The potential of fermentation and blanching in improving bioaccessibility and bioavailability of iron and zinc in African leafy vegetables

Abstract

Introduction - Iron and zinc deficiencies remain a significant problem in sub-Saharan Africa. Consumption of African leafy vegetables (ALVs) have shown potential in alleviating these deficiencies. However, ALVs also contain high levels of antinutrients which have some inhibitory effects on the absorption of iron and zinc. Processing methods such as blanching and spontaneous fermentation are effective in reducing antinutrients. The reduction of antinutrients in ALVs could improve iron and zinc bioaccessibility and bioavailability. The study aimed to compare mineral and antinutrients content of cooked ALVs to blanched-dried and spontaneously fermented-dried ALVs and with the overall aim to evaluate the bioaccessibility of iron and zinc from these processed ALVs and to predict their iron and zinc bioavailability using the Caco-2 cell assay model. Methods - Five ALVs, viz. spider plant (Cleome gynandra, amaranth (Amaranthus hybridus), wild jute (Corchorus olitorius), cowpea leaves (Vigna unguiculata) and Ethiopian kale (Brassica carinata A. Baun) were procured in Botswana. The ALVs were separated into three processing techniques: cooked (as typically prepared and consumed in households when fresh), blanched-dried, and spontaneously fermented-dried (consumed in households when out of season). Mineral (iron, zinc, calcium, phosphorus, magnesium and aluminium) contents were analysed by inductively coupled plasma optical emission spectroscopy (ICP-OES). Total phenolics, condensed tannins, total oxalates and phytate content were analysed. The bioaccessibility of iron and zinc from differently processed ALVs were also analysed using in vitro dialysability assay, and results were used to calculate contributions these ALV could make towards iron and zinc absolute requirements for women aged 19 to 50 years. The two ALVs with the best iron and zinc bioaccessibility were chosen to predict iron and zinc bioavailability using the in vitro digestion combined with the use of Caco-2 cells assay to simulate mineral uptake. Iron uptake by Caco-2 cells from processed ALVs was measured by determining ferritin formation in the cells using the Human ferritin ELISA kit. In contrast, zinc uptakes by the Caco-2 cells were analysed using inductively coupled plasma mass spectrometry (ICP-MS).

Results - Compared to their cooked counterparts the mineral content of spontaneously fermented-dried ALVs was significantly (p ≤ 0.05) lower for iron in wild jute (44%) and spider plant (14%); for magnesium in spider plant (14%), Ethiopian kale (7%), and amaranth (8%) and calcium in amaranth (13%). In contrast, compared to the cooked ALVs, the mineral content of spontaneously fermented-dried ALVs was significantly (p ≤ 0.05) higher for iron in amaranth (11%) and Ethiopian kale (83%); for zinc in cowpea leaves (17%), wild jute (22%) and Ethiopian kale (21%), for magnesium in wild jute (12%) and cowpea (14%) and calcium in cowpea (21%), spider plant (8%) and wild jute (12%). Compared to their cooked counterparts the mineral content of blanched-dried ALVs was significantly (p ≤ 0.05) lower for iron in cowpea (39%), spider plant (35%), wild jute (65%) and amaranth (12%), for magnesium in spider plant (14%) and Ethiopian kale (21%), for calcium in Ethiopian kale (23%), and for phosphorus in amaranth (30%) and Ethiopian kale (8%). In contrast, compared to cooked counterparts, mineral contents of blanched-dried ALVs were significantly (p ≤ 0.05) higher for iron in Ethiopian kale (49%), for zinc in Ethiopian kale (24%) and amaranth (6%), for calcium in wild jute (72%), phosphorus in amaranth (19%). Aluminium levels of cooked ALVs varied from 9.71 to 75.8 mg/100 g, db. However, the aluminium content of blanched-dried and spontaneously fermented-dried ALVs was significantly (p ≤ 0.05) lower for all ALVs compared to their cooked counterparts except in Ethiopian kale which had significantly (p ≤ 0.05) higher contents in the blanched-dried and spontaneously fermented-dried samples. Cooked spider plant had the highest aluminium content (75.8 mg/100 g, db) compared to other cooked ALVs. Condensed tannins were significantly (p ≤ 0.05) lower in spontaneously fermented-dried spider plant (14%), wild jute (40%), cowpea (58%) and amaranth (46%) than in their cooked counterparts. Significantly (p ≤ 0.05) higher levels of condensed tannins were found in spontaneously fermented-dried Ethiopian kale (52%) than its cooked counterpart. Significantly (p ≤ 0.05) lower contents of condensed tannin contents were found in blanched-dried spider plant (19%) and wild jute (47%) compared to their cooked counterparts. In comparison, significantly (p ≤ 0.05) higher contents were found in blanched-dried cowpea (27%) than in its cooked counterpart. Significantly higher (p ≤ 0.05) total phenolic content (TPC) was found in spontaneously fermented-dried cowpea (29%) and spider plant (48%), than in their cooked counterparts while no significant differences (p > 0.05) were found in TPC in all blanched-dried ALVs compared to their cooked counterparts except for wild jute with significantly (p ≤ 0.05) lower contents. Significantly lower (p ≤ 0.05) total oxalate contents were found in all spontaneously fermented-dried and blanched-dried ALVs compared to their cooked counterparts. All blanched-dried ALVs also had significantly (p ≤ 0.05) lower soluble oxalates contents compared to their

