Influence of Coating and Size of Magnetic Nanoparticles on Cellular Uptake for In Vitro MRI

Round 1

Reviewer 1 Report

The manuscript describes an investigation of uptake by cells of iron oxide nanoparticles (IONPs). While the experiments have been well described, the findings and conclusions will be more clear and better supported after some technical issues have been clarified. After a revision, the manuscript may be considered for publication.

1. The methodology described on pp. 7-9 relies on indirect methods to infer the interactions that take place at surfaces of the nanoparticles. It would be beneficial to compare the interpretation against the quantitative assessments performed by multiple complementary techniques on similar systems, e.g., of protein attachment in DOI 10.1039/c6nr01732k. IONPs in a similar size range have been prepared and characterized (including using magnetic techniques) as described in DOI 10.1039/C5RA08200E and DOI 10.1016/j.jcis.2016.03.040, followed by their use in imaging and tracking of brain cells (DOI 10.1021/acs.bioconjchem.6b00522). Comparison to these studies that included a broader range of methods can help to elucidate, for example, the relative contributions from adsorbates and from colloidal stability when interpreting the DLS-derived hydrodynamic sizes.
2. Centrifugation is a useful technique, but it requires careful interpretation when applied to heterogeneous systems, such as mixtures of cells and nanoparticles, e.g., as discussed in DOI 10.1039/c8en01378k. In particular, as DLS suggested some degree of aggregation in the cell media, did the authors check for any additional aggregation induced during centrifugation? Even colloidally stable suspensions of IONPs can become destabilized under centrifugation at as low as 1000 rpm (DOI 10.1021/ac503835a).
3. The SEM images in Figure S2d are unclear. It would be helpful to show them at a larger size and resolution, so that the features related to IONPs can be easier to see. In particular, it is not clear why the objects, which appear at different sizes and roughness in different SEM panels, support the conclusion that the IONPs are internalized. SEM images in the previously mentioned DOI 10.1021/ac503835a, for example, show a "rough" appearance produced by IONPs external to the cells. Optical microscopy and TEM, for example, in DOI 10.1116/1.5009989, also show that IONPs can be both internalized by mammalian cells and attach externally as aggregates. It would be important to evaluate whether the evidence provided in this manuscript supports unambiguous and exclusive internalization of the investigated IONPs, or if some combination of internalization and external interactions may be at play (aggregates external to cells are clearly seen in some of the optical images in Figure S2).

Author Response

Referee1:

(responses with figures are included in attached Word file)

The manuscript describes an investigation of uptake by cells of iron oxide nanoparticles (IONPs). While the experiments have been well described, the findings and conclusions will be clearer and better supported after some technical issues have been clarified. After a revision, the manuscript may be considered for publication.

1. The methodology described on pp. 7-9 relies on indirect methods to infer the interactions that take place at surfaces of the nanoparticles. It would be beneficial to compare the interpretation against the quantitative assessments performed by multiple complementary techniques on similar systems, e.g., of protein attachment in DOI 10.1039/c6nr01732k. IONPs in a similar size range have been prepared and characterized (including using magnetic techniques) as described in DOI 10.1039/C5RA08200E and DOI 10.1016/j.jcis.2016.03.040, followed, followed by their use in imaging and tracking of brain cells (DOI 10.1021/acs.bioconjchem.6b00522). Comparison to these studies that included a broader range of methods can help to elucidate, for example, the relative contributions from adsorbates and from colloidal stability when interpreting the DLS-derived hydrodynamic sizes.

We thank the reviewer for drawing our attention to references 10.1039/c6nr01732k and 10.1021/acs.bioconjchem.6b00522. They have been useful to improve and expand the discussion on page 7 with the following added text: “It is known that the interaction of nanoparticles with proteins can increase the D_hyd either by causing aggregation or by simple adhesion. The later has been purposely employed for engineering nanoparticles and, alone, cannot account for large increases in D_hyd values like the ones observed here. On the other hand, aggregation of nanoparticles in protein-rich media has been suggested to rise the local concentration of superparamagnetic iron oxide nanoparticles in mesenchymal stem cells. This feature must be controlled or monitored, since it can complicate cell tracking by MRI.”

