In vivo MRI is sensitive to remyelination in a nonhuman primate model of multiple sclerosis

Article activity feed

1. 

   Version published to 10.7554/elife.73786 on eLife

   Apr 21, 2023
2. 

   eLife

   Mar 17, 2022

   Author Response

   Reviewer #1 (Public Review):

   Lee et al report the incidence of remyelination in the non-human primate model of multiple sclerosis. EAE was induced in marmosets and serial 7 tesla MRI identified cerebral white matter lesions. Thirty-six focal lesions classified as demyelinated or remyelinated based upon protondensity images were assessed by histopathology. Fifty-one % of these lesions were identified as remyelinated. These studies have implications for preclinical testing of pro-myelinating agents in individuals with MS.

   Strengths: The MRI data presented is of high quality and demonstrates the value of multisequence 7-Telsa MRI. The sequential imaging clearly identify alterations in signal intensity on proton densityweighted images that are consistent with demyelination and remyelination.

   We thank the reviewer for …

   Author Response

   Reviewer #1 (Public Review):

   Lee et al report the incidence of remyelination in the non-human primate model of multiple sclerosis. EAE was induced in marmosets and serial 7 tesla MRI identified cerebral white matter lesions. Thirty-six focal lesions classified as demyelinated or remyelinated based upon protondensity images were assessed by histopathology. Fifty-one % of these lesions were identified as remyelinated. These studies have implications for preclinical testing of pro-myelinating agents in individuals with MS.

   Strengths: The MRI data presented is of high quality and demonstrates the value of multisequence 7-Telsa MRI. The sequential imaging clearly identify alterations in signal intensity on proton densityweighted images that are consistent with demyelination and remyelination.

   We thank the reviewer for commenting on the quality of our high-resolution MRI, which is one of the biggest strengths of the current manuscript.

   Weaknesses: While the MRI aspects of this study are of high quality, the pathological correlates of these MRI abnormalities need better confirmation. The lesions identified by MRI are very small (often less than a mm3). Some indication of how MRI and histological lesion sites were co-registered would be helpful. How was the brain sliced? Where the lesions visible macroscopically on the fixed slices? How many sections were cut to identify the lesion?

   We thank the reviewer for inquiring about the size of the lesions as well as the radiologypathology correlation. Our group has developed a reliable pipeline to correlate MRI findings to pathology slides by performing an ultrahigh resolution, ex-vivo, 3D MRI of the brains once they are extracted. This scan is used to create individualized brain cradles with a 3D printer, which allows us to cut the brains into 2–4 mm slabs that are in an extremely close plane and axis to the in-vivo MRI. Further descriptions can be found in our previous manuscripts, Luciano et al., 2016 JoVE, and Absinta et al., 2014 J Neuropathol Exp Neurol. Some of the lesions that are >1mm3 can be visualized grossly after formalin fixation, and the remainder can be reliably located using the aforementioned 3D printing method and the high-resolution MRI. Each of the 2–4 mm slabs is then cut into 4– 8 µm sections, yielding approximately 500 glass slides. This part and the 2 manuscripts mentioned are mentioned in our Methods section under “Histopathology of EAE Lesions.”

   Lesions were characterized as acute or chronic. Acute lesions in MS brains contain an abundance of macrophages/monocytes, lack of myelin, and on rare occasions myelin protein debris. The acute lesion shown in Fig 2 is difficult to classify. The lesion should be space occupying and convincingly demonstrated by a low magnification image that includes both lesional and nonlesional areas. The lesion area should have an accumulation of Iba1-positive cells and a dramatic reduction in PLP staining compared to surrounding normal appearing white matter. The staining for oligodendrocytes (ASPA and Olig2) in Fig 2 may identify a small centrally located decrease in oligo number, but this area does not correspond to differences in Iba1 or PLP staining. Scale bars are needed on the histological figs and some comment on lesions size would be helpful. Is the size of the lesions similar by MRI and pathology?

   We fully agree with the reviewer’s description of acute MS lesions, especially regarding the abundance of Iba1-positive cells and the dramatic reduction in PLP staining compared to surrounding white matter. To better demonstrate this, we have now added low-magnification images. The size of the lesions, from our experience working with multiple EAE marmosets, has been consistent between MRI and histopathology. We have also included scale bars in all of our figures for better appreciation of the lesions’ true sizes.

   There are similar concern regarding remyelinated lesions. What is the size of the lesions in stained sections? What percentage of the lesional area is occupied by myelin? Are the myelin internodes shorter and thinner than myelin in normal appearing white matter.

