Isolation and Characterisation of Quercitrin as a Potent Anti-Sickle Cell Anaemia Agent from Alchornea cordifolia

Alchornea cordifolia Müll. Arg. (commonly known as Christmas Bush) has been used traditionally in Africa to treat sickle cell anaemia (a recessive disease, arising from the S haemoglobin (Hb) allele), but the active compounds are yet to be identified. Herein, we describe the use of sequential fractionation coupled with in vitro anti-sickling assays to purify the active component. Sickling was induced in HbSS genotype blood samples using sodium metabisulphite (Na2S2O5) or through incubation in 100% N2. Methanol extracts of A. cordifolia leaves and its sub-fractions showed >70% suppression of HbSS erythrocyte sickling. The purified compound demonstrated a 87.2 ± 2.39% significant anti-sickling activity and 93.1 ± 2.69% erythrocyte sickling-inhibition at 0.4 mg/mL. Nuclear magnetic resonance (NMR) spectra and high-resolution mass spectroscopy identified it as quercitrin (quercetin 3-rhamnoside). Purified quercitrin also inhibited the polymerisation of isolated HbS and stabilized sickle erythrocytes membranes. Metabolomic comparisons of blood samples using flow-infusion electrospray-high resolution mass spectrometry indicated that quercitrin could convert HbSS erythrocyte metabolomes to be like HbAA. Sickling was associated with changes in antioxidants, anaerobic bioenergy, and arachidonic acid metabolism, all of which were reversed by quercitrin. The findings described could inform efforts directed to the development of an anti-sickling drug or quality control assessments of A. cordifolia preparations.


Introduction
Sickle cell anaemia (SCA) is an autosomal recessive genetic blood disorder arising from the S allele of haemoglobin (Hb). SCA is prevalent in the tropics and especially in sub-Saharan Africa, where the S allele may confer some tolerance to malaria. The World Health Organization (WHO) estimates that 300,000 children are born with SCA annually, 75% of whom are in sub-Saharan Africa [1].
SCA arises from a single amino acid substitution of glutamic acid with hydrophobic valine in the Hb β-globin chain. This results in an altered haemoglobin tetramer (α 2 β 2 ), haemoglobin S (HbS) [2]. HbS polymerises under hypoxic conditions, due to hydrophobic interactions between β6Val on two deoxy-HbS molecules [3]. The polymer of helical fibers lengthen and stiffen to cause the characteristic sickle shape of HbSS erythrocytes [4]. Polymerisation is linked to a dysregulation of cation homeostasis resulting from the activation of some ion channels, particularly the K + /Cl − co-transport system and the Ca 2+ dependent K + channel (Gardos channel). Ca 2+ activation of the Gardos channel increases H 2 O and K + efflux, leading to the dehydration of sickled erythrocytes [5]. Haemoglobin is denatured to

Blood Sampling
HbSS blood samples from a single clinically diagnosed SCA sufferer (the first author (female, 32 years old), subject to informed consent) was used to evaluate the anti-sickling activities of the plant extracts. Samples of 5 mL of blood were extracted using a lavender topped vacutainer (BD Vacutainer tubes, ISS Ltd., Bradford, UK), which uses dipotassium/tripotassium salts of EDTA as an anticoagulant. Control HbAA samples were taken from a single volunteer (the corresponding author (male, 55 years old), subject to informed consent). Separate blood samples were taken for each experiment presented. The whole blood (4 mL) was centrifuged at 2800× g for 10 min at 4 • C to sediment the erythrocytes. The plasma supernatant was removed, and erythrocytes underwent sequential centrifugations (2800× g for 10 min at 4 • C) involving washing three times in phosphate buffered saline (PBS) at pH 7.4. The cells were finally resuspended in 5 volumes PBS to 1 packed volume of erythrocytes and were used immediately.

