Possible Mimics of Duffy Binding Protein-II for Plasmodium vivax Binding Endothelial Cells or Binding Plasmodium falciparum by Mimicking Epitope on Erythrocyte Binding Antigen-175 A

Two molecules from the same source, Κ casein, are suggested as treatments to prevent infection by Plasmodium vivax, malaria virus, through the prevention of Duffy Binding Protein II (DBPII) monomers 1 and 2 from binding. By preventing DBPII monomer 1 and 2 binding there would be half the Duffy Antigen Receptor for Chemokines (DARC) binding to the DBPII trimer. This may prevent infection by this virus. Here the essential, structural, characterization of two potential binding mimics of the DBPII monomer 2 to prevent its binding to DBPII monomer 1, are provided. Κ casein is treated with NaBH4 in 1 N NH4OH (pH 11.4) to produce the two molecules; N-acetamido deoxy neuraminyl (α 2->3’) N-acetamido deoxy neuraminyl (α 2->6’) galactosyl 4’ (di-hydrido) sulfo (β 1->3) N-acetamido deoxy galactosaminyl 6 (di-hydrido) sulfo di(hydrido) di-phospho serinyl (di-hydrido) sulfo tyrosine dipeptide and N-acetamido deoxy neuraminyl (α 2->3’) N-acetamido deoxy neuraminyl (α 2->6’) 4’ (di-hydrido) sulfo galactosyl (β 1-> 3) 6 (d-hydrido) sulfo 1,5 anhydro N-acetamido deoxy galactosaminitol. Only the first of the two molecules presents a (di-hydrido) sulfo tyrosine, yet both have two (di-hydrido) sulfo groups. Still the dipeptide may mimic another sulfo tyrosine on the DBPII monomer 2, thought to be away from T266 (threonine 266) a significant distance, possibly Y363 of DBPII monomer 2. These molecules may also mimic the appropriate epitope on Erythrocyte Binding Antigen (EBA)-175 of Plasmodium falciparum to prevent infection by this virus.


Introduction
New treatments for malaria are necessary because of the expense of the treatment regimen required over the life-time of the disease etiology. Because of the cost of these treatments those who are poverty stricken cannot afford complete treatment and the disease gains strength and develops into drug resistant malaria. Clearly, lower cost medicines to fight against malaria are needed.
Here the target is the probable requirement of a sulfotyrosine of DBPII monomer 2, Y278 (tyrosine 278), and T266 (threonine 266) of DBPII monomer I, to be in close proximity to each other through hydrogen bonding. The DBPII monomers 1 and 2 are locked and ready for correct quaternary structure, of the hetero-trimer, in order to bind the second Duffy Antigen/Receptor of Chemokines (DARC) [1], [2] for infection by malaria virus.
Another target to prevent malaria infection is the Neu (α 2-> 3) Gal (β 1-> 4) epitope, sequences found in glycophorin A, which has been found on the malaria erythrocyte binding antigen on Plasmodium falciparum. [3] As a mimic of this epitope Κ casein oligosaccharide di-phospho serinyl sulfo tyrosine after treatment with NaBH 4 in 1N NH 4 OH of the whole glycoprotein from Κ casein is proposed.
The authors [1,2] do their calculations without sulfate ester substitution of tyrosine but this does not preclude the need for sulfate because there is a known antibody to sulfo-tyrosine substituted DBP, [4] available commercially (Millipore-Sigma), which prevents DBP-II:DARC binding.
Another tyrosine is not far away in the monomer 2 DBPII binding cleft, Y (tyrosine) 271. [1,2] These authors, also found that exchanging Y363 (tyrosine 363) for D (aspartic acid) in monomer 1 results in complete loss of DARC binding. This tyrosine, Y363, may be an effective candidate for mimicry by the molecules from Κ casein isolate, the Κ casein having been treated with NaBH 4 in pH 11.4 NH 4 OH.
Also, others [3,5] have discovered that the Neu (2->3) Gal (1->4) Glc epitope, from glycophorin A, is important in erythrocyte binding and infection by malaria virus. We hope to use the molecules originating from Κ casein to substitute. They could be similar in structure to the N-glycan of DARC, and, be important in DARC binding to DBPII monomer 1.
Κ casein, treated with NaBH 4 in pH 11.4 NH 4 OH, yields possible DBPII monomer 2 mimics and these mimics could prevent attachment of Plasmodium vivax to red blood cells and may prevent infection by this virus. Essential to this work is the structural characterization of these two molecules from Κ casein. [7] Workers have proposed partial structural characterization of Κ casein O-linked oligosaccharide using traditional Carlson β-elimination, NaBH 4 in 1N NaOH, presumably pH 14, for overnight or longer. Evidence for a more complete view of the structures of the two title molecules from novel chemistry using NaBH 4 in pH 11.4 NH 4 OH (1N) is shown here.
