Effects of 5' untranslated region diversity on the posttranscriptional regulation of the human reduced folate carrier

Abstract

The human RFC (hRFC) gene is regulated by five major 5' non-coding exons, characterized by alternate transcription start sites and splice forms. The result is up to 14 hRFC transcripts for which different 5' untranslated regions (UTRs) are fused to a common coding sequence. By in vitro translation assays with hRFC constructs corresponding to the major transcript forms, most of the forms were translated poorly. Upon expression of the 5'UTR-hRFC constructs in hRFC-null HeLa cells, a range of steady state hRFC proteins and transcripts were detected that reflected relative transcript stabilities and, to a lesser extent, translation efficiencies. Transcripts including 5' UTRs derived from non-coding exon A encoded a modified hRFC protein translated from an upstream initiation site. When this modified hRFC protein was expressed in hRFC-null K562 cells, there were only minor differences in surface targeting, stability, or transport function from wild type hRFC. Our results demonstrate an important role for posttranscriptional determinants of cellular hRFC levels and activity.

Fig. 1. Schematic representation of the human reduced folate carrier (hRFC) upstream region showing alternate non-coding exons

Panel A: The upper panel shows the upstream region of the hRFC gene including five major non-coding regions (A1/A2 and A–D) and the first coding exon (Exon 1) including the major ATG translation start site (position +1). The lower panel shows the origin of the 14 alternate 5’UTRs from the five major non-coding regions including 3 for A1/A2, 5 for A, 2 for B, 1 for C, and 3 for D. Each 5’UTR is linked to a common 49 bp non-coding sequence (positions −1 to −49) and a common hRFC ORF (not shown). The locations of uAUGs in the A1/A2 and A non-coding regions that initiate translation resulting in alternate hRFC proteins with additional N-teminal sequence are noted. Panel B: In the upper panel, the nucleotide sequence of the AV non-coding region is shown in its entirety (positions −3563 to −3505) which is spliced to the 5' sequence of coding exon 1 at position −49. The uAUG in exon A and the downstream AUG (ATG) at position +1 are in bold. All numbering corresponds to the distance from the translation start site with the A of the ATG in exon 1 representing +1. In the lower panel, the amino acid sequence encoded from the uAUG within non-coding exon A, and continuing into the common coding sequence for hRFC is shown. Italicized amino acids designate the additional residues encoded from exon A and the translation initiation methionines are underlined.

Fig. 2. In vitro transcription and translation of 5’ UTR hRFC expression constructs

Various 5’ UTR hRFC fusion constructs cloned into pCDNA3 (1 μg each) were transcribed from the T7 promoter and in vitro translated. Ten microliters of each reaction were separated on a 7.5% SDS polyacrylamide gel, transferred to polyvinylidene difluoride membranes, and probed with HA-specific primary antibody. Panel A shows a representation of the hRFC expression constructs, including 5’UTR sequences as defined in Figure 1, the common 49 bp sequence and the full length hRFC open reading frame with a C-terminal HA epitope at Gln587. Panel B shows a Western blot of in vitro translation products for all of the 5’UTR-hRFC fusion constructs. Densitometry values are shown below each lane. In panel C are shown the in vitro translation products of the BI-hRFC and A1/A2III-hRFC constructs such that the alternate (72 kDa) hRFC protein translated from an uAUG within non-coding exon A1/A2 is detected. In panel D are shown in vitro translation products of the AII-hRFC, AI-hRFC, BI-hRFC, AV-hRFC, and kmAV-hRFC constructs such that the alternate hRFC protein translated from an uAUG within non-coding exon A is detected (uAUG is present in the AI, AV, and kmAV constructs but not the AII or BI constructs). The arrows indicate the sizes of the traditional hRFC (65 kDa) and alternate (~67 kDa) A-hRFC isoforms, as described in the text. A greater sensitivity setting for the Odyssey detection system was used for the images shown in panels C and D. The results shown are representative of 3 experiments.

Fig. 3. Effects of the hRFC 5’ UTRs on the steady state hRFC transcripts and proteins in transfected HeLa cells

5’UTR-hRFC fusion constructs (3.6 μg each) were transfected into R5 HeLa cells, after which plasma membranes were isolated and hRFC protein levels were measured by Western blotting or total RNAs were isolated and hRFC transcript levels were measured by Northern blotting. In panel A is shown a Western blot of hRFC protein levels for the 5’UTR-hRFC fusion constructs, as measured with HA-specific antibody. Equal loading was established by stripping the membrane and reprobing with a Na,K ATPase antibody (not shown). hRFC migrates as a large complex due to its N-glycosylation site at asparagine 58 [8]. In panel B, total RNAs (10 μg) were fractionated on a 0.8% agarose/formaldehyde gel, capillary transferred to nitrocellulose, and probed with a ³²P-labeled hRFC cDNA probe (top panel). Equal loading was established by ethidium bromide staining (lower panel). For both the Westerns and Northerns, intensities of the signals were measured by densitometry and the values reported below each lane. This experiment was repeated three times. Representative data are shown.

Fig. 4. Effects of hRFC 5’ UTRs on the hRFC transcript stabilities

The 5’UTR hRFC constructs were transfected into R5 HeLa cells. The cells were treated with Actinomycin D (10 μg/ml), and total RNA samples (10 μg) were prepared at 0, 2, 4, 6, and 8 h, and analyzed by Northern blotting using a ³²P-labeled hRFC cDNA probe (upper panels). Loading was established by ethidium bromide staining 18S and 28S ribosomal RNAs (lower panels). Densitometry was performed on multiple exposures of film and hRFC transcript levels were normalized to relative RNA loading (from densitometry of ethidium bromide-stained 28S RNA bands) for calculation of half lives of first order transcript turnover. Half lives from the experiment shown are in parentheses. The images shown are representative of duplicate experiments which gave identical results.

Fig. 5. Expression of a hRFC-A protein isoform from an uAUG within non-coding exon

A. The 11-hRFC, AV-hRFC, and kmAV-hRFC constructs were stably transfected into the K500E (K562) hRFC-null cell line, as described in the Materials and Methods section. hRFC expression levels were measured by Western blotting, protein localization was determined by confocal microscopy, and rates of protein turnover were determined by treating the cells with cycloheximide and measuring protein expression on Western blots over 18 h. Panel A: hRFC proteins from crude plasma membranes from the stable clones and hRFC null K500 cell line were measured by Western blotting using an anti-HA primary and IRDye™ 800-conjugated secondary antibody (top panel). hRFC proteins were deglycosylated with N-glycosidase F prior to Western blot analysis (lower panel). The arrows indicate the approximate sizes for the different hRFC proteins isoforms. Panel B: Confocal microscopy of hRFC was used to demonstrate localization to the plasma membrane for the wild type protein (11-hRFC) and the alternate A-hRFC protein isoform (kmAV-hRFC). Cells were fixed with 3.3% paraformaldehyde, permeabilized with 0.1% Triton X-100, incubated with mouse anti-HA primary antibody, followed by anti-mouse IgG-Alexa Fluor 488-conjugated secondary antibody, and spun onto microscope slides. Slides were visualized with a Zeiss laser scanning microscope 310 using a x63 water immersion lens. Panel C: hRFC protein turnover was determined for the wild type carrier (11-hRFC) and the alternate hRFC isoform (kmAV-hRFC). Stable clones were treated with cycloheximide (0.2 mg/ml), and membrane fractions were isolated over 18 h, after which the membrane fractions were analyzed by Western blotting. For both 11-hRFC and kmAV-hRFC, approximately 50% of the hRFC protein decreased over 16 h.