Expression of the FPN1B transcript increased in the duodenum of mice maintained on a low-iron diet. (A) Northern blot and (B) qRT-PCR were used to check the expression of FPN1B and other iron genes in different parts of intestinal tract. The results revealed that FPN1B expression increased on a low-iron diet treatment. Data are presented as mean ± SD. * P<0.05 and ** P<0.01 comparing with corresponding high-iron diet.

The FPN1B promoter was regulated by putative GATA and KLF transcriptional elements. (A) Alignment of promoter region of mouse and human FPN1B. Sequences in black boxes indicated the consensus transcription factor binding sites, the sequence in gray indicated the new exon1, “*” indicated conserved sequences between human and mouse. (B) The FPN1B promoter construct displayed high activity in MEL and K562 cells, whereas FPN1A promoter displayed promoter activity in all of the cells tested. Empty pGL4-luciferase vector was used as a control. Data are presented as mean ± SD. (C) Serial deletion assays in MEL cells showed that the FPN1B promoter was located within 133bp region before the transcription initiation site. (D) Site-specific mutations in MEL cells showed that a GATA-114 site, and a EKLF/GKLF/SP1-71 site were essential for FPN1B promoter activity. The left panels showed the full sequence inserted in the pGL4 luciferase vector, from 1000bp before to 50bp after the transcription initiation site (TIS), the small black boxes indicated by arrows indicate the putative transcription factor binding sites, and white boxes indicate sites at which a consensus transcription factor binding site was mutagenized. The wide type (WT) construct was taken as 100% promoter activity. Data are presented as mean ± SD.

Expression of FPN1 increased in the duodenum of mice maintained on low-iron diet. (A) FPN1 expression (red) was detected with a polyclonal anti-FPN1 antibody and visualized by immunofluorescence in the duodenal sections of low- and high-iron diet mice. The nucleus was counterstained with DAPI (blue). (B) Prussian blue staining revealed more iron (blue) in the intestinal mucosa and spleen of high-iron diet mice than that of low-iron diet mice, counterstained with fast red for nucleus (red) and cytoplasm (pink). (C) Western blots revealed high expression of FPN1 in the lysate from duodenum of mice on a low-iron diet (40μg per lane). Levels of DMT1 and TfR1 increased on the low-iron diet, whereas FtL and FtH decreased. (D) Northern blots revealed that mRNA expression of FPN1, DMT1, TfR1 and Dcytb increased in duodenum of mice on a low-iron diet, whereas FtL mRNA did not change.

A transcript of FPN1 that lacked an IRE was identified in mouse. (A) Schematic representation of genomic (top) and transcript structure (bottom) of FPN1. FPN1 has a total of nine exons (boxes 1b, 1a, 2 through 8, introns depicted as lines), and the IRE and AUG codon are located in exon 1a. By alternative usage of exon 1a or 1b, two FPN1 transcripts are generated. The FPN1A transcript extends from exon1a, and includes exons 2 to 8, producing a transcript that contains an IRE in its 5'UTR. The FPN1B transcript commences with exon 1b and includes a portion of exon 1a that is 3' of the IRE (from nucleotide -145, -102 or -64 depending on different acceptor sites), as well as exons 2 to 8, producing a transcript that does not have an IRE. Primers F1, F2 and R, as indicated on the graphic were used for PCR amplification of FPN1A and FPN1B. (B) PCR amplification of FPN1A (lanes 2, 5 and 8) and FPN1B (lanes 3, 6 and 9) using cDNA generated from spleen, kidney and duodenum revealed three bands which corresponded to the usage of three distinct acceptor sites within exon 1a of FPN1B. FPN1 was amplified by a pair of primers located within the ORF (lane 4, 7 and 10). The expected sizes for each of the bands were 596bp (FPN1), 356bp (FPN1A), and 312bp, 269bp, 231bp (FPN1B). PCR products were separated on a 2% agarose gel. (C) Expression of the FPN1B reporter construct was not repressed in iron starvation conditions. The 5'UTRs of FPN1A and FPN1B were cloned into the pGL3 luciferase vector, and then 1 μg of those vectors with 0.1 μg of renilla luciferase vector were transfected into IEC6 cells. One day later those cells were treated with either 200 μM, 100 μM FAC, 100 μM, or 200 μM DFO for 24h. The luciferase activity was measured by a dual luciferase activity assay. FPN1B1, B2 and B3 represent the longest, middle and shortest transcripts, respectively. Data are presented as mean ± SD.

