Share this page: More

Heng Zhu

Heng Zhu

Department Affiliation: Primary: Pharmacology and Molecular Sciences; High Throughput Biology Center (HIT Center)
Degree: Ph.D., Clemson University
Rank: Professor
Telephone Number: 410-502-0878
Fax Number: 410-502-1872
E-mail address:

School of Medicine Address:  Room 327, Edward D. Miller Research Building, 733 N. Broadway, Baltimore, MD 21205

Systems biology, construction of networks and pathways, transcription regulation, and pathogen-host interactions.

1. Established Functional Protein Microarray Technology

A functional protein microarray is usually constructed by immobilizing individually purified proteins on glass slides at high density. To this end, my group has fabricated functional protein microarrays in various model organisms, including yeast (Lin et al, Cell 2009); herpesviruses (Zhu et al, J Virol 2009); E. coli (Chen et al, Nat Methods 2007); rice blast fungus (Li et al, PLoS One 2011) and humans (Hu et al, Cell 2009; Jeong et al, Mol Cell Proteomics 2012). Among these protein microarrays, the human proteome microarray is the world’s largest and is comprised of ~17,000 individually purified human proteins in full-length. My group has also developed seven types of binding assays and seven types of covalent reactions for protein microarrays. The applications of some of these assays are described below.

To further improve protein microarray technology, we reported two new strategies that greatly simplify the process of peptide/protein chips construction by combining protein translation, purification, and immobilization into a single step. We efficiently fabricated both peptide and protein microarrays at high-density using an in vitro translation system (Tao & Zhu, Nat Biotech 2006). In another example, we coupled mass spectrometry analysis as a detection method for binding assays performed on a protein microarray. We showed that various epitope peptides could be specifically detected by their corresponding antibodies spotted on a porous gold surface (Evans-Nguyen et al, Anal Chem 2008). In a third example, we constructed a lectin microarray and applied it to determine glycan signatures of live cells (Tao et al, Glycobiology 2008).

2. Virion Display (VirD) Technology

Multipath transmembrane proteins encompass a large portion of the human proteome and serve a wide variety of critical roles. For example, G protein-coupled receptors (GPCRs) forms a largest tranmembrane protein family, consisted of seven transmembrane domains. This complex structure allows them to bind to a variety of ligands, ranging from protons, ATP, amino acids, peptides, proteins, and to many other unidentified ligands. To date, ~40% of the FDA-approved drugs target GPCRs. Since the lipid bilayer is required to maintain the conformation of GPCRs, purification attempts often disrupt the GPCR conformation. To overcome this hurdle, we developed VirD technology by replacing a viral envelope gene in herpes simplex virus-1 (HSV-1) with an ORF encoding a human transmembrane protein (Hu et al., Anal Chem 2013). The production of this recombinant virus from mammalian cells allowed the human receptor to be embedded in the viral envelope with correct conformation and function. More importantly, these recombinant viruses were arrayed on a glass slide to facilitate high-throughput screenings. Recently, we expended the VirD technology to cover most of the non-odorant GPCRs (e.g., 315) for further biochemical interrogation (Syu et al., Nat Commun 2019). We demonstrated that the GPCR-VirD array is useful to profile specificity of mAbs. Of the 20 commercial mAbs tested, 10 mAbs were determined to be ultra-specific. The rest failed to recognize their intended targets or failed to show specificity entirely. Importantly, we demonstrated that nine of the 10 GPCRs, which failed in the mAb binding assays, could still recognize their canonical ligands in VirD format, suggesting that they are folded correctly. By probing the VirD-GPCR arrays with known ligands, we discovered several off-targets for a peptide hormone, somatostatin-14. Two selected off-targets along with the canonical GPCR were validated with virion nano-oscillators, which is a real-time, label-free detection method (Ma et al., J Am Chem Soc 2018), and showed significant binding affinity. Lastly, we probed the GPCR-VirD array with a clinical strain involved in neonatal meningitis (Group B Streptococcus K79) and identified five potential GPCR targets. CysLTR1 was further validated in vitro and in vivo as a host receptor for K79 invasion. We believe that the VirD array is a robust platform to profile many different kinds of membrane protein interactions.

