Author response:
Public Reviews:
Reviewer #1 (Public review):
The authors try to investigate how the population of microtubules (LSPMB) that originate from sporozoite subpellicular microtubules (SSPM) and are remodelled during liver-stage development of malaria parasites. These bundles shrink over time and help form structures needed for cell division. The authors have used expansion microscopy, live-cell imaging, genetically engineered mutants, and pharmacological perturbation to study parasite development with liver cells.
A major strength of the manuscript is the live cell imaging and expansion microscopy to study this challenging liver stage of parasite development. It gives important knowledge that PTMs of α-tubulin, such as polyglutamylation and tyrosination/detyrosination, are crucial for microtubule stability. Mutations in α-tubulin reduce the parasite's ability to move and proliferate in the liver cells. The drug oryzalin, which targets microtubules, also blocks parasite development, showing how important dynamic microtubules are at this stage.
The major problem in the manuscript was the way it flows, as the authors keep shifting from the liver stage to the sporogony stages and then back to the liver stages. It was very confusing at times to know what the real focus of the study is, whether sporozoite development or liver stage development. The flow of the manuscript could be improved. Some of the findings reported here substantiate the previous electron microscopy.
Overall, the study represents an important contribution towards understanding cytoskeletal remodelling during liver stage infection. The study suggests that tubulin modifications are key for the parasite's survival in the liver and could be targets for new malaria treatments. This is also the stage that has been used for vaccine development, so any knowledge of how parasites proliferate in the liver cells will be beneficial towards intervention approaches.
We would like to express our sincere gratitude to Reviewer #1 for the positive and encouraging feedback on our manuscript. We are delighted that the reviewer found our experimental design and methodologies appropriate and that our study represents an important contribution to understanding cytoskeletal remodelling during liver stage infection, a critical phase for vaccine development. We are also grateful to the reviewer for highlighting the issue with the manuscript's flow. We acknowledge this limitation and will significantly improve the narrative structure and logical progression in the revised manuscript to ensure clarity and avoid any potential confusion. Thank you again for your thoughtful and constructive comments.
Reviewer #2 (Public review):
Summary:
The authors investigated microtubule distribution and their possible post-translational modifications (PTM) in Plasmodium berghei during development of the liver stage, using either hepatocytes or HeLa cells as models. They used conventional immunofluorescence assays and expansion microscopy with various antibodies recognising tubulin and, in the second part of the work, its candidate PTMs, as well as markers of Plasmodium, in addition to live imaging with a fluorescent marker for tubulin. In the third part of the study, they generated 3 mutants deprived of either the last four residues or the last 11 residues, or where a candidate polyglutamylation site was substituted by an alanine residue.
Strengths:
In the first part, microtubules are monitored by a combination of two approaches (IFA and live), revealing nicely the evolution of the sporozoite subpellicular microtubules (SSPM, the sporozoite is the developmental stage present in salivary glands of the mosquitoes and that infects hepatocytes) into a different structure termed liver-stage parasite microtubule bundle (LSPMB). The LSPMB shrinks during the course of parasite development and finally disappears while hemi-spindles emerge over time. Contact points between these two structures are observed frequently in live cells and occasionally in fixed cells, suggesting the intriguing possibility that tubulin might be recycled from the LSPMB to contribute to hemi-spindle formation.
In the second part, antibodies recognising (1) the final tyrosine found at the C-terminal tail and (2) a stretch of 3 glutamate residues in a side chain are used to monitor these candidate PTMs. Signals are positive at the SSPM, and while it remains positive for polyglutamylation, it becomes negative for the final tyrosine at the LSPM, while a positive signal emerges at hemi-spindles at later stages of development.
In the last part, the three mutants are fed to mosquitoes, where they show reduced development, the one lacking the alpha-tubulin tail even failing to reach the salivary glands. However, the two other mutants infect HeLa cells normally, whereas sporozoites with the C-terminal tail deletion recovered from the haemolymph did not develop in these cells.
The first part provides convincing evidence that microtubules are extensively remodelled during the infection of hepatocytes and HeLa cells, in agreement with the spectacular Plasmodium morphogenetic changes accompanying massive and rapid proliferation. The third part brings further confirmation that the C-terminal tail of alpha-tubulin is essential for multiple stages of parasite development, in agreement with previous work (50). Since it is the region where several post-translational modifications take place in other organisms (detyrosination, polyglutamylation, glycylation), it makes sense to propose that the essential function is related to these PTMs also in Plasmodium.
