Fig 1.
LaNt α31 is upregulated in invasive ductal carcinoma.
Serial sections from formalin-fixed paraffin-embedded human breast tissue processed for immunohistochemistry with mouse monoclonal antibodies against laminin α3, LaNt α31, or mouse IgG- isotype control, (A) uninvolved breast tissue, (B) invasive ductal carcinoma, and (C) ER-, PR-, Her- invasive ductal carcinoma. Dashed boxed regions are shown at higher magnification in columns to the right. Arrowheads indicate regions of anti-LaNt α31 immunoreactivity not recognised by anti-laminin α3. Scale bars: 500 μm.
Fig 2.
LaNt α31 is upregulated in ductal carcinoma and in lymph node metastases.
Formalin-fixed paraffin-embedded human breast tissue microarray sections processed for immunohistochemistry with mouse monoclonal antibodies against LaNt α31. Two separate arrays were used; (A) uninvolved with paired invasive/ in situ ductal carcinoma tissues (N = 25), (B) invasive ductal carcinoma with paired node metastases (N = 29). Cores were scored as either decreased, no change, or increased staining intensity relative to the paired uninvolved (A) or primary tumour (B) core from the same donor (representative images shown). Stacked columns of percentage of cases in each category were plotted and Wilcoxon signed ranks test used to describe observed relationship. Scale bars: 500 μm.
Table 1.
Patient ages for each study cohort.
Fig 3.
LaNt α31 upregulation in invasive ductal carcinoma does not correlate with nodal involvement or tumour grade.
Formalin-fixed paraffin embedded human breast tissue microarray sections processed for immunohistochemistry with mouse monoclonal antibodies against LaNt α31. Cores were scored based on LaNt α31 staining intensity from 0–3. Scores of 0 or 1 were combined and designated as low LaNt α31 expression, score 2 as medium expression, and 3 as high expression. (A) Representative example of core scoring. (B-D) Stacked column graphs of percentage of cases with each staining intensity segregated by tumour grade I, II, or III (B), Ki67 expression (C), or by nodal involvement (D). Somers’ D was used to describe observed relationships between LaNt α31 staining intensity and the independent variables. (E) Kaplan–Meier survival curve, where LaNt α31 staining intensity was simplified to low or high by pooling medium and high cores. Logrank test was used to determine hazard ratio and chi square for significance. Scale bar in (A): 300 μm.
Fig 4.
Positive correlation between LaNt α31 staining intensity and EGFR expression.
Formalin-fixed paraffin embedded human breast tissue microarray sections processed for immunohistochemistry with mouse monoclonal antibodies against LaNt α31. Cores were scored based on LaNt α31 staining intensity from 0–3. Scores of 0 or 1 were combined and designated as low LaNt α31 expression (light grey bars), scores 2 as medium expression (dark grey bars), and 3 as high expression (black bars). Stacked column graph of percentage of cases that fall into group after segregation based on pathologist provided grading of immunohistochemistry markers; (A) EGFR, (B) Her2, (C) ER, (D) PR, or (E) ER- PR- Her2- cases. Somers’ D was used to describe observed relationship between LaNt α31 staining intensity and independent variable.
Fig 5.
LaNt α31 overexpression reduces proliferation rates.
MCF-7 or MDA-MB-231 cells, either left untreated, treated with 20 ng mL-1 nocodazole, or transduced with increasing doses of eGFP (+eGFP) or LAMA3LN1-eGFP (+LaNt α31-eGFP) adenoviral particles, were cultured for 96 h following transduction and replating. (A) Immunoblots from total cell lysates for MCF-7 or MDA-MB-231 cells taken after 24 h were probed with antibodies against eGFP, with ponceau S total protein-stained membrane shown below. (B) Hoechst 33342 was added to the culture media, and the cell nuclei imaged after 20 min. Each dot represents an experimental repeat consisting of the mean of 3 fields of view per well for 3 technical replicates.
Fig 6.
LaNt α31 overexpression does not significantly affect 2D migration or 3D invasion of MCF-7 cells.