Extraction of A. cordifolia Leaves
Pulverized air-dried leaves of A. cordifolia (1 kg) were extracted 12.5 L dichloromethane (CH 2 Cl 2 ; DCM), and then sequentially with 51.5 L of 75% methanol (75% MeOH; ALM (Alchornea methanol extract)), and two rounds of 10 L of sterile deionised water (H 2 O; "Aqueous", "Aqueous2") at room temperature with constant stirring. Each extraction occurred for a 72-h period. After filtration, the extracts were concentrated under reduced pressure at 40 • C and were stored at −20 • C until further use. The aqueous extract was freeze-dried.
Following the anti-sickling assays, ALM7T was further separated by preparative HPLC using a C18 3.5 m, 4.6 × 50 mm column (Waters, Borehamwood, UK). The eluting gradient was as follows: 90% water and 10% acetonitrile for 2.5 min, then 100% acetonitrile at 8.5 min and continuing at 100% until 13 min. There was 0.01% trifluoroacetic acid present throughout and the flow rate was 1.5 mL/min. This yielded eight peaks, separated into single fractions, ALM7T1-ALM7T8.

Ultra High-Performance Liquid Chromatography-High Resolution Mass Spectrometry (UHPLC-HRMS)
The fractions were analysed on an Exactive Orbitrap (Thermo Fisher Scientific, Waltham, MA, USA) mass spectrometer, which was coupled to an Accela Ultra High-Performance Liquid Chromatography (UHPLC) system (Thermo Fisher Scientific). Chromatographic separation was performed on a reverse phase (RP) Hypersil Gold C18 1.9 µm, 2.1 × 150 mm column (Thermo Scientific) using H 2 O using 0.1% formic acid (v/v, pH 2.74) as the mobile phase solvent A and ACN/isopropanol (10:90) with 10 mM ammonium acetate as mobile phase solvent B. Each sample (10 µL) was analysed using a 0-20% gradient of B from 0.5 to 1.5 min, and then to 100% in 10.5 min. After 3 min of being isocratic at 100% B, the column was re-equilibrating with 100% A for 7 min.

Nuclear Magnetic Resonance
NMR spectra were obtained using a Bruker Ultra shield-500 NMR spectrophotometer ( 1 H-NMR 500 MHz, 13 C-NMR 100 MHz) using MeOD as the solvent reference.

In Vitro Sickling-Inhibition and Reversibility Assays
The sickling-inhibition assay consisted of 100 μL of HbSS erythrocytes, 100 μL of PBS, and 100 μL of the test extract, and was incubated at 37 °C for 2 h. To induce sickling, freshly prepared 2% (w/v) Na2S2O5 solution (300 μL) was incubated with the cells for an additional hour in a water bath at 37 °C . The cells were then fixed with 3 mL of 5% (w/v) buffered formalin solution. A total of 10 µ L of the incubated cells were transferred to a haemocytometer and five fields were counted on each slide using a Leica ATC 2000 Binocular Phase Contrast Microscope at 40× magnification. The cells were classified as either normal or sickled. Each assay was repeated five times to generate the data presented.
For the reversibility assays, the cells were prepared as above and were incubated with 2% (w/v) Na2S2O5 at 37 °C for 1 h. Then, 100 μL of each sample was added and incubated at 37 °C for an additional 4 h period. The cells were fixed and mounted on a haemocytometer and were counted as described earlier.

Erythrocyte Leakage Assay
Sterile microcentrifuge tubes were each filled with 1 mL HbSS erythrocytes and centrifuged at 2800× g for 10 min at 4 °C . The supernatant was discarded, and the erythrocytes were washed three times with phosphate buffered saline (PBS; 0.01 M pH 7.4) and resuspended in 4% v/v PBS. Samples of 90 μL were added to the wells on a 96-well plate. Then, 10 µ L of the fraction or chemical composition tested (at 10× final concentration) was added to wells of the first row. These were then serially (1 in 10) diluted down to row 6. Row 7 contained erythrocyte suspensions with 0.1% Triton-X 100 (Sigma-Aldrich UK) as a positive control and row 8 contained only erythrocytes as the negative control. The plate was

