Separation and Concentration of Astaxanthin and Lutein from Microalgae Liquid Extracts Using Magnetic Nanoparticles
, Thalia Tsiaka 2, Panagiotis Zoumpoulakis 2,‡, Emanouella Maggiorou 1, Konstantinos Tyrovolas 1, Antonio Molino 3, Evangelos Hristoforou 1,* and Angelo Ferraro 1,*
Round 1
Reviewer 1 Report
The manuscript is devoted to the use of magnetic particles for extraction of certain carotenoids from algae isolates. Authors suggest that aminopropyl-functionalized particles may prove as a viable alternative to the chromatographic isolation methods.
While this approach deserves attention and further studies, it is debatable whether magnetic particles can be considered as a cheaper alternative to chromatography. Currently the cost of magnetic beads only justifies their small-scale use.
General comments:
How was the calibration done, by analysis of standard solutions with known concentration in the absence of matrix? In this case ISTD is mandatory to compensate for matrix effects when measurements are done on natural extracts. If no ISTD was used, it needs to be explicitly stated and no claims should be made about the method being quantitative.
The aminopropyl-functionalized particles do not and cannot have intrinsic specificity to carotenoids and would inevitable coextract other matrix components such as sugars, for example, or other polar analytes and weak acids. This needs to be made clear in the manuscript.
How the conclusion about the selectivity of extraction was made if only 2 analytes were included in the method? Since Orbitrap was used in MS/MS mode, it does not seem that any other carotenoids, let alone other compound classes, would have been detected.
More specific comments:
L85: supercritical fluid extraction
L103: please rephrase; using GRAS solvents and CO2 supercritical fluid extraction
L133: should be [26]
L138: check if the reference [24] is accurate
L148-166: this paragraph combines both the methods used and the results obtained; the Experimental should only describe methods, but not the results such as “Regression coefficient (R2) were 0.996 for astaxanthin and 0.997 for lutein, respectively, verifying the linearity of the method” or “confirming that the developed LC-MS/MS method was precise since %RSD of QC samples was lower than 15%”.
L157-158: why 20-fold difference in the lowest level of calibration curves for the two carotenoids?
L172: what buffer?
L175: provide equipment details here (instead of L191-192).
L180: would immiscibility of hexane with water pose any issue in this experiment? Please comment or remove.
L182-183: please rephrase or explain “The concentrated extracts were used as 10% of the total volume”
L183-184: what time and what temperature?
L185: should read “was placed in a special magnetic rack”
L204: should read “recovery from solution”
L209-214: do not repeat what’s already given in Experimental
L215: 1 mL acetone or 0.5 mL (see L174)?
Figs 1 and 2 should use the same units of measure for the amount of magnetic particles (µg vs µL).
Given that 10 µL volume is suggested as optimal, why the recovery experiment was conducted only with the low levels of carotenoid (50 ng)? Would 250 µg magnetic particles provide the same recovery for 15 µg carotenoid?
Table 3: what is the exact nature of the control sample? What are samples S1 and S2, two replicates of the same experiment? There are way too many significant figures and standard deviation appears to be unrealistically low. Is this accurate that SD was calculated from only 2 parallel measurements? SD does not add any benefit and should be removed.
L280: what is the extract volume taken for magnetic separation? Given that the linear range of the LCMS method is up to 15 µg/mL and carotenoid content in the samples is over 300 µg/mL there must be a dilution at some point.
L297: how the ratio can be negative? please explain or rephrase.
L298: it is not sufficient to say a “high yield”. The exact method recovery should be provided in Results and Discussion. It is noted that L397 presents a value of >90%. Is it true that recovery is based on spectrophotometric experiment on a standard solution?
L350: should read “can be obtained”
L361-364: when chromatography is used for isolation of carotenoids, why would the low sensitivity be of concern? On preparative scale the sensitivity is not an issue. Rephrase or remove.
L372-374: please rephrase, “…since carotenes do not form ester linkages and so can be directly extracted by lipophilic solvents, while xanthophylls form ester linkages and can readily be extracted into lipophilic solvents.”
Author Response
All authors would like to thank the three reviewers for the useful and constructive comments and for their time spent on our manuscript. Detailed response to the Reviewer’s comment is given below (the page and line numbers in the answer refers to the pdf file with track changes). We have reproduced the comments of the Reviewer in black font and our point-by-point response in red font. We have revised the manuscript (track changes has been used) according to the Reviewers’ suggestions. SMS method is in its infancy, it needs improvement and a lot of work to find the best ligand for each target molecule. In this work we just want to show that it is possible to remove from a solution a specific class of molecules in an easy way. We are aware that there are weakness points on which we are working, hoping that we will soon find a solution.