MCF-7 cells were either left untreated or transduced with eGFP (+eGFP), or LAMA3LN1-eGFP (+LaNt α31-eGFP). For gap closure assays, 24 h after transduction, cells were seeded into ibidi® 2-well culture inserts and allowed to attach for 6 h, the inserts were then removed, and the gap margin imaged at 0 h and 16 h. For single cell migration assays, 24 h after transduction, cells were seeded onto tissue culture plastic and the migration paths of individual cells tracked over a four-hour period. (A) Representative images from immediately after removing chamber (T0 upper panels) and after 16 h (T16 lower panels), yellow lines delineate wound margins. (B) Gap closure was measured as a percentage relative to starting gap area. (C) Vector diagrams showing representative migration paths of 10 individual cells with each colour representing a single cell. (D) Migration speed was measured as total distance migrated over time. Each point on the associated dot plots represents an independent experiment with 2–3 technical replicates per experiment for gap closures assays or 20–40 cells per low density migration assay. For invasion assays, cells were plated onto the outside of a transwell membrane. 10 ng mL-1 epidermal growth factor was used to stimulate invasion through the membrane and into collagen I or Matrigel. After 48 h, the cells were fixed and stained with DAPI then imaged at 5 μm intervals using a spinning disk confocal microscope. (E and G) Representative images of invasion into collagen I or Matrigel from 10–40 μm presented at equal intervals. (F and H) Absolute invasion depth was measured where cell count ≥1. Treatment with GM6001 MMP inhibitor was included as an invasion inhibiting control. Each point on the graphs represents an independent experiment, with 2–3 technical replicates per assay. Statistical tests of differences relative to controls were performed using one-way rANOVA followed by Bonferroni’s post hoc analyses; p>0.05 in all comparisons. Scale bar in (a) represents 100 μm.
Fig 7.
LaNt α31 overexpression does not significantly affect 2D migration MDA-MB-231 cells.
MDA-MB-231 cells were either left untreated or transduced with eGFP (+eGFP), or LAMA3LN1-eGFP (+LaNt α31-eGFP). For gap closure assays, 24 h after transduction, cells were seeded into ibidi® 2-well culture inserts and allowed to attach for 6 h, the inserts were then removed, and the gap margin imaged at 0 h and 16 h. For single cell migration assays, 24 h after transduction, cells were seeded onto tissue culture plastic and the migration paths of individual cells tracked over a four-hour period. (A) Representative images from immediately after removing chamber (T0 upper panels) and after 16 h (T16 lower panels), yellow lines delineate wound margins. (B) Gap closure was measured as a percentage relative to starting gap area. (C) Vector diagrams showing representative migration paths of 10 individual cells with each colour representing a single cell. (D) Migration speed was measured as total distance migrated over time. Each point on the associated dot plots represents an independent experiment with 2–3 technical replicates per experiment for gap closures assays or 20–40 cells per low density migration assay. Statistical tests of differences relative to controls were performed using one-way rANOVA followed by Bonferroni’s post hoc tests; * p<0.05. Scale bar in (a) represents 100 μm.
Fig 8.
LaNt α31 overexpression causes a small reduction in invasion of MDA-MB-231 cells into Matrigel.
MDA-MB-231 cells, either left untreated or transduced with eGFP (+eGFP), or LAMA3LN1-eGFP (+LaNt α31-eGFP), were plated onto the outside of a transwell membrane. 10 ng mL-1 epidermal growth factor was used to stimulate invasion through the membrane into collagen I (A-B) or Matrigel (C-D). After 48 h, the cells were fixed and DAPI stained then imaged at 5 μm intervals using a spinning disk confocal microscope. (A) and (C) Representative images from 10–40 μm depth presented at equal intervals, with an additional slice at 80 μm in (C). Absolute invasion depth was measured where cell count ≥1. Treatment with GM6001 MMP inhibitor was included as an invasion inhibiting control. Each point on the graphs in (B) and (D) represents an independent experiment, with 2–3 technical replicates per assay. * represents p<0.05 between bracketed groups as determined by one-way ANOVA followed by Bonferroni’s post hoc analyses.
Fig 9.
LaNt α31 overexpression causes a change in mode of invasion of MDA-MB-231 cells into Matrigel.
(A) Maximum intensity projection of planes from 20–60 μm from the same assays in Fig 8C. (B) Image analyses method for determining entropy as a measure of cell clustering/cohesiveness; each stack of images was processed using an automated processing algorithm, where cell count and entropy score after a threshold was measured for each image in the stack. (C) Entropy score versus depth graph with points representing mean and SD from 3 independent experiments. Shaded regions indicate where comparisons lack value due to either high cell counts (0–45 μm) or differences in cell numbers between conditions (>70 μm). * denote statistically significant differences between +LaNt α31-GFP cells and both MDA-MB-231 and +GFP conditions by 2-way ANOVA with Tukey’s post hoc test.
Fig 10.
High LaNt α31 expression is associated with low tumour cohesion in invasive ductal carcinoma.
Formalin-fixed paraffin embedded human breast tissue microarray sections processed for immunohistochemistry with mouse monoclonal antibodies against LaNt α31. Tumour cohesion was graded as either cohesive (tight tumour islands), or non-cohesive (chord-like) in tumour cores scored as having high LaNt α31 expression. (A) Representative example of core grading. (B) Stacked column graphs of percentage of cases that are either cohesive or non-cohesive. Scale bars: 300 μm.