A comparison of prostate cancer cell transcriptomes in 2D monoculture vs 3D xenografts identify consistent gene expression alterations associated with tumor microenvironments

BACKGROUND

Prostate cancer (PC) research has relied heavily on patient-derived cell lines, which may be used for in vitro (two-dimensional [2D]) studies or cultivated as three-dimensional (3D) xenografts in mice. These approaches are likely to have differential impacts on cell phenotypes, with implications for experimental outcomes. Therefore, defining and comparing the transcriptional signatures associated with 2D and 3D approaches may be useful for designing experiments and interpreting research results.

METHODS

In this study, LNCaP, VCaP, and 22Rv1 human PC cells were either cultivated in monolayers or as xenografts in NOD SCID mice, and their gene transcription profiles were quantitated and compared using microarray and real-time polymerase chain reaction techniques. Immunohistochemistry was used to evaluate protein expression in cancer cell xenografts.

RESULTS

Comparisons of gene expression profiles of tumor cells grown in 2D vs 3D environments identified gene sets featuring similar expression patterns in all three cancer cell lines and unique transcriptional signatures associated with 3D vs 2D growth. Pathways related to cell-cell interactions, differentiation, and the extracellular matrix were enriched in 3D conditions. Immunohistochemical analyses confirmed that gene upregulation in xenografts occurred in implanted cancer cells and not in mouse stromal cells. Cultivating cells in vitro in the presence of mouse, rather than bovine serum failed to elicit the gene transcription profile observed in xenografts, further supporting the hypothesis that this profile reflects 3D growth and enhanced microenvironmental interactions, rather than exposure to species-specific serum factors.

CONCLUSIONS

Overall, these findings define the expression profiles observed in PC cells cultivated in 2D monolayers and in 3D xenografts, highlighting differentially regulated pathways in each setting and providing information for interpreting research results in model systems.

Abstract IA21: AR is ablated: Now what? Targeting the double-negative phenotype

Background/Introduction: Androgen receptor (AR) signaling is a distinctive feature of prostate cancer (PC) and represents the major therapeutic target for treating metastatic prostate cancer (mPC). Though highly effective, AR antagonism can produce tumors that bypass a functional requirement for AR, often through neuroendocrine (NE) transdifferentiation. Through the molecular assessment of mPCs over two decades, we find that a phenotypic shift has occurred in mPC with the emergence of an AR-null NE-null phenotype. We sought to identify drivers of these “double-negative” PCs (DNPC), with the ultimate objective of developing therapeutic strategies.

Methods: We used genome-wide profiling of copy number (array CGH), mutations (whole-exome sequencing-WES), and gene expression (RNA-seq) to compare the molecular landscapes of AR-active PC (ARPC), NEPC, and DNPC in tumors obtained from 84 consecutive men with mPC undergoing a rapid autopsy. We developed model systems including DNPC patient-derived xenografts (PDX) and cell lines that are devoid of AR activity and lack NE features. Using these models, we tested therapeutics hypothesized to inhibit the activity of pathways preferentially active in DNPC.

Results: In the era prior to the approval of the AR pathway antagonists enzalutamide and abiraterone, most CRPCs were ARPCs (85%) with rare NEPCs (10%) and rarer DNPCs (5%). In the contemporary era (2012-2016), we observed a shift in tumor phenotypes with a higher representation of DNPCs. Gene expression programs of the tumors classified by IHC supported these distinct subtypes. DNPCs did not exhibit GR activity or PI3K/AKT activation but were notable for high MAPK and FGF pathway activity. Models systems recapitulated these findings. FGFR antagonists differentially repressed the growth of DNPC tumors in vitro and in vivo.

Conclusions: Though the majority of mPCs that resist AR targeting retain AR signaling, an increasingly common subtype of mPC exhibits a DNPC phenotype. Our results indicate that at least a subset of these DNPCs are driven by FGF signaling and MAPK activation. Targeting the FGF axis may represent a therapeutic approach for those cancers resistant to AR-directed therapies and may circumvent treatment resistance if combined with initial AR pathway blockade.