2. Centrifugation is a useful technique, but it requires careful interpretation when applied to heterogeneous systems, such as mixtures of cells and nanoparticles, e.g., as discussed in DOI 10.1039/c8en01378k. In particular, as DLS suggested some degree of aggregation in the cell media, did the authors check for any additional aggregation induced during centrifugation? Even colloidally stable suspensions of IONPs can become destabilized under centrifugation at as low as 1000 rpm (DOI 10.1021/ac503835a).

            We appreciate the reviewer's comment about this subject. CMI method was carried out in our group and published by Ocampo, et al. (Reference 26 in the new and edited manuscript). We studied different cells and IONPs concentrations, centrifugal forces and times. Figure 4B shows the labelling efficiency of using different centrifugation times and concentrations. They observed that IONPs were uptaken by the cells and that the labelling efficiency increased with the centrifugal speed and time.

Figure 4B: Ocampo, et al. Scientific Report, 2015.

Our DLS measurements showed that IONPs suspended in DMEM without FBS go under aggregation, increasing the hydrodynamic size (Figure 1B). Therefore, for an initial concentration of 50 µg/ml, we choose a centrifugation time of 1 minute for IONPs that were suspended in DMEM + 0% FBS and 5 minutes for cells suspended in DMEM + 10%FBS (See method for cellular uptake in the manuscript). We reduced the centrifugal time for IONPs resuspended in a media without FBS (from 5 to 1 minute) to avoid aggregations, but we also wanted to have good labelling efficiency as Ocampo, et al. observed for centrifugal times of 1 minute. 

Moreover, we also published a mathematical model explaining the spatial-temporal dynamics of IONPs and their interaction with cells when they are under centrifugation or gravity forces, Fernández-Calvo, et al., (reference 27 in the new and edited manuscript). In this model, we performed simulations to study the sedimentation and diffusion coefficients for IONPs under centrifugal forces using the Lamm equation. The first simulations took into consideration how NPs interact with cellular media and under gravitational forces (no cells). Since our experimental method uses centrifugal forces, the simulations depend on the angular velocity (w). Depending on its value there are two pathways, sedimentation velocity and sedimentation equilibrium. For the first one, IONPs usually are under higher angular speeds (w>10⁴), producing precipitation of the IONPs and can be completely separated from the solvent. The second pathway uses lower speeds, making the diffusion forces more important. Since for our experiments we used w=1500 rpm, we followed the second pathway. In fact, Sousa, et al. (DOI 10.1021/ac503835a) showed a destabilization of the IONPs at centrifugation of 4000 rpm, higher speeds than the one we used in our study. To fit the model with the experiments, we calculated the sedimentation and diffusion coefficients from our measured hydrodynamics sizes from the DLS measurements. Here are the parameters used on the simulations (Table 1, Fernández-Calvo, et al. 2020).

Table 1 and 2, G.F. Calvo, et al., Applied Mathematical Modelling, 2020

The following plots represent the concentration of IONPs over time, using the same parameters as described in Table 1 and 2. Here is clear the exponential growth of the IONPs concentration at the end of the container, something we need to have the IONPs close to the cellular membrane to induce their internalization.

Figure 2, G.F. Calvo, et al. Applied Mathematical Modelling, 2020.

We also compared the simulations with our experiments. Figure 3 from Fernández-Calvo, et al. shows the fitting between the experimental data and the numerical simulations from all the IONPs with different coatings that were under centrifugal forces by CMI method.

Figure 3, G.F. Calvo, et al. Mathematical Modelling, 2020. a) naked-NP, b) NP-D, c) NP-AD, d) NP-CMD, e) NP-APS, f) NP-DMSA.

