Cell surface mucins, such as MUC1 and MUC16, are so consistently upregulated in epithelial cancers that they are considered reliable biomarkers of the disease. (Bast et al., 2009; Rahn et al., 2001) Despite their importance for diagnosis and prognosis, the mechanisms by which mucins might promote malignancy remain largely speculative. (Kufe, 2009a) To date, the majority of studies exploring the role of mucins in determining tumor phenotype have focused on the signaling function of these molecules, which resides within their short cytoplasmic tail. Yet, a striking and defining feature of cell surface mucins is their long, densely glycosylated ectodomain, which can extend hundreds of nanometers from the plasma membrane. (Hattrup and Gendler, 2008) Consequently, mucins can profoundly enhance the bulk physical properties of the glycocalyx, which is a composite of the polysaccharides and glycoproteins that project from the plasma membrane and decorate all cells.

We recently showed that the bulk physical properties of the glycocalyx can exert profound effects on cellular behavior. These effects include a reorganization of cell surface receptors such as integrins. This reorganization alters their activation state and influences not only their ability to interact with extracellular matrix (ECM) ligands but also their synergistic downstream signaling with growth factor receptors. (Paszek et al., 2014; Freeman et al., 2016) By occluding the majority of binding sites, an expansive glycocalyx paradoxically promotes integrin clustering by creating a kinetic funnel in which integrins are most likely to bind to the ECM where bonds are already formed. (Paszek et al., 2009) In this regard, in vitro studies revealed that increasing the bulkiness of the glycocalyx enhanced the ability of nonmalignant mammary epithelial cells to grow and survive under minimally adhesive conditions whereas otherwise the cells underwent anoikis. (Paszek et al., 2014; Woods et al., 2015; Desgrosellier and Cheresh, 2010) Whether a bulky glycocalyx similarly modulates cell growth and survival in vivo, however, has yet to be unambiguously tested.

Importantly, we and others have also noted that tumor aggression is frequently associated with an increased expression of mucins which enhance the bulkiness of the glycocalyx and, similarly, that circulating epithelial tumor cells typically express an abundance of mucins. (Paszek et al., 2014) Concomitantly, primary tumor xenografts that overexpress MUC1 grow and metastasize more aggressively (Wang et al., 2015) and mice deficient in MUC1 resist formation of spontaneous tumors (Spicer et al., 1995). Together, these observations raise the intriguing possibility that tumor associated mucin expression may foster tumor progression and aggression by enhancing either tumor cell growth and survival or both by modulating integrin and growth factor receptor signaling.

Herein, we report that increasing the thickness of the glycocalyx promotes tumor metastasis by permitting cell cycle progression to facilitate cell proliferation at the metastatic site. We modulated glycocalyx thickness both genetically through ectopic expression of a tailless (signaling defective) MUC1 and more definitively by coating cells with synthetic glycopolymers with long-term cell surface retention that were designed to emulate the structure of mucin ectodomains (Figure 1A). (Woods et al., 2015) The longevity of the synthetic glycopolymers was facilitated by developing the chemistry so that the tail terminated in a cholesterylamine lipid at one end, which effectively drives spontaneous insertion into cell membranes based on mass action and the hydrophobic effect. We previously found that this particular lipid confers a long cell surface residence time to glycopolymer conjugates by mediating continuous recycling from an internal reservoir. (Woods et al., 2015) The core of the glycopolymer comprises N-acetylgalactosamine (GalNAc) residues affixed to a poly (methyl vinyl ketone [MVK]) backbone via oxime linkages. The spacing afforded by this structure approximates that of GalNAc-α-Ser/Thr clusters within native mucins. Likewise, GalNAc-modified poly (MVK) glycopolymers emulate the physical characteristics of native mucins such as length and stiffness. (Paszek et al., 2014) However, unlike the more complex glycans found in native mucins, which can engage in biochemical interactions, (Hudak et al., 2014) the GalNAc residues we used here are considered biochemically inert in mammalian systems. Thus, these synthetic structures allowed us to model mucins’ biophysical contribution to the glycocalyx without any confounding contributions from their cytosolic or extracellular biochemical activity. As well, the living polymerization chemistry used to generate the glycodomains enabled precise control of polymer length with high homogeneity. (Godula et al., 2009) Living polymerization techniques allow for the synthesis of polymers of very homogeneous lengths, a necessity for the comparison of the effects of long and short polymers. In this work, we used short glycopolymers of ~3 nm in length (degree of polymerization (DP) = 36 monomer units) and long glycopolymers that were ~90 nm in length (DP = 719 monomer units). As a point of comparison, integrin heterodimer pairs extend roughly 20 nm from the cell membrane (Figure 1A). (Eng et al., 2011)

Figure 1 with 3 supplements see all

Download asset Open asset

Metastasis is a multistep process that requires cancer cells to overcome many checks to their irregular growth – limited not only by ability to disseminate but also to survive and thereafter to grow and expand at the metastatic site. The nonmalignant mammary epithelial cell line, MCF-10A, offers a unique model for oncogenic progression. (Miller et al., 1993) These cells, though immortal, lack the ability to grow or survive efficiently in a soft matrix, and therefore serve as an ideal cell line to evaluate the effects of the glycocalyx on mammary epithelial cell behavior in the context of a metastatic site. (Paszek et al., 2005) Our prior studies suggested that a bulky glycocalyx permitted not only the survival but also the growth of these mammary epithelial cells in an in vitro model of a minimally adhesive microenvironment. Thus, we asked whether a bulky glycocalyx would permit both the survival and the growth and expansion of disseminated mammary epithelial cells in vivo. MCF-10A cells were painted with our glycopolymers, then injected into the venous system of nude mice, where they accumulated in the lungs regardless of glycopolymer length. We found that those cells treated with long glycopolymer mucin-mimetics survived longer, as demonstrated by delayed activation of the apoptosis marker cleaved caspase-3 (CC3), than those decorated with short glycopolymers (Figure 1—figure supplement 1), consistent with our in vitro studies. Nevertheless, the cells eventually died and therefore ultimately failed to form metastatic lesions, leaving us unable to determine the potential effect of the glycocalyx on proliferation. Thus, to unambiguously address whether or not a bulky glycocalyx might influence tumor behavior and metastasis by regulating cell proliferation, independent of cell survival, we assessed the influence of the glycopolymers on the growth and metastatic efficiency of the dormant-prone tumor cell line from the syngeneic 4TO7/Balbc model.