   This is an excellent point, regarding how much of the remyelinated lesion is occupied by the newly formed myelin. In our previous work, published in Lee et al., JCI 2019, we demonstrated that remyelinated lesions and extralesional white matter have significantly similar that stain for both PLP and LFB. Semithin toluidine blue-stained sections and EM images are included in Supplementary Figure 2 of that paper and show the expected findings. Because we prioritize preparation for traditional histology (and, more recently, -omic studies) relative to electron microscopy, and because the marmoset brain is quite large relative to the mouse brain or spinal cord, we have not been able to systematically track internode length and myelin thickness. Such analysis would also be complicated by the fact that the white matter tracts affected by lesions in our model usually run obliquely to the planes of section.

   Reviewer #2 (Public Review):

   The identification of an in vivo imaging strategy to follow demyelination and remyelination in multiple sclerosis (MS) and MS -like experimental lesions is a critical goal for regenerative medicine. MS represents one of the best target diseases for regenerative therapies, with clear evidence for an endogenous regenerative process to target and recognition that the progressive disability in patients with chronic disease results from the axonal degeneration consequent to regenerative failure. There is considerable controversy as to the best strategy for MR imaging in assessing remyelination. This results, in part, from the gulf between rodent models, where the CNS repairs rapidly and efficiently following demyelination and the diseased human CNS where any regeneration can be much slower and complicated by chronic inflammation.

   A potential solution to this is the development of a large animal model that better recapitulates the human cellular pathology and enables the development of imaging protocols that can be used in the clinic. This study does exactly that, studying lesions in six marmosets following induction of acute inflammatory demyelination (EAE) as occurs in MS. Brains were examined by MR imaging after EAE induction, and lesions identified and followed with serial imaging before histological examination to confirm the cellular phenotype. The results show that a high percentage of the lesions undergo spontaneous remyelination and that this can be detected by the change in the demyelination associated signal visualized by proton density weighted (PDw) MRI. Despite the inevitably small number of animals studied, the result is robust although the intriguing findings the steroids had no effect and that remyelination is stronger in males do probably need larger numbers.

   We thank the reviewer for the enthusiasm regarding the utility of the marmoset EAE model and the combined use of high-resolution, in-vivo MRI and histopathological correlation to investigate the pathobiology of remyelination. We agree that the findings regarding the steroid pulse treatment not having any significant effects on the prevalence of remyelination, as well as sex differences, is intriguing but would benefit from a larger sample of experimental animals.

   Reviewer #3 (Public Review):

   Lee, Sati and colleagues investigated whether remyelination can be detected non-invasively using MRI in common marmosets with experimental autoimmune encephalomyelitis (EAE). The authors subjected the marmosets to serial MRI during the course of the pathology. The results of PDw and MTR sequences were compared to those of histopathological analyses performed on brain tissue after the animals reached the end point. They found that PDw was more efficient in detecting remyelinated lesions than MTR. The authors also found that early treatment with methylprednisolone had no effect on remyelination. Moreover, the authors observed less remyelination in females compared to males.

   Strengths: These experiments are valuable as non-invasive detection of remyelination in preclinical models is a indispensable for testing the efficacy of pro-remyelinating agents prior to clinical studies. Moreover, the animal model used (marmoset EAE) is probably the one that mimicks MS lesions the best, which further supports the importance of the results presented. In addition, the manuscript, particularly the discussion section, is well written, and it suitably addresses and clarifies some issues relevant to the experimental design (low animal numbers, the comparability of different paradigms used to induce EAE, and the potential impact of applying corticosteroids early after lesion detection).

   We thank the reviewer for the positive comments regarding the strength of the marmoset EAE model, as well as the discussion section.

   Weaknesses: The main caveat of this manuscript is that histopathological analyses performed appear insufficient to validate the MRI findings regarding demyelination/remyelination, as well as the activity of demyelinating lesions (acute demyelinating versus chronic). This could be improved by addressing the following points:

   1. The criteria to define different lesion types should be clearly presented (numbers/nature of inflammatory cells, their positivity for different myelin antigens, numbers of oligodendrocytes / OPCs, axonal markers), referenced, and applied when performing histological classification.

