NANOG modulates stemness in human colorectal cancer.

NANOG is a stem cell transcription factor that is essential for embryonic development, reprogramming normal adult cells and malignant transformation and progression. The nearly identical retrogene NANOGP8 is expressed in multiple cancers, but generally not in normal tissues and its function is not well defined. Our postulate is that NANOGP8 directly modulates the stemness of individual human colorectal carcinoma (CRC) cells. Stemness was measured in vitro as the spherogenicity of single CRC cells in serum-free medium and the size of the side population (SP) and in vivo as tumorigenicity and experimental metastatic potential in NOD/SCID mice. We found that 80% of clinical liver metastases express a NANOG with 75% of the positive metastases containing NANOGP8 transcripts. In all, 3-62% of single cells within six CRC lines form spheroids in serum-free medium in suspension. NANOGP8 is translated into protein. The relative expression of a NANOG gene increased 8- to 122-fold during spheroid formation, more than the increase in OCT4 or SOX2 transcripts with NANOGP8 the more prevalent family member. Short hairpin RNA (shRNA) to NANOG not only inhibits spherogenicity but also reduces expression of OCT4 and SOX2, the size of the SP and tumor growth in vivo. Inhibition of NANOG gene expression is associated with inhibition of proliferation and decreased phosphorylation of G2-related cell-cycle proteins. Overexpression of NANOGP8 rescues single-cell spherogenicity when NANOG gene expression is inhibited and increases the SP in CRC. Thus, NANOGP8 can substitute for NANOG in directly promoting stemness in CRC.

Mechanistic studies of the effects of anti-factor H antibodies on complement-mediated lysis.

We have recently reported that complement factor H, a negative regulator of complement-mediated cytotoxicity, is produced and secreted by most bladder cancers. This observation was exploited in the development of the BTA stat and BTA TRAK diagnostic assays, both of which make use of two factor H-specific monoclonal antibodies in sandwich format. Here we show that both antibodies exert interesting effects on the biochemistry of complement activation in in vitro systems. Antibody X13.2 competes with C3b for association with factor H and strongly inhibits factor H/factor I-mediated cleavage of C3b, thereby evidently inactivating a negative regulator of complement; yet, the antibody strongly inhibits complement-mediated lysis as well. Conversely, antibody X52. 1, which does not compete with C3b and has no effect on solution-phase cleavage of C3b, is capable of enhancing complement-mediated lysis of various cell types, including cancer cells, by over 10-fold. Our observations indicate that it is possible to deconvolute the biochemical roles of factor H in complement by means of appropriate inhibitors, a finding with potentially valuable implications for both basic research and cancer therapy.

Complement factor H or a related protein is a marker for transitional cell cancer of the bladder.

The BTAstat and BTA TRAK tests are new immunoassays that detect and measure an antigen in the urine of individuals diagnosed with bladder cancer. As described in this report, the monoclonal antibodies used in these kits were developed by immunizing mice with partially purified protein preparations derived from the urine of patients with bladder cancer. The antigen that is recognized by the monoclonal antibodies was purified from the urine of bladder cancer patients by immunoaffinity chromatography and identified as being either complement factor H (FH) or a closely related protein (CFHrp) by partial amino acid sequence analysis. Like serum FH, the urine antigen was demonstrated to have a complement factor C3b binding site and to accelerate the degradation of C3b in the presence of complement factor I. The culture supernatants from several human bladder, cervical, and renal cancer cell lines contained antigen as determined by immunoassay, and antigen affinity-purified from HeLaS3 culture media was shown to have FH activity. Moreover, the cell lines were shown to make products of the expected sizes by reverse transcription-PCR using FH-specific primers. In contrast, normal human epithelial keratinocytes, a myeloid leukemia cell line, and the colon cancer line LS174T were negative for production of a FH-like protein (CFHrp). We propose that the expression of proteins with FH-like activities may confer a selective growth advantage to cancer cells in vivo by decreasing complement activity, thus aiding their escape from lysis by immune surveillance. Identification of these proteins as cancer products also suggests avenues of chemotherapy or immunotherapy of some cancers.

A very sensitive coupled luminescent assay for cytotoxicity and complement-mediated lysis.

