Loss-of-Function but Not Gain-of-Function Properties of Mutant TP53 Are Critical for the Proliferation, Survival, and Metastasis of a Broad Range of Cancer Cells.

Mutations in the tumor suppressor TP53 cause cancer and impart poor chemotherapeutic responses, reportedly through loss-of-function, dominant-negative effects and gain-of-function (GOF) activities. The relative contributions of these attributes is unknown. We found that removal of 12 different TP53 mutants with reported GOFs by CRISPR/Cas9 did not impact proliferation and response to chemotherapeutics of 15 human cancer cell lines and colon cancer-derived organoids in culture. Moreover, removal of mutant TP53/TRP53 did not impair growth or metastasis of human cancers in immune-deficient mice or growth of murine cancers in immune-competent mice. DepMap mining revealed that removal of 158 different TP53 mutants had no impact on the growth of 391 human cancer cell lines. In contrast, CRISPR-mediated restoration of wild-type TP53 extinguished the growth of human cancer cells in vitro. These findings demonstrate that LOF but not GOF effects of mutant TP53/TRP53 are critical to sustain expansion of many tumor types. SIGNIFICANCE: This study provides evidence that removal of mutant TP53, thereby deleting its reported GOF activities, does not impact the survival, proliferation, metastasis, or chemotherapy responses of cancer cells. Thus, approaches that abrogate expression of mutant TP53 or target its reported GOF activities are unlikely to exert therapeutic impact in cancer. See related commentary by Lane, p. 211 . This article is featured in Selected Articles from This Issue, p. 201.

Unified framework for patient-derived, tumor-organoid-based predictive testing of standard-of-care therapies in metastatic colorectal cancer.

Predictive drug testing of patient-derived tumor organoids (PDTOs) holds promise for personalizing treatment of metastatic colorectal cancer (mCRC), but prospective data are limited to chemotherapy regimens with conflicting results. We describe a unified framework for PDTO-based predictive testing across standard-of-care chemotherapy and biologic and targeted therapy options. In an Australian community cohort, PDTO predictions based on treatment-naive patients (n = 56) and response rates from first-line mCRC clinical trials achieve 83% accuracy for forecasting responses in patients receiving palliative treatments (18 patients, 29 treatments). Similar assay accuracy is achieved in a prospective study of third-line or later mCRC treatment, AGITG FORECAST-1 (n = 30 patients). "Resistant" predictions are associated with inferior progression-free survival; misclassification rates are similar by regimen. Liver metastases are the optimal site for sampling, with testing achievable within 7 weeks for 68.8% cases. Our findings indicate that PDTO drug panel testing can provide predictive information for multifarious standard-of-care therapies for mCRC.

Low-viscosity Matrix Suspension Culture for Human Colorectal Epithelial Organoids and Tumoroids.

Three-dimensional culture of human normal colorectal epithelium and cancer tissue as organoids and tumoroids has transformed the study of diseases of the large intestine. A widely used strategy for generating patient-derived colorectal organoids and tumoroids involves embedding cells in domes of extracellular matrix (ECM). Despite its success, dome culture is not ideal for scalable expansion, experimentation, and high-throughput screening applications. Our group has developed a protocol for growing patient-derived colorectal organoids and tumoroids in low-viscosity matrix (LVM) suspension culture. Instead of embedding colonic crypts or tumor fragments in solid ECM, these are grown suspended in medium containing only a low percentage of ECM. Compared with dome cultures, LVM suspension culture reduces the labor and cost of establishing and passaging organoids and tumoroids, enables rapid expansion, and is readily adaptable for high-throughput screening. Graphic abstract: Generation of organoids and tumoroids from human large intestine using LVM suspension culture (Created with BioRender.com).

Low-viscosity matrix suspension culture enables scalable analysis of patient-derived organoids and tumoroids from the large intestine.

Cell embedment into a solid support matrix is considered essential for the culture of intestinal epithelial organoids and tumoroids, but this technique presents challenges that impede scalable culture expansion, experimental manipulation, high-throughput screening and diagnostic applications. We have developed a low-viscosity matrix (LVM) suspension culture method that enables efficient establishment and propagation of organoids and tumoroids from the human large intestine. Organoids and tumoroids cultured in LVM suspension recapitulate the morphological development observed in solid matrices, with tumoroids reflecting the histological features and genetic heterogeneity of primary colorectal cancers. We demonstrate the utility of LVM suspension culture for organoid and tumoroid bioreactor applications and biobanking, as well as tumoroid high-throughput drug sensitivity testing. These methods provide opportunities for the study and use of patient-derived organoids and tumoroids from the large intestine.

