Article Text

Exploiting altered patterns of choline kinase-alpha expression on human prostate tissue to prognosticate prostate cancer
  1. Amarnath Challapalli1,
  2. Sebastian Trousil1,
  3. Steve Hazell2,
  4. Kasia Kozlowski1,
  5. Mihir Gudi2,
  6. Eric O Aboagye1,
  7. Stephen Mangar1
  1. 1Department of Surgery and Cancer, Imperial College London, London, UK
  2. 2Department of Pathology, Imperial College London/ Imperial College Healthcare NHS Trust, London, UK
  1. Correspondence to Professor Eric O Aboagye, Department of Surgery and Cancer, Imperial College London, Hammersmith Hospital, Du Cane Road, London W12 0NN, UK; eric.aboagye{at}


Aims Malignant transformation results in overexpression of choline-kinase (CHK) and altered choline metabolism, which is potentially detectable by immunohistochemistry (IHC). We investigated the utility of CHK-alpha (CHKA) IHC as a complement to current diagnostic investigation of prostate cancer by analysing expression patterns in normal (no evidence of malignancy) and malignant human prostate tissue samples.

Methods As an initial validation, paraffin-embedded prostatectomy specimen blocks with both normal and malignant prostate tissue were analysed for CHKA protein and mRNA expression by western blot and quantitative reverse transcriptase PCR (qRT-PCR), respectively. Subsequently, 100 paraffin-embedded malignant prostate tumour and 25 normal prostate cores were stained for both Ki67 (labelling-index: LI) and CHKA expression.

Results The validity of CHKA-antibody was verified using CHKA-transfected cells and siRNA knockdown. Immunoblotting of tissues showed good resolution of CHKA protein in malignant prostate, verifying use of the antibody for IHC. There was minimal qRT-PCR detectable CHKA mRNA in normal tissue, and conversely high expression in malignant prostate tissues. IHC of normal prostate cores showed mild (intensity) CHKA expression in only 28% (7/25) of samples with no Ki67 expression. In contrast, CHKA was expressed in all malignant prostate cores along with characteristically low proliferation (median 2% Ki67-LI; range 1–17%). Stratification of survival according to CHK intensity showed a trend towards lower progression-free survival with CHK score of 3.

Conclusions Increased expression of CHKA, detectable by IHC, is seen in malignant lesions. This relatively simple cost-effective technique (IHC) could complement current diagnostic procedures for prostate cancer and, therefore, warrants further investigation.


Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Patients with localised prostate cancer are risk stratified based on the probability of recurrence or progression after primary local treatment, using parameters such as T-stage, Gleason score and prostate-specific antigen (PSA).1 However, there still remains considerable heterogeneity in treatment response even with this risk stratification. To improve predictive value, additional secondary factors such as PSA velocity and tumour volume have been considered, but the value of these as prognostic markers is still debatable.2 Furthermore, several protein or nucleic acid-based transcripts are being explored as prognostic biomarkers to discriminate indolent from aggressive cancer. Depending on the sample type (prostate tissue, blood or urine), different techniques including immunohistochemistry (IHC) and reverse transcriptase-PCR (RT-PCR) are used for detection.3 However, most of these biomarkers have not been sufficiently evaluated to demonstrate their additional prognostic value.

Malignant transformation causes prostate epithelial cells to lose their differentiation in conjunction with increased turnover of cells, marked by elevated proliferation of tumour cells. This results in an increase of choline-containing molecules within the prostate gland,4 ,5 which are obligate precursors for cell membrane biosynthesis.6 The progression of normal cells to malignant phenotype is postulated to be associated with changes in transport, incorporation and use of choline as well as increased choline kinase (CHK) activity.4 ,7 CHK is the first enzyme of the Kennedy pathway responsible for de novo synthesis of phosphatidylcholine, the most abundant membrane lipid.8 CHK exists in three isoforms, alpha1 (CHKA), alpha2 and beta (CHKB), of which only the alpha isoforms have been implicated in oncogenesis.9 Upon uptake of free choline via various transporters, CHK is responsible for the generation of phosphocholine. Deregulated CHKA expression and activity have been extensively linked to cell proliferation and human carcinogenesis.10–12 Furthermore, CHKA overexpression is a frequent event occurring in a variety of human tumours such as breast, lung, colorectal and prostate tumours12 ,13 and has also been described as a new relevant prognostic factor in lung cancer.14 ,15 This biology has led to the development of CHKA inhibitors16 ,17 and the increasing use of choline PET as an imaging biomarker.18–20 Furthermore, magnetic resonance spectroscopy (MRS) has demonstrated the feasibility of distinguishing healthy (normal) prostate tissue, which has high concentration of citrate and low concentration of free choline-containing molecules, from malignant prostate tissue, which has decreased concentration of citrate and increased concentration of free choline-containing molecules.4 ,21

