QTL-mapping is the process of utilizing genetically mapped varieties coupled with a biological measurement (as Fe bioavailability) and then utilizing statistics to correlate that measurement with genetic markers. QTL-mapping revealed that Fe concentration in maize grain was under the control of at least 10 regulatory factors on 6 of the 10 chromosomes of maize 9 . However, Fe-bioavailability was regulated by fewer, larger QTL, which suggested that this trait might be easier to manipulate. Furthermore, Fe concentration and bioavailability had only a small positive association between them indicating that Fe concentration differences between members of the mapping population were not driving the differences in Fe bioavailability 9 . Derivation of the High-Fe and Low-Fe bioavailability maize hybrids was previously described 8 9 . Briefly, The Caco-2 bioassay was the guiding tool for the measure of Fe bioavailability in the maize grain 9 . Statistical analysis was used to identify molecular markers (i.e. QTL) associated with Fe bioavailability. These markers were used to select sister lines that contrasted for the 3 largest effect QTL in order to create new varieties that were highly genetically similar but different (high or low) for Fe-bioavailability. As sister lines were created in both of the parental genetic backgrounds used in the mapping population, nearly isogenic hybrids were made by crossing the parents lines (high with high and low with low). These hybrids were heterozygous everywhere except the 3 Fe-bioavailability QTL 9 and were similar except for bioavailable-Fe in the whole grain 8 (Figure 1). The High-Fe and Low-Fe maize were produced using standard agronomic practices at the Cornell University Research Farm (Poplar Ridge, NY) in the summer of 2009. Plots were mechanically planted and harvested. Grain was dried to ~12% moisture, processed in bulk (~ 800 Kg of each variety), and stored at 4°C until the feeding study began. In preparation for the in vivo trial, maize grains were thoroughly washed in _ddH2O prior to cooking and freeze drying. Maize varieties were ground prior to mixing the diets.

One hundred and twenty fertile Cornish cross broiler eggs were obtained from a commercial hatchery (Moyer’s Chicks, Quakertown, PA). Eggs were incubated under optimal conditions at the Cornell University Animal Science poultry farm incubator. Upon hatching (92% hatchability), chicks were allocated into 4 treatment groups on the basis of body weight, gender and blood hemoglobin concentration (aimed to ensure equal distribution between groups, n=10): 1. "High + Fe": 75% High-Fe bioavailability cooked maize with added Fe based diet (65 μg/g Fe). 2. "High": 75% High-Fe bioavailability cooked maize with no Fe added based diet (24 μg/g Fe). 3. "Low + Fe": 75% Low-Fe bioavailability cooked maize with added Fe based diet (66 μg/g Fe). 4. "Low": 75% Low-Fe bioavailability cooked maize with no Fe added based diet (23 μg/g Fe) (Table 1). Cooked/raw maize were compared as in vitro pilot studies indicated that cooking may increase the difference in Fe bioavailability between the two lines. Chicks were housed in a total-confinement building (1 chick per 0.5 m² cage). Birds were under indoor controlled temperatures and were provided 16 h of light. Cages were equipped with an automatic nipple drinker and manual self feeder. All birds were given ad libitum access to water (Fe concentration was 0.379±0.012 μg/g). Iron concentrations in the water and diets were determined by an inductively-coupled argon-plasma/atomic emission spectrophotometer (ICAP 61E Thermal Jarrell Ash Trace Analyzer, Jarrell Ash Co. Franklin, MA) following wet ashing. Feed intakes were measured daily (from day 1). Iron intakes were calculated from feed intakes and Fe concentration in the diets.

¹Vitamin and mineral premix provided/kg diet (330002 Chick vitamin mixture; 230000 Salt mix for chick diet; Dyets Inc. Bethlehem, PA).

²Dietary iron concentrations analysis is described in the materials and methods section.

³Method for determining phytate contents are described in the materials and methods section.

⁴Values are means±SEM. ^a,bWithin a row, means without a common letter are significantly different, P < 0.05.

Blood samples were collected from the wing vein (n=10,~100 μL) using micro-hematocrit heparinized capillary tubes^a (Fisher, Pittsburgh, PA). Samples were collected following an 8 h overnight feed deprivation. Samples were analyzed for Hb concentration (see below). Body weights (BW) and Hb concentrations were measured weekly.