For all coatings, there was an exponential increase of IONPs at the bottom end of the container. Once gravity forces started, after 3h or 10 hours, we saw different behaviors, showing an increase of the concentration of IONPs after 3h and 10 hours, more prominent for naked NPs. Since these NPs do not have any coating, presented less stability and an exponential increase of the concentration at the bottom of the container. 

One of the differences between our method and other methods using centrifugation is that instead of having cells and NPs in the same solution under centrifugal forces, we first centrifugate the cells, and once they reach the bottom of the container, we centrifuge the NPs. In order to simulate this, we took into consideration different forces that play a main role in this movement. We referred to flux density as the key parameter to understand the interaction between NPs and the cellular membrane. The flux density could increase due to a larger sedimentation coefficient (S) or to a larger angular velocity (w, the centrifugal velocity). If the flux density increase, then the adsorption of IONPs is faster, limited to the mass of adsorbed IONPs. We observed that IONPs coated with small molecules as APS and DMSA, affect the sedimentation and diffusion coefficients, increasing the concentration of NPs will be in contact with the cellular membrane. In the case of D and CMD, since they had good steric properties, our models concluded that less amount of NPs will be in contact with the cellular membrane.

            In conclusion, with our previous studies from Ocampo, et al. and Fernández-Calvo, et al. we concluded that from an experimental and theoretical point of view there was no additional aggregation under the centrifugal forces and the experimental conditions (concentration, time) that we used in the present study.

3. The SEM images in Figure S2d are unclear. It would be helpful to show them at a larger size and resolution, so that the features related to IONPs can be easier to see. In particular, it is not clear why the objects, which appear at different sizes and roughness in different SEM panels, support the conclusion that the IONPs are internalized. SEM images in the previously mentioned DOI 10.1021/ac503835a, for example, show a "rough" appearance produced by IONPs external to the cells. Optical microscopy and TEM, for example, in DOI 10.1116/1.5009989, also show that IONPs can be both internalized by mammalian cells and attach externally as aggregates. It would be important to evaluate whether the evidence provided in this manuscript supports unambiguous and exclusive internalization of the investigated IONPs, or if some combination of internalization and external interactions may be at play (aggregates external to cells are clearly seen in some of the optical images in Figure S2).

            We thank the reviewer comments and bibliography regarding the SEM images.  Our SEM images were performed to corroborate the formation of agglomeration on cellular surfaces and discard the CMI parameters for those IONPs. We never conclude these SEM images were to corroborate the internalization of the IONPs because it was for the opposite purpose. We improve the explanation in the manuscript that can lead to a wrong idea (see manuscript).

Here is a better explanation of the SEM experiment. In the first part of the study, we used an initial concentration of 50 µg/ml of IONPs and we studied how cells were labelled by IONPs and if those IONPs were on the surface or intracellular. Figure S2 shows all trials we did to improve the CMI method. First using a filter + DMEM with 10%FBS (Figure S2a), then using UV (no filter) + DMEM with 10% FBS (Figure S2b), and finally, UV + DMEM with 0%FBS (Figure S2c). From the Prussian blue images, we were able to see some agglomerates on NP-AD (Figure S2a), NP-APS (Figure S2c) and NP-DMSA (Figure S2c). We did not study NP-AD from Figure S2.b because of the clear appearance of IONPs agglomeration on the surface. We decided to do SEM images to corroborate the formation of IONPs agglomerations on the cellular surface. Because of that, we decided to use different approaches for those IONPs, in the case of NP-AD were unfiltered with 0%FBS, APS unfiltered with 10%FBS and DMSA filtered and with 10%FBS. We added some arrows to the SEM images on the supplementary file to make it clear where we found the IONPs on top of the cells.

Figure S2.D: Prussian blue and SEM images of cells and IONPs.

We also acquired Energy Dispersive X-ray Spectroscopy (EDX) spectra of the areas close to the arrows to corroborate the iron content. See the following figure:

Figure not included in the manuscript or SI: EDX spectra from SEM images (Figure S2.C) showing iron peaks on the areas close to the arrows. No signal of iron on the controls samples was found.

