Local FK506 delivery induces osteogenesis in in vivo rat bone defect and rabbit spine fusion models

Bone grafting procedures are commonly used for the repair, regeneration, and fusion of bones in in a wide range of orthopaedic surgeries, including large bone defects and spine fusion procedures. Autografts are the clinical gold standard, though recombinant human bone morphogenetic proteins (rhBMPs) are often used, particularly in difficult clinical situations. However, treatment with rhBMPs can have off-target effects and significantly increase surgical costs, adding to patients' already high economic and mental burden. Recent studies have identified that FDA-approved immunosuppressant drug, FK506 (Tacrolimus), can also activate the BMP pathway by binding to its inhibitors. This study tested the hypothesis that FK506, as a standalone treatment, could induce osteogenic differentiation of human mesenchymal stromal cells (hMSCs), as well as functional bone formation in a rat segmental bone defect model and rabbit spinal fusion model. FK506 potentiated the effect of low dose BMP-2 to enhance osteogenic differentiation and mineralization of hMSCs in vitro. Standalone treatment with FK506 delivered on a collagen sponge, produced consistent bone bridging of a rat critically-sized femoral defect with functional mechanical properties comparable to naïve bone. In a rabbit single level posterolateral spine fusion model, treatment with FK506 delivered on a collagen sponge successfully fused the L5-L6 vertebrae at rates comparable to rhBMP-2 treatment. These data demonstrate the ability of FK506 to induce bone formation in human cells and two challenging in vivo models, and indicate FK506 can be utilized either as a standalone treatment or in conjunction with rhBMP to treat a variety of spine disorders.

Sclerostin small-molecule inhibitors promote osteogenesis by activating canonical Wnt and BMP pathways

Background

The clinical healing environment after a posterior spinal arthrodesis surgery is one of the most clinically challenging bone-healing environments across all orthopedic interventions due to the absence of a contained space and the need to form de novo bone. Our group has previously reported that sclerostin in expressed locally at high levels throughout a developing spinal fusion. However, the role of sclerostin in controlling bone fusion remains to be established.

Methods

We computationally identified two FDA-approved drugs, as well as a single novel small-molecule drug, for their ability to disrupt the interaction between sclerostin and its receptor, LRP5/6. The drugs were tested in several in vitro biochemical assays using murine MC3T3 and MSCs, assessing their ability to (1) enhance canonical Wnt signaling, (2) promote the accumulation of the active (non-phosphorylated) form of β-catenin, and (3) enhance the intensity and signaling duration of BMP signaling. These drugs were then tested subcutaneously in rats as standalone osteoinductive agents on plain collagen sponges. Finally, the top drug candidates (called VA1 and C07) were tested in a rabbit posterolateral spine fusion model for their ability to achieve a successful fusion at 6 wk.

Results

We show that by controlling GSK3b phosphorylation our three small-molecule inhibitors (SMIs) simultaneously enhance canonical Wnt signaling and potentiate canonical BMP signaling intensity and duration. We also demonstrate that the SMIs produce dose-dependent ectopic mineralization in vivo in rats as well as significantly increase posterolateral spine fusion rates in rabbits in vivo, both as standalone osteogenic drugs and in combination with autologous iliac crest bone graft.

Conclusions

Few if any osteogenic small molecules possess the osteoinductive potency of BMP itself - that is, the ability to form de novo ectopic bone as a standalone agent. Herein, we describe two such SMIs that have this unique ability and were shown to induce de novo bone in a stringent in vivo environment. These SMIs may have the potential to be used in novel, cost-effective bone graft substitutes for either achieving spinal fusion or in the healing of critical-sized fracture defects.

Funding

This work was supported by a Veteran Affairs Career Development Award (IK2-BX003845).

Multifocal heterotopic ossification in a man with germline variants of LIM Mineralization Protein-1 (LMP-1)

A 54-year-old man with a history of unimelic, post-traumatic multifocal heterotopic ossification (HO) and normal genetic analysis of ACVR1 and GNAS had variants of unknown significance (VUS) in PDLIM-7 (PDZ and LIM Domain Protein 7), the gene encoding LMP-1 (LIM Mineralization Protein-1), an intracellular protein involved in the bone morphogenetic protein (BMP) pathway signaling and ossification. In order to determine if the LMP-1 variants were plausibly responsible for the phenotype observed, a series of in vitro experiments were conducted. C2C12 cells were co-transfected with a BMP-responsive reporter as well as the LMP-1 wildtype (wt) construct or the LMP-1^T161I or the LMP-1^D181G constructs (herein designated as LMP-161 or LMP-181) corresponding to the coding variants detected in the patient. A significantly increased BMP-reporter activity was observed in LMP-161 or LMP-181 transfected cells compared to the wt cells. The LMP-181 variant exhibited BMP-reporter activity with a four-fold increase over the LMP-1 wt protein. Similarly, mouse pre-osteoblastic MC3T3 cells transfected with the patient's LMP-1 variants expressed higher levels of osteoblast markers both at mRNA and protein levels and preferentially mineralized when stimulated with recombinant BMP-2 compared to control cells. Presently, there are no pathogenic variants of LMP-1 known to induce HO in humans. Our findings suggest that the germline variants in LMP-1 detected in our patient are plausibly related to his multifocal HO (LMP1-related multifocal HO). Further observations will be required to firmly establish this gene-disease relationship.

Correction: Sangadala, S. et al. FK506 Induces Ligand-Independent Activation of the Bone Morphogenetic Protein Pathway and Osteogenesis. Int. J. Mol. Sci. 2019, 20, 1900

The authors wish to make the following corrections to this paper [...]

Noggin regulates foregut progenitor cell programming, and misexpression leads to esophageal atresia

Esophageal atresia (EA/TEF) is a common congenital abnormality present in 1 of 4000 births. Here we show that atretic esophagi lack Noggin (NOG) expression, resulting in immature esophagus that contains respiratory glands. Moreover, when using mouse esophageal organoid units (EOUs) or tracheal organoid units (TOUs) as a model of foregut development and differentiation in vitro, NOG determines whether foregut progenitors differentiate toward esophageal or tracheal epithelium. These results indicate that NOG is a critical regulator of cell fate decisions between esophageal and pulmonary morphogenesis, and its lack of expression results in EA/TEF.

Link protein N-terminal peptide binds to bone morphogenetic protein (BMP) type II receptor and drives matrix protein expression in rabbit intervertebral disc cells

Intervertebral disc (IVD) degeneration and associated spinal disorders are leading sources of morbidity, and they can be responsible for chronic low back pain. Treatments for degenerative disc diseases continue to be a challenge. Intensive research is now focusing on promoting regeneration of degenerated discs by stimulating production of the disc matrix. Link protein N-terminal peptide (LPP) is a proteolytic fragment of link protein, an important cross-linker and stabilizer of the major structural components of cartilage, aggrecan and hyaluronan. In this study we investigated LPP action in rabbit primary intervertebral disc cells cultured ex vivo in a three-dimensional alginate matrix. Our data reveal that LPP promotes disc matrix production, which was evidenced by increased expression of the chondrocyte-specific transcription factor SOX9 and the extracellular matrix macromolecules aggrecan and collagen II. Using colocalization and pulldown studies we further document a noggin-insensitive direct peptide-protein association between LPP and BMP-RII. This association mediated Smad signaling that converges on BMP genes leading to expression of BMP-4 and BMP-7. Furthermore, through a cell-autonomous loop BMP-4 and BMP-7 intensified Smad1/5 signaling though a feedforward circuit involving BMP-RI, ultimately promoting expression of SOX9 and downstream aggrecan and collagen II genes. Our data define a complex regulatory signaling cascade initiated by LPP and suggest that LPP may be a useful therapeutic substitute for direct BMP administration to treat IVD degeneration and to ameliorate IVD-associated chronic low back pain.