Fe-bioavailability was calculated as hemoglobin maintenance efficiency (HME) 23 24 25 26 27 28 29 :

HME = Hb Fe , mg final − Hb Fe , mg initial Total Fe Intake , mg × 100

Where Hb-Fe (index of Fe absorption) = total body hemoglobin Fe. Hb-Fe was calculated from hemoglobin concentrations and estimates of blood volume based on BW (a blood volume of 85 mL per kg body weight is assumed) 23 24 25 28 :

Hb − Fe mg = BW kg × 0.085 L blood / kg × Hb g / L × 3 . 35 mg Fe / g Hb .

Fe intakes were calculated from feed intake data and Fe concentrations in the feed.

Blood Hb concentrations were determined spectrophotometrically using the cyanmethemoglobin method (H7506-STD, Pointe Scientific Inc. Canton, MI) following the kit manufacturer’s instructions.

At the end of the experiment (day 42), birds were euthanized by carbon-dioxide exposure. The digestive tracts and livers were quickly removed and separated. Tissue samples were taken from the small intestine and liver (~ 1–2 cm; ~2-3 g, respectively). The samples were immediately frozen in liquid nitrogen, and then stored in a -80°C freezer until analysis.

All animal protocols were approved by the Cornell University Institutional Animal Care and Use Committee.

Total RNA was extracted from 30 mg of the proximal duodenal tissue (n=10) using Qiagen RNeasy Mini Kit (RNeasy Mini Kit, Qiagen Inc.,Valencia, CA) according to the manufacturer’s protocol. Total RNA was eluted in 50 μL of RNase free water. All steps were carried out under RNase free conditions. RNA was quantified by absorbance at A _260/280 . Integrity of the 28S and 18S ribosomal RNAs was verified by 1.5% agarose gel electrophoresis followed by ethidium-bromide staining. DNA contamination was removed using TURBO DNase treatment and removal kit from AMBION (Austin, TX, USA).

As previously described 23 24 25 27 30 , Divalent metal transporter-1 (DMT1); Duodenal cytochrome-B (DcytB) and Ferroprtin mRNA levels in duodenal mucosa were analyzed by quantitative real-time RT-PCR (20 μL reactions); values were normalized to 18S expression. The total RNA was reverse-transcribed to complementary DNA in a 25 μL volume containing 1 μg of extracted RNA. Reverse-transcription was carried out using the Superscript-First Strand Synthesis Kit for reverse-transcription PCR according to the manufacturer’s protocol (Invitrogen, Carlsbad, CA). Gene-specific primers were designed using Primer Express software (Applied Biosystems, Carlsbad, CA) chosen from the fragment of the chicken (Gallus gallus) duodenal DMT1 gene (GeneBank database; GI 206597489) (forward: 5’-AGC CGT TCA CCA CTT ATT TCG-3’; reverse: 5’-GGT CCA AAT AGG CGA TGC TC-3’), DcytB gene (GI 20380692) (forward: 5’-GGC CGT GTT TGA GAA CCA CAA TGT T-3’; reverse: 5’-CGT TTG CAA TCA CGT TTC CAA AGA T-3’) and Ferroportin gene (GI 61098365) (forward: 5’-GAT GCA TTC TGA ACA ACC AAG GA’; reverse: 5’-GGA GAC TGG GTG GAC AAG AAC TC-3’). Ribosomal 18S was used to normalize the results (GI 7262899) (forward: 5’- CGA TGC TCT TAA CTG AGT-3’; reverse: 5’-CAG CTT TGC AAC CAT ACT C-3’). Real-time PCR was performed in a 7500 Real-Time PCR system instrument (Applied Biosystem, Carlsbad, CA). The 20 μL PCR mixture consisted of 10 μL of POWER SYBR Green PCR Master Mix (Applied Biosystem, Carlsbad, CA), 5 μL of water, and 1 μL of each primer that was added to 3 μL of the cDNA diluted 1:25. All reactions were performed in duplicates and under the following conditions: 50°C for 2 min, 95°C for 2 min, 42 cycles of 95°C for 30 s, and 60°C for 1 min. Also, to ensure amplification of a single product, a dissociation curve was determined under the following conditions: 95°C for 1 min, 55°C for 30 s, and 95°C for 30 s. Specificity of the product was also confirmed by running samples on a 1.5% agarose gel, excising for purification using the QIAquick Gel Extraction Kit (QIAGEN, Valencia, CA). Calculations of threshold cycles, amplification efficiencies, and R0 values (the starting fluorescence value that is proportional to the relative starting template concentration) were performed using the data analysis for real-time PCR Excel workbook and as previously described 31 .