The FPN1B transcript was highly expressed in duodenum and bone marrow. (A) Total RNA of multiple tissues was probed with FPN1B or 1A specific probes in northern blots. (B) qRT-PCR was used to check the expression of FPN1B and FPN1A in cDNA of mouse tissues. The results revealed that FPN1B was specifically expressed in duodenum and bone marrow. Data are presented as mean ± SD. (C) qRT-PCR was used to check the expression of FPN1B, FPN1A, total FPN1, DMT1, Dcytb, TfR1 and FtL in cDNA from different parts of intestinal tract. The expression level was normalized to beta-actin. Data are presented as mean ± SD. (D) In situ hybridization to check the expression of the FPN1B and 1A transcripts revealed that both were expressed in duodenal mucosa cells.

FPN1B was specifically expressed in erythroid precursor cells and decreased with cell differentiation. (A) Northern blot showed that FPN1B expressed in bone marrow and MEL cells was identical in size to the approximately 3.4 kb duodenal FPN1B transcript. (B) PCR experiments verified that all of the three FPN1B sub-transcripts were expressed in mouse Ter119-positive erythroid cells. (C) qRT-PCR to check the expression of FPN1B, FPN1A and FPN1 mRNA expression in MEL cells, G1E cells and mouse bone marrow, normalized with actin, revealed high FPN1B expression in these two erythroid precursor cell lines. Data are presented as mean ± SD. (D) qRT-PCR results showed that FPN1B was specifically expressed in Ter119-positive erythroid cells (Ery) rather than CD11b-positive macrophages (Mac) or cells depleted of erythroid cells and macrophages. Mac1 (macrophage antigen alpha) and Hbb (Hemoglobin beta chain) were used as controls to verify the isolation. The expression levels of these genes were normalized with actin. Data are presented as mean ± SD. (E) Western blot showed that ferroportin was highly expressed in erythroid cells. TfR1 was used as a positive control for erythroid cells, and alpha-tubulin was used as a loading control. (F) FPN1B expression significantly decreased during erythroid cell differentiation. qRT-PCR was used to check the gene expression in G1E-ER4 cell at 6h, 12h, 24h, 36h, 48h, 60h and 72h after the addition of 10nM of estradiol to induce cell differentiation. The expression levels of these genes were normalized with actin. Data are presented as mean ± SD. (G) Western blot revealed that ferroportin protein decreased with differentiation of G1E-ER4 cells after addition of estradiol, whereas TfR1 and FtL expression increased with differentiation.

A model for the physiological roles of the FPN1B transcript in the duodenal enterocytes and erythroid precursor cells. In duodenal enterocytes, FPN1 transports iron across the basolateral membrane to bind to transferrin in the blood stream, facilitated by the ferroxidase, hephaestin (Hp). (A) In iron-replete conditions, the liver senses body iron stores as reflected by transferrin saturation levels, and hepatocytes secrete high amounts of the iron-regulatory hormone hepcidin. In the duodenum, hepcidin is thought to bind to FPN1, and to thereby induce FPN1 internalization and degradation. Intracellularly, IRPs are inactivated in iron-replete animals, and both FPN1A and FPN1B can freely be translated into FPN1 protein. (B) In iron-deficient conditions, the liver ceases to produce hepcidin in response to the low levels of circulating iron. Therefore the hepcidin dependent degradation of FPN1 is eliminated. However, in iron-depleted cells, IRPs bind to the 5'IRE of FPN1A and repress the synthesis of FPN1 protein. Nevertheless, as levels of FPN1B transcript increase in iron deficiency, translation of FPN1B is not repressed by IRPs, and continued FPN1 expression allows sufficient iron export to satisfy the systemic iron demands of iron-deficient animals. In erythroid precursor cells, the Tf/TfR1 cycle actively transports iron into cells, and FPN1 generated mainly from FPN1B exports iron to Tf, facilitated by ceruloplasmin (CP). By active uptake and export of iron, erythroid precursor cells can precisely sense systemic iron levels and can determine whether erythropoietic expansion is warranted. (C) In iron-replete conditions, the Tf/TfR1 cycle transports more iron into the intracellular iron pool, and increased hepcidin expression likely blocks FPN1 from exporting iron out of cells, which increases intracellular iron levels and enhances red cell production and heme synthesis. (D) In iron-depleted conditions, because of decreased iron uptake from the Tf/TfR1 cycle and increased iron export from FPN1, diminished iron availability results in microcytic anemia characterized by decreased production of red blood cells and heme.