An analogue of the Prolactin Releasing Peptide reduces obesity and promotes adult neurogenesis

  1. Sara KM Jörgensen
  2. Alena Karnošová
  3. Simone Mazzaferro
  4. Oliver Rowley
  5. Hsiao-Jou Cortina Chen
  6. Sarah J Robbins
  7. Sarah Christofides
  8. Florian T Merkle
  9. Lenka Maletínská
  10. David Petrik

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Hypothalamic Adult Neurogenesis (hAN) has been implicated in regulating energy homeostasis. Adult-generated neurons and adult Neural Stem Cells (aNSCs) in the hypothalamus control food intake and body weight. Conversely, diet-induced obesity (DIO) by high fat diets (HFD) exerts adverse influence on hAN. However, the effects of anti-obesity compounds on hAN are not known. To address this, we administered a lipidized analogue of an anti-obesity neuropeptide, Prolactin Releasing Peptide (PrRP), so-called LiPR, to mice. In the HFD context, LiPR rescued the survival of adult-born hypothalamic neurons and increased the number of aNSCs by reducing their activation. LiPR also rescued the reduction of immature hippocampal neurons and modulated calcium dynamics in iPSC-derived human neurons. In addition, some of these neurogenic effects were exerted by another anti-obesity compound, Liraglutide. These results show for the first time that anti-obesity neuropeptides influence adult neurogenesis and suggest that the neurogenic process can serve as a target of anti-obesity pharmacotherapy.

Article activity feed

  1. Version published to 10.1038/s44319-023-00016-2
  2. Review Commons

    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

    General Statement

    We thank the reviewers for a thorough review that will help us to improve the manuscript in the revision process. In our opinion, all three reviewers found the manuscript interesting, novel, and relevant for a broader readership. The reviewers suggested performing additional analyses of cell quantification from existing brain tissue or from newly generated tissue. All reviewers identified several shared concerns that we are happy to address by additional experiments and analyses to improve our manuscript. The reviewers suggested including the Control Diet + LiPR treatment group to further characterize the effects of LiPR on adult neurogenesis outside the context of the High Fat Diet. Also, the reviewers suggested including built upon the analysis of tanycytes and their proliferation. Some of these analyses will require generating new experimental animals, however, most analyses can be performed from already available brain tissue or previously collected confocal microscope images. Because we had anticipated some of the possible concerns, we have placed mice in the experiment already in February 2023. These mice are in the 4-month treatment group of Control Diet + LiPR. We will collect the brain tissue at the end of May 2023 and will analyze it in June and July 2023. In April and May 2023, we will work on analyses from existing tissue or images as described in detail below. We estimate that the suggested analyses are all feasible and should be manageable in 3 months. In fact, we are pleasantly surprised by the favorable nature of the reviews, especially from the reviewer 1 and 3, which allowed us to address around 50% of comments already as demonstrated in this revision plan (see section 3). Therefore, we are confident that we will be able to address the remaining concerns to full satisfaction of all relevant reviewers’ comments.

    Reviewer 1

    *In this manuscript, Jorgensen and colleagues describe their findings on the action of a palmitoylated form of prolactin-release peptide (LiPR) on neural stem cells (NSC) in the adult mouse hypothalamus and adult mouse hippocampus. Their main conclusion is that LiPR can counteract the effects of high-fat diet (HFD) and rescue some of the adverse effects of HFD. Specifically, the authors provide evidence that:

    • Exposure to HFD reduces the number of presumptive adult neural stem cells (NSCs) in the adult hypothalamus, whereas exposure to LiPR reverses this trend.
    • The results suggest that LiPR reduces the proliferation of alpha-tanycytes and/or their progeny in the hypothalamus in the context of HFD, with Liraglutide acting similarly. In contrast, while LiPR also suppresses proliferation in the SGZ, Liraglutide works there in the opposite direction.
    • LiPR also helps the survival of adult-born hypothalamic neurons.
    • Reduction of proliferation by LiPR suggests a model where LiPR increases the number of NSCs presumably by reducing their rate of activation.
    • The results suggest that LiPR promotes expression of PrRP receptors in the hypothalamic neurons, suggesting that PrRP may act directly on such neurons (and tanycytes?) in vivo.
    • The authors also show that HFD and LiPR alter gene expression profiles of the MBH cells, with HFD, but not LiPR, inducing myelination-related genes.
    • Finally, they show that PrRP stimulates an increase in Ca2+ in in vitro-derived human hypothalamic neurons.
    • The authors conclude that LiPR may be reducing activation and proliferation of the hypothalamic stem cells and thereby preserve their pool from exhaustion, which was stimulated by HFD. The manuscript presents interesting data and is clearly written. There are several comments, mainly editorial.*

    RESPONSE: We thank the reviewer for the favorable and positive assessment of our manuscript and for finding our study to be interesting to a broad audience and well written, with most comments described by the reviewer as “editiorial”. Below, we address the reviewer’s concerns in a detailed revision plan.

      • It is unclear why most of the experiments do not include the control+LiPR group. Even though the focus of the study was the action of LiPR in the context of HFD, questions remain regarding the action of LiPR per se. Is LiPR (or Liraglutide, for that matter) completely inactive on the normal diet background, with respect to neurogenesis in the hypothalamus and the hippocampus? Whether the Response is positive or negative, it would give a much better understanding of the action of LiPR - does it regulate neurogenesis in various physiological contexts, or does it only kick in with a particular type of diet? In fact, this was examined (see Supplementary figures), but only for the cells in culture and, when performed with animals, was limited to 7 and 21 days, rather than 4 months, which would have been much more informative.* RESPONSE: We thank the reviewer for this suggestion. We agree that including the Control Diet + LiPR group for the 4-month HFD group would complement the results from the 7 and 21 days. We will generate this treatment group for the 4-month HFD group and analyze the effect of LiPR on aNSC and adult-generated neurons. These mice in the 4-month treatment are in the experiment already from February 2023 and we plan to analyze their brain sections in June and July 2023.
    • The question above is also relevant when considering the conclusions on the potential depletion of the stem cell pool (again, whether in the hypothalamus or the hippocampus), particularly at the 4-month time point. The mice are ~6 months old by that time, and neurogenesis in both regions is expected to decrease by that time. Are LiPR or Liraglutide able to suppress or exacerbate this decrease? Can they be used to mitigate this decrease when mice are on a regular diet?*

    RESPONSE: This concern will be addressed by analyzing the Control + LiPR mice for the 4-month HFD group (see our response to the point 1 above). We will analyze neural stem cells in the Hypothalamic Ventricular Zone and neural progenitors in the Median Eminence of these mice to address whether LiPR treatment changes the time-dependent decrease in both cell populations.