Comments and Suggestions for Authors
The manuscript is devoted to the use of magnetic particles for extraction of certain carotenoids from algae isolates. Authors suggest that aminopropyl-functionalized particles may prove as a viable alternative to the chromatographic isolation methods. While this approach deserves attention and further studies, it is debatable whether magnetic particles can be considered as a cheaper alternative to chromatography. Currently the cost of magnetic beads only justifies their small-scale use.
General comments:
How was the calibration done, by analysis of standard solutions with known concentration in the absence of matrix? In this case ISTD is mandatory to compensate for matrix effects when measurements are done on natural extracts. If no ISTD was used, it needs to be explicitly stated and no claims should be made about the method being quantitative.
Answer: Although, there are many recent published works, where the LCMS quantification of carotenoids and other natural compound (i.e. polyphenols) [1–3] was performed without an ISTD, we have stated in the manuscript that no ISTD was used in the current study. In addition, we have replaced the terms related to ‘’quantification’’ with terms that refer to ‘’content estimation’’.
The aminopropyl-functionalized particles do not and cannot have intrinsic specificity to carotenoids and would inevitable coextract other matrix components such as sugars, for example, or other polar analytes and weak acids. This needs to be made clear in the manuscript.
Answer: We partially agree with the reviewer, amino groups in theory do present intrinsic specificity to carotenoids and would inevitable interact with other molecules. However, we see that coloured solutions turn almost transparent after mixing them with our functionalized nanoparticles. Furthermore, even from the abstract we say that SMS is a method that should be applied on partially purified extracts (second step of purification), which may reduce cross reaction. We also state in the manuscript that this technology must be improved. In order to follow review’s advice, we added at the beginning of the discussion the following sentence: “The aminosilane functionalized particles should not have intrinsic specificity for carotenoids, however, based on our experiments, by mixing them with commercial astaxanthin as well as with carotenoids from H. pluvialis microalgae extracts we observed a clear separation of carotenoids from the solution since such solution discoloured (Figures 3 and 4)”.
How the conclusion about the selectivity of extraction was made if only 2 analytes were included in the method? Since Orbitrap was used in MS/MS mode, it does not seem that any other carotenoids, let alone other compound classes, would have been detected.
Answer: The LC-MS/MS analysis was a targeted and not an untargeted analysis focusing on the determination only of the trans-forms of the two carotenoids. Therefore, it can provide an estimation of fold changes in the content of these two carotenoids that can be related to the selectivity of SMS.
More specific comments:
L85: supercritical fluid extraction
Answer: Corrected
L103: please rephrase; using GRAS solvents and CO2 supercritical fluid extraction
Answer:Corrected
L133: should be [26]
Answer:Corrected
L138: check if the reference [24] is accurate
Answer:It was not; corrected. We thank the reviewer.
L148-166: this paragraph combines both the methods used and the results obtained; the Experimental should only describe methods, but not the results such as “Regression coefficient (R2) were 0.996 for astaxanthin and 0.997 for lutein, respectively, verifying the linearity of the method” or “confirming that the developed LC-MS/MS method was precise since %RSD of QC samples was lower than 15%”.
Answer: We transferred this paragraph in Section 1.3, which describes the results of the LCMS analysis.
L157-158: why 20-fold difference in the lowest level of calibration curves for the two carotenoids?
Answer: As we have mentioned in the manuscript, the LCMS method used for the determination of carotenoids, is an already developed and validated method of our group, showing excellent linearity in a wide concentration range (from 0.5-15 ug/ml). Based on the peak area of astaxanthin in the algae samples, we had to expand the calibration curve of astaxanthin by adding and analysing extra standards solutions of lower concentrations in order to construct an adequate calibration curve.
L172: what buffer?
Answer: The buffer was ethanal, we added in the sentence the words “dissolved in” in order to emphasize the fact that we used ethanol as buffer
L175: provide equipment details here (instead of L191-192).
Answer: The equipment detail has been moved in the suggested place
L180: would immiscibility of hexane with water pose any issue in this experiment? Please comment or remove.
Answer: The following sentence was added to the text to comment about the miscibility of hexane: “After testing raw extract in hexane it was noticed that such solvent was immiscible with the water-based PBS buffer, therefore only raw extracts in ethanol were used for all the test.”