A novel low-molecular-weight compound enhances ectopic bone formation and fracture repair

BACKGROUND

Use of recombinant human bone morphogenetic protein-2 (rhBMP-2) is expensive and may cause local side effects. A small synthetic molecule, SVAK-12, has recently been shown in vitro to potentiate rhBMP-2-induced transdifferentiation of myoblasts into the osteoblastic phenotype. The aims of this study were to test the ability of SVAK-12 to enhance bone formation in a rodent ectopic model and to test whether a single percutaneous injection of SVAK-12 can accelerate callus formation in a rodent femoral fracture model.

METHODS

Collagen disks with rhBMP-2 alone or with rhBMP-2 and SVAK-12 were implanted in a standard athymic rat chest ectopic model, and radiographic analysis was performed at four weeks. In a second set of rats (Sprague-Dawley), SVAK-12 was percutaneously injected into the site of a closed femoral fracture. The fractures were analyzed radiographically and biomechanically (with torsional testing) five weeks after surgery.

RESULTS

In the ectopic model, there was dose-dependent enhancement of rhBMP-2 activity with use of SVAK-12 at doses of 100 to 500 μg. In the fracture model, the SVAK-12-treated group had significantly higher radiographic healing scores than the untreated group (p = 0.028). Biomechanical testing revealed that the fractured femora in the 200 to 250-μg SVAK-12 group were 43% stronger (p = 0.008) and 93% stiffer (p = 0.014) than those in the control group. In summary, at five weeks the femoral fracture group injected with SVAK-12 showed significantly improved radiographic and biomechanical evidence of healing compared with the controls.

CONCLUSIONS

A single local dose of a low-molecular-weight compound, SVAK-12, enhanced bone-healing in the presence of low-dose exogenous rhBMP-2 (in the ectopic model) and endogenous rhBMPs (in the femoral fracture model).

CLINICAL RELEVANCE

This study demonstrates that rhBMP-2 responsiveness can be enhanced by a novel small molecule, SVAK-12. Local application of anabolic small molecules has the potential for potentiating and accelerating fracture-healing. Use of this small molecule to lower required doses of rhBMPs might both decrease their cost and improve their safety profile.

Osteoinductive LIM mineralization protein-1 suppresses activation of NF-kappaB and selectively regulates MAPK pathways in pre-osteoclasts

LIM mineralization protein-1 (LMP-1) is an intracellular regulator of bone formation and has been shown to be osteoinductive in vitro and in vivo. The effect of LMP-1 on other aspects of bone homeostasis has not been previously studied. In a pilot study we observed that LMP-1 decreased nitric oxide (NO) production in pre-osteoclasts. Here we report a new anti-inflammatory effect of LMP-1 and define its mechanism of action in lipopolysaccharide (LPS)-stimulated RAW 264.7 pre-osteoclasts. We found that LMP-1 significantly inhibited LPS-induced NO production. LMP-1 also effectively inhibited the expression of inducible nitric oxide synthase (iNOS), potently suppressed the transcriptional activity and nuclear translocation of nuclear factor kappa B (NF-kappaB), and prevented the phosphorylation of inhibitor of kappa B (IkappaB). Interestingly, LMP-1 had no effect on Receptor-Activator of Nuclear Factor B Ligand (RANKL)-induced activation of NF-kappaB. Furthermore, LMP-1 had no effect on the LPS-induced phosphorylation of extracellular signal-regulated kinase 1/2 (ERK1/2), whereas it did attenuate the phosphorylation of c-Jun NH2-terminal kinase (JNK) while enhancing phosphorylation of p38 mitogen-activated protein kinases (p38 MAPK). These results suggest that LMP-1 has an anti-inflammatory effect, and this effect is, at least in part, due to the inhibition of NO production by the suppression of NF-kappaB activation and selective regulation of mitogen-activated protein kinase (MAPK) pathways.

Development and optimization of a cell-based assay for the selection of synthetic compounds that potentiate bone morphogenetic protein-2 activity

The requirement of large amounts of the recombinant human bone morphogenetic protein-2 (BMP-2) produces a huge translational barrier for its routine clinical use due to high cost. This leads to an urgent need to develop alternative methods to lower costs and/or increase efficacies for using BMP-2. In this study, we describe the development and optimization of a cell-based assay that is sensitive, reproducible, and reliable in identifying reagents that potentiate the effects of BMP-2 in inducing transdifferentiation of C2C12 myoblasts into the osteoblastic phenotype. The assay is based on a BMP-responsive Smad1-driven luciferase reporter gene. LIM mineralization protein-1 (LMP-1) is a novel intracellular LIM domain protein that has been shown by our group to enhance cellular responsiveness to BMP-2. Our previous report elucidated that the binding of LMP-1 with the WW2 domain in Smad ubiquitin regulatory factor-1 (Smurf1) rescues the osteogenic Smads from degradation. Here, using the optimized cell-based assay, we first evaluated the activity of the recombinantly prepared proteins, LMP-1, and its mutant (LMP-1DeltaSmurf1) that lacks the Smurf1-WW2 domain-binding motif. Both the wild type and the mutant proteins were engineered to contain an 11-amino acid HIV-TAT protein derived membrane transduction domain to aid the cellular delivery of recombinant proteins. The cell-based reporter assay confirmed that LMP-1 potentiates the BMP-induced stimulation of C2C12 cells towards the osteoblastic phenotype. The potentiating effect of LMP-1 was significantly reduced when a specific-motif known to interact with Smurf1 was mutated. We validated the results obtained in the reporter assay by also monitoring the expression of mRNA for osteocalcin and alkaline phosphatase (ALP) which is widely accepted osteoblast differentiation marker genes. Finally, we provide further confirmation of our results by measuring the activity of alkaline phosphatase in support of the accuracy and reliability of our cell-based assay. Direct delivery of synthesized protein can be limited by high cost, instability or inadequate post-translational modifications. Thus, there would be a clear benefit for a low cost, cell penetrable chemical compound. We successfully used our gene expression-based assay to choose an active compound from a select group of compounds that were identified by computational screenings as the most likely candidates for mimicking the function of LMP-1. Among them, we selected SVAK-3, a compound that showed a dose-dependent potentiation of BMP-2 activity in inducing osteoblastic differentiation of C2C12 cells. We show that either the full length LMP-1 protein or its potential mimetic compound consistently exhibit similar potentiation of BMP-2 activity even when multiple markers of the osteoblastic phenotype were parallely monitored.

