Transcriptome analysis of hematopoiesis was the very first case of success in that direction conducted in China. It reflected to a certain extent the aggregated solid foundation in China in areas of experimental and clinical hematology over the years, albeit the transcriptome strategies were applied only in a limited field associated with normal physiology.

Hematopoietic stem/progenitor cells (HSPCs) possess the potentials of self-renewal, proliferation, and differentiation toward different lineages of blood cells. These cells not only play a primordial role in hematopoietic development but also have important clinical applications. Characterization of the gene expression profile in CD34-positive HSPCs may lead to a better understanding of the regulation of normal and pathological hematopoiesis. Zhu Chen and his colleagues in the Shanghai Institute of Hematology at Ruijin Hospital chose specific CD34-positive cells as a model to initiate the first transcriptome project in China. Since 1996, gene expression profiles of CD34-positive cells derived from human umbilical cord blood, bone marrow and leukemia have been established by the EST sequencing strategy.

Among 9866 ESTs obtained from umbilical cord blood, 4697 (47.6%) showed identity to known genes in the GenBank database, 2603 (26.4%) matched to the ESTs previously deposited in a public domain database, while 1415 (14.3%) were previously undescribed ESTs at that time. Integration of the data of ESTs of known genes generated a catalog including 855 genes that could be divided into different categories according to their functions. Some (8.2%) of the genes in this catalog were considered related to early hematopoiesis [1]. Similarly, 3424 meaningful ESTs obtained from healthy adult bone marrow were integrated into 1630 clusters, representing 622 known genes, 522 dbEST entries and 486 novel sequences [2]. Although the CD34-positive cells of this study were obtained from different tissues, the two gene expression profiles of cord blood and adult bone marrow share about 60% genes and ESTs. However, CD34-positive cells from cord blood have more copies of ESTs corresponding to hemoglobin gamma and HLA class II, less ESTs representing thymosin beta-4 and interleukin-8 than that of adult bone marrow, which might reflect the hematopoietic shift between pre- and post-natal processes. In addition, 1985 EST clusters were produced based on 4321 ESTs of CD34-positive cells derived from patients with acute myeloid leukemia, which included 711 known genes, 657 known ESTs and 617 novel sequences. 58 genes showed statistically significant differences in EST frequency hit by comparing EST datasets between bone marrow and leukemia (P<0.05). The distinct gene expression patterns in leukemia cells as compared to normal control cells may contribute to the development and/or maintenance of the malignant phenotypes of leukemia cells [2].

Importantly, nearly 60% of the cDNA clones of mRNA under 2 kb in HSPCs cDNA libraries had 5′ ends upstream of the first ATG codon of the open reading frame (ORF). It implied that the cDNA libraries were enriched for full-length cDNAs and might be used to develop an efficient working system for full-length cDNA isolation. Subsequently, 300 cDNAs containing putative entire ORFs for previously undefined genes were obtained from CD34+ hematopoietic stem/progenitor cells, based on EST cataloging, clone sequencing, in silico cloning, and rapid amplification of cDNA ends (RACE). The sizes of these cDNAs range from 360 to 3496 bp and the corresponding ORFs may encode peptides of 58–752 amino acid residues. Public database search indicated that 225 cDNAs exhibited sequence similarities to genes identified across a variety of species. Homology analysis led to the recognition of 50 basic structural motifs/domains among these cDNAs. Genomic exon-intron organization could be established in 243 genes by integration of cDNA data with genome sequence information. Chromosomal localizations were obtained using electronic mapping for 192 genes and with radiation hybrid (RH) for 38 genes. cDNA microarray technique was applied to screen the gene expression patterns in five hematopoietic cell lines (NB4, HL60, U937, K562, and Jurkat) and a number of genes with differential expression were found. The resource work has provided a wide range of information useful not only for expression genomics and annotation of genomic DNA sequence, but also for further research on the function of genes involved in hematopoietic development and differentiation [3].