3. Global Analysis of the Human Protein-DNA Interactome

Protein-DNA interaction plays a crucial role in many important biological processes, such as transcription, DNA damage repair, DNA replication, chromatin assembly and dynamics, recombination, and defense, to name a few. Understanding the molecular principles underlying protein-DNA interactions will greatly facilitate the research in these fields and beyond. During the past eight years, we developed and applied a human transcription factor (TF) microarray, in conjunction with bioinformatics, to systematically analyze protein-DNA interactions in humans.  This human TF array is comprised of ~1,400 TFs (>85% coverage) and 2,800 additional proteins.  We then screened them with 460 DNA motifs, which resulted in identifying 200 new consensus motifs.  However, an unexpected discovery was that hundreds of proteins not previously annotated as TFs also showed sequence-specific protein-DNA interactions.  These proteins represent a broad range of functional groups, including chromatin-associated proteins, transcriptional co-regulators, RNA-binding proteins, mitochondrial proteins, and protein kinases.  By focusing on a well-studied MAP kinase, Erk2, we demonstrated that it acts as a transcription repressor in the interferon-gamma (IFN-g) signaling pathway via direct binding to promoters of IFN-g induced genes. The conceptual breakthrough is that a large number of unconventional DNA-binding proteins that act as transcription factors exists, suggesting an unexpectedly complex protein-DNA interaction landscape in humans. Part of this was published in a 2009 Cell paper.

DNA methylation, especially CpG methylation at promoter regions, has been generally considered a potent epigenetic modification that prohibits transcription factor (TF) recruitment, resulting in transcription suppression.  So far, only MeCP2, MBD1, MBD2, and a few zinc finger proteins have been identified as bona fide methylated DNA-binding proteins.  Important questions are whether members of other TF subfamilies can specifically recognize methylated CpG (mCpG) and, if they do, what the physiological consequences are.  To address these questions, we recently selected 154 of the 460 DNA motifs (see above) that carry at least one CpG, methylated them, and screened against the human TF microarrays in the presence of excess unmethylated motif DNAs as competitors.  This competitive binding assay resulted in the identification of numerous human TFs with mCpG-dependent DNA-binding activities across many TF subfamilies.  Interestingly, some TFs exhibit a dual specificity in that they can bind to methylated and unmethylated DNA motifs of distinct sequences.  To elucidate the underlying mechanism, we focused on Krüppel-like factor 4 (KLF4) and decoupled its mCpG- and CpG-binding activities via site-directed mutagenesis.  In addition, using ChIP-bisulfite sequencing, we showed that the endogenous KLF4 binds specific methylated or unmethylated motifs in human embryonic stem cells in vivo. The conceptual breakthrough of this study is that mCpG-dependent TF binding activity is a widespread phenomenon that might play an active role in transcription regulation. Our study provides a new framework to understand the role and mechanism of TFs in epigenetic regulation of gene transcription. Part of this work has been accepted by eLife.

4.  Protein Posttranslational Modifications (PTMs)

A prominent advantage of protein microarray technology is its ability to identify protein substrates of various types of PTM enzymes.  In the past, my group has developed a variety of enzymatic reactions to dissect functions of lysine modification enzymes, including protein acetyltransferases, ubiquitin E3 and SUMO E3 ligases, and protein methylases.  The knowledge gained is expected to help us build complex protein regulation networks and pathways.

