Insights into metabolic heterogeneity of colorectal cancer gained from fluorescence lifetime imaging

  1. Anastasia D Komarova
  2. Snezhana D Sinyushkina
  3. Ilia D Shchechkin
  4. Irina N Druzhkova
  5. Sofia A Smirnova
  6. Vitaliy M Terekhov
  7. Artem M Mozherov
  8. Nadezhda I Ignatova
  9. Elena E Nikonova
  10. Evgeny A Shirshin
  11. Liubov E Shimolina
  12. Sergey V Gamayunov
  13. Vladislav I Shcheslavskiy
  14. Marina V Shirmanova  Is a corresponding author
  1. Institute of Experimental Oncology and Biomedical Technologies, Privolzhsky Research Medical University, Russian Federation
  2. Institute of Biology and Biomedicine, Lobachevsky State University of Nizhny Novgorod, Russian Federation
  3. Nizhny Novgorod Regional Oncologic Hospital, Russian Federation
  4. Laboratory of Clinical Biophotonics, Sechenov First Moscow State Medical University, Russian Federation
  5. Faculty of Physics, Lomonosov Moscow State University, Russian Federation
  6. Becker&Hickl GmbH, Germany
4 figures, 3 tables and 3 additional files

Figures

Figure 1 with 2 supplements
FLIM of NAD(P)H in monolayer cell cultures.

(A) Representative FLIM images of colorectal cancer cell lines. Scale bar = 50 μm. For FLIM: ex. 750 nm, reg. 450–490 nm. (B) The relative contribution of free NAD(P)H (a1, %) in cell cultures. Box shows the median and the quartiles Q1 and Q3, whiskers show minimum and maximum. Dots indicate individual cells (n=280 for HT29 cells, n=185 for HCT116 cells, n=146 for CaCo2 cells, n=138 for CT26 cells). p-values are shown in Supplementary file 1. (C) The distribution of the NAD(P)H-a1 for the cell lines. The bimodality index (BI-a1) is shown on each diagram.

Figure 1—figure supplement 1
FLIM of NAD(P)H in monolayer cell cultures.

The distribution of the NAD(P)H-τm for the cell lines. The bimodality index (BI-τm) is shown on each diagram.

Figure 1—figure supplement 2
Autofluorescence of cofactors FAD and NAD(P)H in cultured cells HT29, HCT116, CaCo2 and CT26.

Representative fluorescence intensity images of flavins (FAD, ex. 900 nm, em. 500–550 nm) and NAD(P)H (ex. 750 nm, em. 450–490 nm). Scale bar = 50 μm.

Figure 2 with 2 supplements
FLIM of NAD(P)H in mouse tumors in vivo.

(A) FLIM images of NAD(P)H of tumor cells in mouse models in vivo. Scale bar = 50 μm. For FLIM: ex. 750 nm, reg. 450–490 nm. (B) Representative histological slices of tumors, hematoxylin/eosin (HE) staining, initial magnification 20×. Scale bar = 50 μm. (C) The relative contribution of free NAD(P)H (a1, %) in three representative tumors (numbered 1–3) obtained from different cell lines. Box shows the median and the quartiles Q1 and Q3, whiskers show minimum and maximum. Dots indicate individual cells (n=280 for HT29, n=340 for HCT116, n=160 for CaCo2, n=350 for CT26). p-values are shown in Supplementary file 1. (D) Representative distributions of the NAD(P)H-a1 for each type of tumor. The bimodality index (BI-a1) is shown on the diagrams.

Figure 2—figure supplement 1
Immunohistochemical analysis of the expression of EpCAM (green, epithelial cells marker) and vimentin (red, mesenchymal cells marker) in mouse tumors.

(A) Immunofluorescence images of HT29, HCT116, CaCo2 and CT26 mouse tumors. Scale bar = 100 μm. (B) The ratio of EpCAM-positive to vimentin-positive areas in tumor sections.

Figure 2—figure supplement 2
Autofluorescence of cofactors FAD and NAD(P)H in HT29 and HCT116 tumor xenografts in vivo.

Representative fluorescence intensity images of flavins (FAD, ex. 900 nm, em. 500–550 nm) and NAD(P)H (ex. 750 nm, em. 450–490 nm). Scale bar = 50 μm.

Figure 3 with 3 supplements
FLIM of NAD(P)H in patients’ tumor samples ex vivo.

(A) Representative FLIM images of patient tumors. Scale bar = 50 μm. For FLIM: ex. 750 nm, reg. 450–490 nm. (B) Histopathology of tumors, hematoxylin/eosin (HE) staining, initial magnification 20×. Scale bar = 50 μm. (C) The relative contribution of free NAD(P)H (a1, %) in patients’ tumors (numbered 1–29). Box shows the median and the quartiles Q1 and Q3, whiskers show minimum and maximum. Dots are the measurements from the individual cells. (D) Representative distributions of the NAD(P)H-a1 for patients’ tumors. The bimodality index (BI-a1) is shown on the diagrams.

Figure 3—source data 1

The dataset (NAD(P)H-a1 values) used to plot the charts shown in Figure 3C.

https://cdn.elifesciences.org/articles/94438/elife-94438-fig3-data1-v1.xlsx
Figure 3—figure supplement 1
Immunohistochemical analysis of the expression of EpCAM (green, epithelial cells marker) and vimentin (red, mesenchymal cells marker) in patients’ tumors.

(A) Immunofluorescence images of patients’ tumors (numbered from 6 to 21). Scale bar = 100 μm. (B) The ratio of EpCAM-positive to vimentin-positive areas in tumor sections. (C) The bimodality index (BI-a1) and dispersion (D-a1) plotted against the EpCAM/vimentin ratio in patient tumor samples. Pearson correlation r is shown on the plots.