Virtual screening and selection of drug-like compounds to block noggin interaction with bone morphogenetic proteins

Noggin is a major natural extracellular antagonist to bone morphogenetic proteins (BMPs) which binds to BMPs and blocks binding of them to BMP-specific receptors and thus negatively regulates BMP-induced osteoblastic differentiation. Bone morphogenetic proteins (BMPs) signal through heteromeric protein complexes composed of type I and type II serine/threonine kinase receptors. Preventing the BMP-2/noggin interaction will preserve free BMP-2 and enhance the efficacy of BMP-2 to induce bone formation. This work is an attempt to use the current understanding of BMP-2, and its interaction with its receptors and antagonist to design an inhibitor of BMP-2/noggin interaction with the goal of lowering the dose of BMP-2 required in clinical applications. The crystal structure of the BMP-7/noggin complex, the BMP-2/BMP receptor IA ectodomain complex and the extracellular domain of BMP receptor II monomer are known. We modeled the BMP-2 based on the structure of its homologue BMP-7 and its binding complex with noggin. We also modeled a complex of BMP-2/BMPRIA/BMPRII by modeling BMPRII and replacing ActRIIB in the BMP-2/BMPRIA/ActRIIB complex. We then identified the binding region of noggin with BMP-2 and the receptors with BMP-2. From the analysis of structures of these complexes and modeling we identified the key amino acids present in the entire interacting surfaces among these proteins that play important physiological role in the regulation of cell differentiation and bone metabolism. By in silico screening we selected and ranked several compounds that have high theoretical scores to bind to noggin to block BMP-noggin interaction.

Engineering, cloning, and functional characterization of recombinant LIM mineralization protein-1 containing an N-terminal HIV-derived membrane transduction domain

Short peptide sequences known as protein transduction domains have become increasingly prevalent as tools to internalize molecules that would otherwise remain extracellular. Here, we determine whether a purified recombinant mammalian intracellular osteogenic factor delivered by a HIV-derived TAT-peptide tag is indeed capable of intracellular localization in a form accessible to interaction with other proteins. We engineered and bacterially expressed a TAT-fusion-cDNA construct of a known osteogenic factor, LIM mineralization protein-1 (LMP-1) involved in the bone morphogenetic protein (BMP) pathway that has the potential to serve as an enhancer of BMP-2 efficacy. The expressed recombinant protein contains an N-terminal (His)(6)-tag, a hemagglutinin(HA)-tag, and an 11-amino acid HIV-derived TAT-membrane transduction domain and was purified to homogeneity by Sephacryl S-100 molecular exclusion and Ni(2+)-affinity chromatography. The purified TAT-LMP-1 protein was chemically labeled with fluorescein, and its time and concentration dependent entry into rabbit blood cells was monitored by flow cytometry. We demonstrate the accumulation of TAT-tagged LMP-1 both in cytoplasmic and nuclear compartments. By performing affinity pull-down assays, we confirm our earlier findings that the recombinant TAT-LMP-1, when used as molecular bait to identify the intracellular binding proteins, interacts with Smurf1, a known binding partner of LMP-1. We also show potentiation of BMP-2 activity using the purified TAT-LMP-1 in mouse muscle C2C12 cells by assaying a heterologous luciferase-reporter construct containing multiple copies of a BMP-responsive sequence motif. Finally, we also confirm the biological activity of the purified TAT-LMP-1 by showing enhancement of BMP-2 induced increase of alkaline phosphatase mRNA and protein by RT-PCR and enzyme activity, respectively.

Modeling and analysis of MH1 domain of Smads and their interaction with promoter DNA sequence motif

The Smads are a group of related intracellular proteins critical for transmitting the signals to the nucleus from the transforming growth factor-beta (TGF-beta) superfamily of proteins at the cell surface. The prototypic members of the Smad family, Mad and Sma, were first described in Drosophila and Caenorhabditis elegans, respectively. Related proteins in Xenopus, Humans, Mice and Rats were subsequently identified, and are now known as Smads. Smad protein family members act downstream in the TGF-beta signaling pathway mediating various biological processes, including cell growth, differentiation, matrix production, apoptosis and development. Smads range from about 400-500 amino acids in length and are grouped into the receptor-regulated Smads (R-Smads), the common Smads (Co-Smads) and the inhibitory Smads (I-Smads). There are eight Smads in mammals, Smad1/5/8 (bone morphogenetic protein regulated) and Smad2/3 (TGF-beta/activin regulated) are termed R-Smads, Smad4 is denoted as Co-Smad and Smad6/7 are inhibitory Smads. A typical Smad consists of a conserved N-terminal Mad Homology 1 (MH1) domain and a C-terminal Mad Homology 2 (MH2) domain connected by a proline rich linker. The MH1 domain plays key role in DNA recognition and also facilitates the binding of Smad4 to the phosphorylated C-terminus of R-Smads to form activated complex. The MH2 domain exhibits transcriptional activation properties. In order to understand the structural basis of interaction of various Smads with their target proteins and the promoter DNA, we modeled MH1 domain of the remaining mammalian Smads based on known crystal structures of Smad3-MH1 domain bound to GTCT Smad box DNA sequence (1OZJ). We generated a B-DNA structure using average base-pair parameters of Twist, Tilt, Roll and base Slide angles. We then modeled interaction pose of the MH1 domain of Smad1/5/8 to their corresponding DNA sequence motif GCCG. These models provide the structural basis towards understanding functional similarities and differences among various Smads.

Expression, purification and mass spectrometric analysis of LIM mineralization protein-1 in human lung epithelial cells

LIM mineralization protein-1 (LMP-1) is a novel osteoinductive protein that has been cloned and shown to induce bone formation both in vitro and in vivo. Detection and evaluation of the possible presence of carbohydrate structures in LMP-1 is an important regulatory consideration for the therapeutic use of recombinantly expressed protein. The sequence of LMP-1 contains a highly conserved N-terminal PDZ domain and three C-terminal LIM domains. The sequence analysis of LMP-1 predicts two potential N-glycosylation sites and several O-glycosylation sites. Here, we report the cloning and overexpression of LMP-1 in human lung carcinoma (A549) cells. Even though our group already reported the sequence of LMP-1 cDNA, we undertook this work to clarify whether or not the overexpressed protein undergoes any glycosylation in vivo. The expressed full-length recombinant protein was purified and subjected to chemical analysis and internal sequencing. The absence of any hexosamines (N-acetyl glucosamine or N-acetyl galactosamine) in chemical composition analysis of LMP-1 protein revealed that there is little or no post-translational glycosylation of the LMP-1 polypeptide in lung carcinoma cells (A549). We performed in-gel trypsin digestion on purified LMP-1, and the resulting peptide digests were analyzed further using matrix-assisted laser desorption and ionization mass spectrometry for peptide mass finger printing, which produced several exact matches with the corresponding LMP-1 peptides. Separation by high performance liquid chromatography and purification of the desired peptides followed by N-terminal sequencing resulted in many exact LMP-1 matches for several purified peptides, thus establishing the identity of the purified protein as LMP-1.