We followed previously described procedures 23 24 32 33 . Briefly, 1 g of sample was diluted into 1 mL of 50 mM Hepes buffer, pH 7.4, and homogenized on ice for 2 min (5000 g). One mL of each homogenate was subjected to heat treatment for 10 min at 75°C to aid isolation of ferritin (other proteins are not stable at that temperature). Subsequently, samples were immediately cooled down on ice for 30 min. Thereafter, samples were centrifuged for 30 min (13000 g) at 4°C until a clear supernatant was obtained and the pellet containing most of the insoluble denaturated proteins was discarded. Iron concentrations in the liver samples were determined by an inductively-coupled argon-plasma/atomic emission spectrophotometer (ICAP 61E Thermal Jarrell Ash Trace Analyzer, Jarrell Ash Co. Franklin, MA) following wet ashing.

Native polyacrylamide gel electrophoresis was conducted using a 6% separating gel and a 5% stacking gel. Samples were run at a constant voltage of 100 V. Thereafter, gels were treated with either of the two stains: Coomasie blue G-250 stain, specific for proteins, or potassium ferricyanide (K₃Fe(CN)₆) stain, specific for Fe. The corresponding band found in the protein and Fe stained gel was considered to be ferritin 23 24 32 33 .

Measurements of the bands were conducted using the Quantity-One-1-D analysis program (Bio-Rad, Hercules, CA). The local background was subtracted from each sample. Horse spleen ferritin (Sigma Aldrich Co., St. Louis, MO) was used as a standard for calibrating ferritin protein and Fe concentrations of the samples. Dilutions of the horse spleen ferritin were made and treated similarly to the liver supernatant samples in order to create a reference line for both protein and Fe-stained gels 23 24 32 33 .

An in vitro digestion/Caco-2 cell culture model 19 23 24 25 26 27 28 34 35 was used to assess Fe-bioavailability. The maize only samples (High- Fe bioavailability maize; Low-Fe bioavailability maize and control-commercial maize) and the diets (High diet; Low diet; High + Fe diet; Low + Fe diet) were subjected to simulated gastric and intestinal digestion. Briefly, the intestinal digestion is carried out in cylindrical inserts closed on the bottom by a semipermeable membrane and placed in wells containing Caco-2 cell monolayers bathed in culture medium. The upper chamber was formed by fitting the bottom of Transwell insert ring (Corning) with a 15000 Da molecular weight cut off (MWCO) membrane (Spectra/Por 2.1, Spectrum Medical, Gardena, CA). The dialysis membrane was held in place using a silicone ring (Web Seal, Rochester, NY).

Iron uptake by the Caco-2 cell monolayers was assessed by measuring ferritin concentrations in the cells. Six replicates of each Fe bioavailability measurement were performed. In terms of materials for the study, Caco-2 cells were obtained from the American Type Culture Collection (Rockville, MD) at passage 17 and used in experiments at passage 29. Cells were seeded at densities of 50,000 cells/cm² in collagen-treated 6 well plates (Costar Corp., Cambridge, MA). The integrity of the monolayer was verified by optical microscopy. The cells were cultured at 37°C in an incubator with 5% CO₂ and 95% air atmosphere at constant humidity, and the medium was changed every 48 h.

The cells were maintained in Dulbecco’s modified Eagle medium plus 1% antibiotic/antimycotic solution, 25 mmol/L HEPES, and 10% fetal bovine serum. 48 h prior the experiment, the growth medium was removed from culture wells, the cell layer was washed, and the growth medium was replaced with minimum essential media (MEM) at pH 7.0. The MEM was supplemented with 10 mmol/L PIPES, 1% antibiotic/antimycotic solution, 4 mg/L hydrocortisone, 5 mg/L insulin, 5 μg/L selenium, 34 μg/L triiodothyronine, and 20 μg/L epidermal growth factor. This enriched MEM contained less than 80 μg Fe/L.

All ingredients and supplements for cell culture media were obtained from GIBCO (Rockville, MD). The cells were used in the Fe uptake experiment at 13 days post seeding. In these conditions, the amount of cell protein measured in each well was highly consistent between wells. On experiment day, 1.5 mL of the digested sample was added to the insert’s upper chamber and incubated for 2 h. Then, inserts were removed and 1 mL of MEM was added. Cell cultures were incubated for 22 h at 37°C.

It was previously shown that intracellular ascorbic acid status might influence ferritin formation (i.e. cellular Fe uptake), and Fe related transporters and enzyme expression in Caco-2 cells 23 24 34 . In the current study, samples were not added with ascorbic acid when Fe bioavailability was tested in vitro.