Fetal liver consists of hepatic cells and hematopoietic stem/progenitor cells. Human fetal liver aged 22 weeks of gestation corresponds to the turning point between immigration and emigration of the hematopoietic system. To gain further molecular insight into its developmental and functional characteristics, Fuchu He and his colleagues of Beijing Institute of Radiation Medicine in Beijing studied the gene expression profile of the 22-week fetal liver by generating ESTs and analyzing the compiled expression profiles of liver at distinct developmental stages. Among the available 13 077 ESTs, 5819 (44.5%) matched to known genes, and 5460 (41.8%) exhibited no significant homology to known genes. Integration of ESTs of known human genes generated an expression profile including 1660 genes that could be divided into different gene categories according to their functions. Liver-specific genes and ESTs associated with hematopoiesis were highly expressed, reflecting the unique physiological characteristics of fetal liver. By comparing the expression profiles, six gene groups that were associated with development of liver, tumorigenesis, physiological functions of Itoh cells against the other types of hepatic cells, and fetal hematopoiesis were identified. Meanwhile, cloning 110 full-length cDNAs of novel genes in that work contributed to our understanding of the unique functional characteristics of the human fetal liver [4].

Zebrafish is a useful model for studying on embryogenesis and development, in particular hematopoiesis. Recently, Zhu Chen and his colleagues started to catalog gene expression by EST sequencing from hematopoietic tissue, i.e. kidney marrow of zebrafish. A total of 20 672 ESTs, including 2150 3′-ESTs, were generated and integrated into 13 555 clusters, of which 11 757 ESTs matched with the public zebrafish EST database while 8915 (5617 clusters) are novel ESTs. Interestingly, 463 EST sets have homology with entries in OMIM database at amino acids level (>50 amino acid residues, >45% Identity) [5].

In conclusion, a systematical survey of hematopoiesis gene expression profile in human fetal liver, umbilical cord blood, adult bone marrow, and even zebrafish for the study of normal cell physiology and leukemia pathology represents a meaningful contribution to the progress of transcriptomics. The EST data and full-length cDNA isolated from these studies laid down a solid foundation for further researches, in both functional genomics and hematology. These works, particularly the transcriptomics of CD34-positive cells, is a pioneering milestone of transcriptome study in China. It is also served as a reference in the following studies on HSPCs worldwide [6–8].

Hypothalamus and pituitary, together with adrenals, constitute the major neuroendocrine axis (H-P-A axis) responsible for the maintenance of homeostasis and the response to the perturbations in the environment. Potential new hormones or novel functions of the axis as well as the molecular mechanisms of the regulatory networks need to be uncovered by new strategies. In 2000, Zeguang Han and Jialun Chen and their colleagues at the Chinese Human Genome Center at Shanghai (CHGCS) and Shanghai Institute of Endocrinology established a primary gene expression profile for the human H-P-A axis by generating a large number of ESTs, followed by bioinformatics analysis. Totally, 20 626 EST sequences of high quality were obtained from cDNA libraries of the hypothalamus, pituitary, and adrenal glands. These ESTs could be assembled into 9175 clusters (3979, 3074, and 4116 clusters in hypothalamus, pituitary, and adrenal glands, respectively) when overlapping ESTs were integrated. Of these clusters, 2777 (30.3%) correspond to known genes, 4165 (44.8%) to registered ESTs, and 2233 (24.3%) to novel ESTs. The gene expression profiles reflect well the functional characteristics of the three levels in the hypothalamus-pituitary-adrenal axis, because most of the 20 genes with highest expression showed statistical difference in terms of tissue distribution, including a group of tissue-specific functional markers. The order of gene expression levels of the 5 classical hormones in pituitary coincides with their protein levels in the serum, while all enzymes associated with biosynthetic pathway of corticosteroids are encountered in EST data of adrenal gland. Some findings were made with regard to the physiology of the axis. Except for producing the known hormones, hypothalamus and pituitary could secrete many other cytokines that may play important role in interface between neuroendocrine and immune systems. In addition to the gene expression profile, this work contributed 200 full-length cDNAs of novel genes for which comparative genomics analysis revealed interesting features [9].