The murine cell line 4TO7 is a Balb/c syngeneic mammary carcinoma that is able to efficiently disseminate and survive for prolonged periods of time at metastatic sites such as the lung. (Aslakson and Miller, 1992) However, the tumor cells in this model form predominately dormant tumors because they are unable to proliferate efficiently at the target site. These 4TO7 cells, therefore, provided us the ideal model to specifically address the contribution of a bulky glycocalyx on cell proliferation and G1 cell cycle transit in vivo.

A thick glycocalyx drives cell cycle progression via the PI3K-Akt axis

To gain mechanistic insight into this apparent proliferative effect, we tested the effects of modulating glycocalyx thickness on cell growth in a soft matrix in vitro model. We used fibronectin-functionalized polyacrylamide (PA) gels whose stiffness can be controlled by varying both acrylamide concentration and crosslinking frequency. (Lakins et al., 2012) We painted MCF-10A cells with short or long mucin-mimetic glycopolymers, or with vehicle (PBS), then plated them at low density on fibronectin-functionalized soft PA gels (Young’s modulus = 400 Pa) so that nearly all cells were isolated, minimizing cell-cell interactions. After 72 hr, we imaged the gels and quantified the number of cells in each observed colony. The long glycopolymer-painted cells formed colonies with, on average, twice as many cells as colonies formed by the short glycopolymer-treated cells or PBS-treated controls (Figure 2A–C). And the percentage of colonies growing beyond two cells nearly tripled for long-glycopolymer-treated cells verses PBS-treated controls, indicating that this increase in mean colony size was not due merely to a few large colonies (Figure 2D).

Figure 2 with 5 supplements see all

Download asset Open asset

To test that this phenotype required cell-surface residence of the long glycopolymers, we compared the effects of the cholesterylamine-anchored molecules, which persist on the plasma membrane for several days even after cell division, with those of a phospholipid-functionalized glycopolymer that is lost from the cell surface within several hours. (Woods et al., 2015) Cells treated with these short-lived but equally long glycopolymers (DPPE) did not produce larger colonies compared to PBS treatment alone (Figure 2—figure supplement 1). Thus, the effects measured in Figure 2C and D are likely mediated by glycopolymers resident on the cell surface.

To further test our hypothesis that a bulky glycocalyx permits proliferation in the metastatic niche, we analyzed markers of cell cycle progression. Cells were painted with short or long glycopolymers, or with vehicle (PBS), then plated on either soft (Young’s modulus = 400 Pa) or stiff (Young’s modulus = 60,000 Pa) fibronectin-functionalized PA gels, the latter being a positive control for integrin activation. (Discher et al., 2005) After allowing them to adhere for 6 hr, cells were lysed and the lysates analyzed by immunoblotting. We found that MCF-10A cells plated on stiff PA gels, which offer a rich adhesion setting, produced higher levels of cyclin D (a protein required for cell cycle progression) and showed higher levels of pH3 than those on adhesion-poor soft gels (Figure 2E and F). Long glycopolymer-treated cells also demonstrated increased cyclin D and pH3 levels, compared to short glycopolymer- and PBS-treated cells, even on soft gels, suggesting that a thick glycocalyx drove proliferation and cell cycle progression in this poor adhesion setting.

Integrin activation and growth factor signaling form, in logic terms, an ‘AND gate’ for cell cycle progression in nontransformed cells. (Kuwada and Li, 2000) We sought to determine whether a bulky glycocalyx alters this fundamental control mechanism. After serum-starving MCF-10A cells for 72 hr, we lifted them and painted them with long or short glycopolymers, or with vehicle. The cells were plated on gels as in Figure 2E, except now in serum-free media, and then lysed after 6 hr. A lack of growth factor signaling reduced cyclin D1 and pH3 levels to below the detection limit, even when cells were grown on stiff gels or endowed with a bulky glycocalyx (Figure 2—figure supplement 2). The underlying drivers of proliferation therefore, including growth factor signaling requirements, appeared to be unaltered by glycocalyx thickness in this model.

To discern the factors responsible for increased cell cycle progression in cells with thicker glycocalyces, we examined Erk and Akt, the mitogen activated protein kinases (MAPKs) thought to be central to adhesion-mediated control of cell cycle progression. (Moreno-Layseca and Streuli, 2014) After serum starvation for 72 hr, cells were coated with glycopolymers or vehicle, and plated on PA gels in serum-free media for 6 hr. The cells were then challenged with epidermal growth factor (EGF) for 15 min and then lysed. A bulky glycocalyx increased Akt phosphorylation (pAkt), a proxy for activation, on soft gels relative to that in cells with short glycopolymers, while Erk phosphorylation, however, remained unchanged (Figure 2G and H). The same effect on pAkt was seen on 4TO7 cells in vitro (Figure 2I). As well, IF analysis of tissue sections from the experiments in Figure 1 revealed increased Akt activation, as measured by staining for phosphorylated Akt substrate, in vivo in long glycopolymer-bearing tumor cells when compared to short (Figure 2—figure supplement 3). Consistent with this finding, we repeated the colony formation experiments and found that proliferation of long glycopolymer-painted cells was blocked by an inhibitor of phosphoinositol-3 kinase (PI3K), an upstream activator of Akt, or an inhibitor of MEK—a MAPK downstream of growth factor receptors (Figure 2—figure supplement 4). A bulky glycocalyx, therefore, seems to drive cell cycle progression via the PI3K-Akt axis.

While 4TO7 cells in vitro did not experience significantly increased cyclin D1 expression with a thicker glycocalyx (Figure 2—figure supplement 5), we were able to confirm the role of glycocalyx-driven cell cycle progression in promoting metastasis in vivo by IF staining of the 4TO7 lung mets from the experiments in Figure 1. Metastatic 4TO7 cells in vivo expressed significantly more of the cell cycle progression marker cyclin D1 when bearing long glycopolymers verses short (Figure 2J and K). We posit that transformed cells, such as 4TO7, in rich culture media in vitro may see only a negligible increase in proliferative signal from a bulky glycocalyx as evidenced by the robust cyclin D1 signal in the negative control: PBS treated 4TO7 cells on soft gels.

A thick glycocalyx stimulates integrin-FAK mechanosignaling

Next we sought to test our hypothesis that this phenotype is driven by integrin- focal adhesion kinase (FAK) signaling. FAK disseminates adhesion information from focal adhesions to the rest of the cell via autophosphorylation at tyrosine 397, as well as increased phosphorylation at tyrosine 925, presumably through Src kinase activation. (Mitra et al., 2005) Western blotting with phospho-specific antibodies revealed that possession of a thick glycocalyx led to enhanced activation of FAK (Figure 3A and B). These effects on FAK were mirrored by adhesion of untreated cells to a stiff matrix. By contrast, cells coated with short glycopolymers, or left untreated, and plated on a soft matrix showed comparatively less phosphorylation of FAK. Similar results were seen with 4TO7 cells in vitro (Figure 3C). Reciprocally, cells treated with long glycopolymers along with a nonlethal dose of a specific inhibitor of FAK were unable to form colonies on soft matrices, whereas vehicle (DMSO)-treated, long glycopolymer-coated cells retained colony forming activity (Figure 3—figure supplement 1).