    • A somewhat related issue is that, in most cases, only the percentage or the density of cells are shown on the graphs, rather than the absolute numbers (at least for some cases). This sometimes complicates the comparisons; for instance, does the surface of the hypothalamus change between 2 and 6 months of age? The tanycytes' number stays, apparently, the same (e.g., Fig. 2) but the production of new neurons is supposed to fall dramatically.*

    RESPONSE: We thank the reviewer for this comment. We agree that the quantification of absolute number of cells is the preferable approach that we have used in our previous publications on subventricular (SVZ) or subgranular (SGZ) neurogenesis. However, hypothalamic adult neurogenesis is dispersed over much larger volume of tissue than neurogenesis in the SVZ or SGZ, which is confined to narrow tissue compartments. As we do not have access to a confocal microscope with stereological software, absolute quantification in entire MBH is not feasible. Nevertheless, we believe that our quantification of cell density provides an unbiased and informative approach that allowed us to compare the effects of LiPR and diet on the neurogenic process.

    • The authors write "LiPR may prevent stem cells from exhaustion, induced by HFD" - but it is not clear that HFD indeed leads to exhaustion - there is no statistically significant difference in the number of the stem cells (alpha-tanycytes) between the control and HFD or between HFD at 1, 3, or 12 weeks.*

    RESPONSE: We thank the reviewers for their insights. We adjusted the interpretation to better reflect our results. On line 442, we replaced the original statement “The lower cell activation may protect the stem cell pool from exhaustion elicited by the HFD“ with a new one, “The lower cell activation may protect the stem cell pool from exhaustion elicited by the HFD“.

    • Numerous papers show that the rate of production of new adult hypothalamic neurons (mainly those derived from beta-tanycytes) drops drastically within the first several weeks of mouse life. Does HFD accelerate, and LiPR mitigate, this decrease? Perhaps one can calculate the numbers from the graphs, but it would help if this is explained in the text of the manuscript. Also, it is not always clear whether specific experiments were performed with the zones of the hypothalamic wall that only contain alpha-tanycytes.*

    RESPONSE: Our results show that LiPR rescues the HFD-induced reduction in adult-generated hypothalamic neurons only in the context of 4-month HFD but not in the 7- and 21-day HFD. In the methods (line 877), we specify that “the Region of Interest (ROI) quantified included the MBH parenchyma with the Arcuate (Arc), DMN and Ventromedial (VMN) Nuclei and the Medial Eminence (ME)”. In the results of the revised manuscript (lines 301-303), we highlighted the areas of the ROI. Upon the request of Reviewer 3 (comment 14), we included new data on quantification of BrdU+ neurons in the Arcuate Nucleus (S.Fig.5O). This data show that 21d HFD increases the number of new neurons in ArcN, which is reversed by LiPR or Liraglutide (text added to results and discussion on lines 309-313 and 468-474, respectively). Finally, in the discussion (lines 464-488), it is stated that HFD and/or LiPR had no effect on number of new hypothalamic neurons or cells in the MBH parenchyma in the 7- and 21-day groups and this is discussed in the context of relevant literature.

    • A sharp increase in PCNA+ cells in the hippocampus at the 21-day time point, both in the control and in the HFD and HFD/LiPR groups (Fig. S2f) is a little puzzling because neither the Dcx+ nor the Ki67+ cells show this increase.*

    RESPONSE: We agree with the reviewer that this increase in the number of PCNA+ cells is puzzling. We quantified the number of PCNA+ cells twice by two different people, always getting the same result. Given that this is a minor result in a supplementary figure, we would prefer not analyzing this again, unless the reviewer would insist on it.

    • The study deals with several agents and several processes; a simple scheme that summarizes authors' conclusions might help to better understand the relationships between those agents and processes.*

    RESPONSE: We thank the reviewer for this useful suggestion. We included a summarizing schematic in the revised manuscript as the new Figure 6. We will update the schematic for the final revised manuscript, when we will incorporate the new analyses.

    ***Referee cross-commenting**

    I agree, the lack of the LiPR group complicates the interpretation of the results. I also agree that the experiments with vimentin staining, calcium increase, and even with neurospheres do not add much to the main questions that this study attempts to Response, and I'd rather see a more thorough analysis of the activation and differentiation data. I also want to reiterate that the concept of LiPR/PrRP preventing the exhaustion of the hypothalamic stem cell pool is not clear, because it is not shown that this pool does actually get exhausted under normal or HFD conditions. This latter issue again requires the LiPR-alone group. Also, as a clarification - I wrote about 1 month required to compete the revision assuming that the authors actually have the data on the Control+LipR group or at least the specimens available, mainly because the supplementary material shows results with this group, at least with the neurospheres. If this group is fully missing, then the effort will obviously take a longer time.

    Reviewer #1 (Significance (Required)):

    The provided evidence suggests, for the first time, that PrRP prevents the loss of the neural stem cells population in the adult hypothalamus that was diminished by obesity and HFD. This finding might be interesting to a broad audience.

    Reviewer 2

    *The authors examine the effect of an anorexigenic drug, LiPR in the context of treatment with high fat diet (HFD) and with a special focus on hypothalamic neural stem/progenitor cells and neurogenesis. The work is mostly based on mice and a barrage of different techniques (confocal imaging, cell cultures with time lapse, gene expression...) are used. The results are interesting because they address the yet-poorly understood implication of hypothalamic neurogenesis in food intake and energy balance. The results point at complex effects at different levels (neural stem cells, neurons, division, survival...). The experimental approach is sometimes thorough in the treatment of details on the one hand, it also lacks of consistency on the other, and as a result the conclusions lack strength. There is a number of experiments that sometimes seem unrelated and this hurts the comprehension of the manuscript, specially in lieu of the complexity of the results obtained.

