Single-cell profiling and zebrafish avatars reveal LGALS1 as immunomodulating target in glioblastoma
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (Review Commons)
Abstract
Glioblastoma (GBM) remains the most malignant primary brain tumor, with a median survival rarely exceeding 2 years. Tumor heterogeneity and an immunosuppressive microenvironment are key factors contributing to the poor response rates of current therapeutic approaches. GBM-associated macrophages (GAMs) often exhibit immunosuppressive features that promote tumor progression. However, their dynamic interactions with GBM tumor cells remain poorly understood. Here, we used patient-derived GBM stem cell cultures and combined single-cell RNA sequencing of GAM-GBM co-cultures and real-time in vivo monitoring of GAM-GBM interactions in orthotopic zebrafish xenograft models to provide insight into the cellular, molecular, and spatial heterogeneity. Our analyses revealed substantial heterogeneity across GBM patients in GBM-induced GAM polarization and the ability to attract and activate GAMs – features that correlated with patient survival. Differential gene expression analysis, immunohistochemistry on original tumor samples, and knock-out experiments in zebrafish subsequently identified LGALS1 as a primary regulator of immunosuppression. Overall, our work highlights that GAM-GBM interactions can be studied in a clinically relevant way using co-cultures and avatar models, while offering new opportunities to identify promising immune-modulating targets.
Article activity feed
-
Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Reply to the reviewers
Reviewer #1
Evidence, reproducibility and clarity____:
Summary: the paper suggested a new approach to study in vivo possible interaction between glioblastoma cells and glioblastoma associated macrophages. By using single cells transcriptome profiling and in vitro and in vivo functional experiments the authors also suggested LGALS1 as possible key factor in the suppression of the immune system and a new target for immune modulation in glioma patients. The experimental plan is well described, and the results are beautifully presented using images, clear drawings, and videos.
Major comments: none
Minor comments:
The number of zebrafish embryos …
Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Reply to the reviewers
Reviewer #1
Evidence, reproducibility and clarity____:
Summary: the paper suggested a new approach to study in vivo possible interaction between glioblastoma cells and glioblastoma associated macrophages. By using single cells transcriptome profiling and in vitro and in vivo functional experiments the authors also suggested LGALS1 as possible key factor in the suppression of the immune system and a new target for immune modulation in glioma patients. The experimental plan is well described, and the results are beautifully presented using images, clear drawings, and videos.
Major comments: none
Minor comments:
The number of zebrafish embryos analyzed after the xenograft is highly variable (e.g. 3-18; 4-22 in Figure 6). These numbers can be reported in the results section (not only in the legends) and the authors may comment on them in the discussion. The reproducibility of thexenotransplant experiments is always challenging as it is quite difficult to inject the same number of cells in every embryo and to have the same survival rate of injected cells and of transplanted embryos. For these reasons the volume of each xenograft can vary significantly in different embryos and in different experimental session. Accordingly, the number of macrophages associated to the tumor can vary and the statistical analysis can be deeply influenced by the number of replicates for each experimental group (a group with 3 embryos is very different in term of quality and quantity of information in respect to a group of 18 embryos). It could be useful for the reader, who has no experience in this technique, be aware of the advantages and disadvantages of the procedure including the possible influence of the temperature (34°C instead of 37°C) on the embryo survival and the replication rate of glioma cells or macrophages behavior. Comment on these aspects does not weaken the power and the relevance of the model but unveil the critical aspects that every scientist has to evaluate before planning these kinds of experiments.
__Response: __We agree with Reviewer #1 that the zebrafish avatar model is challenging, and it is difficult to obtain reproducible tumor sizes and survival rates. To be even more transparent about this, we have added a few sentences about the variable n number in the Results section and a critical comment about it in the Discussion section.