IHC is an investigative tool that provides complementary information to the routine morphological assessment of tissues. Its use in studying cellular markers that define specific phenotypes has provided important diagnostic, prognostic and predictive information about the disease status and tumour biology. IHC against CHKA has been used with encouraging results in human breast and lung cancer tissue samples.22 However, there is a paucity of data describing the patterns of CHKA expression in human prostate cancer tissue samples. In this study, we evaluate the patterns of CHKA expression in normal (no evidence of malignancy) and malignant human prostate tissue samples using commercially available reagents. The prediction of clinical outcome stratified according to the CHK intensity scores was also evaluated.

Materials and methods

Tissue culture and transfections

U2OS cells were cultured in Dulbecco's modified Eagle's medium and HCT116 cells in Roswell Park Memorial Institute medium supplemented with 10% fetal bovine serum, glutamine and antibiotics. Cells were maintained at 37°C in a humidified atmosphere containing 5% CO2. U2OS cells were transfected with GFP-tagged ORF clone of homo sapiens CHKA2 or CHKB using Lipofectamine-2000 according to manufacturer's instructions. After selection with 0.5 mg/mL G418, individual clones were picked and used for western blot analysis against CHKA. For siRNA-mediated CHKA knockdown, HCT116 cells were transfected with 12.5 nM scramble control or CHKA-targeting siRNA by reverse wet transfection with Lipofectamine RNAiMAX according to manufacturer's instructions.

Patient samples

Initially, normal and malignant prostate tissue from paraffin-embedded prostatectomy specimen blocks of five patients were analysed for CHKA protein and mRNA expression by western blot23 and quantitative (q)RT-PCR,24 respectively. One hundred paraffin-embedded malignant prostate tumour cores (including 20 cores from patients whose primary and IHC data have been previously published,20 and 10 cores from patients recruited to the choline response study19) and 25 normal prostate cores were analysed. The 70 malignant cores were obtained from consecutive patients who had neoadjuvant androgen deprivation and radical radiotherapy for their prostate cancer. Similarly, the 25 normal prostate cores were obtained from the patients who had bladder cancer and underwent cystoprostatectomy. The patients recruited in the two above-mentioned studies gave consent for their tissue samples to be used for further research.

Western blot analysis

Cell lysates were obtained by incubation with radioimmunoprecipitation assay buffer containing protease and phosphatase inhibitors. Protein and mRNA of formalin-fixed paraffin embedded (FFPE) prostate tissues were extracted using Qproteome FFPE Tissue Kit and RNeasy FFPE Kit. After protein extraction, 20 μg of total protein was subjected to sodium dodecyl sulfate sample buffer and reducing agent and heated to 70°C for 10 min. Proteins were separated by Biorad 4–15% Mini-Protean TX-gels, transferred to a polyvinylidene difluoride membrane using Transblot Turbo Transfer system and blocked with 1% milk in tris-buffered saline containing 0.1% Tween20 (TBST) for 1 h. The membranes were subjected to CHKA, CHKB, turbo-GFP and β-actin primary antibodies. After overnight exposure to antibody, membranes were washed and incubated with horseradish peroxidase-labelled secondary antibodies (antimouse, antirabbit). The western blot reactions were detected by enhanced chemiluminescence.


The extracted RNA was reverse transcribed to cDNA using QuantiTect reverse transcription kit. Gene expression was quantified on a 7900HT-Fast-Real-Time-PCR system using TaqMan Fast-Advanced-Master-Mix and TaqMan gene expression assays detecting CHKA, CHKB and glyceraldehyde 3-phosphate dehydrogenase. Relative expression was calculated by a comparative (CT) method.25

IHC and tissue staining

All the paraffin-embedded prostate cores were stained for H&E, Ki67-labelling index and CHKA expression. For IHC, all the tumour paraffin-fixed sections were initially deparaffinised in xylene and then serially dehydrated using decreasing grades of ethanol (100–70%). For detection, a labelled-streptavidin-biotin method was used, which involved the use of a biotinylated secondary antibody that links primary antibodies to a streptavidin-peroxidase conjugate.26

H&E staining

After deparaffinisation, the adjacent sections to those used for CHKA IHC were stained for H&E as per manufacturer's instructions.