Defining the susceptibility of colorectal cancers to BH3-mimetic compounds.

Novel targets are required to improve the outcomes for patients with colorectal cancers. In this regard, the selective inhibitor of the pro-survival protein BCL2, venetoclax, has proven highly effective in several hematological malignancies. In addition to BCL2, potent and highly selective small molecule inhibitors of its relatives, BCLxL and MCL1, are now available, prompting us to investigate the susceptibility of colorectal cancers to the inhibition of one or more of these pro-survival proteins. While targeting BCLxL, but not BCL2 or MCL1, on its own had some impact, most (15/17) of the immortalized colorectal cancer cell lines studied were efficiently killed by the combined targeting of BCLxL and MCL1. Importantly, these in vitro findings were confirmed in a xenograft model and, interestingly, in all (5/5) patient derived tumor organoids evaluated. Our results lend strong support to the notion that BCLxL and MCL1 are highly promising targets for further evaluation in efforts to improve the treatment of colorectal cancers.

Loss of Bcl-G, a Bcl-2 family member, augments the development of inflammation-associated colorectal cancer.

Gastrointestinal epithelial cells provide a selective barrier that segregates the host immune system from luminal microorganisms, thereby contributing directly to the regulation of homeostasis. We have shown that from early embryonic development Bcl-G, a Bcl-2 protein family member with unknown function, was highly expressed in gastrointestinal epithelial cells. While Bcl-G was dispensable for normal growth and development in mice, the loss of Bcl-G resulted in accelerated progression of colitis-associated cancer. A label-free quantitative proteomics approach revealed that Bcl-G may contribute to the stability of a mucin network, which when disrupted, is linked to colon tumorigenesis. Consistent with this, we observed a significant reduction in Bcl-G expression in human colorectal tumors. Our study identifies an unappreciated role for Bcl-G in colon cancer.

Imbalance of the renin-angiotensin system may contribute to inflammation and fibrosis in IBD: a novel therapeutic target?

OBJECTIVE: We evaluated the influence of the renin-angiotensin system (RAS) on intestinal inflammation and fibrosis. DESIGN: Cultured human colonic myofibroblast proliferation and collagen secretion were assessed following treatment with angiotensin (Ang) II and Ang (1-7), their receptor antagonists candesartan and A779, and the ACE inhibitor captopril. Circulating and intestinal RAS components were evaluated in patients with and without IBD. Disease outcomes in patients with IBD treated with ACE inhibitors and angiotensin receptor blockers (ARBs) were assessed in retrospective studies. RESULTS: Human colonic myofibroblast proliferation was reduced by Ang (1-7) in a dose-dependent manner (p<0.05). Ang II marginally but not significantly increased proliferation, an effect reversed by candesartan (p<0.001). Colonic myofibroblast collagen secretion was reduced by Ang (1-7) (p<0.05) and captopril (p<0.001), and was increased by Ang II (p<0.001). Patients with IBD had higher circulating renin (mean 25.4 vs 18.6 mIU/L, p=0.026) and ACE2:ACE ratio (mean 0.92 vs 0.69, p=0.015) than controls without IBD. RAS gene transcripts and peptides were identified in healthy and diseased bowels. Colonic mucosal Masson's trichrome staining correlated with Ang II (r=0.346, p=0.010) and inversely with ACE2 activity (r=-0.373, p=0.006). Patients with IBD who required surgery (1/37 vs 12/75, p=0.034) and hospitalisation (0/34 vs 8/68, p=0.049) over 2 years were less often treated with ACE inhibitors and ARBs than patients not requiring surgery or hospitalisation. CONCLUSIONS: The RAS mediates fibrosis in human cell cultures, is expressed in the intestine and perturbed in intestinal inflammation, and agents targeting this system are associated with improved disease outcomes.

Colon organoid formation and cryptogenesis are stimulated by growth factors secreted from myofibroblasts.