An aspect that could be interesting to address, to further validate the avatar model, is to monitor the level of pro-inflammatory cytokines (Tumor Necrosis Factor and Interleukin 1, 6, and 8) that are expressed at basal level in the early developing zebrafish embryos. Do their expression level increase after the xenotransplantation? Can the zebrafish cytokines affect the behavior of glioma associated macrophages (i.e. macrophages polarization)?
__Response: __This is an interesting point, indeed. We have injected murine melanoma (B16) cells into Tg(mpeg1:mCherry-F); Tg(TNFa:eGFP-F) embryos, a TNFa reporter line. Some (but not all) macrophages expressed TNFa and their expression decreased over time, which is consistent with previous reports (Póvoa et al, 2021). We further observed that TNFa-expressing macrophages mostly had a round, “tumor-attacking” phenotype. This is in line with our hypothesis that the tumor induces a phenotype switch in GAMs. Of note, we did not see TNFa expression in the rest of the brain tissue. We would be happy to add this data if deemed useful.
We did not investigate other cytokines in the developing zebrafish, but we believe this is not essential for the following reasons: We are mainly interested in the differences between the patient-derived GBM stem cell cultures (GSSCs), and since they are all used in the same avatar model, we expect that if zebrafish cytokines would have an effect on GAMs and their polarization, this effect would be consistent in all avatars, and can thus be ignored when comparing different GSCCs. More importantly, our findings in the zebrafish avatar model were consistent with those in the in vitro model. We observed the same phenotype switch in the co-culture model, indicating that the key interaction is between tumor cells and macrophages.
Significance____:
Strengths and limitation. The manuscript is the result of a well-orchestrated effort to dissect a biological problem by complementary approaches and provide new data with high impact translational value. The image processing pipeline developed by the authors is a step forward in the in vivo analysis of cells interaction in living embryos. The identification of LGALS1 as a potential target for immune modulation can support the development of new therapeutical strategy implementing chemo- or immunotherapy protocols. The described zebrafish avatar can represent a new tool for personalized drug testing recapitulating in a in vivo model the heterogeneity of GBM found in patients.
Audience: All the scientist interested in cell biology, cancer cell biology, imaging techniques, translational medicine, in vivo models for cancer research, precision medicine.
Reviewer expertise: applied developmental biology
Reviewer #2
Evidence, reproducibility and clarity____:
Finotto et al aim to address the polarisation of macrophages within GBM in their study. To do this, they have developed two different models. The first model is an in-vitro co-culture model of patient derived GSC lines and human monocyte derived macrophages. This model was used for single cell sequencing to understand the transcriptomic changes of macrophages upon contact to GBM cells. The second model is a zebrafish xenograft model. Here GFP labeled GBM cells were transplanted into the larval zebrafish ventricle. These experiments were done in the transgenic mpeg zebrafish which allowed to monitor responses of macrophages in vivo.
In my opinion both models are not sophisticated enough to draw solid conclusions on macrophage polarisation in GBM. The in vitro model is highly artificial and is far from the complex situation in GBM. Within GBM the GAM population represents a heterogenous mix of resident microglia and infiltrating macrophages. These are influenced by the heterogeneous environment (which consists of tumour cells but also other host cells) and show diverse transcriptomic adaptations as shown in rodent models as well as sequencing studies of patient derived tumour samples. Studying monocyte derived macrophages in vitro does not provide any reliable insight.
