Molecular Dynamics Study of Supramolecular Complexes to the Carbohydrate-Cation System

Dhaidan Khalaf kafi; Ali Taher Mohi

doi:10.12688/f1000research.173339.1

Home Browse Molecular Dynamics Study of Supramolecular Complexes to the Carbohydrate-Cation...

ALL Metrics

-

Views

-

Downloads

Get PDF

Get XML

Export

▬

✚

Research Article

Molecular Dynamics Study of Supramolecular Complexes to the Carbohydrate-Cation System

[version 1; peer review: 4 approved with reservations, 1 not approved]

Dhaidan Khalaf kafi ¹, Ali Taher Mohi²

PUBLISHED 10 Apr 2026

Author details Author details

¹ Medical Physics, University of Fallujah, Fallujah, Iraq, 41001, Iraq
² Department of Physics, Mustansiriyah University, Baghdad, Iraq, 10001, Iraq

Dhaidan Khalaf kafi
Roles: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Project Administration, Resources, Software, Supervision, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing

Ali Taher Mohi
Roles: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Project Administration, Resources, Software, Supervision, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing

OPEN PEER REVIEW

REVIEWER STATUS

This article is included in the Fallujah Multidisciplinary Science and Innovation gateway.

Abstract

Background

It is known that calcium ions lead to aggregation of disaccharides, while magnesium ions do not. In this study, simulation results indicated that the cations primarily bind to the sulfate groups of the disaccharides.

Methods

From molecular dynamics simulations, the differences between the calcium and magnesium cation systems can be explained. Additionally, when a modified water model is used in the case of the magnesium system, in which the charges of the water atoms are reduced, the magnesium ions bind the hydration shell less strongly than in the ‘normal’ water model, which facilitates complex formation with the disaccharides. To verify whether the stability differences between the two saccharides are also observed in the simulations, the simulations of the calcium system were repeated with the monosaccharide.

Results

Moreover, result in the formation of carbohydrate-cation-carbohydrate complexes, with the disaccharides primarily binding to the glucose unit, which is attached to the sulfate group. If the cation is not present, the sulfate groups repel each other and complex formation is not possible. The differences between the two cation systems can be explained as follows: Magnesium ions bind much weaker to the disaccharide than calcium ions. The reason is that magnesium ions bind more strongly to the hydration shell due to their smaller ion radius compared to calcium ions. This is energetically more expensive in the case of magnesium ions, which leads to a kinetic hindrance of complex formation compared to the calcium system. The other significant difference between the two cation complexes is that in the case of calcium ions, both disaccharides bind directly to the cation through the sulfate groups.

Conclusions

As expected, the binding energy of the monosaccharide complex was found to be lower than that of the disaccharide complex (6 kJ/mol). The difference in binding energies between the two complexes was relatively small. A significantly lower rupture force was found in the simulations for an individual complex.

Keywords

molecular dynamic; supramolecular complexes; carbohydrate-cation system; radial distribution function

Corresponding author: Dhaidan Khalaf kafi

Competing interests: No competing interests were disclosed.

Grant information: The author(s) declared that no grants were involved in supporting this work.

Copyright: © 2026 kafi DK and Mohi AT. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite: kafi DK and Mohi AT. Molecular Dynamics Study of Supramolecular Complexes to the Carbohydrate-Cation System [version 1; peer review: 4 approved with reservations, 1 not approved]. F1000Research 2026, 15:503 (https://doi.org/10.12688/f1000research.173339.1) First published: 10 Apr 2026, 15:503 (https://doi.org/10.12688/f1000research.173339.1) Latest published: 10 Apr 2026, 15:503 (https://doi.org/10.12688/f1000research.173339.1)

Introduction

Divalent cations frequently play a crucial role in stabilizing supramolecular complexes across various natural systems. These cations act as counterions, facilitating the proximity of negatively charged groups, a key factor in processes such as cell adhesion and aggregation. For instance, cadherins, a group of adhesion proteins, utilize calcium ions to bind together, forming short chains capable of adhering to cell membranes. This mechanism enables the interconnection and networking of multiple cells.^1,2 In other systems, the cations ensure that negatively charged groups can approach and react to one another. For instance, when building DNA, the phosphate group of one nucleotide is linked to the hydroxyl group of another nucleotide. Because these two groups normally repel each other due to negative charges, the reaction is catalyzed by DNA polymerase, which has two magnesium ions in its active center.³ Sea sponge cells self-adhere through a complex that stabilizes through ion pair bonds. This self-adhesion is mediated by aggregation factors, which consist of two proteoglycans. Proteoglycans are macromolecules that consist of a central protein to which one or several glycosaminoglycans are attached. One group of proteoglycans binds to receptors located on the cell surface and ensures that the aggregation factors are bound to the individual cells. The actual adhesion between the cells is achieved by the second class of proteoglycans via calcium ions mediated noncovalent bonding. The bonds of the aggregation factors are very specific, and only bonds between the factors of the same sponge type are observed when the aggregation factors of different sponges are mixed.⁴ Furthermore, the presence of cations is absolutely necessary for cell adhesion, whereby other metal ions can partially replace calcium in its function.^5,6 The aggregation factors were isolated and identified for various sea sponges, and for the sea sponge Microciona prolifera, the sulfated disaccharide GlcpNAc3S(β1–3) Fucp is relevant for the complex that is stabilized by the calcium ions.

Fig. 1 presents 2D and 3D structure of the sulfated disaccharide GlcpNAc3S(β1–3) Fucp, which is responsible for the carbohydrate-carbohydrate complex bridged by calcium ions. ‘Glc’ is the abbreviation for a glucose sugar and ‘Fuc’ means a fucose sugar. ‘Ac’ stands for an acetyl group (-CO-CH3) and the curved line indicates the binding to the rest of the proteoglycan. In the simulations, an alkyl chain with 5 carbon chain units was used.

Figure 1. 2D and 3D structure of the sulfated disaccharide GlcpNAc3S(β1–3)Fucp.

Previously, the aggregation behavior of the disaccharide was investigated as a function of various cations. The results demonstrated that while magnesium ions did not induce aggregation, calcium ions did lead to aggregation. Furthermore, cadmium ions also caused aggregation, although to a lesser extent compared to calcium ions. In a separate study, the disaccharide-cation system (involving the cations Mg2+, Ca2+, and Cd2+) was investigated using atomic force microscopy (AFM). The cantilever tip and surface were coated with the disaccharide. When brought into contact, the sugars and ions had the potential to form complexes. In an aqueous environment without additional cations, repetitive measurements were conducted at small distances (before the tip and surface made direct contact), resulting in a repulsive interaction attributed to the negatively charged disaccharide monolayers. The addition of magnesium ions suppressed this repulsive interaction, but no complex formation was observed. In contrast, binding interactions occurred in 95% of measurements when calcium ions were used, and the rupture force for a single complex was determined to be 30 ± 6 pN. In experiments using cadmium ions, binding events were observed in only about 30% of cases.

Following previous findings,⁷ the purpose of this study is to investigate the binding mechanism of calcium ion carbohydrate-carbohydrate bond and to determine why no analogous magnesium complex exists. The individual simulations and their results are presented below. All molecular dynamics (MD) simulations were performed using the DFTB+ version 23.1.⁸ A similar protocol was used as for the simulations of the many different biological systems.^9,10 The 3ob Slater-Koster files were used for modeling the interactions between the atoms.¹¹ The 3ob parameters are optimized for bio and organic molecules. In all simulations, carbohydrates and ions were solvated with water.

In this study, simulations were performed in which multiple disaccharide molecules and cations were modeled in an aqueous medium, with the excess charge of the cations balanced by chlorine ions. Subsequently, different radial distribution functions g(r) were considered to analyze to which extent different groups accumulate in the simulation.

Simulation methods

The sulfated disaccharide presents a relatively large system, posing a challenge for DFT methods due to the substantial computational demands^12,13 associated with such scale. However, the density functional-based tight-binding method (DFTB) emerges as a rapid and efficient semi-empirical quantum-chemical computational approach. It enables molecular dynamics simulations of systems containing several thousand atoms, with simulation durations in the picosecond range, all without necessitating extensive computational time or significantly augmented computing resources.^14–17 This characteristic renders DFTB a distinctive and highly adaptable tool for exploring not just structure but also dynamics. Our investigations involved conducting molecular dynamics simulations (MD) utilizing the dispersion-corrected density functional-based tight-binding theory (DC-DFTB).¹³

The following protocol was used in the actual MD simulations. First, the isolated structures were subjected to energy minimization. Subsequently, the system was solvated and water molecules were randomly exchanged with ions. A second energy minimization was performed for the whole system. In the first equilibration phase, the system was subjected to a ‘velocity rescaling’ thermostat 1 with temperature 300 K and a time constant of 0.25 ps for 50 ps. The MD trajectories were generated in the NVT ensemble using the Berendsen thermostat. The simulation box has a size of about 5.4 nm × 4.4 nm × 4.4 nm.

The binding energy of the ions with the system was calculated with MP2 methods using Gaussian simulation package.¹⁸ Using the MP2 method to study the full system is very costly in terms of computation. A more feasible and accurate approach is to consider only a part of the system that the MP2 method can handle. This part is called a cluster, and its edges are saturated with hydrogen atoms.

In the preliminary study, a single solvated disaccharide molecule and 20 cations (Ca2+ and Mg2+, respectively) in a cubic simulation box with an edge length of 3.7 nm were modeled for 20 ps. To neutralize the system, 39 chlorine ions were added as counterions.

In simulations with many saccharide molecules (mono- or disaccharide) from which radial distribution functions were obtained, a cubic simulation box with an edge length of 6.5 nm was used, and the system was simulated for 15 ps. Study the full systems with DFTB method in range from nanoseconds to microseconds, with some studies reaching milliseconds for more complex systems, tends to be prohibitively expensive at the current technical level. A compromise between computational cost and accuracy can be achieved by considering only a cluster of the system or educing simulation time. It is assumed that 15 ps is sufficient for observing the specific dynamics of cation-mediated disaccharide aggregation. For many molecular processes, significant conformational changes or aggregation events can occur on the order of picoseconds to nanoseconds, making 165 ps potentially sufficient for capturing initial interactions and dynamics.^1–6

Table 1 shows the number of saccharide molecules, cations (Ca2+ and Mg2+, respectively), and counterions (Cl−) used in individual simulations. In the case of the magnesium system, simulations were repeated with the modified water model. The traditional models for water cannot accurately capture the strength and nature of ion-water interactions, especially in systems with high ionic concentrations or specific ions like calcium and magnesium. This change is aimed at improving the representation of electrostatic interactions, reducing excessive polarization, or better mimicking experimental observations related to hydration properties. This modification lead to improved accuracy in simulating ion solvation and aggregation processes. By adjusting the charge distribution, the model more accurately reflect the dynamics of hydration shells around cations. In the simulations with the modified water model, the charges of the individual atoms of a water molecule were reduced to 83.6% of their original value.

Table 1. Number of saccharide molecules, cations (Ca2+ and Mg2+,) and counterions (Cl−) used in the simulations.

saccharides	Cations/Counterions
5	15/25
5	20/35
5	25/45
5	40/75
10	20/30
10	40/70
10	60/110
10	80/150
15	20/25
15	40/65
15	60/105
15	80/145

Three additional simulations were conducted for this simulation box in which the negative charge of disaccharides was balanced by sodium ions, and no other ions were added. These simulations contained 5, 10, and 15 disaccharide molecules and as many sodium ions. These systems were also simulated for 15 ps.

For the calcium system, three more simulations were performed with a larger simulation box (9 nm edge length). Table 2 shows the number of disaccharides, cations (Ca2+), and counterions (Cl-) used in the simulations. The same parameters were used as in the other simulations despite the different number of particles and larger simulation box.