    RESPONSE: We thank the reviewer for finding our results interesting and relevant. We will strive to improve the consistency of our results in the revised manuscript to satisfy the reviewer’s concerns.

    1. A major issue is the lack of a LiPR-only group, which would much facilitate the interpretation of the results. The effect of LiPR alone is however tested, but only in comparison with the Control in one of the in vitro experiments (S.Fig. 3) RESPONSE: We agree with the reviewer that expanding on the LiPR-only effect would facilitate the interpretation of the results (see concern 1 and 2 of reviewer 2). We want to emphasize, however, that we analyzed the HFD-independent LiPR effects not only in vitro but also in vivo by quantifying the number of BrdU+ cells and neurons in the MBH of mice exposed to 21-day HFD (S.Fig. 5 O-Q) and by including the Control Diet + LiPR in the RNAseq experiment (Fig.5C). Nevertheless, we will analyze the number of alpha tanycytes and proliferating cells for the 21-day Control Diet + LiPR treatment group. And we will generate mice treated with Control Diet + LiPR to complement the 4-month group. In this Control Diet + LiPR group, we will quantify the number of tanycytes and number of BrdU+ cells and neurons.
    • As plotted, in Fig 1B is difficult to interpret the effect of HFD and LiPR, might be using percentage and noting the statistical differences as in the other would help. It looks like HFD has no effect compared to control on weight and only at the end LiPR could have an effect. On the other hand, after 4 months, HFD mice are clearly above the controls and it is then, albeit when weight gain has reached a plateau, that LiPR has an effect. The election of these arbitrary paradigms and their drawbacks has to be better explained.*

    RESPONSE: We thank the reviewer for the comment. We analyzed the effect of HFD and/or LiPR on the body weight for the 21-day group (Fig.1B) in the original manuscript (lines 111-115). The two-way, repeated measure ANOVA revealed no effect of the treatment on the body weight in the 7-day group, however, it revealed the effect of the duration of treatment on the body weight in the 21-day group. As suggested by the reviewer, we included the Control Diet + LiPR in the 21-day group (Fig.1B). We analyzed the data with ANOVA and found that the treatment has a statistically significant effect on the body weight, however, without any statistical difference between treatment groups (lines 112-116 in the revised manuscript). In addition, we will include the Control Diet + LiPR in the 4-month group.

    Why was the proportion of GPR10+BrdU+MAP2+ cells only assessed in control mice and no in the experimental groups if its expression in overall neurons changes? This suggests that the receptor is expressed in neurons. Interestingly, exposure to 21d HFD reduced density of GPR10, which was rescued by LiPR administration (Fig.1L). Why was this time point chosen and not the longer-term one? What is the consequence of the alterations in the potential number of GPR10, specially in relation to the administration of LiPR? This clarification is important because a 14-day treatment was chosen for the in vitro experiments in which LiPR, but not HFD, seems to have an effect on cell proliferation. Might be it would have been more useful to use a paradigm in which HFD has an effect to better compare with in vivo work and for the rationale of the work. "Besides GPR10, we co-localized neuronal cytoskeleton structures with NPFFR2 in the MBH (Fig.1O-P)..." Why were not GPR10 and NPFF2 analyzed in a similar and consistent manner ? It is confusing.

    RESPONSE: The proportion of GPR10+BrdU+Map2+ neurons was quantified to address whether new neurons express the PrRP receptor. We chose to analyze the proportion of GPR10+BrdU+Map2+ neurons at the 21d time-point because we had the most robust data for this or related time points in vitro and in vivo. We will emphasize this in the text. But we prefer not to analyze the effect of LiPR on the density or expression of GPR10 or NPFF2 for all time points. We consider this to be beyond the scope and focus of the manuscript.

    The number of GFAP+ α-tanycytes is not significantly changed by HFD therefore LiPR does not rescue, but rather increases the number of GFAP+ α-tanycytes in the 7-day setting. There are no differences among groups later, the effect is lost by 21 days, therefore there is a transient excess of GFAP+ α-tanycytes which later "disappear" in the LiPR group. The authors state that LiPR rescues the decrease in "htNSCs", but after 21 days the number of the GFAP+ α-tanycytes is the same in all groups without the need of LiPR. There is no experimental follow up (addressing proliferation and survival of these cells) and the conclusions stated in the text (results and discussion) are not really supported by the data. The in vitro experiments could be a complement, but are no substitute for the missing in vivo exploration.

    RESPONSE: We thank the reviewer for this comment. We agree that we did not correctly interpret the data. On line 158, we replaced the original statement “This suggests that short LiPR rescues HFD-induced reduction in the number of htNSCs” with a new one that reflects of date correctly, “This suggests that short LiPR increases the number of htNSCs. In our revision plan, we will quantify the number of proliferating tanycytes to complement our in vitro results.

    • The fact that cell division is "rarely found" (Rax GFAP) experiments also push for further investigation. It is difficult to see that relevance of the inclusion of the vimentin staining experiment if there is no further exploration. The effect of LiPR is only transient, in the 7-day paradigm and as the parameter evaluated is the proportion of vimentin+ tanycytes among GFAP+ tanycytes it could only be reflecting increased expression of the filament. "Nevertheless, we did not observe a statistically different change in the area occupied by Rax+ tanycytes (Fig.2H)." Why did the authors use Rax only for this experiment if "GFAP+ α-tanycytes which are considered the putative htNSCs?" What is the justification for not seeing changes in relation to the results reported in Fig 2D-F? "Because Vimentin is associated with nutrient transport in cells and with metabolic response to HFD 52-54, we quantified the proportion of GFAP+ tanycytes expressing Vimentin (Fig.2F)." It is difficult to see that relevance of the inclusion of the vimentin staining experiment if there is no further exploration. The effect of LiPR is only transient, in the 7-day paradigm and as the parameter evaluated is the proportion of vimentin+ tanycytes among GFAP+ tanicytes it could only be reflecting increased expression of the filament.*

    RESPONSE: Because Vimentin is a marker of neural stem cells and alpha tanycytes, we quantified the number of GFAP+Vimentin+ tanycytes to complement the quantification of GFAP+ alpha tanycytes. We are sorry that this was not clear, and we highlighted this connection in the revised manuscript (line 165). Because Rax is expressed in alpha tanycytes, we expected that LiPR will increase Rax in the Hypothalamic Ventricular Zone (HVZ). We agree with the reviewer that further investigation may be useful, and we will quantify the number of alpha tanycytes positive for Rax instead of determining only the volume of Rax+ tissue. We will quantify Rax+GFAP+ neural stem cells in the HVZ and Rax+GFAP+ neural progenitors (so-called beta tanycytes) in the Median Eminence to improve characterization of the cell dynamics in vivo.