Response: We understand the reviewer’s concern about the complexity of our in vitro model. However, these simple models are needed to gain more insight into the complex in vivo situation. Others have demonstrated their usefulness in the past (C. Jayakrishnan et al, 2019; Zhou et al, 2022; Hubert et al, 2016; Chen et al, 2020; Coniglio et al, 2016; Li et al, 2022). Moreover, it may be advantageous to look at only two different cell types and unravel their reciprocal interaction, without the influence of other cell types, making it too complex to draw conclusions. We acknowledge that GAMs are a heterogeneous mix of both microglia and bone marrow-derived macrophages. Considering that bone marrow-derived macrophages have been shown to play an important role in tumor progression and are by far the most abundant immune cell population in GBM tumors (which even increases in recurrent GBM) (Pombo Antunes et al, 2021; Abdelfattah et al, 2022), we chose to focus initially on bone marrow-derived macrophages. Notably, it has already been reported that microglia were associated with significantly better survival, suggesting that they are anti-tumorigenic, whereas macrophages were associated with worse survival, suggesting that they are pro-tumorigenic (Pombo Antunes et al, 2021; Abdelfattah et al, 2022). This justifies our approach to focus on this cell type. Furthermore, although this model may be rather simplistic, it allowed us to screen different GSCCs side by side in a standardized way, through which we found an apparent phenotype switch within the macrophages, even without the complex interplay with other cell types. Because the results obtained using the in vitro model were also confirmed in GBM patient material and KO experiments in the zebrafish avatar model, our work shows that reliable and important insights can be derived. This, combined with its simplicity, makes our co-culture model an exceptionally relevant model that is scalable, screenable and allows us to study the effect of perturbations. Finally, the immunosuppressive role of the target we identified using this model, LGALS1, has been previously demonstrated by others (Verschuere et al, 2014; Van Woensel et al, 2017; Chen et al, 2019), which proves our approach is valid.
Although the zebrafish can be a great model to understand the progression of tumours and the role of immune cells, I don't think that the model developed by the authors is suitable to address their questions. Transplantation of GBM cells into the the ventricle of larval zebrafish doesn't seem to be the right approach here. The poor survival of the transplanted cells is a clear indication of that. Many other groups have reported growth and proliferation of human cancer cells in the larval zebrafish. Direct transplantation into the brain parenchyma would be the better approach here. The brain parenchyma would provide the right environment for the GBM cells including a resident microglial population. This would also allow to study the complex mix of microglia and infiltrating macrophages in the context of GBM.
Response: The reviewer does not specify which articles have reported growth and proliferation of human cancer cells in zebrafish larvae. Most research groups reporting this, did not follow tumor growth/proliferation over time or used immortalized cell lines (Vargas-Patron et al, 2019; Pan et al, 2020; Pudelko et al, 2018; Breznik et al, 2017; Vittori et al, 2017; Hamilton et al, 2016), which obviously have a much higher proliferation rate than the patient-derived cell lines used in this work. Second, although the number of patient-derived tumor cells decreases over time, we observed a clear invasive and migratory behavior, indicating that the human tumor cells reside well in the zebrafish microenvironment. Furthermore, it is important to note that the zebrafish avatars are grown at 34°C, a temperature that is suboptimal for tumor cell growth. The tumor cells still proliferate, albeit at a lower rate than at 37°C.
To our knowledge, there is only one publication that reports the growth of patient-derived GBM tumors over time (Almstedt et al, 2022). However, here, zebrafish embryos were grown at 33°C. Also, prior to injection, patient-derived GBM cells were resuspended in medium containing polyvinylpyrrolidone, a polymer that enhances extracellular matrix deposition and cell proliferation. Furthermore, the authors observed substantial differences in proliferative capacity, ranging from growth to decline of signal, and represented only two patient-derived cell lines with growing tumors. Similar to our findings, another article has demonstrated that injected patient-derived GBM tumor cells progressively underwent mitotic arrest, while maintaining an invasive and aggressive growth pattern (Rampazzo et al, 2013).
Although the tumor cells are injected into the hindbrain ventricle, they end up in the brain parenchyma, as evidenced by the presence of the typical brain vasculature of the zebrafish embryo. Notably, in Tg(mpeg1:mCherryF)ump2 zebrafish embryos, both macrophages and microglia are labeled with mCherry, meaning that we have studied both cell types in our zebrafish avatar model. Therefore, we consider the reviewer’s comment to be unfounded.