Figure 3 with 1 supplement see all

Download asset Open asset

To determine whether enhancement of adhesion signaling due to a bulky glycocalyx also occurs in vivo, we probed FAK phosphorylation status in the mets from Figure 1. Long glycopolymer-treated cells formed tumors with significantly more FAK activation than tumors formed from short glycopolymer-bearing cells (Figure 3D and E). These data, taken with the results in Figure 2, support a model for glycocalyx-driven enhancement of proliferation in minimal-adhesion settings. A bulky glycocalyx drives integrin clustering through kinetic funneling, leading to increased signaling from focal adhesion-associated proteins such as FAK. In combination with growth factor signaling, this leads to enhanced MAPK activation, such as Akt phosphorylation, which then promotes G1 cell cycle progression through expression of proteins like the cyclins (Figure 3F).

The MUC1 ectodomain is sufficient to increase the metastatic potential of 4TO7 cancer cells

Lastly, for physiological relevance, we evaluated the effects of a natural mucin ectodomain on the metastatic potential of 4TO7 cells. We utilized a cytoplasmic truncation of MUC1 (MUC1ΔCT), the mucin most commonly overexpressed in breast and ovarian cancers, to limit any cytoplasmic signaling which might confound our results. We transfected mApple-luciferase expressing 4TO7 cells with MUC1ΔCT under the control of a doxycycline (dox)-inducible promoter. We grew these cells to subconfluence, treated them with doxycycline (200 ng/ml) or vehicle (PBS) for 24 hr, then lifted them and immediately injected them into Balb/c mouse tail veins (Figure 4A). As previously, mice were monitored every 4–5 days via BLI measurements and sacrificed on day 15.

Figure 4 with 2 supplements see all

Download asset Open asset

BLI photon flux revealed that MUC1ΔCT-expressing cells formed mets that were growing more rapidly than those formed from MUC1ΔCT-negative cells (Figure 4—figure supplement 1). Whole lung sectioning, H and E staining and IHC on mApple revealed a dramatic qualitative difference in tumor burden (Figure 4B and C). Wet lung mass was nearly doubled in mice injected with MUC1ΔCT-expressing cells versus control (Figure 4D). Nearly three times the number of mets were formed in the lungs of mice injected with cells expressing the mucin ectodomain compared to those that were not (Figure 4E). pH3 IF staining revealed that more than twice the number of nuclei in MUC1ΔCT-expressing cells were actively mitotic. And while the values did not reach statistical significance (p=0.06), CC3 levels were trending towards lower values in MUC1ΔCT-expressing mets (Figure 4F–H). Staining for pY397-FAK, cyclin D1 and pAkt substrate demonstrated that the MUC1 ectodomain increases mechanosignaling, cell cycle progression and MAPK activity in lung mets, further supporting our model (Figure 4I–K and Figure 4—figure supplement 2). These results dramatically demonstrate the effect that the mucin ectodomain can have on metastatic cells in vivo.

Mucin overexpression is commonly observed in epithelial cancers. MUC1 in particular is overexpressed in ~64% of tumors of all types diagnosed each year in the U.S. (Kufe, 2009b), rendering MUC1 one of the most prominently dysregulated genes in cancer. As a point of reference, Ras (K-, H- and N-RAS combined) mutations are estimated to occur in 9–30% of all cancers. (Cox et al., 2014) Hypotheses regarding the mechanism by which MUC1 overexpression drives cancer progression have focused almost entirely on biochemical interactions of its 72-residue cytosolic domain (Kufe, 2009b), which represents <10% of the overall protein sequence. The bulk of MUC1 resides outside the cell where it dominates the physical properties of the glycocalyx. In previous work, we showed that this ectodomain profoundly influences focal adhesion formation, integrin signaling, and survival in a minimal adhesion setting. (Paszek et al., 2014) But this effect alone cannot explain the striking effect of MUC1 ectodomain expression on metastatic burden that we observed in this study (Figure 4). Our data herein show that a bulky glycocalyx, achieved either with synthetic or natural mucins, also promotes proliferation in the metastatic niche. The mucin ectodomain promotes mechanosignaling and enhances cell cycle progression via the PI3K-Akt axis. This model unifies the structure of mucins with their consistent overexpression in metastatic disease (Horm and Schroeder, 2013) and the correlation of their overexpression with poor prognosis. (Rahn et al., 2001; Duffy et al., 2000; Retterspitz et al., 2010; McGuckin et al., 1995) As well, importantly, our results imply that drugs targeting the cytoplasmic tail of MUC1 will be missing a key pathophysiologic mechanism.

It should be noted that in addition to their physical influence, the glycans on mucins have been found to participate in biochemical interactions. For example, sialylated mucin-associated glycans engage the Siglec family of immunomodulatory receptors and may therefore tune the response of critical effector cells in the tumor microenvironment. (Belisle et al., 2010; Ohta et al., 2010; Beatson et al., 2016) Thus, mucins’ influence on cancer likely reflects many functional modalities, each contributing differentially to various facets of disease progression. From this vantage point, mucins are prime targets for therapeutic intervention and warrant increased focus on avenues for their disruption.

1. 1. Aslakson CJ
   2. Miller FR

   (1992)

   Selective events in the metastatic process defined by analysis of the sequential dissemination of subpopulations of a mouse mammary tumor

   Cancer Research 52:1399–1405.

   • PubMed
   • Google Scholar
2. 1. Bast RC
   2. Hennessy B
   3. Mills GB

   (2009) The biology of ovarian cancer: new opportunities for translation

   Nature Reviews Cancer 9:415–428.

   https://doi.org/10.1038/nrc2644

   • PubMed
   • Google Scholar
3. 1. Beatson R
   2. Tajadura-Ortega V
   3. Achkova D
   4. Picco G
   5. Tsourouktsoglou TD
   6. Klausing S
   7. Hillier M
   8. Maher J
   9. Noll T
   10. Crocker PR
   11. Taylor-Papadimitriou J
   12. Burchell JM

   (2016) The mucin MUC1 modulates the tumor immunological microenvironment through engagement of the lectin Siglec-9

   Nature Immunology 17:1273–1281.

   https://doi.org/10.1038/ni.3552

   • PubMed
   • Google Scholar
4. 1. Belisle JA
   2. Horibata S
   3. Jennifer GA
   4. Petrie S
   5. Kapur A
   6. André S
   7. Gabius HJ
   8. Rancourt C
   9. Connor J
   10. Paulson JC
   11. Patankar MS

   (2010) Identification of Siglec-9 as the receptor for MUC16 on human NK cells, B cells, and monocytes

   Molecular Cancer 9:118.

   https://doi.org/10.1186/1476-4598-9-118

   • PubMed
   • Google Scholar
5. 1. Chambers AF
   2. Groom AC
   3. MacDonald IC

   (2002) Dissemination and growth of cancer cells in metastatic sites

   Nature Reviews Cancer 2:563–572.

   https://doi.org/10.1038/nrc865

   • PubMed
   • Google Scholar
6. 1. Cox AD
   2. Fesik SW
   3. Kimmelman AC
   4. Luo J
   5. Der CJ

   (2014) Drugging the undruggable RAS: Mission possible?