Weaknesses:
The significance of tubulin PTM relies on two antibodies whose reactivity to Plasmodium tubulins is unclear (see below). The interpretation of the literature on detyrosination and polyglutamylation is confusing in several places, meaning that the statements about the possible role of these PTMs need to be carefully revisited.
The authors use the term "tyrosination" but the alpha1-tubulin studied here possesses the final tyrosine when it is synthesised, so it is "tyrosinated" by default. It could potentially be removed by a tyrosine carboxypeptidase of the vasoinhibin family (VASH) as reported in other species. After removal, this tyrosine can be added again by a tubulin-tyrosine ligase (TTL) enzyme. It is therefore more appropriate to talk about detyrosination-retyrosination rather than tyrosination (this confusion is unfortunately common in the literature, see Janke & Magiera, 2020).
The difficulty here is that there is so far no evidence that detyrosination takes place in Plasmodium. Neither VASH nor TTL could be identified in the Plasmodium genome (ref 31, something we can confirm with our unsuccessful BLAST analyses), and mass spectrometry studies of purified tubulin, albeit from blood stages, did not find evidence for detyrosination (reference 43). Western blots using an antibody against detyrosinated tubulin did not produce a positive signal, neither on purified tubulin, nor on whole parasites (43). Of course, the situation could be different in liver stages, but the question of the detyrosinating enzyme is still there. The existence of a unique Plasmodium system for detyrosination cannot be formally ruled out but given the high degree of conservation of these PTMs and their associated enzymes, it sounds difficult to imagine.
The fact that the anti-tyrosinated antibody still produced a signal in the cell line where the final tyrosine is deleted raises issues about its specificity. A cross-reactivity with beta-tubulin is proposed, but the Plasmodium beta-tubulin does not carry a final tyrosine, further raising concerns about antibody specificity.
The interpretation of these results should therefore be considered carefully. There also seems to be some confusion in the function of detyrosination cited from the literature. It is said in line 229 that "tyrosination has been associated with stable microtubules" (33, 34, 50, 55). References 33 and 34 actually show that tyrosinated microtubules turn over faster in neurons or in epithelial cells, respectively, while references 50 and 55 do not study de/retyrosination. The general consensus is that tyrosinated microtubules are more dynamic (see reference 24).
The situation is a bit different for polyglutamylation since several candidate poly- or mono-glutamylases have been identified in the Plasmodium genome, and at least mono-glutamylation of beta-tubulin has been formally proven, still in bloodstream stages (ref 43). The authors propose that the residue E445 is the polyglutamylation site. To our knowledge, this has not been demonstrated for Plasmodium. This residue is indeed the favourite one in several organisms such as humans and trypanosomes (Eddé et al., Science 1990; Schneider et al., JCS, 1997), and it is tempting to propose it would be the same here. However, TTLLs bind the tubulin tails from their C-terminal end like a glove on a finger (Garnham et al., Cell, 2015), and the presence of two extra residues in Plasmodium tubulins would mean that the reactive glutamate might be in position E447 rather than E445. This is worth discussing.
On the positive side, it is encouraging to see that signals for both anti-tyrosinated tail and poly-glutamylated side chain are going down in the various mutants, but this would need validation with a comparison for alpha-tubulin signal.
Line 316: polyglutamylation "is commonly associated with dynamic microtubule behavior (78-80)". Actually, references 78 and 79 show the impact of this PTM on interaction with spastin, and reference 80 discusses polyglutamylation as a marker of stable microtubules in the context of cilia and flagella. The consensus is that polyglutamylated microtubules tend to be more stable (ref24).
Conclusion:
The first and the third parts of this manuscript - evolution of microtubules and importance of the C-terminal tails for Plasmodium development - are convincing and well supported by data. However, the presence and role of tubulin PTM should be carefully reconsidered.
Plasmodium tubulins are more closely related to plant tubulins and are sensitive to inhibitors that do not affect mammalian microtubules. They therefore represent promising drug targets as several well-characterised compounds used as herbicides are available. The work produced here further defines the evolution of the microtubule network in sporozoites and liver stages, which are the initial and essential first steps of the infection. Moreover, Plasmodium has multiple specificities that make it a fascinating organism to study both for cell biology and evolution. The data reported here are elegant and will attract the attention of the community working on parasites but also on the cytoskeleton at large. It will be interesting to have the feedback of other people working on tubulin PTMs to figure out the significance of this part of the work.