The demand for convenient and sensitive means of measuring cytotoxicity and complement-mediated killing is likely to be increased by the recent identification of Complement Factor H, an important regulatory protein of both the classical and alternate pathways of complement, as a tumor-associated antigen. Here we describe a simple luminometric assay capable of detecting the death of approximately 0.03 nucleated human-cell equivalent or approximately 1 rabbit-erythrocyte equivalent. The assay measures the release of glyceraldehyde-3-phosphate dehydrogenase (G3PDH) from dead or damaged cells by coupling its enzymatic activity to production of ATP, which in turn is measured by well-known methods involving firefly luciferase. This is accomplished by means of a reaction series in which the activity of G3PDH is coupled with that of phosphoglycerate kinase, the next enzyme in the glycolytic pathway. As described, the assay uses inexpensive, commercially available reagents. This coupled assay was used to demonstrate that an anti-factor-H antibody is capable of enhancing complement-mediated killing of the Raji cancer cell line by > 1000%.

Modified method for determining carcinoembryonic antigen in the presence of human anti-murine antibodies.

The increasing use of monoclonal antibodies (MAbs) for disease diagnosis and therapy has created a class of patients at risk for systematic error in clinical testing due to interference by human anti-murine antibodies (HAMA). HAMA interference is often difficult to detect and can cause either an increase or a decrease in apparent concentrations of antigen present. We undertook a clinical study to test a HAMA-resistant enzyme immunoassay (EIA) format for carcinoembryonic antigen (CEA) determination. Using the Food and Drug Administration-approved CEA-EIA Monoclonal One-Step Assay (Abbott) with the addition of an acid/heat extraction of patients' specimens, we found that the resulting CEA values accurately reflected the patients' status. We demonstrated that the acid/heat-extracted specimens yield linear dilution curves and show analytical recoveries of added CEA in the range of 76-123% in HAMA-positive specimens and 86-103% in HAMA-negative specimens. The correlation of CEA values in extracted vs unextracted specimens from 184 patients and control subjects was 0.9963. The CEA detection limit of the assay was 1.6 micrograms/L for the extracted samples.

Human antibody response to the intravenous and intraperitoneal administration of the F(ab')2 fragment of the OC125 murine monoclonal antibody.

We have characterized the human immune response against murine monoclonal antibodies (HAMA) in 18 patients following administration of the F(ab')2 fragment of the murine monoclonal antibody OC125. OC125 is directed against the CA125 antigen, present on the surface of many human ovarian cancers. An affinity matrix was used to separate serum into immunoglobulin-containing and immunoglobulin-free fractions. HAMA titer was determined on the immunoglobulin fraction with an OC125 sandwich enzyme-linked immunosorbent assay (ELISA). All patients developed an HAMA response, despite the use of F(ab')2 fragments and small amounts (1-4 mg) of antibody. It may be that the intraperitoneal (i.p.) route provides a more marked HAMA response. Enzyme-linked sandwich immunoassays were also used to determine anti-isotype and anti-idiotype titers. Anti-isotype titers were analyzed with antigen irrelevant, isotype-matched murine antibodies and OC125-HRPO. Anti-idiotypes titers were assessed in a sandwich assay that utilized F(ab')2 and F(ab') fragments of OC125. The anti-isotype response tended to be of low titer and short duration, while the anti-idiotype response was of high titer and remarkably persistent. HAMA interfered in an unpredictable manner with the correct measurement of serum levels of CA125 in an enzyme immunoassay using OC125. Corrected values of CA125 could be obtained by measurement of antigen in the immunoglobulin-free fraction of serum. The response of one patient, who developed a markedly elevated anti-idiotype titer after serial i.v./i.p. injections, was further characterized and found to contain an antibody consistent with an anti-anti-idiotype to the CA125 antigen.

Preparation of synthetic polypeptide domains of carcinoembryonic antigen and their use in epitope mapping.