   Nature Reviews Drug Discovery 13:828–851.

   https://doi.org/10.1038/nrd4389

   • PubMed
   • Google Scholar
7. 1. Desgrosellier JS
   2. Cheresh DA

   (2010) Integrins in cancer: biological implications and therapeutic opportunities

   Nature Reviews Cancer 10:9–22.

   https://doi.org/10.1038/nrc2748

   • PubMed
   • Google Scholar
8. 1. Discher DE
   2. Janmey P
   3. Wang YL

   (2005) Tissue cells feel and respond to the stiffness of their substrate

   Science 310:1139–1143.

   https://doi.org/10.1126/science.1116995

   • PubMed
   • Google Scholar
9. 1. Duffy MJ
   2. Shering S
   3. Sherry F
   4. McDermott E
   5. O'Higgins N

   (2000)

   CA 15-3: a prognostic marker in breast cancer

   The International Journal of Biological Markers 15:330–333.

   • PubMed
   • Google Scholar
10. 1. Eng ET
    2. Smagghe BJ
    3. Walz T
    4. Springer TA

    (2011) Intact alphaIIbbeta3 integrin is extended after activation as measured by solution X-ray scattering and electron microscopy

    The Journal of Biological Chemistry 286:35218–35226.

    https://doi.org/10.1074/jbc.M111.275107

    • PubMed
    • Google Scholar
11. 1. Freeman SA
    2. Goyette J
    3. Furuya W
    4. Woods EC
    5. Bertozzi CR
    6. Bergmeier W
    7. Hinz B
    8. van der Merwe PA
    9. Das R
    10. Grinstein S

    (2016) Integrins form an expanding diffusional barrier that coordinates phagocytosis

    Cell 164:128–140.

    https://doi.org/10.1016/j.cell.2015.11.048

    • PubMed
    • Google Scholar
12. 1. Godula K
    2. Umbel ML
    3. Rabuka D
    4. Botyanszki Z
    5. Bertozzi CR
    6. Parthasarathy R

    (2009) Control of the molecular orientation of membrane-anchored biomimetic glycopolymers

    Journal of the American Chemical Society 131:10263–10268.

    https://doi.org/10.1021/ja903114g

    • PubMed
    • Google Scholar
13. 1. Hattrup CL
    2. Gendler SJ

    (2008) Structure and function of the cell surface (tethered) mucins

    Annual Review of Physiology 70:431–457.

    https://doi.org/10.1146/annurev.physiol.70.113006.100659

    • PubMed
    • Google Scholar
14. 1. Horm TM
    2. Schroeder JA

    (2013) MUC1 and metastatic cancer: expression, function and therapeutic targeting

    Cell Adhesion & Migration 7:187–198.

    https://doi.org/10.4161/cam.23131

    • PubMed
    • Google Scholar
15. 1. Hudak JE
    2. Canham SM
    3. Bertozzi CR

    (2014) Glycocalyx engineering reveals a Siglec-based mechanism for NK cell immunoevasion

    Nature Chemical Biology 10:69–75.

    https://doi.org/10.1038/nchembio.1388

    • PubMed
    • Google Scholar
16. 1. Kufe DW

    (2009a) Mucins in cancer: function, prognosis and therapy

    Nature Reviews Cancer 9:874–885.

    https://doi.org/10.1038/nrc2761

    • PubMed
    • Google Scholar
17. 1. Kufe DW

    (2009b) Functional targeting of the MUC1 oncogene in human cancers

    Cancer Biology & Therapy 8:1197–1203.

    https://doi.org/10.4161/cbt.8.13.8844

    • PubMed
    • Google Scholar
18. 1. Kuwada SK
    2. Li X

    (2000) Integrin alpha5/beta1 mediates fibronectin-dependent epithelial cell proliferation through epidermal growth factor receptor activation

    Molecular Biology of the Cell 11:2485–2496.

    https://doi.org/10.1091/mbc.11.7.2485

    • PubMed
    • Google Scholar
19. 1. Lakins JN
    2. Chin AR
    3. Weaver VM

    (2012) Exploring the link between human embryonic stem cell organization and fate using tension-calibrated extracellular matrix functionalized polyacrylamide gels

    Methods in Molecular Biology 916:317–350.

    https://doi.org/10.1007/978-1-61779-980-8_24

    • PubMed
    • Google Scholar
20. 1. McGuckin MA
    2. Walsh MD
    3. Hohn BG
    4. Ward BG
    5. Wright RG

    (1995) Prognostic significance of MUC1 epithelial mucin expression in breast cancer

    Human Pathology 26:432–439.

    https://doi.org/10.1016/0046-8177(95)90146-9

    • PubMed
    • Google Scholar
21. 1. Miller FR
    2. Soule HD
    3. Tait L
    4. Pauley RJ
    5. Wolman SR
    6. Dawson PJ
    7. Heppner GH

    (1993) Xenograft model of progressive human proliferative breast disease

    JNCI Journal of the National Cancer Institute 85:1725–1732.

    https://doi.org/10.1093/jnci/85.21.1725

    • PubMed
    • Google Scholar
22. 1. Mitra SK
    2. Hanson DA
    3. Schlaepfer DD

    (2005) Focal adhesion kinase: in command and control of cell motility

    Nature Reviews Molecular Cell Biology 6:56–68.

    https://doi.org/10.1038/nrm1549

    • PubMed
    • Google Scholar
23. 1. Moreno-Layseca P
    2. Streuli CH

    (2014) Signalling pathways linking integrins with cell cycle progression

    Matrix Biology 34:144–153.

    https://doi.org/10.1016/j.matbio.2013.10.011

    • PubMed
    • Google Scholar
24. 1. Ohta M
    2. Ishida A
    3. Toda M
    4. Akita K
    5. Inoue M
    6. Yamashita K
    7. Watanabe M
    8. Murata T
    9. Usui T
    10. Nakada H