Although small intestinal epithelial stem cells form crypts when using intestinal culture conditions, colon stem cells usually form colonospheres. Colon mesenchymal cell feeder layers can stimulate colon crypts to form organoids and produce crypts. We have investigated whether conditioned medium from colon mesenchymal cells can also stimulate colonosphere and organoid cryptogenesis. We prepared conditioned medium (CM) from WEHI-YH2 cells (mouse colon myofibroblasts); the CM stimulated both colonosphere formation and organoid cryptogenesis in vitro. The colon organoid-stimulating factors in WEHI-YH2 CM are inactivated by heating and trypsin digestion and proteins can be concentrated by ultrafiltration. Both the colonosphere- and organoid cryptogenesis- stimulatory effects of the CM are independent of canonical Wnt and Notch signaling. In contrast, bone morphogenetic protein 4 (BMP4) abolishes colonosphere formation and organoid cryptogenesis. The Transforming Growth Factor beta (TGFß) Type I receptor kinase inhibitor (A83-01) stimulates colonosphere formation, whereas the Epidermal Growth Factor receptor (EGFR) kinase inhibitor (AG1478) reduces the formation of colonospheres, but in the presence of EGF, a "just-right" concentration of AG1478 increases colon organoid cryptogenesis.

A hypermorphic epithelial ß-catenin mutation facilitates intestinal tumorigenesis in mice in response to compounding WNT-pathway mutations.

Activation of the Wnt/ß-catenin pathway occurs in the vast majority of colorectal cancers. However, the outcome of the disease varies markedly from individual to individual, even within the same tumor stage. This heterogeneity is governed to a great extent by the genetic make-up of individual tumors and the combination of oncogenic mutations. In order to express throughout the intestinal epithelium a degradation-resistant ß-catenin (Ctnnb1), which lacks the first 131 amino acids, we inserted an epitope-tagged ΔN(1-131)-ß-catenin-encoding cDNA as a knock-in transgene into the endogenous gpA33 gene locus in mice. The resulting gpA33(ΔN-Bcat) mice showed an increase in the constitutive Wnt/ß-catenin pathway activation that shifts the cell fate towards the Paneth cell lineage in pre-malignant intestinal epithelium. Furthermore, 19% of all heterozygous and 37% of all homozygous gpA33(ΔN-Bcat) mice spontaneously developed aberrant crypt foci and adenomatous polyps, at frequencies and latencies akin to those observed in sporadic colon cancer in humans. Consistent with this, the Wnt target genes, MMP7 and Tenascin-C, which are most highly expressed in benign human adenomas and early tumor stages, were upregulated in pre-malignant tissue of gpA33(ΔN-Bcat) mice, but those Wnt target genes associated with excessive proliferation (i.e. Cdnn1, myc) were not. We also detected diminished expression of membrane-associated α-catenin and increased intestinal permeability in gpA33(ΔN-Bcat) mice in challenge conditions, providing a potential explanation for the observed mild chronic intestinal inflammation and increased susceptibility to azoxymethane and mutant Apc-dependent tumorigenesis. Collectively, our data indicate that epithelial expression of ΔN(1-131)-ß-catenin in the intestine creates an inflammatory microenvironment and co-operates with other mutations in the Wnt/ß-catenin pathway to facilitate and promote tumorigenesis.

Analysis of Wnt signalling dynamics during colon crypt development in 3D culture.

Many systems biology studies lack context-relevant data and as a consequence the predictive capabilities can be limited in developing targeted cancer therapeutics. Production of colon crypt in vitro is ideal for studying colon systems biology. This report presents the first production of, to our knowledge, physiologically-shaped, functional colon crypts in vitro (i.e. single crypts with cells expressing Mucin 2 and Chromogranin A). Time-lapsed monitoring of crypt formation revealed an increased frequency of single-crypt formation in the absence of noggin. Using quantitative 3D immunofluorescence of ß-catenin and E-cadherin, spatial-temporal dynamics of these proteins in normal colon crypt cells stimulated with Wnt3A or inhibited by cycloheximide has been measured. Colon adenoma cultures established from APC(min/+) mouse have developmental differences and ß-catenin spatial localization compared to normal crypts. Quantitative data describing the effects of signalling pathways and proteins dynamics for both normal and adenomatous colon crypts is now within reach to inform a systems approach to colon crypt biology.

Partial inhibition of gp130-Jak-Stat3 signaling prevents Wnt-ß-catenin-mediated intestinal tumor growth and regeneration.