cooked counterparts. Phytate contents were significantly (p ≤ 0.05) lower in spontaneously fermented-dried amaranth (22%) and spider plant (25%) compared to their cooked counterparts. There were no significant (p > 0.05) differences found between phytate contents of all blanched-dried ALVs and their cooked counterparts except for spider plant which had 24% lower phytate content than its cooked counterpart. The Principal Component Analysis (PCA) graph showed that total oxalates, soluble salts, and magnesium were positively correlated. Moreover, there was a strong positive correlation between phytate and magnesium contents. Iron, zinc, and aluminium were also closely associated. The percentage of iron bioaccessibility from cooked, blanched-dried and spontaneously fermented-dried ALVs ranged from 1.6 to 15.3%, 1.8 to 7.8% and 1.5 to 11.5%, respectively. The amount of iron bioaccessible from processed ALVs ranged from 0.47 to 3.20 mg/100 g, db, 0.38 to 2.06 mg/100 g, db and 0.37 to 3.81 mg/100 g, db respectively. Significantly (p ≤ 0.05) higher percentages of iron bioaccessibility were found in spontaneously fermented-dried Ethiopian kale compared to its cooked counterpart. Compared to cooked counterparts, percentage iron bioaccessibility and amount of iron bioaccessible in all the blanched-dried ALVs showed no significant (p > 0.05) differences except for Ethiopian kale. Significantly (p ≤ 0.05) lower amounts of bioaccessible iron were found in blanched-dried amaranth compared to its cooked counterpart. The percentage zinc bioaccessibility from cooked, blanched-dried and spontaneously fermented-dried ALVs ranged from 8.4 to 40.9%, 16.6 to 25.1% and 12.2 to 30.8%, respectively. The amount of zinc bioaccessible from cooked, blanched-dried and spontaneously fermented-dried ALVs ranged from 0.47 to 3.20 mg/100 g, db, 0.38 to 2.06 mg/100 g, db and 0.37 to 3.81 mg/100 g, db. This study found no significant (p > 0.05) differences in percentage zinc bioaccessibility between the same type of cooked and spontaneously fermented-dried ALVs while significantly (p ≤ 0.05) higher amount of bioaccessible zinc was observed in blanched-dried amaranth (14.5%) than in its cooked counterpart. The percentage iron contribution estimation that could be made by 300 g (as consumed) of ALV cooked, blanched-dried and spontaneously fermented-dried towards iron absolute requirements for women aged 19 to 50 years were 14.5% to 50.3%, 6.1% to 40.6% and 3.9% to 48.1%, respectively. The estimated zinc contributions cooked, blanched, and spontaneously fermented-dried ALVs could make towards zinc absolute requirements for women aged 19 to 50 years were 2.9% to 34.9%, 8.5% to 51.9% and 6.0% to 57.2%, respectively. Amaranth has the potential to contribute over 40% (0.58 mg/day) towards AR for iron irrespective of processing. About zinc, the spider plant is estimated to contribute between 34.9 to 57.1% (0.61 to 1.0 mg/day) towards zinc absolute requirements for women aged 19 to 50 years. Still, these contributions are just relative

not actual values as physiological situations in vivo are not similar to in vitro situations; hence the need for human intervention studies ascertain iron and zinc bioavailability from these ALVs. Ferritin formation in Caco-2 cells exposed to simulated gastrointestinal digestates of differently processed ALVs ranged from 0.31 ng ferritin/mg cell protein to 2.35 ng ferritin/mg cell protein. Compared to cooked ALVs, the amount of ferritin formation in cells exposed to spontaneously fermented-dried and blanched-dried ALVs were not significantly (p > 0.05) different. Concerning zinc, cooked amaranth (31.2%) had significantly (p ≤ 0.05) higher percentages of zinc retained by Caco-2 cells compared to blanched-dried amaranth (3.1%). At the same time, no significant differences were found between differently processed spider plants. Zinc transport percentages in Caco-2 cells varied from 1.8 to 15.8%. Zinc uptake by the Caco-2 cells (calculated as the sum of cell retention and transport) was significantly (p ≤ 0.05) lower from blanched-dried amaranth (approximately five times less) compared to cooked counterpart. The amount of zinc uptake in both spontaneously fermented-dried amaranth and spider plant compared to their cooked counterparts was not significant (p > 0.05). Conclusions - ALVs had high levels of iron and zinc, so their consumption could be beneficial to rural communities with limited access to food and that are at risk of iron and zinc deficiencies. Spontaneously fermented-dried ALVs had similar levels of iron, zinc, and phosphorus as cooked vegetables and suggest that processing ALVs by fermentation could be as good as freshly cooked and therefore, consumption of spontaneously fermented-dried ALVs can be encouraged. Blanching and spontaneous fermentation further removed some of the soil contamination from the ALVs which remained after regular cleaning as shown by low levels of aluminium contents. Though blanched-dried and spontaneously fermented-dried ALVs had lower contents of antinutrients it did not translate to enhanced iron and zinc bioaccessibility compared to their cooked counterparts; hence other ALV components probably influenced mineral bioaccessibility. The uptake of iron and zinc by Caco-2 cells from ALVs is seriously impaired by the presence of high levels of antinutrients. Most antinutrients are dose-dependent, hence levels which were present after processing still had some inhibitory effects on iron and zinc uptake. Consumption of ALVs could be encouraged concurrently with other locally available foods which are rich sources of enhancers of iron and zinc (tomatoes, oranges and Mopani worms) to influence iron and zinc absorption. There is a need to deepen knowledge on the bioavailability of iron and zinc from ALVs and the magnitude of effects of antinutrients on iron and zinc bioavailability mostly when ALVs are consumed concurrently with cereal staples. The findings of the current in vitro studies can only be used to predict the direction of iron and zinc bioavailability from processed ALVs but not

the actual amount of mineral available for absorption, as in vitro models cannot simulate all in