From clones of cells to cloned genes and the proteinpaedia.

This brief description follows the association of the author with Ivan Lefkovits from 1971 until this volume. Sketches of our mutual interests are included. Time periods in California, Basel and Texas are described. Decisions about preparing new tools for clonal analysis are elucidated, and experimental approaches leading to the x-omic revolution are described.

Proteomic analysis of rare molecular species of translated polypeptides from a mouse fetal thymus cDNA library.

Proteomic patterns from an ordered cDNA library of mouse fetal thymus origin of an overall complexity of 1536 clones are described. Patterns have been analyzed at 96, 12, or 8 clones in a mixture, or as individual clones. Clones yield in some instances a single spot, in other cases a complex cluster or family of spots is formed. The determination of the clonal address (a six-digit number, indicating the interception of three pooling dimensions) by inspection of pools of 8, 12 or 16 clones is a reliable approach; nevertheless a complete proof consists in retrieving the clone and then submitting to transcription, translation and proteomic analysis. The spot clusters are meaningful clone identifiers; cluster components (families of polypeptides) are characteristic of individual clones and are independent of clones coexisting (and being co-expressed) in a given pool. A 'cluster' originates from a single cloned message and might be due to post-translational modification (offered by the reticuleocyte machinery) or as a result of programmed degradation. Thirteen clones or families of clonal products are shown, and the heterogeneity of the 'appearance' of clones is documented. In about half, the assignment of a clonal polypeptide product to a storage well position has failed; this might be due to a variety of considerations elaborated herein. The arguments are presented, that analysis of abundant proteins of a cell (activated lymphocyte) comprises 'classical proteomics', but for the analysis of the rare molecular species of proteins, Poissonian approaches of replicable material have to be used.

Proteome, transcriptome and genome: top down or bottom up analysis?

Biological systems are comprised of protein components found at a wide variety of abundances from millions of molecules of a single species per cell to less than one copy per cell. Because of this wide range of concentrations, measurement or a full accounting of each system is presently unavailable. Conventional separation and analytical methods (two-dimensional gel electrophoresis and mass spectrometry) allow identification and quantitation of many of the most abundant gene products (top down methods); and the majority of gene products, which are found at low abundance, can be neither identified nor measured in complex mixtures at present. The gene products that are found at low levels can be characterized and their properties analyzed by preparing ordered gene libraries of limited complexity from mRNA. When such preparations are expressed in cell free systems and analyzed by two-dimensional gel electrophoresis, the features of the gene products are available for analysis. This 'bottom up' approach allows identification of gene product properties so that analytical procedures can be devised and applied to complex mixtures.

Global analysis of gene expression in cells of the immune system I. Analytical limitations in obtaining sequence information on polypeptides in two-dimensional gel spots.

We have outlined various aspects and limitations of the collective analysis of protein species of a cell (lymphocyte). We have indicated research directions that, in to our opinion, deserve more attention. We have evaluated mainly the approach used in our laboratory and we recognize that a bulk of important research on the interface of proteomics and genomics remains to be dealt with. It is of great value that we can proceed in our quest by trial and error. But as much as the human genome initiative was not implemented by trial and error, but by formulating new technological approaches, we hope that our approach can be incorporated in the mainstream of proteomics. We need several integrating research directions, some of which are outlined in this communication, namely the use of ordered cDNA libraries, cell-free expression systems, high density filter hybridization, identification of two-dimensional (2-D) gel spots in terms of their amino acid composition through biosynthetic labeling and identification of restriction sites in the corresponding coding sequences. In the accompanying paper the cDNA ordered library approach will be described in some detail.

Global analysis of gene expression in cells of the immune system II. Cell-free translation products and high-density filter hybridization data.

We have developed an experimental system for linking information on cell-free transcription and translation products from cDNA clones with data obtained from hybridization signals from complex probes. The work described in this paper consists of two distinct processes, one being the construction of a system of clonal addresses and the other the identification of expressed genes involved in the studied processes. We describe the use of this system to identify genes involved in thymus development. Complex probes from fetal thymuses (GD15, 17 and newborn) of Balb/c mice were used to identify genes, which are up- or downregulated during the process of differentiation. The full set of information is available in the Clone-base of the Basel Institute for Immunology and will be retrievable from the website of the collaborating laboratories.