We thank Reviewer #2 for the thoughtful and detailed evaluation of our manuscript. We are pleased that the reviewer found our study elegant and believe it will attract the attention of the broader scientific community, both those working on parasites and those focused on cytoskeleton biology. We also acknowledge the concerns raised regarding the specificity of the antibodies used to detect tubulin post-translational modifications (PTMs), as well as the interpretation of their signals and the current lack of identified detyrosination enzymes in the Plasmodium genome. We agree that these are important limitations, and we will address them thoroughly in the revised manuscript. This includes clarifying our interpretation of tyrosination versus detyrosination, adjusting our claims regarding polyglutamylation sites, and carefully revisiting the literature cited to ensure accurate contextualization of PTM function in microtubule stability.
We are grateful for the reviewer’s close reading and critical feedback, which will help us substantially improve the clarity, precision, and strength of our manuscript.
Reviewer #3 (Public review):
Summary:
The manuscript by Atchou et al. investigates the role of the microtubule cytoskeleton in sporozoites of Plasmodium berghei, including possible functions of microtubule post-translational modifications (tyrosination and polyglutamylation) in the development of sporozoites in the liver. They also assessed the development of sporozoites in the mosquito. Using cell culture models and in vivo infections with parasites that contain tubulin mutants deficient in certain PTMs, they show that may aspects of the life cycle progression are impaired. The main conclusion is that microtubule PTMs play a major role in the differentiation processes of the parasites.
However, there are a number of major and minor points of criticism that relate to the interpretation of some of the data.
We thank Reviewer #3 for the overall positive assessment of our study and for recognizing its contribution to advancing our understanding of Plasmodium biology and malaria pathogenesis. We appreciate the reviewer’s constructive feedback, particularly regarding the interpretation of some of our data. These comments have been very helpful in guiding our revisions, and we have worked to improve both the clarity of our presentation and the precision of our interpretations in the revised manuscript.
Below, we respond in detail to each of the reviewer’s points.
Comments:
(1) The first paragraph of "Results" almost suggests that the presence of a subpellicular MT-array in sporozoites is a new discovery. This is not the case, see e.g. the recent publication by Ferreira et al. (Nature Communications, 2023).
We thank the reviewer for pointing this out and fully agree that the subpellicular microtubule (SPM) array in sporozoites is well established, as documented in earlier work (e.g., Cyrklaff et al., 2007) and more recently by Ferreira et al. (Nat. Commun., 2023). Our intention was not to suggest that the existence of the SSPM is a novel finding. Rather, our study builds on this existing knowledge by demonstrating that these sporozoite-derived microtubules are not disassembled upon hepatocyte entry but are repurposed into a newly described structure, the liver stage parasite microtubule bundle (LSPMB). This reorganization, its persistence into liver stage development, and its dynamic role in microtubule remodeling and nuclear division are, to our knowledge, novel observations. We will revise the manuscript to make this distinction clearer in the introduction and the results section.
(2) Why were HeLa cells and not hepatocytes (as in Figure 3) used for measuring infection rates of the mutants in Figure 5H and 5L? As I understand, HeLa cells are not natural host cells for invading sporozoites. HeLa cells are epithelial cells derived from a cervical tumour. I am not an expert in Plasmodium biology, but is a HeLa infection an accepted surrogate model for liver stage development?
We appreciate the opportunity to clarify our experimental model. While HeLa cells are not the natural host cells, they are a well-established and validated in vitro model for studying Plasmodium berghei liver stage development in our lab and others. In this system, the parasite completes its full development and generates infectious merozoites. Numerous studies have successfully used HeLa cells as a liver stage infection model, with key findings subsequently validated in primary hepatocytes or in vivo, confirming its utility as a representative model. We employed this cell line primarily to reduce animal usage in accordance with the 3Rs principles (Replacement, Reduction, Refinement). Importantly, to ensure the biological relevance of our discoveries in HeLa cells, we validated our key findings in primary mouse hepatocytes, as shown in Figure 3. Furthermore, we confirmed the in vivo infectivity of mutant parasite lines that produced typical salivary gland sporozoites through an in vivo infection assay, presented in Figure S4C.
(3) The tubulin staining in Figures 1A and 1B is confusing and doesn't seem to make sense. Whereas in 1A the antibody nicely stains host and parasite tubulin, in 1B, only parasite tubulin is visible. If the same antibody and the same host cells have been used, HeLa cytoplasmic microtubules should be visible in 1B. In fact, they should be the predominant antigen. The same applies to Figure 2, where host microtubules are also not visible.
