell interaction, as the T cell dose response curve indicates that no cooperation occurs between T cells, and any metabolic contribution by the target cell was eliminated by ultraviolet irradiation. This interaction triggers the transcription of lymphokine-encoding mRNA. The process of lymphokine release can be divided into 4 steps: Antigen binds to the T cell; a signal is transferred to the cell nucleus; transcription of lymphokine-encoding mRNA occurs; and intact lymphokine is synthesized and secreted. CsA inhibits antigen triggered lymphokine production. However, it does not inhibit lymphokine release from the constitutively producing tumor cell lines WEHI-3 (which releases IL-3) and MLA 144 (which produces IL-2). Thus CsA has no effects on the lymphokine secretion process or any direct action upon lymphokine-coding mRNA. CsA does not affect antigen recognition during cell-mediated cytotoxicity. Therefore, CsA acts after antigen binding and before transcription of lymphokine-encoding mRNA. That is CsA blocks the transmission of the antigen signal. This information is used to show that this CsA-sensitive signal is required continuously to maintain the T cell in a lymphokine-secreting state.

The importance of efficacy and partial agonism in evaluating models of B lymphocyte activation.

Immunologists have developed a range of in vitro techniques for probing the receptor mediated response of cells comprising the immune system. An important and ubiquitous method is the use of antibodies in either soluble or aggregated form to engage cell surface receptors and transmit a signal. Models of cell and molecular interactions, derived from the use of these antibodies, form the basis of our efforts to understand and explain the corresponding in vivo systems. However, interpreting in vitro experiments and distinguishing between alternative models is difficult. This complexity is illustrated here using B cell stimulation by surface immunoglobulin and CD40. The fluorescent cell labelling dye carboxyfluorescein, diacetate, succinimidyl ester (CFSE) is used to show that many anti-Ig and CD40 stimulatory agents, used to assess the role of B cells and lymphokines, are partial agonists. By modelling each step in B cell signalling, activation and division it is possible to show that small changes in signal contributed by a second receptor can generate numerous distinct dose response curves that are highly dependent on the "efficacy" of signal transmission by the primary ligand and the number of cell divisions taken in culture. Differences in dose response curves become particularly striking if the primary activating stimulus is a partial agonist. Although exemplified here with B cell stimulation the conclusions are applicable to other in vitro activation systems and suggest ways to improve both the design and interpretation of in vitro experiments.

Murine cytomegalovirus binds reversibly to mouse embryo fibroblasts: implications for quantitation and explanation of centrifugal enhancement.

In a study of the infection of mouse embryo fibroblasts with murine cytomegalovirus (MCMV), we found that plaque number is directly related to virus concentration and not to the total amount of virus contained in the inoculum. These results suggested that virus binding was reversible and that during infection a binding equilibrium is established which limits the amount of bound virus. Further analysis revealed three categories of plaque based on reversibility after virus adsorption. One group was removed simply by washing cell monolayers after virus removal. A second group of plaques was lost gradually with time, giving complete reversal after 5 min at 37 degrees C. The rate of reversal was temperature dependent, and probably represented true virus dissociation. The final group was irreversible plaques, the number of which increased with increasing infection time. The number of reversible plaques remained constant with time of infection, and represented about 70% of the total plaques after 1 h of virus adsorption. Centrifugation of the virus inoculum onto the fibroblast monolayer at 1000 X g increased plaque numbers up to 100-fold, but had little effect on plaque number when carried out after the virus inoculum was removed. In contrast centrifugation increased the number of reversible plaques, suggesting an increase in the number of virus particles attached to the cell monolayers. We suggest that centrifugation enhances MCMV infection by three mechanisms related to reversibility of binding: (1), it increases the rate of virus association; (2), it decreases the rate of dissociation; (3), by increasing the length of time each virus particle is bound it increases the probability of virus being taken into the cell.

Quantal and graded stimulation of B lymphocytes as alternative strategies for regulating adaptive immune responses.

Lymphocytes undergo a typical response pattern following stimulation in vivo: they proliferate, differentiate to effector cells, cease dividing and predominantly die, leaving a small proportion of long-lived memory and effector cells. This pattern results from cell-intrinsic processes following activation and the influence of external regulation. Here we apply quantitative methods to study B-cell responses in vitro. Our results reveal that B cells stimulated through two Toll-like receptors (TLRs) require minimal external direction to undergo the basic pattern typical of immunity. Altering the stimulus strength regulates the outcome in a quantal manner by varying the number of cells that participate in the response. In contrast, the T-cell-dependent CD40 activation signal induces a response where division times and differentiation rates vary in relation to stimulus strength. These studies offer insight into how the adaptive antibody response may have evolved from simple autonomous response patterns to the highly regulable state that is now observed in mammals.