Modeling and analysis of molecularinteraction between Smurf1-WW2 domain and various isoforms of LIM mineralization protein

LIM Mineralization Protein-1 (LMP-1) has been cloned and shown to be osteoinductive. Our efforts to understand the mode of action of LMP-1 led to the determination that LMP-1 interacts with Smad Ubiquitin Regulatory Factor-1 (Smurf1). Smurf1 targets osteogenic Smads, Smad1/5, for ubiquitin-mediated proteasomal degradation. Smurf1 interaction with LMP-1 or Smads is based on the presence of unique WW-domain interacting motif in these target molecules. By performing site-directed mutagenesis and binding studies in vitro on purified recombinant proteins, we identified a specific motif within the osteogenic region of several LMP isoforms that is necessary for Smurf1 interaction. Similarly, we have identified that the WW2 domain of Smurf1 is necessary for target protein interaction. Here, we present a homology-based modeling of the Smurf1 WW2 domain and its interacting motif of LMP-1. We performed computational docking of the interacting domains in Smurf1 and LMPs to identify the key amino acid residues involved in their binding regions. In support of the computational predictions, we also present biochemical evidence supporting the hypothesis that the physical interaction of Smurf1 and osteoinductive forms of LMP may prevent Smurf1 from targeting osteogenic Smads by ubiquitin-mediated proteasomal degradation.

Truncated human LMP-1 triggers differentiation of C2C12 cells to an osteoblastic phenotype in vitro

LIM mineralization protein-1 (LMP-1) is a novel intracellular osteoinductive protein that has been shown to induce bone formation both in vitro and in vivo. LMP-1 contains an N-terminal PDZ domain and three C-terminal LIM domains. In this study, we investigated whether a truncated form of human LMP-1 (hLMP-1[t]), lacking the three C-terminal LIM domains, triggers the differentiation of pluripotent myoblastic C(2)C(12) cells to the osteoblast lineage. C(2)C(12) cells were transiently transduced with Ad5-hLMP-1(t)-green fluorescent protein or viral vector control. The expression of hLMP-1(t) RNA and the truncated protein were examined. The results showed that hLMP-1(t) blocked myotube formation in C(2)C(12) cultures and significantly enhanced the alkaline phosphatase (ALP) activity. In addition, the expressions of ALP, osteocalcin, and bone morphogenetic protein (BMP)-2 and BMP-7 genes were also increased. The induction of these key osteogenic markers suggests that hLMP-1(t) can trigger the pluripotent myoblastic C2C12 cells to differentiate into osteoblastic lineage, thus extending our previous observation that LMP-1 and LMP-1(t) enhances the osteoblastic phenotype in cultures of cells already committed to the osteoblastic lineage. Therefore, C(2)C(12) cells are an appropriate model system for the examination of LMP-1 induction of the osteoblastic phenotype and the study of mechanisms of LMP-1 action.

Molecular interaction between Smurf1 WW2 domain and PPXY motifs of Smad1, Smad5, and Smad6--modeling and analysis

The ubiquitin-proteasome proteolytic pathway is essential for various important biological processes including cell cycle progression, gene transcription, and signal transduction. One of the important regulatory mechanisms by which the bone-inducing activity of the bone morphogenetic protein (BMP) signaling is modulated involves ubiquitin-mediated proteasomal degradation. The BMP induced receptor signal is transmitted intracellularly by phosphorylation of Smad proteins by the activated receptor I. The phosphorylated Smads 1, 5, and 8 (R-Smads) oligomerize with the co-Smad (Smad4). The complex, thus, formed translocates to the nucleus and interacts with other cofactors to regulate the expression of downstream target genes. R-Smads contain PPXY motif in the linker region that interacts with Smad ubiquitin regulatory factor 1 (Smurf1), an E3 ubiquitin ligase that catalyzes ubiquitination of target proteins for proteasomal degradation. Smurf1 contains a HECT domain, a C2 domain, and 2 WW domains (WW1, WW2). The PPXY motif in target proteins and its interaction with Smurf1 may form the basis for regulation of steady-state levels of Smads in controlling BMP-responsiveness of cells. Here, we present a homology-based model of the Smurf1 WW2 domain and the target octa-peptides containing PPXY motif of Smurf1-interacting Smads. We carried out docking of Smurf1 WW2 domain with the PPXY motifs of Smad1, Smad5, and Smad6 and identified the key amino acid residues involved in interaction. Furthermore, we present experimental evidence that WW2 domain of Smurf1 does indeed interact with the Smad proteins and that the deletion of WW2 domain of Smurf1 results in loss of its binding to Smads using the purified recombinant proteins. Finally, we also present data confirming that the deletion of WW2 domain in Smurf1 abolishes its ubiquitination activity on Smad1 in an in vitro ubiquitination assay. It shows that the interaction between the WW domain and Smad PPXY motif is a key step in Smurf1-mediated ubiquitination of its natural targets such as Smad1, Smad5, and Smad6. This work facilitates further strategies to unravel the biological function of such interactions and help in designing effective mimetic compounds that either mimic or disrupt the specific interaction.

LIM mineralization protein-1 potentiates bone morphogenetic protein responsiveness via a novel interaction with Smurf1 resulting in decreased ubiquitination of Smads

Development and repair of the skeletal system and other organs is highly dependent on precise regulation of bone morphogenetic proteins (BMPs), their receptors, and their intracellular signaling proteins known as Smads. The use of BMPs clinically to induce bone formation has been limited in part by the requirement of much higher doses of recombinant proteins in primates than were needed in cell culture or rodents. Therefore, control of cellular responsiveness to BMPs is now a critical area that is poorly understood. We determined that LMP-1, a LIM domain protein capable of inducing de novo bone formation, interacts with Smurf1 (Smad ubiquitin regulatory factor 1) and prevents ubiquitination of Smads. In the region of LMP responsible for bone formation, there is a motif that directly interacts with the Smurf1 WW2 domain and can effectively compete with Smad1 and Smad5 for binding. We have shown that small peptides containing this motif can mimic the ability to block Smurf1 from binding Smads. This novel interaction of LMP-1 with the WW2 domain of Smurf1 to block Smad binding results in increased cellular responsiveness to exogenous BMP and demonstrates a novel regulatory mechanism for the BMP signaling pathway.

Cloning, expression and properties of porcine trachea UDP-galnac: polypeptide N-acetylgalactosaminyl transferase

A UDP-GalNAc:polypeptide N-acetyl-galactosaminyl transferase which catalyses the transfer of GalNAc from UDP-GalNAc to serine and threonine residues in mucin polypeptide chains was purified to homogeneity from swine trachea epithelium (Mendicino J, Sangadala S: Mol Cell Biochem 185: 135-145, 1998). Peptides obtained by proteolysis of the purified enzyme were isolated, sequenced and used to prepare degenerate oligonucleotide primers. Amplified segments of a gene encoding GalNAc transferase were synthesised using the primers and a swine trachea epithelial cDNA library. Selected cDNA fragments were then used to screen the cDNA library, and a clone containing an open reading frame encoding 559 amino acids was isolated. The predicted amino acid sequence contains type II transmembrane region, three potential N-glycosylation sites as well as all of the isolated peptide sequences. The nucleotide sequence and predicted primary protein structure of the transferase were very similar to those of type T-1 GalNAc transferases. The isolated clone was transiently expressed in COS 7 cells and the recombinant enzyme, which contained an N-terminal hexa-histidine tag, was purified to homogeneity and its enzymatic properties were examined. The Vmax of the recombinant enzyme, 2.08 micromol/(min mg), was nearly the same as the native enzyme, 2.12 micromol/(min mg), when assayed with partially deglycosylated mucins as glycosyl acceptors. Both enzymes showed much higher activities when assayed with peptides prepared by limited acid hydrolysis of incompletely deglycosylated Cowper's gland, swine, and human respiratory mucins and tryptic peptides isolated from deglycosylated mucin polypeptide chains. However, as noted earlier (Mendicino J, Sangadala S: Mol Cell Biochem 185: 135-145, 1998), these enzymes showed very little activity with completely deglycosylated mucin polypeptide chains. When completely deglycosylated polypeptide chains were partially glycosylated by incubation with microsome preparations they were again good glycosyl acceptors for the T1-GalNAc transferases isolated from swine trachea. These results show for the first time that multiple isoforms of GalNAc transferases acting in sequence may be required for the complete O-glycosylation of mucin polypeptide chains, and those acting on the nacent polypeptide chain synthesize intermediates which can serve as glycosyl acceptors for other isoforms of the enzyme.