We thank the reviewer for this careful observation regarding the α-tubulin staining in Figures 1A and 1B. The same host cell type (HeLa) and α-tubulin antibody were indeed used in both experiments. Figure 1A shows results from conventional immunofluorescence assays, where both host and parasite microtubules are clearly stained. In contrast, Figure 1B shows the outcome of ultrastructure expansion microscopy (U-ExM), where parasite microtubules appear prominently, while host microtubules are less visible.
This effect appears to be a technical outcome of the U-ExM protocol, which can differentially preserve or reveal microtubule epitopes. We consistently observed stronger parasite signal across various cell types, including primary hepatocytes (Figure 3A,B). The lack of visible host microtubules in some U-ExM images does not reflect their absence, but rather reduced signal intensity relative to the parasite structures. This is not observed with all antibodies, e.g., host microtubules stain strongly with anti-tyrosinated α-tubulin (Figure 3B), likely reflecting their high tyrosination state.
To overcome this limitation, we employed PS-ExM and combined PS-ExM/U-ExM approaches (as described in reference 56), which allowed simultaneous high-resolution visualization of both host and parasite microtubule networks. These combined methods are now being used in follow-up studies to investigate host–parasite microtubule interactions in more detail.
We will clarify this point in the revised manuscript to avoid confusion.
(4) In Figures 2A and B, the host nuclei appear to have very different sizes in the DMSO controls and in the drug-treated cells. For example, in the 20 µM (-) image (bottom right), the nuclei are much larger than in the DMSO (-) control (top left). If this is the case, expansion microscopy hasn't worked reproducibly, and therefore, quantification of fluorescence is problematic. The scalebar is the same for all panels.
The expansion microscopy methods used in this study have been rigorously validated for both reproducibility and isotropicity. However, as the reviewer rightly notes, host cell nuclei can vary in size due to several factors, including cell cycle stage, infection status, and the extent of parasite development, all of which can influence host nuclei morphology and size.
Importantly, the quantifications relevant to our conclusions were focused specifically on parasite structures. We did not rely on host nuclear size or host fluorescence intensity as a quantitative readout in this context. While we acknowledge the observed variability in host nuclear dimensions, it does not compromise the accuracy or reproducibility of the parasite specific measurements central to our study.
We will clarify this point in the revised figure legend and manuscript.
(5) I don't quite follow the argument that spindles and the LSPMB are dynamic structures (e.g., lines 145, 174). That is a trivial statement for the spindle, as it is always dynamic, but beyond that, it has only been shown that the structure is sensitive to oryzalin. That says little about any "natural" dynamic behaviour. Any microtubule structure can be destroyed by a particular physical or chemical treatment, but that doesn't mean all structures are dynamic. It also depends on the definition of "dynamic" in a particular context, for example, the time scale of dynamic behaviour (changes within seconds, minutes, or hours).
We agree that sensitivity to chemical depolymerization alone does not necessarily indicate dynamic behavior, particularly in the absence of data on turnover kinetics or temporal changes.
Our interpretation was based on two observations: first, that the LSPMB, which derives from the highly stable sporozoite subpellicular microtubules (known to be drug-resistant), becomes susceptible to depolymerization during the liver stage; and second, that the LSPMB gradually shrinks over time during parasite development. These features suggested a transition toward a more dynamic state compared to its origin. However, we fully agree that “dynamic” is a context-dependent term and that direct evidence such as turnover rates or structural changes on short time scales, is required to rigorously define microtubule dynamics.
We will revise the manuscript to clarify our use of this term and explicitly acknowledge the need for further studies to characterize the timescale and mechanisms underlying LSPMB remodeling.
(6) I am not sure what part in the story EB1 plays. The data are only shown in the Supplements and don't seem to be of particular relevance. EB1 is a ubiquitous protein associated with microtubule plus ends. The statement (line 192) that it "may play a broader role..." is unsubstantiated and cannot be based merely on the observation that it is expressed in a particular life cycle stage.
We agree that EB1 is a ubiquitous microtubule plus-end binding protein and that its presence alone does not imply a novel function. Previous studies (e.g., Maurer et al., 2023; Yang et al., 2023; Zeeshan et al., 2023) have focused on its role during Plasmodium sexual stages, while its expression during liver and mosquito stages has not been previously documented.