Citation Format: Colm Morrissey, Eric G. Bluemn, Ilsa M. Coleman, Jared M. Lucas, Roger T. Coleman, Robin Tharakan, Daniella Bianchi-Frias, Ruth F. Dumpit, Arja Kaipainen, Alexandra N. Corella, Yu Chi Yang, Michael D. Nyquist, Elahe Mostaghel, Xiaotun Zhang, Eva Corey, Lisha G. Brown, Holly M. Nguyen, Michael Schweizer, Lawrence D. True, Paul S. Rennie, Robert L. Vessella, Peter S. Nelson. AR is ablated: Now what? Targeting the double-negative phenotype [abstract]. In: Proceedings of the AACR Special Conference: Prostate Cancer: Advances in Basic, Translational, and Clinical Research; 2017 Dec 2-5; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(16 Suppl):Abstract nr IA21.

Androgen Receptor Pathway-Independent Prostate Cancer Is Sustained through FGF Signaling

Androgen receptor (AR) signaling is a distinctive feature of prostate carcinoma (PC) and represents the major therapeutic target for treating metastatic prostate cancer (mPC). Though highly effective, AR antagonism can produce tumors that bypass a functional requirement for AR, often through neuroendocrine (NE) transdifferentiation. Through the molecular assessment of mPCs over two decades, we find a phenotypic shift has occurred in mPC with the emergence of an AR-null NE-null phenotype. These "double-negative" PCs are notable for elevated FGF and MAPK pathway activity, which can bypass AR dependence. Pharmacological inhibitors of MAPK or FGFR repressed the growth of double-negative PCs in vitro and in vivo. Our results indicate that FGF/MAPK blockade may be particularly efficacious against mPCs with an AR-null phenotype.

Complexity of inflammatory responses in endothelial cells and vascular smooth muscle cells determined by microarray analysis

To better understand the molecular basis of vascular cell system behavior in inflammation, we used gene expression microarrays to analyze the expression of 7,075 genes and their response to IL-1beta and TNFalpha in cultures of coronary artery endothelium and smooth muscle derived from a single coronary artery. The most noticeable difference between the cell types was the considerably greater magnitude and complexity of the transcriptional response in the endothelial cells. Two hundred and nine genes were regulated in the endothelium and only 39 in vascular smooth muscle. Among the 209 regulated genes in the endothelium, 99 have not been previously associated with endothelial cell activation and many implicate the endothelium in unconventional roles. For example, the induced genes include several that have only been associated with leukocyte function (e.g., IL-7 receptor, EBI-3 receptor) and others related to antiviral and antibacterial defense (e.g., oligoadenylate synthetase, LMP7, toll-like receptor 4, complement component 3). In addition, 43 genes likely to participate in signal transduction (eg. IL-18 receptor, STK2 kinase, STAF50, ANP receptor, VIP receptor, RAC3, IFP35) were regulated providing evidence that a major effect of TNFalpha and IL-1beta is to alter the potential of the endothelial cell to respond to various other external stimuli.

Comparison between cefprozil and penicillin to eradicate pharyngeal colonization of group A beta-hemolytic streptococci

BACKGROUND

Our objective was to perform a prospective, randomized, double blinded study of cefprozil and penicillin therapy to eradicate group A beta-hemolytic streptococci (GABHS) in children who were bacteriologic failures after receiving a standard 10-day course of penicillin treatment for GABHS pharyngitis.

METHODS

Children and adolescents 2 to 18 years of age were eligible for the study. From 3 to 7 days after completing oral penicillin therapy for pharyngitis caused by GABHS, the study was explained, informed consent was obtained, a history and physical examination were completed and a throat culture was performed. Children with throat cultures positive for GABHS were randomized to receive either cefprozil or penicillin for 10 days. Children who were bacteriologic failures after administration of the first study drug were crossed over to receive the alternate drug.

RESULTS

Of 180 enrolled children 66 (37%) had throat cultures positive for GABHS. Seventeen were excluded from the study, leaving 49 who completed the protocol. Of the 49 participants 26 received cefprozil initially whereas 23 received penicillin. GABHS were eradicated from the pharynx of 73% of children who received cefprozil as the first antibiotic compared with 39% of penicillin recipients (chi square, 5.748, 0.01 < P < 0.025). After crossover of failures, the final efficacy rate for cefprozil was 65% compared with 36.7% for penicillin (chi square, 5.523, 0.01 < P < 0.025).