Genes encoding the four principal polypeptide domains (N, A1-B1, A2-B2, and A3-B3) of carcinoembryonic antigen (CEA) were synthesized and expressed in Escherichia coli as fusion products with bacterial CMP-KDO synthetase (CKS). The four synthetic fusion proteins were purified in high yield and used as targets in Western blots for 11 anti-CEA MAbs and to compete with immobilized CEA for binding to four of these MAbs. Each of the MAbs showed strong binding to one or more of the fusion proteins. In Western blots, MAbs H19C91 and 4230 bound only to CKS-N. MAbs H8C2 and H11C35 bound only CKS-A1-B1, and MAbs T84.66, H46C136, and H21C83 appeared to be specific for CKS-A3-B3. None of the MAbs tested bound only to CKS-A2-B2. However, two MAbs bound both CKS-A1-B1 and CKS-A3-B3 and one MAb (3519) bound to all three of the repeated domains. Since these three domains exhibit over 90% amino acid sequence homology, the latter results were not surprising. The competition studies largely confirmed the results of Western blots but did show some MAb-fusion protein interactions not observed in Western blots. These competition studies also allowed estimation of the relative affinities of the MAbs for the synthetic domains and for native CEA. These studies demonstrated that epitopes in CEA recognized by the MAbs in this study are peptide in nature and that the fusion proteins are of utility in the localization of the epitopes on the polypeptide chain of CEA.

Mucin glycoproteins as tumor markers.

Extensive biochemical studies have shown that mucin tumor antigens have a range of molecular sizes from 200 to greater than 1000 kDa. The molecular size of mucin antigens can be dramatically affected by the source and method of purification. Mucin antigens vary from 24 to 80% in carbohydrate content and their density is usually greater than 1.40 g/ml. Galactose and N-acetyl glucosamine are the predominant sugar residues in many mucins, whereas mannose is usually present in low levels or absent. The amino acids serine, threonine, alanine, glycine, and proline are abundant in mucins. An O-glycosidic linkage between the carbohydrate and protein of mucins is the most common linkage encountered. The gene encoding the core peptide for at least one mucin tumor marker, HMFG, has been identified, sequenced, and expressed. These findings may lead to a better understanding of the multiepitope nature of mucin tumor markers. The advent of hybridoma technology has yielded several monoclonal antibodies that have been used to identify the presence of tumor-associated mucins in the sera of cancer patients. Elevated levels of mucin antigens have been found in the serum of most patients with advanced adenocarcinomas. Many studies have shown that tumor-associated markers are useful in monitoring patients following cancer treatment. Clinically useful immunoassays have been developed for monitoring patients with ovarian, breast, and pancreatic adenocarcinomas. Although individual mucin tumor markers show limited utility in detecting early adenocarcinoma, recent studies using multiple mucin markers have suggested that early detection, at sensitivities greater than 50%, can be achieved.

Purification of T47D human progesterone receptor and immunochemical characterization with monoclonal antibodies.

In order to obtain steroid-independent probes for human progesterone receptor (PR), the A [88-93 kilodalton (kDa)] and B (109-119 kDa) forms of PR from T47D human breast cancer cells were partially purified and used to generate a series of 14 monoclonal antibodies. Initially, unoccupied PR was isolated from cytosol extracts by steroid affinity chromatography, followed by chromatography on diethylaminoethyl Bio-Gel. The partially pure (3-15%) PR consisted of two steroid-binding components that migrated at 89 kDa and 109 kDa in reducing sodium dodecyl sulfate gels after being photoaffinity labeled with the synthetic progestin [3H]R5020. Two unique monoclonal antibodies to PR were derived from a male Lewis rat immunized with this material. One of these antibodies (JU601) was coupled to Sepharose 4B and used to purify T47D nuclear PR for additional immunizations. Highly purified (30-70%) PR migrated as 93 kDa and 119 kDa progestin-binding proteins in sodium dodecyl sulfate gels. In all, thirteen monoclonal antibodies were obtained that recognized epitopes shared by both receptor forms. One mouse immunoglobulin G (KC146) was completely specific for the larger B form. Interestingly, the epitope for this antibody was present on all PRs tested, including the B form of PR from chicken oviduct, whereas nine other antibodies recognized only human PR and the remaining four cross reacted with rabbit PR. With the exception of the JU145 and JU601 rat immunoglobulin Ms, all antibodies appeared to be completely specific for the A or B forms of PR. Each recognized the cytosol and nuclear forms of occupied as well as unoccupied PR. Although the relationship between B and A was not established, it is clear that an amino-terminal region of B is not present in A, and that a significant portion of A and B are either identical or very similar in amino acid sequence.

Purification of a cell growth inhibitor from bovine cerebral cortex cells.