Our data extend this knowledge by showing that EB1 is also expressed during liver stage development, particularly during the highly mitotic schizont phase. While we agree that this observation alone does not prove functional involvement, it raises the possibility of a broader role for EB1 in regulating microtubule dynamics beyond sexual stages. To avoid overinterpretation, we have presented these findings in the supplementary material and will revise the manuscript to tone down speculative statements and clearly frame this as a preliminary observation that warrants further investigation.
(7) Line 196 onwards: The antibody IN105 is better known in the field as polyE. Maybe that should be added in Materials and Methods. Also, the antibody T9028 against tyrosinated tubulin is poorly validated in the literature and rarely used. Usually, researchers in this field use the monoclonal antibody YL1/2. I am not sure why this unusual antibody was chosen in this study. In fact, has its specificity against tyrosinated α-tubulin from Plasmodium berghei ever been shown? The original antigen was human and had the sequence EGEEY. The Plasmodium sequence is YEADY and hence very different. It is stated that the LSPMB is both polyglutamylated and tyrosinated. This is unusual because polyglutamylated microtubules are usually indicative of stable microtubules, whereas tyrosinated microtubules are found on freshly polymerised and dynamic microtubules. However, a co-localisation within the same cell has not been attempted. This is, however, possible since polyE is a rabbit antibody and T9028 is a mouse antibody. I suspect that differences or gradients along the LSPMB would have been noticed. Also, in lines 207/208, it is said that tyrosination disappears after hepatocyte invasion, which is shown in Figure 3. However, in Figure 3A, quite a lot of positive signals for tyrosination are visible in the 54 and 56 hpi panels.
First, we acknowledge that the IN105 antibody is more widely known as "polyE" in the field. We will update the Materials and Methods section accordingly to reflect this nomenclature.
Regarding the use of the T9028 antibody against tyrosinated α-tubulin: we agree that this monoclonal antibody is less commonly used than YL1/2, and we appreciate the reviewer drawing attention to this. The original antigen for T9028 is based on the mammalian C-terminal sequence EGEEY, which differs from the Plasmodium α1-tubulin sequence (YEADY). Like many in the field, we face the challenge that most available antibodies are raised against mammalian epitopes, and specificity in Plasmodium can vary. Nonetheless, the literature (e.g., Hirst et al., 2022; Fennell et al., 2008) has demonstrated that tyrosination occurs in Plasmodium α1-tubulin, using anti-tyrosination antibodies including YL1/2.
Following the reviewer’s excellent suggestion, we are currently repeating the key experiments using the YL1/2 antibody to compare staining patterns directly with those obtained using T9028. We will include these results in the revised manuscript.
Concerning the potential co-localization of polyglutamylation and tyrosination on the LSPMB: we agree that this is an interesting and testable hypothesis. In the current manuscript, Figures 3A and 3B were generated from independent experiments, and thus co-localization was not assessed. However, as the reviewer correctly notes, polyE and T9028 antibodies are raised in rabbit and mouse, respectively, making co-staining feasible. We will follow up on this experimentally and, if feasible within our revision timeline, include data in the revised version or highlight this as a future direction.
Finally, with regard to Figure 3 and the observation that tyrosination appears to persist at 54 and 56 hpi (Figure 3B): the reviewer is correct that tyrosination signal is still detectable at these time points. Our statement that tyrosination “disappears after hepatocyte invasion” was intended to refer to an overall decrease in signal intensity during early liver stage development, with a reappearance at later stages (e.g., cytomere formation). We will rephrase this section for greater clarity and ensure that figure annotations and legends unambiguously reflect the dynamics observed.
(8) In line 229, it is stated that tyrosination "has previously been associated with stable microtubule in motility". This statement is not correct. In fact, none of the cited references that apparently support this statement show that this is the case. On the contrary, stable microtubules, such as flagellar axonemes, are almost completely detyrosinated. Therefore, tyrosination is a marker for dynamic microtubules, whereas detyrosinated microtubules are indicative of stable microtubules. This is an established fact, and it is odd that the authors claim the opposite.
We fully agree that in canonical eukaryotic systems, tyrosinated microtubules are generally markers of dynamic microtubule populations, whereas detyrosinated microtubules are typically associated with stability particularly in structures such as flagellar axonemes.