Denaturation of either Manduca sexta aminopeptidase N or Bacillus thuringiensis Cry1A toxins exposes binding epitopes hidden under nondenaturing conditions

The effect of polypeptide denaturation of Bacillus thuringiensis Cry1A toxins or purified Manduca sexta 120-kDa aminopeptidase N on the specificities of their interactions was investigated. Ligand and dot blotting experiments were conducted with (125)I-labeled Cry1Ac, Cry1Ac mutant (509)QNR-AAA(511) (QNR-AAA), or 120-kDa aminopeptidase N as the probe. Mutant QNR-AAA does not bind the N-acetylgalactosamine moiety on the 120-kDa aminopeptidase. Both (125)I-Cry1Ac and (125)I-QNR-AAA bound to 210- and 120-kDa proteins from M. sexta brush border membrane vesicles and purified 120-kDa aminopeptidase N on ligand blots. However, on dot blots (125)I-QNR-AAA bound brush border vesicles but did not bind purified aminopeptidase except when aminopeptidase was denatured. In the reciprocal experiment, (125)I-aminopeptidase bound Cry1Ac but did not bind QNR-AAA. (125)I-aminopeptidase bound Cry1Ab to a limited extent but not the Cry1Ab domain I mutant Y153D or Cry1Ca. However, denatured (125)I-aminopeptidase detected each Cry1A toxin and mutant but not Cry1Ca on dot blots. The same pattern of recognition occurred with native (nondenatured) (125)I-aminopeptidase probe and denatured toxins as the targets. The broader pattern of toxin-binding protein interaction is probably due to peptide sequences being exposed upon denaturation. Putative Cry toxin-binding proteins identified by the ligand blot technique need to be investigated under native conditions early in the process of identifying binding proteins that may serve as functional toxin receptors.

Polydispersity of Bacillus thuringiensis Cry1 toxins in solution and its effect on receptor binding kinetics

Dynamic light scattering and surface plasmon resonance techniques were used to investigate the influence of ionic strength, buffer composition and pH on the multimerization of trypsin-activated Cry1Ac and Cry1C toxins over time and the subsequent effects of the different multimers on receptor binding models. In carbonate buffer at pH 10.5, Cry1Ac and Cry1C assumed a monomeric state. After 24 h, a complete conversion of monomeric toxin to a dimeric or trimeric form was observed only for Cry1Ac under low ionic strength condition. Cry1C and Cry1Ac in high ionic strength buffer remained monomeric. Substitution of CAPS pH 11 for carbonate buffer suppressed this Cry1Ac oligomerization effect. Once Cry1Ac toxin was in an aggregated form, increases in ionic strength failed to revert the aggregated toxin back to a monomeric form. Monomeric Cry1Ac bound to a purified 115 kDa aminopeptidase N receptor from Manduca sexta in a 2:1 molar ratio thus confirming the existence of two binding sites on this receptor. Binding rates of dimeric or higher aggregated Cry1Ac toxin forms were different from those generated using the monomeric form and could not be fitted to existing binding models. In summary, our results confirm that the M. sexta 115 kDa aminopeptidase N receptor possesses two Cry1Ac binding sites. They further suggest that although high pH and low salt conditions promote Cry1Ac aggregation, this observation cannot be applied universally to other members of the Cry family.

Carbohydrate analyses of Manduca sexta aminopeptidase N, co-purifying neutral lipids and their functional interactions with Bacillus thuringiensis Cry1Ac toxin

Bacillus thuringiensis Cry1Ac insecticidal toxin binds specifically to 120kDa aminopeptidase N (APN) (EC 3.4.11.2) in the epithelial brush border membrane of Manduca sexta midguts. The isolated 120-kDa APN is a member of a functional Cry1 toxin receptor complex (FEBS Lett. 412 (1997) 270). The 120-kDa form is glycosyl-phosphatidylinositol (GPI) anchored and converted to a 115-kDa form upon membrane solubilization. The 115-kDa APN also binds Cry1A toxins and Cry1Ac binding is inhibited by N-acetylgalactosamine (GalNAc). Here we determined the monosaccharide composition of APN. APN is 4.2mol% carbohydrate and contains GalNAc, a residue involved in Cry1Ac interaction. APN remained associated with non-covalently bound lipids through anion-exchange column purification. Most associated lipids were separated from APN by hydrophobic interaction chromatography yielding a lipid aggregate. Chemical analyses of the lipid aggregate separated from APN revealed neutral lipids consisting mostly of diacylglycerol and free fatty acids. The fatty acids were long, unsaturated chains ranging from C:14 to C:22. To test the effect of APN-associated lipids on Cry1Ac function, the lipid aggregate and 115-kDa APN were reconstituted into phosphatidylcholine (PC) vesicles. The lipid aggregate increased the amount of Cry1Ac binding, but binding due to the lipid aggregate was not saturable. In contrast the lipid aggregate promoted Cry1Ac-induced release of 86Rb(+) at the lowest Cry1Ac concentration (50nM) tested. The predominant neutral lipid component extracted from the lipid aggregate promoted Cry1Ac-induced 86Rb(+) release from membrane vesicles in the presence of APN.

Binding of Bacillus thuringiensis Cry1Ac toxin to Manduca sexta aminopeptidase-N receptor is not directly related to toxicity

Bacillus thuringiensis Cry1Ac delta-endotoxin specifically binds a 115-kDa aminopeptidase-N purified from Manduca sexta midgut. Cry1Ac domain III mutations were constructed around a putative sugar-binding pocket and binding to purified aminopeptidase-N and brush border membrane vesicles (BBMV) was compared to toxicity. Q509A, R511A, Y513A, and 509-511 (QNR-AAA) eliminated aminopeptidase-N binding and reduced binding to BBMV. However, toxicity decreased no more than two-fold, indicating activity is not directly correlated with aminopeptidase-N binding. Analysis of toxin binding to aminopeptidase-N in M. sexta is therefore insufficient for predicting toxicity. Mutants retained binding, however, to another BBMV site, suggesting alternative receptors may compensate in vivo.