    • Why there is no Ki67 experiment in the 7-day paradigm if that is the timepoint in which changes in the number or proportion of GFAP+ tanycytes are observed? PCNA was then used but only in the 21-day paradigm. What is the interpretation and relevance of these data? What are the non-htNSCs proliferating cells, whose dynamics are different from the changes in the number or proportion of htNSCs that could be potentially related to changes in mitosis? Again, I think it would be much useful for the work to explore in detail the changes in the putative htNSCs than investing in experiments that only add confusion.*

    __RESPONSE: __We apologize if the data presentation is confusing. We will include the quantification of the Ki67+ cells for the 7-day time point. In the MBH, many cell types undergo mitosis, including the oligodendrocyte precursor cells, microglia, astrocytes, and infiltrating macrophages. However, characterizing the identify of all these different cell types in response to the HFD and/or LiPR is beyond the scope of this study. To resolve whether HFD and/or LiPR influence proliferating aNSCs, we will quantify the proliferating cells in the HVZ, which will allow us to separate the proliferating aNSCs from all other proliferating cell types in the MBH.

    • The inclusion of Liraglutide + HFD, (not Liraglutide alone) only in some of the experiments is pointless if there is no direct comparison with LiPR and a timepoint is missing. In S.Fig 3, Fig. 5 and S.Fig 7 LFD (low fat diet?) is used in several occasions as in: "on reducing number of PCNA+ cells in 21d protocol (one-way ANOVA (OWA), F(2,12) = 16.66, p = 0.0003) when compared to both LFD and HFD groups". Is this the control diet?*

    RESPONSE: We apologize for the confusion caused by labelling the conditions of the Control Diet inconsistently. In some figures (e.g., Fig.2, S.Fig.3, Fig.4), we labelled the Control Diet as “Control”, whereas in some other figures (e.g., Fig.5, S.Fig.7) we labelled the Control Diet as “LFD” (Low Fat Diet). In all experiments and figures, the used Control Diet was identical. We unified the labelling of the Control Diet in all figures and in the text of the revised manuscript. Respectfully, we do not agree that including the Liraglutide data is pointless. We included the Liraglutide in the context of the HFD as a direct comparison with the HFD + LiPR group to demonstrate that the two anti-obesity compounds exert differential effects on adult neurogenesis. Such comparison has not been done before in analyzing adult neurogenesis and is valuable for better understanding of functions of these anti-obesity compounds.

    • The final experiment shows that application of hPrRP31, a variation of LiPR, causes an immediate calcium increase in human induced pluripotent stem cell-derived hypothalamic nucleus. This finding is interesting in itself because it brings light about the function of the receptor/s. It would have been very useful to test what other receptors mentioned to bind LiPR is mediating the effect. In any case, the focus of the work are the neural stem/progenitor cells responsible for neurogenesis and the changes in their properties because of HFD and LiPR, therefore I would trade these experiments for a more thorough and detailed dissection of these effects.*

    RESPONSE: We thank the reviewer for recognizing the relevance of the experiments with the hiPSC-derived neurons. As described in the comments above, we will conduct additional experiments to address the effect of LiPR on aNSCs and proliferation to more thoroughly as suggested by the reviewer.

    __Minor points: __ A.* Introduce "GLP-1RA"*

    __RESPONSE: __We thank the reviewer for identifying this omission. We introduced the term in the revised manusript (line 50).

      • "HFD-induced inflammation and astrogliosis in the hypothalamus 45,46, whereas the long (4mo) protocol leads to DIO" Are these notions exclusive?* __RESPONSE: __This statement emphasized that HFD-induced inflammation and astrogliosis precede obesity. We prefer to leave the statement as it is.
      • LiPR displays no effects on astrocytes" "Displays" is not the correct term.* RESPONSE: We replaced the term “display” with the word “show” in the revised manuscript (line 342).

    ***Referee cross-commenting**

    I think we all referees agree for the most part. The main concern stated by all of us is the lack of a LiPR-alone group. The rest of the concerns are also related or complementary. In my opinion the mostly common view by the referees is reasuring.

    Reviewer #2 (Significance (Required)):

    The strengths of the work are its novelty in the field and the variety of techniques employed. The work has the potential of unveiling mechanistic insight into the regulation of neural stem/progenitor cells and neurogenesis. The main audience of this work would be the community working on this field. The lack of experiments testing that the changes observed actually participate in food intake prevent the work from being of relevance for a broader audience (food intake, energy balance, obesity...). The limitations are the descriptive nature of the work and the lack of a consistent and systematic experimental design that would allow to extract solid conclusions upon to which build upon future research.

    Reviewer 3

    The work of Jörgensen et al describes the effect of a lipidized analogue of the prolactin releasing peptide (LiPR) on the mouse metabolism in response to high fat diet (HFD) and on hypothalamic and subgranular zone (SGZ) neurogenesis. They conclude that LiPR reduces body weight and improves metabolic parameters affected by HFD as well as it concomitantly stimulates neurogenesis in both niches the SGZ and the hypothalamus. The link between both effects is not demonstrated. The work is well conducted, the hypothesis is interesting and the experimental approach is adequate. The scope is wide and results are interesting, however a few aspects need to be further clarified. The manuscript is well written although the modification of some aspects would facilitate the reading such as the use of non described abbreviations for example.

    RESPONSE: We thank the reviewer for the positive assessment of our manuscript and for recognizing its novelty and importance for the research in neurogenesis, endocrinology, and metabolism. We will strive to clarify and facilitate our conclusions to improve the manuscript.