Our original statement will be corrected. In our study, we observed that tyrosinated microtubules are prevalent in invasive stages (sporozoites and merozoites), while detyrosinated forms become more prominent during intracellular liver stage development. This pattern is consistent with the established link between tyrosination and dynamic microtubules.
What is particularly intriguing in Plasmodium is the apparent cycling of tyrosination despite the absence of known tubulin tyrosine ligase (TTL) homologs in the genome. This suggests either a highly divergent enzyme or the involvement of host cell factors, a hypothesis supported by the reappearance of tyrosinated microtubules during liver stage schizogony (Figure 3B).
We will revise the relevant text and the Discussion section to reflect these mechanistic considerations more accurately and to avoid misrepresenting established principles of microtubule biology.
(9) Line 236 onwards: Concerning the generation of tubulin mutants, I think it is necessary to demonstrate successful replacement of the wild-type allele by the mutant allele. I am sure the authors have done this by amplification and subsequent sequencing of the genomic locus using PCR primers outside the plasmid sequences. I suggest including this information, e.g., by displaying the chromatograph trace in a supplementary figure. Or are the sequences displayed in Figure S3B already derived from sequenced genomic DNA? This is not described in the Legend or in Materials and Methods. The left PCR products obtained for Figure S3 B would be a suitable template for sequencing.
Indeed, these data are presented in Figure 4B and the corresponding sequence data are shown in Figure S3B. We appreciate the reviewer’s suggestion, which will help improve the transparency and reproducibility of our methodology.
(10) It is also important to be aware of the fact that glutamylation also occurs on β-tubulin. This signal will also be detected by polyE (IN105). Therefore, it is surprising that IN105 immunofluorescence is negative on the C-term Δ cells (Figure S3 D). Is there anything known about confirmed polyglutamylation sites on both α- and β-tubulins in Plasmodium, e.g., by MS? In Toxoplasma, both α- and β-tubulin have been shown to be polyglutamylated.
Indeed, polyglutamylation is known to occur not only on α-tubulin but also on β-tubulin in many organisms, including Toxoplasma gondii, and the polyE (IN105) antibody is expected to detect polyglutamylation on both tubulin isoforms.
The parasites shown in Figure S3D correspond to mutant lines originally generated by Spreng et al. (2019): the IntronΔ mutant (with deletion of introns in the Plasmodium α1-tubulin gene) and the C-termΔ mutant (with deletion of the final three C-terminal residues: ADY). As the reviewer correctly notes, this particular C-terminal deletion does not include the predicted polyglutamylation site (E445 or E447, depending on alignment), and thus should not abolish all polyglutamylation. However, in our experiments, the IN105 signal is substantially reduced in this mutant. This may suggest that structural alterations in the tubulin tail affect accessibility of the polyglutamylation epitope or influence the modification itself though we cannot exclude other possibilities, including changes in antibody recognition.
To date, polyglutamylation sites in Plasmodium tubulins have not been definitively confirmed by mass spectrometry. However, a recent MS-based study (reference 43) detected monoglutamylation on β-tubulin in blood stage parasites. Direct MS evidence for polyglutamylation of either α- or β-tubulin in Plasmodium liver stages is still lacking. We will clarify these points in the revised manuscript to avoid potential confusion and to highlight the need for future biochemical validation of PTM sites.
(11) Figure S3 is very confusing. In the legend, certain intron deletions are mentioned. How does this relate to posttranslational tubulin modifications? The corresponding section in Results (lines 288-292) is also not very helpful in understanding this.
The parasite lines shown in Figure S3D were originally generated by Spreng et al. (2019) and are not directly part of the main set of PTM-targeted mutants described in our study. Specifically, the IntronΔ line carries deletions in introns of the Plasmodium α1-tubulin gene, while the C-termΔ line lacks the final three C-terminal residues (ADY). These lines were included for comparative purposes to explore whether structural changes in α-tubulin could impact polyglutamylation signal, as detected by the polyE (IN105) antibody.
We acknowledge that the figure legend and corresponding text (lines 288–292) did not adequately explain the rationale for including these control lines. We will revise both the legend and Results section to more clearly describe the origin, purpose, and relevance of these mutants to the overall study.
(12) Figure 4E doesn't look like brightfield microscopy but like some sort of fluorescent imaging. In Figure 4C, were the control (NoΔ) cells with an integrated cassette, but no mutations, or non-transgenic cells?
