Advances in single-cell RNA sequencing (scRNA-Seq) have allowed for comprehensive analyses of single cell data. However, current analyses of scRNA-Seq data usually start from unsupervised clustering or visualization. These methods ignore prior knowledge of transcriptomes and the probable structures of the data. Moreover, cell identification heavily relies on subjective and possibly inaccurate human inspection afterwards. To address these analytical challenges, we developed SCINA (Semi-supervised Category Identification and Assignment), a semi-supervised model that exploits previously established gene signatures using an expectation–maximization (EM) algorithm. SCINA is applicable to scRNA-Seq and flow cytometry/CyTOF data, as well as other data of similar format. We applied SCINA to a wide range of datasets, and showed its accuracy, stability and efficiency, which exceeded most popular unsupervised approaches. SCINA discovered an intermediate stage of oligodendrocytes from mouse brain scRNA-Seq data. SCINA also detected immune cell population changes in cytometry data in a genetically-engineered mouse model. Furthermore, SCINA performed well with bulk gene expression data. Specifically, we identified a new kidney tumor clade with similarity to FH-deficient tumors (FHD), which we refer to as FHD-like tumors (FHDL). Overall, SCINA provides both methodological advances and biological insights from perspectives different from traditional analytical methods.

Clear cell renal cell carcinoma (ccRCC) is characterized by BAP1 and PBRM1 mutations, which are associated with tumors of different grade and prognosis. However, whether BAP1 and PBRM1 loss causes ccRCC and determines tumor grade is unclear. We conditionally targeted Bap1 and Pbrm1 (with Vhl) in the mouse using several Cre drivers. Sglt2 and Villin proximal convoluted tubule drivers failed to cause tumorigenesis, challenging the conventional notion of ccRCC origins. In contrast, targeting with PAX8, a transcription factor frequently overexpressed in ccRCC, led to ccRCC of different grades. Bap1-deficient tumors were of high grade and showed greater mTORC1 activation than Pbrm1-deficient tumors, which exhibited longer latency. Disrupting one allele of the mTORC1 negative regulator, Tsc1, in Pbrm1-deficient kidneys triggered higher grade ccRCC. This study establishes Bap1 and Pbrm1 as lineage-specific drivers of ccRCC and histologic grade, implicates mTORC1 as a tumor grade rheostat, and suggests that ccRCCs arise from Bowman capsule cells. Significance: Determinants of tumor grade and aggressiveness across cancer types are poorly understood. Using ccRCC as a model, we show that Bap1 and Pbrm1 loss drives tumor grade. Furthermore, we show that the conversion from low grade to high grade can be promoted by activation of mTORC1. Cancer Discov; 7(8); 900–17. ©2017 AACR. See related commentary by Leung and Kim, p. 802. This article is highlighted in the In This Issue feature, p. 783

Most kidney cancers are metabolically dysfunctional^1–4, but how this dysfunction affects cancer progression in humans is unknown. We infused ¹³C-labelled nutrients in over 80 patients with kidney cancer during surgical tumour resection. Labelling from [U-¹³C]glucose varies across subtypes, indicating that the kidney environment alone cannot account for all tumour metabolic reprogramming. Compared with the adjacent kidney, clear cell renal cell carcinomas (ccRCCs) display suppressed labelling of tricarboxylic acid (TCA) cycle intermediates in vivo and in ex vivo organotypic cultures, indicating that suppressed labelling is tissue intrinsic. [1,2-¹³C]acetate and [U-¹³C]glutamine infusions in patients, coupled with measurements of respiration in isolated human kidney and tumour mitochondria, reveal lower electron transport chain activity in ccRCCs that contributes to decreased oxidative and enhanced reductive TCA cycle labelling. However, ccRCC metastases unexpectedly have enhanced TCA cycle labelling compared with that of primary ccRCCs, indicating a divergent metabolic program during metastasis in patients. In mice, stimulating respiration or NADH recycling in kidney cancer cells is sufficient to promote metastasis, whereas inhibiting electron transport chain complex I decreases metastasis. These findings in humans and mice indicate that metabolic properties and liabilities evolve during kidney cancer progression, and that mitochondrial function is limiting for metastasis but not growth at the original site.