To support the internationalization of IONPs by CMI, Ocampo et al. did a full study about it. Figure 3A showed transmission electron microscopy (TEM) images after 4, 24 and 72 hours after CMI method. After 4 hours IONPs were internalized in open vesicles with membranes that sometimes presented not a close vesicle perimeter. After 24 and 72 hours, the vesicles evolved into an earlier and later endosome. Then, EDX measurements confirmed the presence of IONPs in the vessels suggesting an endocytosis-independent pathway for the internalization of IONPs using CMI method (Figure S7, Ocampo, et al.). To corroborate this, they studied the presence of receptor-mediated endocytosis in the presence or absence of chlorpromazine (CPZ), and inhibitor of clathrin-independent endocytosis.

Figure 3, S.M. Ocampo, et al., Scientific Report, 2015.

Figure S7, S.M. Ocampo et al., Scientific Report, 2015.

Taking into consideration Ocampo, et al.'s studies and our different approaches to induce CMI with all the IONPs in our study, we concluded that the IONPs can internalize into the cells. We also change our SEM discussion in the manuscript to make it clear to the reader. We changed the word “internalized” or “uptake” for “labelling” in the section of IONPs with different core size and when using the highest concentration (75 µg/ml). This will clarify to the reader when some of the IONPs can be internalized or when IONPs can labelled the cells, a combination of internal and external interactions.  

Author Response File: Author Response.pdf

Reviewer 2 Report

Authors have studied the influence of iron oxide nanoparticles (IONPs) coating and size on cell uptake for in vitro MRI. Prepared IONPs were characterized well and evaluated for their influence on coating and on cell uptake for MRI. The proposed work is interesting for the publication in Nanomaterials. However, the manuscript needs some revisions which are suggested below:

Title of article: IONPs is not a standard abbreviated form of iron oxide nanoparticles. Authors are advised to this abbreviation in the title of the paper for better understanding to the readers.

Abstract: Authors are advised to include more quantitative information in order to enhance the readability of the article.

The term cell uptake is not good. It is better to use the term cellular uptake instead of cell uptake throughout the manuscript.

1. 47-52: Authors should provide the reference numbers for clinical trials instead of providing clinical trial registry numbers. The same references should be included in reference list.

Introduction: The rationale and objective of studies should be clearly indicated in introduction section.

Materials and methods: Please include a separate section for materials and include complete source of each material (manufacturer, city, country etc.) in proper way.

Figure 1: kindly break this figure into two different figures.

Figure 2: kindly break this figure into two different figures.

Reference list: please remove the DOI numbers.

Author Response

(responses with figures are included in attached Word file)

Referee2:

Authors have studied the influence of iron oxide nanoparticles (IONPs) coating and size on cell uptake for in vitro MRI. Prepared IONPs were characterized well and evaluated for their influence on coating and on cell uptake for MRI. The proposed work is interesting for the publication in Nanomaterials. However, the manuscript needs some revisions which are suggested below:

1. Title of article: IONPs is not a standard abbreviated form of iron oxide nanoparticles. Authors are advised to this abbreviation in the title of the paper for better understanding to the readers.

The reviewer is right, the acronym IONPs is far less common than, for example, MRI. We have changed the title to correct it. See the main manuscript.

2. Abstract: Authors are advised to include more quantitative information in order to enhance the readability of the article.

We thank the reviewer for this suggestion. See the changes in the manuscript.

3. The term cell uptake is not good. It is better to use the term cellular uptake instead of cell uptake throughout the manuscript.

The term cell uptake has been changed to cellular uptake throughout the manuscript.

4. 47-52: Authors should provide the reference numbers for clinical trials instead of providing clinical trial registry numbers. The same references should be included in reference list.

We added the clinical trials to the reference list. See references 9-22.