To this date, these data, which constituted the majority of the ESTs from human H-P-A axis in the public database, is still a unique resource for hormonal genomics approach. The data may facilitate the isolation of novel genes involved in hormonal signaling as well as their characterization, such as chromosomal location, polymorphism, splicing variants, differential expression, and physiological functions. These data may also enhance our insights into the complexity of intercellular communications, and understand the physiology and pathology of the human neuroendocrine system [10].

Further studies with regard to the gene expression profiles in the axis are underway. That includes: EST sequencing of 17 000 cDNA clusters after fingerprinting 54 000 clones from the pituitary library; isolation of more previously undefined full-length cDNAs; identifying regulatory/response proteins by functional genomics and systems biology strategies; medical application of the knowledge and platform of transcriptome analysis, such as analyzing the cell and tissue effect of Traditional Chinese Medicine (TCM) at the level of transcriptome.

Zhenyi Wang and his colleagues in the Shanghai Institute of Hematology were the first to successfully treat patients with acute promyelocytic leukemia (APL) using all-trans-retinoic acid (ATRA) [11]. The revolutionary strategy of cancer differentiation therapy inspired numerous scientists to address the molecular mechanisms of ATRA treatment. In an attempt to identify the events downstream of the targeting of aberrant RA receptor by ligand, Zhu Chen and his co-workers compared the gene expression patterns in the APL cell line NB4 before and after ATRA treatment using cDNA array, suppression-subtractive hybridization, and differential display-polymerase chain reaction (DD-PCR). Among 169 genes modulated by ATRA estimated to represent 1–2% of all genes expressed in APL cells, 100 genes, including 8 novel ones, were up-regulated while 69 ones were down-regulated by ATRA. The ATRA-induced gene expression profiles were in high accord with the differentiation and proliferation status of the NB4 cells. The time courses of their modulation were interesting. The 100 up-regulated genes occurred most frequently 12–48 hours after ATRA treatment, while 59 of 69 down-regulated genes had their expression suppressed within 8 h. The transcriptional regulation of 8 induced and 24 repressed genes was not blocked by cycloheximide, which suggests that these genes may be direct targets of the ATRA signaling pathway. A balanced functional network seemed to emerge, forming the foundation of decreased cellular proliferation, maintenance of cell viability, increased protein modulation, and promotion of granulocytic maturation. Several cytosolic signaling pathways, including JAKs/STAT and MAPK, may also be implicated in the symphony of differentiation [12].

This large-scale transcriptome analysis for APL cells upon treatment with ATRA provided the first global illustration on the molecular characteristics of ATRA-induced differentiation. A model of gene regulatory network modulation during this process could be obtained and precious datasets were developed for further studies in order to ultimately clarify the mechanism of both APL leukemogenesis and the differentiation therapy. This kind of knowledge will certainly accelerate the identification of drug targets for novel drug discovery.

Human hepatocellular carcinoma (HCC) is one of the most common cancers worldwide, particularly in China. Hepatitis viruses (HBV and HCV) infection is the major factor contributing to hepatocarcinogenesis in China. Zeguang Han and his colleagues conducted a comprehensive characterization of gene expression profiles for hepatitis B virus-positive HCC. A large set of 5′-read EST clusters (11 065 in total) from HCC and non-cancerous liver samples were established, and then a homemade membrane based cDNA microarray system containing 12 393 genes/ESTs was applied for further analysis. The Atlas^TM human cancer 1.2 array cDNA microarray from Clontech Inc., which contains 1176 known genes related to oncogenesis, was also used for profiling gene expression. Integrated data from the above approaches identified 2253 genes/ESTs as candidates with differential expression. A number of genes related to oncogenesis and hepatic function/differentiation were selected for further semi-quantitative reverse transcriptase-PCR analysis in 29 paired HCC/non-cancerous liver samples. Many genes involved in cell cycle regulation such as cyclins, cyclin-dependent kinases, and cell cycle negative regulators were deregulated in most patients with HCC. Aberrant expression of the Wnt-beta-catenin pathway and enzymes for DNA replication also could contribute to the pathogenesis of HCC. The alteration of transcription levels was noted for a large number of genes implicated in metabolism, whereas a profile change of others might represent a status of dedifferentiation of the malignant hepatocytes, both considered as potential markers of diagnostic value. Notably, the altered transcriptome profiles in HCC could be correlated to a number of chromosome regions with amplification or loss of heterozygosity, providing one of the underlying causes of the transcription anomaly of HCC [13].