Because of the source of the title molecules, simply isolated from cow's milk, prevalent, even in the third world, the two components could be simply prepared from the milk isolate and used as anti-malaria medicines.

Preparation of Samples
A sample of Κ casein (< 0.300 g, Sigma-Aldrich, St. Louis, MO, USA) is treated with NaBH 4 (3µL, 4N solution) in NH 4 OH (1.oo mL, 2N, pH 11.4-11.6) for 8 hours at ambient temperature. The reaction mixture was pushed through an NH 4 + cation exchange cartridge (Thermo Fisher, Sunnyvale, CA USA). The NH 4 + cartridge is prepared from an H + cation exchange cartridge (ThermoFisher, Sunnyvale, CA, USA) by pushing 2N NH 4 OH (20mL). [8,9] The effluent was immediately frozen and only partially thawed for analysis. The O-linked oligosaccharide di-peptide and 1, 5 anhydro N-acetamido deoxy galactosaminitol derived tetrasaccharide doubly substituted with (di-hydrido) sulfo groups are in the same sample and are isolated as noted below.

Isolation of HPAEC-PAD of Components
A sample of the treated Κ casein components was taken and chromatographed and each component collected by retention time on an HPAEC-PAD using a CarboPac MA 1 column (ThermoFisher, Co., Sunnyvale, CA, USA) with a flow rate of 150 µL/ min. with a 310 mM NaOH isocratic elution for at least 190 min. A chromatogram, figure 1, and mass spectra, figures 2 and 3, were obtained and appear after the Results and Discussion section.  . This is the mass spectrum generated from isolating the smaller of two peaks from the MA1 chromatogram; K Casein oligosaccharide 1, 5 anhydro N acetamido deoxy galactosaminitol.

API Mass Spectrometry and MALDI-TOF Mass Spectrometry
The effluent from the HPAEC-PAD was pushed through an NH 4 + cartridge, prepared as noted above and the effluent was evaporated to no less than 200 µL, H 2 O (1.00 mL, 18 megaΩ resistivity) added to the vial and the solution pushed through a Na + cartridge (ThermoFisher, Sunnyvale, CA, USA) prepared by pushing NaOH (freshly prepared from 50% NaOH, 20 ml, 1N) through the H + cation exchange cartridge, ThermoFIsher, Sunnyvale, CA, USA, and then water was pushed through the Na + cartridge (using 18 megaΩ resistivity H 2 O, Milipore Co, MA, USA)). The effluents from the Κ casein isolates are immediately frozen. The samples should not in any way be concentrated to dryness. A sample of the partially frozen sample was also taken, both peaks (shorter and longer retention times) isolated, singly, for the mass spectrometer, from the HPAEC-PAD system and analyzed by API 2000 triple quadrapole ms as noted. [9,10] The samples for the API 2000 mass spectrometer were infused with 1% methanolic H 2 O. A MALDI mass spectrum, figure 4, was also obtained for the mixture of the oligosaccharide dipeptide and 1, 5 anhydro oligosaccharides.
By ms, from Κ-casein, noted above in the materials and methods section, a Gal-GalNAc and both Neu (α 2->3) and Neu (α 2->6) as monosaccharide components of the isolated dipeptide and 1, 5 anhydro alditol are supported. The dipeptide contains not only this derivatized, sulfated carbohydrate, but also, the isolated dipeptide includes a tyrosine substituted with derivatized sulfate.
An interpretation of our mass spectral data for Κ casein has been published. [10] It suggested a sulfo tyrosine substituent of a dipeptide linked to serine tetrasaccharide. The linkage between the tetrasaccharide and serine must not be a direct linkage because the Κ casein was treated with pH 11.4 NH 4 OH which is two to three orders of magnitude less base than the 1 N NaOH at a much lower time than with the standard Carlson β-elimination. The di-phospho linkage between Κ casein tetrasaccharide and serine sulfo tyrosine dipeptide are proposed here The 1, 5 anhydro molecule is the result of novel nucleophilic attack of hydride anion on the anomeric carbon, C-1, linked to (hydrido) phosphate group. These molecules may be helpful in preventing the binding of DBPII monomer 2 to DBPII monomer 1 or EBA-175 binding and prevent malaria virus progression.