In a collaborative effort with the Berger (U Penn), Boeke (JHU), and Zhao (U Chicago) groups, we developed protein acetylation reaction on the yeast proteome microarray to identify substrates of the NuA4 and SAGA complexes.  We were among the first to identify >90 non-histone substrates in yeast, many of which are cytosolic proteins, including metabolic enzymes and signaling components.  Using a combination of genetics, biochemistry, and molecular and cell biology approaches, we performed in-depth in vivo analyses of two NuA4 substrates, namely phosphoenolpyruvate carboxykinase (Pck1p) and the beta subunit of the SNF1 kinase complex (Sip2p).  We found that acetylation on K514 of Pck1p by the NuA4 complex elevates Pck1’s enzymatic activity, which is reversed by a HDAC Sir2.  More strikingly, both acetylation mimetics (K514-to-Q) of PCK1 and sir2D mutant showed a much enhanced survival rate compared with the WT in the chronological longevity assay.  Therefore, a new ageing pathway via gluconeogenesis pathway was identified (Lin et al, Cell 2009).  By following another NuA4 substrate, Sip2, we found that its acetylation, controlled by antagonizing NuA4 acetyltransferase and Rpd3 deacetylase, enhances interaction with Snf1 kinase, the catalytic subunit of Snf1 complex. Sip2-Snf1 interaction inhibits Snf1 activity, thus decreasing phosphorylation of a downstream target, Sch9 (homolog of Akt/S6K), and ultimately leading to slower growth but extended replicative life span. Sip2 acetylation mimetics are more resistant to oxidative stress.  We further demonstrated that the anti-aging effect of Sip2 acetylation is independent of extrinsic nutrient availability and TORC1 activity.  We propose a protein acetylation-phosphorylation cascade that regulates Sch9 activity, controls intrinsic aging, and extends replicative life span in yeast (Lu et al, Cell 2011).  The conceptual breakthrough of our studies is the identification the characterization of two new parallel metabolic pathways involved in yeast ageing via acetylation regulation.

In parallel, my group was among the first to develop ubiquitylation reactions to identify substrates of ubiquitin E3 ligases.  We employed a HECT domain ubiquitylation E3 ligase, Rsp5, to develop in vitro ubiquitylation reactions on the yeast proteome chips and identified 86 candidate substrates of Rsp5. Using genetic screens and in vitro and in vivo biochemical analyses, we validated eight new substrates in vivo.  We then used a combination of genetic, biochemical, and cell biological analyses to determine whether these ubiquitylation events were physiologically relevant and found two novel functions of Rsp5: it positively regulates ribonucleotide reductase-dependent DNA synthesis by managing the cellular localization of Rnr2 and influences chromosome segregation through regulating Nkp2 and Nsl1 activity.  We believe that the same paradigm can be extended to study RING domain E3 ligases, the study of which will eventually accumulate enough data to build comprehensive regulation networks of ubiquitylation (Lu et al, Mol Cell Proteomics 2008).

Protein phosphorylation, mediated by protein kinases, is one of the most widespread regulatory mechanisms in eukaryotes. Recently, several high-throughput studies designed to analyze the global properties of phosphorylation networks in various model organisms have been reported. These studies, which employed approaches based on protein-protein interactions, genetic interactions, gene expression profiling, and motif-based predictions, have uncovered important clues about the organization and regulation of kinase-mediated signaling pathways. However, they are each limited in their ability to identify direct enzymatic interactions between kinases and their substrates, which requires biochemical analysis of kinase-substrate relationships (KSRs) using purified protein components. However, global analysis of activity-based phosphorylation networks, built upon direct KSRs, is lacking in higher eukaryotes. Indeed, only ~2,000 human KSRs have been experimentally identified to date. In contrast, >70,400 in vivo phosphorylated serine, threonine, and tyrosine residues have been characterized by mass spectrometry. Together, this implies that, for the vast majority of identified in vivo phosphorylation sites, the specific kinase(s) responsible for the phosphorylation event remains unknown.

In a collaborative effort with the J Qian and J Zhang groups, we developed a new strategy, called CEASAR, based on functional protein microarrays and bioinformatics to experimentally identify substrates for 289 unique kinases, resulting in 3,656 high-quality KSRs. We then generated consensus phosphorylation motifs for each of the kinases and integrated this information, along with information about in vivo phosphorylation sites determined by mass spectrometry, to construct a high-resolution map of phosphorylation networks that connects 230 kinases to 2,591 in vivo phosphorylation sites in 652 substrates. The value of this dataset is demonstrated through the discovery of a new role for PKA downstream of Btk during B cell receptor signaling. Overall, these studies provide global insights into kinase-mediated signaling pathways and promise to advance our understanding of cellular signaling processes in humans.  Part of this work has been published in a 2013 Mol Syst Biol paper.