Gengxi Hu and his colleagues of the Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, also studied the expression profiles in HCC using a cDNA array representing 14 000 cDNA clusters. They found that 72 genes (including 30 novel genes) were down-regulated and 84 genes (including 48 novel genes) were up-regulated in a majority of the cancer samples. It was conspicuous that 21 of 38 HCC down-regulated genes studied previously were regulated by a group of liver-enriched transcription factors (LETFs). Reexamination of the cDNA array data further revealed that most of the genes known to be regulated by LETFs were down regulated in at least a portion of the HCC samples. Among the LETFs, the expression level of CCAAT/enhancer-binding protein (C/EBP) alpha was down-regulated in cancer, whereas hepatocyte nuclear factor 1 (HNF-1), HNF-3beta, HNF-4alpha, and HNF-4gamma were up-regulated. The expression profiling thus suggested multiple regulatory pathways involved in HCC, especially that related to LETFs [14].

Although a comprehensive and systematic understanding of hepatocarcinogenesis and metastasis is yet to be achieved, the transcriptomics approach for HCC conducted by Chinese scientists is one of the earliest trials in that direction. This work provided not only significant amount of relevant publicly available ESTs but also intriguing unique concepts to explore in hepatocarcinogenesis. These works clearly facilitated the process of isolation and identification of novel genes associated with HCC, and thus have laid down the foundation for the discovery of HCC related biomarkers and drug targets [15–18].

Chronic neuropathic pain, as common symptom associated with many diseases, disturbs continually the normal life of patients. Phenotypic modification of dorsal root ganglion (DRG) neurons represents an important mechanism underlying neuropathic pain. However, the nerve injury-induced molecular changes are not fully identified. To determine the molecular alterations in a broader way, Dr. Zhang's and Dr. Han's groups in Shanghai carried out EST sequencing and cDNA array analysis on the genes mainly represented in the cDNA libraries of lumbar DRGs of normal rats and of rats 14 days after peripheral axotomy. Of the 7523 examined genes and ESTs, the expression of 122 genes and 51 ESTs is strongly changed. These genes encompass a large number of members of distinct families, including neuropeptides, receptors, ion channels, signal transduction molecules, synaptic vesicle proteins, and others. Of particular interest is the up-regulation of gamma-aminobutyric acid (A) receptor alpha5 subunit, peripheral benzodiazepine receptor, nicotinic acetylcholine receptor alpha7 subunit, P2Y1 purinoceptor, Na(+) channel beta2 subunit, and L-type Ca(2+) channel alpha2delta-1 subunit. Those findings therefore reveal dynamic and complex changes in molecular diversity among DRG neurons after axotomy [19].

Human physiological and pathological characteristics, particularly for diseases with high incidence in China, will be approached initially by profiling gene expression patterns. Based on our previous experiences, systematic sample collection, sophisticated experiment design and innovative technology application are the main factors affecting the outcome of these costly studies. In order to understand the mechanism of human physiology and disease, appropriate model systems are essential. The transcriptome studies should be coordinated with the development of model animals, in particular, mouse and primates. Another major goal for the study of this field is isolation of more full-length cDNAs from different tissues and cells such as hypothalamus, pituitary, adrenal gland, liver, esophagus cancer, hematopoietic tissues and immune cells. Meanwhile, systematic survey of RNA isoforms will gradually become a more important goal in human transcriptome studies.