      • One concern in this study is the experimental groups. Authors analyze three groups control,HFD and HFD treated with LiPR. Authors conclude that the effects of LiPR are diet independent. However, given the results obtained by the authors on the effect of LiPR, the main question that arises in here is whether LiPR would have an effect on control mice. It seems tha a group is missing in the experimental design in which control ,mice are treated with LiPR during 7, 21 and the last two weeks of the 4 months. Author must include this information or at least argue the election of the experimental design.* RESPONSE: We thank the reviewer for this insight. We agree that including the Control Diet + LiPR in some of our analyses would improve the revised manuscript as also noted by Reviewer 2 (comment 1 and 2) and by Reviewer 2 (comment 1 and 2). In the original manuscript, we included the quantification of BrdU+ cells in the MBH for the Control Diet + LiPR in the 21-day group. To expand on these results, we will quantify the effects of LiPR on alpha tanycytes in the 21-day group. In addition, we will generate Control Diet + LiPR mice for the 4-month group to complement the HFD and HFD + LiPR data.
    • Body weight is found reduced by LiPR as well as other metabolic parameters in mice treated with LiPR during the last two weeks of the 4 Mo HFD. However, no effects on hypothalamic or SGZ neurogenesis are not observed in this experimental group. How do authors explain this results?*

    __RESPONSE: __The 4-month group contains animals that are over 6-month-old, which display very low levels of cell proliferation and differentiation in comparison with the 7 and 21-day groups that contain mice that are 2 and 2.5 months old, respectively. It is possible that these low levels of neurogenesis did not allow us to detect any pro-neurogenic effects of LiPR. Alternatively, the low neurogenesis in older animals precludes us from detecting the adverse effects of the HFD, which are rescued by LiPR in younger animals.

    • In figure 1 I-K images are not clear and better resolution images would help.*

    RESPONSE: We provided images with higher resolution for Figure 1I-K of the revised manuscript.

    • Authors conclude that LiPR is increasing the number of NSC by reducing their activation. However, authors show an induced increase in htNSC only in mice fed HFD for 7 days and not in the 21 day fed mice or the 4 mo fed mice (fig 2 d-f). In addition, authors test for the number of cells expressing Ki67 (fig 2 L), however, the number of Ki67+ alpha tanicytes is not shown.*

    RESPONSE: We thank the reviewer for this insight. In the revised manuscript (line 158), we corrected the inaccurate statement that LiPR increased the number of aNSCs and did not rescue their number, which was also noted by Reviewer 1 (comment 5) and by Reviewer 2 (comment 4). In addition, we will quantify the number of Ki67+ cells in the Hypothalmic Ventricular Zone (HVZ), which will address whether LiPR affects proliferation of aNSCs. This concern parallels comment 6 of Reviewer 2.

    • On figure 2B it seems that is alpha 2 tanicytes that are missing in response to HFD.*

    RESPONSE: Indeed, the panel in Figure 2B shows that the HFD reduces the number of alpha tanycytes, including the alpha 2 tanycytes. This representative image supports our quantification results in Figure 2D-E.

    • Are Fig 2 A-C images representative of mice fed HFD for 7 days?*

    __RESPONSE: __Yes, the representative images in panels of Fig. 2A-C are from the 7-day group. However, the legend states that these images are from the 21-day group. This is an error that we corrected in the revised manuscript in the legend of Figure 2 (line 572). We apologize for this and thank the reviewer for double-checking.

    • By looking at figure 2B it seems like the proportion of alpha tanicytes is higher in HFD since no or very few tanicytes are observed and almost all of them are alpha tanicytes.*

    RESPONSE: Indeed, 7 days of HFD reduced the number of alpha 2 tanycytes, which occupy the ventral-lateral aspect of the 3rd ventricle. This reduction of alpha 2 tanycytes drives the lover proportion of GFAP+ alpha-tanycytes out of all GFAP+ tanycytes. We emphasized this in the text of the revised manuscript (line 435-437).

    • In fig 2 d-f, an increase in the number of GFAP+ alpha tanicytes and its proportion as well as labelled with vimentin is observed in control mice fed with normal diet for 7 days compared with mice fed normal diet for 21 days. How do authors explain this difference?*

    RESPONSE: There is no difference in the number of GFAP+ alpha tanycytes or proportion of GFAP+ alpha tanycytes between 7-day and 21-day Control Diet mice. We used the two-way, repeated measure ANOVA with the Bonferroni’s pots-hoc test and did not observe any statistical difference between these 2 quantifications for the Control Diet mice at 7 and 21 days. There is a statistical difference between 7-day and 21-day Control Diet mice in the proportion of GFAP+Vimentin+ tanycytes. This could be due to expansion of the Vimentin+ tanycytes in relatively young adult mice. Given that this is not a major point, we prefer not expanding its discussion in the manuscript.

    • In fig 2 Why are the differences in RAX, KI67 and PCNA only present in mice fed HFD for 21 days?*

    RESPONSE: We thank the reviewer for this question, which reflects a similar comment 6 of Reviewer 2. To improve consistency of the presented data, we will quantify the proliferating cells also for the 7-day time point. In addition, we will quantify the number of proliferating cells in the HVZ, which will allow us to address whether HFD and/or LiPR alter proliferation of tanycytes.

    • Authors test for adult hippocampal neurogenesis in the three groups. DO images in fig S2 correspond to the 21 day treatment group?*

    RESPONSE: Yes, the representative images in the Supplementary Figure 2 are from the 21-day group. This is stated in the figure legend.

    • On fig S2 C, it seems that in HFD fed mice treated with LiPR newly generated neuroblasts are more differentiated have authors looked at DCX+ cell morphology?*

    RESPONSE: We thank the reviewer for this observation. We have not analyzed the morphology of DCX+ cells or DCX+ neuroblasts in the SGZ. As the manuscript focuses on the hypothalamic and not hippocampal neurogenesis, we prefer not to analyze the morphology in the revised manuscript.

    • In this same figure, it seems like the number of DCX+ neuroblasts and the number of newly generated neurons is reduced in mice of the 21 d group compared to the 7 day group. Is this statistically significant?*

    RESPONSE: We used the two-way, repeated measure ANOVA with the Bonferroni’s pots-hoc test to analyze the DCX+ neuroblasts and neurons. We observed a statistically very significant effect of LiPR treatment on the number of DCX+ neuroblasts and neurons (page 10 of the original manuscript). However, the Bonferroni’s test did not reveal any difference between 7-day and 21-day treatment groups.