CHKA immunohistochemistry

The sections were subjected to heat-induced antigen retrieval, quenching of endogenous peroxidase activity and blocking of non-specific binding and stained for CHK as described previously.20 Tumour slides were then independently scored by two experienced accredited pathologists (MG and SH). The intensities were scored as 1+, mild intensity; 2+, moderate intensity; 3+, high intensity, including nuclear staining as compared to the positive control. Photomicrographs were obtained using BX51 Olympus microscope (Olympus Optical, Tokyo, Japan) at 400× magnifications.

Ki67-labelling index

After antigen retrieval, quenching of endogenous peroxidase activity and blocking of non-specific binding as described above, sections were stained for Ki67 as per manufacturer's instructions. Tumour cells and Ki67-positive cells were then manually counted in four randomly selected tumour fields of view to establish the Ki67 labelling index ((Ki67-positive tumour cells×100)/total tumour cells).

The details of the reagents used in the transfection, western blot, qRT-PCR and the detailed methodology of the HC techniques are given in the online supplementary information.

CHK expression and survival

All the patients were followed up with PSA as per the standard departmental protocol and the survival was calculated. The progression-free survival (PFS) was calculated as the time from start of treatment to progression on CT, and overall survival (OS) was calculated as the time from diagnosis to death.

Statistical considerations

The associations between CHK expression, PSA and Gleason score were evaluated using Pearson's correlation test. Analysis of PFS and OS was performed by using Kaplan–Meier estimates and the log-rank test. A two-sided p value <0.05 was considered statistically significant. Analyses were performed using GraphPad (Prism software V. 4).


CHKA protein and gene expression in normal and malignant prostate tissue

To validate the immunoreactivity of anti-CHKA antibody, U2OS cells were transfected with turboGFP-tagged CHKA2 or CHKB and immunoblotted. The antibody could detect recombinant, endogenous human CHKA, but not CHKB, in U2OS cells. Transfection with GFP-tagged CHKA2 or CHKB resulted in marked overexpression of respective proteins with an expected shift in molecular weight of 30 kDa (figure 1A). Importantly, despite profound CHKB overexpression (see turboGFP band) and sequence similarity, the CHKA antibody did not cross-react with CHKB. Conversely, transient target knockdown by siRNA in HCT116 cells resulted in decreased detection of CHKA (figure 1B). Immunoblotting of tissues from two patients showed good resolution of CHKA protein in the malignant prostate (figure 1C). Analysis of CHKA mRNA in the tissues of five patients by qRT-PCR showed minimal CHKA gene expression in normal tissue and elevated CHKA expression in two out of five malignant prostate tissues (figure 1D). The prostate tissues were not micro-dissected, and therefore, the malignant tissues could be confounded by normal tissue. However, this readout provided initial validity of the antibody and justified examination of larger numbers of patients by IHC.

Figure 1

Expression of choline kinase-α (CHKA) protein and genes in the normal (N) and malignant (M) prostate tissue. (A) Western blot analysis of U2OS wild-type cells, U2OS cells transfected with turboGFP-tagged CHKA2 or CHK-β (CHKB). Note the turbo-GFP tag increases the molecular weight by ca. 30 kDa. (B) CHKA expression was abrogated in HCT116 cells using siRNA and protein detected by western blot. (C) Western blot analysis of CHKA in two normal and malignant prostate tissues showed good resolution of bands in the malignant prostate tissue. (B) Quantitative expression of CHKA mRNA showed no significant differential expression in the normal and malignant prostate tissues, as determined by qRT-PCR. Ctr, control; Scr, scramble non-targeting control; siCHKA, targeting siRNA.

CHKA expression and Ki67 index

Normal prostate cores

There was no CHKA expression demonstrable in a majority of normal prostate cores (figure 2). However, there was mild (intensity) CHKA expression in 28% (7/25) of normal prostate cores. This suggests that CHKA expression per se may not be restricted to malignant prostate tissues. Furthermore, there was no Ki67 expression seen in any of the normal cores.

Figure 2

Typical choline kinase-α (CHKA), Ki67 and H&E stained immunohistochemistry cores from a normal prostate gland. The sections indicate absence of CHKA and Ki67 expression in one set of cores (A) and mild CHKA expression and no Ki67 expression in another set of cores (B).