Model organisms are especially useful systems for comparative genomics studies. Besides bacteria, China has been emphasizing studies of the rice genome, a model organism for monocotyl plants. Transcriptome research for rice and other plants is developing quickly in China. It is an important strategy for isolation of genes combining with the genomic information. It will also facilitate the identification of genes associated with diseases and specific characters of plants. Microbes associated with human and plant diseases will also be studied in priority, emphasizing their interaction with their host and environment.

For animals, the metazoan parasite Schistosoma japonicum, as the major blood fluke in China, remains a public health problem even today. Gaining knowledge about the S. japonicum genome will be important for understanding the biology of the worms with respect to their evolution and pathogenesis. Schistosomes are dioecious digenetic trematode under the phylum platyhelminthes. Unlike Caenorhabditis elegans, schistosomes have sexual dimorphism and larger males gripping smaller females in a gynecophoric canal. Recently, The S. japonicum genome project was initiated and large-scale EST data, representing about 43 000 ESTs and 13 000 clusters, was generated from male, female, mixed sexes adult worms, and their eggs by EST sequencing a first step.

Recently Chinese scientists were actively involved in the international consortium for the genomics research of Lepidoptera, moths and butterflies. The centerpiece of this strategy is the production of a draft genome sequence of Bombyx mori, the domesticated silk moth, by the end of 2004. The silk moth is regarded as the central model for lepidopteran genomics and genetics, and is an important source of livelihood for subsistence farmers engaged in silk production in at least 55 countries. Along with the efforts in genomic sequencing, gene expression profiles are being studied by Chinese, Japanese and French scientists with respect to its development, particularly the development of the posterior silk gland employing techniques of EST sequencing and SAGE. Proteomics work has also started in Japan and China. It is expected that this study will be greatly improved with the help of the whole genome draft sequence.

Although the transcriptome study was initiated by large-scale EST sequencing analysis, it will be hard to achieve the essential high-throughput in order to generate meaningful data sets without the technology of cDNA microarray and biochips. However, the biochip technology is obviously still under development and it will be crucial to make this technology accurate and reliable enough for high quality research. It is also crucial to make it economically affordable for most laboratories working on functional genomics. In that direction, SAGE technology, along with the improvement of its protocol as well as the easier availability of large-scale sequencing platforms, is becoming more and more popular. It will become a nice compromise to complement the weaknesses of the biochip technology.

Other technologies under development are emphasizing the detection of protein-protein, protein-nucleic acid interactions. For other purposes, the interaction between biomolecules with small molecules (ligands) as well as large-scale detection of molecular interaction in tissues and cells will soon be detected by certain kinds of ‘biochip’ technology, e.g., tissue chips or cell chips. In that direction, the long-waited protein chips and antibody chips are the most important.

Bioinformatics has always been considered as the key component as well as the tool for any ‘ome’ research, as for the transcriptome studies. Innovations in bioinformatics algorithms should be emphasized while consistent efforts in sophisticated interrogative database development must be encouraged.

An important direction is to combine biology with mathematics, physics, chemistry, informatics, computing technology and engineering. Many biomedical and systems biology centers and groups have been established in large cities of China, particularly in Shanghai and Beijing. These centers have the advantage of aggregating resources of many universities, institutions and even companies in order to establish inter-disciplinary research platforms and environment. Scientists working in these centers will be fostered to integrate information and technology ranging from genomics, transcriptomics, proteomics, to metabolomics, to get novel insight into the nature of complex biological systems and diseases.

Besides joining the international HGP, Chinese scientists will contribute continually their data to the public databases accessible by their colleagues worldwide. Human full-length cDNA will be deposited in GenBank and exchanged by collaboration with the international mammalian gene collection (MGC) project. The data of transcriptome analysis will follow the MIAME guidelines of the MGED Society and will be shared with other scientists via the SYSTEMOSCOPE International Consortium.