    (2010) Immunomodulation of monocyte-derived dendritic cells through ligation of tumor-produced mucins to Siglec-9

    Biochemical and Biophysical Research Communications 402:663–669.

    https://doi.org/10.1016/j.bbrc.2010.10.079

    • PubMed
    • Google Scholar
25. 1. Paszek MJ
    2. Boettiger D
    3. Weaver VM
    4. Hammer DA

    (2009) Integrin clustering is driven by mechanical resistance from the glycocalyx and the substrate

    PLoS Computational Biology 5:e1000604.

    https://doi.org/10.1371/journal.pcbi.1000604

    • PubMed
    • Google Scholar
26. 1. Paszek MJ
    2. DuFort CC
    3. Rossier O
    4. Bainer R
    5. Mouw JK
    6. Godula K
    7. Hudak JE
    8. Lakins JN
    9. Wijekoon AC
    10. Cassereau L
    11. Rubashkin MG
    12. Magbanua MJ
    13. Thorn KS
    14. Davidson MW
    15. Rugo HS
    16. Park JW
    17. Hammer DA
    18. Giannone G
    19. Bertozzi CR
    20. Weaver VM

    (2014) The cancer glycocalyx mechanically primes integrin-mediated growth and survival

    Nature 511:319–325.

    https://doi.org/10.1038/nature13535

    • PubMed
    • Google Scholar
27. 1. Paszek MJ
    2. Zahir N
    3. Johnson KR
    4. Lakins JN
    5. Rozenberg GI
    6. Gefen A
    7. Reinhart-King CA
    8. Margulies SS
    9. Dembo M
    10. Boettiger D
    11. Hammer DA
    12. Weaver VM

    (2005) Tensional homeostasis and the malignant phenotype

    Cancer Cell 8:241–254.

    https://doi.org/10.1016/j.ccr.2005.08.010

    • PubMed
    • Google Scholar
28. 1. Rahn JJ
    2. Dabbagh L
    3. Pasdar M
    4. Hugh JC

    (2001)

    The importance of MUC1 cellular localization in patients with breast carcinoma: an immunohistologic study of 71 patients and review of the literature

    Cancer 91:1973–1982.

    • PubMed
    • Google Scholar
29. 1. Retterspitz MF
    2. Mönig SP
    3. Schreckenberg S
    4. Schneider PM
    5. Hölscher AH
    6. Dienes HP
    7. Baldus SE

    (2010)

    Expression of {beta}-catenin, MUC1 and c-met in diffuse-type gastric carcinomas: correlations with tumour progression and prognosis

    Anticancer Research 30:4635–4641.

    • PubMed
    • Google Scholar
30. 1. Spicer AP
    2. Rowse GJ
    3. Lidner TK
    4. Gendler SJ

    (1995) Delayed mammary tumor progression in Muc-1 null mice

    The Journal of Biological Chemistry 270:30093–30101.

    https://doi.org/10.1074/jbc.270.50.30093

    • PubMed
    • Google Scholar
31. 1. Wang X
    2. Lan H
    3. Li J
    4. Su Y
    5. Xu L

    (2015)

    Muc1 promotes migration and lung metastasis of melanoma cells

    American Journal of Cancer Research 5:2590–2604.

    • PubMed
    • Google Scholar
32. 1. Woods EC
    2. Yee NA
    3. Shen J
    4. Bertozzi CR

    (2015) Glycocalyx engineering with a recycling glycopolymer that increases cell survival in vivo

    Angewandte Chemie International Edition 54:15782–15788.

    https://doi.org/10.1002/anie.201508783

    • PubMed
    • Google Scholar
33. 1. Yang J
    2. Mani SA
    3. Donaher JL
    4. Ramaswamy S
    5. Itzykson RA
    6. Come C
    7. Savagner P
    8. Gitelman I
    9. Richardson A
    10. Weinberg RA

    (2004) Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis

    Cell 117:927–939.

    https://doi.org/10.1016/j.cell.2004.06.006

    • PubMed
    • Google Scholar

Thank you for submitting your article "A bulky glycocalyx fosters metastasis formation by promoting G1 cell cycle progression" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Johanna Ivaska as the Reviewing Editor and Philip Cole as the Senior Editor. The following individual involved in review of your submission has agreed to reveal his identity: Ron Schnaar (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

Aberrant expression of MUC1 is found in the majority of carcinomas and in hematological cancers. It has been demonstrated earlier by the authors of this manuscript that addition of ~80nm glycoprotein mimetics or enforced expression of MUC1 stimulates survival and proliferation of non-transformed MECs on soft substrates that would otherwise not support survival (Nature 2015). It was also found that oncogenic transformation, which allows these MECs to proliferate on soft substrates, is accompanied by increased expression of MUC1 on these cells. This earlier work indicated that a thicker glycocalyx augments growth-promoting integrin signaling through tension applied to ECM-bound integrins.

In the current manuscript, the authors extend their work and show that expression of a tailless MUC1 mutant or coating with a ~90nm synthetic GalcNAc polymer is sufficient to support outgrowth of IV-injected 4TO7 breast cancer cells in the mouse lung. This represents a new finding, demonstrating that manipulation of the glycocalyx at the extracellular surface, rather than signaling by MUCs drives metastatic outgrowth. Mechanistic studies in vitro revealing that MCF10A cells coated with large polymers show proliferation even on soft substrates is largely a confirmation of previous work (Nature 2015). The idea that FAK and PI3K-AKT signaling as well as cyclin D accumulation is increased on soft substrates and in metastatic lesions in the presence of the large polymers, extends the previous work but in its current form is not fully convincing.

Overall the reviewers find the work interesting, largely novel, and the data in the manuscript to be mainly of good quality. However, additional controls are required and further exploration of alternative explanations of the data should be considered. In addition, more Introduction/Discussion regarding the kinetic funneling concept are needed.

Essential revisions:

1) Figure 1 – controls are required showing what the metastatic potential of parental 4T07 cells are in this experimental set up. There is significant metastatic formation following treatment of the cells with the short polymer even though there is no signal activation in the in vitro model with this polymer. Therefore it is necessary to show whether this polymer is also having a significant effect in vivo.

2) Figure 1: It is important to rigorously rule out that the bulky glycocalyx in fact increases the propensity of IV injected tumor cells to get stuck in the lung vasculature. Figure 1D shows that the number of macroscopically detectable lung mets is higher in the presence of the long polymer. Is this in any way related to more cells getting trapped? For this, lung tissue can be isolated at early time points post injection (days 1, 2, 3, 4, 5) and the number of trapped tumor cells quantitatively determined by qPCR.