Most colon cancers arise from somatic mutations in the tumor suppressor gene APC (adenomatous polyposis coli), and these mutations cause constitutive activation of the Wnt-to-ß-catenin pathway in the intestinal epithelium. Because Wnt-ß-catenin signaling is required for homeostasis and regeneration of the adult intestinal epithelium, therapeutic targeting of this pathway is challenging. We found that genetic activation of the cytokine-stimulated pathway mediated by the receptor gp130, the associated Jak (Janus kinase) kinases, and the transcription factor Stat3 (signal transducer and activator of transcription 3) was required for intestinal regeneration in response to irradiation-induced damage in wild-type mice and for tumorigenesis in Apc-mutant mice. Systemic pharmacological or partial genetic inhibition of gp130-Jak-Stat3 signaling suppressed intestinal regeneration, the growth of tumors in Apc-mutant mice, and the growth of colon cancer xenografts. The growth of Apc-mutant tumors depended on gp130-Jak-Stat3 signaling for induction of the polycomb repressor Bmi-1, and the associated repression of genes encoding the cell cycle inhibitors p16 and p21. However, suppression of gp130-Jak-Stat3 signaling did not affect Wnt-ß-catenin signaling or homeostasis in the intestine. Thus, these data not only suggest a molecular mechanism for how the gp130-Jak-Stat3 pathway can promote cancer but also provide a rationale for therapeutic inhibition of Jak in colon cancer.

Analysis of Wnt signaling ß-catenin spatial dynamics in HEK293T cells.

BACKGROUND: Wnt/ß-catenin signaling is involved in different stages of mammalian development and implicated in various cancers (e.g. colorectal cancer). Recent experimental and computational studies have revealed characteristics of the pathway, however a cell-specific spatial perspective is lacking. In this study, a novel 3D confocal quantitation protocol is developed to acquire spatial (two cellular compartments: nucleus and cytosol-membrane) and temporal quantitative data on target protein (e.g. ß-catenin) concentrations in Human Epithelial Kidney cells (HEK293T) during perturbation (with either cycloheximide or Wnt3A). Computational models of the Wnt pathway are constructed and interrogated based on this data. RESULTS: A single compartment Wnt pathway model is compared with a simple ß-catenin two compartment model to investigate Wnt3A signaling in HEK293T cells. When protein synthesis is inhibited, ß-catenin decreases at the same rate in both cellular compartments, suggesting diffusional transport is fast compared to ß-catenin degradation in the cytosol. With Wnt3A stimulation, the total amount of ß-catenin rises throughout the cell, however the increase is initially (~first hour) faster in the nuclear compartment. While both models were able to reproduce the whole cell changes in ß-catenin, only the compartment model reproduced the Wnt3A induced changes in ß-catenin distribution and it was also the best fit for the data obtained when active transport was included alongside passive diffusion transport. CONCLUSIONS: This integrated 3D quantitation imaging protocol and computational modeling approach allowed cell-specific compartment models of the signaling pathways to be constructed and analyzed. The Wnt models constructed in this study are the first for HEK293T and have suggested potential roles of inter-compartment transport to the dynamics of signaling.

Colonic myofibroblast cell line stimulates colonoid formation.

A stable and efficient system for the culture of murine colon epithelial cells or crypts is required to facilitate studies of the dynamics and factors affecting colon stem cell niche and crypt formation. Survival of colonic epithelial cells or crypts in vitro was not established until recently, when it was found that exogenous Wnt3A and R-spondin could promote cell survival and formation of spheroids (colonospheres) or some advanced organoids with well-developed crypts (colonoids). However, after 6-8 days in these culture conditions, only small numbers of colonospheres form organoids with crypt-like structures (colonoids). This study describes the use of a myofibroblast cell line and a coculture system that increases the efficiency of colonoid formation from isolated crypts. The enhanced coculture system has significantly improved colonoid-forming efficiency compared with results from previous systems. Crypt formation can be detected as early as day 2. The coculture system will facilitate the characterization of the colon stem cell niche and the changes that occur as a result of perturbations or mutations in colon stem or epithelial cells, such as those that favor precancerous adenoma or cancer.

Colon cryptogenesis: asymmetric budding.