In Vitro Sickling-Inhibition and Reversibility Assays
The sickling-inhibition assay consisted of 100 µL of HbSS erythrocytes, 100 µL of PBS, and 100 µL of the test extract, and was incubated at 37 • C for 2 h. To induce sickling, freshly prepared 2% (w/v) Na 2 S 2 O 5 solution (300 µL) was incubated with the cells for an additional hour in a water bath at 37 • C. The cells were then fixed with 3 mL of 5% (w/v) buffered formalin solution. A total of 10 µL of the incubated cells were transferred to a haemocytometer and five fields were counted on each slide using a Leica ATC 2000 Binocular Phase Contrast Microscope at 40× magnification. The cells were classified as either normal or sickled. Each assay was repeated five times to generate the data presented.
For the reversibility assays, the cells were prepared as above and were incubated with 2% (w/v) Na 2 S 2 O 5 at 37 • C for 1 h. Then, 100 µL of each sample was added and incubated at 37 • C for an additional 4 h period. The cells were fixed and mounted on a haemocytometer and were counted as described earlier.

Erythrocyte Leakage Assay
Sterile microcentrifuge tubes were each filled with 1 mL HbSS erythrocytes and centrifuged at 2800× g for 10 min at 4 • C. The supernatant was discarded, and the erythrocytes were washed three times with phosphate buffered saline (PBS; 0.01 M pH 7.4) and resuspended in 4% v/v PBS. Samples of 90 µL were added to the wells on a 96-well plate. Then, 10 µL of the fraction or chemical composition tested (at 10× final concentration) was added to wells of the first row. These were then serially (1 in 10) diluted down to row 6. Row 7 contained erythrocyte suspensions with 0.1% Triton-X 100 (Sigma-Aldrich UK) as a positive control and row 8 contained only erythrocytes as the negative control. The plate was incubated for 1 h at 37 • C. Following a 5 min centrifugation at 2800× g at 4 • C, 70 µL of supernatant from each well was transferred to a transparent, flat bottom 96-well plate. Changes in absorbance (OD 450 nm), indicating haemoglobin leakage, were measured using the Hidex Sense Plate Reader (LabLogic, Sheffield, UK). This absorbance was used to calculate the percentage haemolysis (0.1% triton X-100 = 100% and the negative control = 0%). Experiments were performed in quadruplicate.

Hb Polymerisation Assay
Hb polymerisation assays [34] were adapted for 96-well plates to assess polymerized Hb SS turbidity. A haemolysate was prepared by adding 2 mL of ice-cold distilled water to packed, washed erythrocytes, and then the cellular debris was pelleted by centrifugation at 6000× g for 20 min at 4 • C. Then, 220 µL of 2% Na 2 S 2 O 5 , 20 µL of the test compound (at five different concentrations and using PBS as the control), and 50 µL HbSS haemolysate (1:5 v/v dilution in PBS) were added into a 96-well plate and incubated. The 96-well plate was shaken and the absorbance at 700 nm was taken in 30 s intervals for period of 20 min (Hidex sense microplate reader, LabLogic, UK). The tests were carried out in quadruplicate.

Scanning Electron Microscopy (SEM)
Erythrocytes were fixed in 2% glutaraldehyde PBS for 30 min, and were rinsed three times in a 0.075 M sodium phosphate buffer (pH 7.4). The samples were then incubated with 2% OsO 4 in PBS (pH 7.4) for 2 h at 4 • C and then rinsed thrice in a 0.075 M sodium phosphate buffer (pH 7.4). Subsequently, the sample was dehydrated in 30%, 50%, 70%, 90%, and finally, underwent three changes of 100% ethanol. For SEM, 200 µL of the samples in hexamethyldisilazane (HMDS) were air dried on coverslips, coated with carbon, and imaged using a Zeiss Ultra plus FEG SEM.

