Metabolic imaging across scales reveals distinct prostate cancer phenotypes

Hyperpolarised magnetic resonance imaging (HP-13C-MRI) has shown promise as a clinical tool for detecting and characterising prostate cancer. Here we use a range of spatially resolved histological techniques to identify the biological mechanisms underpinning differential [1-13C]lactate labelling between benign and malignant prostate, as well as in tumours containing cribriform and non-cribriform Gleason pattern 4 disease. Here we show that elevated hyperpolarised [1-13C]lactate signal in prostate cancer compared to the benign prostate is primarily driven by increased tumour epithelial cell density and vascularity, rather than differences in epithelial lactate concentration between tumour and normal. We also demonstrate that some tumours of the cribriform subtype may lack [1-13C]lactate labelling, which is explained by lower epithelial lactate dehydrogenase expression, higher mitochondrial pyruvate carrier density, and increased lipid abundance compared to lactate-rich non-cribriform lesions. These findings highlight the potential of combining spatial metabolic imaging tools across scales to identify clinically significant metabolic phenotypes in prostate cancer.


Field-specific reporting
Please select the one below that is the best fit for your research.If you are not sure, read the appropriate sections before making your selection.

Life sciences
Behavioural & social sciences Ecological, evolutionary & environmental sciences For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

Life sciences study design
All studies must disclose on these points even when the disclosure is negative.

Data exclusions
The authors declare that the clinical and imaging data supporting the findings of this study are available within the article and its Supplementary Information.The open-source TCGA-PRAD data used in this study are available through the NCI GDC Data Portal, with additional single-cell and spatial RNA-sequencing analyses performed on two publicly available datasets, EGAS00001005787 and GSE176031, respectively.Data used to generate plots in Fig. 2-7 along with the DESI-MSI metabolite data are provided in the Source Data file.The authors defer raw DESI-MSI and clinical MRI data deposition to ensure compliance with legal requirements of the University of Cambridge and Cambridge University Hospitals NHS Foundation Trust and avoid compromising privacy of the study participants.Requests for raw data can be referred to the corresponding author (N.S.); these will be reviewed within ten working days in consultation with the institutional R&D which will determine the terms of a data transfer agreement between the recipient institution, the University of Cambridge, and Cambridge University Hospitals NHS Foundation Trust.
Due to the prostate cancer-specific focus of this study, it only included male patients whose sex was specified in their clinical records.
We have not reported any of these categories in our manuscript.
This study included two prospective cohorts of prostate cancer patients who underwent robot-assisted radical prostatectomy in our centre.The study included a total of 21 patients (median age 67 years, interquartile range 63-69 years) who harboured a total of 30 biopsy-proven prostate tumours.The study population included two matched prospective surgical cohorts who were matched based on tumour histopathological characteristics, as outlined in Table 1 of the main manuscript.
Patients in both cohorts were selected at clinical multi-disciplinary team meetings or clinics and approached by clinical staff involved in their routine care.If patients agreed, they were then approached by the research staff when the nature of the two studies (MISSION-Prostate or DIAMOND, as detailed in the below section) was explained and they were given written information to read.Patients were given at least 24 hours to consider the study before consenting.For the purpose of this pilot physiological study, with the primary objective of identifying biological mechanisms underpinning differential [1-13C]lactate labelling between the benign and malignant prostate, as well as between cribriform and non-cribriform intermediate-risk tumours, insufficient prior knowledge is available to perform a formal sample size calculation.The defined sample size of the MISSION-Prostate study has therefore been determined based on pragmatic considerations of the anticipated recruitment rates over the specified study duration, with the final number of included patients impacted severely by the COVID-19 pandemic that led to the cancellation of all routine imaging studies.The size of the matched DIAMOND cohort was determined by the number and histopathological characteristics of primary tumours (n = 15) available for analysis as part of the MISSION-Prostate study.
In the hyperpolarised MRI cohort, two tumours were excluded from the imaging analysis due to technical failure of the HP 13C-MRI.Tissue