The reviewer is absolutely correct: Figure 4E shows a fluorescent image acquired using widefield microscopy and not a brightfield image. We will revise the figure legend accordingly to avoid confusion. The “BF” (brightfield) label applies only to the left panel in Figure 4C, which depicts oocysts imaged using transmitted light.
Regarding the controls labeled "NoΔ" in Figure 4C, we confirm that these parasites contain the integrated selection cassette but do not harbor any mutations in the target gene. They serve as proper integration controls, allowing us to distinguish the effects of the point mutations or deletions introduced in the experimental lines.
(13) It is difficult to understand why the TyΔ and the CtΔ mutants still show quite a strong signal using the anti-tyrosination antibody. If the mutants have replaced all wild-type alleles, the signal should be completely absent, unless the antibody (see my comment above concerning T9028) cross-reacts with detyrosinated microtubules. Therefore, the quantitation in Figures 5F and 5G is actually indicative of something that shouldn't be like that. The quantitation of 5F is at odds with the microscopy image in 5D. If this image is representative, the anti-Ty staining in TyΔ is as strong as in the control NoΔ.
We agree that the persistence of anti-tyrosination signal in the TyΔ and CtΔ mutant lines is unexpected, given that all wild-type alleles were replaced. This discrepancy has led us to further investigate the specificity of the T9028 antibody, as raised in the reviewer’s earlier comment. To address this concern, we are currently repeating the key experiments using the well-established YL1/2 monoclonal antibody, which is widely accepted for detecting tyrosinated α-tubulin in other systems.
We also acknowledge that Figure 5F shows residual tyrosination signal, and the reviewer is correct that this should not occur if the modified residues are the exclusive PTM sites. One possible explanation is that adjacent residues or even alternative tubulin isoforms may serve as substrates. While α1-tubulin is the dominant isoform in Plasmodium, low-level expression of α2-tubulin has been detected in liver stages based on transcriptomic data, and it may contribute to the observed signal.
Regarding the apparent discrepancy between the quantification in Figure 5F and the representative image in Figure 5D, we will revise the figure legend to clarify that image selection aimed to show detectable signal, not necessarily the average phenotype. We will also reassess and, if needed, repeat the quantification with improved image sets to ensure accuracy and consistency.
We will revise the manuscript to reflect these points and include a more nuanced interpretation of the residual staining in the mutant lines.
(14) The statement that the failure of CtΔ mutants to generate viable sporozoites is due to the lack of microtubule PTMs (lines 295-296) is speculative. The lack of the entire C-terminal tail could have a number of consequences, such as impaired microtubule assembly or failure to recruit and bind associated proteins. This is not necessarily linked to PTMs. Also, it has been shown in yeast that for microtubules to form properly and exquisite regulation (proteostasis) of the ratio between α- and β-tubulin is essential (Wethekam and Moore, 2023). I am not sure, but according to Materials and Methods (line 423), the gene cassettes for replacing the wild-type tubulin gene with the mutant versions contain a selectable marker gene for pyrimethamine selection. Are there qPCR data that show that expression levels of mutant α-tubulin are more or less the same as the wild-type levels?
We agree that attributing the developmental failure of the CtΔ mutants solely to the absence of microtubule post-translational modifications (PTMs) is speculative. As the reviewer rightly points out, deletion of the entire C-terminal tail may have multiple effects, including impaired microtubule assembly, altered α/β-tubulin stoichiometry, or disruption of interactions with essential microtubule-associated proteins (MAPs). These consequences may arise independently of PTMs.
That said, we note that PTMs particularly polyglutamylation, can modulate MAP binding by altering the surface charge of microtubules (Genova et al., 2023; Mitchell et al., 2010). Therefore, while PTM loss may be a contributing factor, we acknowledge that the phenotype likely results from a combination of mechanisms. We will revise the relevant section of the manuscript to present a more cautious and balanced interpretation.
Regarding the reviewer’s question on expression levels: although the replacement constructs include a pyrimethamine resistance cassette, we have not yet quantified α-tubulin transcript levels by qPCR. In the interim, the study by Spreng et al. (2019) (reference 50) on a related α1-tubulin nutations provides valuable insight. They observed no difference in mRNA levels in day 12 oocysts, yet reported fainter microtubule staining and shorter sporozoites, suggesting a post-transcriptional mechanism affecting protein expression or function in later stages. Furthermore, the phenotypic spectrum across their mutant panel (Suppl. Fig. 3 D and E) implies that robust α-tubulin regulation is highly sensitive to specific sequences.