L182-183: please rephrase or explain “The concentrated extracts were used as 10% of the total volume”
Answer: Thanks for this comment, we removed the sentence because it was a redundant information, indeed a few lines above 184-185, we say that 0.1 ml of extract was added to 0.9 ml of PBS buffer.
L183-184: what time and what temperature?
Answer: Time and temperature details are added
L185: should read “was placed in a special magnetic rack”
Answer: The sentence was rephrased
L204: should read “recovery from solution”
Answer: The sentence was corrected
L209-214: do not repeat what’s already given in Experimental
Answer: Technical sentences have been removed
L215: 1 mL acetone or 0.5 mL (see L174)?
Answer: Thanks for the comment. The correct volume is 1 ml, we corrected the number at line 174.
Figs 1 and 2 should use the same units of measure for the amount of magnetic particles (µg vs µL).
Answer: We corrected the text describing figures 1 and 2
Given that 10 µL volume is suggested as optimal, why the recovery experiment was conducted only with the low levels of carotenoid (50 ng)? Would 250 µg magnetic particles provide the same recovery for 15 µg carotenoid?
Answer: The low level of carotenoid of 50ng was used essentially to save astaxanthin standard and give us the possibility to repeat several times the experiments. The highly pure astaxanthin standard we used was very expensive so we just kept its use to the minimum.
Table 3: what is the exact nature of the control sample? What are samples S1 and S2, two replicates of the same experiment? There are way too many significant figures and standard deviation appears to be unrealistically low. Is this accurate that SD was calculated from only 2 parallel measurements? SD does not add any benefit and should be removed.
Answer: Control sample is the raw microalgae extract the first bottle on the left of figure 7; S1 and S2 represent the control sample that undergo to magnetic separation and the reaction is repeated twice. Though we have done it several times which allowed us to calculate the SD. We prefer to leave the SD bars since other reviewers, from a different journal, insisted on putting it.
L280: what is the extract volume taken for magnetic separation? Given that the linear range of the LCMS method is up to 15 µg/mL and carotenoid content in the samples is over 300 µg/mL there must be a dilution at some point.
Answer: The calibration curve provides the concentration of carotenoids in the sample solutions used for the LCMS analysis, not the final concentration of carotenoids in the initial samples. The results presented in Table 3 refer to the content of carotenoids in the initial samples. Since the only processing step prior to the LCMS analysis was a dilution of a sample aliquot in the injection solvent, we provided in Section 2.4 of the manuscript the dilution factor of all the samples for LCMS.
L297: how the ratio can be negative? please explain or rephrase.
Answer: Thanks for this comment, indeed the ratio is not negative, we used the symbol “~” that might be confused with the symbol minus “-“. To avoid any further misinterpretation the symbol “~” was removed and the words “…..between to roughly….” was added .
L298: it is not sufficient to say a “high yield”. The exact method recovery should be provided in Results and Discussion. It is noted that L397 presents a value of >90%. Is it true that recovery is based on spectrophotometric experiment on a standard solution?
Answer: We modified the sentence at line 298 by adding the value of >90%. In the same sentence is reported that such yield is obtained by microalgae extracts. The recovery is not based on spectrophotometric experiment on a standard solution, but by comparing the concentration of carotenoids before and after magnetic separation on microalgae extracts. This is reported on the conclusion in the sentence at line 397.
L350: should read “can be obtained”
Answer: Thanks for the comment, we corrected the word obtained
L361-364: when chromatography is used for isolation of carotenoids, why would the low sensitivity be of concern? On preparative scale the sensitivity is not an issue. Rephrase or remove.
Answer: The sentence part expressing concern was removed from the text.
L372-374: please rephrase, “…since carotenes do not form ester linkages and so can be directly extracted by lipophilic solvents, while xanthophylls form ester linkages and can readily be extracted into lipophilic solvents.”
Answer: Thank for all the comments and in particular this one which improved a lot the manuscript quality. We rephrased the sentence and added all suggested references.
References
- Todorović, B.; Grujić, V.J.; Krajnc, A.U.; Kranvogl, R.; Ambrožič-Dolinšek, J. Identification and Content of Astaxanthin and Its Esters from Microalgae Haematococcus Pluvialis by HPLC-DAD and LC-QTOF-MS after Extraction with Various Solvents. Plants 2021, 10, 2413, doi:10.3390/plants10112413.