Table 2. Number of disaccharide molecules, cations (Ca2+ or Mg2+) and counterions (Cl−), that were used in the simulations with the larger simulation box.

saccharides	Cations/Counterions
30	120/210
30	150/270
30	180/330

Results and discussion

Preliminary investigation

Simulations were conducted to study the carbohydrate-cation system, which involved modeling several disaccharide molecules and divalent cations (Ca2+ and Mg2+).¹⁹ Before presenting the results of these simulations, a preliminary investigation was carried out to determine to which groups or atoms within the disaccharide magnesium ions bind to. In this investigation, two systems were studied, each containing a solvated disaccharide molecule and 20 cations (Ca2+ or Mg2+). To compensate for the excess positive charge of the cations, 39 chlorine ions were added as counterions1. The simulations showed that the magnesium ions bind to the sulfate group and carbon group of the disaccharide, but not to the acid atoms of the disaccharide, to which the calcium ions have a higher affinity.²⁰ The reason for this is the smaller ionic radius of the magnesium ion compared to the calcium ion, which causes a solvated magnesium ion to bind the water molecules of the hydrate shell to itself more strongly than a calcium ion does.²¹ The difference between the two ions is also reflected in their hydration enthalpies and the rate constants for the water exchange.²² Table 3 in the source gives the experimentally determined values as well as the cation radius of both types of cations. The cation radius,²³ hydration enthalpy ∆Hhydra²⁴ and rate constant for water exchange kex for the two types of cation (Ca2+ and Mg2+). For the Values of kex see the following references: Mg2+²⁵ and Ca2+.²⁶

Table 3. The experimentally determined values as well as the cation radius of both types of cations.

Cation type	Rcation/Å	∆Hhydra/kJ/mol	kex/s − 1
Mg2+	0.60	−1922	5.3 · 105
Ca2+	0.95	−1577	(6–9) · 108

Magnesium has a higher (more negative) hydration enthalpy than calcium, which means that magnesium binds the hydrate shell more strongly than calcium 1 2. This results in slower water exchange in the magnesium system, as the hydrate shell must first be partially broken for the cations to bind to the oxygen atoms of the disaccharide.²⁷ The formation of the cation-disaccharide bond is kinetically inhibited due to the stronger binding of the hydrate shell to magnesium1. To investigate whether the strength of the hydrate shell binding can explain the differences in the affinity of the cation types, further simulations were carried out with magnesium ions1. In these simulations, the water model used was modified by scaling the partial charges of the water molecules by the factor rMg/rCa = 0.84, which reduced the strength of the Coulomb interaction between the magnesium ion and the water molecule1. As a result, the hydrate shell should bind less strongly to the magnesium ion, allowing the cation to attach more readily to the disaccharide.²⁸ The simulations with the modified water model are designated as ‘Mg*’.

Determination of the RDF

Multiple simulations were conducted to investigate the binding structure and stability of cation-carbohydrate complexes, varying the numbers of disaccharide molecules and cations. For each cation type (Ca, Mg, and Mg*), three disaccharide concentrations with four cation concentrations were examined, resulting in a total of 12 simulations. Table 4 details the specific quantities of disaccharide molecules and cations utilized in the simulations. Chlorine ions were employed to balance the excess charge of the cations in each simulation. Additionally, simulations were performed for the three disaccharide concentrations where the negative charge of the disaccharides was neutralized by sodium ions without the addition of other ions. These simulations included 5, 10, and 15 sodium ions, resulting in only 3 simulations compared to the 12 simulations with other cation types. These sodium ion simulations represented a system without cations, although the presence of sodium ions was necessary for charge balance in the simulation. Experimental evidence suggests that no disaccharide complex forms with sodium ions due to repulsion between deprotonated sulfate groups. Consequently, no complex formation was observed in the MD simulations of the sodium system, serving as a ‘negative blank sample’. The term ‘Na’ was introduced for simulations involving sodium ions. All simulations utilized a cubic box with an edge length of 6.5 nm and were solvated in water. The system was coupled to a thermostat and barostat to regulate temperature (T = 300 K) and pressure (p = 1 bar). Each simulation ran for 15 ps.

Table 4. The quantities of disaccharide molecules and cations employed in the simulations.

Disaccharides	cations (Ca, Na, Mg, Mg*)
5	15, 20, 25, 40
10	20, 40, 60, 80
15	20, 40, 60, 80

Table 4 displays the quantities of disaccharide molecules and cations employed in the simulations. Each cation type was studied with three disaccharide concentrations, each paired with four distinct cation concentrations. The quantities of disaccharide molecules and cations for the four cation concentrations are presented in a single line.

RDF of sugar rings

Figure 2 illustrates three potential binding modes for disaccharide molecules to cations. The cations are not depicted, as the focus is on the relative arrangement of disaccharides. In cases (1) and (2), conformations are also considered where one of the two disaccharides is bent out of the plane, resulting in an angled arrangement rather than a linear one. In case 3, the glucose rings to which the sulfate groups are bound are colored red, while the fucose rings to which the alkyl chains are bound are colored green. In cases (1) and (2), where two identical sugar rings bind to a cation, the sugar rings are connected by corresponding colored lines. Interactions between different sugar rings are highlighted in blue, as seen in case (3). For the determination of radial distribution functions, the distances between the centers of the sugar rings (excluding the side groups) were measured. Henceforth, the glucose ring will be referred to with the index ‘Glc’, and the fucose ring with the index ‘Fuc’.By analyzing the three possible radial distribution functions gGlc-Glc, gFuc-Fuc, and gGlc-Fuc, it can be determined whether one of the three possible arrangements (1, 2, or 3) is preferred in the complex. Figure 3 presents the three radial distribution functions for all cation types, using the same color code as in Figure 2. The radial distribution functions were averaged over all simulations for the corresponding cation type. In the calculation of gGlc-Fuc, only the intermolecular contributions to the radial distribution function were considered, as these are the only ones relevant for the subsequent analysis.

Figure 2. Possible arrangements of disaccharides in the complex.

Other similar conformations can also be generated can be produced by rotation of a disaccharide molecule. The colored lines indicate contacts between the same (red or green) and different (blue) sugar rings.

Figure 3. Radial distribution function with respect to the centers of gravity of the sugar rings.

The Color scheme is the same as in Figure 2. With the index ‘Glc’, the glucose rings and by ‘Fuc’ meant the fucose rings.

Comparing the three radial distribution functions of the sodium system with those of other cation types, two things stand out. On the one hand, the three distribution functions have smaller maximum values (gX-Y < 3) than those of the other cation types (gX-Y > 4). This means that in the case of the sodium system, the disaccharides do not approach each other as much as in the other cation systems. This is especially true for the glucose units, as can be seen from gGlc-Glc (Na) < gFuc-Fuc (Na); gGlc-Fuc (Na) for the sodium system, while the opposite trend is observed for the other cation systems. This trend can be explained by the deprotonated sulfate groups that are bound to the glucose units. The few cations in the sodium system cannot balance the negative charge of the sulfate groups in the same way as the many divalent cations. As a result, no pronounced approach of the disaccharides is observed in the sodium system. It can therefore be assumed that a certain amount of (divalent) cations is essential for the aggregation of the disaccharides.

In the case of calcium ions, a pronounced peak is found in gGlc-Glc (Ca) at about r = 0.6 nm, and it is evident that arrangement (1) from Figure 2 is clearly preferred over the other two arrangements. The divalent cations act here as a kind of “buffer” that balances the negative charge of the sulfate groups and thus enables the approach of these groups. Moving on to magnesium ions (with the normal water model), this pronounced peak is missing in gGlc-Glc (Mg). Nevertheless, gGlc-Glc (Mg) has a larger value for small distances (r ≈ 0.55 nm) than gFuc-Fuc (Mg) and gGlc-Fuc (Mg), which is why a slight preference for arrangement (1) can also be assumed here. The reason for the lack of pronounced peak structure is likely the fact that magnesium ions bind the hydration shell more strongly. As a result, the magnesium ions indirectly bind to the sulfate group via the hydration shell, resulting in a weaker bond than in the calcium system. This leads to the complex being less stable and occurring less frequently over time than in Ca, resulting in gGlc-Glc (Mg) < gGlc-Glc (Ca). When considering the radial distribution functions in the case of the magnesium system with the modified water model gX-Y (Mg*), the relevance of the binding strength of the hydration shell is confirmed. In this system, the hydration shell is bound less strongly than in the magnesium system with the normal water model (Mg). As a result, the magnesium ions can more easily bind to the sulfate group, and especially for small distances, gGlc-Glc (Mg*) > gGlc-Glc (Mg).

When comparing gFuc-Fuc (cation) and gGlc-Fuc (cation), they are quite similar for all types of cations. However, compared to gGlc-Glc (cation), the maxima are found at slightly larger distances. In summary, it can be said that especially for Ca and Mg*, a complex is preferred in which the glucose units bind to each other. However, for such a complex, divalent cations are essential to balance the negative charge of the sulfate groups, which would otherwise repel each other (see gGlc-Glc (Na)).

RDF of sulfate groups and cations

Since the disaccharides primarily bind through the glucose units, it is worthwhile to examine the radial distribution function of the cations and sulfate oxygen atoms, as the sulfate groups are attached to these glucose units. Furthermore, the preliminary investigation discussion noted that the cations predominantly bind to the sulfate groups. Although the carbonyl group of the N-acetyl group is also a preferred binding partner compared to the other oxygen atoms of the disaccharide, it is less preferred than the sulfate group. Before presenting the results, a technical detail should be addressed. When comparing the values of gSul-Kation (rmax), where rmax indicates the position of the first peak, for different cation concentrations at a given disaccharide concentration, it is found that the values decrease with increasing cation concentration. This can be explained by the fact that the number of cations that bind to the sulfate groups does not increase to the same extent as the number of cations modeled in the simulation. This becomes most apparent when considering a single disaccharide molecule. At a certain cation concentration, the sulfate group is saturated with cations. If the cation concentration is increased further, the number of cations that bind to the sulfate group remains unchanged. However, the number of cations that are further away from the sulfate group increases. This leads to an increased probability of not finding a specific cation in the immediate vicinity of the sulfate group, and consequently, the probability of a specific cation binding to the sulfate group decreases. Since the simulations of the different cation types are equally affected by this effect, the results of a given disaccharide and cation concentration can be compared with each other for different cation types. Figure 4 shows the radial distribution function of the sulfate oxygen atoms and cations, gSul-Kation, for the three cation types (Ca, Mg, Mg*), where the results were obtained from simulations with 10 disaccharides and 60 cations. Qualitatively similar results are obtained for the other simulations, and the same trend regarding the affinity of the cation types to bind to the sulfate group is observed.

Figure 4. Radial distribution function with respect to the cations and the acidic atoms of the Sulfate groups.

In the case of calcium ions, gSul-Ca exhibits two peaks at r = 0.24 nm and 0.46 nm. The first peak corresponds to calcium ions that bind directly to the sulfate oxygen atoms. The distance of 0.24 nm is also consistent with the distances found between calcium and oxygen in crystalline sugar-calcium complexes. The second peak corresponds to cations that are indirectly bound to the surrounding sulfate oxygen atoms. This includes calcium ions that bind to the sulfate group via a water molecule. Additionally, this peak is also due to the fact that the sulfate group contains three oxygen atoms to which the cations can bind. If the calcium ion binds to one of the three atoms, it is located near the other two atoms, but the distance is greater. In the case of magnesium ions, the two peaks in the radial distribution function are shifted to smaller distances regardless of the water model used, as smaller binding distances are possible due to the smaller ionic radius of magnesium compared to calcium. In summary, this confirms the assumption that the formation of a bond between the cation and disaccharide in the case of the magnesium system is kinetically hindered due to the more strongly bound hydration shell compared to the calcium system.