Classification and properties of 64 multiplexed microsphere sets.

We describe a practical method for the analysis of multiple analytes in a single sample. The vehicle for each separate measurement consists of a set of microspheres identifiable by characteristic fluorophores embedded in the particles. The use of robust, bench-top flow cytometers (flow microfluorimeters) for the analysis of the multiple sets of microspheres is facilitated by hardware and software, which acquire the data from the cytometer, classify the microspheres according to sets, and collate measurement information from each microsphere set in real time. This measurement system can analyze up to 64 analytes in a single sample. The advantages of multiplexed assays using flow cytometry include robust measurements, because each microsphere set is measured repeatedly. The advantage of the assay's is consistent with simultaneous measurement of many parameters as well as the speed with which the flow microfluorimeter (cytometer) makes measurements (many hundreds per second). Here, we describe the properties of the microspheres, the calibration of the cytometer, and the influence of the properties of the microspheres on the sensitivity of measurements.

Advanced multiplexed analysis with the FlowMetrix system.

The FlowMetrix System is a multiplexed data acquisition and analysis platform for flow cytometric analysis of microsphere-based assays that performs simultaneous measurement of up to 64 different analytes. The system consists of 64 distinct sets of fluorescent microspheres and a standard benchtop flow cytometer interfaced with a personal computer containing a digital signal processing board and Windows95-based software. Individual sets of microspheres can be modified with reactive components such as antigens, antibodies, or oligonucleotides, and then mixed to form a multiplexed assay set. The digital signal-processing hardware and Windows95-based software provide complete control of the flow cytometer and perform real-time data processing, allowing multiple independent reactions to be analyzed simultaneously. The system has been used to perform qualitative and quantitative immunoassays for multiple serum proteins in both capture and competitive inhibition assay formats. The system has also been used to perform DNA sequence analysis by multiplexed competitive hybridization with 16 different sequence-specific oligonucleotide probes.

Analysis of a randomly assorted cDNA library from BW 5147 lymphoid cells. Frequency estimates based on computer aided concatenation matching of 2D gel polypeptide products.

A cDNA library was divided into 291 sectors of low complexity, 800-1000 recombinant phage plaques per sector. Aliquots of DNA from sectors prepared from high titer phage were subjected to in vitro transcription using T7 polymerase and thereafter RNA was translated in a cell-free rabbit reticulocyte system in the presence of [35S] methionine. The resulting polypeptides were separated by 2D gel electrophoresis. Radiofluorographs of the gels prepared from 10 sectors were subjected to scanning and image analysis. A multilevel spot matching procedure was employed allowing us to recognise spot occurrence in the 10 sectors. The number of detected polypeptide spots per sector varied between 147 and 325. Overall, 1082 different spots were counted totalling 1983 spots considering multiple entries. The distribution of the spots and the frequency distribution of the RNA population among the 10 sectors are shown. The highest detected frequency is 10(-2), while one half of the RNA molecular species are present at a frequency of 10(-3) or lower.

Interaction of 2C T cells with a hybrid Ld molecule bearing an alpha 3 domain derived from the class IB molecule, Qa-2.

The CD8 co-receptor interacts with nonpolymorphic residues on class I molecules. LQ3, a laboratory engineered Ld molecule bearing an alpha 3 domain derived from Q7 (Qa-2), interacts poorly with anti-Ld CD8-dependent T cells. 2C TCR transgenic mice bear a receptor specific for the p2Ca peptide bound to Ld. The authors show that although this peptide interacts with LQ3, LQ3 APC fail to activate splenic 2C CD8 T cells in vitro in the absence of IL-2, while control Ld APC do. The authors have used this receptor ligand pair to examine negative selection within the thymus of (B6 x C3H.Ld)F1 versus (B6 x C3H.LQ3)F1 radiation chimeras repopulated with 2C bone marrow cells. While positive selection occurs normally in (B6 x C3H)F1 chimeras, animals expressing either Ld or LQ3 fail to generate 2C CD8+ cells. Thus, either CD8 is not required for negative selection of this TCR or a weak interaction of CD8 with LQ3 is sufficient. TSA-1, a developmentally regulated marker, was used to follow the process of negative selection. The results show that deletion of 2C T cells does not occur until thymocytes reach the double positive (DP) stage. Furthermore, the authors noted a small population of DP TSA-1hi cells remains, while DP TSA-1int and TSA-1lo cells are absent. These data support the notion that thymocytes either reach a particular stage of development or locate in an appropriate intrathymic compartment before they undergo negative selection.