CONCLUSIONS

Cefprozil was more effective than penicillin in treating children who were bacteriologic failures after a standard 10-day course of oral penicillin.

Cloning and characterization of human tissue inhibitor of metalloproteinases-3

The tissue inhibitors of metalloproteinases (TIMPs) comprise a family of proteins, of which two members have so far been described in humans. We have cloned and sequenced a third human TIMP (hTIMP-3) from phorbol ester-differentiated THP-1 cells stimulated with bacterial lipopolysaccharide. The open reading frame encodes a 211-amino-acid precursor including a 23-residue secretion signal. The mature polypeptide has a calculated molecular weight of 21.6 kD and includes an N-linked glycosylation site near the carboxyl terminus. The protein is quite basic, having a predicted isoelectric point of 9.04. We have mapped the single gene encoding human TIMP-3 to chromosome 22. By Northern analysis, transcripts for TIMP-3 were identified in a broad cross-section of tissues examined from both embryonic and adult origin. In all tissues except the placenta, the predominant transcript was 5.0 kb in size, with minor bands around 2.4 and 2.6 kb comprising no more than about 10% of the signal. In the placenta, the smaller bands accounted for close to 50% of the signal. Human TIMP-3 shows slightly closer amino acid sequence similarity to TIMP-2 (44.3%) than to TIMP-1 (38.4%), but is most closely related to a recently reported chicken TIMP, chIMP-3 (80.8% amino acid; 77.7% nucleic acid similarity.

Prevention of neonatal group B streptococcal infections: advances in maternal vaccine development

OBJECTIVE

We describe the current state of prevention of neonatal group B streptococcal infections and present recent advances toward the development of a maternal vaccine for prevention of this disease.

DATA SOURCES

We used a MEDLINE search of the Index Medicus from 1976-1992 for articles regarding group B streptococcus classification and immunology. Group B streptococcus was also cross-referenced with bacterial antigens, antibodies, and vaccines. Relevant textbooks were reviewed.

METHODS OF STUDY SELECTION

Fifty-seven articles were selected as providing important background and new findings pertinent to this topic.

DATA EXTRACTION AND SYNTHESIS

The literature supports prophylactic use of intrapartum antibiotics in mothers who are known carriers of group B streptococcus but highlights the need for more sensitive rapid screening techniques to identify this high-risk population. The promise of intravenous immunoglobulin for neonatal prophylaxis has not been borne out, although hyperimmune and monoclonal preparations offer renewed hope for prophylaxis and adjuvant therapy. Native bacterial polysaccharides, conjugated oligosaccharides and polysaccharides, and C proteins have been investigated as antigens for candidate vaccines. Antibodies elicited in human and animal studies provide protection against bacterial strains possessing these determinants. The theoretical existence of a "universal antigen" is significant because polysaccharide and C protein formulations are required to be polyvalent.

CONCLUSIONS

The development of a vaccine for prevention of neonatal group B streptococcal sepsis is an attainable goal. Further study of the immunogenic properties of bacterial-cell-wall polysaccharides and their conjugates, C proteins, and the potential universal antigen is required.

Expression of CA III in rodent models of obesity

To achieve a better understanding of the biochemical basis of obesity, we have undertaken comparative analyses of adipose tissue of lean and obese mice. By two-dimensional gel analysis, carbonic anhydrase-III (CA III) has been identified as a major constituent of murine adipose tissue. Quantitative comparisons of CA III protein and mRNA levels indicate that this enzyme is expressed at lower levels in adipose tissue from animals that were either genetically obese or had experimentally induced obesity compared to levels in the corresponding lean controls. This decrease in CA III expression was unique to adipose tissue, since other CA III-containing organs and tissues did not show a change when lean and obese animals were compared. Additionally, levels of CA III in adipose tissue from obese animals responded to acute changes in energy balance of the animal. These results are discussed in light of possible metabolic roles for CA III.