The process of crypt formation and the roles of Wnt and cell-cell adhesion signaling in cryptogenesis are not well described; but are important to the understanding of both normal and cancer colon crypt biology. A quantitative 3D-microscopy and image analysis technique is used to study the frequency, morphology and molecular topography associated with crypt formation. Measurements along the colon reveal the details of crypt formation and some key underlying biochemical signals regulating normal colon biology. Our measurements revealed an asymmetrical crypt budding process, contrary to the previously reported symmetrical fission of crypts. 3D immunofluorescence analyses reveals heterogeneity in the subcellular distribution of E-cadherin and ß-catenin in distinct crypt populations. This heterogeneity was also found in asymmetrical budding crypts. Singular crypt formation (i.e. no multiple new crypts forming from one parent crypt) were observed in crypts isolated from the normal colon mucosa, suggestive of a singular constraint mechanism to prevent aberrant crypt production. The technique presented improves our understanding of cryptogenesis and suggests that excess colon crypt formation occurs when Wnt signaling is perturbed (e.g. by truncation of adenomatous polyposis coli, APC protein) in most colon cancers.

Wnt signalling pathway parameters for mammalian cells.

Wnt/ß-catenin signalling regulates cell fate, survival, proliferation and differentiation at many stages of mammalian development and pathology. Mutations of two key proteins in the pathway, APC and ß-catenin, have been implicated in a range of cancers, including colorectal cancer. Activation of Wnt signalling has been associated with the stabilization and nuclear accumulation of ß-catenin and consequential up-regulation of ß-catenin/TCF gene transcription. In 2003, Lee et al. constructed a computational model of Wnt signalling supported by experimental data from analysis of time-dependent concentration of Wnt signalling proteins in Xenopus egg extracts. Subsequent studies have used the Xenopus quantitative data to infer Wnt pathway dynamics in other systems. As a basis for understanding Wnt signalling in mammalian cells, a confocal live cell imaging measurement technique is developed to measure the cell and nuclear volumes of MDCK, HEK293T cells and 3 human colorectal cancer cell lines and the concentrations of Wnt signalling proteins ß-catenin, Axin, APC, GSK3ß and E-cadherin. These parameters provide the basis for formulating Wnt signalling models for kidney/intestinal epithelial mammalian cells. There are significant differences in concentrations of key proteins between Xenopus extracts and mammalian whole cell lysates. Higher concentrations of Axin and lower concentrations of APC are present in mammalian cells. Axin concentrations are greater than APC in kidney epithelial cells, whereas in intestinal epithelial cells the APC concentration is higher than Axin. Computational simulations based on Lee's model, with this new data, suggest a need for a recalibration of the model.A quantitative understanding of Wnt signalling in mammalian cells, in particular human colorectal cancers requires a detailed understanding of the concentrations of key protein complexes over time. Simulations of Wnt signalling in mammalian cells can be initiated with the parameters measured in this report.

Signal therapy of human pancreatic cancer and NF1-deficient breast cancer xenograft in mice by a combination of PP1 and GL-2003, anti-PAK1 drugs (Tyr-kinase inhibitors).

The majority of cancers are caused by mutations of a few signal transducers such as the GTPase RAS, the kinase Src and the tumor suppressor p53. Thus, a group of specific chemical compounds called 'signal therapeutics', that block or reverse selectively these abnormally activated signaling pathways would be very useful for the treatment of these signally disordered cancers. More than 90% of human pancreatic cancers are associated with oncogenic mutations of RAS, in particular K-RAS at codon 12. We have previously shown that, PAK1, the Rac/CDC42-dependent Ser/Thr kinase, is essential for RAS/estrogen-induced transformation and neurofibromatosis (NF). Furthermore, we and others have demonstrated that the growth of mouse RAS-induced sarcomas allografts in mice is almost completely suppressed by either FK228 or a combination of two complimentary Tyr-kinase inhibitors, PP1 and AG 879, all of which block the RAS-induced activation of PAK1. Since, so far no effective therapeutic is available for the treatment of pancreatic cancer patients, we have examined the therapeutic potential of either FK228, the combination of these two Tyr-kinase inhibitors or GL-2003, a water-soluble derivative of AG 879, on human pancreatic cancer (Capan-1) xenograft in mice. Among these PAK1-blocking approaches, the PP1/GL-2003 combination is the most effective in the therapy of this cancer xenograft model. Its therapeutic potential is equivalent to those of gemcitabine and kigamicin D which suppress by 70-80% the growth of a similar human pancreatic cancer xenograft model. Also, this PP1/GL-2003 combination therapy has been proven to be very effective to suppress the estrogen-independent growth of an NF1-deficient multidrug/FK228-resistant human breast cancer (MDA-MB-231) xenograft in mice.