3) Figure 1, 4 versus 2, 3: Why do the authors switch from 4TO7 to MCF10A here? The in vivo work is with 4TO7, the in vitro experiments with MCF10A. We agree that MCF10A is a good model but are the same results seen with 4TO7 cells? Are 4TO7 cells indeed not showing proliferation on soft substrates and is FAK, PI3K-AKT signaling, cyclin D accumulation in those cells under those conditions stimulated by the long polymers? Not all experiments, but at least confirmation of a few key findings across these two models would be required to directly relate in vitro work in Figure 2 and 3 to the in vivo experiments in Figure 1 and 4.

4) How valid is it to compare the signaling events in the in vitro model with those seen in vivo? For example activation of AKT signaling was reported in vitro but not shown in vivo. pAKT IHC should be carried out on lung metastases in the experiments described for the polymer coated cells and also those where MUC1ΔCT-expressing cells were used. Only then can the authors claim that the PI3K/AKT pathway plays a role in the lung environment.

5) Figure 2 and 3: Increased pH3 levels (Figure 2) and FAK-pY925 (Figure 3) on soft substrate coated with FN in presence of long polymers is clear. However, increased cyclin D levels (compared to PBS and short polymer in Figure 2E) and increased AKT phosphorylation (compared to PBS control in Figure 2G) are not convincing at all and increased FAK-pY397 is very modest at best (compared to PBS and short polymer in Figure 3A). Western blots are indicated by the authors to be representative of at least 2 (Figure 2) or 3 (Figure 3) experiments. If this is the best example, this data is not convincing. Notably, the effects on pH3 and colony size are clear so the effect on proliferation is well-demonstrated. However, specifically for cyclin D, p-AKT, and p-Y397FAK data, the authors should present WB data where differences are convincing and rigorous quantification of 3 or more biologically independent experiments need to be provided.

6) pH3 analysis was carried out on 3-4 tumors per mouse. Does the level of positivity depend on the size of the individual tumors? As this analysis was carried out at a single time point it provides a snapshot of proliferation at that time. It may be that reduced proliferation in the short polymer treated cells reflects a delay in their initial growth (and therefore size of the individual tumors) and not one of reduced proliferation per se. The same argument could be made for the cyclin D1 analysis.

7) Figure 4: were the mice fed doxycline? What are the levels of MUC1ΔCT at the time of sacrifice? If the expression is not maintained for the duration of the experiment it could again suggest that the effects of MUC1ΔCT may be mediated at very early stages of tumor cell arrest and extravasation in the lung.

8) Cells were "painted.… with glycopolymers" but the methods used are not apparent within the manuscript. Details sufficient for reproducing this aspect of the experimental approach would be welcome.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Thank you for submitting your article "A bulky glycocalyx fosters metastasis formation by promoting G1 cell cycle progression" for consideration by eLife. Evaluation of your article has been overseen by a Reviewing Editor and Philip Cole as the Senior Editor.

After discussion, the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The authors have revised the manuscript considerably and addressed many of the previous points raised by the reviewers. However, it is felt that two of the main points need further work (see below).

Essential revisions:

"Point 3) Figure 1, 4 versus 2, 3: Why do the authors switch from 4TO7 to MCF10A here? The in vivo work is with 4TO7, the in vitro experiments with MCF10A. We agree that MCF10A is a good model but are the same results seen with 4TO7 cells? Are 4TO7 cells indeed not showing proliferation on soft substrates and is FAK, PI3K-AKT signaling, cyclin D accumulation in those cells under those conditions stimulated by the long polymers? Not all experiments, but at least confirmation of a few key findings across these two models would be required to directly relate in vitro work in Figure 2 and 3 to the in vivo experiments in Figure 1 and 4."

To their credit, the authors have explained why MCF10A cells were not used for the metastasis in vivo studies and have added data that show the limitations in this regard. However, it was not clear to the reviewers why 4T07 cells were not studied 'in vitro' in terms of cell signaling. It would seem that these cells can be analyzed in culture since the authors discuss growing them to sub-confluency. We thus ask the authors to investigate the FAK, PI3K-AKT, and cyclin D status in 4T07 as this would be important to demonstrate the generality and robustness of the main findings.

"Point 5) However, specifically for cyclin D, p-AKT, and p-Y397FAK data, the authors should present WB data where differences are convincing and rigorous quantification of 3 or more biologically independent experiments need to be provided."

We appreciate that the authors have now added densitometry and error bars to the western blot results in the revised manuscript. However, given the limitations and challenges of western blots, it is desirable to perform at least three replicates and calculate p values to determine the reliability and significance of the findings. We hope that the authors can therefore add at least one additional replicate to the experiments on cyclin D, pAKT and pY397FAK to ensure the robustness of the results.

Essential revisions:

1) Figure 1 – controls are required showing what the metastatic potential of parental 4T07 cells are in this experimental set up. There is significant metastatic formation following treatment of the cells with the short polymer even though there is no signal activation in the in vitro model with this polymer. Therefore it is necessary to show whether this polymer is also having a significant effect in vivo.

The reviewers bring up a very important point: our negative control (short glycopolymer-bearing cells) requires validation. We have added a new supplementary figure to illustrate that validation: Figure 1—figure supplement 3. We can determine the effect of short glycopolymers on metastasis by comparing the metastatic potential of cells treated with these glycopolymers or merely mock treated with PBS. The negative control in Figure 4 – doxycycline withheld cells – is effectively mock-treated naïve 4TO7 cells, so we constructed this new figure by compiling the data from this control experiment with the one from Figure 1, which uses short glycopolymer-bearing cells. The figure demonstrates that short glycopolymer-bearing cells cause no increase in tumor burden as measured by quantitative bioluminescence imaging (A), lung mass (B), or number of frank mets (C) compared to PBS-treated cells. We have added reference to this new figure in the third paragraph of the subsection “The glycocalyx increases metastatic potential in a size dependent manner”.

2) Figure 1: It is important to rigorously rule out that the bulky glycocalyx in fact increases the propensity of IV injected tumor cells to get stuck in the lung vasculature. Figure 1D shows that the number of macroscopically detectable lung mets is higher in the presence of the long polymer. Is this in any way related to more cells getting trapped? For this, lung tissue can be isolated at early time points post injection (days 1, 2, 3, 4, 5) and the number of trapped tumor cells quantitatively determined by qPCR.