Metabolomic Analyses
Samples of 500 µL washed HbSS erythrocytes, 500 µL of PBS, and 500 µL of quercitrin (0.5 mg/mL in final volume) were mixed and incubated at 37 • C for 2 h. Sickling was induced through chronic deoxygenation, with the reaction mixture (1500 µL) in an anaerobic tube deoxygenated by gently bubbling 100% N 2 through in a 37 • C water bath for 12 h. After the incubation period, the morphology of the cells was confirmed using light microscopy.
Erythrocyte extractions were carried out using published protocols [35,36]. Extracts were transferred to a 2 mL microcentrifuge tube, and were dried using a SpeedVac at 4 • C. The pellets were then resuspended in 100 µL of 50% methanol, in a HPLC vial containing a 0.2 mL flat bottom micro insert for flow infusion electrospray ion high resolution mass spectrometry (FIE-HRMS) analysis. FIE-HRMS was performed with an Exactive HCD mass analyser equipped with an Accela UHPLC system (Thermo-Scientific, UK). Data acquisition for each individual sample was conducted in alternating positive and negative ionisation mode, over four scan ranges (15-110, 100-220, 210-510, and 500-1200 m/z) with an acquisition time of 5 min. Individual metabolite m/z values were normalised as a percentage of the total ion count for each sample. The derived data are provided in Table S1. Data were normalised to total ion count and log 10 -transformed. Metabolites and pathway identification were performed by the MetaboAnalyst 4.0 MS peaks to pathway algorithm [37] (tolerance = 5 ppm, reference library: Homo sapiens). This involved metabolites being annotated using the KEGG database, considering the following possible adducts: For each m/z, the annotation was made using a 5 ppm tolerance on their accurate mass and considering the different adducts formed for each metabolite.

Statistical Analysis
The statistical analyses used SPSS version 26.0 software and XLSTAT_2020.1.1.64525. One way analysis of variance (ANOVA) coupled with Tukey's post-hoc test were used to compare the data and to identify means with significant differences; p values of <0.05 were considered significant.

Sample Extraction and Anti-Sickling Activity of A. cordifolia Crude Extracts
Sequential extractions using DCM, 75% MeOH, and 100% water from the dried leaves derived samples of 2.47, 9.65, and 8.28% dry weight. The sickling-inhibitory activities of A. cordifolia leaf extracts were compared at 1 mg/mL in PBS on a haemocytometer ( Figures S2A-C and 2A,B). The MeOH extract exhibited a significantly (p < 0.01) higher sickling-inhibitory activity (91.4%) than any other extract (Figure 2A,B). MeOH extract also exhibited a better sickling reversibility activity than any other extract ( Figure 2C). The cytotoxicity of aqueous and MeOH leaf extracts of A. cordifolia on HbSS erythrocytes was evaluated using an erythrocyte leakage assay ( Figure S3). Some haemolysis was observed in both the MeOH and aqueous extracts at 10 mg/mL, but this was only < 2% at 1 mg/mL. considered significant.

Sample Extraction and Anti-Sickling Activity of A. cordifolia Crude Extracts
Sequential extractions using DCM, 75% MeOH, and 100% water from the dried leaves derived samples of 2.47, 9.65, and 8.28% dry weight. The sickling-inhibitory activities of A. cordifolia leaf extracts were compared at 1 mg/mL in PBS on a haemocytometer ( Figures S2A-C and 2A,B). The MeOH extract exhibited a significantly (p < 0.01) higher sickling-inhibitory activity (91.4%) than any other extract (Figure 2A,B). MeOH extract also exhibited a better sickling reversibility activity than any other extract ( Figure 2C). The cytotoxicity of aqueous and MeOH leaf extracts of A. cordifolia on HbSS erythrocytes was evaluated using an erythrocyte leakage assay ( Figure S3). Some haemolysis was observed in both the MeOH and aqueous extracts at 10 mg/mL, but this was only < 2% at 1 mg/mL.
Sub-fraction ALM7T was further separated by preparative-HPLC into individual compounds to yield ALM7T1-ALM7T8. Screening for anti-HbSS erythrocyte sickling
Sub-fraction ALM7T was further separated by preparative-HPLC into individual compounds to yield ALM7T1-ALM7T8. Screening for anti-HbSS erythrocyte sickling properties indicated that ALM7T5 showed the best sickling-inhibition activity (Figure 3). Representative haemocytometer images can be seen in Figure S2. NMR was used to identify the only chemical detected within these peaks. The chemical in ALM7T5 was identified as quercitrin (quercetin-3-rhamnoside), a flavonol glycoside, based on high resolution LC-MS and NMR data. at 0.05 mg/mL (Figure 4).
Different concentrations of quercitrin (0.25-4 mg/mL) were tested for their ability to inhibit the polymerisation of deoxygenated HbS over 20 min ( Figure 5). Deoxy HbSS and HbAA haemolysates (with Na 2 S 2 O 5 and phosphate buffers) were used as the controls. Quercitrin prevented HbS polymerisation over all of the tested concentrations so that there was no significant difference in HbAA results ( Figure 5A). A comparison was also made with PHBA, known to inhibit sickling in erythrocytes [20,40] (Figure 5B). PHBA showed a similar ability to suppress the exhibited HbSS polymerization, but at 0.5 and 0.25 mg/mL, suppression was significantly (p > 0.05) less effective than quercitrin ( Figure 5B). Quercetin exhibited no ability to prevent the polymerisation of HbSS, with no significant difference to the positive control ( Figure S7). Taken together, these data indicate that quercitrin is a potent inhibitor of in vitro HbS polymerisation and this is most likely to represent its main mode of action in inhibiting and reversing HbSS-erythrocyte sickling. HbS not subjected to deoxygenation. DeoxyHbA (HbAA) did not polymerise and represents a negative control. This represents data obtained from three typical independent experiments performed in quadruples.