Reviewer #3
__ Evidence, reproducibility and clarity: __
In this study, Finotto and colleagues developed patient-derived Glioblastoma (GBM) stem cell cultures from 7 patients. These GBM stem cell cultures were either co-cultured in vitro with human macrophages combined with single-cell RNA sequencing or injected into the orthotopic zebrafish xenograft to study live GBM-macrophage/microglia interactions. Authors aimed at studying tumor heterogeneity and GBM-associated macrophages (GAMs) which often exhibit immunosuppressive features that promote tumor progression. Their analyses revealed substantial heterogeneity across GBM patients in GBM-induced macrophages polarization and the ability to attract and activate GAMs - features that correlated with patient survival. Also authors show 3 distinct macrophage subclusters (MC1-3), highlighting that the simple M1/M2 polarization phenotypes is too reductive and there are no clear "markers". Authors associate these profiles with morphology and macrophage behaviour. Differential gene expression analysis, immunohistochemistry on original tumor samples, and knock-out experiments in zebrafish subsequently identified / confirmed that LGALS1 as a primary regulator of immunosuppression.
Cheng et ( DOI: 10.1002/ijc.32102) had previously shown the immunosuppression effect of LGALS1 - but this work shows as a proof of concept that the authors approach is a valuable and interesting approach to find immune regulators.
Response: We fully agree with Reviewer #3. In fact, the immunosuppressive role of LGALS1 has already been described by several research groups (Van Woensel et al, 2017; Verschuere et al, 2014), which indeed proves that our approach is valid. The reference cited by the reviewer was already included in the manuscript, along with other references.
Major comments:
In general claims are supported by date - very carefully presented and well characterized data with numbers, stats. It is an interesting descriptive study that illustrates the complexity and diversity of glioblastoma and the induced TME. I just have a few comments or clarifications that I would like to have elucidated:
I did not understand why not single cell sequence the original tumor - without in vitro passaging and have the original patient population of MACs/microglia and monocytes sequenced? In other words why sequence the in vitro system-with its inherent caveats of in vitro culturing and not the original tumor? Can you please clarify.
Response: We agree with Reviewer #3 that our in vitro model does indeed have caveats inherent to patient-derived cell culture models. However, we chose this model to specifically focus on the reciprocal interaction between GBM tumor cells and macrophages in a way that also allows us to investigate how perturbations affect these interactions. This is not possible when using original tumors (e.g. we cannot make KO cells, as we did for LGALS1, and study the effects of genes of interest). (See also the response to the comment of Reviewer #2)
We do have scRNAseq data from one original tumor sample (LBT123) that is currently being analyzed. Unfortunately, scRNAseq is not available for the other tumor samples. Also, for some of the patients, there is no original material left to use for sequencing. For LBT123, we will compare the scRNAseq data from the original tumor with the in vitro data from the co-culture model.
Mac signatures - out of curiosity- authors could not find TNFa and IFN signatures in any population?
Response: Our analyses did not reveal TNF or IFN as cluster signature genes. However, we did find that TNF expression was slightly higher in MC2, the pro-inflammatory macrophages, although still at low levels. We did not find IFN expression in the macrophage subclusters, but we did find low expression of some IFN receptors. We found a gradient for IFNGR1 with the highest expression in MC3, followed by MC1 and the lowest expression in MC2. IFNGR2 was expressed at slightly higher levels in MC1 compared to the other subclusters. IFNAR1 and IFNAR2 were expressed at comparable low levels in all subclusters. Finally, IFNLR1 expression was higher in MC3 compared to the other two macrophage subclusters. Considering the overall low expression of IFN receptors, we believe that the differences in expression are rather negligible. Furthermore, it has been previously shown that IFN exerts its anti-tumor effect primarily through the responsiveness of endothelial cells and not of myeloid cells, such as macrophages (Kammertoens et al, 2017). Since vascular cells were not present in the co-culture model, low IFN receptor expression is not surprising. We are happy to investigate this in more detail and include it if deemed useful.