The reviewers bring up an important potential explanation. The experiments in the paper do actually address this question, but perhaps in a way that would not benefit the reader as originally written. The graphs in Figure 1—figure supplement 1 and Figure 4—figure supplement 1 show that we measured bioluminescence of the mice on day 0, just 4 hours after injection when, as seen in Figure 1D, only those cells stuck in the lungs are left after clearing from the rest of the vasculature. By comparing photon flux at this time point, we can quantitatively measure the number of trapped tumor cells (since the mice themselves generate essentially zero background photons). The treatment groups show no significant increase in accumulated tumor cells relative to controls – see below. We have added a sentence explaining and hopefully making clear this finding in the first paragraph of the subsection “The glycocalyx increases metastatic potential in a size dependent manner”.

Author response image 1

Download asset Open asset

3) Figure 1, 4 versus 2, 3: Why do the authors switch from 4TO7 to MCF10A here? The in vivo work is with 4TO7, the in vitro experiments with MCF10A. We agree that MCF10A is a good model but are the same results seen with 4TO7 cells? Are 4TO7 cells indeed not showing proliferation on soft substrates and is FAK, PI3K-AKT signaling, cyclin D accumulation in those cells under those conditions stimulated by the long polymers? Not all experiments, but at least confirmation of a few key findings across these two models would be required to directly relate in vitro work in Figure 2 and 3 to the in vivo experiments in Figure 1 and 4.

To build upon the work in the 2014 Paszek et al.paper, in which we used MCF-10A cells as an in vitro model, we tested the ability of these same cells to seed metastases in our tail-vein-to-lung model. At early time points, we observed a delay in the onset of anchorage-dependent cell death based on glycocalyx size, just as in our in vitro model. Long glycopolymers seem to delay the onset of cell death as shown by the apoptosis marker cleaved caspase-3 (CC3). Being nontransformed cells, however, the MCF-10A cells were incapable of forming frank metastases. Eventually, the cells did die, likely because metastasis formation also relies upon additional factors including activation of an angiogenic response and immune cell evasion – two parameters extending well beyond the scope of the current proposal. This prevented us from being able to assess any effect the glycocalyx may have had on proliferation specifically.

Thus in order to study the effect of glycocalyx bulkiness on G1 cell cycle progression and cell proliferation, we moved to a new model – that of the syngeneic 4TO7/Balbc model. Upon tail vein injection, the cells in this model survive for extended periods of time in the lung. They fail, however, to proliferate and thus have been used extensively as a model of micrometastatic tumor dormancy. This dormancy "prone" tumor cell line, therefore, permitted us to specifically address the contribution of a bulky glycocalyx on cell proliferation and G1 cell cycle transit in vivo.

In our newly revised manuscript, we have outlined the rationale for the use of these different breast epithelial models including their complementary strengths/weaknesses and expanded the discussion of their use in the first paragraph of the Results. To demonstrate, as the reviewers requested, key findings of integrin activation across these two models, in our revised article we now include data from the aforementioned MCF-10A in vivo studies in a new supplementary figure, Figure 1—figure supplement 1. This important experiment not only validates critical in vitro findings in an in vivo setting, but the data also demonstrate the utility of the 4TO7/Balbc dormancy model, which permits a more direct exploration of the impact of the glycocalyx on cell proliferation in the metastatic settingin vivo.

4) How valid is it to compare the signaling events in the in vitro model with those seen in vivo? For example activation of AKT signaling was reported in vitro but not shown in vivo. pAKT IHC should be carried out on lung metastases in the experiments described for the polymer coated cells and also those where MUC1ΔCT-expressing cells were used. Only then can the authors claim that the PI3K/AKT pathway plays a role in the lung environment.

Indeed, the pAkt signaling should be apparent in the in vivo model as well. To address this, we performed IF-IHC on tissue sections from the experiments described in Figures 1 and 4. We have included these data as Figure 2—figure supplement 3 and Figure 4—figure supplement 2 as well as made reference to their inclusion in the fifth paragraph of the subsection “A thick glycocalyx drives cell cycle progression via the PI3K-Akt axis” and in the last paragraph of the subsection “The MUC1 ectodomain is sufficient to increase the metastatic potential of 4TO7 cancer cells”.

5) Figure 2 and 3: Increased pH3 levels (Figure 2) and FAK-pY925 (Figure 3) on soft substrate coated with FN in presence of long polymers is clear. However, increased cyclin D levels (compared to PBS and short polymer in Figure 2E) and increased AKT phosphorylation (compared to PBS control in Figure 2G) are not convincing at all and increased FAK-pY397 is very modest at best (compared to PBS and short polymer in Figure 3A). Western blots are indicated by the authors to be representative of at least 2 (Figure 2) or 3 (Figure 3) experiments. If this is the best example, this data is not convincing. Notably, the effects on pH3 and colony size are clear so the effect on proliferation is well-demonstrated. However, specifically for cyclin D, p-AKT, and p-Y397FAK data, the authors should present WB data where differences are convincing and rigorous quantification of 3 or more biologically independent experiments need to be provided.

As the reviewers have requested, we have quantified all of the western blot data for cyclin D1, pAkt and pFAKY397. The quantitative results confirm our subjective conclusions, and illustrate them in a more objective manner. This quantification was accomplished as follows: densitometry was performed on the blots, absolute intensity was normalized relative to the tubulin loading control and then normalized to the positive control lane (in each case, this was the band from cells loaded on stiff gels). Normalized values from replicate blots were averaged and are plotted in new supplementary figures as follows: cyclin D and pAkt were added as Figure 2—figure supplement 2, and pFAKY397 quantitation was added as Figure 3—figure supplement 1. We also added references to these new figures in the third paragraph of the subsection “A thick glycocalyx drives cell cycle progression via the PI3K-Akt axis” and in the first paragraph of the subsection “A thick glycocalyx stimulates integrin-FAK mechanosignaling”. These data bolster our in vivo findings illustrating the role of these markers in the tumor model.

6) pH3 analysis was carried out on 3-4 tumors per mouse. Does the level of positivity depend on the size of the individual tumors? As this analysis was carried out at a single time point it provides a snapshot of proliferation at that time. It may be that reduced proliferation in the short polymer treated cells reflects a delay in their initial growth (and therefore size of the individual tumors) and not one of reduced proliferation per se. The same argument could be made for the cyclin D1 analysis.

We have accounted for this possible bias in our image analysis. Image analysis software was used to quantify the number of nuclei present in the tumor in each image. The number of ‘positive’ pH3 nuclei – determined by a signal intensity threshold algorithmically in ImageJ – is then divided by the number of total nuclei. So the values are normalized for the number of nuclei present in the image (i.e. the size or density of the tumor) and they, therefore, do not affect our quantification. We have added clarification to this effect in the Materials and methods subsection “Quantification of immunofluorescence”.