Assessing the Impact on Quercitrin on HbSS Sickling Using Metabolomics
In using metabolomics to assess quercitrin's mode of action, it was not possible to use 2% Na2S2O3 to induce the sickling phenotype, as this would dominate the subsequent metabolite profile. Thus, an alternative approach was employed where anoxic conditions were induced through the replacement of air with 100% N2. This was proven to be a highly successful approach, as demonstrated using SEM ( Figure S8). Imposition of N2-induced anoxia led to sickling in HbSS erythrocytes, but not HbAA erythrocytes (data not shown). HbSS erythrocytes treated with 0.5 mg/mL quercitrin exhibited morphologies that were similar but not identical to the HbAA erythrocytes. Interestingly, erythrocytes treated with 0.5 mg/mL PHBA did all exhibit the normal discoid phenotype, with some sickled cells observed.
In the metabolomic treatments, HbSS erythrocytes were either (1) maintained under

Assessing the Impact on Quercitrin on HbSS Sickling Using Metabolomics
In using metabolomics to assess quercitrin's mode of action, it was not possible to use 2% Na 2 S 2 O 3 to induce the sickling phenotype, as this would dominate the subsequent metabolite profile. Thus, an alternative approach was employed where anoxic conditions were induced through the replacement of air with 100% N 2 . This was proven to be a highly successful approach, as demonstrated using SEM ( Figure S8). Imposition of N 2 -induced anoxia led to sickling in HbSS erythrocytes, but not HbAA erythrocytes (data not shown). HbSS erythrocytes treated with 0.5 mg/mL quercitrin exhibited morphologies that were similar but not identical to the HbAA erythrocytes. Interestingly, erythrocytes treated with 0.5 mg/mL PHBA did all exhibit the normal discoid phenotype, with some sickled cells observed.
In the metabolomic treatments, HbSS erythrocytes were either (1) maintained under normoxic conditions (SS), (2) deoxygenated with N 2 (SS-N) only, or (3) incubated with quercitrin followed by deoxygenation with N 2 (SS-Q). Controls consisted of HbAA erythrocytes (AA) with no quercitrin treatment ( Figure 6). Principal component analysis (PCA) indicated that the different experimental classes formed two clusters; one broadly associated with sickled cells the other with non-sickled cells ( Figure 6A). SS and SS-N samples were both closely clustered across principal component 1 (PC1), which describes the major source of variation. This suggested pre-existing metabolomic changes in the HbSS erythocytes, even under normoxia. However, by adding quercitrin, the hypoxic HbSS metabolomes shifted so that the samples clustered with the HbAA group.
The sources of variation between sickled and non-sickled erythrocyte samples were determined ( Figure 6B). We annotated metabolites using the KEGG database and their relative abundances were discriminated between the sickled and non-sickled phenotypes, as shown using a heatmap. The sickled group showed relative increases in metabolites such as arachidonic, stearic, myristic, and linolenic acids, which were suggestive of lipid processing. The sickled cells also appeared to be relatively deficient in glucose and fructose. These effects were all reversed by quercitrin treatment. To provide functional information for these differences, biochemical enrichment analyses were conducted. The over-representation analysis (ORA) method was used to evaluate pathway-level enrichment based on significant features (p-value is measured with Fisher's exact test) [41], combined with gene set enrichment analysis (GSEA), which extracts biological meaning from a ranked metabolite list [42]. A total of six metabolic pathways were identified to be significantly enriched (Table S2).
HbAA (AA) and HbSS (SS) erythrocytes were maintained under normoxic conditions or under N 2 (N), and, in some cases, treated with quercitrin (0.5 mg/mL) (SS-Q) imposition of hypoxic conditions. (A) PCA of the derived metabolite profiles for each treatment (note the separation of the samples into two main clusters, reflecting sickled and non-sickled groups) (B) heatmap based on significant metabolite differences between the non-sickled (cluster 1) and sickled (cluster 2) groups These suggested that the metabolomic switches in the erythrocyte between the sickled and non-sickled states apparently involves redox changes (ascorbate), thiol metabolism (cysteine), fatty acid processing, and haem metabolism (porphyrin).

