DIAGNOSTIC BIOMARKERS FOR COLORECTAL CANCER

1
DIAGNOSTIC BIOMARKERS FOR COLORECTAL CANCER

• Identifying Protein based Serum Plasma
  Biomarkers using a Proteomics-based Approach

Ms. Aisha Q. Butt Prof. Martin Clynes Dr. Paul
Dowling National Institute for Cellular
Biotechnology (NICB)
2
Objectives

• Colorectal Cancer Facts Diagnostic Limitation
• Hypothesis Aims
• Research Strategy
• Biomarker Discovery
• Cell Culture
• Nano-HPLC Mass Spectrometry
• Bioinformatics
• Biomarker Verification
• ELISA
• Immunohistochemistry
• Conclusion Future Work

3
Colorectal Cancer Facts Figures

• 2nd most common cause of death in Republic of
  Ireland.
• Approximately, 2000 new CRC cases diagnosed each
  year.
• Ireland Highest bowel cancer incidence, lower
  survival rate and higher mortality rates in
  Western Europe both in men and women.

• 1 Million new cases 1/2 million deaths p/yr
• 3rd most common malignancy in world.
• Primary Cancer mortality cause in USA, Europe
  Korea

• Prediction By 2020, new CRC cases projected to ?
  by 79 in men 56 in women.

4

• CRC Diagnosing Limitations

• Our Hypothesis Aims

• CRC highest cure rates if detected early.
• Colonoscopy
• Gold standard for diagnosis of colonic neoplasia
• But its not a perfect test
• Lengthy waiting lists for colonoscopy services in
  Ireland
• Lack of sensitivity low compliance Most of
  current diagnostics detect cancer after it has
  already spread to other parts of the body

• Devise screening programme based on
• Panel of Protein Biomarkers
• Used in combination with Colonoscopy
• To offer most accurate test for early diagnosis
  of CRC
• Novel high value biomarker diagnostics test kit
  based on blood or serum could
• Prescreen high risk groups
• Eliminate waiting lists
• Permit rapid access to suspects

5
Research Strategy

• Cancer Secretome comprising all proteins
  released by tumour cell Attracting much
  attention recently.
• Our Approach Collection of Conditioned Media
  from 4 CRC cell lines HCT116, HT-29, SW480 WiDr
  using Cell Culturing Techniques.
• Proteins identified from conditioned media using
  Mass Spectrometry evaluated using specific
  criteria to identify most likely candidate for
  further investigations.
• Criteria Being detectable in all cell lines
• Likelihood that proteins are shed/secreted by
  tumour cells
• Proteins primary localisation molecular
  function within the cell
• Literature search
• Bioinformatics allowed data sets to be cross
  compared from the shortlisted data, few
  proteins were verified using
• ELISA kit
• Immunohistochemistry

6
Biomarker Discovery

• Cell Culture
• Four CRC cell lines cultured in respective media-
  HCT116, HT-29, SW480 WiDr
• Assessment of Cell Viability performed on derived
  Conditioned Media
• Protein Concentration
• Proteins in CMs centrifuged down in CENTRICONS
  (Sartorius Stedim Biotech)
• Concentrated proteins by 40 fold (4mls to 100µl)
• Protein Precipitation Clean Up
• Using Ready PrepTM 2-D Clean-Up Kit protein
  samples subjected to precipitation and cleaning
  using buffers and reagents yielding 30µl sample
  for each cell line
• In Solution Digestion / Double Digestion
• Double Digestion carried out using LysC Trypsin
  for denaturation digestion of proteins into
  peptides.
• Sample Clean Up
• Using PepClean TM C-18 Resin Spin Columns,
  tryptic peptide mixtures were further cleaned up
  for removal of any solvents and buffers.
• Sample speed vacuumed down and resuspended in
  Mass Spec. compatible buffer.

7
LC-MS Bioinformatics

• Ettan MDLC system (GE Healthcare, Piscataway, NJ)
  - applied for desalting and separation of LysC
  and tryptic peptide mixtures.
• Two experiments set up
• 10µl tryptic peptide sample loaded onto LC column
  with 1hour gradient time
• 20µl tryptic peptide sample loaded onto LC column
  with 3hour gradient time

• MDLC-MS results analysed using Bioworks Browser
  software suite TM (Thermo Fisher Scientific, USA)
• Proteins in samples searched against Swiss Prot
  human protein database (Dec, 2009 65,533
  entries 20,339 human proteins).
• The stringent protein identification criteria
  based on multiple filters
• Distinct peptides
• Delta CN (0.040)
• Xcorr vs. Charge state (1.50, 2.20, 2.50, and
  3.00)
• Peptide Probability (0.05)
• Number of distinct peptides (2)

8
Results Discussion

• Conditioned Media Cell Viability
• Mean Average 89.875 Viable Cells 10.125 Dead
  Cells
• Cell counts performed using haemocytometer and
  trypan blue dye on CM derived from 4 CRC cell
  lines using NICB-SOP-003-01.