Antibodies
Antibodies used Validation samples from one patient were excluded from the analysis based on the impact of the interval androgen deprivation therapy on physiological mechanisms investigated in this study.This exclusion could not be foreseen because the interval androgen deprivation therapy was prescribed due to cancellation of all elective surgeries due to COVID-19 pandemic and was, therefore, not part of a standard of care.These data exclusions are described in the manuscript.
The hyperpolarised imaging data was acquired following a single injection and could not be repeated.However, images were acquired at multiple timepoints following injection which reduced the effects of noise or artefact in a single image.IHC, DESI-MSI, and RNAscope experiments were not repeated, however, extensive validation of the methods was conducted as described in the main manuscript and elsewhere in this Reporting Summary.
Randomization was not undertaken as there was no therapeutic intervention.Consecutive patients fulfilling the required inclusion and exclusion criteria and consenting to be enrolled were included to avoid investigator bias.Patient selection in the DIAMOND cohort was guided by specific histopathological characteristics of tumours derived from the primary cohort, as described in the main text.
The IHC, RNAscope, and DESI-MSI data were analyzed in a blinded fashion.The clinical clinical data was anonymized prior to analysis although given the small patient size, this was not blinded.All antibodies have been previously validated in our centre using positive and negative tissue controls under the supervision of specialist pathologists.
For the MCT1 antibody validation, we initially tested the antibody in a brain cell line that served as a positive control.We tested the antibody at a single dilution (1:100) with three different pre-treatments (sodium citrate, Tris EDTA, and enzyme), along with a no primary control for each retrieval.We saw strong cell surface signal with both sodium citrate and Tris EDTA, as expected MCT1 which is most commonly localised at cell membrane (predominant cell membrane staining in the prostate samples can be clearly seen in Fig. 2i and Fig. 5d of the main text).The initial tissue tests of the antibody were subsequently carried out in a human xenograft in a rat model (Rat 3(2)_SP13) with a range of antibody titrations.Following the review of positive and negative control sections by a consultant neuropathologist, the antibody dilution to 23.36 µg/mL was deemed optimal; this was subsequently confirmed in primary human tissue.The antibody has been used routinely since it was validated in 2014.
For CD31, we tested the antibody at a single dilution (1:50) with three different pre-treatments (sodium citrate, Tris EDTA, and enzyme) in FFPE human tonsil, along with a no primary control for each retrieval condition.Sodium citrate pre-treatment was taken forward and tested on FFPE human breast cancer TMA offcuts.Results were reviewed by an expert human breast pathologist, with optimal conditions (of a 1:50 antibody dilution and sodium citrate antigen retrieval buffer) agreed.The antibody has been used routinely since it was validated in 2018.Specific vascular staining is clearly demonstrated in Fig. 2i and Fig. 5d on the main text.
For FASN, we tested the antibody at a single dilution (1:100) with three different pre-treatments (sodium citrate, Tris EDTA, and enzyme) in C42b (a FAS over-expressing) and a FAS knockdown FFPE human cell-line, with a no primary control for each retrieval.Sodium citrate and Tris EDTA antigen retrieval yielded positive staining so both antigen retrieval conditions were tested in FFPE human prostate TMA off-cuts next, again including a no primary control for each retrieval condition.Tris EDTA was settled upon as giving the best result with positive staining in prostate cancer tissue and negative staining in normal prostate.The antibody has been used routinely since it was validated in 2011.Cytosolic staining is clearly demonstrated in Fig. 3b, Fig. 5d, and Fig. 6c of the main text.For AR, we we tested the antibody at at two dilutions (1:100 & 1:250) with two different pre-treatments (sodium citrate & Tris EDTA) in in FFPE LNCaps (a (a human prostate cell line), along with a no no primary control for each retrieval condition.Tris EDTA pre-treatment only was taken forward with a 1:50, 1:100 & 1:250 dilution on on FFPE human prostate cancer TMA offcuts, alongside the LNCaps, and with no no primary controls.Optimal conditions of of a 1:50 antibody dilution and Tris EDTA antigen retrieval were settled upon in in agreement with an an expert genitourinary pathologist.The antibody has been used routinely since it was validated in in 2008.Exclusively nuclear staining is is clearly demonstrated in in Fig. 3b, Fig. 5d, and Fig. 6c 6c of of the main text.Importantly, in in Fig. 5b, nuclear AR AR staining is is exclusively seen in in luminal epithelial cells in in the benign gland, with no no staining detected in in basal epithelium.
For HIF-1a, we we tested the antibody at at a single dilution (1:100) with three different pre-treatments (sodium citrate, Tris EDTA, and enzyme) in in FFPE human breast tissue (normal versus tumour), with a no no primary control for each retrieval.Sodium citrate and Tris EDTA antigen retrieval yielded positive staining so so both antigen retrieval conditions were tested in in FFPE human breast TMA off-cuts next using a dilution of of 1:50 & 1:100 for sodium citrate and 1:100 & 1:200 for Tris EDTA.Results were reviewed by by an an expert human breast pathologist and the antibody dilution of of 1:100 (23.36 µg/mL) with sodium citrate retrieval deemed optimal as as it it differentiated positive and negative cases clearly.The antibody has been used routinely since it was validated in in 2018.Both cytosolic and nuclear staining is is clearly demonstrated in in Fig. 6c 6c of of the main text.
For the RNAscope experiment, prior to to the analysis, we we used spare tissue sections to to run the negative control slides (4 (4 Plex DapB to to ensure that DapB is is in in every channel) to to assess background staining, along with the positive control slides (POLR2A for channel 1 and PPIB for channel 2) 2) to to determine good RNA quality.In In the analysis optimisation, we we used the negative controls to to set the thresholds for positive signal in in the test slides.The described routine in-house RNAscope antibody validation process, along with the subsequent analysis, was performed by by an an experienced member of of our dedicated Histopathology Core Facility (see Acknowledgments) with 11years' experience of of using all RNAscope automated kits available for the Leica Bond Rx Rx (Single Plex, Duplex, 3 Plex, 4 Plex, BaseScope, and RNAscope Plus), as as well as as manual HiPlex kits, for more than 50 50 separate projects in in a variety of of tissues and species, including human and murine breast, brain, kidney, lung, and liver.
This study falls under the remit of of a physiological study rather than a formal clinical trial.This has been approved by by the Medicines and Healthcare products Regulatory Agency (MHRA) in in the UK.
Both MISSION-Prostate and DIAMOND study protocols will be be made available on on a local publicly available site at at the close of of the study.
clinical studiesAll manuscripts should comply with the ICMJEguidelines for publication of clinical research and a completedCONSORT checklist must be included with all submissions.
We require information from authors about some types of materials, experimental systems and methods used in many studies.Here, indicate whether each material, system or method listed is relevant to your study.If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.