5. Pathogen-Host Interactions

Although most viruses encode limited genetic information, they fulfill their life cycles by hijacking the host cell machinery, mostly via direct physical contact with the host components.  However, these physical contacts are of different flavors, such as protein-protein, protein-RNA, protein-DNA, and enzyme-substrate interactions, which make it rather challenging to study using traditional methods.  Therefore, we envisioned that functional protein microarrays could serve as a versatile screening platform for identifying direct, physical interactions between a pathogen and host at the molecular level.  Indeed, in collaboration with several virology experts, we identified several new host pathways that were proved important for the life cycles of the pathogens of interest.

Our studies in human herpesvirus-host interactions serve as a good example.  The human a, b, and g herpesviruses infect different tissues and cause distinct diseases; however, they each confront many of the same challenges in infecting their hosts, including reprogramming cellular gene expression and reactivating the lytic life cycle to produce new virions and spread infection.  Therefore, an attractive hypothesis was that the shared substrates targeted by these orthologous viral kinases would reveal host pathways critical for replication across the herpesvirus family.  To test this hypothesis, we performed phosphorylation reactions on the human protein microarrays with four orthologous kinases encoded by EBV, KSHV, HCMV, and HSV-1 and identified 110 shared substrates.  Strikingly, 15 proteins formed a highly connected cluster and were involved in the DNA damage response.  By focusing on a protein acetyltransferases TIP60, an upstream master regulator, we showed that TIP60 is required for lytic replication in all four viruses.  Using a series of cell-based assays, we further demonstrated that upon EBV lytic induction, TIP60 is activated by EBV BGLF4 kinase and recruited to the EBV genome to induce expression of three lytic genes.  This work established a new paradigm to identify a conserved host pathway targeted by related pathogens by probing the host protein microarrays with orthologous pathogen factors. Part of this work was published in a 2011 Cell Host Path paper.

6. Biomarker Identification

A rapidly growing application of the protein microarray technology in translational medicine is biomarker identification, which focuses on the diagnostic identification of antibodies associated with a disease. These antibodies can be produced as part of an immune response to an infection, against a foreign protein, or even against a person’s own proteins. When proteins on a protein microarray are viewed as potential antigens, protein microarrays become a platform to identify autoantibodies that show statistically significant association with an infection or with a disease of interest. In the past eight years, my group has published a series of papers in the area of biomarker identification. For example, we have applied the E. coli K12 proteome microarrays (Chen et al, Nat Methods 2007) to identify new biomarkers that could differentiate Crohn’s disease (CD) from ulcerative colitis (UC) in inflammatory bowel disease (IBD) patients (Chen et al, Mol Cell Proteomics 2009). A similar approach has been applied to profile humoral immune responses of both IgG and IgA isotypes to two human herpesviruses, Epstein-Barr virus (EBV) and Kaposi's sarcoma-associated herpesvirus (KSHV). Isotype-specific biomarkers were identified in patients’ plasma samples (Zheng et al, J Infect Dis 2011).

More important, we have developed a two-phase strategy to identify biomarkers in primary biliary cirrhosis (PBC) using the human protein microarrays. Although human protein microarray can serve as a powerful tool to discover novel biomarkers for human autoimmune diseases, a major hurdle is the high cost of protein microarrays, especially when a large number of samples are used. In many cases, smaller cohorts were used to reduce the cost. However, overfitting can be a serious problem when dealing with a complex disease that involves multiple factors. To reduce the cost, we screened a smaller cohort (i.e., 30 PBC versus 30 control sera) on the high-cost human protein microarray to discover candidate biomarkers in Phase I. To alleviate the overfitting problem, in Phase II we fabricated a focused protein microarray with those candidate biomarkers and screened with a much larger cohort (i.e., 50 PBC versus 194 control sera). This approach allowed us to validated three new biomarkers associated with PBC (Hu et al, Mol Cell Proteomics 2012).