   We thank the reviewer for identifying the need to have better-defined lesion categorization. The lesion categorization was mainly based on how the two experienced raters for histopathological analysis, blinded to each other and to the MRI dataset, rated the lesions based on pre-existing criteria used to categorize MS lesions (i.e., Kuhlmann et al., 2017 Acta Neuropathol) and on our experience with marmoset EAE lesions (as detailed in previous publications, in which we extensively characterized histopathology of lesions in marmoset EAE; see references elsewhere in this response document). We do not believe that it is necessary for the purposes of this study — which largely focuses on the ability of MRI to follow lesion repair/remyelination — to fully recapitulate the pathology studies in prior work. With respect to timeline, the lesions were categorized as acute when younger than 5 weeks, and as either chronic demyelinated or remyelinated if older than 5 weeks; this timeline also corresponds to our previous studies.

   1. Quantification of histological parameters is lacking. Statements such as "decrease in numbers of oligodendrocytes/oligodendrocyte depletion/axonal loss" etc should be corroborated by quantification of specific cellular/axonal markers in lesion areas, as compared to normally myelinated tissue (normally appearing white matter). For example, in Fig 2, based on the image of ASPA/Olig2 labeling, the authors mention "loss of oligodendrocytes", but such loss is apparent only in the small area in the center of the lesion, while the remaining area negative for PLP contains many ASPA+ cells. Quantification of ASPA in the lesions versus NAWM would unequivocally clarify this issue. The same is true for Bielschowsky staining and inflammatory cell markers-quantification would provide solid data. Quantification of inflammatory cells (possibly using additional markers), and co-labelings with different myelin antigens would be very helpful in distinguishing between acute demyelinating and chronic demyelinated lesions (histologically).

   We thank the reviewer for highlighting the need for quantification of histological markers. Given that our recent publication (Lee et al., JCI 2019) included all the quantification data, we thought repeating the quantification for this manuscript, which focuses mainly on the utility of serial in vivo MRI to detect spontaneous lesion remyelination, was redundant. To briefly summarize, Lee et al., JCI 2019 demonstrated that remyelinated lesions showed significant reappearance of both oligodendrocyte precursor cells and mature oligodendrocytes (based on ASPA/Olig2). Increased density of PLP-positive and LFB-positive myelin sheaths signified return-to-baseline as compared to extralesional white matter. We also presented results of axonal staining with Bielschowsky’s silver method. Our preliminary data and staining for this paper did include different variants of myelin staining, including MBP, MOG, Sudan Black, LFB, and PLP, but given the redundancy of the results, we opted to only include 1 protein and 1 lipid marker (PLP and LFB, respectively) here; these are very commonly used markers of myelin in both MS and EAE. To clarify this point, we added an additional description on our discussion section (paragraph 4).

   1. Regarding remyelinated lesions, it would be useful to see Luxol Fast Blue staining pattern at lower power and appreciate paler staining of the remyelinated area as compared to nondemyelinated white matter.

   We absolutely agree with the statement and have included lower magnification LFB-PAS staining in Figure 4 (now Figure 3).

   Additional information: Related to the above-mentioned point, it would be interesting to present additional histological data/discussion for the animals treated with methylprednisolone (MP). From Figure 5, it seems that MP treatment (applied at week 24-25) resolved demyelination of the first lesion in M#5, but the second lesion in M#5 presented developed after MP treatment was completed (around week 32). This suggests that no differences in the total percentage of remyelinated/total lesions in MPtreated versus MP non-treated animals were observed because most lesions in MP-treated animals developed after MP treatment was completed. It would be interesting to find out whether there were any histological particularities in early lesions in MP-treated animals (even though there should be very few of these). These data would fit nicely with the relevant paragraph presented in Discussion.

   We thank the reviewer for pointing this out. We did not identify differences between methylprednisolone-treated and untreated lesions. It remains possible that subtle differences would have been missed due to power limitations. We have also included this statement on the Results section, under “Remyelination is independent of corticosteroid administration.”

   Read the original source
3. 

   eLife

   Mar 17, 2022

   Evaluation Summary:

   Reich and colleagues have combined MRI imaging and histopathology to study the remyelination of brain lesions in an EAE marmoset model of multiple sclerosis. This work addresses in a non-human primate a missing link in the neuropathology of myelin repair, because in human MS it is virtually impossible to study the lesion dynamics by MRI (in live patients) and demyelination by histology (upon brain autopsy). The present manuscript would be improved by adding further histological evidence of remyelination and clarifying open questions of data acquisition.