A glycopeptide, isolated from bovine cerebral cortex cells and added in only nanogram levels to cells in culture, has been shown to inhibit both cell protein synthesis and cell division. When purified by gel filtration and Ulex europaeus lectin affinity chromatography, the radioiodinated preparation was subjected to high resolution isoelectric focusing and shown to contain three species of macromolecules. The glycopeptide focusing at pH 8.1 comprised over 75% of the radioiodinated material and possessed inhibitory activity against both cell protein synthesis and cell division. A second species that focused at pH 8.3 was also found to be inhibitory to cell metabolism and may have represented a variant of the major glycopeptide.

A monoclonal antibody to a unique cell surface growth regulatory glycopeptide.

A glycopeptide, isolated from bovine cerebral cortex cells and added to cells in only nanogram/ml levels, has been shown to inhibit both cell protein synthesis and cell division. A monoclonal antibody was used to show that the inhibitory component originated from the cell surface. Incubation of the M1 IgG monoclonal antibody with partially purified bovine glycopeptide preparations and Staphyloccocus protein A removed the inhibitor from solution. Intact mouse cerebral cortex cells were found to have a similar epitope on their surfaces. In contrast, normal rat kidney cells (NRK) did not react with the monoclonal antibody. An analysis of mouse cerebral cortex membrane preparations, incubated with the monoclonal antibody, confirmed that the primary source of the antigenic determinant was the plasma membrane.

G2 cell cycle arrest induced by glycopeptides isolated from the bovine cerebral cortex.

The ability of glycopeptides, isolated from bovine cerebral cortex, to alter cell division was studied by cell-cycle analyses. The results showed that glycopeptides arrested baby hamster kidney (BHK)-21 cells and Chinese hamster ovary (CHO) cells in the G2 phase of the cell cycle. Upon removal of the growth inhibition from arrested BHK-21 cells, the mitotic index in colchicine-treated cultures increased from 5 to 40% within 6 h and the increase in mitotic activity was accompanied by a complete doubling of all arrested cells within this 6-h time period. Determination of DNA content in growth-arrested BHK-21 cells showed that growth-arrested cells contained about twice the DNA of control cell cultures. Although CHO cells treated in a like manner with growth inhibitor could not be arrested for the same length of time as BHK-21 cells (18 h vs. 72 h before initiation of escape) and to the same degree (60% of the cell population vs. 99% of BHK-21 cells), the escape kinetics of CHO cells did indicate a G2 arrest. Approximately 3.5 h after escape began, CHO cell numbers in treated cultures attained the cell numbers found in control cultures. This rapid growth phase occurring in less than 4 h indicated that the growth inhibitor induced a G2 arrest-point in CHO cells that was not lethal since the entire arrested cell population divided.

Selective protection of nonmalignant cells by a novel cell surface glycopeptide.

A novel glycopeptide inhibitor of cell division, isolated from bovine cerebral cortex cell surfaces, was shown to selectively protect nonmalignant cells from the cytoxic action of 5-bromo-2-deoxyuridine (5-BrdUrd). When mouse LM-22 cells (nonmalignant and devoid of gangliosides) were preincubated with GM1 ganglioside (3.0 micrograms/ml), the cell surface glycopeptide inhibitor effectively arrested cell division. In contrast to LM-22 cells, transformed mouse fibrosarcoma (No. 1316) cells were insensitive to the glycopeptide inhibitor whether or not they were preincubated with GM1 ganglioside. Mixed cultures of LM-22 cells preincubated with GM1 ganglioside and 1316 fibrosarcoma cells at an approximate ratio of 1:1 were established. Since LM-22 cells are resistant and 1316 fibrosarcoma cells are sensitive to 3.0 mM ouabain, the identity of surviving cells following BrdUrd treatment could easily be determined. Three hr after the establishment of the mixed cell population, 250 ng protein per ml of the purified bovine glycopeptide inhibitor was added to selectively arrest the mitosis of the LM-22 cells. After an additional 3 hr of incubation, 5-BrdUrd was added to a final concentration of 5.0 mM. Twelve hr later, cells were serially diluted and seeded into duplicate plates with and without 3.0 mM ouabain. LM-22 cells were effectively protected from the cytotoxic action of 5-BrdUrd (92 to 94% survival) while the majority of the 1316 fibrosarcoma cells were killed (21 to 30% survival). The selective protection of LM-22 cells was shown to be independent of differences in plating efficiency, cytotoxicity of 5-BrdUrd in the absence of the glycopeptide inhibitor, and the generation time of the two cell lines.