The larger of the two molecules are linked to oligosaccharide to serine, and subsequently sulfo tyrosine, by di-(hydrido) di-phospho group. This linkage is suggested because the original Carlson β-elimination of O-linked oligosaccharides from peptides, using 1 N NaOH containing NaBH 4 with a presumed pH of approximately 14, requires, also, an overnight treatment. In our case we use pH 11.4 to pH 11.6. The amount of hydroxide used is nearly three orders of magnitude less than traditional Carlson β-elimination. This higher pH and more time is required because of a very high pKa, possibly higher than 13, for the abstraction of the serine α hydrogen which is required to initiate β-elimination.
In this protocol the molecule would remain intact, to include the di-phospho groups converted to the di-(hydrido) di-phospho groups and the sulfo groups converted to the (dihydrido) sulfo substituents, when using NaBH 4 in NH 4 OH. There would be minimal Carlson β-elimination at pH 11.6 or using pH 11.4 NaOH, without NaBH 4 , and shorter times, [14] except if the carbohydrate protein linkage is through a diphospho linkage, as in non-reductive β-elimination method. [14] Since the reaction conditions require an uncapped reaction vial, the pH will drop permitting hydrolysis of peptide linkages driven by the release of NH 3 to the atmosphere. This phenomenon is known to occur in hydrazine hydrate, pKa ~ 8, hydrolysis of glycan N acetamido groups from glycoproteins.
The second molecule found has two (di-hydrido) sulfo substitutions on the oligosaccharide with a 1, 5 anhydro Nacetamido deoxy galactosaminitol on the 'reducing end' of the oligosaccharide. We demonstrate this structure as shown by the cumulative support for ion structures for ten ions from the ms, figure 3, of the smaller peak, and longer retention time, from the HPAEC-PAD chromatography, a trace (blue) is shown in figure 1.
A MALDI spectrum in figure 4 which is from the mixture of substituted oligosaccharides from Κ casein shows two peaks one at m/z 1036.7596 and m/z 698.8596. Figures 5 and  13 describe pictorially the identities of two components of standard Κ casein after reaction under the new conditions.
The smaller of the two peaks, smaller peak in HPAEC-PAD as well ( figure 1, blue trace), is shown to have two (dihydrido) sulfo substituents at O-6 of N-acetamido deoxy 1, 5 anhydro galactosaminitol and O-4' of the galactose residue. It is also substituted at the O-3' and O-6' positions of the galactosyl residue with N-acetamido deoxy neuraminyl groups.
The chemistry for Hnucleophilic substitution of the hydrido phospho substituted anomeric carbon of Nacetamido galactose, C-1, is novel. We suggest a possible mechanism to involve steroelectronics of the O-5, C-1, O-1 and P-1 atoms of the whole di (di-hydrido) sulfo oligosaccharide di-(hydrido) di-phospho serinyl (di-hydrido) tyrosine dipeptide molecule. [15,16] An O-1 non-bonding electron orbital can be anti-periplanar to the P-1 to O double bond. Along with the H atom on the P atom, O-1 donates electron density preventing development of partial positive charge on the P-1 atom, δ + . This weakens the O-1 to C-1 bond. An O-5 non-bonding electron orbital is in an antiperiplanar alignment with the C-1 atom giving the O-5-C-1 bond, double bond character which may lead to an SN 1 nucleophilic Hsubstitution at C-1 and while breaking the C-1 to O-1 bond resulting in the 1, 5 anhydro alditol.
This would be also important as an inhibitor of DBPII monomer 1 and monomer 2 binding to prevent Plasmodium vivax infection. This molecule could help determine the importance of, remote to the Y 278 to T266 interaction, another sulfo group possible requirement for DBPII monomer 1 and monomer 2 binding and also the contribution of the 1, 5 anhydro glycan to the possible inhibition of malaria infection.