Human lymphocyte cDNA ordered library analyzed by 2D gel electrophoresis. 1. Pooling strategy and matching of gel patterns.

We have analyzed an ordered library of 4,608 cDNA clones from the CEM human leukemic cell line. The aim was to facilitate gene retrieval, to enable immediate access to cDNA clones and to provide information on the protein expression of the individual clones in a 2D gel readout. The matrix array of 24 x 16 x 12, each position of which contained lambda jacII phage from one plaque, enabled us to establish pools of clones along the three axes (24 pools of complexity 192 cloned entities, 16 pools of complexity 288 and 12 pools of complexity 384). The total cDNA complexity is here reduced to such a level that spots which in more complex gels served as landmark spots are not present in each pool, and thus cannot serve as landmarks anymore. The image analysis of such gels and especially the matching of spots is not reliable under these circumstances. In order to achieve reliable matching, additional samples were created, such that pools were co-electrophoresed according to a special concatenation scheme; these samples then contained over-lapping elements (e.g., pools 1 + 2 + 3 and 3 + 4 + 5 have at least those spots in common, which originate from pool 3). This approach turned out to be feasible and we have completed the matching of one half of the ordered library. Already from the present stage of analysis we have obtained valuable information on the cDNA library and on the distribution of clones in this library.

The amino acid composition of 350 lymphocyte proteins.

We determined the amino acid composition of proteins of Sp2 hybridoma cells by a procedure which assembles the information on the polypeptides upon two-dimensional gel electrophoresis, such that biosynthetic labeling with 20 different 3H amino acids provides the data--spot intensities--on the relative representation of the detected polypeptides. The gels were impregnated with 2,5-diphenyloxazol (PPO) and suitably exposed radiofluorographs were selected for analysis. The images originating from the 12 cultures labeled with amino acids R, A, H, I, L, K, M, F, P, S, T and Y were analysed with the Kepler image analysis system. The spot volume data of the 12 analysed patterns were corrected for the unequal labeling efficiencies of the 3H amino acids and for the various exposure times. This correction is performed by applying calibration factors based on the amino acid determination of a hydrolysate of the analysed cells. After the calibration step was applied to the data files we used the amino acid compositions of nine proteins taken from a database to establish for each of these proteins the correlation coefficients with the image analysis derived amino acid compositions of all spots. The correlation coefficients allow us to tentatively identify polypeptide spots on two-dimensional gels, while the amino acid composition of 350 investigated two-dimensional gel spots is usable as an identification tag in the gene retrieval from our cDNA libraries.

Transplantable polyomavirus-induced salivary gland epitheliomas are populated by immature T cells derived from the host.

Transplantable, polyomavirus-induced salivary gland epitheliomas were passaged through F1-hybrid mice of C3H/Bittner with BALB/cJ or AKR/J strains and congenic nu/nu athymic mice on the BALB/c background. By analysis of the Thy-1 and Ly-2 (CD8a) alloantigens of the T cells that infiltrated the transplanted tumors, it has been determined that the lymphocytes were derived from the host and not the donor of the epithelial tumor. It is known that the tumor masses in the F1-hybrid and athymic mice were not the result of a secondary tumor induction or transformation of host tissue because the epithelioma could be transplanted back into the strain of origin, C3H/Bittner. Intratumor lymphocytes resembled thymocytes on the basis of their intense staining by anti-Thy-1 reagents, the presence of large fractions of cells that bore both CD4 and CD8a, and the cell-size distribution of the phenotypes. Nu/nu mice carrying epitheliomas infiltrated with host-origin T lymphocytes were not immunologically reconstituted, as judged by the lack of T cells in the spleens of such tumor hosts. The epithelial tumors acted as sites of host T-lymphocyte maturation, but the process was incomplete, and mature T cells did not emigrate to the periphery.

Lymphocyte Proteinpaedia stage two: T-cell polypeptides from a partitioned cDNA library revealed by the dual decay method.