Significance: Despite the discovery of the von Hippel–Lindau ( VHL ) gene in 1993, and that inactivating germ-line mutations of VHL cause multiple kidney lesions, including clear-cell renal cell carcinoma (ccRCC), Vhl inactivation in the mouse does not lead to ccRCC and a mouse model has been lacking. We discovered that the BRCA1-associated protein-1 ( BAP1 ) two-hit tumor suppressor gene is mutated in ccRCC, and one BAP1 allele is frequently somatically codeleted with VHL in tumors. In the mouse, Vhl and Bap1 are on different chromosomes. We show that SIX homeobox 2 ( Six2 ) -Cre;Vhl ^F/F ;Bap1 ^{F/ +} mice develop premalignant lesions and malignant ccRCC resembling VHL syndrome. More broadly, differences in tumor predisposition across species may result from differences in the location of two-hit tumor suppressor genes across the genome.

Renal cell carcinoma (RCC) is characterized by a particularly broad metastatic swath, and, enigmatically, when the pancreas is a destination, the disease is associated with improved survival. Intrigued by this observation, we sought to characterize the clinical behavior, therapeutic implications, and underlying biology. While pancreatic metastases (PM) are infrequent, we identified 31 patients across 2 institutional cohorts and show that improved survival is independent of established prognostic variables, that these tumors are exquisitely sensitive to antiangiogenic agents and resistant to immune checkpoint inhibitors (ICIs), and that they are characterized by a distinctive biology. Primary tumors of patients with PM exhibited frequent PBRM1 mutations, 3p loss, and 5q amplification, along with a lower frequency of aggressive features such as BAP1 mutations and loss of 9p, 14q, and 4q. Gene expression analyses revealed constrained evolution with remarkable uniformity, reduced effector T cell gene signatures, and increased angiogenesis. Similar findings were observed histopathologically. Thus, RCC metastatic to the pancreas is characterized by indolent biology, heightened angiogenesis, and an uninflamed stroma, likely underlying its good prognosis, sensitivity to antiangiogenic therapies, and refractoriness to ICI. These data suggest that metastatic organotropism may be an indicator of a particular biology with prognostic and treatment implications for patients.

Background Programmed death-ligand 1 (PD-L1) expression in metastatic renal cell carcinoma (RCC) correlates with a worse prognosis, but whether it also predicts responsiveness to anti-PD-1/PD-L1 therapy remains unclear. Most studies of PD-L1 are limited by evaluation in primary rather than metastatic sites, and in biopsy samples, which may not be representative. These limitations may be overcome with immuno–positron emission tomography (iPET), an emerging tool allowing the detection of cell surface proteins with radiolabeled antibodies. Here, we report iPET studies of PD-L1 in a preclinical tumorgraft model of clear cell RCC (ccRCC) from a patient who had a favorable response to anti-PD-1 therapy. Case presentation A 49-year-old man underwent a cytoreductive nephrectomy in 2017 of a right kidney tumor invading into the adrenal gland that was metastatic to the lungs and a rib. Histological analyses revealed a ccRCC of ISUP grade 4 with extensive sarcomatoid features. IMDC risk group was poor. Within two hours of surgery, a tumor sample was implanted orthotopically into NOD/SCID mice. Consistent with an aggressive tumor, a renal mass was detected 18 days post-implantation. Histologically, the tumorgraft showed sarcomatoid differentiation and high levels of PD-L1, similar to the patient’s tumor. PD-L1 was evaluated in subsequently transplanted mice using iPET and the results were compared to control mice implanted with a PD-L1-negative tumor. We labeled atezolizumab, an anti-PD-L1 antibody with a mutant Fc, with zirconium-89. iPET revealed significantly higher ⁸⁹Zr-atezolizumab uptake in index than control tumorgrafts. The patient was treated with high-dose IL2 initially, and subsequently with pazopanib, with rapidly progressive disease, but had a durable response with nivolumab. Conclusions To our knowledge, this is the first report of non-invasive detection of PD-L1 in renal cancer using molecular imaging. This study supports clinical evaluation of iPET to identify RCC patients with tumors deploying the PD-L1 checkpoint pathway who may be most likely to benefit from PD-1/PD-L1 disrupting drugs.