8 please show controls side by side with the KO
Response: We thank Reviewer #3 for this comment. We are not quite sure which panel the reviewer is referring to. If it is panel F, we agree with Reviewer #3 and have changed the order of the bars in the revised version. If it is panel E, the corresponding control images are shown in Figure 5I. Since we believe that these images should not be repeated, we have added a figure reference to Figure 5I in the figure legend of Figure 8, in addition to the figure reference already provided in the text. Furthermore, images of all embryos are presented side by side in Figure S8D-E.
Figure 5:* if each pair of images are separated and have the legend on top would be easier to* *read and follow. *
Response: We appreciate the comment that the figure should be intuitively easy to read and follow. However, we have chosen a compromise between overview and visibility of details (e.g. morphological features of GAMs). Since this figure already has the maximum width, the images would become smaller if they needed to be separated. Reducing the size would compromise the visibility of important details.
Significance:
It is a very interesting study, carefully designed and performed that highlights the heterogeneity of glioblastoma and how GBM can modulate the macrophage population into 3 different subsets. This study constitutes a proof of concept of the combination of and in vitro approach and an in vivo approach to find new players and treatments in glioblastoma. I believe that it would be important and interesting to have a the original tumor sequenced to compare to the in vitro platform and understand how the in vitro selection impacts on the tumor biology and even if it changes the heterogeneity and differential composition of the tumor and macrophage profiles.
References:
Abdelfattah N, Kumar P, Wang C, Leu JS, Flynn WF, Gao R, Baskin DS, Pichumani K, Ijare OB, Wood SL, et al (2022) Single-cell analysis of human glioma and immune cells identifies S100A4 as an immunotherapy target. Nat Commun13
Almstedt E, Rosen E, Gloger M, Stockgard R, Hekmati N, Koltowska K, Krona C & Nelander S (2022) Real-time evaluation of glioblastoma growth in patient-specific zebrafish xenografts. Neuro Oncol 24: 726–738
Breznik B, Motaln H, Vittori M, Rotter A & Turnšek TL (2017) Mesenchymal stem cells differentially affect the invasion of distinct glioblastoma cell lines. Oncotarget 8: 25482–25499
Jayakrishnan P, H. Venkat E, M. Ramachandran G, K. Kesavapisharady K, N. Nair S, Bharathan B, Radhakrishnan N & Gopala S (2019) In vitro neurosphere formation correlates with poor survival in glioma. IUBMB Life 71: 244–253
Chen JWE, Lumibao J, Leary S, Sarkaria JN, Steelman AJ, Gaskins HR & Harley BAC (2020) Crosstalk between microglia and patient-derived glioblastoma cells inhibit invasion in a three-dimensional gelatin hydrogel model. J Neuroinflammation 17
Chen Q, Han B, Meng X, Duan C, Yang C, Wu Z, Magafurov D, Zhao S, Safin S, Jiang C, et al (2019) Immunogenomic analysis reveals LGALS1 contributes to the immune heterogeneity and immunosuppression in glioma. Int J Cancer145: 517–530
Coniglio S, Miller I, Symons M & Segall JE (2016) Coculture assays to study macrophage and microglia stimulation of glioblastoma invasion. Journal of Visualized Experiments 2016
Hamilton L, Astell KR, Velikova G & Sieger D (2016) A zebrafish live imaging model reveals differential responses of microglia toward glioblastoma cells in vivo. Zebrafish 13: 523–534
Hubert CG, Rivera M, Spangler LC, Wu Q, Mack SC, Prager BC, Couce M, McLendon RE, Sloan AE & Rich JN (2016) A three-dimensional organoid culture system derived from human glioblastomas recapitulates the hypoxic gradients and cancer stem cell heterogeneity of tumors found in vivo. Cancer Res 76: 2465–2477
Kammertoens T, Friese C, Arina A, Idel C, Briesemeister D, Rothe M, Ivanov A, Szymborska A, Patone G, Kunz S, et al(2017) Tumour ischaemia by interferon-γ resembles physiological blood vessel regression. Nature 545: 98–102