DraI and ScaI RFLPs at human tyrosine hydroxylase (TH) gene locus

Images

EcoRI and NsiI RFLPs at a human PLA2 gene on chromosome 1

Images

Human estrogen receptor (ESR) gene locus: PssI dimorphism

Images

Bsm I RFLP at the human apolipoprotein AI-CIII gene complex locus

Images

NcoI dimorphic site located 8kb 3' to the human apolipoprotein AIV (APOA4) gene

Images

A BanII dimorphic site located in the third intron of the human apolipoprotein AI (APOA1) gene

Images

Human natriuretic peptides (ANP) gene locus: BsmI RFLP

Images

Human apolipoprotein CI (ApoC1) gene locus: BglI dimorphic site

Images

BanI dimorphic site in the third intron of the human apolipoprotein AI gene (Apo A1)

Images

Dimorphic markers for the human apolipoprotein CII gene locus

Two restriction-site polymorphisms (RSP) have been detected when using a human apolipoprotein CII (apoCII) cDNA clone as a hybridization probe. These include a BanI and a TaqI RSP. Frequencies of the more common allele have been determined in a German population of 100 individuals and are 0.66 (BanI RSP) and 0.56 (TaqI RSP). Corresponding polymorphic information content (PIC) values are 0.36 and 0.37 for individual sites, and 0.58 for the BanI-TaqI pair of sites, making this locus a very informative (PIC-rich) marker for this region of chromosome 19. Haplotype studies also indicate the presence of allelic association (linkage disequilibrium) at the human apoCII gene locus.

Human atrial natriuretic peptides (ANP) gene locus: BglI RFLP

Images

DraI RFLP in the human apolipoprotein AI-CIII-AIV gene complex

Images

NcoI RFLP at the human apolipoprotein CII gene locus

Images

Polymorphisms in the apolipoprotein AI-CIII gene complex

We report the presence of five restriction fragment length polymorphisms in the apolipoprotein AI-CIII (apo AI-CIII) gene complex as well as their respective allelic frequencies in a German population (pi) that we studied: an SstI (BanII) polymorphism in the 3' non-coding region of apo CIII q1(pi) = 0.13); an MspI polymorphism in the third intron of apo AI (q2(pi) = 0.12); a PvuII polymorphism in the first intron of apo CIII (q3(pi) = 0.30); and two XmnI polymorphisms in the 5' side of apo AI (q4(pi) = 0.20 and q5(pi) = 0.05).

Pvu II RFLP in the 5' of the human apolipoprotein B gene

Images

Molecular processing of adenovirus proteins

Late in adenovirus infection, a virus-encoded protease processes several viral structural proteins. The maturation cleavages are a prerequisite for full viral infectivity. The peptide fragment removed during processing is located at the amino end of the major core protein VII. The structure of the precursor peptide sequence was determined by both protein and nucleotide sequencing. Two processing events were elucidated. First, during protein biosynthesis, the initiator methionyl residue is removed and the penultimate seryl residue is acetylated. Second, the resulting NH2-terminal 23-residue fragment is removed during virus assembly. The specificity of the viral endoprotease was investigated by isolating and characterizing another viral proprotein precursor, Pro-VI. The propeptide of VI was also found to be extended at the amino end of the molecule. Comparison of the two propeptide sequences at the cleavage site revealed a consensus amino acid sequence of Gly-Gly-Ala. In addition, there is extensive similarity in the precursor sequences of both proteins. The analogous constitution of the precursor fragments in Pro-VI and Pro-VII suggests that a common mechanism is implicated in controlling the reorganization of VI and VII during virion assembly.

Gene and protein sequences of adenovirus protein VII, a hybrid basic chromosomal protein

The sequences of both the gene and the corresponding protein of adenovirus major core protein VII have been determined. The precise location of this gene is between 43.37 and 44.90 map coordinates on the viral genome. Protein VII is 173 residues long and has a molecular weight of 19,258. Detailed analysis of its sequence has revealed four basic domains separated by several predicted alpha helices. It is proposed that intrachain folding of protein VII is driven by hydrophobic interactions of the alpha helices, leaving the basic domains of the protein to interact with DNA phosphates. Protein monomers may further associate with each other in the formation of hexameric nucleosome-like particles. The displacement and replacement of protein VII during the viral infectious cycle in the host cell appears to mimic the biology of nucleoprotamine during the processes of spermatogenesis and fertilization. The presence of a protamine-like domain affirms a hybrid histone/protamine molecular structure for protein VII, although it may resemble the protamine in function.