Malignant prostate cores

CHKA was expressed in all malignant prostate cores (n=100) and its intensity varied from 1 to 3 (predominantly cytoplasmic and some nuclear staining) compared with positive control. The different intensities of CHKA staining along with adjacent H&E stained sections are shown in figure 3. Various patterns of staining were observed across the sections (figure 4). Areas of prostatic intraepithelial neoplasia (PIN) consistently showed nuclear staining (figure 4). Ki67 staining (figure 4) revealed that most of the primary prostate tumours had a low proliferation index (median 2%, range 1–17%).

Figure 3

Choline kinase-α (CHKA) staining and matched H&E sections in malignant prostate cores. The panels show different staining intensities: (A) mild staining (score 1), (B) moderate staining (score 2) and (C) strong cytoplasmic and nuclear staining (score 3). There is increasing intensity of staining with increasing Gleason score 3 (A) and Gleason score 5 (C). GL, Gleason score.

Figure 4

Patterns of choline kinase-α staining in malignant cores (A–C) and Ki67 labelling index (D–F). Areas of PIN showed nuclear staining (A), cytoplasmic and nuclear staining in malignant glands but not in benign gland (B) and cytoplasmic and nuclear staining in both malignant and benign glands (C). Ki67 staining in positive control: tonsil (D), core with lowest Ki67 index and (E) core with highest Ki67 index. B, benign; M, malignant; PIN, prostatic intraepithelial neoplasia.

Figure 5

Kaplan–Meier survival curves for predictive value of choline kinase (CHK) scores, (A) Progression-free survival (PFS) and (B) overall survival (OS). When stratified according to the CHK scoring intensity, there was a trend towards lower PFS in patients with high CHK scoring intensity. The median survival was not reached.

In 28 patients, cores from one of the lobes of the prostate gland were benign and those from the other lobe were malignant. The benign cores in 7 of these 28 patients showed that there was increased CHKA expression (mild staining), but no Ki67 expression (supplementary figure 2) which is similar to the pattern seen in the cores from the normal prostate gland.

There was moderate correlation between the intensity of the CHKA staining and PSA and good association between intensity of the CHKA staining and degree of differentiation (Gleason score) in the malignant cores (table 1). There was no association between Ki67-labelling index and Gleason scores or PSA.

Table 1

Correlation of CHKA and Ki67 scores with PSA and GS in malignant prostate cores

Prediction of clinical outcome

Of the 100 patients, survival data were available only for 62 patients. At the time of analysis, only 10 patients had relapsed and 1 patient died due to other causes, after a mean follow-up of 37 months (range 12–59 months). When the survival was stratified according to the CHK intensity scores, the median survival was not reached but there was a trend towards lower PFS in those patients with a CHK score of 3 (figure 5). There was no difference in OS.


To improve the predictive value of the risk stratification in prostate cancer, additional prognostic biomarkers are being explored to discriminate indolent from aggressive cancer27 and also to assess treatment response. Metabolic changes in prostate cancer present as a reduction in citrate and an increase in choline-containing compound compared with normal prostate cells.28 These metabolic alterations in prostate cancer are increasingly applied in diagnostic tools like magnetic resonance spectroscopy and positron emission tomography to improve characterisation of the disease and to monitor response to treatment. The gene expression changes underlying these metabolic aberrations include changes in phospholipase A2 and CHKA.29 These key regulatory enzymes play important roles in prostate cancer progression and could be possible targets for prostate cancer-specific therapy. Therefore, this study was done to develop an IHC method for monitoring CHK expression in human prostate samples.

The CHKA primary antibody used in this study was able to identify cytoplasmic and nuclear CHKA expression. Immunoblotting showed good differential CHKA protein expression in the malignant and normal prostate, and the detection of CHKA mRNA using qRT-PCR indicated low CHKA gene expression in both normal and some malignant prostate tissues and overexpression on two out of five samples. This supports the mild degree of CHKA expression (28%) seen in the normal prostate cores and could be the basis for the differential [11C]choline uptake seen in the normal and tumour prostate.19