Synthesis of sulfated oligosaccharides by cystic fibrosis trachea epithelial cells

The mucin glycoproteins in tracheal mucus of patients with cystic fibrosis is more highly sulfated than the corresponding secretions from healthy individuals [16]. In order to further characterize these differences in sulfation and possibly also glycosylation patterns, we compared the structures of sulfated mucin oligosaccharides synthesized by continuously cultured human tracheal cells transformed by simian virus 40. The synthesis of highly sulfated oligosaccharide chains in mucins secreted by normal human epithelial and submucosal cell lines were compared with mucins formed by cystic fibrosis tracheal epithelial and submucosal cell lines. The epithelial cell lines from cystic fibrosis trachea showed a higher rate of sulfate uptake and a significantly higher rate of synthesis and sulfation of high molecular weight chains. Mucins synthesized by each cell line in the presence of 35SO4 were isolated and oligosaccharide chains were released by beta-elimination and separated by ion exchange chromatography and gel filtration. The sulfated high molecular weight chains synthesized by the cystic fibrosis cell lines were characterized by methylation analysis and sequential glycosidase digestion before and after desulfation. Carbohydrate analysis yielded Fuc, Gal and GlcNAc in a ratio of 1:2:2.2 and only one galactosaminitol residue for about every 150-200 sugar residues present. The average molecular size of oligosaccharide chains in these fractions was between 30,000-40,000 daltons. These studies show that increased sulfation of oligosaccharides in mucins synthesized by cells from cystic fibrosis trachea is accompanied by a significant increase in the extension of a basic branched structure present in many of the lower molecular weight oligosaccharides.

Domain III of the Bacillus thuringiensis delta-endotoxin Cry1Ac is involved in binding to Manduca sexta brush border membranes and to its purified aminopeptidase N

Three types of binding assays were used to study the binding of Bacillus thuringiensis delta-endotoxin Cry1Ac to brush border membrane vesicle (BBMV) membranes and a purified putative receptor of the target insect Manduca sexta. Using hybrid proteins consisting of Cry1Ac and the related Cry1C protein, it was shown that domain III of Cry1Ac is involved in specificity of binding as observed by all three techniques. In ligand blotting experiments using SDS-PAGE-separated BBMV proteins as well as the purified putative receptor aminopeptidase N (APN), the presence of domain III of Cry1Ac in a hybrid with Cry1C was necessary and sufficient for specific binding to APN. Using the surface plasmon resonance (SPR) technique with immobilized APN, it was shown that the presence of domain III of Cry1Ac in a hybrid is sufficient for binding to one of the two previously identified Cry1Ac binding sites, whereas the second site requires the full Cry1Ac toxin for binding. In addition, the role of domain III in the very specific inhibition of Cry1Ac binding by the amino sugar N-acetylgalactosamine (GalNac) was determined. Both in ligand blotting and in surface plasmon resonance experiments, as well as in binding assays using intact BBMVs, it was shown that the presence of domain III of Cry1Ac in a toxin molecule is sufficient for the inhibition of binding by GalNAc. These and other results strongly suggest that domain III of delta-endotoxins play a role in insect specificity through their involvement in specific binding to insect gut epithelial receptors.

Purification and characterization of UDP-GalNAc:polypeptide N-acetylgalactosaminyl transferase from swine trachea epithelium

UDP-GalNac: polypeptide N-acetylgalactosaminyltransferase from swine trachea epithelium was purified to homogeneity by procedures which included affinity chromatography on Sepharose 4B columns containing bound deglycosylated Cowper's gland mucin. The enzyme, purified 12,000-fold from microsomes with a yield of 40%, showed only a single band on dodecyl sulfate polyacrylamide gel electrophoresis. The homogenous enzyme has an apparent molecular mass of 70,000 Da, as determined by gel electrophoresis or gel filtration. The transferase has a broad pH optimum between 6.7-7.8 with maximal activity at pH 7.2, and required Mn2+ for activity with maximal activity at 5-7.5 mM. Higher concentrations of Mn2+, inhibited the enzyme. The purified transferase was specific for UDPGalNAc and glycosylated both threonine and serine residues in tryptic peptides prepared from deglycosylated Cowper's gland and swine and human trachea mucins. The apparent Km of the transferase for UDPGalNAc was 6.3 microM, and the Km values for deglycosylated Cowper's gland and human and swine trachea mucins were 0.83, 1.12 and 0.94 mg/ml, respectively. The Vmax of the purified enzyme was 2.1 micromol/min/mg with deglycosylated Cowper's gland mucin, as the glycosyl acceptor. However, the activities with peptides prepared from deglycosylated mucins by limited acid hydrolysis were 20-fold greater than the intact glycoprotein under identical conditions. The deglycosylated mucins and larger peptides aggregated with time of storage and precipitated from solution. Aggregation was accompanied by a corresponding loss of enzymatic activity even after dispersion of the aggregate by sonication. The deglycosylated mucins which were prepared by chemical treatment and periodate oxidation still contained about 20% of the N-acetylgalactosamine present in the intact mucin. When this residual amino sugar was removed by periodate oxidation the completely deglycosylated mucins became very poor substrates for the purified transferase. Data obtained in the current study indicate that the accessibility of serine and threonine in the polypeptide chains of mucin glycoproteins significantly influences the rate of glycosylation of these amino acids. The best substrates and affinity ligand for the enzyme were fragments of incompletely deglycosylated mucin polypeptide chains.

The heliothis virescens 170 kDa aminopeptidase functions as "receptor A" by mediating specific Bacillus thuringiensis Cry1A delta-endotoxin binding and pore formation

The relationship between Bacillus thuringiensis Cry1Aa, Cry1Ab and Cry1Ac delta-endotoxin binding and pore formation was investigated using a purified 170 kDa aminopeptidase N (APN) from Heliothis virescens brush border membranes. Aminopeptidases with molecular sizes of 110, 140 and 170 kDa were eluted from a Cry1Ac toxin affinity column using N-acetylgalactosamine. The 140 kDa aminopeptidase has a cross-reacting determinant typical of a cleaved glycosyl-phosphatidylinositol anchor. After mild base treatment to de-acylate the glycosyl-phosphatidylinositol linkage and incubation in phosphatidyl inositol phospholipase C, anti-cross-reacting determinant antibody recognized the 170 kDa protein. Kinetic binding characteristics of Cry1A toxins to purified 170 kDa APN were determined using surface plasmon resonance. Cry1Aa, Cry1Ab and Cry1Ac, but not Cry1C and Cry1E toxins recognized 170 kDa APN. Each Cry1A toxin recognized two binding sites: a high affinity site with KD ranging from 41 to 95 nM and a lower affinity site with KD in the 325 to 623 nM range. N-acetylgalactosamine inhibited Cry1Ac but not Cry1Aa and Cry1Ab binding to 170 kDa APN. When reconstituted into phospholipid vesicles, the 170 kDa APN promoted toxin-induced 86Rb+ release for Cry1A toxins, but not Cry1C toxin. Furthermore Cry1Ac, the Cry protein most toxic to H. virescens larvae, caused 86Rb+ release at lower concentrations, and to a greater extent than Cry1Aa and Cry1Ab toxins. The correlation between toxin-binding specificity and 86Rb+ release strongly suggests that the purified 170 kDa APN is the functional receptor A in the H. virescens midgut epithelial cell brush border membranes.