   (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

   Read the original source
4. 

   eLife

   Mar 17, 2022

   Reviewer #1 (Public Review):

   Lee et al report the incidence of remyelination in the non-human primate model of multiple sclerosis. EAE was induced in marmosets and serial 7 tesla MRI identified cerebral white matter lesions. Thirty-six focal lesions classified as demyelinated or remyelinated based upon proton-density images were assessed by histopathology. Fifty-one % of these lesions were identified as remyelinated. These studies have implications for preclinical testing of pro-myelinating agents in individuals with MS.

   Strengths:
   The MRI data presented is of high quality and demonstrates the value of multisequence 7-Telsa MRI. The sequential imaging clearly identify alterations in signal intensity on proton density-weighted images that are consistent with demyelination and remyelination.

   Weaknesses:
   While the MRI aspects of this study …

   Reviewer #1 (Public Review):

   Lee et al report the incidence of remyelination in the non-human primate model of multiple sclerosis. EAE was induced in marmosets and serial 7 tesla MRI identified cerebral white matter lesions. Thirty-six focal lesions classified as demyelinated or remyelinated based upon proton-density images were assessed by histopathology. Fifty-one % of these lesions were identified as remyelinated. These studies have implications for preclinical testing of pro-myelinating agents in individuals with MS.

   Strengths:
   The MRI data presented is of high quality and demonstrates the value of multisequence 7-Telsa MRI. The sequential imaging clearly identify alterations in signal intensity on proton density-weighted images that are consistent with demyelination and remyelination.

   Weaknesses:
   While the MRI aspects of this study are of high quality, the pathological correlates of these MRI abnormalities need better confirmation. The lesions identified by MRI are very small (often less than a mm3). Some indication of how MRI and histological lesion sites were co-registered would be helpful. How was the brain sliced? Where the lesions visible macroscopically on the fixed slices? How many sections were cut to identify the lesion? Lesions were characterized as acute or chronic. Acute lesions in MS brains contain an abundance of macrophages/monocytes, lack of myelin, and on rare occasions myelin protein debris. The acute lesion shown in Fig 2 is difficult to classify. The lesion should be space occupying and convincingly demonstrated by a low magnification image that includes both lesional and non-lesional areas. The lesion area should have an accumulation of Iba1-positive cells and a dramatic reduction in PLP staining compared to surrounding normal appearing white matter. The staining for oligodendrocytes (ASPA and Olig2) in Fig 2 may identify a small centrally located decrease in oligo number, but this area does not correspond to differences in Iba1 or PLP staining. Scale bars are needed on the histological figs and some comment on lesions size would be helpful. Is the size of the lesions similar by MRI and pathology?

   There are similar concern regarding remyelinated lesions. What Is the size of the lesions in stained sections? What percentage of the lesional area is occupied by myelin? Are the myelin internodes shorter and thinner than myelin in normal appearing white matter.

   Read the original source
5. 

   eLife

   Mar 17, 2022

   Reviewer #2 (Public Review):

   The identification of an in vivo imaging strategy to follow demyelination and remyelination in multiple sclerosis (MS) and MS -like experimental lesions is a critical goal for regenerative medicine. MS represents one of the best target diseases for regenerative therapies, with clear evidence for an endogenous regenerative process to target and recognition that the progressive disability in patients with chronic disease results from the axonal degeneration consequent to regenerative failure. There is considerable controversy as to the best strategy for MR imaging in assessing remyelination. This results, in part, from the gulf between rodent models, where the CNS repairs rapidly and efficiently following demyelination and the diseased human CNS where any regeneration can be much slower and complicated by …

   Reviewer #2 (Public Review):

   The identification of an in vivo imaging strategy to follow demyelination and remyelination in multiple sclerosis (MS) and MS -like experimental lesions is a critical goal for regenerative medicine. MS represents one of the best target diseases for regenerative therapies, with clear evidence for an endogenous regenerative process to target and recognition that the progressive disability in patients with chronic disease results from the axonal degeneration consequent to regenerative failure. There is considerable controversy as to the best strategy for MR imaging in assessing remyelination. This results, in part, from the gulf between rodent models, where the CNS repairs rapidly and efficiently following demyelination and the diseased human CNS where any regeneration can be much slower and complicated by chronic inflammation.

   A potential solution to this is the development of a large animal model that better recapitulates the human cellular pathology and enables the development of imaging protocols that can be used in the clinic. This study does exactly that, studying lesions in six marmosets following induction of acute inflammatory demyelination (EAE) as occurs in MS. Brains were examined by MR imaging after EAE induction, and lesions identified and followed with serial imaging before histological examination to confirm the cellular phenotype. The results show that a high percentage of the lesions undergo spontaneous remyelination and that this can be detected by the change in the demyelination associated signal visualized by proton density weighted (PDw) MRI. Despite the inevitably small number of animals studied, the result is robust although the intriguing findings the steroids had no effect and that remyelination is stronger in males do probably need larger numbers.