Li H, Yan X & Ou S (2022) Correlation of the prognostic value of FNDC4 in glioblastoma with macrophage polarization. Cancer Cell Int 22
Pan H, Xue W, Zhao W & Schachner M (2020) Expression and function of chondroitin 4-sulfate and chondroitin 6-sulfate in human glioma. FASEB Journal 34: 2853–2868
Pombo Antunes AR, Scheyltjens I, Lodi F, Messiaen J, Antoranz A, Duerinck J, Kancheva D, Martens L, De Vlaminck K, Van Hove H, et al (2021) Single-cell profiling of myeloid cells in glioblastoma across species and disease stage reveals macrophage competition and specialization. Nat Neurosci 24: 595–610
Póvoa V, Rebelo de Almeida C, Maia-Gil M, Sobral D, Domingues M, Martinez-Lopez M, de Almeida Fuzeta M, Silva C, Grosso AR & Fior R (2021) Innate immune evasion revealed in a colorectal zebrafish xenograft model. Nat Commun12
Pudelko L, Edwards S, Balan M, Nyqvist D, Al-Saadi J, Dittmer J, Almlöf I, Helleday T & Bräutigam L (2018) An orthotopic glioblastoma animal model suitable for high-throughput screenings. Neuro Oncol 127: 415
Rampazzo E, Persano L, Pistollato F, Moro E, Frasson C, Porazzi P, Della Puppa A, Bresolin S, Battilana G, Indraccolo S, et al (2013) Wnt activation promotes neuronal differentiation of glioblastoma. Cell Death Dis 4
Van Woensel M, Mathivet T, Wauthoz N, Rosière R, Garg AD, Agostinis P, Mathieu V, Kiss R, Lefranc F, Boon L, et al(2017) Sensitization of glioblastoma tumor micro-environment to chemo- and immunotherapy by Galectin-1 intranasal knock-down strategy. Sci Rep 7: 1–14
Vargas-Patron LA, Agudelo-Dueñãs N, Madrid-Wolff J, Venegas JA, González JM, Forero-Shelton M & Akle V (2019) Xenotransplantation of human glioblastoma in zebrafish larvae: in vivo imaging and proliferation assessment. Biol Open 8
Verschuere T, Toelen J, Maes W, Poirier F, Boon L, Tousseyn T, Mathivet T, Gerhardt H, Mathieu V, Kiss R, et al (2014) Glioma-derived galectin-1 regulates innate and adaptive antitumor immunity. Int J Cancer 134: 873–884
Vittori M, Breznik B, Hrovat K, Kenig S & Lah TT (2017) RECQ1 helicase silencing decreases the tumour growth rate of U87 glioblastoma cell xenografts in zebrafish embryos. Genes (Basel) 8
Zhou F, Shi Q, Fan X, Yu R, Wu Z, Wang B, Tian W, Yu T, Pan M, You Y, et al (2022) Diverse macrophages constituted the glioma microenvironment and influenced by PTEN status. Front Immunol 13
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #3
Evidence, reproducibility and clarity
In this study, Finoto and colleagues developed patient-derived Glioblastoma (GBM) stem cell cultures from 7 patients. These GBM stem cell cultures were either co-cultured in vitro with human macrophages combined with single-cell RNA sequencing or injected into the orthotopic zebrafish xenograft to study live GBM-macrophage/microglia interactions. Authors aimed at studying tumor heterogeneity and GBM-associated macrophages (GAMs) which often exhibit immunosuppressive features that promote tumor progression. Their analyses revealed substantial heterogeneity across GBM patients in GBM-induced macrophages polarization and the ability to …
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #3
Evidence, reproducibility and clarity
In this study, Finoto and colleagues developed patient-derived Glioblastoma (GBM) stem cell cultures from 7 patients. These GBM stem cell cultures were either co-cultured in vitro with human macrophages combined with single-cell RNA sequencing or injected into the orthotopic zebrafish xenograft to study live GBM-macrophage/microglia interactions. Authors aimed at studying tumor heterogeneity and GBM-associated macrophages (GAMs) which often exhibit immunosuppressive features that promote tumor progression. Their analyses revealed substantial heterogeneity across GBM patients in GBM-induced macrophages polarization and the ability to attract and activate GAMs - features that correlated with patient survival. Also authors show 3 distinct macrophage subclusters (MC1-3), highlighting that the simple M1/M2 polarization phenotypes is too reductive and there are no clear "markers". Authors associate these profiles with morphology and macrophage behaviour.