    • There is a large reduction in the number of DCX+ cells from control 21 d treated mice to control 4 month treated mice. Is this statistically significat? How do authors explain this dramatic reduction?*

    RESPONSE: Yes, there is statistically significant reduction in the number of DCX+ cells and DCX+ neurons in the SGZ between the 21-day and 4-month group S.Fig.2). This reduction is most likely a result of aging. The mice of the 21-day group were around 2.5 months of age when culled, whereas the 4-month group month mice were over 6.5-month-old. The decline in SGZ neurogenesis with age is well documented. Because this decrease in DCX+ cells in the SGZ is an obvious consequence of the animals’ age and because the hippocampal neurogenesis is not the primary focus of this manuscript, we prefer not to discuss this feature in the manuscript.

    • Authors do not show the effect of HFD on BrdU+ neurons in the Arcuate. However, all data need to be shown.

    RESPONSE: We stated (on page 12 of the original manuscript) that in the Arcuate Nucleus of the 21-day group, there was “a statistically significant increase of BrdU+ neurons by HFD compared to Control (data not shown)”. To satisfy reviewer’s comment, we incorporated this data in the S.Fig.5 as the new panel S.Fig.5O and added the following text (lines 309-313) to the revised manuscript: “However, in the ArcN, the primary nutrient and hormone sensing neuronal nucleus of MBH 4, there was a statistically significant difference in number of BrdU+ neurons due to treatment (OWA, F(3,15) = 3.97, p = 0.0029). Exposure to 21d HFD significantly increased the number of BrdU+ neurons in the ArcN, which was reversed by co-administration of LiPR or Liraglutide (S.Fig.5O).” In addition, we adjusted the relevant discussion (lines 468-472): “Our results show that the short and intermediate exposure to HFD does not change the number of newly generated, BrdU+ cells, neurons, or astrocytes in the MBH parenchyma, however, it increases the number of BrdU+ neurons in the primary sensing ArcN, which is reversed by the con-current administration of LiPR or Liraglutide” and (lines 474-476): “In addition, our results show that while LiPR does not change the number of new cells in the MBH parenchyma, it can rescue the increased production of new neurons in the ArcN in the context of the intermediate HFD exposure.”

    *Reviewer #3 (Significance (Required)):

    In general the manuscript includes a great amount of work to demonstrate the effect of LiPR on neurogenesis (hippocampal and hypothalamic). The scope is wide, and the hypothesis is really interesting. Authors may need to solve some issues in order to completely demonstrate their claims and conclusions, but once the work is done, it will be very valuable to understand the effect of pharmacological agents used in the field of endocrinology to treat metabolic disorders such as type 2 diabetes di type 2 diabetes. So far, no studies have been done in which the effect of this molecules have been described on SGZ and hypothalamic neurogenesis. Both the field of endocrinology and metabolism as well as the field of adult neurogenesis may benefit of a study of this type.*

  3. Review Commons

    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

    The work of Jörgensen et al describes the effect of a lipidized analogue of the prolactin releasing peptide (LiPR) on the mouse metabolism in response to high fat diet (HFD) and on hypothalamic and subgranular zone (SGZ) neurogenesis. They conclude that LiPR reduces body weight and improves metabolic parameters affected by HFD as well as it concomitantly stimulates neurogenesis in both niches the SGZ and the hypothalamus. The link between both effects is not demonstrated. The work is well conducted, the hypothesis is interesting and the experimental approach is adequate. The scope is wide and results are interesting, however a few aspects need to be further clarified. The manuscript is well written although the modification of some aspects would facilitate the reading such as the use of non described abbreviations for example.

    Major comments:

    1. One concern in this study is the experimental groups. Authors analyze three groups control,HFD and HFD treated with LiPR. Authors conclude that the effects of LiPR are diet independent. However, given the results obtained by the authors on the effect of LiPR, the main question that arises in here is whether LiPR would have an effect on control mice. It seems tha a group is missing in the experimental design in which control ,mice are treated with LiPR during 7, 21 and the last two weeks of the 4 months. Author must include this information or at least argue the election of the experimental design.
    2. Body weight is found reduced by LiPR as well as other metabolic parameters in mice treated with LiPR during the last two weeks of the 4 Mo HFD. However, no effects on hypothalamic or SGZ neurogenesis are not observed in this experimental group. How do authors explain this results?
    3. In figure 1 I-K images are not clear and better resolution images would help.
    4. Authors conclude that LiPR is increasing the number of NSC by reducing their activation. However, authors show an induced increase in htNSC only in mice fed HFD for 7 days and not in the 21 day fed mice or the 4 mo fed mice (fig 2 d-f). In addition, authors test for the number of cells expressing Ki67 (fig 2 L), however, the number of Ki67+ alpha tanicytes is not shown.
    5. On figure 2B it seems that is alpha 2 tanicytes that are missing in response to HFD.
    6. Are Fig 2 A-C images representative of mice fed HFD for 7 days?
    7. By looking at figure 2B it seems like the proportion of alpha tanicytes is higher in HFD since no or very few tanicytes are observed and almost all of them are alpha tanicytes.
    8. In fig 2 d-f, an increase in the number of GFAP+ alpha tanicytes and its proportion as well as labelled with vimentin is observed in control mice fed with normal diet for 7 days compared with mice fed normal diet for 21 days. How do authors explain this difference?
    9. In fig 2 Why are the differences in RAX, KI67 and PCNA only present in mice fed HFD for 21 days?
    10. Authors test for adult hippocampal neurogenesis in the three groups. DO images in fig S2 correspond to the 21 day treatment group?
    11. On fig S2 C, it seems that in HFD fed mice treated with LiPR newly generated neuroblasts are more differentiated have authors looked at DCX+ cell morphology?
    12. In this same figure, it seems like the number of DCX+ neuroblasts and the number of newly generated neurons is reduced in mice of the 21 d group compared to the 7 day group. Is this statistically significant?
    13. There is a large reduction in the number of DCX+ cells from control 21 d treated mice to control 4 month treated mice. Is this statistically significat? How do authors explain this dramatic reduction?
    14. Authors do not show the effect of HFD on BrdU+ neurons in the Arcuate. However, all data need to be shown.