A mixture of Manduca sexta aminopeptidase and phosphatase enhances Bacillus thuringiensis insecticidal CryIA(c) toxin binding and 86Rb(+)-K+ efflux in vitro

CryIA(c) delta-endotoxin, a member of the CryI family of Bacillus thuringiensis insecticidal proteins, specifically recognizes and binds with high affinity to target proteins in the midgut of susceptible insects. Protein blots of Manduca sexta brush-border membranes probed with 125I-CryIA(c) identify a major binding protein of 120 kDa and a minor binding protein of 65 kDa. Monoclonal antibodies were raised against the 120-kDa toxin binding protein. Using isoelectric focusing and monoclonal antibodies (2B3, 8G1, and 12B8) 120- and 65-kDa brush-border proteins were isolated. Labeled CryIA(c) and monoclonal antibodies probed to blots of the affinity-selected proteins recognized the 120- and 65-kDa proteins. When reconstituted into phospholipid vesicles, antibody-selected proteins increased toxin binding (35%) and enhanced toxin-induced 86Rb+ release up to 1000-fold. The 120-kDa protein was identified as aminopeptidase N (EC 3.4.11.2). A CryIA(c)-sensitive phosphatase was also present in the 120/65-kDa protein mixture. These findings provide the first identification of B. thuringiensis toxin binding proteins, although confirmation is needed in vivo.

Structures of sulfated oligosaccharides in human trachea mucin glycoproteins

The structures of high molecular weight sulfated oligosaccharide chains in mucins purified from the sputum of a patient with cystic fibrosis and blood group H determinant were established. Reduced oligosaccharides released by treatment with alkaline borohydride were separated by ion exchange chromatography on DEAE-Agarose and a fraction containing multisulfated chains was further purified by lectin affinity chromatography to completely remove small amounts of sialylated chains. A major sulfated oligosaccharide fraction containing chains with an average of 160 to 200 sugar residues was isolated by gel filtration on BioGel P-10 columns and individual subfractions were characterized by methylation analysis, periodate oxidation and sequential glycosidase digestion before and after desulfation. Carbohydrate analysis yielded Fuc, Gal and GldNAc in a ratio of 1:2:2.1 and only one galactosaminitol residue for every 160- to 200 sugar residues. The average molecular weight of oligosaccharide chains in these fractions was between 27,000 and 40,000 daltons. Structural analysis showed that these high molecular weight chains contained varying amounts of the repeating unit shown in the following oligosaccharide. Only one in about every 10 repeating units contained sulfate esters. Several shorter chains which contain 2 to 3 sulfate esters were also isolated from this multisulfated oligosaccharide fraction. The structures proposed for these oligosaccharides indicate that they are lower molecular weight chains with the same general structure as those found in the high molecular weight sulfated oligosaccharides. Taken collectively, the results of these studies show that a major sulfated oligosaccharide fraction in respiratory mucin purified from the mucus of patients with cystic fibrosis contains high molecular weight branched chains that consist of a repeating oligosaccharide sequence with sulfate linked to the 6 positions of galactose and possibly GlcNAc residues in the side chains.

Quantitation and structures of oligosaccharide chains in human trachea mucin glycoproteins

Human respiratory mucin glycoproteins from patients with cystic fibrosis were purified and oligosaccharide chains were released by treatment with alkaline borohydride. A neutral oligosaccharide alditol fraction was isolated from mucin obtained from a patient with A blood group determinant by chromatography on DEAE-cellulose and individual oligosaccharide chains were then isolated by gel filtration on BioGel P-6 columns and high performance liquid chromatography with gradient and isocratic solvent systems. The structures of the purified oligosaccharides were determined by methylation analysis, sequential glycosidase digestion and 'H-NMR spectroscopy. The amount of each chain was determined by compositional analysis. A wide array of discrete branched oligosaccharide structures that contain from 3 to 22 sugar residues were found. Many of the oligosaccharides are related and appear to be precursors of larger chains. The predominant branched oligosaccharides which accumulate contain terminal blood group H (Fuc alpha 2Ga1 beta 4) or blood group A (Fuc alpha 2(Ga1NAc alpha 3) (Ga1 beta 4) determinants which stop further branching and chain elongation. The elongation of oligosaccharide chains in respiratory mucins occurs on the beta 3-linked G1cNAc at branch points, whereas the beta 6-linked G1cNAc residue ultimately forms short side chains with a Fuc alpha 2(Ga1NAc alpha 3) Ga1 beta 4 G1cNAc beta 6 structure in individuals with A blood group determinant. The results obtained in the current studies further suggest that even higher molecular weight oligosaccharide chains with analogous branched structures are present in some human respiratory mucin glycoproteins. Increasing numbers of the repeating sequence shown in the oligosaccharide below is present in the higher molecular weight chains. [formula: see text] This data in conjunction with our earlier observations on the extensive branching of these oligosaccharide chains helps to define and explain the enormous range of oligosaccharide structures found in human and swine respiratory mucin glycoproteins. Comparison of the relative concentrations of each oligosaccharide chain suggest that these oligosaccharides represent variations of a common branched core structure which may be terminated by the addition of alpha 2-linked fucose to the beta 3/4 linked galactose residue at each branch point. These chains accumulate and are found in the highest concentrations in these respiratory mucins.

Characterization of mucin glycoprotein-specific translation products from swine and human trachea, pancreas and colon

RNA was isolated from cultured swine trachea epithelial cells and mucus-secreting tumor cell lines from human pancreas, lung and colon by extraction with guanidine isothiocyanate. Poly(A)+mRNA rich fractions were purified by repeated chromatography on oligo (dT)-cellulose columns and they were translated in a cell-free rabbit reticulocyte system. Translation products labelled with 35S-methionine were isolated by immunoprecipitation with specific antibodies to the polypeptide chains of mucin glycoproteins and they were analyzed by SDS-PAGE and fluorography. A single principal polypeptide band of 67 kDa was found in all cases when the immunoprecipitates were washed with buffer containing bovine serum albumin and unlabeled deglycosylated mucin glycoprotein. The intensity of the 67 kDa band decreased when unlabeled deglycosylated mucin glycoprotein was added to the translation mixture before immunoprecipitation. Affinity purified monospecific antibodies elicited against chemically deglycosylated polypeptide chains of purified mucin glycoproteins from human and swine trachea and Cowper's gland were all equally effective in immunoprecipitating the 67 kDa translation product. Monospecific antibodies directed against the glycosylated and unglycosylated regions of the polypeptide chain yielded single bands with a molecular size of 67 kDa in each case. Peptide profiles obtained by digestion of the 67 kDa translation product with S. aureus V-8 protease were identical to those obtained with deglycosylated human and swine trachea mucin glycoproteins. These studies clearly demonstrate that the translation product of swine trachea and human lung, colon and pancreatic mucin glycoprotein gene is a single polypeptide chain of 67 kDa. The relative size and properties of the translation products synthesized with poly (A)+RNA isolated from mucus-secreting cells derived from three different tissues are similar to those of mucin glycoproteins purified directly from mucus secretions of human and swine trachea epithelium.

