MicroRNA signatures and Foxp3+ cell count correlate with relapse occurrence in follicular lymphoma

First line drug treatment of follicular lymphoma (FL) patients is followed by a highly variable disease-free time before relapse in about one third of patients. No molecular marker is able to predict efficiently the risk of relapse. We investigated the expression profile of microRNAs (miRNAs) by microarrays and of the tumor microenvironment by immunohistochemistry in 26 FLs and 12 reactive lymph nodes (rLN) as reference. Twenty-nine miRNAs were differentially expressed in FLs compared to rLNs and some of them discriminated grade 1 from 3a FLs. Both FLs and rLNs displayed molecular heterogeneity. FLs grouped into two clusters mostly driven by the tumor T-cell content. Among 21 drug-treated FL patients with an average follow-up of 13.5 years, eight cases relapsed. Twenty-six miRNAs discriminated between relapsed and non-relapsed FLs. Ten miRNAs also correlated with Foxp3+ cells number. Notably, Foxp3+ cells were significantly less in relapsed patients and lower Foxp3+ cell number associated with shorter time-to-relapse. Foxp3+ cells did not co-expressed follicular helper T-cell markers and were therefore classified as regulatory T cells rather than follicular regulatory T-cells. These findings introduce new knowledge about the relationship between miRNA alterations and infiltrating immune cells and show that Foxp3+ cells might be predictive of disease relapse.


Supplementary materialS nucleic acids
Five μm slides of FL and rLN frozen tissues were cut at the cryostat. Tumor cellularity of FL, as assessed by an expert pathologist, was higher than 70%. Total RNA was isolated from frozen tissues and sorted cells using TRIZOL reagent (Invitrogen) according to manufacturer protocol. RNA concentration and integrity was determined respectively by spectrophotometer and agarose gel separation.

Quantitative rt-pCr (qrt-pCr)
For each sample and miRNA, 5 ng of total RNA was converted to cDNA by TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems) using the miRNA specific primer contained in the TaqMan MicroRNA assays (Applied Biosystems). MiRNA expression was evaluated in 1× TaqMan Universal Master Mix (Applied Biosystems). For mRNA expression analysis, 1 μg of total RNA was converted to cDNA using the cDNA Universal Reverse Transcription Kit (Applied Biosystems). MRNA expression was evaluated in 1x Power Sybr Green Master Mix (Applied Biosystems). TaqMan miRNA assays and oligonucleotide primers used in qRT-PCR analysis are listed in Supplementary Table 1. QRT-PCR analysis was performed by 7900HT SDS instrument (Applied Biosystems). Expression differences among samples were determined by the comparative method according to User Bulletin #2 (Applied Biosystems) using the average level of noncoding RNA RNU44 and U47 for miRNAs and GAPDH for mRNAs as reference.

immunohistochemistry (iHC)
Number of Foxp3 + (mAb 221D/D3, Serotec, 1:200) PD1 + (mAb NAT105, UCS Diagnostics, 1:100), CD68 + (mAb KP1, DAKO, 1:50) and CD8 + (mAb C8/144B, DAKO, 1:200) cells were quantified in whole-tissue sections using an automated scanning microscope and image analysis system (S.CORE Web Based Image Analysis, S.CO LifeScience, Germany). For double staining Foxp3-PD1 and Foxp3-CXCL13, Foxp3 was revealed by horsereadish peroxidase and the second antibody revealed by alkaline phosphatase. CXCL13 polyclonal antibody (RD Systems) was diluted 1:50. A hematoxylin and eosin stain was prepared using routine methods. Immunostaining was performed using a Dako autostainer. The number of positive cells in each FL and rLN case was the average of cells counted in two consecutive slides. The average number of positive cells used for subsequent analysis was normalized on the whole number of nuclei present in a slide. Immunoarchitectural distribution of the different markers was determined in relation to the neoplastic follicles. For each slide, "intrafollicular" and "interfollicular" and total positive cells were determined.

microarrays
MicroRNA labeling and hybridization were performed using 5 μg total RNA, as described by Liu G et al., PNAS, 2008. We used a multi-species microarray platform containing 2,284 probes, 1,256 for human and 1,028 for mouse targets, respectively. A total of 353 human mature or pre-miRNAs were detectable by the microarray. Each human target was matched by at least two probes, with an average of 4.3 probes for each target. Hybridization signals were detected with Streptavidin-Alexa647 conjugate and scanned images (Axon 4000B) were quantified using the Genepix 6.0 software (Axon Instruments). To minimize the possible batch effect on miRNAs expression, samples of the same category were randomized through different batches. MiRNAs were named according with the old nomenclature reported in miRBase (www.mirbase.org). MiR-X was the 5p or 3p form of a miRNA according to the old nomenclature and mir-X is the 5p or 3p remaining form of MiR-X.

Data analysis
Expression data from microarrays were normalized and transformed using the vsn package for R. Microarray data sets are available on the ArrayExpress under http:// www.ebi.ac.uk/arrayexpress/experiments/E-MTAB-5844. The spots were subsequently classified based on their target sequence regardless of the original designation. The expression measures for probes matching the same miRNA sequence were summarized using the medpolish algorithm, in order to obtain a unique expression fi gure for each target. Clustering analysis was performed using the hclust function and the inverse Pearson correlation as a distance metric, for both genes and arrays. All the clusters were visualized using the Java TreeView software (http://jtreeview.sourceforge.net). To select differentially expressed genes, we performed either Anova or t-tests. To take into account multiple hypothesis testing, the FDR (false discovery rate) was calculated using the q-value package for R. All the calculations were performed using the R statistical software (http://www.r-project. org). Correlation was calculated using Deming linear regression. Pathway analysis was performed by Gene Set Enrichment Analysis. Experimentally validated, strong evidences, miRNA targets were taken from MiRTarBase at http://mirtarbase.mbc.nctu.edu.tw and were submitted to http://software.broadinstitute.org/gsea/msigdb/index.jsp. More signifi cant Immunological Signatures and Hallmark Gene Sets were considered.