Discussion
The high prevalence of SCA in the developing nations of West Africa is driving the requirement for cost-effective means to treat the disease. In Sub-Saharan Africa, medicinal plants are used widely to manage SCA, although relatively few have been validated scientifically. In the case of A. cordifolia, an aqueous extract showed an anti-sickling activity [43,44], but the bioactives had not been previously defined. In this study, we followed a bioactivity guided fractionation and purification strategy to define quercitrin as the main active anti-sickling agent in A. cordifolia.
Many anti-sickling drug leads effect the Hb gene [54] or the HbS molecule, or are erythrocyte membrane modifiers [55]. Our methods did not assess Hb gene modification, but we have provided clear evidence that quercitrin affected the HbS protein to suppress polymerisation. This was important, as clinically, the delay of HbS polymerization during the transit of erythrocytes through post-capillary venules is necessary for SCA disease remediation [56]. Our assessments of extracted Hb polymerisation showed the expected results for the controls, i.e., no polymerisation with HbA (HbAA), oxygenated HbS (HbSS), but deoxyHbS did polymerise. However, with quercitrin, deoxyHb polymerisation was suppressed in a concentration-dependent manner, comparable to PHBA. Crucially, a recent study used multi-spectroscopic and molecular simulation techniques and showed that isoquercitrin had an anti-sickling activity through a direct interaction with haemoglobin molecules [57]. Thus, quercitrin could act by stereospecific covalent or non-covalent attachment to the HbS molecule [58]. Additionally, quercitrin was able to partially reverse erythrocytes sickling after a two-hour incubation period, suggesting quercitrin was able to modify already polymerized HbS and thus have some ability to reverse polymerisation. This partial effect may reflect the rate at which quercitrin is transported across the membrane.
The pharmacological activities of a given bioactive compound are dependent on its chemical structure [59]. Previous investigations have credited the biological activity of quercitrin to the aglycone, quercetin [38,39], but the possibility of the glycosidic residue being crucial for its effects as the glycoside activity has also been suggested [39,60]. The quercetin aglycone or glucoside is not found in human plasma, however conjugates, such as quercetin-3-glucuronide, quercetin-3 -sulfate, and isorhamnetin-3-glucuronide, have been found [61]. It is generally thought that flavonoid glycosides such as quercitrin, enter the colon and are hydrolysed to the aglycone by quercitrinase found in Enterobacteria [62]. The aglycone is then absorbed in the large intestine easily because of its lipophilicity, and is then metabolized in the liver by O-methylation, glucuronidation, and/or sulfation [63]. We assessed if quercetin could be an active anti-sickling agent, as suggested by Muhammad et al. [64]. However, we failed to demonstrate any appreciable anti-sickling activities for quercetin, thus the substitution of an alpha-L-rhamnosyl moiety at position 3 via a glycosidic linkage in quercitrin would be important for its biological activity. Confirmation of this mode of action could focus on use of X-ray crystallography of Hb-quercitrin to demonstrate this.