7) Figure 4: were the mice fed doxycline? What are the levels of MUC1ΔCT at the time of sacrifice? If the expression is not maintained for the duration of the experiment it could again suggest that the effects of MUC1ΔCT may be mediated at very early stages of tumor cell arrest and extravasation in the lung.

The reviewers are correct that the mice are not fed doxycycline, and we have updated the Materials and methods section to make this more clear. The rational was that the glycopolymers, which emulate the effects of mucins such as MUC1, cannot be continuously administered, and so they also dilute out over time (and proliferative generations). While the effects of the MUC1 ectodomain may indeed be mediated at early stages of metastasis, the in vitro experiments demonstrating a proliferative phenotype were performed at these early time points with this in mind.

8) Cells were "painted.… with glycopolymers" but the methods used are not apparent within the manuscript. Details sufficient for reproducing this aspect of the experimental approach would be welcome.

The authors apologize for not making this clear in the original manuscript, since the glycopolymers play such an integral role to our experimentation. Their use was detailed in individual experimental Materials and methods sections, but we have added a new subsection to the Materials and methods entitled ‘Glycopolymer loading onto cell surfaces’ with the inclusion of reference to a paper describing both the synthesis and characterization of these glycopolymers. We hope this clarifies the procedure for readers.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Essential revisions:

"Point 3) Figure 1, 4 versus 2, 3: Why do the authors switch from 4TO7 to MCF10A here? The in vivo work is with 4TO7, the in vitro experiments with MCF10A. We agree that MCF10A is a good model but are the same results seen with 4TO7 cells? Are 4TO7 cells indeed not showing proliferation on soft substrates and is FAK, PI3K-AKT signaling, cyclin D accumulation in those cells under those conditions stimulated by the long polymers? Not all experiments, but at least confirmation of a few key findings across these two models would be required to directly relate in vitro work in Figure 2 and 3 to the in vivo experiments in Figure 1 and 4."

To their credit, the authors have explained why MCF10A cells were not used for the metastasis in vivo studies and have added data that show the limitations in this regard. However, it was not clear to the reviewers why 4T07 cells were not studied 'in vitro' in terms of cell signaling. It would seem that these cells can be analyzed in culture since the authors discuss growing them to sub-confluency. We thus ask the authors to investigate the FAK, PI3K-AKT, and cyclin D status in 4T07 as this would be important to demonstrate the generality and robustness of the main findings.

Indeed, 4TO7 cells can be analyzed in culture, and so we performed the experiments as requested above. 4TO7 cells were cultured, painted with long or short glycopolymers, then plated on soft (400 Pa) polyacrylamide gels with or without EGF challenge before being analyzed by immunoblotting. We found that, like MCF-10A cells, 4TO7 cells demonstrated increased FAK phosphorylation and increased Akt activation when long glycopolymers were used to bolster their glycocalyces when compared to short glycopolymers. Cyclin D1 expression levels, however, did not change in vitro. We posit that this is due to 4TO7 being a transformed cell line which gains only a marginal increase in cell-cycle driving forces with an increased glycocalyx in the rich environment of cell culture. This is demonstrated by our negative control conditions: cyclin D1 expression in PBS treated cells on soft gels. MCF-10A cells in these conditions demonstrate very little cyclin D1 expression, whereas 4TO7 cells show robust levels of cyclin D1 expression, even on soft gels with mock treatment (Figure 2—figure supplement 5). In addition, the nontransformed cell line MCF-10A does benefit from the increase in glycocalyx size in vitro, and our analysis of 4TO7 metastases in vivo demonstrates a glycocalyx length-dependency on cyclin D1 expression. For all of these results, we repeated the experiments at least thrice and quantified the results along with calculated p values. The data have been added as a combination of new panels in Figures 2 and 3 as well as the new supplementary figure: Figure 2—figure supplement 5. Furthermore, the manuscript has been updated with references to these data in the subsection “A thick glycocalyx drives cell cycle progression via the PI3K-Akt axis”.

"Point 5) However, specifically for cyclin D, p-AKT, and p-Y397FAK data, the authors should present WB data where differences are convincing and rigorous quantification of 3 or more biologically independent experiments need to be provided."

We appreciate that the authors have now added densitometry and error bars to the western blot results in the revised manuscript. However, given the limitations and challenges of western blots, it is desirable to perform at least three replicates and calculate p values to determine the reliability and significance of the findings. We hope that the authors can therefore add at least one additional replicate to the experiments on cyclin D, pAKT and pY397FAK to ensure the robustness of the results.

We have repeated these key results as requested so that the data reflect a minimum of three biological replicate experiments (n = 4 for pAkt and n = 5 for pFAK) for these three markers and have included calculated p values for each of the desitometric quantifications. Figures 2 and 3 and the text in the subsection “A thick glycocalyx drives cell cycle progression via the PI3K-Akt axis” has been updated to reflect these added data in this revision.

Author details

1. Elliot C Woods

   Department of Chemistry, Stanford University, California, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2920-8334

2. FuiBoon Kai

   Center for Bioengineering and Tissue Regeneration, Department of Surgery, University of California, San Francisco, San Francisco, United States

   Contribution

   Data curation, Validation, Investigation, Methodology

   Competing interests

   No competing interests declared

3. J Matthew Barnes

   Center for Bioengineering and Tissue Regeneration, Department of Surgery, University of California, San Francisco, San Francisco, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

4. Kayvon Pedram

   Department of Chemistry, Stanford University, California, United States

   Contribution

   Data curation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8365-1826

5. Michael W Pickup

   Center for Bioengineering and Tissue Regeneration, Department of Surgery, University of California, San Francisco, San Francisco, United States

   Contribution

   Data curation, Validation, Investigation, Methodology

   Competing interests

   No competing interests declared

6. Michael J Hollander

   Department of Chemistry, Stanford University, California, United States

   Contribution

   Data curation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

7. Valerie M Weaver

   1. Center for Bioengineering and Tissue Regeneration, Department of Surgery, University of California, San Francisco, San Francisco, United States
   2. Department of Anatomy, University of California, San Francisco, San Francisco, United States
   3. Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States
   4. Department of Radiation Oncology, University of California, San Francisco, San Francisco, United States
   5. Eli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research, University of California, San Francisco, San Francisco, United States
   6. UCSF Helen Diller Comprehensive Cancer Center, University of California, San Francisco, San Francisco, United States

   Contribution

   Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

8. Carolyn R Bertozzi

   1. Department of Chemistry, Stanford University, California, United States
   2. Howard Hughes Medical Institute, Stanford University, California, United States

   Contribution

   Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   bertozzi@stanford.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4482-2754

• 3,243

  Page views

• 545

  Downloads

• 57

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.