    Significance

    In general the manuscript includes a great amount of work to demonstrate the effect of LiPR on neurogenesis (hippocampal and hypothalamic). The scope is wide, and the hypothesis is really interesting. Authors may need to solve some issues in order to completely demonstrate their claims and conclusions, but once the work is done, it will be very valuable to understand the effect of pharmacological agents used in the field of endocrinology to treat metabolic disorders such as type 2 diabetes di type 2 diabetes.

    So far, no studies have been done in which the effect of this molecules have been described on SGZ and hypothalamic neurogenesis. Both the field of endocrinology and metabolism as well as the field of adult neurogenesis may benefit of a study of this type.

  4. Review Commons

    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

    The authors examine the effect of an anorexigenic drug, LiPR in the context of treatment with high fat diet (HFD) and with a special focus on hypothalamic neural stem/progenitor cells and neurogenesis. The work is mostly based on mice and a barrage of different techniques (confocal imaging, cell cultures with time lapse, gene expression...) are used.

    The results are interesting because they address the yet-poorly understood implication of hypothalamic neurogenesis in food intake and energy balance. The results point at complex effects at different levels (neural stem cells, neurons, division, survival...). The experimental approach is sometimes thorough in the treatment of details on the one hand, it also lacks of consistency on the other, and as a result the conclusions lack strength. There is a number of experiments that sometimes seem unrelated and this hurts the comprehension of the manuscript, specially in lieu of the complexity of the results obtained.

    These are the detailed comments:

    A major issue is the lack of a LiPR-only group, which would much facilitate the interpretation of the results. The effect of LiPR alone is however tested, but only in comparison with the Control in one of the in vitro experiments (S.Fig. 3)

    As plotted, in Fig 1B is difficult to interpret the effect of HFD and LiPR, might be using percentage and noting the statistical differences as in the other would help. It looks like HFD has no effect compared to control on weight and only at the end LiPR could have an effect. On the other hand, after 4 months, HFD mice are clearly above the controls and it is then, albeit when weight gain has reached a plateau, that LiPR has an effect. The election of these arbitrary paradigms and their drawbacks has to be better explained.

    Why was the proportion of GPR10+BrdU+MAP2+ cells only assessed in control mice and no in the experimental groups if its expression in overall neurons changes?

    This suggests that the receptor is expressed in neurons. Interestingly, exposure to 21d HFD reduced density of GPR10, which was rescued by LiPR administration (Fig.1L). Why was this time point chosen and not the longer-term one? What is the consequence of the alterations in the potential number of GPR10, specially in relation to the administration of LiPR?

    This clarification is important because a 14-day treatment was chosen for the in vitro experiments in which LiPR, but not HFD, seems to have an effect on cell proliferation. Might be it would have been more useful to use a paradigm in which HFD has an effect to better compare with in vivo work and for the rationale of the work.

    "Besides GPR10, we co-localized neuronal cytoskeleton structures with NPFFR2 in the MBH (Fig.1O-P)..." Why were not GPR10 and NPFF2 analyzed in a similar and consistent manner ? It is confusing.

    The number of GFAP+ α-tanycytes is not significantly changed by HFD therefore LiPR does not rescue, but rather increases the number of GFAP+ α-tanycytes in the 7-day setting. There are no differences among groups later, the effect is lost by 21 days, therefore there is a transient excess of GFAP+ α-tanycytes which later "disappear" in the LiPR group. The authors state that LiPR rescues the decrease in "htNSCs", but after 21 days the number of the GFAP+ α-tanycytes is the same in all groups without the need of LiPR. There is no experimental follow up (addressing proliferation and survival of these cells) and the conclusions stated in the text (results and discussion) are not really supported by the data. The in vitro experiments could be a complement, but are no substitute for the missing in vivo exploration.

    The fact that cell division is "rarely found" (Rax GFAP) experiments also push for further investigation.

    It is difficult to see that relevance of the inclusion of the vimentin staining experiment if there is no further exploration. The effect of LiPR is only transient, in the 7-day paradigm and as the parameter evaluated is the proportion of vimentin+ tanycytes among GFAP+ tanycytes it could only be reflecting increased expression of the filament. "Nevertheless, we did not observe a statistically different change in the area occupied by Rax+ tanycytes (Fig.2H)."

    Why did the authors use Rax only for this experiment if "GFAP+ α-tanycytes which are considered the putative htNSCs?" What is the justification for not seeing changes in relation to the results reported in Fig 2D-F?

    "Because Vimentin is associated with nutrient transport in cells and with metabolic response to HFD 52-54, we quantified the proportion of GFAP+ tanycytes expressing Vimentin (Fig.2F)." It is difficult to see that relevance of the inclusion of the vimentin staining experiment if there is no further exploration. The effect of LiPR is only transient, in the 7-day paradigm and as the parameter evaluated is the proportion of vimentin+ tanycytes among GFAP+ tanicytes it could only be reflecting increased expression of the filament.

    Why there is no Ki67 experiment in the 7-day paradigm if that is the timepoint in which changes in the number or proportion of GFAP+ tanycytes are observed? PCNA was then used but only in the 21-day paradigm. What is the interpretation and relevance of these data? What are the non-htNSCs proliferating cells, whose dynamics are different from the changes in the number or proportion of htNSCs that could be potentially related to changes in mitosis? Again, I think it would be much useful for the work to explore in detail the changes in the putative htNSCs than investing in experiments that only add confusion.

    The inclusion of Liraglutide + HFD, (not Liraglutide alone) only in some of the experiments is pointless if there is no direct comparison with LiPR and a timepoint is missing. In S.Fig 3, Fig. 5 and S.Fig 7 LFD (low fat diet?) is used in several occasions as in: "on reducing number of PCNA+ cells in 21d protocol (one-way ANOVA (OWA), F(2,12) = 16.66, p = 0.0003) when compared to both LFD and HFD groups". Is this the control diet?

    The final experiment shows that application of hPrRP31, a variation of LiPR, causes an immediate calcium increase in human induced pluripotent stem cell-derived hypothalamic nucleus. This finding is interesting in itself because it brings light about the function of the receptor/s. It would have been very useful to test what other receptors mentioned to bind LiPR is mediating the effect. In any case, the focus of the work are the neural stem/progenitor cells responsible for neurogenesis and the changes in their properties because of HFD and LiPR, therefore I would trade these experiments for a more thorough and detailed dissection of these effects.