Sichuan pepper extracts block the PAK1/cyclin D1 pathway and the growth of NF1-deficient cancer xenograft in mice.

There is increasing evidence that more than 70% of cancers including pancreatic, breast and prostate cancers as well as neurofibromatosis (NF) are highly addicted to abnormal activation of the Ser/Thr kinase PAK1 for their growth. So far FK228 is the most potent among the HDAC (histone deacetylase) inhibitors that block the activation of both PAK1 and another kinase AKT, downstream of PI-3 kinase. However, FK228 is still in clinical trials (phase 2) for a variety of cancers (but not for NF as yet), and not available for most cancer/NF patients. Thus, we have been exploring an alternative which is already in the market, and therefore immediately useful for the treatment of those desperate cancer/NF patients. Here we provide the first evidence that extracts of Chinese/ Japanese peppercorns (Zanthoxyli Fructus) from the plant Zanthoxylum piperitum called "Hua Jiao"/"Sansho", block selectively the key kinase PAK1, leading to the downregulation of cyclin D1. Unlike FK228, these extracts do not inhibit AKT activation at the concentrations that block either cancer growth or PAK1 activation. The Chinese pepper extract selectively inhibits the growth of NF1-deficient malignant peripheral nerve sheath tumor (MPNST) cells, without affecting the growth of normal fibroblasts, and suppresses the growth of NF1-deficient human breast cancer (MDA-MB-231) xenograft in mice. Our data suggest that these peppercorn extracts would be potentially useful for the treatment of PAK1-dependent NF such as MPNST, in addition to a variety of PAK1-dependent cancers including breast cancers.

Signal therapy of breast cancers by the HDAC inhibitor FK228 that blocks the activation of PAK1 and abrogates the tamoxifen-resistance.

PAK1, a Rac/CDC42-dependent Ser/Thr kinase, is required for both neurofibromatosis (NF) and RAS transformation in vivo. FK228, a histone deacetylase (HDAC) inhibitor, activates a very specific set of genes such as the tumor suppressor WAF1, an inhibitor of cyclin-dependent kinases (CDKs), and suppresses the growth of these tumors. In addition, this drug downregulates cyclin D1, which is upregulated by RAS through PAK1, in breast cancers. In this study, we demonstrate that FK228 at 0.1-1 nM significantly reduces the kinase activity of PAK1 in these cells, without affecting the protein level of PAK1. Interestingly, estrogen receptor (ER) and PAK1 mutually activate each other in breast cancers. Here we provide an evidence suggesting that breast cancers require PAK1 for their estrogen-dependent growth. Moreover, the treatment with FK228 strongly inhibits the estrogen-dependent growth of human breast cancers (both tamoxifen-sensitive and resistant cell lines) in vivo, suggesting that FK228 and other anti-PAK1 drugs would be useful for the treatment of breast cancers which become resistant to currently used estrogen antagonists such as tamoxifen.

Signal therapy of NF1-deficient tumor xenograft in mice by the anti-PAK1 drug FK228.

PAK1, a Rac/CDC42-dependent Ser/Thr kinase, is required for the malignant growth of RAS transformants as well as both NF1-deficient and NF2-deficient cancer cells. FK228, a histone deacetylase (HDAC) inhibitor, suppresses the growth of more than 70% of human cancers in vivo including RAS transformants, breast cancers and prostate cancers by activating a set of genes including the tumor suppressors gelsolin and p21(WAF1), that block upstream and downstream of PAK1, respectively. Here we demonstrate that (1) the anti-PAK1 drug FK228 (0.1 nM) completely blocks the growth of both NF1-deficient and NF2-deficient cancer cells in vitro, and that (2) FK228 (2.5 mg/kg, i.p., twice a week) causes the complete regression of an NF1-deficient human malignant peripheral nerve sheath tumor (MPNST) xenograft in nude mice. This is the very first case where a chemical drug in clinical trials for cancers has ever worked so effectively on neurofibromatosis (experimental neurofibromas) in vivo.