Differential gene expression analysis, immunohistochemistry on original tumor samples, and knock-out experiments in zebrafish subsequently identified / confirmed that LGALS1 as a primary regulator of immunosuppression. Cheng et ( DOI: 10.1002/ijc.32102) had previously shown the immunosuppression effect of LGALS1 - but this work shows as a proof of concept that the authors approach is a valuable and interesting approach to find immune regulators.
Major comments
In general claims are supported by date - very carefully presented and well characterized data with numbers, stats. It is an interesting descriptive study that illustrates the complexity and diversity of glioblastoma and the induced TME. I just have a few comments or clarifications that I would like to have elucidated:
- I did not understand why not single cell sequence the original tumor - without in vitro passaging and have the original patient population of MACs/microglia and monocytes sequenced? In other words why sequence the in vitro system-with its inherent caveats of in vitro culturing and not the original tumor? Can you please clarify
- Mac signatures - out of curiosity- authors could not find TNFa and IFN signatures in any population?
- Fig. 8 please show controls side by side with the KO
- Figure 5 if each pair of images are separated and have the legend on top would be easier to read and follow.
Significance
It is a very interesting study, carefully designed and performed that highlights the heterogeneity of glioblastoma and how GBM can modulate the macrophage population into 3 different subsets. This study constitutes a proof of concept of the combination of and in vitro approach and an in vivo approach to find new players and treatments in glioblastoma.
I believe that it would be important and interesting to have a the original tumor sequenced to compare to the in vitro platform and understand how the in vitro selection impacts on the tumor biology and even if it changes the heterogeneity and differential composition of the tumor and macrophage profiles.
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #2
Evidence, reproducibility and clarity
Finotto et al aim to address the polarisation of macrophages within GBM in their study. To do this, they have developed two different models. The first model is an in-vitro co-culture model of patient derived GSC lines and human monocyte derived macrophages. This model was used for single cell sequencing to understand the transcriptomic changes of macrophages upon contact to GBM cells. The second model is a zebrafish xenograft model. Here GFP labeled GBM cells were transplanted into the larval zebrafish ventricle. These experiments were done in the transgenic mpeg zebrafish which allowed to monitor responses of macrophages in vivo.
I…
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #2
Evidence, reproducibility and clarity
Finotto et al aim to address the polarisation of macrophages within GBM in their study. To do this, they have developed two different models. The first model is an in-vitro co-culture model of patient derived GSC lines and human monocyte derived macrophages. This model was used for single cell sequencing to understand the transcriptomic changes of macrophages upon contact to GBM cells. The second model is a zebrafish xenograft model. Here GFP labeled GBM cells were transplanted into the larval zebrafish ventricle. These experiments were done in the transgenic mpeg zebrafish which allowed to monitor responses of macrophages in vivo.
In my opinion both models are not sophisticated enough to draw solid conclusions on macrophage polarisation in GBM. The in vitro model is highly artificial and is far from the complex situation in GBM. Within GBM the GAM population represents a heterogenous mix of resident microglia and infiltrating macrophages. These are influenced by the heterogeneous environment (which consists of tumour cells but also other host cells) and show diverse transcriptomic adaptations as shown in rodent models as well as sequencing studies of patient derived tumour samples. Studying monocyte derived macrophages in vitro does not provide any reliable insight.