Immunological characterization of deglycosylated human and swine trachea and Cowper's gland mucin glycoproteins

Antibodies were raised in rabbits against purified swine and human trachea and Cowper's gland mucin glycoproteins and their deglycosylated polypeptide chains. Three monospecific antibody fractions that recognize the carbohydrate, the deglycosylated or unglycosylated regions of the polypeptide chains in these glycoproteins, were isolated by immunoaffinity chromatography. The human and swine trachea mucin glycoproteins showed extensive immunological homology in both their carbohydrate and polypeptide chains. The carbohydrate chains and unglycosylated region of the polypeptide chain in Cowper's gland mucin glycoprotein showed little or no cross-reaction with comparable regions in respiratory mucin glycoproteins. However, the polypeptide chains in the deglycosylated regions of all three mucin glycoproteins showed extensive homology. Five bands with molecular masses ranging from 40 to 46 kDa that differed by a constant molecular mass of approximately 1.5 kDa were detected in hydrolysates of all of the polypeptide chains after treatment with S. aureus V8 protease. Monospecific antibodies to the deglycosylated region of these chains reacted with the peptides, whereas those directed against the unglycosylated region did not. The results suggest that these chains contain tandem repeating sequences of amino acids.

Subunit structure of deglycosylated human and swine trachea and Cowper's gland mucin glycoproteins

The oligosaccharide chains in human and swine trachea and Cowper's gland mucin glycoproteins were completely removed in order to examine the subunit structure and properties of the polypeptide chains of these glycoproteins. The carbohydrate, which constitutes more than 70% of these glycoproteins, was removed by two treatments with trifluoromethanesulfonic acid for 3 h at 3 degrees and periodate oxidation by a modified Smith degradation. All of the sialic acid, fucose, galactose, N-acetylglucosamine and N-acetylgalactosamine present in these glycoproteins was removed by these procedures. The deglycosylated polypeptide chains were purified and characterized. The size of the monomeric forms of all three polypeptide chains were very similar. Data obtained by gel filtration, release of amino acids during hydrolysis with carboxypeptidase B and gel electrophoresis in the presence of 0.1% dodecyl sulfate showed that a major fraction from each of the three mucin glycoproteins had a molecular size of about 67 kDa. All of the deglycosylated chains had a tendency to aggregate. Digestion with carboxypeptidases showed that human and swine trachea mucin glycoproteins had identical carboxyl terminal sequences, -Val-Ala-Phe-Tyr-Leu-Lys-Arg-COOH. Cowper's gland mucin glycoprotein had a similar carboxyl terminal sequence, -Val-Ala-Tyr-Leu-Phe-Arg-Arg-COOH. The yield of amino acids after long periods of hydrolysis with carboxypeptidases showed that at least 85% of the polypeptide chains in each of the deglycosylated preparations have these sequences. These results suggested that the polypeptide chains in these deglycosylated mucin glycoprotein preparations were relatively homogeneous. The deglycosylated polypeptide chains as well as the intact mucin glycoproteins had blocked amino terminii. The purified polypeptide chains were digested with trypsin-TCPK, and S. aureus V8 protease and the resulting peptides were isolated by gel electrophoresis in the presence of 0.1% dodecyl sulfate and by HPLC. Two partial amino acid sequences from swine trachea mucin glycoprotein, two partial sequences from human trachea mucin glycoprotein and three partial sequences from Cowper's gland mucin glycoprotein were determined. The partial amino acid sequences of the peptides isolated from swine trachea mucin glycoprotein showed more than 70% sequence homology to a repeating sequence present in porcine submaxillary mucin glycoprotein. Five to eight immunoprecipitable bands with sizes ranging from about 40 kDa to 46 kDa were seen when the polypeptide chains were digested with S. aureus V8 protease. All of the bands had blocked amino terminii and differed by a constant molecular weight of about 1.5 kDa.(ABSTRACT TRUNCATED AT 400 WORDS)

UDP-GlcNAc: Gal beta 3GalNAc-mucin: (GlcNAc----GalNAc) beta 6-N-acetylglucosaminyltransferase and UDP-GlcNAc: Gal beta 3(GlcNAc beta 6) GalNAc-mucin (GlcNAc----Gal)beta 3-N-acetylglucosaminyltransferase from swine trachea epithelium

Two specific beta-N-acetylglucosaminyltransferases involved in the branching and elongation of mucin oligosaccharide chains, namely, a beta 1,6 N-acetylglucosaminylsaminyltransferase that transfers N-acetylglucosamine from UDP-N-acetylglucosamine to Gal beta 3GalNAc-mucin to yield Gal beta 3(GlcNAc beta 6)GalNAc-Mucin and a beta 3-N-acetylglucosaminyl transferase that transfers N-acetylglucosamine from UDP-N-acetylglucosamine to Gal beta 3(GlcNAc beta 6)GalNAc-mucin to yield GlcNAc beta 3Gal beta 3 (GlcNAc beta 6)GalNAc-Mucin were purified from the microsomal fraction of swine trachea epithelium. The beta 1,6-N-acetylglucosaminyltransferase was purified about 21,800-fold by procedures which included affinity chromatography on DEAE columns containing bound asialo Cowper's gland mucin glycoprotein with Gal beta 1,3GalNAc side chains. The apparent molecular weight estimated by gel filtration was found to be about 60 Kd. The purified enzyme showed a high specificity for Gal beta 1,3GalNAc chains and the most active substrates were mucin glycoproteins containing these chains. The apparent Km of the beta 6-glucosaminyltrans-ferase for Cowper's gland mucin glycoprotein containing Gal beta 1,3GalNAc chains was 0.53 microM; for UDP-N-acetylglucosamine, 12 microM; and for Gal beta 1,3GalNAc alpha NO2 phi, 4 mM. The activity of the beta 6-glucosaminyltransferase was dependent on the extent of glycosylation of the Gal beta 3GalNAc chains in Cowper's gland mucin glycoprotein. The best substrate for the partially purified beta 3-Glucosaminyltransferase was Cowper's gland mucin glycoprotein containing Gal beta 1,3(GlcNAc beta 6)GalNAc side chains. This enzyme showed little or no activity with intact sialylated Cowper's gland mucin glycoprotein or derivatives of this glycoprotein containing GalNAc or Gal beta 1,3GalNAc side chains. The radioactive oligosaccharides formed by these enzymes in large scale reaction mixtures were released from the mucin glycoproteins by treatment with alkaline borohydride, isolated by gel filtration on Bio-Gel P-6 and characterized by methylation analysis and sequential digestion with exoglycosidases. The oligosaccharide products formed by the beta 6- and beta 3-glucosaminyltransferases were shown to be Gal beta 3(GlcNAc beta 6) GalNAc and GlcNAc beta 3 Gal beta 3(GlcNAc beta 6)GalNAc respectively. Taken collectively, these results demonstrate that swine trachea epithelium contains two specific N-acetylglucosaminyltransferases which catalyze the initial branching and elongation reactions involved in the synthesis of O-linked oligosaccharide chains in respiratory mucin glycoproteins. The first enzyme a beta 6-glucosaminyltransferase converts Gal beta 3GalNAc chains in mucin glycoproteins to Gal beta 3(GlcNAc beta 6)GalNAc chains.(ABSTRACT TRUNCATED AT 400 WORDS)