   Read the original source
6. 

   eLife

   Mar 17, 2022

   Reviewer #3 (Public Review):

   Lee, Sati and colleagues investigated whether remyelination can be detected non-invasively using MRI in common marmosets with experimental autoimmune encephalomyelitis (EAE). The authors subjected the marmosets to serial MRI during the course of the pathology. The results of PDw and MTR sequences were compared to those of histopathological analyses performed on brain tissue after the animals reached the end point. They found that PDw was more efficient in detecting remyelinated lesions than MTR. The authors also found that early treatment with methylprednisolone had no effect on remyelination. Moreover, the authors observed less remyelination in females compared to males.

   Strengths: These experiments are valuable as non-invasive detection of remyelination in pre-clinical models is a indispensable for testing …

   Reviewer #3 (Public Review):

   Lee, Sati and colleagues investigated whether remyelination can be detected non-invasively using MRI in common marmosets with experimental autoimmune encephalomyelitis (EAE). The authors subjected the marmosets to serial MRI during the course of the pathology. The results of PDw and MTR sequences were compared to those of histopathological analyses performed on brain tissue after the animals reached the end point. They found that PDw was more efficient in detecting remyelinated lesions than MTR. The authors also found that early treatment with methylprednisolone had no effect on remyelination. Moreover, the authors observed less remyelination in females compared to males.

   Strengths: These experiments are valuable as non-invasive detection of remyelination in pre-clinical models is a indispensable for testing the efficacy of pro-remyelinating agents prior to clinical studies. Moreover, the animal model used (marmoset EAE) is probably the one that mimicks MS lesions the best, which further supports the importance of the results presented. In addition, the manuscript, particularly the discussion section, is well written, and it suitably addresses and clarifies some issues relevant to the experimental design (low animal numbers, the comparability of different paradigms used to induce EAE, and the potential impact of applying corticosteroids early after lesion detection).

   Weaknesses: The main caveat of this manuscript is that histopathological analyses performed appear insufficient to validate the MRI findings regarding demyelination/remyelination, as well as the activity of demyelinating lesions (acute demyelinating versus chronic). This could be improved by addressing the following points:

   1. The criteria to define different lesion types should be clearly presented (numbers/nature of inflammatory cells, their positivity for different myelin antigens, numbers of oligodendrocytes/OPCs, axonal markers), referenced, and applied when performing histological classification.

   2. Quantification of histological parameters is lacking. Statements such as "decrease in numbers of oligodendrocytes/oligodendrocyte depletion/axonal loss" etc should be corroborated by quantification of specific cellular/axonal markers in lesion areas, as compared to normally myelinated tissue (normally appearing white matter). For example, in Fig 2, based on the image of ASPA/Olig2 labeling, the authors mention "loss of oligodendrocytes", but such loss is apparent only in the small area in the center of the lesion, while the remaining area negative for PLP contains many ASPA+ cells. Quantification of ASPA in the lesions versus NAWM would unequivocally clarify this issue. The same is true for Bielschowsky staining and inflammatory cell markers-quantification would provide solid data. Quantification of inflammatory cells (possibly using additional markers), and co-labelings with different myelin antigens would be very helpful in distinguishing between acute demyelinating and chronic demyelinated lesions (histologically).

   3. Regarding remyelinated lesions, it would be useful to see Luxol Fast Blue staining pattern at lower power and appreciate paler staining of the remyelinated area as compared to non-demyelinated white matter.

   Additional information:
   Related to the above-mentioned point, it would be interesting to present additional histological data/discussion for the animals treated with methylprednisolone (MP). From Figure 5, it seems that MP treatment (applied at week 24-25) resolved demyelination of the first lesion in M#5, but the second lesion in M#5 presented developed after MP treatment was completed (around week 32). This suggests that no differences in the total percentage of remyelinated/total lesions in MP-treated versus MP non-treated animals were observed because most lesions in MP-treated animals developed after MP treatment was completed. It would be interesting to find out whether there were any histological particularities in early lesions in MP-treated animals (even though there should be very few of these). These data would fit nicely with the relevant paragraph presented in Discussion.

   Read the original source
7. 

   Version published to 10.1101/2021.10.27.466044v1 on bioRxiv

   Oct 28, 2021