A complex population of gene products was analyzed by combining the great resolving power of two-dimensional (2D) protein electrophoresis with a detailed dissection of individual protein species afforded by cDNA cloning. A cDNA library from BW5147 was partitioned by random sampling into sectors (of a complexity of 500 phase plaques/sector). From each of the unique sets of cDNA clones present in the sector, 2D gels were prepared. Patterns of spots were analyzed using the Kepler software system, and the compiled data both from the sectors and from the natural lymphocyte populations have been established to serve as a guide for gene retrieval. Use of the dual decay method, which compares two exposures of a 2D gel co-electrophoretic pattern, obtained from samples with (35S)-polypeptides mixed with (3H)-polypeptides, is described and scrutinized. It was determined that of 268 spots (three sectors combined), 118 spots were shared with the natural lymphocyte population, the remaining 150 spots occur only in the sectors (and not detected in the cell population). All cDNA populations are available for retrieval. Information obtained throughout the work was entered into the database.

Autoantibodies to the GLUT-2 glucose transporter of beta cells in insulin-dependent diabetes mellitus of recent onset.

Purified immunoglobulin G (IgG) from the serum of patients with insulin-dependent diabetes mellitus (IDDM) of recent onset inhibits high-Km uptake of 3-O-methyl-beta-D-glucose by rat pancreatic islets. To determine if the inhibition is the result of antibodies against GLUT-2, the high-Km glucose transporter of beta cells, we incubated IDDM sera with rat islet cells and with AtT-20ins cells transfected to express GLUT-2. IDDM sera inhibited glucose uptake in islet cells and in GLUT-2-expressing AtT-20ins cells but not in AtT-20ins cells transfected to express the low-Km isoform, GLUT-1. In 24 of 30 (77%) patients with newly diagnosed IDDM, IgG binding as measured by immunofluorescence and flow cytometry of the cells transfected to express GLUT-2 was > 2 standard deviations from the mean of the nondiabetic population; 29 of 31 (96%) of nondiabetic children were negative (P < 0.0001). Increased IgG binding could be removed by absorption with GLUT-2-expressing cells but not with GLUT-1-expressing cells. We conclude that most patients with IDDM of recent onset have autoantibodies to GLUT-2.

Restriction sites as identification tags for the gene catalog: a 2D gel model.

In our effort to collect, organize and assemble data from lymphocyte cDNA libraries, we assign DNA restriction sites collectively to the spots on two-dimensional (2D) gel patterns. In order to test the efficiency and reliability of such an approach, we have modeled the restriction analysis of cDNA libraries with a panel of restriction endonucleases. The work has two parts. In the first, we have chosen 255 proteins from the EMBL data base and determined whether or not their coding sequences contain restriction sites for the enzymes of our choice. In order to apply a sufficient discriminatory power we decided to use a relatively large number of cleaving enzymes with low and high cutting frequencies. In total, 13 restriction enzymes were chosen, which could distinguish 2(13) or 8192 different restriction site combinations. We have compiled a table in which the absence or presence of restriction sites yields a pattern of 'zeros' and 'ones'. Such a restriction pattern can be read as a binary number. The binary numbers with maximally 13 digits would uniquely assign each of the 255 proteins if the nucleotide sequences would be truly at random. As the restriction sites are not randomly distributed, the 'typing' does not yield a unique assignment. The choice of sequences was not random either. In fact, there are some human nucleotide sequences which possess the same cut number (the decimal equivalent of the binary number representing the restriction pattern). In spite of this redundancy, 141 coding sequences could uniquely be distinguished by the above treatment. In the second part of the project we have used the above mentioned coding sequences to prepare two-dimensional maps (plots of charge vs size) of the same kind as one obtains from experimental 2D gels and submitted such a map together with 13 maps of restriction enzyme treated populations to a computer image analysis. Ideally, one would expect results (cut numbers) congruent to those obtained in the first part of the work. In the modeled system we were confronted with 2D maps which closely resembled the experimental situation (e.g. some spots were close together and overlapping) and instances of incorrect spot detection yielding 'false cut numbers'. From 255 proteins we were able to assign unequivocally 161 proteins. To implement the model in an actual experiment we will perform the digestion with the restriction enzymes in duplicate, and only spots assigned the same cut number upon the two independent treatments will be considered as carrying a valid restriction tag.