5. Introduction: The rationale and objective of studies should be clearly indicated in introduction section.

We changed this part in the manuscript. See the last paragraph of the Introduction.

6. Materials and methods: Please include a separate section for materials and include complete source of each material (manufacturer, city, country etc.) in proper way.

Section 2.1 has been completed on this regard.

7. Figure 1: kindly break this figure into two different figures.

Figure1 has been split in two different figures: now Figure 1 and Figure 2.

8. Figure 2: kindly break this figure into two different figures.

Figure2 has been split in two different figures: now Figure 3 and Figure 4.

9. Reference list: please remove the DOI numbers.

All DOI numbers have been removed.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have responded to my comments in detail. I find their arguments in the response to be persuasive, however, I note that multiple paragraphs of response text, including deep citations of literature and details of figures, have not been reflected in a significant way in the revised text. While the authors clearly have deep knowledge of the relevant cited work, the readers are unlikely to have the same background expertise.

Accordingly, I recommend to provide an expanded (a couple of sentences) explanation whenever the discussion relies on very specific findings of the cited references (i.e., to summarize the relevant findings that the authors detailed in the response).

Author Response

Manuscript ID: nanomaterials-1407792

Title: Title: Influence of coating and size of magnetic nanoparticles on cellular uptake for in vitro MRI

Dear Dr. Milica Nikolic,

Please find below the response to the review. The changes in the manuscript are marked using the “Track Changes” function. We thank again for the reviewer comments that lead in an improvement on the manuscript discussion. Thank you.

Comments and Suggestions for Authors

Referee1:

The authors have responded to my comments in detail. I find their arguments in the response to be persuasive, however, I note that multiple paragraphs of response text, including deep citations of literature and details of figures, have not been reflected in a significant way in the revised text. While the authors clearly have deep knowledge of the relevant cited work, the readers are unlikely to have the same background expertise.

Accordingly, I recommend to provide an expanded (a couple of sentences) explanation whenever the discussion relies on very specific findings of the cited references (i.e., to summarize the relevant findings that the authors detailed in the response).

We agree with the reviewer that readers will have different background expertise and an extended summary of the previous discussion and revision should be included in the manuscript. Please, find few sentences on pages 11 and 12 of the newly submitted manuscript version. 

• Extended paragraph discussion regarding question 2 on previous revision (effect of centrifugation forces and possible aggregation and destabilization of IONPs):

“Higher speeds and times of incubation can produce additional aggregations and IONPs destabilization when they are under centrifugation forces [59]. Since the elimination of FBS in the media can increase the hydrodynamic size of IONPs (Figure 1B), the centrifugation time was reduced to 1 min, instead of 5 min, to avoid additional IONPs aggregations during CMI but still obtain high labelling efficiency [26]. CMI method was performed using 1500 rpm, where the sedimentation and diffusion forces play an important role without producing any precipitation. Theoretical simulations and our experiments support there is no additional aggregation or IONPs destabilization under the CMI parameters used in this study [27].”

• Extended paragraph discussion for question 3 on previous revision (SEM, TEM and aggregations):

“Depending on CMI parameters, Prussian blue images showed a possible IONPs aggregation on the cellular membrane surface. SEM and transmission electron microscopy (TEM) have been used to deeply understand if the interaction of IONPs with cells is external (aggregates on the surface), internal (IONPs uptake by cells), or both, internal and external interactions [59,60]. Previous studies used TEM to corroborate the internalization of IONPs by CMI method [26]. In addition, we observed the formation of agglomerates on top of the cellular membrane by SEM images, that were correlated with higher D_hyd, for NP-AD (561 nm), NP-APS (1342 nm) and NP-DMSA (1636 nm) coatings (see SI and Figure S2D). Based on this and previous observations, we fine-tuned the FBS concentration and the type of sterilization protocol until we reached the values that yielded the best performance between CMI-mediated IONPs internalization and avoiding the formation of agglomerates”.

Author Response File: Author Response.pdf