Representative Publications:

  • Hu, S., Wan, J., Su, Y., Song, Q., Zeng, Y., Nguyen, H.N., Shin, J., Cox, E., Rho, H.S., Woodard, C., Xia, S., Liu, S., Lu, H., Ming, G., Wade, H., Song, H., Qian, J., Zhu, H*.  DNA methylation presents distinct binding sites for human transcription factors. eLife 2013; in press.
  • Newman, R.H., Hu, J., Rho, H.S., Xie, Z., Woodard, C., Neiswinger, J., Hwang, W., Shirley, M., Hu, S., Cooper, C., Jeong, J.S., Wu, G., Lin, J., Gao, X., Ni, Q., Dalby, K., Ji, H., Desiderio, S., Birnbaum, M.J., Cole, P.A., Knapp, S., Ryazanov, A., Zack, D.J., Blackshaw, S., Pawson, T., Gingras, A.-C., Pandey, A., Turk, B.E., Zhang, J., Zhu, H*., Qian, J.  Construction of human activity-based phosphorylation networks. Mol Syst Biol 2013; 9:655. Pub Med Reference
  • Jeong, J.S., Jiang, L., Albino, E., Marrero, J., Rho, H.S., Hu, J., Hu, S., Vera, C., Bayron-Poueymiroy, D., Rivera-Pacheco, Z.A., Ramos, L., Torres-Castro, C., Qian, J., Bonaventura, J., Boeke, J.D., Yap, W.Y., Pino, I., Eichinger, D.J., Zhu, H*., Blackshaw, S.  Rapid identification of monospecific monoclonal antibodies using a human proteome microarray. Mol Cell Proteomics 2012; 11(6):O111.016253. Pub Med Reference
  • Lu, J.-Y., Lin, Y.-Y., Sheu, J.-C., Wu, J.-T., Lee, F.-J., Chen, Y., Lin, M.-I., Chiang, F.-T., Tai, T.-Y., Berger, S.L., Zhao, Y., Tsai, K.-S., Zhu H*., Chuang, L.-M., and Boeke, J.D.  Acetylation of AMPK controls intrinsic aging independently of caloric restricion. Cell 2011; 146:969-79. Pub Med Reference
  • Hu S, Xie Z, Onishi A, Yu X, Jiang L, Lin J, Rho H-S, Woodard C, Wang H, Jeong J-S, Long S, He X, Wade H, Blackshaw S, Qian J, Zhu H*. Profiling the Human Protein-DNA Interactome Reveals MAPK1 as a Transcriptional Repressor of Interferon Signaling. Cell 2009; Oct 30;139(3):610-22. Pub Med Reference
  • Kung L, Tao, S-C, Qian, J, Smith M, Snyder M, Zhu H*. Global analysis of the glycoproteome in S. cerevisiae reveals new roles for protein glycosylation. Mol Syst Biol 2009; 5:308. Pub Med Reference
  • Lin Y-y, Lu J-y, Zhang J, Walter W, Dang W, Wan J, Tao SC, Qian J, Zhao Y, Boeke JD, Berger SL, Zhu H*. Protein acetylation microarray reveals NuA4 controls key metabolic target regulating gluconeogenesis.Cell 2009; 136:1073-1084. Pub Med Reference
  • Chen C-S, Korobkova E, Chen H, Zhu J, Jian X, Tao S-C, He H, Zhu H*. A proteome chip approach reveals new DNA damage recognition activities in Escherichia coli. Nature Methods 2008; 5:69-74. Pub Med Reference
  • Tao, S.-C. and Zhu, H*. Protein chip fabrication by capture of nascent polypeptides. Nature Biotech. 24:1253-1254, 2006. Pub Med Reference
  • Li R, Zhu J, Xie Z, Liao G, Liu J, Chen M-R, Hu S, Woodard C, Lin J, Taverna S, Desai P, Ambinder R, Hayward GS, Qian J, Zhu H*, Hayward SD. Conserved herpesvirus kinases target the DNA damage response pathway and TIP60 histone acetyltransferase to promote virus replication. Cell Host Microbe 2011; 10(4):390-400. Pub Med Reference


Other graduate programs in which Dr. Zhu participates:

BCMB Graduate Program
Chemistry-Biology Interface Program