RDF of sulfate groups

The radial distribution function of the sulfate groups is of particular interest for investigating the stability of carbohydrate-cation-carbohydrate complexes. It has already been established that the cations primarily bind to the sulfate group. The results indicate that the disaccharides primarily aggregate through the glucose units to which the sulfate groups are bound. It can therefore be assumed that in the carbohydrate-cation-carbohydrate complex, the disaccharides are bridged by the sulfate groups and cations. Thus, the distance between two sulfate groups is a suitable choice for the reaction coordinate of complex formation or dissociation. Furthermore, it can be ensured that two disaccharides are bridged by a cation when they come very close together, as the groups would repel each other due to their negative charge without the compensating charge of the cation.

The PMF will be presented, as it is essentially a logarithmic representation of the radial distribution function, and therefore, the same information can be obtained from both quantities in principle. However, first, the general trend of gSul-Sul (Ca) for different disaccharide concentrations and different sizes of the simulation box will be briefly discussed. This discussion is more of technical interest; nevertheless, it shows that gSul-Sul (Ca) was adequately determined.

When comparing the radial distribution functions of the sulfate groups for different disaccharide and cation concentrations, it is noticeable that they do not always yield exactly the same result, and different values are found. However, unlike gSul-cation, no dependence on the disaccharide or cation concentration can be found. If one averages over the different cation concentrations at a given disaccharide concentration, the results for the three disaccharide concentrations are very similar. This can be seen for the calcium system on the left side of Figure 5. To check whether gSul-Sul (Ca) is affected by the size of the simulation box, further simulations were performed for the calcium system with a larger simulation box (9 nm edge length instead of 6.5 nm). Here, 30 disaccharides and 120, 150, or 180 calcium ions were investigated. On the right side of Figure 5. 6, gSul-Sul (Ca) is shown for the two simulation boxes. In both cases, gSul-Sul (Ca) represents the average of the results for all disaccharide and cation concentrations.

Figure 5. gSul-Sul for the calcium system.

Left: gSul-Sul (Ca) at a given disaccharide concentration, where the four calcium concentrations were averaged out. Right: Comparison of gSul-Sul (Ca) for two different simulation boxes, with the smaller box using to determine the radial distribution functions.

In both figures, it can be seen that the averaged radial distribution functions differ only slightly from each other. Especially since the averaged radial distribution functions for the two different simulation boxes show only small deviations from each other, it can be assumed that gSul-Sul (Ca) was reliably determined, and the same applies to the other cation types. The curves of VCa, VMg, and VMg* have different shapes. VCa has two dips, while VMg and VMg* have only one dip each. To understand why this is the case, it is important to look at the binding structures of each complex. But first, let’s compare VMg* with the other two curves. VMg* is similar to VMg, but it has lower energies overall. This means that the modified water model makes the complex more stable and more frequent than the normal water model. However, VMg* also reaches zero at larger distances than VCa and VMg. This implies that the sulfate groups attract each other more strongly in the modified water model than in the normal water model. This is an unintended effect of using the modified water model, which reduces the partial charges and the dipole moment of water molecules. As a result, the dielectric constant of water decreases and the modified water screens electrostatic interactions less effectively than ‘normal’ water.

Potential of mean force

A potential of mean force can be calculated from a radial distribution functionusing an equation that describes the free energy along the reaction coordinate r, which is the distance between two sulfate groups.^29,30 The term V cation(r) will be introduced for the PMF of a particular cation system. The PMF can be obtained in Monte Carlo or molecular dynamics simulations to examine how a system’s energy changes as a function of some specific reaction coordinate parameter. It can be a geometrical coordinate or a more general energetic (solvent) coordinate. The PMF is related to the radial distribution function of the system, g(r), by the equation:

(1)

V_{pmf} (r) = - K_{B} T \ln g (r)

where

V_{pmf} (r)

is the potential of mean force. The PMF gives the average force over all the configurations of all the particles acting on a particle j at any fixed configuration keeping fixed a set of particles. Potentials of mean force can be calculated from both simulations and experiments run under equilibrium conditions by histogramming the values of the chosen parameter.

Comparing VCa(r) from Figure 6 with the radial distribution functions from Figure 5, it can be seen that both quantities are equivalent in terms of the essential characteristics of the curve shapes. The maxima of gSul-Sul (Ca) are reflected as minima in VCa(r), and both curves have a constant value from r ≈ 1 nm onwards.

Figure 6. Potential of mean for the carbohydrate-cation-carbohydrate complex, where reaction coordinate, the distance between the sulfate groups is used.

When considering the PMF of the sodium system VNa(r), it is noticeable that it practically does not have a minimum compared to the other cation systems and already increases strongly at larger distances. Since these simulations only contain a few monovalent cations, the curve can be explained by the electrostatic repulsion of the sulfate groups. In the three other systems, the divalent cations can compensate for the negative charge of the sulfate groups, and as a result, the sulfate groups can approach each other further. This can be seen from the fact that the repulsive range of VKation(r) for Ca, Mg, and Mg* is shifted to smaller distances compared to VNa(r). Furthermore, negative values are observed in VKation(r) for the divalent cations, meaning that the sulfate groups experience an attractive interaction at these distances. The distances at which the minima are found correspond to the distance between the two sulfate groups in the bound complex. The minimum of the PMF corresponds to the negative binding energy. For the three systems of divalent cations, the binding energies are found in descending order: 8.5 kJ/mol (Ca), 6.9 kJ/mol (Mg*), and 3.9 kJ/mol (Mg) as Table 5.

Table 5. Quantitative summary of binding energies and characteristic distances from RDF and PMF simulations.

System	Binding Energy (kJ/mol)	g_Sul-Sul Peak Position (nm)	g_Sul-Cation 1st Peak (nm)
Ca²⁺	−8.5	~0.6	0.24
Mg²⁺	−3.9	~0.55	0.22
Mg*	−6.9	~0.54	0.21

The trend of the binding energies is similar to the trend of the affinity of the cation types to bind to the sulfate group. To estimate an experimental value for the lower bound of the binding energy, the results of pull experiments are suitable. In reference, the mean rupture force of a single calcium complex was determined to be F ≈ 30 ± 6 pN. The distance between the bound state and the transition state was estimated in reference³¹ to be ∆x ≈ 0.3 nm. The product of these two values corresponds to the amount by which the energy barrier is lowered on average before the transition occurs. Since the transition generally occurs through thermal activation, the energy barrier and thus the binding energy are greater than the amount of the lowering. For the lowering of the energy barrier, one obtains ∆V ≈ F · ∆x = 5.4 kJ/mol, and it can be assumed that the error is about 3 kJ/mol. The value of this lower bound is at least of the same order of magnitude as the binding energy determined here, but it cannot be said by how much the actual value is greater. To achieve better comparability with the experiments, pull experiments were simulated using Brownian simulations. The rupture force that can be determined from these can be compared with the experimental rupture forces and thus provides a better comparison value than the binding energies.

In all three cases, the binding energy is in the order of magnitude of the thermal energy kBT ≈ 2.5 kJ/mol (for T = 300 K). The difference in binding energies between the calcium and magnesium complexes can probably explain why the latter is not observed in experiments. For the magnesium complex, the binding energy is not even twice the thermal energy, meaning that the complex should be easily opened by thermal fluctuations. Thermal fluctuations as the variations in energy that occur due to thermal motion at a given temperature. These fluctuations can affect molecular interactions and stability, making it important to consider them when analyzing complex formation. If thermal fluctuations are large relative to binding energies, this could lead to transient disruptions in complex formation, affecting overall stability and aggregation behavior. In all three cases, the binding energy is in the order of magnitude of the thermal energy kBT ≈ 2:5 kJ/mol (for T = 300 K). The difference in the binding energies for the calcium and magnesium complex can probably explain why the latter is not observed in the experiments. For the magnesium complex, the binding energy is not even twice the thermal energy, which means that the complex should be able to be opened very easily by thermal fluctuations. In the case of the calcium complex, however, the binding energy is more than a factor of 3 greater than the thermal energy, which is why this complex should be somewhat more stable than the magnesium complex. Since the binding energy is nevertheless quite small, it is also to be expected that the calcium complex can dissociate even through thermal fluctuations. However, a single complex is negligible for cell aggregation, and only many such complex bonds bind the cells together.

In the case of the calcium complex, however, the binding energy is more than a factor of 3 greater than the thermal energy, so this complex should be slightly more stable than the magnesium complex. However, since the binding energy is still quite small, it can be expected that the calcium complex can also dissociate due to thermal fluctuations. For cell aggregation, however, a single complex can be neglected, and it is only many such “complex bindings” that bind the sponge cells together.

Comparison with a monosaccharide system

The previous results show that sulfate groups, which are bound to the glucose unit of the disaccharide, are mainly responsible for complex formation. Experimentally, it was shown in^32,33 that a monosaccharide consisting only of the glucose unit of the disaccharide presented here does not form a stable carbohydrate-calcium-carbohydrate complex. In the experiments, gold nanoparticles were coated with the disaccharide or monosaccharide. When calcium ions were added to a solution of such nanoparticles, aggregation of the nanoparticles was observed in the case of the disaccharide system. Since this aggregation was absent in the case of the monosaccharide system, it was concluded that no carbohydrate-calcium-carbohydrate complex is formed with the monosaccharide.

In MD simulations, differences in complex formation between the two saccharides should also be detectable. Therefore, the same simulations as for the disaccharide system were performed with the monosaccharide shown in Figure 7.

Figure 7. Monosaccharide that consists only of the glucose unit of the disaccharide.

Figure 8 shows the PMF in the case of calcium ions for the two saccharide systems. Ignoring the numerical values of Vcation(r), the course of the PMF for both systems is the same. This is expected because the immediate environment of the sulfate groups, whose distance serves as a reaction coordinate, is the same in both systems. However, in the case of the monosaccharide, a smaller binding energy is observed (≈ 6 kJ/mol instead of 8.5 kJ/mol). This is thus one unit of thermal energy kBT smaller than for the disaccharide system. The fact that the monosaccharide system is less stable than the disaccharide system is positive in terms of comparability with the experiments. Nevertheless, the relatively small energy difference is striking.

Figure 8. Potential of mean force for the carbohydrate-calcium-carbohydrate complex for the disaccharide and monosaccharide system.

Finally, it is worth mentioning that trajectory lengths of DFTB simulations of 15 ps may not potentially enough for the studied system. However, one may emphasize that only preliminary analyses have been conducted in this study (e.g., energy convergence and structural stability), which suggest that the system reaches a near-equilibrium state within this time frame. Longer simulations could provide further insights into dynamic behavior, hence exploring options for extending simulations in future work is demanded.

Conclusion

This research investigates the cation-mediated aggregation of disaccharides, which is relevant for the self-adhesion of cells in the sea sponge Microciona prolifera. Experimentally, it is known that calcium ions induce aggregation of disaccharides, while magnesium ions do not. The goal of this study was to determine the differences in binding behavior between the two cation species. Initial findings revealed that the cations primarily bind to the sulfate groups of the disaccharides, leading to the formation of carbohydrate-cation-carbohydrate complexes. In these complexes, the disaccharides predominantly bind to the glucose unit attached to the sulfate group. This binding preference is particularly pronounced in the calcium system. To assess the stability of the complexes, a potential of mean force was determined using the distance between the sulfate groups as the reaction coordinate, as small distances between these groups indicate that they are bridged by a cation. In the absence of cations, the sulfate groups repel each other, preventing complex formation. The binding energy for the calcium complex was found to be 8.5 kJ/mol, while in the case of the magnesium complex, it is 3.9 kJ/mol. In both cases, the binding energy is relatively small, consistent with experimental observations that the calcium complex is not very stable and the magnesium complex is not observed. The binding energies are significantly higher in the presence of certain cations (like Ca²⁺ or Mg²⁺), this could indicate that these ions play a critical role in stabilizing disaccharide interactions.