Although the zebrafish can be a great model to understand the progression of tumours and the role of immune cells, I don't think that the model developed by the authors is suitable to address their questions. Transplantation of GBM cells into the the ventricle of larval zebrafish doesn't seem to be the right approach here. The poor survival of the transplanted cells is a clear indication of that. Many other groups have reported growth and proliferation of human cancer cells in the larval zebrafish. Direct transplantation into the brain parenchyma would be the better approach here. The brain parenchyma would provide the right environment for the GBM cells including a resident microglial population. This would also allow to study the complex mix of microglia and infiltrating macrophages in the context of GBM.
Significance
In my opinion both models are not sophisticated enough to draw solid conclusions on macrophage polarisation in GBM.
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #1
Evidence, reproducibility and clarity
Summary: the paper suggested a new approach to study in vivo possible interaction between glioblastoma cells and glioblastoma associated macrophages. By using single cells transcriptome profiling and in vitro and in vivo functional experiments the authors also suggested LGALS1 as possible key factor in the suppression of the immune system and a new target for immune modulation in glioma patients. The experimental plan is well described, and the results are beautifully presented using images, clear drawings, and videos.
Major comments: none
Minor comments: The number of zebrafish embryos analyzed after the xenograft is highly …
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #1
Evidence, reproducibility and clarity
Summary: the paper suggested a new approach to study in vivo possible interaction between glioblastoma cells and glioblastoma associated macrophages. By using single cells transcriptome profiling and in vitro and in vivo functional experiments the authors also suggested LGALS1 as possible key factor in the suppression of the immune system and a new target for immune modulation in glioma patients. The experimental plan is well described, and the results are beautifully presented using images, clear drawings, and videos.
Major comments: none
Minor comments: The number of zebrafish embryos analyzed after the xenograft is highly variable (e.g. 3-18; 4-22 in Figure 6). These numbers can be reported in the results section (not only in the legends) and the authors may comment on them in the discussion. The reproducibility of the xenotransplant experiments is always challenging as it is quite difficult to inject the same number of cells in every embryo and to have the same survival rate of injected cells and of transplanted embryos. For these reasons the volume of each xenograft can vary significantly in different embryos and in different experimental session. Accordingly, the number of macrophages associated to the tumor can vary and the statistical analysis can be deeply influenced by the number of replicates for each experimental group (a group with 3 embryos is very different in term of quality and quantity of information in respect to a group of 18 embryos). It could be useful for the reader, who has no experience in this technique, be aware of the advantages and disadvantages of the procedure including the possible influence of the temperature (34{degree sign}C instead of 37{degree sign}C) on the embryo survival and the replication rate of glioma cells or macrophages behavior. Comment on these aspects does not weaken the power and the relevance of the model but unveil the critical aspects that every scientist has to evaluate before planning these kinds of experiments. An aspect that could be interesting to address, to further validate the avatar model, is to monitor the level of pro-inflammatory cytokines (Tumor Necrosis Factor and Interleukin 1, 6, and 8) that are expressed at basal level in the early developing zebrafish embryos. Do their expression level increase after the xenotransplantation? Can the zebrafish cytokines affect the behavior of glioma associated macrophages (i.e. macrophages polarization)?
Significance
Strengths and limitation. The manuscript is the result of a well-orchestrated effort to dissect a biological problem by complementary approaches and provide new data with high impact translational value. The image processing pipeline developed by the authors is a step forward in the in vivo analysis of cells interaction in living embryos. The identification of LGALS1 as a potential target for immune modulation can support the development of new therapeutical strategy implementing chemo- or immunotherapy protocols. The described zebrafish avatar can represent a new tool for personalized drug testing recapitulating in a in vivo model the heterogeneity of GBM found in patients.
Audience: All the scientist interested in cell biology, cancer cell biology, imaging techniques, translational medicine, in vivo models for cancer research, precision medicine.
Reviewer expertise: applied developmental biology
-