- Ullah, Z.; Ali, S.; Hussain, A.; Öztürk, M.; Ertaş, A.; Alamzeb, M.; Rashid, M.U.; Ullah, H.; Zaman, R.; Imtiaz, M. In Vitro Antioxidant, Anticholinesterase, Tyrosinase Activity Studies, and LC-MS/MS Simultaneous Determination of 37 Bioactive Compounds in Indigofera Heterantha. South African Journal of Botany 2022, 148, 537–545, doi:10.1016/j.sajb.2022.05.012.
- Zhang, L.; Wang, S.; Yang, R.; Mao, J.; Jiang, J.; Wang, X.; Zhang, W.; Zhang, Q.; Li, P. Simultaneous Determination of Tocopherols, Carotenoids and Phytosterols in Edible Vegetable Oil by Ultrasound-Assisted Saponification, LLE and LC-MS/MS. Food Chemistry 2019, 289, 313–319, doi:10.1016/j.foodchem.2019.03.067.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The authors presented an interesting method for the separation and concentration of two carotenoids using magnetic nanoparticles. They measured the efficiency of the proposed technology in the function of variable conditions of temperature, the concentration of magnetic nanoparticles, and elution time.
There are some unclear points. I would them to reply to the following questions.
1) In the Introduction paragraph, on lines 78-79, the authors reported the reference n.17 on the importance of astaxanthin, but for lutein?
2) In the Results paragraph, the authors reported the conditions which affected magnetic binding for astaxanthin, but it's not studied for lutein?
3) Why any statistical test not be applied to these results? In table 3, there is an ANOVA test for the interpretation of obtained results, but it is missed in Figures 1 and 2.
Then, there are some minor errors: on line 138, check that reference 24 is appropriate, on line 165 check the unit of measure, what is ug?
Author Response
All authors would like to thank the three reviewers for the useful and constructive comments and for their time spent on our manuscript. Detailed response to the Reviewer’s comment is given below (the page and line numbers in the answer refers to the pdf file with track changes). We have reproduced the comments of the Reviewer in black font and our point-by-point response in red font. We have revised the manuscript (track changes has been used) according to the Reviewers’ suggestions. SMS method is in its infancy, it needs improvement and a lot of work to find the best ligand for each target molecule. In this work we just want to show that it is possible to remove from a solution a specific class of molecules in an easy way. We are aware that there are weakness points on which we are working, hoping that we will soon find a solution.
Comments and Suggestions for Authors
The authors presented an interesting method for the separation and concentration of two carotenoids using magnetic nanoparticles. They measured the efficiency of the proposed technology in the function of variable conditions of temperature, the concentration of magnetic nanoparticles, and elution time.
There are some unclear points. I would them to reply to the following questions.
In the Introduction paragraph, on lines 78-79, the authors reported the reference n.17 on the importance of astaxanthin, but for lutein?
Answer: We would like to thank the reviewer for this comment. We replaced Ref 17 with a more comprehension one.
In the Results paragraph, the authors reported the conditions which affected magnetic binding for astaxanthin, but it's not studied for lutein?
Answer: Yes indeed, we only focused on astaxanthin. The main reason was because in the magnetic extracts of H. pluvialis we were unable to separate them, therefore we used only astaxanthin to optimize the conditions.
Why any statistical test not be applied to these results? In table 3, there is an ANOVA test for the interpretation of obtained results, but it is missed in Figures 1 and 2.
Answer: In the experiments showed in Figure 1 and Figure 2 we evaluate which combination of conditions (amount of nanoparticles, temperature and time) gave the best yield results. Each histogram present replicates of the same condition with standard deviation and we think that the standard deviation give enough information about which combination works better.
Then, there are some minor errors: on line 138, check that reference 24 is appropriate, on line 165 check the unit of measure, what is ug?
Answer: Thanks for these comments which improved the manuscript quality. Ref. 24 has been corrected. Τhe ug is the 'english version' of μg. We are keeping the 'μg' form in the text, so we have corrected the 'ug' to 'μg'.
Author Response File: Author Response.docx
Reviewer 3 Report
The authors report on the use of selective magnetic separation (SMS) of astaxanthin and lutein from extracts of microalga Haematococcus pluvi-alis. Their work is very interesting and the results rather promising, considering the complexity involved in most existing methods for isolating natural products from microalgae liquid extracts. However, there are a few questions deserving appropriate comments in the manuscript:
1) The interaction of astaxanthin with the amino-functionalized magnetic nanoparticles was carried out in PBS. At this pH, the amino groups are protonated, but there is no information on the pKa of astaxanthin. Can it be deprotonated, considering the keto-enolic equilibrium? This could improve their interaction, introducing electrostatic forces to the proposed hydrogen bonding effects.