MD simulations provide insights into the differences between the two cation systems: Magnesium ions bind much more weakly to the disaccharide compared to calcium ions. This can be attributed to the smaller ionic radius of magnesium ions, which results in stronger binding to the hydration shell. For a cation to bind to the disaccharide, the hydration shell must first be broken, which is energetically more costly for magnesium ions, leading to a kinetic hindrance of complex formation relative to the calcium system. When a modified water model is used for the magnesium system, where the charges of the water atoms are reduced, the magnesium ions exhibit weaker binding to the hydration shell, facilitating complex formation with the disaccharides. Another significant difference is that in the calcium system, both disaccharides bind directly to the cation through the sulfate groups, while in the magnesium system, only one disaccharide enters a direct binding, with the other binding indirectly through a water molecule, weakening the overall binding strength. The results indicate that the sulfate group is crucial for complex formation. Experimentally, it is known that a monosaccharide, consisting only of the glucose unit of the disaccharide, does not form a stable carbohydrate-calcium complex when bound to the sulfate group. To verify whether the stability differences between the two saccharides are also observed in the simulations, the calcium system simulations were repeated with a monosaccharide. As expected, the binding energy of the monosaccharide complex was found to be lower than that of the disaccharide complex (6 kJ/mol), but the difference in binding energies between the two complexes was relatively small. Additionally, adhesion clusters of different sizes stabilized through multiple complex bindings were modeled. These simulations allowed for the determination of the average rupture force of an individual complex. Compared to experimental data, the simulations predicted a significantly lower rupture force for an individual complex, which can be attributed to simplifications in the models used.

Moreover, RDF simulation results showed that the cations primarily bind to the sulfate group. it was shown that the disaccharides primarily aggregate through the glucose units to which the sulfate groups are bound. It can therefore be assumed that in the carbohydrate-cation-carbohydrate complex, the disaccharides are bridged by the sulfate groups and cations. Thus, the distance between two sulfate groups is a suitable choice for the reaction coordinate of complex formation or dissociation. Furthermore, it can be ensured that two disaccharides are bridged by a cation when they come very close together, as the groups would repel each other due to their negative charge without the compensating charge of the cation.

Nevertheless, it is remarkable that the simulations qualitatively reproduced the experiments, as differences in the stability of the complexes formed by the different saccharide-cation systems were observed, this study provides a multiscale understanding bridging molecular physics and biological function.

Preprint:

This work is based on our previously preprint, where large sections of this manuscript have appeared on a preprint server, DOI: 10.21203/rs.3.rs-4630635/v1, Posted Date: July 19th, 2024. This study did not involve human participants, human data, or human tissue.

Data Availability

All data generated or analyzed during this study are fully represented in the figures and results presented within this article. No separate underlying numerical datasets exist beyond what is shown in the graphs and tables provided.

References

1. Cailliez F, Lavery R: Cadherin mechanics and complexation: the importance of calcium binding. Biophys. J. 2005; 89: 3895–3903. PubMed Abstract | Publisher Full Text | Free Full Text
2. Sotomayor M, Schulten K: The allosteric role of the Ca2+ switch in adhesion and elasticity of C-cadherin. Biophys. J. 2008; 94: 4621–4633. PubMed Abstract | Publisher Full Text | Free Full Text
3. Steitz TA, Steitz JA: A general two-metal-ion mechanism for catalytic RNA. Proc. Natl. Acad. Sci. 1993; 90: 6498–6502. PubMed Abstract | Publisher Full Text | Free Full Text
4. Mitchell JM, Nichols SA: Diverse cell junctions with unique molecular composition in tissues of a sponge (Porifera). EvoDevo. 2019; 10: 1–16. PubMed Abstract | Publisher Full Text | Free Full Text
5. González-Amaro R, Sánchez-Madrid F: Cell adhesion molecules: selectins and integrins. Crit. Rev. Immunol. 1999; 19: 389–429. PubMed Abstract
6. De Kerchove AJ, Elimelech M: Calcium and magnesium cations enhance the adhesion of motile and nonmotile Pseudomonas aeruginosa on alginate films. Langmuir. 2008; 24: 3392–3399. PubMed Abstract | Publisher Full Text
7. Kafi DK, Mohi AT: Molecular dynamics study of supramolecular complexes to the carbohydrate. Research Square Preprint. 2024. Publisher Full Text
8. Hourahine B, Aradi B, Blum V, et al.: DFTB+, a software package for efficient approximate density functional theory based atomistic simulations. J. Chem. Phys. 2020; 152: 124101. PubMed Abstract | Publisher Full Text
9. Zeng J, Zhang R-Q, Treutlein H: Quantum simulations of materials and biological systems. Springer Science & Business Media; 2012. Publisher Full Text
10. Frauenheim T, Aradi B: Atomistic Simulations Using the Approximate DFT Method DFTB+: Applications to Nanomaterials and Bio-Systems. Multiscale Modelling Methods for Applications in Materials Science: CECAM Tutorial, Forschungszentrum Jülich. 2013; pp. 41.
11. Lourenço MP, da Silva MC , Oliveira AF, et al.: FASP: a framework for automation of Slater–Koster file parameterization. Theor. Chem. Accounts. 2016; 135: 1–12. Publisher Full Text
12. Frauenheim T, Seifert G, Elsterner M, et al.: A Self-Consistent Charge Density-Functional Based Tight-Binding Method for Predictive Materials Simulations in Physics. Chemistry and Biology, physica status solidi (b). 2000; 217: 41–62. Publisher Full Text
13. Zhechkov L, Heine T, Patchkovskii S, et al.: An efficient a posteriori treatment for dispersion interaction in density-functional-based tight binding. J. Chem. Theory Comput. 2005; 1: 841–847. PubMed Abstract | Publisher Full Text
14. Porezag D, Jungnickel G, Frauenheim T, et al.: Theoretical investigations of homo- and heteronuclear bridged fullerene oligomers. Appl. Phys. A. 1997; 64: 321–326. Publisher Full Text
15. Mandumpal J, Gemming S, Seifert G: Curvature effects of nitrogen on graphitic sheets: Structures and energetics. Chem. Phys. Lett. 2007; 447: 115–120. Publisher Full Text
16. Porezag D, Frauenheim T, Köhler T, et al.: Construction of tight-binding-like potentials on the basis of density-functional theory: Application to carbon. Phys. Rev. B. 1995; 51: 12947–12957. PubMed Abstract | Publisher Full Text
17. Albertazzi E, Domene C, Fowler PW, et al.: Pentagon adjacency as a determinant of fullerene stability. Phys. Chem. Chem. Phys. 1999; 1: 2913–2918. Publisher Full Text
18. Zheng GJ, Sonnenberg L, Hada M, et al.: Gaussian 09. Wallingford CT: Gaussian Inc.; 2009; 48.
19. Chen H, Cox JR, Panagiotopoulos AZ: Force fields for carbohydrate–divalent cation interactions. J. Phys. Chem. B. 2016; 120: 5203–5208. PubMed Abstract | Publisher Full Text
20. Remko M, Van Duijnen PT, Broer R: Effect of metal ions (Li+, Na+, K+, Mg 2+ and Ca 2+) and water on the conformational changes of glycosidic bonds in heparin oligosaccharides. RSC Adv. 2013; 3: 9843–9853. Publisher Full Text
21. Yan H, Yuan S-L, Xu G-Y, et al.: Effect of Ca2+ and Mg2+ ions on surfactant solutions investigated by molecular dynamics simulation. Langmuir. 2010; 26: 10448–10459. PubMed Abstract | Publisher Full Text
22. Aakesson R, Pettersson LGM, Sandstroem M, et al.: Theoretical study on water-exchange reactions of the divalent and trivalent metal ions of the first transition period. J. Am. Chem. Soc. 1994; 116: 8705–8713. Publisher Full Text
23. Da Silva JJRF, Williams RJP: The biological chemistry of the elements: the inorganic chemistry of life. Oxford University Press; 2001.
24. Mejıas JA, Lago S: Calculation of the absolute hydration enthalpy and free energy of H+ and OH−. J. Chem. Phys. 2000; 113: 7306–7316. Publisher Full Text
25. Schutte-Smith M, Roodt A, Visser HG: Ambient and high-pressure kinetic investigation of methanol substitution in fac-[Re (Trop)(CO) 3 (MeOH)] by different monodentate nucleophiles. Dalton Trans. 2019; 48: 9984–9997. Publisher Full Text
26. Mason WR: Electronic absorption and magnetic circular dichroism of the linear diaquadimercury (I) cation. Inorg. Chem. 1983; 22: 147–150. Publisher Full Text
27. Huynh UTD, Lerbret A, Neiers F, et al.: Binding of divalent cations to polygalacturonate: A mechanism driven by the hydration water. J. Phys. Chem. B. 2016; 120: 1021–1032. PubMed Abstract | Publisher Full Text
28. Derbyshire W, Hedges ND, Lillford PJ, et al.: The influence of a mixed anionic system on the aggregation behaviour of agarose. Food Hydrocoll. 2001; 15: 153–163. Publisher Full Text
29. Meeron E: Theory of potentials of average force and radial distribution functions in ionic solutions. J. Chem. Phys. 1958; 28: 630–643. Publisher Full Text
30. Kawada S, Fujimoto K, Yoshii N, et al.: Molecular dynamics study of the potential of mean force of SDS aggregates. J. Chem. Phys. 2017; 147. Publisher Full Text
31. Lorenz BR, de Cienfuegos LA , Oelkers M, et al.: Model system for cell adhesion mediated by weak carbohydrate–carbohydrate interactions. J. Am. Chem. Soc. 2012; 134: 3326–3329. PubMed Abstract | Publisher Full Text | Free Full Text
32. Carvalho de Souza A, Halkes KM, Meeldijk JD, et al.: Gold glyconanoparticles as probes to explore the carbohydrate-mediated self-recognition of marine sponge cells. ChemBioChem. 2005; 6: 828–831. PubMed Abstract | Publisher Full Text
33. Chen Q, Cui Y, Zhang T-L, et al.: Fluorescent Conjugated Polyfluorene with Pendant Lactopyranosyl Ligands for Studies of Ca2+−Mediated Carbohydrate− Carbohydrate Interaction. Biomacromolecules. 2010; 11: 13–19. PubMed Abstract | Publisher Full Text

Comments on this article Comments (0)

Version 1

VERSION 1 PUBLISHED 10 Apr 2026

Author details Author details

¹ Medical Physics, University of Fallujah, Fallujah, Iraq, 41001, Iraq
² Department of Physics, Mustansiriyah University, Baghdad, Iraq, 10001, Iraq

Dhaidan Khalaf kafi
Roles: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Project Administration, Resources, Software, Supervision, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing

Ali Taher Mohi
Roles: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Project Administration, Resources, Software, Supervision, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing

Competing interests

No competing interests were disclosed.

Grant information

The author(s) declared that no grants were involved in supporting this work.

Article Versions (1)

version 1

Published: 10 Apr 2026, 15:503

https://doi.org/10.12688/f1000research.173339.1

Copyright

© 2026 kafi DK and Mohi AT. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Download

Export To

metrics

	Views	Downloads
F1000Research	-	-
PubMed Central Data from PMC are received and updated monthly.	-	-

Citations

0

SEE MORE DETAILS

CITE

how to cite this article

kafi DK and Mohi AT. Molecular Dynamics Study of Supramolecular Complexes to the Carbohydrate-Cation System [version 1; peer review: 4 approved with reservations, 1 not approved]. F1000Research 2026, 15:503 (https://doi.org/10.12688/f1000research.173339.1)

NOTE: If applicable, it is important to ensure the information in square brackets after the title is included in all citations of this article.

track

receive updates on this article

Track an article to receive email alerts on any updates to this article.