IHC showed increased expression of CHKA (mild staining intensity) in up to 28% of the normal prostate cores and the benign cores in a malignant prostate. However, there was no Ki67 expression seen in the normal prostate cores. This suggests that the combination of CHK and Ki67 expression could be more specific for cancer. CHKA staining in areas of PIN and the different intensities of positive staining in the malignant prostate cores were consistently seen. The staining in areas of PIN in the malignant cores may represent the range of CHKA expression in premalignant and malignant tissues. The predominant nuclear staining seen in areas of PIN could be due to the morphological and cytological features of PIN such as nuclear enlargement, prominent nucleoli and hyperchromasia.30 Furthermore, the localisation of choline in areas of PIN has also been shown with MRS.31

In certain tissue sections, nuclear staining was observed, particularly in certain high Gleason grade tumours, and although this phenomenon is not fully understood, a possible reason is that CHKA may translocate to the nucleus for synthesis of endonuclear phosphatidylcholine.9 This hypothesis needs further evaluation.

This study also showed that proliferation in prostate tumours was low, as reflected by the low Ki67 index in most tumours (median 2%; range 1–17%), contrary to the high CHKA expression. Possibly, CHKA expression is not directly linked to proliferation and may be an independent marker of the prostate tumour phenotype. This is contrary to the evidence in other cell and tumour types linking CHKA or choline metabolites and proliferation.13 ,32 ,33 However, the Ki67 indices did not correlate with the Gleason score (r=−0.1, p=0.70). This lack of correlation could be due to the inhomogeneous group of the selected patients.

We have previously established a direct relationship between CHKA expression and [11C]choline uptake in prostate tumours.20 This relationship suggests that [11C]choline PET/CT could potentially be used as a surrogate for CHKA expression. In addition, due its non-invasive nature, [11C]choline PET/CT could improve the diagnostic pathway and aid treatment decisions alongside IHC-based approaches. However, due to the high degree of spatial heterogeneity of prostate cancer,34 this needs further evaluation.

Patients with high-risk prostate cancer are a heterogeneous group, and patients with several risk factors have different outcomes than patients with a single risk factor.35 The trend towards a lower PFS in those patients with CHK score of 3 is an interesting observation. This is seen even in the presence of only modest associations between CHK scoring, PSA and Gleason score (probably due to inclusion of both intermediate-risk and high-risk group patients in the study). This suggests that CHK IHC may potentially aid in the prediction of patients with poor outcomes and could complement the current risk stratification model used to define risk groups in prostate cancer.1

The limitations of the study include the following: the prostate tissues were not microdissected, for the analysis of CHKA mRNA, and therefore, the malignant prostate tissues could be confounded by normal prostate tissue, lack of survival data for all patients and inclusion of both intermediate-risk and high-risk group patients in the study.

In summary, a spectrum of CHK expression was seen in the premalignant and malignant lesions. This relatively simple, cost-effective technique (IHC) could be exploited as a biomarker alongside choline PET to screen potential suitability of patients for novel CHK inhibitors and as a prognostic biomarker.

Take home messages

  • Choline kinase-α (CHKA) antibody has been validated with CHKA-transfected cells and siRNA knockdown.

  • Immunohistochemistry of normal prostate cores showed mild CHKA expression with no Ki67 expression.

  • CHKA was expressed in malignant prostate cores along with low Ki67 labelling index.

  • A trend towards shorter progression-free survival with CHK score of 3 was seen.


We would like to thank the tissue bank staff for their assistance with retrieving the prostatectomy specimens, tissue blocks and providing paraffin-fixed tissue sections for immunohistochemistry.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Handling editor Cheok Soon Lee

  • Contributors Study concepts/design: AC, ST, SM and EA. Data acquisition/quality control: AC, KK, ST, MG and EA. Data analysis/ interpretation: AC, ST, KK, SH and MG. Statistical analysis: AC, ST and EA. Manuscript preparation: AC and ST. Manuscript editing/review: AC, ST, KK, SH, MG, EA and SM.

  • Funding  The authors’ work was funded by the UK Medical Research Council (MRC) grant MC-A652-5PY80, Joint Cancer Research UK and Engineering and Physical Sciences Research Council Cancer Imaging Centre at Imperial College London, in association with the MRC and Department of Health (England) grant C2536/A10337, Experimental Cancer Medicine Centres grant C37/A7283 and National Institute for Health Research (NIHR) Biomedical Research Centre award to Imperial College Healthcare NHS Trust and Imperial College London. The sponsor was not involved in the work done for the manuscript or preparation of the manuscript.

  • Competing interests None declared.

  • Ethics approval Imperial Tissue bank, Imperial College, London.

  • Provenance and peer review Not commissioned; externally peer reviewed.