We acknowledge this as a current limitation in our study and will address it in the revised manuscript, noting that direct measurement of transcript levels is a key area for future investigation.
(15) In the Discussion, my impression is that two recent studies, the superb Expansion Microscopy study by Bertiaux et al. (2021) and the cryo-EM study by Ferreira et al. (2023), are not sufficiently recognised (although they are cited elsewhere in the manuscript). The latter study includes a detailed description of the microtubule cytoskeleton in sporozoites. However, the present study clearly expands the knowledge about the structure of the cytoskeleton in liver stage parasites and is one of the few studies addressing the distribution and function of microtubule post-translational modifications in Plasmodium.
Indeed, our work builds upon the established knowledge from Bertiaux et al. (2021) and the cryo-EM study by Ferreira et al. (2023), as rightly mentioned by the reviewer. We agree that these foundational studies, combined with our findings, will significantly expand the understanding of Plasmodium biology and cytoskeleton dynamics across its life cycle and will open the door for further investigations. We are grateful for this suggestion and will ensure these key studies are appropriately acknowledged in the revised manuscript.
(16) I somewhat disagree with the statement of a co-occurrence of polyglutamylated and tyrosinated microtubules. I think the resolution is too low to reach that conclusion. As this is a bold claim, and would be contrary to what is known from other organisms, it would require a more rigorous validation. Given the apparent problems with the anti-Ty antibody (signal in the TyΔ mutant), one should be very cautious with this claim.
This is a very important point to clarify. As mentioned previously, the initial experiments for these modifications were performed independently. It is established that sporozoite subpellicular microtubules exhibit both tyrosination and polyglutamylation. We will revise the manuscript to temper this statement and clearly indicate that the co-occurrence of these PTMs remains a hypothesis that requires more rigorous validation. As suggested, we are now conducting additional co-staining experiments using the better validated YL1/2 antibody to re-express and directly compare the distribution of both PTMs within the same cell. These follow-up experiments will help clarify whether both modifications occur simultaneously on the same microtubule structures in Plasmodium liver stages.
(17) In the Discussion (lines 311 and 377), it is again claimed that tyrosinated microtubules are "a well-known marker of stable microtubules". This statement is completely incorrect, and I am surprised by this serious mistake. A few lines later, the authors say that polyglutamylated is "commonly associated with dynamic microtubule behaviour". Again, this is completely incorrect and is the opposite of what is firmly established in the literature. Polyglutamylation and detyrosination are markers of stable microtubules.
Indeed, in canonical eukaryotic systems, tyrosinated microtubules are generally considered markers of dynamic microtubule populations, whereas detyrosinated and polyglutamylated microtubules are more commonly associated with stability.
We acknowledge this mistake and will revise the Discussion to correct these statements accordingly. In the context of Plasmodium, our observations suggest an unusual regulation of microtubule dynamics, which may reflect parasite-specific adaptations. For example, we observed tyrosinated α-tubulin in the stable subpellicular microtubules of sporozoites structures typically known for their exceptional stability. This atypical association implies either non-canonical roles for tyrosination or parasite-specific mechanisms for modulating microtubule properties. Additionally, the presence of both PTMs at different stages of development and on different microtubule populations suggests tightly regulated spatial and temporal modulation of microtubule function.
We will carefully revise the relevant sections of the manuscript to remove incorrect generalizations and ensure accurate representation of the current consensus in the field, while emphasizing the possibility of Plasmodium-specific adaptations that merit further study.
(18) In line 339, the authors interpret the residual antibody staining after the introduction of the mutant tubulin as a compensatory mechanism. There is no evidence for this. More likely explanations are firstly the quality of the anti-Ty-antibody used (see comment above), and the fact that also β-tubulin carries C-terminal polyglutamylation sites, which haven't been investigated in this study. PTMs on β-tubulin are not compensatory, but normal PTMs, at least in all other organisms where microtubule PTMs have been investigated.
As mentioned above, we are currently repeating the key experiments with the [YL1/2] antibody, as suggested. Furthermore, we fully agree with the reviewer's point regarding polyglutamylation on β-tubulin. The C-terminal tail of β-tubulin does indeed contain polyglutamylation sites. As we noted in the manuscript (Lines 340-352), this aspect has not been investigated in the present study, and we acknowledge it as a valuable direction for future research. We will revise the text accordingly to avoid overinterpretation and to more accurately reflect the limitations of our current data.