Open Peer Review

Current Reviewer Status: ?

Key to Reviewer Statuses VIEW HIDE

ApprovedThe paper is scientifically sound in its current form and only minor, if any, improvements are suggested

Approved with reservations A number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.

Not approvedFundamental flaws in the paper seriously undermine the findings and conclusions

Version 1

VERSION 1

PUBLISHED 10 Apr 2026

Views

6

Reviewer Report 03 Jun 2026

Weiwei He, Chemistry, New York University Abu Dhabi (Ringgold ID: 167632), Abu Dhabi, Abu Dhabi, United Arab Emirates

Approved with Reservations

https://doi.org/10.5256/f1000research.191145.r481856

1. Typographical and formatting issues should be corrected throughout the manuscript.
The manuscript contains multiple typographical, formatting, and citation errors that affect readability. For example, the phrase “magnesium binds the hydrate shell more strongly than calcium 1 2” appears ... Continue reading

1. Typographical and formatting issues should be corrected throughout the manuscript.
The manuscript contains multiple typographical, formatting, and citation errors that affect readability. For example, the phrase “magnesium binds the hydrate shell more strongly than calcium 1 2” appears to contain broken citation formatting. The authors should carefully check the entire manuscript, correct all citation-formatting errors, remove duplicated text, and ensure consistent use of chemical terms, symbols, units, and figure labels.

2. The artificial modification of the Mg²⁺–water interaction is not sufficiently justified.
A central part of the manuscript relies on a modified Mg system, in which water charges are scaled to weaken the Mg²⁺ hydration shell. This modification is highly artificial, because changing water charges affects not only Mg²⁺ hydration, but also the water dipole moment, dielectric screening, hydrogen bonding, ion–ion interactions, and sulfate–sulfate electrostatics.

The Mg* model therefore cannot be used as direct evidence for the proposed mechanism without systematic parameter refinement, benchmarking, and physical validation. The authors are encouraged to perform additional simulations using existing, validated Mg²⁺, Ca²⁺, and Na⁺ ion parameters developed for classical biomolecular force fields, especially those tested for nucleic acid or phosphate/sulfate-rich systems. Relevant examples include DOI: 10.1021/jz201501a, 10.1021/acs.jctc.3c00520, and 10.1002/bip.22868. Such validation is needed to determine whether the reported Ca²⁺/Mg²⁺ selectivity is robust or model-dependent.

3. The computational model and simulation protocol require substantial clarification.
The manuscript states that the MD simulations were performed using DFTB+, but the actual molecular model is not sufficiently described. The authors should clarify whether the simulations are fully DFTB-based, classical MD, or a hybrid workflow involving DFTB-MD and MP2 cluster calculations.

More details are needed on how water, ions, saccharides, and sulfate groups were modeled. For example, what water description was used, and how are the Mg²⁺, Ca²⁺, Na⁺, and Cl⁻ ions treated in the DFTB framework? The authors should also provide the relevant parameters, timestep, convergence criteria, etc. Without these details, the simulations are difficult to reproduce or critically assess.

4. The simulation time is too short to support reliable RDF, PMF, and aggregation conclusions.
The main RDF and PMF analyses appear to be based on very short simulations, approximately 15 ps for the multi-saccharide systems. This is likely insufficient for sampling divalent-ion dehydration, sulfate coordination, carbohydrate reorientation, solvent relaxation, and supramolecular aggregation.

5. The PMF-derived binding energies require stronger justification and uncertainty estimates.
The manuscript derives PMFs directly from RDFs and interprets the resulting minima as binding energies. This treatment requires much stronger justification, especially because the reported binding energies are small and comparable to only a few kBT.

The authors should provide convergence analysis and uncertainty estimates for the RDF-derived PMFs. They should also clarify that an RDF-derived PMF in a multi-component, finite-concentration system is an effective potential rather than a rigorous standard binding free energy. To obtain quantitative binding free energies, enhanced-sampling methods such as umbrella sampling or metadynamics are recommended, ideally using reaction coordinates that distinguish direct cation bridging from water-mediated or non-bridged contacts.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

No
If applicable, is the statistical analysis and its interpretation appropriate?

No
Are all the source data underlying the results available to ensure full reproducibility?

No
Are the conclusions drawn adequately supported by the results?

Partly

Competing Interests: No competing interests were disclosed.

Reviewer Expertise: Computational chemistry, force fields, MD simulations

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

CITE

Report a concern

Respond or Comment

Views

2

Reviewer Report 02 Jun 2026

Lei Dong, Ocean University of China, Qingdao, China

Approved with Reservations

https://doi.org/10.5256/f1000research.191145.r484822

The author presents a molecular dynamics simulation study focusing on the supramolecular complexes of carbohydrate-cation systems. This research aims to elucidate the microscopic mechanisms underlying the cation-mediated aggregation of disaccharides, specifically exploring why calcium ions (Ca²⁺) induce stable disaccharide ... Continue reading

The author presents a molecular dynamics simulation study focusing on the supramolecular complexes of carbohydrate-cation systems. This research aims to elucidate the microscopic mechanisms underlying the cation-mediated aggregation of disaccharides, specifically exploring why calcium ions (Ca²⁺) induce stable disaccharide aggregation while magnesium ions (Mg²⁺) do not. The author proposes a mechanistic framework based on the kinetic hindrance of the ion hydration shell and direct/indirect coordination with sulfate groups. While the manuscript presents interesting findings, major revisions are required to address the following concerns:

There is a critical error in the chemical structure presentation in Figure 1. Despite the heavy emphasis throughout the manuscript on the sulfate group as the core binding site, no sulfate groups are depicted in the chemical structures of the disaccharides in Figure 1. Furthermore, while the figure caption claims to display both 2D and 3D structures, only 2D skeletal formulas are visible in the figure. The author must completely redraw Figure 1 to ensure technical accuracy and consistency between the text and the illustration.
The simulation timescale is severely insufficient. The production run lasted for only 15 ps, which is far too short to achieve thermodynamic equilibrium for flexible carbohydrate chains characterized by rich conformational degrees of freedom. Such limited sampling compromises the reliability and accuracy of the resulting radial distribution functions (RDF) and potentials of mean force (PMF). The author is strongly encouraged to extend the simulation timescale to the nanosecond or microsecond range.
The physical validity of the modified water model is highly questionable. To artificially weaken the hydration shell of magnesium ions, the author scaled down the atomic charges of water to 83.6% (referred to as the Mg* model). This ad hoc modification alters the intrinsic physical properties of bulk water, such as its dipole moment and dielectric constant. The author must provide comprehensive benchmark data evaluating how this modification impacts the fundamental physical properties of water.
There is a logical disconnection between the simulated binding energies and the observed macroscopic phenomena. The simulation indicates that the monosaccharide complex still possesses a binding energy of approximately 6kJ/mo, which is significantly higher than the environmental thermal fluctuation background ( 2.5 kJ/mol); thus, complex formation should theoretically be observable. However, the experimental control showed that monosaccharide-functionalized gold nanoparticles do not undergo any aggregation. The author needs to provide a thorough and rigorous explanation for this quantitative contradiction.
The manuscript suffers from poor text preparation and severe text duplication. For instance, on Page 11, a long paragraph discussing the binding energy being in the same order of magnitude as thermal energy ("In the case of the calcium complex….") is repeated word-for-word in the consecutive text. Additionally, the structure of the conclusions section appears disorganized. A meticulous proofreading, language polishing, and structural refinement are mandatory for the entire manuscript.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

Yes
If applicable, is the statistical analysis and its interpretation appropriate?

Partly
Are all the source data underlying the results available to ensure full reproducibility?

Partly
Are the conclusions drawn adequately supported by the results?

No

Competing Interests: No competing interests were disclosed.

Reviewer Expertise: Controlled Synthesis and Applications of Inorganic Nanomaterials；Synthesis of Acrylic Polymers and Their Applications in Coatings；Development of Environmentally Friendly Marine Protective Coatings

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

CITE

Report a concern

Respond or Comment

Views

5

Reviewer Report 27 May 2026

Thuat T. Trinh, Norwegian University of Science and Technology, Trondheim, Norway

Approved with Reservations

https://doi.org/10.5256/f1000research.191145.r481858

This manuscript presents a molecular dynamics (MD) study investigating the binding mechanisms of cation-mediated carbohydrate-carbohydrate complex formation, specifically examining the differences between calcium (Ca²⁺) and magnesium (Mg²⁺) ions in facilitating aggregation of sulfated disaccharides. The work builds upon experimental observations ... Continue reading

This manuscript presents a molecular dynamics (MD) study investigating the binding mechanisms of cation-mediated carbohydrate-carbohydrate complex formation, specifically examining the differences between calcium (Ca²⁺) and magnesium (Mg²⁺) ions in facilitating aggregation of sulfated disaccharides. The work builds upon experimental observations that calcium ions induce aggregation while magnesium ions do not, and aims to elucidate the molecular-level mechanisms underlying this behavior. The authors employ density functional tight-binding (DFTB) simulations combined with radial distribution function (RDF) and potential of mean force (PMF) analyses to investigate cation-carbohydrate interactions. The study explores various simulation conditions including different cation concentrations, system sizes, and a modified water model for magnesium to test the hydration shell hypothesis.
The manuscript provides a systematic investigation of cation-carbohydrate interactions with reasonable coverage of relevant parameters. However, significant concerns exist regarding the simulation methodology, timescale sufficiency, and rigor of the computational approach that require careful consideration.

Comments:

1. The most significant concern regarding the computational methodology is the extremely short simulation duration of 15 ps for production runs. This time scale is orders of magnitude shorter than typical production MD simulations for biomolecular systems. While DFTB is more computationally demanding than classical force field MD, 15 ps is insufficient to adequately sample conformational space, especially for:

- Cation binding/unbinding events typically require microseconds to milliseconds of sampling for proper convergence.
- While RDFs can converge relatively quickly, the statistical uncertainties are likely large given the short sampling. Block averaging over multiple independent trajectories is essential.
- The carbohydrate-cation-carbohydrate binding energies reported (3.9-8.5 kJ/mol) are small and thermal fluctuations are significant. Longer simulations with independent replicas are needed to establish reliable binding free energy estimates.
The authors should extend the simulation time to at least 100-500 ps and run multiple independent replicas,

2. The use of a modified water model for magnesium simulations, where atomic partial charges are scaled to 83.6% of their original values, is a significant methodological choice that requires substantial justification and validation.
This approach raises several concerns:
-Scaling water charges arbitrarily modifies fundamental electrostatic properties including the dipole moment and dielectric constant. What is the theoretical justification for choosing the factor of 0.83 (rMg/rCa)?
-The authors should validate the modified water model against experimental properties of water and aqueous Mg²⁺ solutions (density, dielectric constant, hydration structure, ion-water binding energies) to demonstrate that it provides improved representation compared to the standard model.
-The modified model likely alters hydrogen bonding networks, proton transfer rates, and other water-mediated processes. The authors acknowledge (line 224) that this may cause unintended effects including reduced dielectric screening. A more rigorous approach would be to systematically study how water model selection affects the key observables.
Without this validation, conclusions drawn from the modified water model simulations must be treated with caution, as they may reflect artifacts of the ad hoc parameter modification rather than genuine physical effects.