The ferritin and total protein contents analyses protocols were previously described 19 23 24 35 . Briefly, growth medium was removed from the culture well by aspiration and the cells were washed twice with a solution containing 140 mmol/L NaCl, 5 mmol/L KCl, and 10 mmol/L PIPES at pH 7.0. The cells were harvested by adding an aliquot of deionized water and placing them in a sonicator (Lab-Line instruments, Melrose Park, IL).

The ferritin and total protein concentrations were determined on an aliquot of the harvested cell suspension with a one-stage sandwich immunoradiometric assay (FER-IRON II Ferritin assay, Ramco laboratories, Houston, TX) and a colorimetric assay (Bio-Rad DC Protein assay, Bio-Rad, Hercules, CA), respectively. Caco-2 cells synthesize ferritin in response to increases in intracellular Fe concentration. Therefore, we used the ratio of ferritin/total protein (expressed as ng ferritin/mg protein) as an index of the cellular Fe-uptake.

A Dionex liquid (Dionex Corp. Sunnyvale, CA) chromatograph system (AS50 autosampler), equipped with conductivity detector model ED50, and gradient pump GS50 were used along with an IonPac AG11 guard column and IonPac AS11 column (4×250 mm) to quantify phytate. PeakNet 6.40 software was used to process chromatographic data. The mobile phases were (A) 200 mmol/L NaOH (carbonate-free) and (B) deionized water, using a flow rate of 1 mL/min. Phytate was extracted from 250 mg of dry, lyophilized diet sample, in 10mL of a 1.25% H₂SO4 solution; the extraction process was 2 h, after which the samples were centrifuged at 3660 g for 10 min. Subsamples were diluted 1:10 with deionized water, and 10 μL was injected and analyzed (n=10).

Results were analyzed by ANOVA using the general linear models procedure of SAS software (SAS Institute Inc. Cary, NC). Differences between treatments were compared by Tukey’s test were considered statisticant at P < 0.050. Values in the text are means ± SEM.

No significant differences were measured in body weights between treatment groups (P > 0.05). However, as from day 21 of the study, hemoglobin (Hb) concentrations were higher (P < 0.05) in the "High" group than in the "Low" group. In addition, as from day 14, Hb-Fe values were higher in the "High" group than in the "Low" group; the increase in total body Hb-Fe from the beginning of the study to the end of the 6th wk was significantly greater in the "High" group than in the "Low" group (12.8 ± 0.5 mg vs. 9.7 ± 0.3 mg, respectively, P < 0.05, Table 2). Significant differences in HME (P < 0.05) were measured between the "High" group and "Low" group on day 21 (P < 0.05). Also, significant differences in HME (P < 0.05) were measured between the "High + Fe" and "Low + Fe" groups on days 28 and 42 (P < 0.05, Table 2).

^a,b,cWithin a column and for each parameter (i.e. Hb, Hb Fe, HME), treatment group means without a common letter differ, P < 0.05.

¹Calculations are described in the materials and methods section.

²Values are means±SEM, n=10.

³The experimental diets are described in the materials and methods section.

Avian ferritins corresponded to a weight of approximately 470 to 500 kDa 23 24 32 33 36 . Liver Fe and ferritin concentrations were higher in the "High" group than in the "Low" group (n=10, P < 0.05, Table 3).

^a,b,cWithin a column and for each parameter (i.e. liver ferritin or liver Fe), treatment group means without a common letter differ, P< 0.05 (values are mean±SEM, n=10).

¹Atomic mass for iron is 55.8 g/mol.

²Liver tissue iron concentrations analysis is described in the materials and methods section.

Gene expression analysis of duodenal DMT-1, Ferroportin and DcytB, with results reported relative to 18S rRNA, revealed greater mRNA levels for DMT1, DcytB and Ferroportin in the "Low" group compared to the "High" group (mean±SEM) (n=10, P < 0.05, Figure 2).

An in vitro digestion/Caco-2 cell culture model was used to evaluate Fe bioavailability from the tested maize only and maize based diets by measuring ferritin formation in the cells (ie. a measure of cell Fe uptake) following exposure to digests of the samples. The amount of bioavailable iron in vitro was significantly higher (P < 0.05) in the "High" and "High + Fe" diets than in the "Low" and "Low + Fe" diets (mean±SEM) (n=6, P < 0.05, Table 4).

¹Values are means ± SEM, n = 6.

^a,b,c,d,e,f Within a column (ferritin or Fe concentrations), means without a common letter differ, P < 0.05.

²Caco-2 bioassay procedures and preparation of the digested samples are described in the materials and methods section.

³Dietary iron concentrations analysis is described in the materials and methods section.

No significant differences in phytate concentration (IP₆) were measured between treatments diets (n=5; P > 0.05, Table 1).