9
Mass Spectrometry Sample Gradient Time

• An approximate gt2.5 fold increase in the total
  proteins identified on MS using 1hr vs. 3hr
  sample gradient time.
• Stringent Multiple filters applied Distinct
  Peptides, Delta CN (0.040), Xcorr vs. Charge
  state (1.50, 2.20, 2.50, and 3.00), Peptide
  Probability (0.05) and Number of distinct
  peptides (2).
• Data Cross compared using Bioinformatics Excel
  tools
• 92 common proteins identified within 4 CRC cell
  lines.

Cell Lines Quantity-Gradient Total Proteins Filtered Proteins
HCT 116 10ul - 1Hour 2076 148
  20ul - 3Hour 8120 603
HT 29 10ul - 1Hour 1851 75
  20ul - 3Hour 3889 185
SW-480 10ul - 1Hour 1953 119
  20ul - 3Hour 3949 255
WiDr 10ul - 1Hour 2086 200
  20ul - 3Hour 5162 384
10
Cellular Localisation Biological Process

• 92 Common proteins further analysed classified
  using Human Protein Reference Database on basis
  of
• Cellular Localisation
• Biological Processes
• Motifs
• 51 Proteins localization is based within
  cytoplasm and 23 proteins are associated with
  cell growth maintenance.

• 92 Common proteins further analysed classified
  using Human Protein Reference Database on basis
  of
• Cellular Localisation
• Biological Processes
• Motifs
• 51 Proteins localization is based within
  cytoplasm and 23 proteins are associated with
  cell growth maintenance.

11
Protein Motifs

• Proteins secreted/shed by the tumour cells (and
  so will be detectable in the circulatory system)
  an important criterion to identify the most
  likely candidate biomarkers for further
  investigations.
• Proteins having signal peptide on their motifs
  are secretory proteins, whereas all other
  proteins are shed by the tumour cells.

12
Biomarker Verification

• Stringent classification of 92 common proteins on
  the basis of
• Proteins Cellular Localisation
• Biological Process
• Proteins Motifs
• Allowed to choose 2 proteins
• Protein X
• Protein Y
• Why these 2 Proteins?
• Presence of signal peptide motif on Protein X
• Relevance and association of Protein Y with other
  cancer types (Literature Search)

13
ELISA Test Serum Plasma Samples

• Protein X levels in serum plasma samples
  measured using double-antibody sandwich ELISA
  system (Bender MedSystems, Austria)
• Serum samples 16 Advanced stage 8 Healthy
  controls
• Plasma samples 9 Early stage 8 Healthy
  controls

(Early Stage )
14
Receiver Operating Characteristics ROC Curve

• ROC curve analysis of Protein X for
  discriminating CRC patients from healthy controls
  for advanced stage serum (AUC - 0.825) and early
  stage plasma samples (AUC - 0.639).

Protein X Diagnostic efficacy in Serum samples
Protein X Diagnostic efficacy in Plasma samples
15
Immunohistochemistry CRC Tissue Sections

• CRC tissue sections compared to normal colon
  tissues via Protein Y expression using
  Immunohistochemical staining.
• Protein Y positive expression observed in a
  moderately differentiated adenocarcinoma
  (intermediate differential grade)
• Protein Y positive staining in the cytoplasm -
  Granular in nature
• A lot less nuclear staining in areas of tumour
  compared to normal tissue

16
Conclusion

• This pilot study has shown that cancer secretome
  from the tumour cells presents a promising
  reservoir of biomarkers with soluble-secreted
  proteins and shed membrane proteins.
• And, the use of these secreted proteins to go
  back on clinical serum or plasma samples to
  distinguish patients with or without CRC is a
  promising diagnostic approach.
• Also, both Protein X and Protein Y could serve as
  potential biomarkers if used in combination with
  few other biomarkers for the early diagnosis of
  colorectal cancer.

17
Further Work...

• Further work and efforts are required to fully
  validate the biomarkers detected in this study
  with
• Large sized clinical samples
• Multiple medical centre samples
• ELISA Test for Protein Y protein on serum
  plasma samples
• These biomarkers detected can be used in
  combination with a panel of other protein
  biomarkers and colonoscopy to improve the overall
  accuracy and speed for detecting CRC at an early
  stage and prioritise high-risk individuals.

18
Summary

• Colorectal Cancer Diagnostic Limitations
• Aim Protein based Biomarkers for early detection
  of CRC
• Biomarker Discovery Research Strategy
• Cell Culture
• Nano-HPLC Mass Spectrometry
• Bioinformatics
• Biomarker Verification
• ELISA
• Immunohistochemistry
• Conclusion Future Work

19
Thank You !