Isolation of cell-surface glycopeptides from bovine cerebral cortex that inhibit cell growth and protein synthesis in normal but not in transformed cells.

Glycopeptides were isolated from the cell surfaces of bovine cerebral cortex that could inhibit increase in cell numbers in tissue culture and cellular protein synthesis. This cell growth inhibition apparently affected all cells exposed, could completely block cell division in a reversible manner and synchronized BHK-21 cell cultures. Polyoma-virus-transformed BHK-21 cells were completely insensitive to the inhibitor. Fractionation of the inhibitor on a Bio-Gel P-100 column revealed two peaks of biologically active material eluting at apparent molecular weights of 45 000 and 10 000 with A 1cm,280 1% 11.0. Affinity purification of the inhibitory fractions on a Ulex europaeus agglutination I lectin column resulted in retention of the inhibitory activity, suggesting the inhibitor material was a glycopeptide. Subsequent elution with 0.10 M-fucose resulted in a 244-fold increase in the specific biological activity over the starting material. Although purified from bovine brain, the material could inhibit baby-hamster kidney cell protein synthesis by 50% at a concentration of 5 x 10(6) molecules per target cell. Analysis by competitive radioimmunoassay or immunoadsorption indicated that the bovine inhibitor was structurally related to, although not necessarily identical with, a similar inhibitory glycopeptide preparation that we had previously isolated from mouse brain.

Isolation of the glycoprotein of vesicular stomatitis virus and its binding to cell surfaces.

The glycoprotein (G) of vesicular stomatitis virus (VSV) was radiolabelled, extracted and purified so that its potential interaction with host cell surfaces could be studied. When BHK-21 cells were incubated with the radiolabelled virus glycoprotein, the virus component rapidly attached to the cell surface. The attachment was shown to be temperature-dependent adn saturated at approx. 3 X 10(5) molecules/cell. The omission of Mg2+ or Ca2+ from the incubation medium had little effect on the glycoprotein binding. Treating the isolated G protein and intact virions with neuraminidase did not significantly decrease their binding to BHK-21 cells. Pre-incubating cells with trypsin did not decrease the attachment of VSV virions nor the binding of purified G protein. Treating cells with phospholipase A or phospholipase C suggested that the binding of the glycoprotein and the intact virion might have been dissimilar. Unlabelled glycoprotein competitively inhibited binding of the labelled molecules although the presence of intact virions did not inhibit attachment of the G protein. Likewise, saturating amounts of the glycoprotein did not decrease binding of VSV to BHK-21 cells. These observations suggested that either the isolated glycoprotein bound to cell surface components that were distinct from the virion receptor or that the manner of the purified glycoprotein attachment differed from the G protein still associated with the intact virion. Chemical crosslinking and diagonal two-dimensional gel electrophoresis were used to identify and to compare the cell surface components responsible for glycoprotein and virion attachment.

Characterization of cell-surface glycopeptides from mouse cerebral cortex that inhibit cell growth and protein synthesis.

Glycopeptides were isolated from the cell surfaces of mouse brain cortical tissue that inhibited both cell division and protein synthesis by cells in culture. The protein-synthesis inhibition appeared to affect most cells exposed and was equally effective against glycoprotein and protein synthesis. The inhibition of protein metabolism was independent of mRNA synthesis and uptake of labelled precursors into intracellular pools, indicating that it was directed at intracellular translational events. Fractionation of chloroform/methanol-extracted preparations of this brain cell-surface substance on Bio-Gel P-100 revealed the material to be quite heterogenous, although inhibitory activity was found only in fractions of mol.wt. 25000--30000 and 6000--10000. Biochemical analysis of these fractions demonstrated that they were 6% carbohydrate and 94% amino acid by weight. The 25000--30000-mol.wt. glycopeptides were shown to inhibit cell growth at concentrations of 2 microgram/ml in cultured cells and to inhibit protein synthesis by 50% at concentrations of 3 microgram/ml. The 25000--30000-mol.wt. brain-cell-surface-substance glycopeptides were further purified by ultrafiltration and affinity chromatography with Ulex europaeus agglutinin, resulting in a 400-fold increase in specific biological activity. The inhibitor was not lethal to cells and was not species- or tissue-specific.