Insights into the Mechanism of Quercitrin Effects Using Metabolomics Studies
To provide a wider appreciation of the effects of quercitrin, we applied an omic approach based on metabolite detection using direct infusion-high resolution mass spectroscopy. Multivariate statistical assessments of the derived data provided a series of important observations. Firstly, it was apparent that even under normoxia, the metabolomes of HbSS erythrocytes were unlike HbAA cells, and, indeed, were more like anoxic cells. This could suggest that even under normoxia, HbSS cells are poised to sickle and, indeed, the SEM of suggested cellular irregularities. However, the addition of quercitrin caused a shift in the HbSS metabolome, so that it became like HbAA. This allowed us to assess the metabolomic difference between sickled and non-sickled erythrocytes that quercitrin was effectively correcting.
Another altered pathway involves bioenergetic metabolism. Glycerolipid pathways include monoacylglycerols (MAGs), diacylglycerols (DAGs), triacylglycerols (TAGs), phosphatidic acids (PAs), and lysophosphatidic acids (LPAs) with functions in energy generation [67]. The hexose monophosphate shunt, which is parallel to the glycolytic pathway, generates NADPH and pentoses, as well as ribose 5-phosphate, a nucleotides-synthesis precursor [68]. Increased D-glucose 1-phosphate, when interconverted to D-glucose 6-phosphate, will also feed through to the glycolytic pathway involved in the generation of ATP in the anabolic generation of intracellular energy [69]. This could indicate that the sickle cells could be exhibiting a tendency towards anaerobic respiration. Reduced glucose 1-phosphate levels in quercitrin-treated samples suggest that the anaerobic shifts are countered by quercitrin.
Sickle cells generate approximately two times more reactive oxygen species compared with normal red blood cells [70], and this is linked to endothelial dysfunction, inflammation, and multiple organ damage [71,72]. Decreased intravascular sickling has been linked with reduced oxidative stress and also increased nitric oxide bioavailability [73]. Such oxidative effects with sickling were suggested from the metabolomic analysis, as "ascorbate and aldarate" metabolism pathways were being affected in our experiments. It should be noted that glutathione, detected in our metabolome, also plays an important role in the antioxidative process. Glutathione protects the red cells from oxidative damage, denaturation of haemoglobin, the formation of Heinz bodies, reduced cell deformability, and intravascular haemolysis [74].

Conclusions
In this work, quercitrin was isolated from A. cordifolia and its anti-sickling activity was characterized. Quercitrin inhibits and marginally reverses HbSS erythrocytes under hypoxic conditions in vitro. Experimental evidence was presented to support its action against two features of SCA. Most importantly, quercitrin inhibited Hb polymerisation, but it also stabilized the HbSS-erythrocytes membrane to reduce erythrocytes' fragility. Metabolomics was used to provide a wider description of the sickling process and the effects of quercitrin. All of these were reversed with quercitrin treatment.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/jcm11082177/s1. Figure S1. Schematic diagram for the purification of anti-sickling activities in Alchornea spp. Figure S2. The effect of Alchornea spp. extracts and fractions on erythrocyte-sickling in Na 2 S 2 O 5 -induced hypoxia. Figure S3. Percentage haemolysis in HbSS RBCs incubated with various concentrations of MeOH and aqueous extracts of Alchornea spp. leaves compared with 0.1% Triton X100. Figure S4. The effect of ALM fractions (1 mg/mL) on erythrocytes-sickling under Na 2 S 2 O 5 -induced hypoxic condition. Figure S5. The effect of ALM fractions (1 mg/mL) on erythrocytes-sickling under Na 2 S 2 O 5 -induced hypoxic conditions. Figure S6, Effect of ALM7 fractions (500-62.5 µg/mL) on erythrocytes-sickling under Na 2 S 2 O 5 -induced hypoxic condition. Figure S7. Inhibitory effects of quercetin on HbSS erythrocyte sickling. Figure S8. Assessing the potential inhibitory effects of quercetin on HbS Polymerisation. Figure S9. The effect of Alchornea spp. extracts and fractions on erythrocyte-sickling in N 2 -induced hypoxia. Table S1. Metabolome data. Table S2. Pathway analysis of m/z features significantly (p < 0.05) different between the treated and untreated groups.