2) Is there any possibility of interaction between the amino groups from apts, and the ketone groups from astaxanthin, e.g. forming iminobonds?
3) It would be very important to quantify the interaction of the astaxanthin and lutein with the silanized nanoparticles using, for instance, Langmuir isotherms.
4) The author should comment further on the recycling of the magnetic nanoparticles, including their relative behavior after the many successive cycles.
Author Response
All authors would like to thank the three reviewers for the useful and constructive comments and for their time spent on our manuscript. Detailed response to the Reviewer’s comment is given below (the page and line numbers in the answer refers to the pdf file with track changes). We have reproduced the comments of the Reviewer in black font and our point-by-point response in red font. We have revised the manuscript (track changes has been used) according to the Reviewers’ suggestions. SMS method is in its infancy, it needs improvement and a lot of work to find the best ligand for each target molecule. In this work we just want to show that it is possible to remove from a solution a specific class of molecules in an easy way. We are aware that there are weakness points on which we are working, hoping that we will soon find a solution.
Comments and Suggestions for Authors
The authors report on the use of selective magnetic separation (SMS) of astaxanthin and lutein from extracts of microalga Haematococcus pluvialis. Their work is very interesting and the results rather promising, considering the complexity involved in most existing methods for isolating natural products from microalgae liquid extracts. However, there are a few questions deserving appropriate comments in the manuscript:
The interaction of astaxanthin with the amino-functionalized magnetic nanoparticles was carried out in PBS. At this pH, the amino groups are protonated, but there is no information on the pKa of astaxanthin. Can it be deprotonated, considering the keto-enolic equilibrium? This could improve their interaction, introducing electrostatic forces to the proposed hydrogen bonding effects.
Answer: We would like to thank the reviewer for this valuable comment. In the present manuscript we analysed only three parameters affecting the binding: nanoparticles concentration, temperature and time of mixing/elution. We didn’t consider the pH and consequently the pKa of both carotenoids involved. We agree with the review that deprotonation may further improve the interaction between amino-functionalized magnetic nanoparticles and carotenoids. Definitely it will be tested in our next work. In order to give the reader a more comprehensive view of the interaction we added a sentence in the discussion that follows the review suggestion: “Further improvement of SMS is necessary especially about selectivity as well as other parameters affecting the like pH and pKa of both carotenoids; indeed it is possible that by changing the reaction pH (in our case roughly 7 because of PBS buffer) the protonation of aminosilane group changes and in turn a more favourable interaction is possible”.
Is there any possibility of interaction between the amino groups from apts, and the ketone groups from astaxanthin, e.g. forming iminobonds?
Answer: As we reported in the discuss several interaction may be possible, we cannot exclude the reaction suggested.
It would be very important to quantify the interaction of the astaxanthin and lutein with the silanized nanoparticles using, for instance, Langmuir isotherms.
Answer: Thanks again for the suggestion. As mentioned in the previous comment, we are already working on a second project to improve the selectivity and binding of amino groups. We will try the suggested experiment with silanized nanoparticles.
The author should comment further on the recycling of the magnetic nanoparticles, including their relative behavior after the many successive cycles.
Answer: We didn’t perform recycle experiments yet. Therefore, we aren’t sure if the aminosilane groups remain functional after elution with acetone or other solvent. In any case we added the following sentences in the discussion that address the recycling topic: “In addition, nanoparticles can be reuse. Even though we didn’t perform recycling experiments; we believe that magnetic nanoparticles can be reuse several times. In fact, assuming that aminosilane groups are damage during the elution process, the nanoparticle core remains intact and can undergo to a second round of functionalization providing again the same binding performance. On the other hand, if aminosilane groups are not damage during the elution process they can be reuse directly in a second round of isolation without any additional step to regenerate them”.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Authors: thanks for your effort to improve the manuscript. It still is strongly suggested that a loadability experiment is conducted to determine the maximum amount of carotenoids per volume of beads that can be adsorbed before the beads get saturated.
The fact that carotenoid solution gets discolored upon treatment with the magnetic beads does not mean that beads are in any way specific for these two carotenoids. Yes, they are extracted from solution, no, it is not selective (and many other compounds would be coextracted, too).
Lastly: ethanol is not a buffer.
The manuscript can be accepted in its present form.
Reviewer 2 Report
No other comments.