3. For a reliable PMF calculation from RDFs, the RDF must be well-converged across all reaction coordinate values. With only 15 ps of simulation time and no independent replicas or block averaging reported, the statistical uncertainties are unknown and likely large.
- Binding energies of 3.9-8.5 kJ/mol are reported with no error bars or confidence intervals. Given the small magnitude relative to thermal energy (kBT ≈ 2.5 kJ/mol), statistical significance cannot be established without error analysis.
- More robust approaches for calculating binding free energies exist, including umbrella sampling, thermodynamic integration, or metadynamics. While these methods are more computationally demanding, they would provide more reliable quantitative estimates.

4. The manuscript compares the simulated binding energies (ΔV ≈ 5.4 kJ/mol lower bound from experimental rupture force) with experimental AFM pull experiments measuring rupture forces from reference 31 (Lorenz et al. 2012).While this comparison is useful, several issues require clarification:
- Rupture force measurements probe kinetic properties (energy barriers) while PMF calculations yield thermodynamic free energies. The relationship between these quantities is complex and depends on the pulling rate, cantilever spring constant, and molecular dynamics of the dissociation pathway.
- The treatment assumes a simple Bell model with a single energy barrier, which may not capture the more complex energy landscape of carbohydrate-cation-carbohydrate interactions.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

Yes
If applicable, is the statistical analysis and its interpretation appropriate?

Partly
Are all the source data underlying the results available to ensure full reproducibility?

Partly
Are the conclusions drawn adequately supported by the results?

No

Competing Interests: No competing interests were disclosed.

Reviewer Expertise: Atomistic simulations

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

CITE

Report a concern

Respond or Comment

Views

5

Reviewer Report 26 May 2026

Blake I. Armstrong, Curtin University, Perth, Australia

Not Approved

https://doi.org/10.5256/f1000research.191145.r481859

This work attempts to study the aggregation thermodynamics of disaccharides with divalent cations in aqueous solution using molecular dynamics simulations. Below, I have included some questions and statements regarding the quality of the work. I summarise my main criticisms at ... Continue reading

This work attempts to study the aggregation thermodynamics of disaccharides with divalent cations in aqueous solution using molecular dynamics simulations. Below, I have included some questions and statements regarding the quality of the work. I summarise my main criticisms at the end.

Intro
I'm a little confused by the lack of a sulphate group in the molecule presented in Figure 1. This seems like a critical error. Likewise, the caption suggests there is a 3D structure as well, but I do not see such a thing.
Simulation methods
What is 'thermostat 1'?
Was there no NPT equilibration before NVT production?
This part "binding energy of the ions with the system" is too vague and does not actually describe what property between what components is computed.
I'm not sure it is technically correct to say it is more "accurate" to only consider a part of the system that the MP2 method can handle.
Can you elaborate on exactly which part is the "cluster" and what it means to saturate the "edges" with hydrogen atoms. Perhaps this is obvious to others but it doesn't mean much to me.
20 cations in a box of size 3.7 nm is highly concentrated. Can the authors provide a justification for why they chose to create their system this way.
The water exchange rate around calcium ions is at least in the hundreds of picoseconds, and for magnesium it is orders of magnitude larger. I would strongly argue that 165 ps is certainly not long enough to probe the dynamics of solvated doubly charged ions; and, as such, their thermodynamic quantities.
I would also disagree with the claim that "traditional models for water cannot accurately capture the strength and nature of ion-water interactions", as there are classical and polarisable models for water with calcium & magnesium that do a sufficient job at reproducing their relevant thermodynamic properties (hydration free energy, solvation structure).
What exactly is "this change" that the authors have made to the model? The order of the sentences in the paragraph starting with "Table 1 shows..." does not flow logically. This "adjusting the charge distribution, the model more accurately reflect the dynamics of hydration shells around cations" is a strong claim and must be backed up by data demonstrating this is true, yet there is none. Given that water dynamics around magnesium are incredibly slow (on MD timescales), and that there is no mention of any enhanced sampling, I find it highly unlikely that this claim has any backing.
Results and discussion
The issue with this reasoning: "The reason for this is the smaller ionic radius of the magnesium ion compared to the calcium ion, which causes a solvated magnesium ion to bind the water molecules of the hydrate shell to itself more strongly than a calcium ion does.", is that, due to the relatively small simulation lengths and no enhanced sampling, the binding of ions to any given functional group will be extremely dependent on the random starting positions of these ions. This is because it takes a long time for ions to move and reorient themselves across large free-energy barriers between thermodynamic end states. Instead, the authors use the already known hydration enthalpies and water exchange rates as a post facto justification for incomplete data.

I believe the rest of the criticisms I will have for this work will just be a restatement of my previous critiques. To conclude and summarise: while the choice to analyse the simulations with radial distribution functions is sensible, the simulations being analysed do not sufficiently sample the thermally accessible phase space relevant to the aggregation of disaccharides with divalent cations. This is not to say that the aim of this work is not possible. I believe that the model chosen in this work is prohibitively computationally demanding for the simulation times required for sufficient analysis. Instead, I believe classical potentials (or perhaps machine-learned potentials) would be more useful, with the caveat that the chosen model be sufficiently parameterised to reproduce relevant solvation thermodynamics of the relevant species. Likewise, even with classical force fields, enhanced sampling techniques like metadynamics or OPES will be required to explore the phase space, especially for magnesium.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

No
Are sufficient details of methods and analysis provided to allow replication by others?

Partly
If applicable, is the statistical analysis and its interpretation appropriate?

No
Are all the source data underlying the results available to ensure full reproducibility?

No
Are the conclusions drawn adequately supported by the results?

No

Competing Interests: No competing interests were disclosed.

Reviewer Expertise: Computational chemistry, force fields, geochemistry, crystal growth, free energy calculations

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.

CITE

Report a concern

Respond or Comment

Views

14

Reviewer Report 04 May 2026

Gonzalo R. Quezada, Universidad del Bío-Bío, Concepción, Chile

Approved with Reservations

https://doi.org/10.5256/f1000research.191145.r475047

The manuscript addresses an interesting question regarding Ca2+ and Mg2+ mediated carbohydrate-cation-carbohydrate complex formation. However, major revisions are required.

The authors should first clarify if there is no problem between this manuscript and a previous preprint version with ... Continue reading

The manuscript addresses an interesting question regarding Ca2+ and Mg2+ mediated carbohydrate-cation-carbohydrate complex formation. However, major revisions are required.

The authors should first clarify if there is no problem between this manuscript and a previous preprint version with the same title and authors.

The computational methodology needs to be described in much greater detail. It is unclear whether the simulations are fully DFTB-based, classical MD, or a combination of both. Key information is missing, including force-field/DFTB parameters, ion treatment, water model, timestep, electrostatics, ensemble, thermostat/barostat, equilibration protocol, number of replicas, and convergence criteria.

The simulation time of 15 ps appears insufficient to get good RDF, PMF graphs. Longer simulations, independent replicas, block averaging, and statistical uncertainties should be provided.

The PMF derived directly from RDFs requires stronger justification and convergence analysis. The reported binding energies are small, so error estimates are essential.

The modified water model for Mg2+ also requires validation, since scaling water charges may alter hydration, dielectric screening, and ion-carbohydrate interactions in a nonphysical way.

Finally, the conclusions should be softened. The results suggest a plausible mechanism, but they do not yet fully support the strong claims made. Source data, input files, structures, and processed RDF/PMF data should be made available to ensure reproducibility.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

No
If applicable, is the statistical analysis and its interpretation appropriate?

No
Are all the source data underlying the results available to ensure full reproducibility?

No
Are the conclusions drawn adequately supported by the results?

Partly

Competing Interests: No competing interests were disclosed.

Reviewer Expertise: chemical engineer science, computer simulations, solid-liquids interfaces, polymer science.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

CITE

Report a concern

Respond or Comment

Comments on this article Comments (0)

Version 1

VERSION 1 PUBLISHED 10 Apr 2026

Open Peer Review

Reviewer Status

Reviewer Reports

	Invited Reviewers
	1	2	3	4	5
Version 1 10 Apr 26	read	read	read	read	read

Gonzalo R. Quezada, Universidad del Bío-Bío, Concepción, Chile
Blake I. Armstrong, Curtin University, Perth, Australia
Thuat T. Trinh, Norwegian University of Science and Technology, Trondheim, Norway
Lei Dong, Ocean University of China, Qingdao, China
Weiwei He, New York University Abu Dhabi (Ringgold ID: 167632), Abu Dhabi, United Arab Emirates

Comments on this article

All Comments(0)

Add a comment

Sign up for content alerts

Browse by related subjects

Back to all reports

Reviewer Report

6 Views

03 Jun 2026 | for Version 1

Weiwei He, Chemistry, New York University Abu Dhabi (Ringgold ID: 167632), Abu Dhabi, Abu Dhabi, United Arab Emirates

6 Views Cite this report Responses(0)

Approved With Reservations

1. Typographical and formatting issues should be corrected throughout the manuscript.
The manuscript contains multiple typographical, formatting, and citation errors that affect readability. For example, the phrase “magnesium binds the hydrate shell more strongly than calcium 1 2” appears to contain broken citation formatting. The authors should carefully check the entire manuscript, correct all citation-formatting errors, remove duplicated text, and ensure consistent use of chemical terms, symbols, units, and figure labels.

2. The artificial modification of the Mg²⁺–water interaction is not sufficiently justified.
A central part of the manuscript relies on a modified Mg system, in which water charges are scaled to weaken the Mg²⁺ hydration shell. This modification is highly artificial, because changing water charges affects not only Mg²⁺ hydration, but also the water dipole moment, dielectric screening, hydrogen bonding, ion–ion interactions, and sulfate–sulfate electrostatics.

The Mg* model therefore cannot be used as direct evidence for the proposed mechanism without systematic parameter refinement, benchmarking, and physical validation. The authors are encouraged to perform additional simulations using existing, validated Mg²⁺, Ca²⁺, and Na⁺ ion parameters developed for classical biomolecular force fields, especially those tested for nucleic acid or phosphate/sulfate-rich systems. Relevant examples include DOI: 10.1021/jz201501a, 10.1021/acs.jctc.3c00520, and 10.1002/bip.22868. Such validation is needed to determine whether the reported Ca²⁺/Mg²⁺ selectivity is robust or model-dependent.

3. The computational model and simulation protocol require substantial clarification.
The manuscript states that the MD simulations were performed using DFTB+, but the actual molecular model is not sufficiently described. The authors should clarify whether the simulations are fully DFTB-based, classical MD, or a hybrid workflow involving DFTB-MD and MP2 cluster calculations.

More details are needed on how water, ions, saccharides, and sulfate groups were modeled. For example, what water description was used, and how are the Mg²⁺, Ca²⁺, Na⁺, and Cl⁻ ions treated in the DFTB framework? The authors should also provide the relevant parameters, timestep, convergence criteria, etc. Without these details, the simulations are difficult to reproduce or critically assess.

4. The simulation time is too short to support reliable RDF, PMF, and aggregation conclusions.
The main RDF and PMF analyses appear to be based on very short simulations, approximately 15 ps for the multi-saccharide systems. This is likely insufficient for sampling divalent-ion dehydration, sulfate coordination, carbohydrate reorientation, solvent relaxation, and supramolecular aggregation.

5. The PMF-derived binding energies require stronger justification and uncertainty estimates.
The manuscript derives PMFs directly from RDFs and interprets the resulting minima as binding energies. This treatment requires much stronger justification, especially because the reported binding energies are small and comparable to only a few kBT.

The authors should provide convergence analysis and uncertainty estimates for the RDF-derived PMFs. They should also clarify that an RDF-derived PMF in a multi-component, finite-concentration system is an effective potential rather than a rigorous standard binding free energy. To obtain quantitative binding free energies, enhanced-sampling methods such as umbrella sampling or metadynamics are recommended, ideally using reaction coordinates that distinguish direct cation bridging from water-mediated or non-bridged contacts.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

No
If applicable, is the statistical analysis and its interpretation appropriate?