Diverse classes of T-cell lymphomas arise in T-lymphocyte-inductive epitheliomas of the mouse salivary gland.

During the course of serial transplantations of polyomavirus-induced C3H-Bittner salivary gland epitheliomas in F1-hybrid mice, three tumor sublines were found which gave rise to T-cell lymphomas of host origin. The lymphomas resembled spontaneous AKR/J thymic lymphomas in their expression of lymphoid differentiation antigens, and they may represent sequential stages in the differentiation of immature T lymphocytes. We found no evidence that polyomavirus directly induced the lymphomas, rather, the lymphomagenic events paralleled those which occur in spontaneous AKR/J thymic lymphomas.

Global analysis of lymphocyte gene expression: perturbation of H-9 cells by infection with distinct isolates of human immunodeficiency virus--an exposition by multivariate analysis of a host-parasite interface.

AIDS is a progressive disease associated with steady loss of helper T cells and several other functions. As the disease evolves, cytopathogenic human immunodeficiency (HIV) variants of increasing virulence can be isolated from the host. The HIV is an unusually variable genome by virtue of a low replication fidelity. In this report we describe our effort to test the hypothesis that there is a correlation between virus variability and cytopathogenicity, and further, that there is an "impact" of the virus infection on the expression of host cellular genes. To search for such a relationship, we infected H-9 cells (human CD4+ lymphoblastoid cell line) with each of 5 isolates of HIV of distinct origin and cytopathogenicity. To measure the influence of the virus infection on the expression of host cellular genes, shortly after infection, (3 h or 13 h), cells were radiolabeled and the radioactive polypeptides separated by two-dimensional gel electrophoresis. Radiofluorographs were prepared and analyzed to determine relative rates of biosynthesis of cellular polypeptides. To organize the large amounts of data found, cluster analysis and principal component analysis were used to expose the data in formats that allowed a model construction. The rates of biosynthesis of many cellular polypeptides were altered upon viral infection in terms of both enhancements and impairment of biosynthesis. Some of the variation in polypeptide synthesis was isolate-specific, while most alterations were of modest magnitude. There appears to be no "overall effect" associated with infection by a cytopathic variant of the virus. Polypeptides affected by the cytopathic variants were determined as targets for further investigation. The method used promotes the measurement of "ensemble" information that is characteristic of the process and it promotes the creation of models of virus action.

Histology of transplantable murine epithelial tumors infiltrated with immature T lymphocytes: neoplastic analogues of the thymus.

Serially transplantable murine thymic and salivary gland epithelial tumors were studied by immunohistologic methods and by electron microscopy. At the ultrastructural level, tumors appear to be identical. The large polygonal epithelial cells lack rough endoplasmic reticulum, Golgi apparatus, and secretory vesicles, and are anchored together by tonofilaments and desmosome junctions. Some areas in the tumors are composed solely of the neoplastic epithelium, while other areas are richly "infiltrated" with lymphocytes having a cortical thymocyte phenotype and supported by a lacy dendritic network of neoplastic epithelial cells. Immunohistochemical staining demonstrates that the epithelial cells express class I and class II major histocompatibility gene complex (MHC) antigens, epidermal keratin, and the ER-TR4 and ER-TR5 thymic epithelial markers. Lymphocyte-rich regions show homogeneous staining for both L3T4 (CD4) and Lyt-2(CD8) T-cell antigens.

Multiple steps in the immortalization of cells of the B lineage.

The ability of A-MuLV to transform bone marrow cells on in vitro culture in agarose is enhanced by inclusion of conditioned media during infection and culture. The conditioned medium of a non-virus producing A-MuLV transformed fibroblast cell line was synergistic with medium from Whitlock-Witte long-term bone marrow cultures, while conditioned medium from modified Dexter-type cultures was not active. These media all contained growth promoting activity for bone marrow cells. There are two types of transformed colonies produced, and transformation of only one type was enhanced by inclusion of conditioned media. Analysis of this type of transformed cell showed them to be pre-B cells. Limiting dilution analysis suggests the transformation process to be dependent on two types of cells, one presumably the target and the second an "accessory cell". Models are presented to account for the factor-dependent in vitro transformation of pre-B cells.