Analysis of prolactin-modulated gene expression profiles during the Nb2 cell cycle using differential screening techniques

Abstract

Background: Rat Nb2-11C lymphoma cells are dependent on prolactin for proliferation and are widely used to study prolactin signaling pathways. To investigate the role of this hormone in the transcriptional mechanisms that underlie prolactin-stimulated mitogenesis, five different techniques were used to isolate differentially expressed transcripts: mRNA differential display, representational difference analysis (RDA), subtractive suppressive hybridization (SSH), analysis of weakly expressed candidate genes, and differential screening of an organized library.

Results: About 70 transcripts were found to be modulated in Nb2 cells following prolactin treatment. Of these, approximately 20 represent unknown genes. All cDNAs were characterized by northern blot analysis and categorized on the basis of their expression profiles and the functions of the known genes. We compared our data with other cell-cycle-regulated transcripts and found several new potential signaling molecules that may be involved in Nb2 cell growth. In addition, abnormalities in the expression patterns of several transcripts were detected in Nb2 cells, including the constitutive expression of the immediate-early gene EGR-1. Finally, we compared the differential screening techniques in terms of sensitivity, efficiency and occurrence of false positives.

Conclusions: Using these techniques to determine which genes are differentially expressed in Nb2 lymphoma cells, we have obtained valuable insight into the potential functions of some of these genes in the cell cycle. Although this information is preliminary, comparison with other eukaryotic models of cell-cycle progression enables identification of expression abnormalities and proteins potentially involved in signal transduction, which could indicate new directions for research.

Figure 1

Cell-cycle analysis of synchronized Nb2 cells stimulated by prolactin. Nb2 cells were serum deprived for 24 hours, then incubated with no PRL or PRL at 20 ng/ml before cells were collected for analysis. DNA content (FL2-A) versus cell number is presented in each panel. (a) Profiles obtained with control cells; (b) profiles obtained with cells incubated with PRL. From left to right, profiles correspond to cells in apoptosis (Apo, area below 400 on the x axis), in G0/G1 (peak centered on 400 on the x axis), or in S/M phase (area above 400 on the x axis).

Figure 2

Expression profiles of various known and unknown transcripts during Nb2 cell cycle progression. Samples of total RNA (10 μg) were loaded per lane and blots were hybridized with the indicated cDNA probes. Ethidium bromide staining (EtBr) of the gels is shown as a control (18S and 28S rRNA). (a) Cdc21 homolog; (b) Ant-2; (c) CRM-1; (d) CD45; (e) unknown DD3; (f) unknown 4-16; (g) unknown 4-15; (h) unknown 6-4.

Figure 3

Analysis of candidate gene expression. (a) General principles. Messenger RNAs from the different cell populations (cells A and B) are reverse transcribed. Multiplex PCR is then performed using specific primer pairs to amplify the cDNAs of interest. The resulting mixture of PCR products is radiolabeled and these complex probes are used to hybridize identical membranes spotted with the candidate gene cDNA targets. After autoradiography, the intensities of the hybridization signals are compared and quantified. Arrows indicate the positions of differentially expressed genes. The absence of hybridization (open circles) indicates that the candidate gene is not expressed. (b) Efficiency of the technique and examples of differentially expressed genes. The expression of different candidate genes was compared in either unsynchronized (UN), growth-arrested (GA), G1 phase (G1), G1/S transition (G1/S) or G2 phase (G2) cultures of Nb2 cells. The efficiency of the technique was controlled using equivalent amounts of rabbit α and β globin cDNAs, which were included on the nylon membranes along with the candidate gene targets. The two globin cDNAs were added in different amounts (50 or 150 ng) to each cDNA population before co-amplification. For each population tested, filters were hybridized with both globin probes, but only representative hybridization signals are shown, for either α (Panel A) or β (Panel B) globin. Numbers 1 and 3 represent the relative amount of the control rabbit globin cDNAs added, and are reflected in the differences in the intensity of the hybridization signals. Thus, a threefold difference in the quantity of a particular transcript in the initial population generates a clear difference in the intensity of the corresponding hybridization signals. Rows C, D, E and F are examples of the results obtained with ganglioside synthase GD3, EGR-1, FAK p125 and Stat3, respectively. Except for Stat3, which is not differentially expressed in probes UN, GA, G1, G1/S and G2, the three other genes showed a clear differential expression. (c) Northern blot analysis showing the constitutive expression of EGR-1 during Nb2 cell-cycle progression. Growth-arrested Nb2 cells were stimulated with ovine prolactin and collected after various periods of stimulation corresponding to different stages of the cell cycle (G1, G1/S and G2). The expression of EGR-1 was evaluated by northern blot using 10 μg of total RNA from the various times following prolactin stimulation. Ethidium bromide (EtBr) staining of the gel is shown as a control (18S and 28S rRNA).

Figure 4

Schematic representation of rat candidate genes on a nylon filter. Squares with names and accession numbers represent the places where the cDNAs were spotted. The solid gray boxes correspond to the controls (rabbit α and β globin). The boxes enclosed in a thick black square represent differentially expressed genes in Nb2 cells; the boxes enclosed in a thin black square represent genes that are repressed, but not differentially in Nb2 cells; and those enclosed in an oval correspond to expression abnormalities in Nb2 cells.