No
Are all the source data underlying the results available to ensure full reproducibility?

No
Are the conclusions drawn adequately supported by the results?

Partly

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

Computational chemistry, force fields, MD simulations

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Respond to this report

Responses (0)

Back to all reports

Reviewer Report

2 Views

02 Jun 2026 | for Version 1

Lei Dong, Ocean University of China, Qingdao, China

2 Views Cite this report Responses(0)

Approved With Reservations

The author presents a molecular dynamics simulation study focusing on the supramolecular complexes of carbohydrate-cation systems. This research aims to elucidate the microscopic mechanisms underlying the cation-mediated aggregation of disaccharides, specifically exploring why calcium ions (Ca²⁺) induce stable disaccharide aggregation while magnesium ions (Mg²⁺) do not. The author proposes a mechanistic framework based on the kinetic hindrance of the ion hydration shell and direct/indirect coordination with sulfate groups. While the manuscript presents interesting findings, major revisions are required to address the following concerns:

There is a critical error in the chemical structure presentation in Figure 1. Despite the heavy emphasis throughout the manuscript on the sulfate group as the core binding site, no sulfate groups are depicted in the chemical structures of the disaccharides in Figure 1. Furthermore, while the figure caption claims to display both 2D and 3D structures, only 2D skeletal formulas are visible in the figure. The author must completely redraw Figure 1 to ensure technical accuracy and consistency between the text and the illustration.
The simulation timescale is severely insufficient. The production run lasted for only 15 ps, which is far too short to achieve thermodynamic equilibrium for flexible carbohydrate chains characterized by rich conformational degrees of freedom. Such limited sampling compromises the reliability and accuracy of the resulting radial distribution functions (RDF) and potentials of mean force (PMF). The author is strongly encouraged to extend the simulation timescale to the nanosecond or microsecond range.
The physical validity of the modified water model is highly questionable. To artificially weaken the hydration shell of magnesium ions, the author scaled down the atomic charges of water to 83.6% (referred to as the Mg* model). This ad hoc modification alters the intrinsic physical properties of bulk water, such as its dipole moment and dielectric constant. The author must provide comprehensive benchmark data evaluating how this modification impacts the fundamental physical properties of water.
There is a logical disconnection between the simulated binding energies and the observed macroscopic phenomena. The simulation indicates that the monosaccharide complex still possesses a binding energy of approximately 6kJ/mo, which is significantly higher than the environmental thermal fluctuation background ( 2.5 kJ/mol); thus, complex formation should theoretically be observable. However, the experimental control showed that monosaccharide-functionalized gold nanoparticles do not undergo any aggregation. The author needs to provide a thorough and rigorous explanation for this quantitative contradiction.
The manuscript suffers from poor text preparation and severe text duplication. For instance, on Page 11, a long paragraph discussing the binding energy being in the same order of magnitude as thermal energy ("In the case of the calcium complex….") is repeated word-for-word in the consecutive text. Additionally, the structure of the conclusions section appears disorganized. A meticulous proofreading, language polishing, and structural refinement are mandatory for the entire manuscript.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

Yes
If applicable, is the statistical analysis and its interpretation appropriate?

Partly
Are all the source data underlying the results available to ensure full reproducibility?

Partly
Are the conclusions drawn adequately supported by the results?

No

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

Controlled Synthesis and Applications of Inorganic Nanomaterials；Synthesis of Acrylic Polymers and Their Applications in Coatings；Development of Environmentally Friendly Marine Protective Coatings

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Respond to this report

Responses (0)

Back to all reports

Reviewer Report

5 Views

27 May 2026 | for Version 1

Thuat T. Trinh, Norwegian University of Science and Technology, Trondheim, Norway

5 Views Cite this report Responses(0)

Approved With Reservations

This manuscript presents a molecular dynamics (MD) study investigating the binding mechanisms of cation-mediated carbohydrate-carbohydrate complex formation, specifically examining the differences between calcium (Ca²⁺) and magnesium (Mg²⁺) ions in facilitating aggregation of sulfated disaccharides. The work builds upon experimental observations that calcium ions induce aggregation while magnesium ions do not, and aims to elucidate the molecular-level mechanisms underlying this behavior. The authors employ density functional tight-binding (DFTB) simulations combined with radial distribution function (RDF) and potential of mean force (PMF) analyses to investigate cation-carbohydrate interactions. The study explores various simulation conditions including different cation concentrations, system sizes, and a modified water model for magnesium to test the hydration shell hypothesis.
The manuscript provides a systematic investigation of cation-carbohydrate interactions with reasonable coverage of relevant parameters. However, significant concerns exist regarding the simulation methodology, timescale sufficiency, and rigor of the computational approach that require careful consideration.

Comments:

1. The most significant concern regarding the computational methodology is the extremely short simulation duration of 15 ps for production runs. This time scale is orders of magnitude shorter than typical production MD simulations for biomolecular systems. While DFTB is more computationally demanding than classical force field MD, 15 ps is insufficient to adequately sample conformational space, especially for:

- Cation binding/unbinding events typically require microseconds to milliseconds of sampling for proper convergence.
- While RDFs can converge relatively quickly, the statistical uncertainties are likely large given the short sampling. Block averaging over multiple independent trajectories is essential.
- The carbohydrate-cation-carbohydrate binding energies reported (3.9-8.5 kJ/mol) are small and thermal fluctuations are significant. Longer simulations with independent replicas are needed to establish reliable binding free energy estimates.
The authors should extend the simulation time to at least 100-500 ps and run multiple independent replicas,

2. The use of a modified water model for magnesium simulations, where atomic partial charges are scaled to 83.6% of their original values, is a significant methodological choice that requires substantial justification and validation.
This approach raises several concerns:
-Scaling water charges arbitrarily modifies fundamental electrostatic properties including the dipole moment and dielectric constant. What is the theoretical justification for choosing the factor of 0.83 (rMg/rCa)?
-The authors should validate the modified water model against experimental properties of water and aqueous Mg²⁺ solutions (density, dielectric constant, hydration structure, ion-water binding energies) to demonstrate that it provides improved representation compared to the standard model.
-The modified model likely alters hydrogen bonding networks, proton transfer rates, and other water-mediated processes. The authors acknowledge (line 224) that this may cause unintended effects including reduced dielectric screening. A more rigorous approach would be to systematically study how water model selection affects the key observables.
Without this validation, conclusions drawn from the modified water model simulations must be treated with caution, as they may reflect artifacts of the ad hoc parameter modification rather than genuine physical effects.

3. For a reliable PMF calculation from RDFs, the RDF must be well-converged across all reaction coordinate values. With only 15 ps of simulation time and no independent replicas or block averaging reported, the statistical uncertainties are unknown and likely large.
- Binding energies of 3.9-8.5 kJ/mol are reported with no error bars or confidence intervals. Given the small magnitude relative to thermal energy (kBT ≈ 2.5 kJ/mol), statistical significance cannot be established without error analysis.
- More robust approaches for calculating binding free energies exist, including umbrella sampling, thermodynamic integration, or metadynamics. While these methods are more computationally demanding, they would provide more reliable quantitative estimates.

4. The manuscript compares the simulated binding energies (ΔV ≈ 5.4 kJ/mol lower bound from experimental rupture force) with experimental AFM pull experiments measuring rupture forces from reference 31 (Lorenz et al. 2012).While this comparison is useful, several issues require clarification:
- Rupture force measurements probe kinetic properties (energy barriers) while PMF calculations yield thermodynamic free energies. The relationship between these quantities is complex and depends on the pulling rate, cantilever spring constant, and molecular dynamics of the dissociation pathway.
- The treatment assumes a simple Bell model with a single energy barrier, which may not capture the more complex energy landscape of carbohydrate-cation-carbohydrate interactions.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

Yes
If applicable, is the statistical analysis and its interpretation appropriate?

Partly
Are all the source data underlying the results available to ensure full reproducibility?

Partly
Are the conclusions drawn adequately supported by the results?

No

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

Atomistic simulations

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Respond to this report

Responses (0)

Back to all reports

Reviewer Report

5 Views

26 May 2026 | for Version 1

Blake I. Armstrong, Curtin University, Perth, Australia

5 Views Cite this report Responses(0)

Not Approved

This work attempts to study the aggregation thermodynamics of disaccharides with divalent cations in aqueous solution using molecular dynamics simulations. Below, I have included some questions and statements regarding the quality of the work. I summarise my main criticisms at the end.

Intro
I'm a little confused by the lack of a sulphate group in the molecule presented in Figure 1. This seems like a critical error. Likewise, the caption suggests there is a 3D structure as well, but I do not see such a thing.
Simulation methods
What is 'thermostat 1'?
Was there no NPT equilibration before NVT production?
This part "binding energy of the ions with the system" is too vague and does not actually describe what property between what components is computed.
I'm not sure it is technically correct to say it is more "accurate" to only consider a part of the system that the MP2 method can handle.
Can you elaborate on exactly which part is the "cluster" and what it means to saturate the "edges" with hydrogen atoms. Perhaps this is obvious to others but it doesn't mean much to me.
20 cations in a box of size 3.7 nm is highly concentrated. Can the authors provide a justification for why they chose to create their system this way.
The water exchange rate around calcium ions is at least in the hundreds of picoseconds, and for magnesium it is orders of magnitude larger. I would strongly argue that 165 ps is certainly not long enough to probe the dynamics of solvated doubly charged ions; and, as such, their thermodynamic quantities.
I would also disagree with the claim that "traditional models for water cannot accurately capture the strength and nature of ion-water interactions", as there are classical and polarisable models for water with calcium & magnesium that do a sufficient job at reproducing their relevant thermodynamic properties (hydration free energy, solvation structure).
What exactly is "this change" that the authors have made to the model? The order of the sentences in the paragraph starting with "Table 1 shows..." does not flow logically. This "adjusting the charge distribution, the model more accurately reflect the dynamics of hydration shells around cations" is a strong claim and must be backed up by data demonstrating this is true, yet there is none. Given that water dynamics around magnesium are incredibly slow (on MD timescales), and that there is no mention of any enhanced sampling, I find it highly unlikely that this claim has any backing.
Results and discussion
The issue with this reasoning: "The reason for this is the smaller ionic radius of the magnesium ion compared to the calcium ion, which causes a solvated magnesium ion to bind the water molecules of the hydrate shell to itself more strongly than a calcium ion does.", is that, due to the relatively small simulation lengths and no enhanced sampling, the binding of ions to any given functional group will be extremely dependent on the random starting positions of these ions. This is because it takes a long time for ions to move and reorient themselves across large free-energy barriers between thermodynamic end states. Instead, the authors use the already known hydration enthalpies and water exchange rates as a post facto justification for incomplete data.

I believe the rest of the criticisms I will have for this work will just be a restatement of my previous critiques. To conclude and summarise: while the choice to analyse the simulations with radial distribution functions is sensible, the simulations being analysed do not sufficiently sample the thermally accessible phase space relevant to the aggregation of disaccharides with divalent cations. This is not to say that the aim of this work is not possible. I believe that the model chosen in this work is prohibitively computationally demanding for the simulation times required for sufficient analysis. Instead, I believe classical potentials (or perhaps machine-learned potentials) would be more useful, with the caveat that the chosen model be sufficiently parameterised to reproduce relevant solvation thermodynamics of the relevant species. Likewise, even with classical force fields, enhanced sampling techniques like metadynamics or OPES will be required to explore the phase space, especially for magnesium.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

No
Are sufficient details of methods and analysis provided to allow replication by others?

Partly
If applicable, is the statistical analysis and its interpretation appropriate?

No
Are all the source data underlying the results available to ensure full reproducibility?

No
Are the conclusions drawn adequately supported by the results?