    Minor points:

    Introduce "GLP-1RA"

    "HFD-induced inflammation and astrogliosis in the hypothalamus 45,46, whereas the long (4mo) protocol leads to DIO" Are these notions exclusive?

    "LiPR displays no effects on astrocytes" "Displays" is not the correct term.

    Referee cross-commenting

    I think we all referees agree for the most part. The main concern stated by all of us is the lack of a LiPR-alone group. The rest of the concerns are also related or complementary. In my opinion the mostly common view by the referees is reasuring.

    Significance

    The strengths of the work are its novelty in the field and the variety of techniques employed. The work has the potential of unveiling mechanistic insight into the regulation of neural stem/progenitor cells and neurogenesis. The main audience of this work would be the community working on this field. The lack of experiments testing that the changes observed actually participate in food intake prevent the work from being of relevance for a broader audience (food intake, energy balance, obesity...).

    The limitations are the descriptive nature of the work and the lack of a consistent and systematic experimental design that would allow to extract solid conclusions upon to which build upon future research.

  5. Review Commons

    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

    In this manuscript, Jorgensen and colleagues describe their findings on the action of a palmitoylated form of prolactin-release peptide (LiPR) on neural stem cells (NSC) in the adult mouse hypothalamus and adult mouse hippocampus. Their main conclusion is that LiPR can counteract the effects of high-fat diet (HFD) and rescue some of the adverse effects of HFD. Specifically, the authors provide evidence that:

    • Exposure to HFD reduces the number of presumptive adult neural stem cells (NSCs) in the adult hypothalamus, whereas exposure to LiPR reverses this trend.
    • The results suggest that LiPR reduces the proliferation of alpha-tanycytes and/or their progeny in the hypothalamus in the context of HFD, with Liraglutide acting similarly. In contrast, while LiPR also suppresses proliferation in the SGZ, Liraglutide works there in the opposite direction.
    • LiPR also helps the survival of adult-born hypothalamic neurons.
    • Reduction of proliferation by LiPR suggests a model where LiPR increases the number of NSCs presumably by reducing their rate of activation.
    • The results suggest that LiPR promotes expression of PrRP receptors in the hypothalamic neurons, suggesting that PrRP may act directly on such neurons (and tanycytes?) in vivo.
    • The authors also show that HFD and LiPR alter gene expression profiles of the MBH cells, with HFD, but not LiPR, inducing myelination-related genes.
    • Finally, they show that PrRP stimulates an increase in Ca2+ in in vitro-derived human hypothalamic neurons.
    • The authors conclude that LiPR may be reducing activation and proliferation of the hypothalamic stem cells and thereby preserve their pool from exhaustion, which was stimulated by HFD. The manuscript presents interesting data and is clearly written. There are several comments, mainly editorial.
    1. It is unclear why most of the experiments do not include the control+LiPR group. Even though the focus of the study was the action of LiPR in the context of HFD, questions remain regarding the action of LiPR per se. Is LiPR (or Liraglutide, for that matter) completely inactive on the normal diet background, with respect to neurogenesis in the hypothalamus and the hippocampus? Whether the answer is positive or negative, it would give a much better understanding of the action of LiPR - does it regulate neurogenesis in various physiological contexts, or does it only kick in with a particular type of diet? In fact, this was examined (see Supplementary figures), but only for the cells in culture and, when performed with animals, was limited to 7 and 21 days, rather than 4 months, which would have been much more informative.
    2. The question above is also relevant when considering the conclusions on the potential depletion of the stem cell pool (again, whether in the hypothalamus or the hippocampus), particularly at the 4-months time point. The mice are ~6 months old by that time, and neurogenesis in both regions is expected to decrease by that time. Are LiPR or Liraglutide able to suppress or exacerbate this decrease? Can they be used to mitigate this decrease when mice are on a regular diet?
    3. A somewhat related issue is that, in most cases, only the percentage or the density of cells are shown on the graphs, rather than the absolute numbers (at least for some cases). This sometimes complicates the comparisons; for instance, does the surface of the hypothalamus change between 2 and 6 months of age? The tanycytes' number stays, apparently, the same (e.g., Fig. 2) but the production of new neurons is supposed to fall dramatically.
    4. The authors write "LiPR may prevent stem cells from exhaustion, induced by HFD" - but it is not clear that HFD indeed leads to exhaustion - there is no statistically significant difference in the number of the stem cells (alpha-tanycytes) between the control and HFD or between HFD at 1, 3, or 12 weeks.
    5. Numerous papers show that the rate of production of new adult hypothalamic neurons (mainly those derived from beta-tanycytes) drops drastically within the first several weeks of mouse life. Does HFD accelerate, and LiPR mitigate, this decrease? Perhaps one can calculate the numbers from the graphs, but it would help if this is explained in the text of the manuscript. Also, it is not always clear whether specific experiments were performed with the zones of the hypothalamic wall that only contain alpha-tanycytes.
    6. A sharp increase in PCNA+ cells in the hippocampus at the 21-day time point, both in the control and in the HFD and HFD/LiPR groups (Fig. S2f) is a little puzzling because neither the Dcx+ nor the Ki67+ cells show this increase.
    7. The study deals with several agents and several processes; a simple scheme that summarizes authors' conclusions might help to better understand the relationships between those agents and processes.

    Referee cross-commenting

    I agree, the lack of the LiPR group complicates the interpretation of the results. I also agree that the experiments with vimentin staining, calcium increase, and even with neurospheres do not add much to the main questions that this study attempts to answer, and I'd rather see a more thorough analysis of the activation and differentiation data. I also want to reiterate that the concept of LiPR/PrRP preventing the exhaustion of the hypothalamic stem cell pool is not clear, because it is not shown that this pool does actually get exhausted under normal or HFD conditions. This latter issue again requires the LiPR-alone group. Also, as a clarification - I wrote about 1 month required to compete the revision assuming that the authors actually have the data on the Control+LipR group or at least the specimens available, mainly because the supplementary material shows results with this group, at least with the neurospheres. If this group is fully missing, then the effort will obviously take a longer time.

    Significance

    The provided evidence suggests, for the first time, that PrRP prevents the loss of the neural stem cells population in the adult hypothalamus that was diminished by obesity and HFD. This finding might be interesting to a broad audience.

  6. Version published to 10.1101/2023.02.28.528936v1 on bioRxiv