Claims for structural assignment by the ions drawn in the following figures 5-12 for the smaller peak and figures 13-20 for the larger peak are supported. The spectrum, figure 3, was generated by isolating the smaller of the two peaks in the HPAEC-PAD chromatogram as noted in the materials and methods section. All ten ions, except m/z 589.1, figure 9, can be drawn with the 1, 5 anhydro N-acetamido deoxy galactosaminitol as the 'reducing end' of the ion, giving credence to the action of this chemistry on Κ casein.  Two of the ions, m/z 465.3 and m/z 526.5, figures 7 and 8, are substituted with one (di-hydrido) sulfo group rather than two (di-hydrido) sulfo group substitutions on the 1, 5 anhydro glycan. This reflects fragmentation not another compound since we isolated this molecule in pure form by HPAEC-PAD preparative chromatography. For one ion, m/z 589.1, figure 9, one of 5acetamido deoxy neuraminyl groups has been fragmented away while the other remains. Yet it still retains the (di-hydrido) sulfo, though it is drawn hydrated.  An exception to the tetrasaccharide remaining intact in all ten ions, the ion, m/z 589.1, is fragmented at the C-3 to O-1' bond, noted in figure 9. The 4' hydrated (di-hydrido) sulfo, 6' N-acetamido deoxy neuraminyl substituted galactose is drawn to represent the ion. Three ions, m/z 498.5, m/z 426.3 and m/z 484.7, figures 10, 11 and 12, are drawn comprising 4' (di-hydrido) sulfo galactosyl (β 1-> 3) 2-acetamido 1, 5 anhydro galactosaminitol structure. The ion, figure 12, m/z 484.7 has these attributes; its (di-hydrido) sulfo group hydrated and ionized as a diammonium ion salt.    The larger of the two MALDI, figure 4, peaks, m/z 698.8596, drawn in figure 13 shows the whole molecule less one (dihydrido) sulfo group. The larger of the two peaks, isolated by chromatography via HPAEC-PAD, figure 1, was subjected to mass spectrometry on an API 2000 triple quadrapole mass spectrometer. The mass spectrum, API 2000 ms, of the isolated component, is shown in figure 2. The ion, m/z 719.1, figure 14, drawn, represents the total isolated molecule less three ketene molecules. The loss of ketene, CH 2 =C= O, is common for Nacetamido substituted carbohydrates.  In figure 17 the structure of an N-acetamido deoxy galactosaminyl di (hydrido) di-phospho (hydrated) serinyl (di-hydrido) sulfo tyrosine ion, m/z 335.0 is drawn. This ion helps to corroborate the dipeptide portion of the molecule isolated from NaBH 4 in NH 4 OH treated Κ casein. In the structure drawn to represent ion m/z 931.0, figure  18, the tetrasaccharide is shown less three ketenes yet intact as a fragment without the di (hydrido) di-phospho serinyl (dihydrido) sulfo tyrosine. The last two ions, m/z 506.9 and m/z 825.0, are drawn, figures 19 and 20, showing a trisaccharide, one 3' linked and one 6' linked, of an N-acetamido deoxy neuraminyl group with linkage to galactosyl N-acetamido deoxy galactosaminyl di (hydrido) di-phospho group. One di (hydrido) phospho group is hydrated, m/z 506.9, and one is dehydrated, m/z 825.0.   Methods are described [9][10][11] to isolate and analyze these derivatives from Κ casein.
A sulfo-tyrosine is asserted as a mediary for binding DARC to Duffy Binding Protein (DBPII). [17] attest to This hypothesis is attested to and this fact is proposed to be the title compounds, from Κ casein, as inhibitors of Plasmodium vivax binding of DARC and subsequent progression by this virus, malaria.
The connection between Duffy Binding Protein II (DBP-RII), monomers 1 and 2 is thought [1,2] mediated by the interaction between T (threonine) 266 and Y (tyrosine) 278, C in Figure 4. [1,2] This interaction, probably bridged by sulfated tyrosine, possibly locks the two binding proteins DBPII monomers 1 and 2 in the correct quaternary structure so that the second DARC molecule can bind to the DBPII trimer and then chemokines.
The di-hydrido derivative of the sulfated title compound may be stable to hydrolysis, due to the difference between the electro-negativities of a hydroxyl versus two hydrogens, making the sulfuryl group, OH substituted and not two H substituted, sulfur develop partial positive charge, δ + , compared to the substituents with two hydrogen substitutions, making the sulfuryl group of sulfate a better nucleofuge, in the case of hydroxyl substitution.

Conclusion
The use of the two molecules, described and structurally established above, may mimic the DBP-RII monomer 2 to provide the sulfo-tyrosine dipeptide oligosaccharide and 1, 5 anhydro tetrasaccharide, both from Κ casein, in order to prevent DBP-RII monomer 2 from binding the DBP-RII monomer 1, in Plasmodium vivax. If this quaternary structure is blocked by the these compounds, Plasmodium vivax cannot effectively bind its DBP-RII monomer 2, preventing Y 278 or Y271 or Y363 sulfate interaction with DBPII monomer 1, and preventing progression toward infection by the malaria virus. The 1, 5 anhydro tetrasaccharide, without the (di-hydrido) sulfo tyrosine epitope, could also bind to Plasmodium vivax because of a report of the presence of 5-Nacetamido neuraminyl-Gal epitope [6] on the N-glycans of recombinant DBPII.