No

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

Computational chemistry, force fields, geochemistry, crystal growth, free energy calculations

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.

Respond to this report

Responses (0)

Back to all reports

Reviewer Report

14 Views

04 May 2026 | for Version 1

Gonzalo R. Quezada, Universidad del Bío-Bío, Concepción, Chile

14 Views Cite this report Responses(0)

Approved With Reservations

The manuscript addresses an interesting question regarding Ca2+ and Mg2+ mediated carbohydrate-cation-carbohydrate complex formation. However, major revisions are required.

The authors should first clarify if there is no problem between this manuscript and a previous preprint version with the same title and authors.

The computational methodology needs to be described in much greater detail. It is unclear whether the simulations are fully DFTB-based, classical MD, or a combination of both. Key information is missing, including force-field/DFTB parameters, ion treatment, water model, timestep, electrostatics, ensemble, thermostat/barostat, equilibration protocol, number of replicas, and convergence criteria.

The simulation time of 15 ps appears insufficient to get good RDF, PMF graphs. Longer simulations, independent replicas, block averaging, and statistical uncertainties should be provided.

The PMF derived directly from RDFs requires stronger justification and convergence analysis. The reported binding energies are small, so error estimates are essential.

The modified water model for Mg2+ also requires validation, since scaling water charges may alter hydration, dielectric screening, and ion-carbohydrate interactions in a nonphysical way.

Finally, the conclusions should be softened. The results suggest a plausible mechanism, but they do not yet fully support the strong claims made. Source data, input files, structures, and processed RDF/PMF data should be made available to ensure reproducibility.

Is the work clearly and accurately presented and does it cite the current literature?

Partly
Is the study design appropriate and is the work technically sound?

Partly
Are sufficient details of methods and analysis provided to allow replication by others?

No
If applicable, is the statistical analysis and its interpretation appropriate?

No
Are all the source data underlying the results available to ensure full reproducibility?

No
Are the conclusions drawn adequately supported by the results?

Partly

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

chemical engineer science, computer simulations, solid-liquids interfaces, polymer science.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Respond to this report

Responses (0)

[1] 1. Cailliez F, Lavery R: Cadherin mechanics and complexation: the importance of calcium binding. Biophys. J. 2005; 89: 3895–3903. PubMed Abstract | Publisher Full Text | Free Full Text

[2] 2. Sotomayor M, Schulten K: The allosteric role of the Ca2+ switch in adhesion and elasticity of C-cadherin. Biophys. J. 2008; 94: 4621–4633. PubMed Abstract | Publisher Full Text | Free Full Text

[3] 3. Steitz TA, Steitz JA: A general two-metal-ion mechanism for catalytic RNA. Proc. Natl. Acad. Sci. 1993; 90: 6498–6502. PubMed Abstract | Publisher Full Text | Free Full Text

[4] 4. Mitchell JM, Nichols SA: Diverse cell junctions with unique molecular composition in tissues of a sponge (Porifera). EvoDevo. 2019; 10: 1–16. PubMed Abstract | Publisher Full Text | Free Full Text

[5] 5. González-Amaro R, Sánchez-Madrid F: Cell adhesion molecules: selectins and integrins. Crit. Rev. Immunol. 1999; 19: 389–429. PubMed Abstract

[6] 6. De Kerchove AJ, Elimelech M: Calcium and magnesium cations enhance the adhesion of motile and nonmotile Pseudomonas aeruginosa on alginate films. Langmuir. 2008; 24: 3392–3399. PubMed Abstract | Publisher Full Text

[7] 7. Kafi DK, Mohi AT: Molecular dynamics study of supramolecular complexes to the carbohydrate. Research Square Preprint. 2024. Publisher Full Text

[8] 8. Hourahine B, Aradi B, Blum V, et al.: DFTB+, a software package for efficient approximate density functional theory based atomistic simulations. J. Chem. Phys. 2020; 152: 124101. PubMed Abstract | Publisher Full Text

[9] 9. Zeng J, Zhang R-Q, Treutlein H: Quantum simulations of materials and biological systems. Springer Science & Business Media; 2012. Publisher Full Text

[10] 10. Frauenheim T, Aradi B: Atomistic Simulations Using the Approximate DFT Method DFTB+: Applications to Nanomaterials and Bio-Systems. Multiscale Modelling Methods for Applications in Materials Science: CECAM Tutorial, Forschungszentrum Jülich. 2013; pp. 41.

[11] 11. Lourenço MP, da Silva MC , Oliveira AF, et al.: FASP: a framework for automation of Slater–Koster file parameterization. Theor. Chem. Accounts. 2016; 135: 1–12. Publisher Full Text

[12] 12. Frauenheim T, Seifert G, Elsterner M, et al.: A Self-Consistent Charge Density-Functional Based Tight-Binding Method for Predictive Materials Simulations in Physics. Chemistry and Biology, physica status solidi (b). 2000; 217: 41–62. Publisher Full Text

[13] 13. Zhechkov L, Heine T, Patchkovskii S, et al.: An efficient a posteriori treatment for dispersion interaction in density-functional-based tight binding. J. Chem. Theory Comput. 2005; 1: 841–847. PubMed Abstract | Publisher Full Text

[14] 14. Porezag D, Jungnickel G, Frauenheim T, et al.: Theoretical investigations of homo- and heteronuclear bridged fullerene oligomers. Appl. Phys. A. 1997; 64: 321–326. Publisher Full Text

[15] 15. Mandumpal J, Gemming S, Seifert G: Curvature effects of nitrogen on graphitic sheets: Structures and energetics. Chem. Phys. Lett. 2007; 447: 115–120. Publisher Full Text

[16] 16. Porezag D, Frauenheim T, Köhler T, et al.: Construction of tight-binding-like potentials on the basis of density-functional theory: Application to carbon. Phys. Rev. B. 1995; 51: 12947–12957. PubMed Abstract | Publisher Full Text

[17] 17. Albertazzi E, Domene C, Fowler PW, et al.: Pentagon adjacency as a determinant of fullerene stability. Phys. Chem. Chem. Phys. 1999; 1: 2913–2918. Publisher Full Text

[18] 18. Zheng GJ, Sonnenberg L, Hada M, et al.: Gaussian 09. Wallingford CT: Gaussian Inc.; 2009; 48.

[19] 19. Chen H, Cox JR, Panagiotopoulos AZ: Force fields for carbohydrate–divalent cation interactions. J. Phys. Chem. B. 2016; 120: 5203–5208. PubMed Abstract | Publisher Full Text

[20] 20. Remko M, Van Duijnen PT, Broer R: Effect of metal ions (Li+, Na+, K+, Mg 2+ and Ca 2+) and water on the conformational changes of glycosidic bonds in heparin oligosaccharides. RSC Adv. 2013; 3: 9843–9853. Publisher Full Text

[21] 21. Yan H, Yuan S-L, Xu G-Y, et al.: Effect of Ca2+ and Mg2+ ions on surfactant solutions investigated by molecular dynamics simulation. Langmuir. 2010; 26: 10448–10459. PubMed Abstract | Publisher Full Text

[22] 22. Aakesson R, Pettersson LGM, Sandstroem M, et al.: Theoretical study on water-exchange reactions of the divalent and trivalent metal ions of the first transition period. J. Am. Chem. Soc. 1994; 116: 8705–8713. Publisher Full Text

[23] 23. Da Silva JJRF, Williams RJP: The biological chemistry of the elements: the inorganic chemistry of life. Oxford University Press; 2001.

[24] 24. Mejıas JA, Lago S: Calculation of the absolute hydration enthalpy and free energy of H+ and OH−. J. Chem. Phys. 2000; 113: 7306–7316. Publisher Full Text

[25] 25. Schutte-Smith M, Roodt A, Visser HG: Ambient and high-pressure kinetic investigation of methanol substitution in fac-[Re (Trop)(CO) 3 (MeOH)] by different monodentate nucleophiles. Dalton Trans. 2019; 48: 9984–9997. Publisher Full Text

[26] 26. Mason WR: Electronic absorption and magnetic circular dichroism of the linear diaquadimercury (I) cation. Inorg. Chem. 1983; 22: 147–150. Publisher Full Text

[27] 27. Huynh UTD, Lerbret A, Neiers F, et al.: Binding of divalent cations to polygalacturonate: A mechanism driven by the hydration water. J. Phys. Chem. B. 2016; 120: 1021–1032. PubMed Abstract | Publisher Full Text

[28] 28. Derbyshire W, Hedges ND, Lillford PJ, et al.: The influence of a mixed anionic system on the aggregation behaviour of agarose. Food Hydrocoll. 2001; 15: 153–163. Publisher Full Text

[29] 29. Meeron E: Theory of potentials of average force and radial distribution functions in ionic solutions. J. Chem. Phys. 1958; 28: 630–643. Publisher Full Text

[30] 30. Kawada S, Fujimoto K, Yoshii N, et al.: Molecular dynamics study of the potential of mean force of SDS aggregates. J. Chem. Phys. 2017; 147. Publisher Full Text

[31] 31. Lorenz BR, de Cienfuegos LA , Oelkers M, et al.: Model system for cell adhesion mediated by weak carbohydrate–carbohydrate interactions. J. Am. Chem. Soc. 2012; 134: 3326–3329. PubMed Abstract | Publisher Full Text | Free Full Text

[32] 32. Carvalho de Souza A, Halkes KM, Meeldijk JD, et al.: Gold glyconanoparticles as probes to explore the carbohydrate-mediated self-recognition of marine sponge cells. ChemBioChem. 2005; 6: 828–831. PubMed Abstract | Publisher Full Text

[33] 33. Chen Q, Cui Y, Zhang T-L, et al.: Fluorescent Conjugated Polyfluorene with Pendant Lactopyranosyl Ligands for Studies of Ca2+−Mediated Carbohydrate− Carbohydrate Interaction. Biomacromolecules. 2010; 11: 13–19. PubMed Abstract | Publisher Full Text

Molecular Dynamics Study of Supramolecular Complexes to the Carbohydrate-Cation System

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Figure 1. 2D and 3D structure of the sulfated disaccharide GlcpNAc3S(β1–3)Fucp.

Simulation methods

Table 1. Number of saccharide molecules, cations (Ca2+ and Mg2+,) and counterions (Cl−) used in the simulations.

Table 2. Number of disaccharide molecules, cations (Ca2+ or Mg2+) and counterions (Cl−), that were used in the simulations with the larger simulation box.

Results and discussion

Preliminary investigation

Table 3. The experimentally determined values as well as the cation radius of both types of cations.

Determination of the RDF

Table 4. The quantities of disaccharide molecules and cations employed in the simulations.

RDF of sugar rings

Figure 2. Possible arrangements of disaccharides in the complex.

Figure 3. Radial distribution function with respect to the centers of gravity of the sugar rings.

RDF of sulfate groups and cations

Figure 4. Radial distribution function with respect to the cations and the acidic atoms of the Sulfate groups.

RDF of sulfate groups

Figure 5. gSul-Sul for the calcium system.

Potential of mean force

(1)

Figure 6. Potential of mean for the carbohydrate-cation-carbohydrate complex, where reaction coordinate, the distance between the sulfate groups is used.

Table 5. Quantitative summary of binding energies and characteristic distances from RDF and PMF simulations.

Comparison with a monosaccharide system

Figure 7. Monosaccharide that consists only of the glucose unit of the disaccharide.

Figure 8. Potential of mean force for the carbohydrate-calcium-carbohydrate complex for the disaccharide and monosaccharide system.

Conclusion

Data Availability

References

Comments on this article Comments (0)

Open Peer Review

Comments on this article Comments (0)

Open Peer Review

Reviewer Status

Reviewer Reports

Comments on this article

Browse by related subjects

Competing Interests Policy

Stay Updated