Figure 3—figure supplement 2
Immunohistochemical analysis of the expression of GLUT3 and LDHA in patients’ tumors.

(A) Representative immunohistochemical images of GLUT3 expression. Scale bar = 50 μm (magnification x200) and 20 μm (magnification x630). (B) Representative immunohistochemical images of LDHA expression. Scale bar = 50 μm (magnification x200) and 20 μm (magnification x630). (С) Semi-quantitative evaluation of the expression level by staining intensity.

Figure 3—figure supplement 3
Autofluorescence of cofactors FAD and NAD(P)H in patient tumor ex vivo (№ 11).

(A) Representative fluorescence intensity images of flavins (FAD, ex. 900 nm, em. 500–550 nm) and NAD(P)H (ex. 750 nm, em. 450–490 nm). Scale bar = 50 μm. (B) The number of photons per pixel recorded by FLIM for flavins in ex vivo patient samples. Mean ± SD, n=7 patients.

Figure 4 with 1 supplement
The relationships between metabolic heterogeneity and clinicopathological characteristics of patients’ tumors.

(A) Plots of SHAP analysis for the built decision tree models to determine the importance of dispersion (D) and bimodality index (BI) of the fluorescence decay parameters of NAD(P)H. The higher the value of the variable, the more red the dot is. (B) Box-plots of D-a1 with highest significance, * p-val <0.05.

Figure 4—figure supplement 1
The relationships between parameters BI-τm (A) and BI-a1 (B) and clinicopathological characteristics of patients’ tumors.

No significant differences were found between the groups.

Tables

Table 1
NAD(P)H fluorescence decay parameters of colorectal cancer cells in monolayer cultures in vitro and in mouse tumors in vivo.
Cell lineτm, nsτ1, nsτ2, nsa1, %BI-τm
Cell lines in vitro
HT290.80±0.050.39±0.032.47±0.1780.11±1.341.05
HCT1160.71±0.050.40±0.022.39±0.2084.13±1.580.90
CaCo20.95±0.100.38±0.052.53±0.2473.48±2.261.30
CT260.57±0.060.35±0.032.03±0.2286.48±1.281.28
Tumors in vivo
HT290.84 [0.81;0.90]0.46 [0.42;0.48]2.66 [2.54;2.76]81.54 [79.93;83.13]0.85±0.35
HCT1160.88 [0.85;0.92]0.47 [0.45;0.48]2.66 [2.56;2.77]80.61 [79.32;81.96]0.91±0.28
CaCo21.02 [0.86;1.19]0.42 [0.38;0.50]2.91 [2.61;3.32]76.89 [74.06;78.52]1.51±0.71
CT260.72 [0.67;0.78]0.39 [0.37;0.41]2.34 [2.21;2.50]82.12 [80.96;84.70]1.20±0.36
  1. τm – mean lifetime, τ1 – short lifetime component, τ2 – long lifetime component, a1 – relative contribution of the short lifetime component, BI-τm – bimodality index of the mean lifetime.

Table 2
The bimodality index BI-a1 and dispersion D-a1 of NAD(P)H in cultured cells, mouse tumors and patients’ tumor samples.
Cell lines in vitro
HT29HCT116CaCo2CT26
BI-a10.160.920.490.63
D-a11.832.013.411.67
Tumors in vivo
HT29HCT116CaCo2CT26
BI-a11.2±0.320.93±0.230.86±0.151.06±0.54
D-a13.202.604.002.81
Patients’ tumors ex vivo
Sample12345678910
BI-a11.250.811.011.161.011.000.860.701.281.25
D-a17.774.188.6811.996.937.839.338.2511.516.96
Sample11121314151617181920
BI-a10.591.451.341.050.890.241.291.320.800.94
D-a14.52.374.942.194.163.346.105.412.235.19
Sample212223242526272829
BI-a11.280.981.531.310.881.581.650.51.29
D-a15.273.296.545.774.875.834.145.676.36
Table 3
Information about patients and their colorectal tumors.
CharacteristicsNumberPercent
Gender
Male1655.17
Female1344.83
Age
Mean ±SD65.28±11.92 y.o.-
Median67 y.o.-
Tumor staging
I310.34
IIA620.69
IIB310.34
IIIB1137.93
IIIC13.45
IV517.25
Tumor site
Cecum colon26.90
Transverse colon1034.48
Hepatic flexure310.34
Sigmoid colon827.59
Rectum620.69
Grade
Low (G1)413.79
Moderate (G2)1965.52
High (G3)620.69

Additional files

Supplementary file 1

Statistical significance of the differences of NAD(P)H a1-% between different cell lines and tumors (p-values).

https://cdn.elifesciences.org/articles/94438/elife-94438-supp1-v1.docx
Supplementary file 2

Clinicopathological characteristics and NAD(P)H fluorescence decay parameters of patients’ tumors.

https://cdn.elifesciences.org/articles/94438/elife-94438-supp2-v1.xlsx
MDAR checklist
https://cdn.elifesciences.org/articles/94438/elife-94438-mdarchecklist1-v1.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Anastasia D Komarova
  2. Snezhana D Sinyushkina
  3. Ilia D Shchechkin
  4. Irina N Druzhkova
  5. Sofia A Smirnova
  6. Vitaliy M Terekhov
  7. Artem M Mozherov
  8. Nadezhda I Ignatova
  9. Elena E Nikonova
  10. Evgeny A Shirshin
  11. Liubov E Shimolina
  12. Sergey V Gamayunov
  13. Vladislav I Shcheslavskiy
  14. Marina V Shirmanova
(2024)
Insights into metabolic heterogeneity of colorectal cancer gained from fluorescence lifetime imaging
eLife 13:RP94438.
https://doi.org/10.7554/eLife.94438.3